Stock Annex 3.2: VIa Cod

Transcription

1 ICES WGCSE REPORT Annex 2: Stock Annexes Stock Annex 3.2: VIa Cod Stock Working Group West of Scotland Cod (Division VIa) Celtic Seas Ecoregion (WGCSE) Date February 2012 Revised by WKROUND/Steven Holmes A. General A.1. Stock definition Cod west of Scotland are believed to comprise of at least two subpopulations of cod that remain geographically separate throughout the year. The latitudinal boundary of these groups is between 57 and N. The southern component is characterised by coastal groups with a tendency towards year round residency, although there is some exchange with the Irish Sea. The northern component appears to inter-mix with cod in IVa at all stages of the life history (ICES 2012, WD 4). A.2. Fishery The minimum landing size of cod in this area is 35 cm. The demersal fisheries in Division VIa are predominantly conducted by otter trawlers fishing for cod, haddock, anglerfish and whiting, with bycatches of saithe, megrim, lemon sole, ling and skate sp. Fishing in the area is conducted mainly by vessels from Scotland, France, Ireland, Norway and Spain with Scottish vessels taking the majority of cod catch. Since 1976, effort by larger Scottish trawlers and seiners has decreased. Records of effort trends since 2000 can be obtained from the (STECF) [ Cod is believed to be no longer targeted in any of the fisheries now operating in ICES Division VIa. Cod are a bycatch in Nephrops and anglerfish fisheries in Division VIa. Nephrops fisheries use a smaller mesh size than the 120 mm mandatory for cod targeted fisheries, but landings of cod are restricted through bycatch regulations and from 2012 all fisheries are restricted to landings of cod through bycatch only (see below). For 2009 Council regulation (EC) No 1342\2008 introduced a cod long-term management plan. The objective of the plan is to ensure the sustainable exploitation of the cod stock on the basis of maximum sustainable yield while maintaining a fishing mortality of 0.4. For stocks above Bpa, but where mortality is above 0.4 the harvest control rule (HCR) requires: 1 ) setting a TAC that achieves a 10% decrease in the fishing mortality in the year of application of the TAC compared to the previous year, or a TAC that achieves a fishing mortality of 0.4, whichever is the higher. 2 ) limiting annual changes in TAC to ± 20%.

2 1372 ICES WGCSE REPORT 2013 For stocks above Blim, the HCR requires: 3 ) setting a TAC that achieves a 15% decrease in the fishing mortality in the year of application of the TAC compared to the previous year, or a TAC that achieves a fishing mortality of 0.4, whichever is the higher. 4 ) limiting annual changes in TAC to ± 20%. For stocks below Blim the Regulation requires: 5 ) setting a TAC that achieves a 25% decrease in the fishing mortality in the year of application of the TAC compared to the previous year. 6 ) limiting annual changes in TAC to ± 20%. In addition the plan states: That if lack of sufficiently accurate and representative information does not allow a TAC affecting fishing mortality to be set with confidence then; If advice is for catches of cod to be reduced to the lowest possible level, the TAC shall be reduced by 25%; In all other cases the TAC shall be reduced by 15% (unless STACF advises this is not appropriate). TACs are to be set net of discards and fish corresponding to other sources of cod mortality caused by fishing. Initial baseline values for effort shall be set for effort groups defined by the Council and then annual effort and cod catch calculated for those effort groups. For effort groups where the percentage cumulative catch is 20% of that for all fleets, maximum allowable effort shall be adjusted by the same amount as the TAC. If STECF advises cod stocks are failing to recover properly the EU Council will set a TAC and maximum allowable effort lower than those derived from the HCR. For 2012 council regulation (EU) No 43/2012 set a zero TAC for cod in VIa and EU and international waters of Vb east of W with the proviso that: Bycatch of cod in the area covered by this TAC may be landed provided that it does not comprise more than 1,5% of the live weight of the total catch retained on board per fishing trip. A.3. Ecosystem aspects Geographic location and timing of spawning Spawning has occurred throughout much of the region in depths <200 m. However, a number of spawning concentrations can be identified from egg surveys in the 1950s, 1992 and from recent surveys of spawning adult distribution. The most commercially important of these, range from the Butt of Lewis to Papa Bank. There are also important spawning areas in the Clyde and off Mull. The relative contribution of these areas is not known. Based on recent evidence there are no longer any significant spawning areas in the Minch. Peak spawning appears to be in March, based on egg surveys (Raitt, 1967). Recent sampling suggests that this is still the case.

3 ICES WGCSE REPORT The main concentrations of juveniles are now found in coastal waters. Fecundity Fecundity data are available from West, 1970 and Yoneda and Wright, Potential fecundity for a given length is higher than in the northern North Sea but lower than off the Scottish east coast (see Yoneda and Wright, 2004). There was no significant difference in the potential fecundity length relationship for cod between 1970 (West, 1970) and (Yoneda and Wright, 2004). B. Data B.1. Commercial catch Raised landings and discards data, ages 1 to 7+. Discard data are available from1978 but sampling was very limited before Discards in years raised according to Millar and Fryer (2005). The following table gives the source of landings data for West of Scotland cod: Kind of data Country Caton (catch in weight) Canum (catchat-age in numbers) Weca (weight-atage in the catch) Matprop (proportion mature-by-age) Length composition in catch UK(NI) UK(E&W) UK(Scotland) Ireland France Norway X X X X X X X X X X X X X B.2. Biological Natural mortality-at-age (M) is assumed weight-dependent after Lorenzen (1996) with mortality assumed to be time invariant, M is calculated by finding the time-series means for stock weights-at-age before applying the Lorenzen parameters, i.e. Ma = W a 3 exp( 0.29) Where Ma is natural mortality-at-age a, W a is the time averaged stock weight-at-age a (in grammes) and the numbers are the Lorenzen parameters for fish in natural ecosystems. Maturities-at-age are given by Age Proportion mature-at-age

4 1374 ICES WGCSE REPORT 2013 Weights-at-age are supplied separately for landings and discards. Catch weights are derived using the sum of products from the landings and discards weights-at-age. Stock weights-at-age are assumed equal to the catch weights-at-age. B.3. Surveys ScoGFS-WIBTS-Q1: Ages 1 to 6 where oldest age is a true age. Fixed station design. ScoGFS-WIBTS-Q4: Ages 1 to 6 where oldest age is a true age. Fixed station design. Modest to poor self-consistency (a weak ability to track cohorts) and very limited influence on exploratory assessment runs means not included in assessment. IGFS-WIBTS-Q4: Ages 0 to 4 where oldest age is a true age. Sufficient non-zero entries only present for ages 1 and 2. Survey only extends to N. Concerns survey not representative of full assessment area means not included in assessment. UKSGFS-WIBTS-Q1: Ages 1 to 6 where oldest age is a true age. Random stratified design. Replaced ScoGFS-WIBTS-Q1. ICES will consider inclusion as a tuning index through an inter-benchmark procedure when 4+ years of data have been gathered. UKSGFS-WIBTS-Q4: Ages 1 to 6 where oldest age is a true age. Random stratified design. Replaced ScoGFS-WIBTS-Q4. ICES will consider inclusion as a tuning index through an inter-benchmark procedure when 4+ years of data have been gathered. B.4. Commercial cpue Not used. B.5. Other relevant data Grey seal consumption of cod data from Hammond and Harris (2006). Supplementary model run only (used to test sensitivity of outcomes to assumptions about natural mortality). C. Assessment: data and method Model used: TSA Software used: NAG library (FORTRAN DLL) and functions in R. Model Options chosen: Weight-dependent M after Lorenzen (1996); natural system values. Mwght.b < Mwght.Mu <- 3.0 Response: landings-at-age, discards-at-age and survey indices-at-age Commercial data : treated as unbiased : age structure only used (with unaccounted mortality estimated) : adjusted to account for misreporting and then treated as unbiased

5 ICES WGCSE REPORT Points given greater variance at WKROUND 2012 landings cvmult-at-age = c(1, 1, 1, 1, 1, 2, 2): extra variability for ages 6 and 7+ landings cvmult = 3 for age 2 in 1987 and 7+ in 1989 discards cvmult = 2 for age 1 in 1988, age 2 in 1988, age 1 in 1992 discards cvmult = 3 for age 2 in 1992 discards cvmult = 5 for age 2 in 1998, age 2 in 2002 Discard model step model: random walk for each age, with a step function allowed : ages 1and 2 modelled : ages 1 to 4 modelled, with a step function for ages 1 and 2 Stock recruit model Ricker Numbers-at-age 1 assumed to be independent and normally distributed with mean η1 S exp( η2 S), where S is the spawning stock biomass at the start of the previous year. To allow recruitment variability to increase with mean recruitment, a constant coefficient of variation is assumed. Large year class: 1986 Mean in Ricker model replaced by 5η1 S exp( η2 S). The factor of 5 was chosen by comparing maximum recruitment to median recruitment from for VIa cod, haddock, and whiting in turn using previous XSA runs. The coefficient of variation is again assumed to be constant. Fishing selection model amat = 4: fishing selection flat (apart from noise) from age 4 gudmundssonh1 = c(2, 1, 1, 1, 1, 1, 1): extra variability for age 1 Survey model (IBTS Q1) full model: separate catchability for each age ages 1 to 6 modelled transitory and persistent changes in catchability allowed Points given greater variance at WKROUND 2012 cvmult = 3 for age 4 in 2001, 2 in 2007, 4 in 2008, 2 in 2010 cvmult = 5 for age 5 in 2001, 3 in 2008 The main diagnostics of the quality of the model fit come from consideration of the objective value (-2*log likelihood), prediction error results and a consideration of how well the model has replicated discard ratios in the input data. As new years of data become available these diagnostics will indicate the need to down-weigh individual data points or that the data, be it landings, discards or survey, for a given age is more or less variable than previously thought. It is therefore important that changes to the variance structures used in the TSA models will be allowed if they improve model diagnostics.

6 1376 ICES WGCSE REPORT 2013 Seal feeding model (supplementary model run only) α M 2 ( y, a) = where: qasyb y M2(y,a) = Seal predation mortality (in year y on age of cod a) qa = Catchability coefficient (varies with age but not year) Sy = Seal numbers in year y By = Total biomass of cod in year y α = Cod biomass (density) dependency term Input data types and characteristics: Type Name Year range Age range Caton Catch in tonnes 1981 onwards?? Landings at age in numbers?? Discards at age in numbers?? Weight at age in the commercial landings?? Weight at age in the commercial discards West Weight at age of the spawning stock at spawning time. Mprop Proportion of natural mortality before spawning Fprop Proportion of fishing mortality before spawning 1 to 7+ Yes Variable from year to year Yes/No (excluded ) 1981 onwards 1 to 7+ Yes (excluded ) 1981 onwards 1 to 7+ Yes (excluded ) 1981 onwards 1 to 7+ Yes 1981 onwards 1 to 7+ Yes Not used Not used Not used Matprop Proportion mature at age 1981 onwards 1 to 7+ No Natmor Natural mortality 1981 onwards 1 to 7+ No For sensitivity analysis only Numbers consumed by seals at age 1985 and to 7+ na Tuning data: Type Name Year range Age range Tuning fleet 1 ScoGFS WIBTS Q to 6 Tuning fleet 2 ScoGFS WIBTS Q4 Not used 1 to 6 Tuning fleet 3 IGFS WIBTS Q4 Not used 1 to 2

7 ICES WGCSE REPORT D. Short-term projection Model used: Age structured Software used: MFDP prediction with management option table and yield-per-recruit routines. MLA suite (WGFRANSW) used for sensitivity analysis and probability profiles. The following configuration was agreed at WGNSDS 2008 Initial stock size: Taken from TSA for age 1 and older. Weight-at-age in the catch: Average weight of the three last years. Weight-at-age in the stock: Average stock weights for last three years. Assumed equal to the catch weight-at-age, (adopted because mean weights-at-age have been relatively stable over the recent past). CVs are calculated from the standard errors on weights-at-age. Maturity: The same ogive as in the assessment is used for all years. F and M before spawning: Set to 0 for all ages in all years. Exploitation pattern: Average of the three last years. Not partitioned to give landings, misreporting and discard F. If further work can solve this problem, this partition should be applied. Intermediate year assumptions: Still open. Stock recruitment model used: None, recruitment in the intermediate year (terminal year year class at age 1) is taken from the TSA assessment, (the value is based largely on the ScoGFSQ1 survey datum from the terminal year). For the TAC year and following year the short-term (ten years to year before terminal year) geometric mean recruitment-at-age 1 is used. Procedures used for splitting projected catches: Still open. E. Medium-term projections Not considered at the WKROUND benchmark. F. Long-term projections Not considered at the WKROUND benchmark.

8 1378 ICES WGCSE REPORT 2013 G. Biological reference points Type Value Technical basis MSY MSY Bpa Btrigger t Approach FMSY 0.19 Provisional proxy by analogy with North Sea cod Fmax. Fishing mortalities in the range are consistent with Fmsy Precautionary Approach (unchanged since: 2010) Blim t Bpa t Blim = Bloss, the lowest observed spawning stock estimated in previous assessments. Considered to be the minimum SSB required to ensure a high probability of maintaining SSB above Blim, taking into account the uncertainty of assessments. This also corresponds with the lowest range of SSB during the earlier, more productive historical period. Flim 0.8 Fishing mortalities above this have historically led to stock decline. Fpa 0.6 This F is considered to have a high probability of avoiding Flim. Since these reference points were established the assessment has adopted weightdependent natural mortalities (M) at age. This has increased M values for younger ages and increased perceptions of SSB and recruitment in years where they were previously estimated using the old values for M. The differences were, however, judged too small to merit a revision of biomass reference points (ICES 2012). The limit and MSY mortality reference points were also confirmed as still valid in 2012 (ICES 2012). The FMSY estimate was derived by WGCSE 2010 using the srmsymc package. Figures showing stochastic fits to three stock recruit relationships, estimates of FMSY and Fcrash and estimates of yield per recruit, together with descriptive text, are given in Appendix 1. H. Other issues H.1. Change of Scottish Research Survey For 2011 the rig and sampling design of the ScoGFS-WIBTS-Q1 survey was changed. A new groundgear capable of tackling challenging terrain was introduced broadly modelled around the rig used by Ireland for the IRGFS-WIBTS-Q4. The move to a more robust groundgear also allowed a move to a random stratified survey (which is again consistent with the IRGFS-WIBTS-Q4) as the previous repeat station survey format consisting of the same series of survey trawl positions being sampled at approximately the same temporal period every year was considered a bias prone method for surveying the area. It is hoped the greater compatibility between Scottish and Irish surveys will facilitate both being used to assess gadoids west of Scotland. New survey strata were designed using cluster analysis on aggregated data from the previous ScoGFS-WIBTS-Q1 data ( ) as well as the data collected from a dedi-

9 ICES WGCSE REPORT cated gadoid survey which took place during quarter 1 of Species considered were cod, haddock, whiting, saithe and hake. Cluster analysis yielded 4 specific clusters. Two additional strata were added; the Clyde area and the windsock which is an area that has been designated as a recovery zone since 2002 and has therefore experienced no mobile gear exploitation during this time. The new strata are shown in Figure H.1. Each individual polygon was treated as a separate stratum and the number of survey stations for each was allocated according to polygon size and the variability of indices within each stratum. Strata were weighted by surface area to build the final indices. H.2. Historical overview of previous assessment methods 2004 to 2011 Model used: TSA Software used: Compaq visual FORTRAN using NAG library. Model Options chosen: Natural mortality (M) 0.2 at all ages. Commercial data : treated as unbiased 1995-AY-1: omitted landings cvmult-at-age = c(1, 1, 1, 1, 1, 2, 2): extra variability for ages 6 and 7+ Discard model : ages 1and 2 modelled 1995-AY-1: omitted Stock recruit model ricker large year class: 1986 Fishing selection model amat = 4: fishing selection flat (apart from noise) from age 4 gudmundssonh1 = c(4, 1, 1, 1, 1, 1, 1): extra variability for age 1 Survey model (IBTS Q1) amat = 4: catchability flat (apart from noise) from age 4 survey catchabilities up to amat assumed to follow a log-linear model survey cvmult-at-age = c(2, 1, 1, 1, 2, 2): extra variability for ages 1, 5 and 6 ages 1 to 6 modelled only transitory changes in catchability allowed; modelled using the additive scale.

10 1380 ICES WGCSE REPORT 2013 Summary of data ranges used in recent assessments (no accepted assessment in 2011): Data 2007 assessment 2008 assessment 2009 assessment 2010 assessment Catch data Years: 1978 (AY-1) Ages: 1 7+ Years: 1978 (AY-1) Ages: 1 7+ Years: 1978 (AY-1) Ages: 1 7+ Years: 1978 (AY-1) Ages: 1 7+ Survey: A_Q1 Years: 1985 AY Ages: 1 6 Years: 1985 AY Ages 1 6 Years: 1985 AY Ages 1 6 Years: 1985 AY Ages 1 6 Survey: B_Q4 Not used Not used Not used Not used Survey: C Not used Not used Not used Not used AY Assessment year I. References Hammond, P. S., and Harris, R. N Grey seal diet composition and prey consumption off western Scotland and Shetland. Final report to Scottish Executive Environment and Rural Affairs Department and Scottish Natural Heritage. ICES Report of the Benchmark Workshop on Western Waters Roundfish (WKROUND). Aberdeen, UK February ICES CM 2012/ACOM: pp. Lorenzen, K The relationship between body weight and natural mortality in juvenile and adult fish: a comparison of natural ecosystems and aquaculture. Journal of Fish Biology, 49, Millar, C., P., and Fryer, R., J Revised estimates of Annual discards-at-age for cod, haddock, whiting and saithe in ICES Subarea IV and Division VIa. Fisheries Research Services internal report No 15/05, July 2005, 23pp. Raitt, D.F. S Cod spawning in Scottish waters. Preliminary investigations. ICES C. M. 1967/F:29. West, W. Q-B The spawning biology and fecundity of cod in Scottish waters. PhD. thesis, Aberdeen University, Aberdeen. Yoneda, M. and Wright, P. J Temporal and Spatial variation in reproductive investment of Atlantic cod Gadus morhua in the northern North Sea and Scottish west coast. Marine Ecology Progress Series, 276:

11 ICES WGCSE REPORT Figure H.1. Sampling strata of UKSGFS-WIBTS-Q1 survey. Figure also shows cpue numbers for fish aged at 1+ by haul for cod in 2011 (numbers standardised to 60 minutes towing).

12 1382 ICES WGCSE REPORT 2013 Appendix 1: Investigations of F MSY using the srmsymc package The same input data files as used for the short-term forecast were used. An alternative run using ten year means for stock weights-at-age and mortality-at-age showed there to be little sensitivity to the averaging period used. Figure A.1 shows the three stock recruit relationships fitted by the package; Ricker, Beverton Holt and smooth hockey stick. Models were fitted using 1000 MCMC re-samples. For all three stock recruit relationships all re-samples allowed FMSY and Fcrash values to be determined. As such, there was no basis to reject any of the recruitment models as unsuitable for this stock. For each of the stock recruit relationships (SRR) Figures A.2 to A.4 show box plots of FMSY and Fcrash together with the values of FPA and Flim. For the Ricker and Beverton Holt SRR the estimated value of Fcrash is very close to Flim. For the smooth hockey-stick SRR Fcrash is estimated between Flim and FPA. For all three SRR the current level of Z-02 is higher than the median Fcrash value. Also the value of FMSY is well defined and considerably lower than FPA for all three SRR. The level of removals possible at the estimated FMSY is poorly defined however. Circles showing the data points show values of Z-0.2 repeatedly in excess of the upper percentile for Fcrash. As expected removals and SSB have declined such that values for both are now inside confidence limits for these metrics at the estimated Z-0.2 mortality rates. Figure A.5 shows estimation of yield per recruit. FMAX is well defined for this stock. Comparison of FMAX to FMSY estimated using the three SRRs (Figures A.2 4) shows FMSY estimated as lower than FMAX for the Beverton Holt model, equal for the smooth hockey stick and higher than FMAX in the Ricker model reflecting the downward slope of the stock recruit relationship at higher SSBs. In conclusion mortalities from removals in the range 0.17 to 0.33 were considered consistent with FMSY.

13 ICES WGCSE REPORT Figure A.1. Cod in Division VIa. Stock recruit relationships fitted by srmsymc package. Models were fitted using 1000 MCMC re-samples. Left hand panels illustrate confidence intervals. Right hand panels present curves plotted from the first 100 re-samples for illustration. The blue line indicates a deterministic estimate, separate from the MCMC chain. The legends for each recruitment model show it was possible to converge on a value of FMSY and Fcrash for all 1000 iterations in each case.

14 1384 ICES WGCSE REPORT 2013 Figure A.2. Cod in Division VIa. srmsymc package. Estimation of F reference points and equilibrium yield and SSB against mortality using Ricker recruitment model. For yield and SSB plots left hand panels illustrate confidence intervals. Right hand panels present curves plotted from the first 100 re-samples for illustration. The blue line indicates a deterministic estimate, separate from the MCMC chain. Circles show datapoints with the most recent year labelled. For VIa cod the model has been run using total removals over and above natural mortality, i.e. the x-axis represents Z-0.2.

15 ICES WGCSE REPORT Figure A.3. Cod in Division VIa. srmsymc package. Estimation of F reference points and equilibrium yield and SSB against mortality using Beverton Holt recruitment model. For yield and SSB plots left hand panels illustrate confidence intervals. Right hand panels present curves plotted from the first 100 resamples for illustration. The blue line indicates a deterministic estimate, separate from the MCMC chain. Circles show datapoints with the most recent year labelled. For VIa cod the model has been run using total removals over and above natural mortality, i.e. the x-axis represents Z-0.2.

16 1386 ICES WGCSE REPORT 2013 Figure A.4. Cod in Division VIa. srmsymc package. Estimation of F reference points and equilibrium yield and SSB against mortality using smooth hockey-stick recruitment model. For yield and SSB plots left hand panels illustrate confidence intervals. Right hand panels present curves plotted from the first 100 re-samples for illustration. The blue line indicates a deterministic estimate, separate from the MCMC chain. Circles show datapoints with the most recent year labelled. For VIa cod the model has been run using total removals over and above natural mortality, i.e. the x-axis represents Z-0.2.

17 ICES WGCSE REPORT Figure A.5. Cod in Division VIa. srmsymc package. F reference points and yield per recruit and SSB per recruit against mortality. For VIa cod the model has been run using total removals over and above natural mortality, i.e. the x-axis represents Z-0.2.

18 1388 ICES WGCSE REPORT 2013 Table A.1. Cod in Division VIa. Output from srmsymc ADMB package. STOCK NAME Cod-6a Sen filename sum_and_sen_files/codvia10runspalyhf075hf0563.sen pf, pm 0 0 Number of iterations 1000 Simulate variation in Biological parameters TRUE SR relationship constrained TRUE Ricker 1000/1000 Iterations resulted in feasible parameter estimates Fcrash Fmsy Bmsy MSY ADMB Alpha ADMB Beta Unscaled Alpha Unscaled Beta AIC Deterministic E Mean E-05 5%ile E-06 25%ile E-06 50%ile E-05 75%ile E-05 95%ile E-05 CV

19 ICES WGCSE REPORT Table A.5 (cont). Cod in Division VIa. Output from srmsymc ADMB package. BEVERTON HOLT 1000/1000 Iterations resulted in feasible parameter estimates Fcrash Fmsy Bmsy MSY ADMB Alpha ADMB Beta Unscaled Alpha Unscaled Beta AIC Deterministic Mean %ile %ile %ile %ile %ile CV Smooth hockeystick 1000/1000 Iterations resulted in feasible parameter estimates Fcrash Fmsy Bmsy MSY ADMB Alpha ADMB Beta Unscaled Alpha Unscaled Beta AIC Deterministic Mean %ile %ile %ile %ile %ile CV

20 1390 ICES WGCSE REPORT 2013 Table A.5 (cont). Cod in Division VIa. Output from srmsymc ADMB package. PER RECRUIT F35 F40 F01 Fmax Bmsypr MSYpr Fpa Flim Deterministic Mean %ile %ile %ile %ile %ile CV

21 ICES WGCSE REPORT Stock Annex 3.3: Haddock VIa Stock Working Group West of Scotland Haddock (Division VIa) Assessment of Northern Shelf Demersal Stock Last updated May 2009 A. General A.1. Stock definition The haddock is widely distributed around the west coast of Scotland and can be caught in most areas within the 200 m depth contour. The stocks occurring off the northwest coast of Scotland are usually identified according to the regions which support a fishery, but genetic and biological marker studies suggest the possibility of different populations of haddock. A continuous population of haddock is thought to extend from the west coast around to the north of Scotland. Results from tagging experiments and larval transport studies suggest that there may be links between west coast haddock and those in the North Sea. A.2. The fishery The minimum landing size of haddock in the human consumption fishery in this area is 30 cm. The demersal fisheries in Division VIa are predominantly conducted by demersal trawlers fishing for cod, haddock, anglerfish and whiting, with bycatches of saithe, megrim, lemon sole, ling and several species of skate. Since 1976, effort by Scottish heavy trawlers and seiners has decreased. Light trawler effort has declined rapidly since 1997 after a long-term increasing trend onwards Emergency measures were introduced in 2001 to allow the maximum number of cod to spawn (see emergency measures below). Council Regulation (EC) No. 423\2004 introduced a cod recovery plan affecting Division VIa. This has been revised and updated (Council Regulation (EC) No. 1342/2008). The measures only take effect east of a line defined in Council Regulation No 51\2006. The days-at-sea limitations associated with the cod recovery plan and this seasonal closure has led some of the Irish Demersal fleet to switch effort away from VIa. Under Council Regulation (EC) No. 51/2006 the use of gillnets has been banned outside 200 m depth. WGFTFB 2006 report that this has greatly reduced effort at depths greater than 200 m in VIa. The measure was aimed to protect monkfish and deepwater shark and it is unclear what effect it will have on haddock. Technical measures The minimum mesh size for vessels fishing for haddock in the mixed demersal fishery in EC Zones 1 and 2 (West of Scotland and North Sea excluding Skagerrak) changed from 100 mm to 120 mm from the start of This came under EU regulations regarding the cod recovery plan (Commission Regulation EC 2056/2001), with a one-year derogation of 110 mm for vessels targeting species other than cod. This derogation was not extended beyond the end of 2002.

22 1392 ICES WGCSE REPORT 2013 Since mid-2000, UK vessels in this fishery have been required to include a 90 mm square mesh panel (SSI 227/2000), predominantly to reduce discarding of the large 1999 year class of haddock. Further unilateral legislation in 2001 (SSI 250/2001) banned the use of lifting bags in the Scottish fleet. Under Council Regulation No. 51/2006 the use of gillnets has been banned outside 200 m depth. Emergency measures and effort limitation Emergency measures were enacted in 2001, consisting of area closures from 6 March 30 April, in an attempt to maximize cod egg production. These measures were retained into 2003 and In 2005 the following area closures were in effect: 1 ) The Greencastle codling fishery from mid-november to mid-february. This closure has been operating since ) A closure in the Clyde for spawning cod from 14th February to 30th April. This closure has been operating since 2001 and was last revised by The Sea Fish (prohibited methods of fishing) (Firth of Clyde) Order ) A closure introduced in 2004 by Council Regulation No. EC 2287\2003, known as the windsock. Effort reductions for much of the international fleet to 16 days-at-sea per month have been imposed since February 2003 (EU 2003\0090). The maximum number-of-days in any calendar month for which a fishing vessel may be absent from port to the West of Scotland varies for particular gears and the allocations since 2003 are given below: Gear Demersal trawls, seines or similar towed gears of mesh size 100 mm except beam trawls Demersal trawls, seines or similar towed gears of mesh size between 70 mm and 99 mm except beam trawls 1 ; Demersal trawls, seines or similar towed gears of mesh size between 16 mm and 31 mm except beam trawls. Maximum Days Allowed 2003: 2004: 2005: 2006: / / /12 1: With mesh size between 80 mm and 99 mm in The documents listing these days-at-sea limitations are, 2004: (EC) No 2287/ : (EC) No 27/2005-Annex IVa 2006: (EC) No 51/2006-Annex IIa A Commission Decision (C (2003) 762) in March 2003 allocated additional days absent from port to particular vessels and Member States. UK vessels were granted 4 additional days-per-month (based on evidence of decommissioning programmes). An additional two days were granted to demersal trawls, seines or similar towed gears (mesh 100 mm, except beam trawls) to compensate for steaming time between home ports and fishing grounds and for the adjustment to the newly installed effort management scheme.

23 ICES WGCSE REPORT For 2006 one extra day was allocated to trawls >=100 mm if the mesh was >120 mm and the net contained a square mesh panel of 140 mm mesh size. Altogether 148 days in the year was allowed for vessels with mesh between 100 and 120 mm if the catch contained <5% cod in This allowance rises to 160 days in the year if the same 140 mm square mesh panel is used together with a mesh size >120 mm. The new effort regulations provided an incentive for some vessels previously using >100 mesh in otter trawls to switch to smaller mesh gears to take advantage of the larger numbers of days-at-sea available. This would also require these vessels to be targeting Nephrops or anglerfish, megrim and whiting with various catch and bycatch composition limits after EC Regulation No 850/98. Decommissioning schemes. Vessel decommissioning has been underway since Information on the number of vessels operating in the cod recovery zone to have been decommissioned in Division VIa between 2001 and 2004 was as follows: Decomm. To Total VIa Percentage Number of vessels >10 m % A.3. Ecosystem aspects Geographic location and timing of spawning Spawning of haddock usually occurs in February and March and in almost any area where the fish are distributed. There is major spawning between the Butt of Lewis and Shetland. Some larvae from the west coast spawning grounds can be transported to the North Sea, which they enter through the Fair Isle/Shetland Gap or to the northeast of Shetland. Young haddock then spend the first few months of life in the upper water layers before adopting the demersal way of life. The survival rate of young haddock is very variable from year to year. Fecundity The majority of haddock mature-at-age two with usually all mature by age three. However, mature age two haddock spawn fewer eggs for a given size than an age three haddock. A three-year-old female of good size is able to produce around eggs in a season and releases her eggs in a number of batches over many weeks. Diet The diet of haddock varies seasonally and according to location and body size. In winter, haddock of all sizes feed mainly on benthic invertebrates, for example, polychaetes, small crustaceans and echinoderms. In spring and summer, fish prey, especially sandeels, are important particularly for larger haddock. Norway pout is also important prey for haddock. During herring spawning seasons, haddock will feed heavily on herring eggs. B. Data B.1. Commercial catch B1.1. Landings The following table gives the source of landings data for West of Scotland haddock:

24 1394 ICES WGCSE REPORT 2013 Kind of data Country Caton (catchin-weight) Canum (catch-at-age in numbers) Weca (weight-atage in the catch) Matprop (proportion mature-byage) Length compositionin-catch UK(NI) UK(E&W) UK(Scotland) Ireland France Norway X X X X X X X X X X X X X Quarterly landings and length/age composition data are supplied from databases maintained by national Government Departments and research agencies. These figures may be adjusted by national scientists to correct for known or estimated misreporting by area or species. Data are supplied in the requested format to a stock coordinator, who compiles the international landings and catch-at-age data and maintains a time-series of such data with any amendments. To avoid double counting of landings data, each UK region supplies data for UK landings into its regional ports, and landings by its fleet into non-uk ports. Quarterly landings are provided by the UK (Scotland), UK (E/W), UK (NI), France and Ireland. The quarterly estimates of landings-at-age by UK (Scotland) and Ireland are raised to include landings by France, UK (NI) and Norway (distributed proportionately over quarters), then summed over quarters to produce the annual landingsat-age. B1.2. Discards EU countries are now required under the EU Data Collection regulation to collect data on discards of haddock and other species. Up to 2003, estimates of discards were available only from UK (Scotland) and Ireland. Observer data are collected using standard at-sea sampling schemes. Results are reported to ICES. The quantity, length and age of haddock discarded by Scottish Nephrops trawlers are collected during observer trips on board commercial vessels. Haddock discarded by boats using other gears (heavy trawl, seine, light trawl and pair trawl) are also collected by Scotland. Haddock discarded by otter board trawl and otter board/twin rig gears are collected by Ireland. Discards from Scottish and Irish boats using several different gear types are estimated by observers. B.2. Biological Natural mortality is assumed to be constant (M=0.2, applied annually) for the whole range of ages and years. There are no direct estimates of M. Proportion mature-at-age is currently assumed constant over the full time-series as follows: Age Proportion mature

25 ICES WGCSE REPORT These maturity values were derived from a French survey carried out in Division VIa in They were first discussed in the 1984 meeting of the North Sea Roundfish Working Group (ICES-NSRWG 1984), and were first used at the 1985 meeting (ICES- NSRWG 1985). Proportions of F and M before spawning were both set to 0.0, in order to generate abundance (and hence SSB) estimates dated to January 1st. B.3. Surveys Four research vessel survey series for haddock in VIa were available to the working group in In all surveys listed the highest age represents a true age not a plus group. Scottish first-quarter west coast groundfish survey (ScoGFSQ1): ages 1 7, years The survey gear is a GOV trawl, and the design is a minimum of one station per rectangle, but with more depending on logistics. Ages are reported from 0 to the maximum obtained. Sex/Maturity-Sex and Maturity (ICES 4-stage scale) are reported. The Scottish groundfish survey has been conducted with a new vessel and gear since The catch rates for the series as presented are corrected for the change on the basis of comparative trawl haul data (Zuur et al., 2001). Irish fourth-quarter west coast groundfish survey (IreGFS): ages 0 3, years The Irish quarter four survey was a comparatively short series. It was discontinued in 2003 and has been replaced by the IRGFS (see below). Scottish fourth quarter west coast groundfish survey (ScoGFSQ4): ages 0 8, years As is the case for the European IBTS surveys (such as ScoGFS Q1 above) the survey gear is a GOV trawl, and the design is a minimum of one station per rectangle, but with more depending on logistics. Ages are reported from 0 to the maximum obtained. Sex/Maturity-Sex and Maturity (ICES 4-stage scale) are reported. The Scottish groundfish survey has been conducted with a new vessel and gear since The catch rates for the series as presented are corrected for the change on the basis of comparative trawl haul data (Zuur et al., 2001). Irish fourth-quarter west coast groundfish survey (IRGFS); ages 0 3, years This survey used the RV Celtic Explorer and is part of the IBTS coordinated western waters surveys. The vessel uses a GOV trawl, and the design is a depth stratified survey with randomized stations. Effort is recorded as minutes towed. There were 41 stations sampled in 2003, 44 in 2004 and 34 in 2005, corresponding to 1229, 1321 and 1010 minutes towed. Plots of the spatial distribution of the ScoGFS Q1 survey mean catch rates per ICES statistical rectangle by age class are given in Figure 1. The numbers caught in the most recent Scottish Groundfish Surveys are indicated in Figure 2. B.4. Commercial cpue Three commercial Scottish cpue series have been made available in recent years. Irish otter-trawl cpue data (IreOTR) were presented for the first time at the 2001 WG meeting. Updated series have been presented to subsequent meetings. Given the current

26 1396 ICES WGCSE REPORT 2013 concerns about misreporting of catch and effort, this series has not been considered further as a tuning fleet. The commercial cpue data available consists of the following: Scottish seiners (ScoSEI): ages 1 6, years Scottish light trawlers (ScoLTR): ages 1 6, years Irish otter trawlers (IreOTR): ages 1 7, years Reported effort has declined in recent years to very low levels in both Scottish fleets for which effort data are available to the WG (pairtrawlers and light trawlers; see Table 1). The historical mean levels of lpue (landings-per-unit-effort) for these fleets were more constant, although variable. However, problems with effort recording mean that these estimates are unlikely to be valid: further details are available in the report of the 2000 meeting of the ICES WG on the Assessment of Demersal Stocks in the North Sea and Skagerrak (ICES-WGNSSK 2000). For this reason, commercial Scottish lpue data has not been used in the current assessment. Data are also available (although not updated to 2007) from the Irish trawler fleet (IreOTB; Table 4.1.8), but are not used in the assessment as a consequence of concerns about targeting leading to hyperstability. B.5. Other relevant data None. C. Historical stock development In 2007 ICES changed its advisory structure: the previous committees (ACE, ACFM and ACME) were merged into a single committee now known as ACOM. Amongst many of the modifications to accompanying working practices, it was intended that all stock assessments conducted by the Expert Groups from 2008 should be update analyses based on the work conducted by the last benchmark meeting. For west of Scotland haddock, a benchmark assessment per se has not taken place for some time. However, at the 2004 WGNSDS, a full and detailed examination of the assessment was carried out following concerns of ACFM about the assumptions and parameter settings implemented in the TSA methods used to assess this stock (ICES, 2004). The investigation used Time Series Analysis (TSA) Extended Survivors Analysis (XSA) and Survey Based Assessment (SURBA) models. Although the results from this investigation were in some ways contradictory, and the WG remained uncertain about the most appropriate model for the stock, subsequent Review Groups concluded that a TSA assessment, using the Scottish Quarter 1 Groundfish Survey and excluding the catch and discard data from 1995 onwards, should be presented as the final assessment in In 2006 this assessment was modified slightly to incorporate an additional survey, the Scottish Quarter 4 Groundfish Survey (western division bottomtrawl survey). In 2007, concerns were raised about the potential impact on management advice of using a plus-group at-age 8 when the dominant large 1999 year class has reached that age in 2007, and also about the removal in the previous assessment of older ages in the Scottish Q4 Groundfish Survey (ScoGFS Q4). Several exploratory analyses were carried out, from which it was concluded that the same procedure should be used in 2007 as was used 2006, but with two additional ages in the ScoGFS Q4 dataset. In 2008, subject to the ACOM request, an update assessment was carried out using the same procedures as in In 2009 an update assessment was carried out using the same procedure as in This used the TSA assessment model and tuning data from the two Scottish Groundfish surveys.

27 ICES WGCSE REPORT Software used: Lowestoft VPA suite; Marine Scotland Science (Marine Lab Aberdeen) TSA and SURBA software. Input data types and characteristics: Type Name Year range Age range Variable from year to year Yes/No Caton Catch in tonnes 1966 last data 1 8+ Yes year Canum Catch-at-age in numbers 1966 last data year 1 8+ Yes Weca West Mprop Fprop Matprop Weight-at-age in the commercial catch Weight-at-age of the stock at spawning time. Proportion of natural mortality before spawning Proportion of fishing mortality before spawning Proportion mature-at-age 1966 last data year 1968 last data year 1978 last data year 1978 last data year 1978 last data year Natmor Natural mortality 1978 last data year Tuning data: 1 8+ Yes 1 8+ Yes: 1 8+ No set to 0 for all ages in all years 1 8+ No set to 0 for all ages in all years 1 8+ No the same ogive for all years 1 8+ No set to 0.2 for all ages in all years Type Name Year range Age range Research Vessel Survey Tuning fleet 1 ScoGFS-Q last data year 1 7 Tuning fleet 3 ScoGFS-Q last data year 1 7 Summary of data ranges used in recent assessments: Data 2006 assessment 2007 assessment 2008 assessment 2009 assessment Catch data Years: Ages: 1 8+ Years: Ages: 1 8+ Years: Ages: 1 8+ Years: Ages: 1 8+ Survey: ScoGFS Q1 Years: Ages: 1 7 Years: Ages 1 7 Years: Ages 1 7 Years: Ages 1 7 Survey: ScoGFS Q4 Years: Ages: 1 5 Years: Ages 1 7 Years: Ages 1 7 Years: Ages 1 7 Survey: IreGFS Not used Not used Not used Not used TSA TSA parameter settings for the analyses.

28 1398 ICES WGCSE REPORT 2013 Parameter Notation Description F (1, 1978) Initial fishing mortality F (2, 1978) Fishing mortality at age a in year y F (4, 1978) Survey selectivities Φ(1) ScoGFS Q1 Φ(2) ScoGFS Q1 survey selectivity at age a Φ(4) Survey selectivities Φ(1) ScoGFS Q4 Φ(2) ScoGFS Q4 survey selectivity at age a Φ(4) Fishing mortality standard deviations Survey catchability standard deviations Measurement coefficients of variation 2003 σf Transitory changes in overall F σu Persistent changes in selection (age effect in F) σv Transitory changes in the year effect in F σy Persistent changes in the year effect in F σω1 Transitory changes in ScoGFS Q1 catchability σβ1 Persistent changes in ScoGFS Q1 catchability * 0.00* 0.00* 0.00* 0.00* 0.00* σω2 Transitory changes in ScoGFS Q4 catchability σβ2 Persistent changes in ScoGFS Q4 catchability * 0.00* 0.00* cv landings Coefficent of variation of landings-at-age data cv discards Coefficent of variation of discards-at-age data cv survey Coefficent of variation of ScoGFS Q1 survey data cv survey Coefficent of variation of ScoGFS Q4 survey data Discard curve parameters Trend parameters Recruitment σp Transitory changes in overall discard proportion σα1 Transitory changes in discard-ogive intercept σν1 Persistent changes in discard-ogive intercept σα2 Transitory changes in discard-ogive slope σν2 Persistent changes in discard-ogive slope θν1 Trend parameter for discard-ogive intercept 0.00* 0.00* 0.00* 0.00* 0.00* 0.00* 0.00* θν2 Trend parameter for discard-ogive slope 0.00* 0.00* 0.00* 0.00* 0.00* 0.00* 0.00* η1 Ricker parameter (slope at the origin) η2 Ricker parameter (curve dome occurs at 1/η2) cv rec Coefficent of variation of recruitment curve D. Short-term projection TSA produces short-term forecasts as part of every standard model run. The recruitment values used in these forecasts have been discussed above. The model will also forecast fishing mortality rates. It does so by iterating forward the time-series model that had been fitted to historical data. These forecast mortalities therefore retain the time-series characteristics of the preceding data. However, it is not clear to the WG what the precise statistical properties of these mortality forecasts are. It is likely that they follow a pattern of damped oscillation towards an eventual steady state, but without further analysis the WG did not feel confident in using them as the basis for a forecast. Model used: Age structured Software used: MFDP prediction with management option table and yield-per-recruit routines. MLA suite (WGFRANSW) used for sensitivity analysis and probability profiles. Initial stock size. Taken from XSA or TSA for age 1 and older. The recruitment-at-age 0 in the last data year is estimated as a GM because of a perceived downward trend in recruitment in recent years. Natural mortality: Set to 0.2 for all ages in all years. Maturity: The same ogive as in the assessment is used for all years. F and M before spawning: Set to 0 for all ages in all years. Weight-at-age in the stock: based on either of simple three-year means or linear model projections: simple three year means are used for the younger ages (1 2) and linear model projections for the older ages (3 8+). Weight-at-age in the catch: as above for stock weights. Exploitation pattern: Average of the three last years. Intermediate year assumptions: status quo F.

29 ICES WGCSE REPORT Stock recruitment model used: TSA estimate of recruits-at-age 1 for intermediate year, Ricker model from TSA used for intermediate year +1 and the long-term geometric mean recruitment-at-age 1 is used for intermediate year +2. E. Medium-term projections Stochastic medium-term projections were not produced for this stock. The reliance of the fishery on intermittent large year classes and the fluid nature of the fishery and related management make the usefulness of medium-term projections questionable in any case. F. Yield and biomass per recruit/long-term projections Model used: yield and biomass per recruit over a range of F values. Software used: MFDP Selectivity pattern: mean F array from last three years of assessment (to reflect recent selection patterns). Stock and catch weights-at-age: mean of last three years. Maturity: Fixed maturity ogive as used in assessment. G. Biological reference points BPA is set at tonnes and is defined as Blim*1.4. Blim is defined as the lowest observed SSB, considered to be tonnes when the current reference points were established in FPA is 0.5 on the technical basis of a high probability of avoiding SSB falling below BPA in the long term. Flim is not defined. In the 2007 ACFM report, FMAX was estimated at 0.44 and F0.1 was 0.2. H. Other issues None. I. References ICES Report of the Working Group on the Assessment of Northern Shelf Demersal Stocks (WGNSDS). ICES CM 2005/ACFM:01.

30 1400 ICES WGCSE REPORT 2013 Table 1. Haddock in Division VIa. Commercial effort and tuning-series made available to the WG. Effort (first column) is given as reported hours fished per year; numbers landed are in thousands. Note that a) these data are not used in the final assessment, and b) 2006 data were not available to the WG in Scottish pair trawl (ScoPTR) Age Year Effort Irish otter trawl (IreOTB) Age Year Effort

31 ICES WGCSE REPORT Table 1. cont. Scottish light trawl (ScoLTR) Age Year Effort

32 1402 ICES WGCSE REPORT 2013 Haddock, West Coast Survey Q age 7 age 6 age 5 age 4 age 3 age 2 age 1 Numbers per 30 min Figure 1. Haddock in Division VIa. Number per 30 min tow, averaged over ICES statistical rectangles from the west of Scotland groundfish Q1 (IBTS) survey , ages 1 7.

33 ICES WGCSE REPORT Haddock, West Coast Survey Q age 7 age 6 age 5 age 4 age 3 age 2 age 1 Numbers per 30 min Figure 1. continued. Haddock in Division VIa. Number per 30 min tow, averaged over ICES statistical rectangles from the west of Scotland groundfish Q1 (IBTS) survey , ages 1 7.

34 1404 ICES WGCSE REPORT ScoGFSQ Haddock > 21 cm Closed areas West_Scot_Mgnt_Line VIa ScoGFSQ Haddock > 13 cm Closed areas West_Scot_Mgnt_Line VIa Figure 2. Haddock in Division VIa. Numbers per 30 min tow from the Scottish groundfish surveys (ScoGFS): Quarter 4 (2008) and Quarter 1 (2009).

35 ICES WGCSE REPORT Stock Annex 3.4: Whiting in Subarea VI Stock Working Group West of Scotland Whiting (Subarea VI) Working Group for the Celtic Seas Ecoregion (WGCSE) Date February 2012 Author Revised by Andrzej Jaworski WKROUND/Andrzej Jaworski A. General A.1. Stock definition Whiting occur throughout Northeast Atlantic waters in a wide range of depths from shallow inshore waters down to 200 m. Adult whiting are widespread throughout Division VIa, while high numbers of juvenile fish occur in inshore areas. Whiting are less common in Division VIb, and it is likely these fish are migrants from VIa, rather than a separate stock. Stock identity in Division VIa has recently been explored in greater detail. Tagging experiments on recruiting fish have shown that the whiting found to the south of 56 N and to the west of Ireland are distinct from those in the Minches, the Clyde and the Irish Sea. Five juvenile nursery areas have been discriminated off the west of Scotland and northern North Sea, three of them being found in VIa. The nursery areas on the Scottish west coast contribute individuals to the spawning aggregations in the Scottish coastal North Sea and Shetland, and there is no evidence of the converse (Tobin et al., 2010). Within VIa, there is little indication of interaction between population components in the south and that off the northwest coast. A.2. Fishery The demersal fisheries in Division VIa are predominantly conducted by otter trawlers fishing for cod, haddock, anglerfish and Nephrops, with bycatch of whiting, saithe, megrim, lemon sole, ling and a number of skate species. Whiting are taken by trawlers using gear with mesh size between 80 mm and 120 mm. Since 1976, effort by Scottish heavy trawlers and seiners has decreased. Light trawler effort has declined rapidly since 1997 after a long-term increasing trend. More recently, days-at-sea limitations associated with the cod recovery plan and the seasonal closure of some areas has led to some switching of effort away from VIa. The demersal whitefish fishery in Subarea VI occurs largely in Division VIa with the UK, Ireland and France being the most important exploiters. Landings from Rockall (Division VIb) are generally less than 10 t. The whiting fishery in VIa is dominated by the UK (Scotland) and Irish fleets. French whiting landings have declined considerably since the late 1980s. Landings of whiting in Division VIa are affected by emergency measures introduced in 2001 as part of the cod recovery programme. Council Regulation 423/2004 introduced a cod recovery plan affecting Division VIa. The measures only take effect, however east of a line defined in Council Regulation No 51/2006. Measures brought in 2002, such as a switch from 100 to 120 mm mesh codends at the start of 2002 (Commission Regulation EC2056/2001), are likely to have had some impact on whit-

36 1406 ICES WGCSE REPORT 2013 ing. The UK implemented a regulation requiring the fitting of a square mesh panel in certain towed gears. Most catch of whiting comes in non-whiting directed fisheries, particularly the Nephrops trawl fishery. The Nephrops trawl fishery in VIa discards significant amounts of small whiting, making whiting landings figures a poor indicator of removals due to fishing. The proportion of whiting discarded has been very high and appears to have increased in recent years. Whiting also has a low market demand, which contributes to increased discarding and highgrading. In terms of the total weight of demersal fish landed by the Scottish fleet from the west coast, whiting is ranked fourth, with an annual value of (in 2009). The minimum landing size of whiting in the human consumption fishery in this area is 27 cm. There have been some problems regarding area misreporting of Scottish landings during the early 1990s, which are linked to area misreporting of other species such as haddock and anglerfish into Division VIb. More recently there has been area misreporting of anglerfish from VIa to IVa, which may have affected the reliability of whiting landings distribution. A.3. Ecosystem aspects Unlike some species, whiting do not form distinct spawning shoals, and both ripe and immature fish are often found together. As the latitude increases, spawning of whiting occurs progressively later. This is closely associated with temperature changes, but spawning activity generally peaks in springtime, just as sea temperatures begin to rise. On the west coast of Scotland whiting spawn between January and June. Within this period, the spawning season of an individual female lasts around fourteen weeks, during which time she releases many batches of eggs. At two years old most whiting are mature and able to spawn. By the time it reaches four years old, a single female fish of reasonable size can produce more than eggs. Like many other fish, whiting spend their first few months of life in the upper water layers before moving to the seabed. Male and female whiting grow very quickly reaching around 19 cm in their first year. After this the growth rate becomes much slower. There are large differences between the growth rates of individual fish and a 30 cm fish can be as young as one year or as old as six. Whiting are active predators. Juvenile fish eat mainly crustaceans (shrimps and crabs) but as whiting grow, the amount of fish in their diet increases. The exact composition of the diet depends on the size of the fish, the area and the time of the year. Whiting is one of the main predators of other commercially important species of fish. Norway pout, sandeels, haddock, cod and even whiting themselves are frequently eaten. It has been estimated that each year the whiting population consumes several hundred thousand tonnes of these species. B. Data B.1. Commercial catch Monthly length frequency distribution data were available from Scotland for Division VIa. A total international catch-at-age distribution for Division VIa was obtained by raising this distribution to the WG estimates of total international catch from this area. Landings officially reported to ICES were used for countries not supplying estimates directly to the WG. The Scottish market sampling length weight relationships

37 ICES WGCSE REPORT (given below) have been used to raise the sampled catch-at-length distribution data Working Group estimates of total landings for Division VIa. Month b a Discard data are available from 1978 but sampling was very limited before To reduced bias and increase precision, discards in years were raised according to the procedure described in Millar and Fryer (2005). Discard age-compositions are generally available from both Scotland and Ireland, but in some recent years (2006 and 2007) lack of access to fishing vessels by Irish observers has meant that no Irish data have been collected. B.2. Biological Natural mortality (M) is assumed to vary and be dependent on fish weight (Lorenzen, 1996). M values are time-invariant and are calculated as: where 0.29 = 3.0 a M a W M a is natural mortality-at-age a, W a is the time averaged stock weight-at-age a (in grammes) and the numbers are the Lorenzen parameters for fish in natural ecosystems. A combined sex maturity is assumed, knife-edged at age 2. The use of a knife-edged maturity ogive has been a source of criticism in previous assessments. However, recent research on gadoid maturity conducted by the UK gives no evidence for substantial change in whiting maturity since the 1950s, although there has been an increase in the incidence of precocious maturity-at-age 1, particularly in males, since 1998, in the Irish Sea. B.3. Surveys Six research vessel survey-series for whiting in VIa were available to the WKROUND In all surveys listed, the highest age represents a true age not a plus group. Scottish first quarter west coast groundfish survey (ScoGFS-WIBTS-Q1): ages 1 7, years ). Scottish fourth quarter west coast groundfish survey (ScoGFS-WIBTS-Q4): ages 0 8, years ). The Q1 Scottish Groundfish survey was running in the period , and this was performed using a repeat station format with the GOV survey trawl together with the west coast groundgear rig, C. Similarly the Q4 Scottish Groundfish survey

38 1408 ICES WGCSE REPORT 2013 was running in , once again using the GOV survey trawl with groundgear C and the fixed station format. In 2011, the Q1 and Q4 Scottish Groundfish surveys were re-designed. The previous repeat station survey format consisting of the same series of survey trawl positions being sampled at approximately the same temporal period every year is considered a rather imprecise method for surveying both these subareas and as such a move towards some sort of random stratified survey design was judged necessary. The largest obstacle preventing an earlier move to a more randomised survey design was the lack of confidence in the C rig to tackle the potentially hard substrates that a new randomised survey was likely to encounter. The first step in the process of modifying the survey design was therefore to design a new groundgear that would be capable of tackling such challenging terrain. The introduction of the new design initiated two new time-series: Scottish first-quarter west coast groundfish survey (no acronym assigned yet): ages 1 7, years ). Scottish fourth quarter west coast groundfish survey (no acronym assigned yet): ages 0 8, years 2011 ). ICES will consider inclusion of the above time-series to produce tuning indices through an inter-benchmark procedure when 4+ years of data have been gathered. The Irish groundfish surveys: Irish fourth-quarter west coast groundfish survey (IreGFS): ages 0 5, years The Irish quarter four survey was a comparatively short series, was discontinued in 2003 and has been replaced by the IGFS. Irish fourth quarter west coast groundfish survey (IGFS-WIBTS-Q4): ages 0 6, years This survey used the RV Celtic Explorer and is part of the IBTS coordinated western waters surveys. The vessel uses a GOV trawl, and the design is a depth stratified survey with randomised stations. Effort is recorded in terms of minutes towed. Further descriptions of these surveys and distribution plots of whiting catch rates obtained on these surveys can be found in the IBTS WG Report of B.4. Commercial cpue Due to a number of concerns regarding the non-mandatory recording of effort in terms of hours fished, the present assessment of the stocks does not make use of commercial catch per unit of effort data. The data are included here for completeness and include: Scottish light trawlers (ScoLTR): ages 1 7 years Scottish seiners (ScoSEI): ages 1 6 years Scottish Neprhops trawlers (ScoNTR): ages 1 6 years Irish Otter Trawlers (IreOTB): ages 1 7 years Data to update these time-series were not available for the recent years.

39 ICES WGCSE REPORT B.5. Other relevant data Fecundity data for a number of areas are available from Hislop and Hall (1974), and was estimated at L 3.25 for whiting in Subarea VI. C. Assessment: data and method Model used: TSA Software used: NAG library (FORTRAN DLL) and functions in R. Model Options chosen: Weight-dependent M after Lorenzen (1996) Mwght.b < Mwght.Mu <- 3.0 Response: landings-at-age, discards-at-age and survey indices-at-age Commercial data : treated as unbiased : age structure only used (with unaccounted mortality estimated) : treated as unbiased landings cvmult-at-age = c(2, 1, 1, 1, 1, 1, 2): extra variability for ages 1 and 7+ discards cvmult-at-age = c(1, 1, 1, 1, 2): extra variability for age 5 discards cvmult = 3 for age 1 in 1981, age 1 in 1987, age 3 in 1991, age 1 in 2000 Discard model full model : ages 1 5 modelled Stock recruit model hockey stick Fishing selection model amat = 4: fishing selection flat (apart from noise) from age 4 gudmundssonh1 = c(2, 1, 1, 1, 1, 1, 1): extra variability for age 1 Survey model (ScoGFS-WIBTS-Q1) full model: separate catchability for each age ages 1 to 6 modelled transitory and persistent changes in catchability allowed cvmult = 3 for age 5 in 1992, age 2 in 1993, age 1 in 2000, age 2 in 2000 cvmult = 5 for age 4 in 1992 Survey model (ScoGFS-WIBTS-Q4) full model: separate catchability for each age ages 1 to 6 modelled transitory and persistent changes in catchability allowed

40 1410 ICES WGCSE REPORT 2013 cvmult-at-age = c(1, 1, 1, 1, 1, 2): extra variability for age 6 cvmult = 3 for age 4 in 2007, age 5 in 2007 Survey model (IGFS Q4 IGFS-WIBTS-Q4) full model: separate catchability for each age ages 1 to 4 modelled years and (year 2007 excluded due to a high prediction error) transitory and persistent changes in catchability allowed Input data types and characteristics: Type Name Year range Age range Variable from year to year Yes/No Caton Catch in tonnes to 7+ Yes Canum Catch at age in to 7+ Not used numbers?? Landings at age in to 7+ Yes numbers?? Discards at age in to 7+ Yes numbers Weca Weight at age in to 7+ Not used the commercial catch?? Weight at age in to 7+ Yes the commercial landings?? Weight at age in to 7+ Yes the commercial discards Weca Weight at age in to 7+ Not used the commercial catch West Weight at age of to 7+ Not used the spawning stock at spawning time. Mprop Proportion of to 7+ No natural mortality before spawning Fprop Proportion of to 7+ No fishing mortality before spawning Matprop Proportion mature to 7+ No at age Natmor Natural mortality to 7+ No

41 ICES WGCSE REPORT Tuning data: Type Name Year range Age range Tuning fleet 1 ScoGFS-WIBTS-Q Tuning fleet 2 ScoGFS-WIBTS-Q Tuning fleet 3 IGFS-WIBTS-Q D. Short-term projection Not done. E. Medium-term projections No medium-term projections are carried out for this stock. F. Long-term projections No long-term projections are carried out for this stock. G. Biological reference points Type Value Technical basis MSY MSY No estimate Btrigger Approach FMSY No estimate Blim t ICES proposition Precautionary Bpa t ICES proposition Approach Flim 1.0 ICES proposition Fpa 0.6 ICES proposition H. Other issues H.1. Historical overview of previous assessment methods Data 2008 assessment 2009 assessment 2010 assessment 2011 assessment Catch data No assessment No assessment Years: Ages: 1 7+ Years: and Ages: 1 7+ Survey: ScoGFS Q1 No assessment No assessment Years: Ages 1 6 Years: Ages 1 6 Survey: No assessment No assessment Not used Not used ScoGFS Q4 Survey: IRGFS Q4 No assessment No assessment Not used Not used

42 1412 ICES WGCSE REPORT 2013 I. References Hislop, J. R. G., Hall, W. B The fecundity of whiting, Merlangius merlangus (L.) in the North Sea, the Minch and at Iceland. Journal du Conseil International pour L'exploration de la Mer, 36: Millar, C. P., Fryer, R., J Revised estimates of Annual discards-at-age for cod, haddock, whiting and saithe in ICES Subarea IV and Division VIa. Fisheries Research Services internal report No 15/05, July 2005, 23 pp. Lorenzen, K The relationship between body weight and natural mortality in juvenile and adult fish: a comparison of natural ecosystems and aquaculture. Journal of Fish Biology, 49: Tobin, D., Wright, P. J., Gibb, F. M., Gibb, I. M The importance of life stage to population connectivity in whiting (Merlangius merlangus) from the northern European shelf. Marine Biology, 157:

43 ICES WGCSE REPORT Stock Annex 3.5: Nephrops in FU 11 North Minch Stock North Minch Nephrops (FU 11) Working Group WGCSE Date: 01/04/2013 Revised by (WKNEPH2013/Carlos Mesquita) A. General A.1. Stock definition The North Minch Functional Unit (FU 11) is located off the northwest coast of Scotland. The northern boundary of the FU is the 59 N line and the boundary with the South Minch FU is at N. The North Minch FU is characterised by numerous islands of varying size and sea lochs occur along the mainland coast and exhibits the most patchy ground amongst west coast FUs. Throughout its distribution, Nephrops is limited to muddy habitat, and requires sediment with a silt & clay content of between % to excavate its burrows. This means that the distribution of suitable sediment defines the species distribution. The sediment data from the British Geological survey is considered incomplete in this FU, therefore the area of the ground is given by the VMS distribution of fishing effort (vessels >15 meters). Results from recent work on mapping the spatial extent of Nephrops habitat in the North Minch sea lochs indicate that the muddy habitat is only a very small proportion of the total Nephrops grounds (see Section ). The total area of the ground is estimated to be 2908 km 2 (Figure B1-4). The North Minch is part of Division VIa and the fishery data for this Functional Unit includes the following statistical rectangles: E3 E4 (Figure B1-3). A.2. Fishery The North Minch Nephrops fishery is predominantly exploited by Nephrops trawlers using single rig gear with an 80 mm mesh, although about 20% of landings are currently made by creel vessels. Landings for this FU are only reported from Scotland. The fleet is mainly formed by smaller trawlers working 1 4 day trips from the main ports of Lochinver, Ullapool, Stornoway and Gairloch. The largest part of the North Minch fleets continued to be based at Stornoway, made up of mostly 15 m length vessels, both single rigged and twin-rigged trawlers. The Barra fleet is more nomadic as the fishing grounds are more exposed which forces the fleet to find shelter on the east side of the North Minch. The Barra vessels are generally bigger than the Stornoway fleet, being all over 15 m in length. The minimum landing size for Nephrops in the North Minch is 20 mm CL, and less than 1% of the animals are landed under size. Discarding takes place at sea, and landings are made by category for whole animals (small, medium and large) and as tails. The main bycatch species is haddock, although whiting and Norway pout also feature significantly in discards. The fishery is exploited throughout the year, with the highest landings usually made in the spring and summer. Vessels usually have a trip duration of one day in the winter, but up to six days in the summer. The current legislation governing Nephrops trawl fisheries on the West coast of Scotland was laid down by the North Sea and West of Scotland cod recovery plan (EC

44 1414 ICES WGCSE REPORT /2001), which established additional measures to EC 850/98. This regulation was amended in 2003 by Annex XVII of EC 2341/2002, which establishes fishing effort and additional conditions for monitoring, inspection and surveillance for the recovery of certain cod stocks. For 2012, this regulation effectively limits vessels targeting Nephrops with mm mesh size to 200 days at sea per year. Additional Scottish legislation (SSI No 2000/226) applies to twin trawlers operating north of 56 N. A mesh size of 100 mm or above must be used without a lifting bag and with not more than 100 meshes round the circumference but with up to 5 mm double twine. By comparison, vessels using a single trawl may use mm mesh with a lifting bag and 120 meshes round the codend but with 4 mm single twine. From 2009 onwards under the west coast emergency measures a square meshed panel of 120 mm was also required (Council Reg. (EU) 43/2009). A.3. Ecosystem aspects No information on the ecosystem aspects of this stock has been collated by the working group. B. Data B.1. Commercial catch Length and sex compositions of Nephrops landed from the North Minch are estimated from port sampling in Scotland. Length data from Scottish sampling are applied to all catches. Rates of discarding by length class are estimated for Scottish fleets by onboard sampling. The proportion of discarded to landed Nephrops changes with year, often determined by strong year classes. Discard sampling started in 1990, and for years prior to this estimates have been made based on later data. Landings and discards-at-length are combined to removals. The discard survival rate for creel caught Nephrops have been shown to be high and a value of 100% is used. Removals are raised separately for each sex. Reported effort by all Scottish trawlers has shown a decreasing trend since 2000 (Figure B1-1). The increase in lpue in 2005 is probably reflecting the increase in reported landings rather than a change in stock abundance. In general, males make the largest contribution to the landings (Figure B1-2). This is likely to be due to the varying seasonal pattern in the fishery and associated relative catchability (due to different burrow emergence behaviour) of male and female Nephrops. This occurs because males are available throughout the year and the fishery is also prosecuted in all quarters. Females on the other hand are mainly taken in the summer when they emerge after egg hatching. The mean size of smaller animals (<35 mm) in the catch (and landings) is also relatively stable through time (Figure B1-1). Trawl and creel fisheries are sampled separately. B.2. Biological Mean weights-at-age for this stock are estimated from fixed Scottish weight length relationships (unpublished data). The size-at-maturity was estimated by Queirós et al. (2013). Relevant biological parameters are as follows: natural mortality was assumed to be 0.3 for males (Morizur, 1982) of all ages and in all years. Natural mortality was assumed to be 0.3 for immature females, and 0.2 for mature females.

45 ICES WGCSE REPORT Summary PARAMETER VALUE SOURCE Discard Survival (trawl) 25 % Charuau et al., 1982; Sangster et al., 1997; Wileman et al., 1999 Discard Survival (creel) 100 % Wileman et al., 1999; Harris and Ulmestrand (2004); Chapman,1981 MALES Growth K 0.16 Adapted from Bailey and Chapman (1983) Growth - L(inf) 70 mm Adapted from Bailey and Chapman (1983) Natural mortality - M 0.3 Morizur, 1982 Length/weight - a Howard and Hall (1983) Length/weight - b 3.24 Howard and Hall (1983) Size at maturity 27 mm Adapted from Bailey and Chapman (1983) FEMALES Immature Growth Growth K 0.16 Adapted from Bailey and Chapman (1983) Growth - L(inf) 70 mm Adapted from Bailey and Chapman (1983) Natural mortality - M 0.3 As for males Size at maturity 22 mm Queirós et al., (2013) Mature Growth Growth K 0.06 Adapted from Bailey and Chapman (1983) Growth - L(inf) 60 mm Adapted from Bailey and Chapman (1983) Natural mortality - M 0.2 Length/weight - a Howard and Hall (1983) Length/weight - b 2.91 Howard and Hall (1983) Discard survival A discard survival of 25% is assumed for the trawl fleet (Charuau et al., 1982; Sangster et al., 1997; Wileman et al., 1999). The discard survival rate for creel caught Nephrops have been shown to be high (see Section ) and a value of 100% is used. B.3. Surveys Abundance indices are available from the following research-vessel surveys: Underwater TV survey (UWTV FU 11): years 1994 present. The survey usually occurs in May/June. The burrowing nature of Nephrops, and variable emergence rates mean that trawl catch rates may bear little resemblance to population abundance. An underwater TV survey has been developed, estimating Nephrops population abundance from burrow density raised to stock area. The methods used in the survey were similar to those employed for UWTV surveys of Nephrops stocks around Scotland and are documented by WKNEPHTV (ICES, 2007) and SGNEPS (ICES, 2010; ICES, 2012). In the assessment, burrow densities are raised to the total estimated area. The survey provides a total abundance estimate, and is not age or length structured. Samples are distributed randomly over the area of suitable sediment. The area calculation was based on the alpha convex-hull method to define and characterize the overall shape of a set of points and is described in ICES (2010). A number of annual polygons based on the VMS distribution of effort ( ) was generated and the union of these used to define the area of Nephrops ground in the North Minch. The VMS area was updated in 2013 at the WKNEPH2013 and estimated to be 2908 km 2 (Section ).

46 1416 ICES WGCSE REPORT 2013 UWTV relative to absolute conversion factors A number of factors are suspected to influence the ability of the surveys to map directly to absolute abundance. In order to use the survey abundance estimate as an absolute it is necessary to correct for these potential biases. The history of bias estimates is given in the following table and are based on simulation models, preliminary experimentation and expert opinion (ICES, 2009). The biases associated with the estimates of Nephrops abundance in the North Minch are: TIME PERIOD EDGE EFFECT DETECTION RATE SPECIES IDENTIFICATION OCCUPANCY CUMULATIVE ABSOLUTE CONVERSION FACTOR FU 11: North Minch B.4. Commercial cpue Landings-per-unit-of-effort time-series are available from the following fleets: Scottish trawl gears: Landings and effort data for Scottish trawl gears are used to generate a non-standardized lpue index. Lpue is estimated using officially recorded effort (days absent from port). Effort data are available for the trawl fleet from There is no account taken of any technological creep in the fleet. Effort data for the creel fleet are not available. B.5. Other relevant data None. C. Assessment: data and method Model used: UWTV Based Approach to generate catch options In 2009 WKNEPH debated the use of the surveys as either an absolute measure of abundance or a relative index (ICES, 2009). Ultimately this led to a consensus that bias corrected survey abundance estimates could be used directly in the formulation of catch advice. Two modelling approaches were used to estimate sustainable stockspecific Harvest Ratio reference points; SCA (a separable LCA model Bell) & Age Structured Simulation model (Dobby) (ICES, 2009). Software used: Age Structured Simulation model per recruit analysis in R 1 ) Survey indices are worked up annually resulting in the TV index. 2 ) Apply the Absolute Conversion Factor (see Section B3). The combined effect of these biases is to be applied to the new survey index. 3 ) Generate mean weight in landings. Check the time-series of mean landing weights for evidence of a trend in the most recent period. If there is no firm evidence of a recent trend in mean weight use an average taken over an appropriate time scale. If, however, there is strong evidence of a recent trend then apply most recent value (don t attempt to extrapolate the trend further in the future). 4 ) The catch option table will include the harvest ratios associated with fishing at F0.1, F35%SpR and FMAX. These values are estimated by benchmark workshops (Section 3.5.5) but may be revised if there are indications of changes to fisheries or biological factors.

47 ICES WGCSE REPORT ) Create catch option table on the basis of a range of harvest ratios ranging from 0 to the maximum observed ratio or the ratio equating to FMAX, whichever is the larger. Insert the harvest ratios from step 4 and also the current harvest ratio. 6 ) Multiply the survey index by the harvest ratios to give the number of total removals. 7 ) Create a landings number by applying the discard ratio (dead discard rate). 8 ) Produce landings biomass by applying mean weight. E. Medium-term projections None presented. F. Long-term projections None presented. G. Biological reference points Under the ICES MSY framework, exploitation rates which are likely to generate high long-term yield (and low probability of overfishing) have been evaluated and proposed for each Nephrops functional unit. Owing to the way Nephrops are assessed, it is not possible to estimate FMSY directly and hence proxies for FMSY have been determined. Three stock-specific candidates for FMSY (F0.1, F35%SPR, and Fmax) were derived from a length-based per recruit analysis (these may be modified following further data exploration and analysis). There may be strong differences in relative exploitation rates between the sexes in many stocks. To account for this, values for each of the candidates have been determined individually for males, females, and the two sexes combined. The combined sex FMSY proxy should be considered appropriate, provided that the resulting percentage of virgin spawner-per-recruit for males or females does not fall below 20%. If this happens a more conservative sex-specific FMSY proxy should be picked instead of the combined proxy. In the North Minch the absolute density observed on the UWTV survey is medium (~0.59 burrow/m 2 ). Historical harvest ratios in this FU have been above that equivalent to fishing at FMAX and landings have been relatively stable in the last thirty years. F35%SpR (combined between sexes) is expected to deliver high long-term yield with a low probability of recruitment overfishing and therefore is chosen as a proxy for FMSY. These calculations assume that the TV survey has a knife-edge selectivity at 17 mm and that the supplied length frequencies represented the population in equilibrium. The MSY Btrigger proposed for North Minch was based on the lowest observed UWTV abundance time-series. The FMSY proxy harvest rate values were updated at WKNEPH2013 from the perrecruit analysis based on input parameters from a combined sex length cohort analysis of catch-at-length data. All FMSY proxy harvest rate and MSY Btrigger values remain preliminary and may be modified following further data exploration and analysis.

48 1418 ICES WGCSE REPORT 2013 HARVEST RATIO REFERENCE POINTS: Male Female Combined Fmax F F35%SpR TYPE VALUE TECHNICAL BASIS MSY MSY Btrigger 541 million individuals Approach FMSY 10.9% harvest rate Bias-adjusted lowest observed UWTV survey estimate of abundance (corrected for the new VMS area estimate) Equivalent to F35%SpR combined sex. FMSY proxy based on length based Y/R. H. Other issues H.1. Historical overview of previous assessment methods Up to 2010 the ground area for the North Minch was based on the British Geological Survey (BGS) and estimated as 1775 km 2. Marine Scotland Science recent access to Vessel Monitoring System data (VMS) has shown that fishing effort for trawlers (length >15 m) clearly extends outside of the BGS area for FU 11, which would imply an underestimate of the stock area. In the 2011 and 2012 assessments, a preliminary VMS based area estimated as 2506 km 2 was used for raising abundances. A correction ratio calculated as 1.41 (VMS area / Sediment area) was applied to the previous sediment abundances estimates to get a rough measure of the abundance raised to the VMS area. As more VMS data became available since 2010, in 2013 at the WKNEPH2013 the sediment area of North Minch was updated to 2908 km 2. This was based on the union of annual polygons produced from the VMS data which was shown to be the best method to define the ground area in FU 11 as it includes the main fishing areas while it excludes some low intensity areas. The correction ratios to be applied to the previous abundance estimates are now 1.64 (new VMS area / Sediment area) for years and 1.16 (new VMS area / preliminary VMS area,) for years I. References Bailey, N., and Chapman, C. J A comparison of density, length composition and growth of two Nephrops populations on the west coast of Scotland. ICES CM 1983/K:42. Chapman C.J Discarding and tailing Nephrops at sea. Scott Fish Bull 46: Charuau A., Morizur Y., Rivoalen J.J Survival of discarded Nephrops norvegicus in the Bay of Biscay and in the Celtic Sea, ICES-CM-1982/B:13. Harris, R. R. and Ulmestrand, M Discarding Norway lobster (Nephrops norvegicus L.) through low salinity layers - mortality and damage seen in simulation experiments. ICES Journal of Marine Science 61, Howard F.G. and Hall, W.B Some observations on the biometrics of Nephrops norvegicus (L.) in Scottish waters. ICES, Doc.ShellfishComm.,CM1983/K:36. ICES Report of the Workshop on the use of UWTV surveys for determining abundance in Nephrops stocks throughout European waters (WKNEPHTV). ICES CM: 2007/ACFM: 14 Ref: LRC, PGCCDBS.

49 ICES WGCSE REPORT ICES Report of the Benchmark Workshop on Nephrops assessment (WKNEPH). ICES CM: 2009/ACOM:33. ICES Report of the Study Group on Nephrops Surveys (SGNEPS), 9 11 November 2010, Lisbon, Portugal. ICES CM 2010/SSGESST: pp. ICES Report of the Study Group on Nephrops Surveys (SGNEPS), 6 8 March 2012, Acona, Italy. ICES CM 2012/SSGESST: pp. Morizur, Y Estimation de la mortalité pour quelques stocks de la langoustine, Nephrops norvegicus (L.). ICES, Doc. Shellfish Comm., CM 1982/K:10 (mimeo). Sangster, G.I., Breen, M., Bova, D.J., Kynoch, R., O Neill, F.G., Lowry. N., Moth-Poulsen, T. Hansen, U.J., Ulmestrand, M., Valentinsson, D., Hallback, H., Soldal, A.V., and Hoddevik, B Nephrops survival after escape and discard from commercial fishing gear. Presented at ICES FTFB Working Group, Hamburg, Germany April, 1997, ICES CM 1997 CM/B. Wileman, D. A., Sangster, G. I., Breen, M., Ulmestrand, M., Soldal, A. V. and Harris, R. R Roundfish and Nephrops survival after escape from commercial fishing gear. Final report to European Commission, Brussels, FAIR-CT Queirós, A.M., Weetman, W., McLay, H.A., Dobby, H., Geographical variation in size at the onset of maturity of male and female Norway lobster Nephrops norvegicus (L., Homarida: Decapoda) in Scottish waters. Fisheries Research 139,

50 1420 ICES WGCSE REPORT 2013 Table B1-1. Nephrops, North Minch (FU11). Nominal Landings of Nephrops, YEAR UK SCOTLAND Trawl landings Creel Total** * * provisional na = not available. ** There are no landings by other countries from this FU.

51 ICES WGCSE REPORT Figure B1-1. Nephrops, North Minch (FU11). Long term landings, effort, lpue and mean sizes. The interpretation of the lpue series is likely to be affected by the introduction of the buyers and sellers regulations in 2006.

52 1422 ICES WGCSE REPORT 2013 Figure B1-2. Nephrops, North Minch (FU11). Landings by quarter and sex from Scottish trawlers.

53 ICES WGCSE REPORT Lattitude VIa South Minch North Minch Cly de VIIa Irish Sea Irish Sea West East Longitude Figure B1-3. Nephrops Functional Units in VIa and VIIa. North Minch (FU11), South Minch (FU12), Clyde (FU13), Irish Sea East (FU14) and Irish Sea West (FU15).

54 1424 ICES WGCSE REPORT 2013 Figure B1-4. Nephrops, North Minch (FU11). Nephrops ground area (shown in red) estimated using VMS data (2908 km 2 ).

55 ICES WGCSE REPORT Stock Annex 3.6: Nephrops FU12, South Minch Stock South Minch Nephrops (FU 12) Date Updated Revised by 09 March 2009 (WKNEPH2009) May 2013 (WGCSE) Sarah Clarke/Carlos Mesquita A. General A.1. Stock definition Throughout its distribution, Nephrops is limited to muddy habitat, and requires sediment with a silt & clay content of between % to excavate its burrows, and this means that the distribution of suitable sediment defines the species distribution. Adult Nephrops only undertake very small scale movements (a few 100 m) but larval transfer may occur between separate mud patches in some areas. In the South Minch area the Nephrops stock inhabits a generally continuous area of muddy sediment extending from the south of Skye to the Stanton Bank, to the south of the Outer Hebrides. The South Minch functional unit (FU12) is located off the west coast of Scotland, and is bounded to the north and south by the and circles of latitude, and to the west by the 8 W meridian. Out with the functional unit, a mixed fishery for gadoids and Nephrops takes place on Stanton Bank, to the south-west of the Outer Hebrides. A.2. Fishery The South Minch Nephrops fishery is predominantly exploited by Nephrops trawlers, although about 20% of landings are made by creel vessels, which has increased in recent years. About 90% of trawler landings are made by vessels targeting Nephrops, and 25% of landings were made by twin-rig vessels in All the creel vessels are local, and roughly half of the trawl landings are made by vessels based between Mallaig and Campbeltown. Visiting vessels originate from the North Minch and the Scottish east coast. The east coast vessels tend to be larger than the local ones, and carry out longer trips. Mean engine power of the local vessels is 200 kw, and their mean length 15.0 m. Most vessels were built between the 1960s and the 1980s. The major landing ports are Oban and Mallaig. The smaller vessels usually have a trip duration of 1 3 days, while larger boats may stay out for 5 6 days. The minimum landing size for Nephrops in the South Minch is 20 mm CL and less than 1% of animals are landed under size. Discarding takes place at sea and landings are made by category for whole animals (small and large) and as tails. The main bycatch species are whiting and haddock, with whiting in particular featuring heavily in discards. Of the non-commercial species caught, poor cod, Norway pout and long rough dab contribute significantly to the discards. The fishery is exploited throughout the year, with the highest landings usually being made in the spring and summer. A seasonal sprat fishery often develops in November and December, which is targeted by vessels of all sizes (including those that usually target Nephrops). Some vessels also turn to scallop dredging when Nephrops catches or prices drop, although the scope for this has been limited in recent years with ASP and PSP closures of the scallop fishery in some areas.

56 1426 ICES WGCSE REPORT 2013 The current legislation governing Nephrops trawl fisheries on the west coast of Scotland was laid down by the North Sea and west of Scotland cod recovery plan (EC 2056/2001), which established measures additional to EC 850/98. This regulation was amended in 2003 by Annex XVII of EC 2341/2002, which establishes fishing effort and additional conditions for monitoring, inspection and surveillance for the recovery of certain cod stocks. For 2012, this regulation effectively limits vessels targeting Nephrops with mm mesh size to 200 days at sea per year. The use of square mesh and headline panels are compulsory in this fishery. Additional Scottish legislation (SSI No 2000/226) applies to twin trawlers operating North of 56 0 N, A mesh size of 100 mm or above must be used without a lifting bag and with not more than 100 meshes round the circumference but with up to 5 mm double twine. By comparison, vessels using a single trawl may use mm mesh with a lifting bag and 120 meshes round the codend but with 4 mm single twine. From 2009 onwards under the west coast emergency measures a square meshed panel of 120 mm was also required (Council Reg. (EU) 43/2009). A.3. Ecosystem aspects No information on the ecosystem aspects of this stock has been collated by the working group. B. Data B.1. Commercial catch Length and sex compositions of Nephrops landed from the South Minch are estimated from port sampling in Scotland. Length data from Scottish sampling are applied to all catches and raised to total international landings. Rates of discarding by length class are estimated for Scottish fleets by on-board sampling, and extrapolated to all other fleets. The proportion of discarded to landed Nephrops changes with year, often determined by strong year classes. Discard sampling started in 1990, and for years prior to this estimates have been made based on later data. Landings and discards at length are combined (assuming a discard survival rate of 25%) to removals. The differences in catchability between sexes have led to the two sexes being assessed separately. And hence removals are raised separately for each sex. Reported Nephrops trawl effort in 2007 was similar to the four previous years, whilst total landings show a marked increase since 2006 (Figure B1-1, Table B1-1), possibly as a result of more accurate reporting since the introduction of the buyers and sellers regulations in the UK in this year. Males contribute more to the landings than females (Figure B1-2), as in all other functional units. Effort is normally highest in the 2nd quarter in this fishery, and generally lowest in the 4th quarter. Lpue has remained relatively stable over the time-series prior to 2005, but shows a marked increase in 2006 and 2007, possibly as a result of the aforementioned introduction of the buyers and sellers regulations. Discarding of undersize and unwanted Nephrops occurs in this fishery, and quarterly discard sampling has been conducted on the Scottish Nephrops trawler fleet since Discarding rates averaged over the period 2010 to 2012 for this stock were 9% by number. This represents a decrease on the 2000 to 2009 period. Mean length data for Nephrops above and below 35 mm CL, are shown in Figure B1-2. This size was chosen for all the Scottish stocks examined as the general size limit

57 ICES WGCSE REPORT above which the effects of discarding practices and the addition of recruits were likely to be small. Trawl and creel fisheries are sampled separately. B.2. Biological Mean weights-at-age for this stock are estimated from fixed Scottish weight length relationships (Howard et al., 1988-citation required). Relevant biological parameters are as follows: natural mortality was assumed to be 0.3 for males (Morizur, 1982) of all ages and in all years. Natural mortality was assumed to be 0.3 for immature females, and 0.2 for mature females. Summary PARAMETER VALUE SOURCE Discard Survival (trawl) 25% Charuau et al., 1982; Sangster et al., 1997; Wileman et al., 1999 Discard Survival (creel) 100% Wileman et al., 1999; Harris and Ulmestrand (2004); Chapman,1981 MALES Growth K 0.16 Adapted from Bailey and Chapman (1983) Growth - L(inf) 66 mm Adapted from Bailey and Chapman (1983) Natural mortality - M 0.3 Morizur, 1982 Length/weight - a Howard and Hall (1983) Length/weight - b 3.24 Howard and Hall (1983) Size at maturity 25 mm Adapted from Bailey and Chapman (1983) FEMALES Immature Growth Growth K 0.16 Adapted from Bailey and Chapman (1983) Growth - L(inf) 66 mm Adapted from Bailey and Chapman (1983) Natural mortality - M 0.3 As for males Size at maturity 25 mm Adapted from Bailey and Chapman (1983) Mature Growth Growth K 0.06 Adapted from Bailey and Chapman (1983) Growth - L(inf) 59 mm Adapted from Bailey and Chapman (1983) Natural mortality - M 0.2 Length/weight - a Howard and Hall (1983) Length/weight - b 2.91 Howard and Hall (1983) Proportion of F and M prior to spawning was specified as zero to give estimates of spawning stock biomass at January 1. B.3. Surveys Abundance indices are available from the following research vessel surveys: Underwater TV survey: years 1995 present. The survey usually occurs in June. The burrowing nature of Nephrops, and variable emergence rates mean that trawl catch rates may bear little resemblance to population abundance. An underwater TV survey has been developed, estimating Nephrops population abundance form burrow density raised to stock area. A random stratified sampling design is used, on the basis of British Geological Survey sediment strata. The survey provides a total abundance estimate, and is not age or length structured (Figure B1-3).

58 1428 ICES WGCSE REPORT 2013 UWTV relative to absolute conversion factors A number of factors are suspected to influence the ability of the surveys to map directly to absolute abundance. In order to use the survey abundance estimate as an absolute it is necessary to correct for these potential biases. The history of bias estimates is given in the following table and are based on simulation models, preliminary experimentation and expert opinion (ICES, 2009). The biases associated with the estimates of Nephrops abundance in the South Minch are: TIME PERIOD EDGE EFFECT DETECTION RATE SPECIES IDENTIFICATION OCCUPANCY CUMULATIVE ABSOLUTE CONVERSION FACTOR FU 12: South Minch <= B.4. Commercial cpue Landings-per-unit-effort time-series are available from the following fleets: Scottish trawl gears: Landings and effort data for Scottish trawl gears are used to generate a non-standardized lpue index. Lpue is estimated using officially recorded effort (days absent from port). Effort data are available for the trawl fleet from There is no account taken of any technological creep in the fleet. Effort data for the creel fleet are not available. B.5. Other relevant data C. Assessment: data and method Model used: UWTV Based Approach to generate catch options In 2009 WKNEPH debated the use of the surveys as either an absolute measure of abundance or a relative index (ICES, 2009). Ultimately this led to a consensus that bias corrected survey abundance estimates could be used directly in the formulation of catch advice. Two modelling approaches were used to estimate sustainable stock specific Harvest Ratio reference points; SCA (a separable LCA model Bell) & Age Structured Simulation model (Dobby) (ICES, 2009). Software used: Age Structured Simulation model per recruit analysis in R 1 ) Survey indices are worked up annually resulting in the TV index. 2 ) Apply the Absolute Conversion Factor (see Section B3). The combined effect of these biases is to be applied to the new survey index. 3 ) Generate mean weight in landings. Check the time-series of mean landing weights for evidence of a trend in the most recent period. If there is no firm evidence of a recent trend in mean weight use an average taken over an appropriate time scale. If, however, there is strong evidence of a recent trend then apply most recent value (don t attempt to extrapolate the trend further in the future). 4 ) The catch option table will include the harvest ratios associated with fishing at F0.1, F35%SpR and FMAX. These values are estimated by benchmark workshops but may be revised if there are indications of changes to fisheries or biological factors.

59 ICES WGCSE REPORT ) Create catch option table on the basis of a range of harvest ratios ranging from 0 to the maximum observed ratio or the ratio equating to FMAX, whichever is the larger. Insert the harvest ratios from step 4 and also the current harvest ratio. 6 ) Multiply the survey index by the harvest ratios to give the number of total removals. 7 ) Create a landings number by applying the discard ratio (dead discard rate). 8 ) Produce landings biomass by applying mean weight. E. Medium-term projections None presented. F. Long-term projections None presented. G. Biological reference points Under the ICES MSY framework, exploitation rates which are likely to generate high long-term yield (and low probability of overfishing) have been evaluated and proposed for each Nephrops functional unit. Owing to the way Nephrops are assessed, it is not possible to estimate FMSY directly and hence proxies for FMSY have been determined. Three stock-specific candidates for FMSY (F0.1, F35%SPR, and FMAX) were derived from a length-based per recruit analysis (these may be modified following further data exploration and analysis). There may be strong differences in relative exploitation rates between the sexes in many stocks. To account for this, values for each of the candidates have been determined individually for males, females, and the two sexes combined. The combined sex FMSY proxy should be considered appropriate, provided that the resulting percentage of virgin spawner-per-recruit for males or females does not fall below 20%. If this happens a more conservative sex-specific FMSY proxy should be picked instead of the combined proxy. In the South Minch the absolute density observed on the UWTV survey is medium (~0.42 burrow/m 2 ). Historical harvest ratios in this FU have been above that equivalent to fishing at FMAX and landings have been relatively stable in the last thirty years. F35%SpR (combined between sexes) is expected to deliver high long-term yield with a low probability of recruitment overfishing and therefore is chosen as a proxy for FMSY. These calculations assume that the TV survey has a knife-edge selectivity at 17 mm and that the supplied length frequencies represented the population in equilibrium. The MSY Btrigger proposed for South Minch was based on the lowest observed UWTV abundance time-series. The FMSY proxy harvest rate values were calculated at the WKNEPH (ICES 2009) from the per-recruit analysis based on input parameters from a combined sex length cohort analysis of catch-at-length data. All FMSY proxy harvest rate and MSY Btrigger values remain preliminary and may be modified following further data exploration and analysis.

60 1430 ICES WGCSE REPORT 2013 HARVEST RATIO REFERENCE POINTS: Male Female Combined Fmax F F35%SpR TYPE VALUE TECHNICAL BASIS MSY MSY Btrigger 1016 million individuals Approach FMSY 12.3% harvest rate Bias-adjusted lowest observed UWTV survey estimate of abundance Equivalent to F35%SpR combined sex. Fmsy proxy based on length based Y/R. H. Other issues None. I. References Bailey, N., and Chapman, C. J A comparison of density, length composition and growth of two Nephrops populations on the west coast of Scotland. ICES CM 1983/K:42. Chapman C.J Discarding and tailing Nephrops at sea. Scott Fish Bull 46: Charuau A., Morizur Y., Rivoalen J.J Survival of discarded Nephrops norvegicus in the Bay of Biscay and in the Celtic Sea, ICES-CM-1982/B:13. Harris, R. R. and Ulmestrand, M Discarding Norway lobster (Nephrops norvegicus L.) through low salinity layers - mortality and damage seen in simulation experiments. ICES Journal of Marine Science 61, Howard F.G. and Hall, W.B Some observations on the biometrics of Nephrops norvegicus (L.) in Scottish waters. ICES, Doc.ShellfishComm.,CM1983/K:36. ICES Report of the Benchmark Workshop on Nephrops assessment (WKNEPH). ICES CM: 2009/ACOM:33. Morizur, Y Estimation de la mortalité pour quelques stocks de la langoustine, Nephrops norvegicus (L.). ICES, Doc. Shellfish Comm., CM 1982/K:10 (mimeo). Sangster, G.I., Breen, M., Bova, D.J., Kynoch, R., O Neill, F.G., Lowry. N., Moth-Poulsen, T. Hansen, U.J., Ulmestrand, M., Valentinsson, D., Hallback, H., Soldal, A.V., and Hoddevik, B Nephrops survival after escape and discard from commercial fishing gear. Presented at ICES FTFB Working Group, Hamburg, Germany April, 1997, ICES CM 1997 CM/B. Wileman, D. A., Sangster, G. I., Breen, M., Ulmestrand, M., Soldal, A. V. and Harris, R. R Roundfish and Nephrops survival after escape from commercial fishing gear. Final report to European Commission, Brussels, FAIR-CT

61 ICES WGCSE REPORT Table B1-1. Nephrops, South Minch (FU12), Nominal Landings of Nephrops, , as officially reported. YEAR UK SCOTLAND OTHER UK IRELAND TOTAL Nephrops trawl Other trawl Creel Sub-total * * provisional na = not available

62 1432 ICES WGCSE REPORT 2013 Figure B1-1. Nephrops, South Minch (FU12). Long-term landings, effort, lpue and mean sizes. The interpretation of the lpue series is likely to be affected by the introduction of the buyers and sellers regulations in 2006.

63 ICES WGCSE REPORT Figure B1-2. Nephrops, South Minch (FU12). Landings by quarter and sex from Scottish trawlers.

64 1434 ICES WGCSE REPORT 2013 Latitude Longitude Figure B1-3. Sediment strata in the South Minch. Light Grey Muddy sand, Grey Sandy mud, Dark Grey Mud. Light dashed lines represent spatial strata imposed on the sampling regime to ensure adequate spatial coverage.

65 ICES WGCSE REPORT Stock Annex 3.7: Nephrops FU13, Clyde Stock Clyde Nephrops (FU 13) Date 09 March 2009 Updated Revised by May 2013 (WGCSE) Sarah Clarke/Carlos Mesquita A. General A.1. Stock definition Throughout its distribution, Nephrops is limited to muddy habitat, and requires sediment with a silt & clay content of between % to excavate its burrows, and this means that the distribution of suitable sediment defines the species distribution. Adult Nephrops only undertake very small scale movements (a few 100 m) but larval transfer may occur between separate mud patches in some areas. In the Clyde area the Nephrops stock inhabits an area of muddy sediment extending throughout the Firth of Clyde, and another smaller area in the Sound of Jura, as shown in Figure B1-3. The two areas are separated by a large area of sandy gravely sediment around the Mull of Kintyre, and are treated as separate populations since they have differing population characteristics. A.2. Fishery Firth of Clyde The Firth of Clyde Nephrops fishery is predominantly exploited by a dedicated Nephrops trawler fleet of approximately 120 vessels, with less than 4% of the landings made by creel vessels. The 90 resident Clyde trawlers make about 90% of the Nephrops landings. Under the Scottish 'Inshore Fishing Order' of 1989 (Prohibition of Fishing and Fishing Methods), fishing with mobile gear is prohibited within the Firth of Clyde over weekends, and with vessels >70 feet (about 21 m) in length. The trawler fleet that fishes the Firth of Clyde mostly consists of vessels between 10 and 20 m in length (mean overall length 14 m), with a mean engine power of 185 kw. Almost half the fleet was built during the 1960s, with less than 20% built after Most vessels use single otter trawls with an 80 mm mesh codend. The regular fleet is comprised of Scottish vessels, but some catches are taken by Northern Ireland and Republic of Ireland vessels. The major landing ports are Troon, Campbeltown, Girvan and Tarbert, but smaller landings are also made at Carradale, Largs and Rothsay. The minimum landing size for Nephrops in the Clyde is 20 mm CL. Compliance with the minimum landing size is good, with samples suggesting only a very small undersized component in the landings (<2%). Nephrops growth varies within the area, with low density animals growing to large sizes in the North, and with higher density animals reaching smaller sizes in the South. Far more Nephrops material (undersized individuals and 'heads' from tailed animals) is discarded in the South. Discarding usually takes place at sea and landings are made by category for whole animals (small, medium and large) and as tails. In poor weather or for the last haul of the day, discarding may take place within the harbour, thus increasing discard mortality.

66 1436 ICES WGCSE REPORT 2013 Only a small fish bycatch is made in the Firth of Clyde, with whiting and cod being the most important species. The composition of the bycatch and discards varies within the Firth of Clyde, with more flatfish (common and long rough dab), echinoderms and crustaceans (other than Nephrops) caught in the north, while more roundfish (particularly whiting) are caught in the south. These differences reflect the different habitats and fish communities in the area. The fishery is exploited throughout the year, with highest landings usually made between July and September. Vessels usually have a trip duration of one day, sailing to shoot before dawn, and carrying out 3 4 hauls of four hours per day. Sound of Jura The fishery for Nephrops in the Sound of Jura constitutes part of the Clyde FU, but is examined separately from the fishery within the Firth of Clyde, because of differences in the biological parameters of the Nephrops populations. The fleet exploiting the Sound of Jura is also different to the Firth of Clyde, with vessels tending to be slightly smaller but more powerful. Most landings are taken by Scottish vessels (which are virtually all local to the area), with a very small proportion taken by boats from the rest of the UK. The local trawler fleet consists of vessels between 9 and 16 m in length, and with a mean engine power of 185 kw. Vessels employing twin-rig gear are generally larger and more powerful than those using single rig trawls (15 m and 220 kw compared to 13 m and 160 kw). The main landing ports are Port Askaig, West Loch Tarbert and Crinan. The minimum landing size for Nephrops in the Sound of Jura is 20 mm CL. Nephrops are found in high densities in this stock, but only grow to relatively small sizes. Discarding takes place at sea (this can be a high proportion of the catch by number, because of the small mean size of the animals caught), and landings are made by category for whole animals (small, medium and large) and as tails. Catches of fish in the Sound of Jura area are generally poor, and Nephrops are clearly the target species, with only small bycatches of whitefish and flatfish. The fishery is exploited throughout the year, with highest landings usually made between April and June. Vessels usually have a trip duration of one day, with 3 4 hauls per day. For both areas the current legislation governing Nephrops trawl fisheries on the West coast of Scotland was laid down by the North Sea and west of Scotland cod recovery plan (EC 2056/2001), which established measures additional to EC 850/98. This regulation was amended in 2003 by Annex XVII of EC 2341/2002, which establishes fishing effort and additional conditions for monitoring, inspection and surveillance for the recovery of certain cod stocks. For 2012, this regulation effectively limits vessels targeting Nephrops with mm mesh size to 200 days at sea per year. The use of square mesh and headline panels are compulsory in this fishery. Additional UK legislation has also been applied in the southern areas of the Firth of Clyde in recent years, aimed at protecting the aggregating cod in the south of the Clyde during February, March and April. A.3. Ecosystem aspects No information on the ecosystem aspects of this stock has been collated by the working group.

67 ICES WGCSE REPORT B. Data B.1. Commercial catch Length and sex compositions of Nephrops landed from the Firth of Clyde are estimated from port sampling in Scotland. Length data from Scottish sampling are applied to all catches and raised to total international landings. Rates of discarding by length class are estimated for Scottish fleets by on-board sampling, and extrapolated to all other fleets. The proportion of discarded to landed Nephrops changes with year, often determined by strong year classes. Discard sampling started in 1990, and for years prior to this estimates have been made based on later data. Landings and discards at length are combined (assuming a discard survival rate of 25%) to removals. Due to differences in catchability between sexes removals are raised separately for each sex. Landings show a sharp increase from 2006 (Figure B1-1 and Table B1-1). However this may be an artefact due to improved reporting of landings data due to the introduction of the buyers and sellers regulations in the UK in In addition, logsheet recording of hours fished is known to be erratic as it is a non-mandatory field on the logsheet It is therefore not clear whether the observed lpue inter-annual trends are actually indicative of real trends in lpue. Males contribute more to the landings than females (Figure B1-2). Mean length data for Nephrops above and below 35 mm CL, are shown in Figure B1-3. This size was chosen for all the Scottish stocks examined as the general size limit above which the effects of discarding practices and the addition of recruits were likely to be small. B.2. Biological Mean weights-at-age for this stock are estimated from fixed Scottish weight length relationships (unpublished data). The size at maturity was estimated by Queirós et al., (2013). Relevant biological parameters are as follows: natural mortality was assumed to be 0.3 for males (Morizur, 1982) of all ages and in all years. Natural mortality was assumed to be 0.3 for immature females, and 0.2 for mature females. Summary PARAMETER VALUE SOURCE Discard Survival (trawl) 25% Charuau et al., 1982; Sangster et al., 1997; Wileman et al., 1999 Discard Survival (creel) 100 % Wileman et al., 1999; Harris and Ulmestrand (2004); Chapman,1981 MALES Growth K 0.16 Adapted from Bailey and Chapman (1983) Growth - L(inf) 73 mm Adapted from Bailey and Chapman (1983) Natural mortality - M 0.3 Morizur, 1982 Length/weight - a Howard and Hall (1983) Length/weight - b 3.24 Howard and Hall (1983) Size at maturity 25 mm Adapted from Bailey and Chapman (1983) FEMALES Immature Growth Growth K 0.16 Adapted from Bailey and Chapman (1983) Growth - L(inf) 73 mm Adapted from Bailey and Chapman (1983) Natural mortality - M 0.3 As for males

68 1438 ICES WGCSE REPORT 2013 PARAMETER VALUE SOURCE Size at maturity 25 mm Queirós et al., (2013) Mature Growth Growth K 0.06 Adapted from Bailey and Chapman (1983) Growth - L(inf) 60 mm Adapted from Bailey and Chapman (1983) Natural mortality - M 0.2 Length/weight - a Howard and Hall (1983) Length/weight - b 2.91 Howard and Hall (1983) Proportion of F and M prior to spawning was specified as zero to give estimates of spawning stock biomass at January 1. B.3. Surveys The burrowing nature of Nephrops, and variable emergence rates mean that trawl catch rates may bear little resemblance to population abundance. An underwater TV survey has been developed, estimating Nephrops population abundance from burrow density raised to stock area. A random stratified sampling design is used on the basis of British Geological Survey sediment strata and latitude (Tuck et al., 1999) (see Figure B1-3). The survey provides a total abundance estimate, and is not age or length structured. A series of annual underwater TV surveys are available since 1995 for the Firth of Clyde and Sound of Jura. Whilst the survey in the Clyde has been continuous, the TV survey for the Sound of Jura was not conducted from 1997 to 2000, 2004 and Such gaps in the series make interpretation of any trends from the data difficult. The number of valid stations in the survey has remained relatively stable throughout the time period. An average of 36 stations have been sampled in each year, and then raised to a stock area of 2081 km 2 for the Firth of Clyde, and an average of ten stations have been considered valid each year for the Sound of Jura (area 383 km 2 ). Confidence intervals around the abundance estimates have remained relatively stable through the time period. UWTV relative to absolute conversion factors A number of factors are suspected to influence the ability of the surveys to map directly to absolute abundance. In order to use the survey abundance estimate as an absolute it is necessary to correct for these potential biases. The history of bias estimates is given in the following table and are based on simulation models, preliminary experimentation and expert opinion (ICES, 2009). The biases associated with the estimates of Nephrops abundance in the Clyde are: TIME PERIOD EDGE EFFECT DETECTION RATE SPECIES IDENTIFICATION OCCUPANCY CUMULATIVE ABSOLUTE CONVERSION FACTOR FU 13: Clyde <= B.4. Commercial cpue Landings-per-unit-of-effort time-series are available from the following fleets: Scottish trawl gears: Landings and effort data for Scottish trawl gears are used to generate a non-standardized lpue index. Lpue is estimated using officially recorded

69 ICES WGCSE REPORT effort (days absent from port). Effort data are available for the trawl fleet from There is no account taken of any technological creep in the fleet. Effort data for the creel fleet are not available. B.5. Other relevant data None. C. Assessment: data and method Model used: UWTV Based Approach to generate catch options In 2009 WKNEPH debated the use of the surveys as either an absolute measure of abundance or a relative index (ICES, 2009). Ultimately this led to a consensus that bias corrected survey abundance estimates could be used directly in the formulation of catch advice. Two modelling approaches were used to estimate sustainable stock specific Harvest Ratio reference points; SCA (a separable LCA model Bell) & Age Structured Simulation model (Dobby) (ICES, 2009). Software used: Age Structured Simulation model per recruit analysis in R 1 ) Survey indices are worked up annually resulting in the TV index. 2 ) Apply the Absolute Conversion Factor (see section B3). The combined effect of these biases is to be applied to the new survey index. 3 ) Generate mean weight in landings. Check the time-series of mean landing weights for evidence of a trend in the most recent period. If there is no firm evidence of a recent trend in mean weight use an average taken over an appropriate time scale. If, however, there is strong evidence of a recent trend then apply most recent value (don t attempt to extrapolate the trend further in the future). 4 ) The catch option table will include the harvest ratios associated with fishing at F0.1, F35%SpR and FMAX. These values are estimated by benchmark but may be revised if there are indications of changes to fisheries or biological factors. 5 ) Create catch option table on the basis of a range of harvest ratios ranging from 0 to the maximum observed ratio or the ratio equating to FMAX, whichever is the larger. Insert the harvest ratios from step 4 and also the current harvest ratio. 6 ) Multiply the survey index by the harvest ratios to give the number of total removals. 7 ) Create a landings number by applying the discard ratio (dead discard rate). 8 ) Produce landings biomass by applying mean weight. E. Medium-term projections None presented. F. Long-term projections None presented.

70 1440 ICES WGCSE REPORT 2013 G. Biological reference points Under the ICES MSY framework, exploitation rates which are likely to generate high long-term yield (and low probability of overfishing) have been evaluated and proposed for each Nephrops functional unit. Owing to the way Nephrops are assessed, it is not possible to estimate FMSY directly and hence proxies for FMSY have been determined. Three stock-specific candidates for FMSY (F0.1, F35%SPR, and Fmax) were derived from a length-based per recruit analysis (these may be modified following further data exploration and analysis). There may be strong differences in relative exploitation rates between the sexes in many stocks. To account for this, values for each of the candidates have been determined individually for males, females, and the two sexes combined. The combined sex FMSY proxy should be considered appropriate, provided that the resulting percentage of virgin spawner-per-recruit for males or females does not fall below 20%. If this happens a more conservative sex-specific FMSY proxy should be picked instead of the combined proxy. For the Firth of Clyde subarea of FU 13, the absolute density observed on the UWTV survey is generally high (average of over 0.8 m -2 for entire series and around 1.0 m -2 for the last five years suggesting the stock has relatively high productivity. In addition, the fishery in this area has been in existence since the 1960s and the population and biological parameters have been studied numerous times (Bailey and Chapman, 1983; Tuck et al., 1997; Tuck et al., 1999). Historical harvest ratios in this FU have been generally high at or above FMAX.. An appropriate FMSY proxy is considered therefore to be the total population FMAX which is predicted to deliver an F35%SpR of about 22% for males; considered precautionary for this species. Yield per recruit analysis is not yet available for the Sound of Jura subarea of this FU and so proxies from the Firth of Clyde (shown in the table above) are used. The absolute density observed on the UWTV survey is generally high (average of about 0.9 m -2 over the time-series and around 1 m -2 over the last five years) suggesting the stock has relatively high productivity. A number of studies have investigated biology and the area is acknowledged as having high abundance for many years. However, the time-series of TV data is more fragmented and sampling is at a relatively low level; confidence intervals are larger. The fishery in this area has been in existence since the 1960s but in recent times has operated at a low level and harvest ratios in this FU have been low. An appropriate FMSY proxy is considered therefore to be the total population F35%SpR which is predicted to deliver an F35%SpR of about 25% for males; above the level considered precautionary for this species (See Section 2.2). The FMSY proxy harvest rate values were calculated at the WKNEPH (ICES 2009) from the per-recruit analysis based on input parameters from a combined sex length cohort analysis of catch-at-length data. All FMSY proxy harvest rate and MSY Btrigger values remain preliminary and may be modified following further data exploration and analysis. Firth of Clyde HARVEST RATIO REFERENCE POINTS: Male Female Combined FMAX F F35%SpR

71 ICES WGCSE REPORT TYPE VALUE TECHNICAL BASIS MSY MSY Btrigger 579 million individuals Approach FMSY 16.4% harvest rate Bias-adjusted lowest observed UWTV survey estimate Equivalent to FMAX combined sex. FMSY proxy based on length based Y/R. Sound of Jura HARVEST RATIO REFERENCE POINTS: Male Female Combined FMAX F F35%SpR TYPE VALUE TECHNICAL BASIS MSY MSY Btrigger Not defined Approach FMSY 14.5% harvest rate Equivalent to F35%SpR combined sex. FMSY proxy based on length based Y/R. H. Other issues None. I. References Bailey, N., and Chapman, C. J A comparison of density, length composition and growth of two Nephrops populations on the west coast of Scotland. ICES CM 1983/K:42. Chapman C.J Discarding and tailing Nephrops at sea. Scott Fish Bull 46: Charuau A., Morizur Y., Rivoalen J.J Survival of discarded Nephrops norvegicus in the Bay of Biscay and in the Celtic Sea, ICES-CM-1982/B:13. Harris, R. R. and Ulmestrand, M Discarding Norway lobster (Nephrops norvegicus L.) through low salinity layers - mortality and damage seen in simulation experiments. ICES Journal of Marine Science 61, Howard F.G. and Hall, W.B Some observations on the biometrics of Nephrops norvegicus (L.) in Scottish waters. ICES, Doc.ShellfishComm.,CM1983/K:36. ICES Report of the Benchmark Workshop on Nephrops assessment (WKNEPH). ICES CM: 2009/ACOM:33. Morizur, Y Estimation de la mortalité pour quelques stocks de la langoustine, Nephrops norvegicus (L.). ICES, Doc. Shellfish Comm., CM 1982/K:10 (mimeo). Sangster, G.I., Breen, M., Bova, D.J., Kynoch, R., O Neill, F.G., Lowry. N., Moth-Poulsen, T. Hansen, U.J., Ulmestrand, M., Valentinsson, D., Hallback, H., Soldal, A.V., and Hoddevik, B Nephrops survival after escape and discard from commercial fishing gear. Presented at ICES FTFB Working Group, Hamburg, Germany April, 1997, ICES CM 1997 CM/B. Tuck, I.D., Chapman C.J. and Atkinson, R.J.A Population biology of the Norway lobster, Nephrops norvegicus (L.) in the Firth of Clyde, Scotland. I: Growth and density. ICES J. Mar.Sci 54,

72 1442 ICES WGCSE REPORT 2013 Tuck, I.D., Bailey, N., Atkinson, R.J.A. and Marrs, S.J Changes in Nephrops density in the Clyde Sea area from UWTV survey data. ICES, Doc. Living Resources Comm., CM 1999/G:13 (mimeo). Wileman, D. A., Sangster, G. I., Breen, M., Ulmestrand, M., Soldal, A. V. and Harris, R. R., Roundfish and Nephrops survival after escape from commercial fishing gear. Final report to European Commission, Brussels, FAIR-CT

73 ICES WGCSE REPORT Table B1-1. Nephrops, Clyde (FU13), Nominal Landings of Nephrops, , as officially reported. YEAR UK SCOTLAND OTHER Nephrops trawl Other trawl Creel Subtotal UK TOTAL** * * Provisional. ** Total also includes Rep. of Ireland.

74 1444 ICES WGCSE REPORT 2013 Figure B1-1. Nephrops, Clyde (FU13), Firth of Clyde subarea. Long-term landings, effort, lpue and mean sizes. The interpretation of the lpue series is likely to be affected by the introduction of the buyers and sellers regulations in 2006.

75 ICES WGCSE REPORT Figure B1-2. Nephrops, Clyde (FU13), Firth of Clyde subarea. Landings by quarter and sex from Scottish trawlers.

76 1446 ICES WGCSE REPORT 2013 Latitude Longitude Figure B1-3. Distribution of suitable sediments in Clyde. Light grey - muddy sand; medium grey - sandy mud; dark grey - mud.

77 ICES WGCSE REPORT Stock Annex 4.2: Cod in Division VIb Rockall Plateau Stock Working Group cod-rock WGCSE Date May 2013 Revised by WGCSE / Steven Holmes A. General A.1. Stock definition Cod at the Rockall plateau. A.2. Fishery No targeted fishery. Cod in Division VIb are taken in otter-trawl fisheries targeting haddock or anglerfish. A.3. Ecosystem aspects In recent years there has been a rise in seawater temperature on the Rockall bank combined with a significant reduction of Calanus finmarchicus which is a significant food item for larval and juvenile cod. In order to protect cold-water corals, four areas (northwest Rockall, Logachev Mounds, west Rockall Mounds, and Empress of British Banks) have been closed since B. Data B.1. Commercial catch Official landings. No discards information. There are doubts on the accuracy of the reported landings as these are reported by vessels operating in both Divisions VIa and VIb. B.2. Biological None. B.3. Surveys None. B.4. Commercial cpue Irish otter-trawl fleet (lpue). Scottish otter-trawl fleet (lpue). B.5. Other relevant data None.

78 1448 ICES WGCSE REPORT 2013 C. Assessment: data and method Model used: None. Software used: Model Options chosen: Input data types and characteristics: TYPE NAME YEAR RANGE AGE RANGE VARIABLE FROM YEAR TO YEAR YES/NO Caton Catch in tonnes na na na Canum Catch-at-age in numbers na na na Weca Weight-at-age in the commercial catch na na na West Weight-at-age of the spawning stock at na na na spawning time. Mprop Proportion of natural mortality before na na na spawning Fprop Proportion of fishing mortality before na na na spawning Matprop Proportion mature-at-age na na na Natmor Natural mortality na na na Tuning data: TYPE NAME YEAR RANGE AGE RANGE Tuning fleet 1 na na na Tuning fleet 2 na na na Tuning fleet 3 na na na D. Short-term projection Model used: None. Software used: Initial stock size: Maturity: F and M before spawning: Weight-at-age in the stock: Weight-at-age in the catch: Exploitation pattern: Intermediate year assumptions: Stock recruitment model used: Procedures used for splitting projected catches: E. Medium-term projections Model used: None.

79 ICES WGCSE REPORT Software used: Initial stock size: Natural mortality: Maturity: F and M before spawning: Weight-at-age in the stock: Weight-at-age in the catch: Exploitation pattern: Intermediate year assumptions: Stock recruitment model used: Uncertainty models used: 1 ) Initial stock size: 2 ) Natural mortality: 3 ) Maturity: 4 ) F and M before spawning: 5 ) Weight-at-age in the stock: 6 ) Weight-at-age in the catch: 7 ) Exploitation pattern: 8 ) Intermediate year assumptions: 9 ) Stock recruitment model used: F. Long-term projections Model used: None. Software used: Maturity: F and M before spawning: Weight-at-age in the stock: Weight-at-age in the catch: Exploitation pattern: Procedures used for splitting projected catches: G. Biological reference points No reference points are defined for this stock. TYPE VALUE TECHNICAL BASIS MSY MSY Btrigger xxx t Explain Approach FMSY Xxx Explain Blim xxx t Explain Precautionary BPA xxx t Explain Approach Flim Xxx Explain FPA Xxx Explain

80 1450 ICES WGCSE REPORT 2013 H. Other issues H.1. Historical overview of previous assessment methods NA. I. References

81 ICES WGCSE REPORT Stock Annex 4.3: Haddock in Division VIb Stock Working Group Haddock in Division VIb WGCSE Date 20 May 2012 Revised by Vladimir Khlivnoy A. General A.1. Stock definition The haddock stock at Rockall is an entirely separate stock from that on the continental shelf of the British Isles (Chuksin and Gerber, 1976; Shestov, 1977; Blacker, 1982; Newton et al., 2008). The TAC for haddock VIb was previously (before 2004) set for Subarea Vb, VI, XII and XIV combined, with a limitation on the amount to be taken in Vb and VIa. In 2004, the TAC for Division VI was split and the VIb TAC for haddock was included with Divisions XII and XIV. This combined TAC has been in place since then. A.2. Fishery The development of the Rockall haddock fishery is documented in the 2001 Working Group Report (ICES-WGNSDS, 2001) and in the Report of the ICES Group meeting on Rockall haddock convened in January 2001 (ICES, WGNSDS, 2002). That meeting was set up to respond to a NEAFC request for information on the Rockall haddock fishery. NEAFC agreed to consider regulation of the international fishery in The Rockall haddock fishery changed markedly in 1999 when a revision of the EU EEZ placed the southwestern part of the Rockall plateau in international waters. This has opened opportunities for other nations, notably Russia, to exploit the fishery in this area. The table of official statistics includes Russian catches from the Rockall area. The Russian fleet started fishing operations in international waters at Rockall in May October The Russian haddock fishery uses bottom trawls with cod-end mesh size of mm (mainly mm) and retains haddock of all length classes in the catch. This fishery targets concentrations of haddock mainly during the spring and the beginning of summer. Russian catches increased from 458 t in 1999 to 2154 t in In 2001, they were markedly reduced to 630 t due to the introduction of a closed area and low density of fish concentrations. Russian catches increased again in from 1630 to 5844 t. In , they decreased from 4708 t to 1282 t, and are estimated to be 1669 t in Prior to 1999, the UK and Ireland fisheries had been principally summer fisheries but in more recent years the Scottish and Irish fishery was conducted throughout the year with the peak in April May. This shift in the fishery appears to have followed the discovery of concentrations of haddock in deeper water to the west of Rockall, at depths between 200 and 400 m. High catch rates attracted effort into the area. However, catch rates in 2000 were reported to be poor in deeper water. Anecdotal evidence suggests that increased discarding has been associated with the deeper-water fishery compared to the traditional fishery at northern Rockall. In , a considerable proportion of EU landings were taken in the international waters. Historical fishing patterns of the Scottish fleet at Rockall are presented by Newton et al. (2004).

82 1452 ICES WGCSE REPORT 2013 There are some indications that, due to a general decline in catches by the Scottish and Irish fleets in Division VIa, there is an increasing focus in the Rockall fishery in Division VIb (ICES, WGFTFB, 2007). Paired gear (both seine and trawl) are to be tested by some Scottish fishermen, which, if it proves successful, can lead to a considerable increase in effective effort in VIb. The fishery at Rockall seems particularly attractive given the lack of effort restrictions in this area. Information on the Russian fishery and biological investigations from commercial vessels fishing in Rockall during 2008 are presented in WD11 to WGCSE An analysis of the spatial and depth distributions of Rockall haddock in association with oceanographic variables is presented by Vinnichenko and Sentyabov (2004), a WD to WGNSDS Changes in distribution have occurred over a period coincidental with changes in oceanographic variables. Information on oceanographic conditions on Rockall bank in spring 2005 was presented by Sentyabov at WGNSDS A.3. Ecosystem aspects In May 2001, the International Waters component of statistical rectangle 42D5, which is mainly at depths less than 200 m, was closed by NEAFC to all fishing activities, except with longlines. That area had the following coordinates: LATITUDE N N N N N LONGITUDE W W W W W In spring 2002, the EU component of this rectangle, again mostly shallow water, was also closed to trawling activities (EC No 2287/2003). The whole Rockall Haddock Box is bounded by the following coordinates: LATITUDE N N N N LONGITUDE W W W W At the 25th Annual Meeting of NEAFC (in November 2006), a closure of three areas on the Rockall Bank to bottom fishery was proposed to protect cold-water corals: North West Rockall, Logachev Mounds and West Rockall Mounds (NEAFC AM, 2006). This measure will be in force for the period January 2007 December In 2007, the ICES prepared advice for NEAFC and arrived at the conclusion about the expediency of establishing a new closed area on the so-called Empress of British Banks and adjusting the boundaries of the currently closed area of Northwest Rockall. At the 26th Annual Meeting of NEAFC (in November 2007), a new closed area (Empress of British Banks) was established, and the boundaries of the Northwest Rockall closure were slightly modified (NEAFC AM, 2007). Due to the complex shape of the boundaries of the Northwest Rockall closure proposed by ICES, which potentially could cause problems with enforcement, the introduced changes differed from the ICES recommendation. NEAFC also requested ICES to continue providing all

83 ICES WGCSE REPORT available new information on distribution of vulnerable habitats in the NEAFC Convention Area and fisheries activities in and in the vicinity of such habitats. WGDEC supported the ICES conclusion on the necessity of revising the boundaries of the Northwest Rockall area established to protect cold-water corals and recommended to consider proposals at the WGNSDS meeting. These recent proposals greatly simplify the boundaries, which would create better conditions for enforcement (see WD8 to WGNSDS, 2008). B. Data B.1. Commercial catch Landings Nominal landings as reported to ICES are given in Table of the main Report, along with working group estimates of total estimated landings. Reported international landings of Rockall haddock in were about t, except for , when they decreased down to about t. In 2006, they were also low at 2760 t, but increased slightly to 3348 in 2007, and 4221 t in Revisions to official catch statistics for previous years are also shown in Table Anecdotal evidence suggests that misreporting of haddock from Rockall have occurred historically (which may have led to discrepancies in assessment), but an estimation of overall magnitude is not possible. Age composition and mean weight-by-age of Scottish and Irish landings were obtained from port sampling. Data on the volume, length-age and weight composition of landings for the period from 1988 to 1998 correspond to values used at this WG. In 2002, there was no sampling of the Russian catch and therefore the length composition has to be estimated for this year. In 2002 and 2003, the structure of the Russian fishery on the Rockall Bank was the same: the same vessels were operating with the same gear in the same fishing areas. The relationship between the haddock length composition obtained from the trawl survey and that in the Russian catches is assumed to be the same for 2002 and 2003; i.e. it is assumed that the length dependent selectivity pattern in 2002 is the same as that in 2003 as there no changes to the fishery in these years. The relationship is described as: P = S p (1) L L L where PL is the proportion of fish with length L in catches, pl is proportion of fish with length L in the stock (survey), and SL is the proportion of fish of length L taken aboard. SL is determined using a theoretical selectivity curve (stock annex, Figure 4.3.2) which may be described by the following formula: S L 1 = 1+ exp( S S ). 1 2L where SL is the proportion of fish of size L taken aboard, L is the size group, S1 and S2 are coefficients. The selectivity curve (stock annex, Figure 4.3.2), fitted to the data on catch measurements in different periods of the Russian fishery in 2003 is described well by Equation 2 with coefficients S1 = and S2 = The estimated length frequency distri- (2)

84 1454 ICES WGCSE REPORT 2013 butions for 2003 are compared with the measured length frequency distributions for this year in stock annex, Figure The size distribution in the Russian catch in 2002 is then estimated by applying the theoretical selectivity curve to the survey length frequency in To determine the age composition in Russian catches in 2002, the combined age length key for all years of Russian catches was used. Discards The haddock catch estimated by landings is underestimated as a result of unaccounted discarding of small individuals in the Scottish and Irish fisheries in most years. On Russian vessels, the whole catch of haddock is retained onboard and therefore, total catch is equivalent to landings. Haddock discards onboard Scottish vessels in 1999 and 2001 and Irish vessels in 1995, 1997, 1998, 2000 and 2001 were determined directly. In other years, indirect estimates of discarding were calculated. The direct estimates from the Scottish trawlers in 1985, 1999 and 2001 showed a higher proportion of discards of small haddock: from 12 to 75% by weight (Table 4.3.1) and up to 80 90% of catch numbers. Discard trips in 1995, 1997, 1998, 2000 and 2001 showed that discarding by Irish fishing vessels also reaches considerable values (Table 4.3.2). Discard trips in 1995, 1997, 1998, 2000 and 2001 showed that discarding by Irish fishing vessels is variable with a mean rate of 30% (Table 4.3.2). Discard data were also obtained by Irish scientists from discard trips in and They showed that 52, 87 and 63% of the catch in numbers, respectively, was discarded. The range of discarded sizes was cm (mean 30 cm). In 2011 the discards are significantly reduced as a result of the small number of young haddock in the population. (Table of main report). It should be noted that these estimates are based on very few trips (one, two and three for 2007, 2008 and 2009 respectively) and should therefore be treated with caution. Total numbers and weight landed and discarded by age on the Scottish observer trips in 1999 and 2001 are presented in stock annex, Tables and The analysis of the discard data collected by Scottish scientists in 1999 and 2001 indicated that only a relatively small proportion of fish taken aboard is landed (stock annex Figure 4.2.3). The probability of being retained increases with increasing fish length (Stratoudakis et al., 1999; Palsson et al., 2002; Palsson, 2003; Sokolov, 2003). The relationship between the number of individuals caught and number discarded may be described by the following relationship: ND = PD NP (3) L L L where NDL is the number of discarded fish with length L, NPL is the number of fish caught at length L, PDL is the portion of discarded fish at length L. The length composition of fish taken onboard by Scottish and Irish trawlers was calculated by applying the logistic selectivity curve (stock annex, Figure 4.3.4) to the haddock stock length composition obtained from the survey. The selectivity parameters were calculated from Scottish and Irish catches taken by trawls with mesh size that are typical for the fleets of those countries operating at Rockall. The parameters were calculated as S1 = and S2 = for the Scottish fleet. S1 = and S2 = were used for Irish catches.

85 ICES WGCSE REPORT The catch-at-length compositions obtained by the theoretical curve of selectivity agree well with available results of catch measurements in 1999 and 2001and the distributions are compared in stock annex, Figure The proportion of fish discarded from catches at different sizes may be determined and modelled using a logistic curve (stock annex, Figure 4.3.6) described by the following equation: PD L 1 = 1+ exp( b( L DL 50 )) (4) where L is size group, DL50 is the fish length at which 50% of this size fish caught are discarded and b is a constant reflecting the angle of curve slope. The parameters were determined from research on discards by Scottish vessels (stock annex, Table 4.3.5). The following values were used in subsequent calculations: DL50 = cm, b = The logistic curve of discards may be found using Equation 2 and the coefficient values: S1 = and S2 = To determine abundance of discards the following procedure was used: a ) A theoretical catch-at-length distribution (%) was calculated by applying the theoretical selectivity curve to the survey length composition. b ) An estimate of total catch-at-length was made by summing the reported landings-by-length to the number of discards-at-length calculated from the assumed discard ogive and the landings-at-length data. c ) An intermediate theoretical catch size distribution in numbers is calculated by dividing the estimate of the total numbers retained (numbers greater than 34 cm) in B by the fraction retained from the theoretical catch length distribution calculated in a). d ) Theoretical discard size frequency is then calculated by applying the theoretical discard ogive to the intermediate theoretical catch size distribution. The spreadsheet containing these calculations can be found in the stock file. Calculations where the discard curve was applied agree well with the results of size composition measurements by Scottish vessels in 1999 and 2001 (stock annex, Figure 4.3.7). Aboard Irish vessels, larger fish are retained (stock annex, Figure 4.3.8). The portion of discards was calculated using Equation 2 with coefficients S1 = and S2 = , from the combined Irish discard trips. The Russian fleet fish in the areas covered only partially by the bottom-trawl surveys. However, Russian vessels retain all haddock and therefore there is no need to calculate discards. There is no information on large-scale fisheries of other countries outside the surveyed area. In addition, available data on the real length composition of catches indicate a correspondence between length composition obtained by the results from surveys and commercial catches, including the catches obtained in the parts of Russian fishery (stock annex, Figures and 4.3.6). The amount of discarded haddock by age was determined using a length age key derived by the data collected during the trawl survey allowing for selectivity of the fishery (stock annex, Figure 4.3.3).

86 1456 ICES WGCSE REPORT 2013 In 1998 and 2000, the trawl survey for haddock in the Rockall Bank area was not carried out. To determine the haddock length composition in these years, the length distribution was calculated from the survey data in the previous and following years. For this purpose, the length age matrices characterizing the stock status in the years before and after the missing data year were obtained. The length age distribution from the year before the missing year was projected forward on the basis of mean growth increment at age and estimated total mortality. Similarly the distribution from the year after was projected backwards. The length composition in the missing year was then calculated from these two estimates. The total loss (Z) used in the calculation described above was determined by minimization of values of deviation square sum between survey age group abundance values in previous and following years by the data from surveys and calculated data. At that, the factor of age effect (Sa) was taken into account. The mean growth increment at-age was also estimated from the survey data. The method of calculation is explained further in WD8 to WGNSD 2004 and a spreadsheet showing the calculations is in the stock file. B.2. Biological Age composition and mean weight-at-age of Scottish and Irish landings were obtained from port sampling. Age composition and mean weight-at-age of Russian landings were obtained by observers onboard commercial fishing vessels. In 2002, there was no sampling of the Russian catch and therefore the length composition for that year had to be estimated (for estimation details, see stock annex). Observer data from commercial vessels are also available for Norwegian landings for In the absence of any direct estimates of natural mortality, M has been set at 0.2 for all ages and years. Natural mortality coefficient and portion of mature individuals by age used for estimation correspond to those adopted by working group before. Previous working groups have adopted a maturity ogive with knife-edge maturityat-age 3 in assessments of this stock (see the Table below). Age Proportion mature The data from new Russian histological examination of haddock gonad samples mass sexual maturation occur at-age of two years with length of 25 cm (WGNSDS WD6 2006). These data agree well with the results of recent Scottish research in compliance with which the majority of fish become mature at the age of 2 years (ICES 2003; Newton et al., 2004). Visual estimation of maturity stage of post-spawning haddock on the Rockall Bank in expeditions leads to considerable errors. For more precise estimation of length and age-at-maturity for haddock it is necessary to conduct investigations in pre-spawning and spawning periods as well as to collect gonads for further histological analysis (see WGNSDS WD for further details). Research on determining more precise values for natural mortality and maturity ogive parameters should be continued and new estimates could be used in future stock assessments.

87 ICES WGCSE REPORT In the absence of any direct estimates of natural mortality, M has been set at 0.2 for all ages and years. MSVPA estimates for the North Sea haddock stock give estimates of M of 2.05 at age 0, 1.65 at age 1, 0.40 at age 2, 0.25 at ages 2 and 4, and 0.20 at ages 5+ (ICES CM 2003/ACFM:02). Similarly, large values of M at the younger ages at Rockall would have implications for interpretation of fishing mortality patterns from surveybased methods such as SURBA which essentially estimate total mortality conditional upon assumptions regarding survey catchability-at-age. ACFM in 2001 encouraged the WG to investigate a more realistic maturity ogive for this stock. At the 2002 working group combined sex maturity ogives were presented to the WG for Russian sampling in and Scottish sampling in In 2003 new sex disaggregated maturity data were supplied to the working group for Russian sampling. The results of all these recent studies indicate that a high proportion of both females and males at-age 2 were mature. B.3. Surveys There is only one research survey index available for VPA assessment of this stock from the Scottish survey conducted annually in September (Figure 4.3.1, Table of main report). However, from 1997 onwards the Scottish survey was only conducted in alternate years. Due to concerns about the haddock stock at Rockall some extra time was allocated to carry out a partial survey in September Full surveys have been conducted since 2005 to improve the quality of assessment. The Scottish survey is currently conducted on about 40 (the target number for a survey) standard trawl stations. However, the survey area and number of stations varied in different years. The majority of stations are within the 200 m depth contour. In 2002 the survey was carried out in the central and northern parts of the bank. In 1999 the survey switched from using an Aberdeen 48 bottom trawl to a GOV trawl and from 60 min tows to 30 min tows. The indices have been adjusted for tow duration, but no calibration has been made for gear changes. A 20 mm mesh size is used on the survey. In spring 2005, the Russian trawl-acoustic survey (TAS) for haddock on the Rockall Bank was conducted for the first time (Oganin et al., 2005). However, no such survey has been carried out in subsequent years. In the 2005 survey, the trawl survey method estimated the total stock number at million individuals and its biomass at t (see the Table below). The acoustic survey yielded a haddock biomass estimate of t with the abundance of million (see the WGNSDS 2006 Report for more details of the trawl-acoustic survey). The estimates of haddock abundance and biomass from the two methods are quite similar. The results of the Russian trawlacoustic survey are summarised in the Table below: Survey type Trawl survey Acoustic survey Area component Area (sq. miles) Total stock Abundance (106) Biomass (103 t) Whole Spawning stock Abundance (106) Biomass (103 t) International waters EU zone Whole * 60.0* * Pelagic component estimated to make up 13.7%.

88 1458 ICES WGCSE REPORT 2013 The Irish Fisheries Board (BIM) and the Marine Institute recently conducted a collaborative series of surveys to assess the length structure of haddock at various locations on the Rockall Bank and tested the selectivity of a number of codend configurations, which are typically used by both the Irish and Russian fleets. B.4. Commercial cpue Commercial cpue series are available for Scottish trawlers, light trawlers, seiners, Irish otter trawlers and Russian trawlers fishing in VIb. The effort data for these five fleets are shown in Figure and Table of main report. Commercial cpue series for the different fleets are shown in Figure of main report. In , the Russian effort in bottom fishery (in hours and number of vessels/days) decreased due to economic reasons. The effort in 2008 increased slightly compared to Haddock catches varied accordingly with the changes in fishing effort. In , fishing efficiency in the Russian haddock fishery (mainly with trawlers of tonnage class 10) increased compared to previous years. In 2008, with trawlers of class 8 and 9 only, it was still high (on average, 12.2 t per fishing day for trawlers of class 9), but lower than the efficiency in 2007 (on average, 16.9 t per fishing day for a trawler of class 10). In the period of the targeted fishery (April May), the mean catch of haddock per hour trawling by a trawler of tonnage class 9 was 0.86 t (in 2007, it was 0.88 t for a trawler of class 10) (Figure of main report). The dynamics of catch per unit of effort for this type of vessels agrees well with year-toyear variations in total biomass of haddock (Figure of main report). The effort data from the Scottish fleets are known to be unreliable due to changes in the practices of effort recording and non-mandatory effort reporting (see the report of WGNSSK 2000, CM 2001/ACFM:07, for further details). It is unknown what proportion of Scottish and Irish effort was applied directly to the haddock fishery. The apparent effort increase may just be the result of more exact reporting of effort due to VMS, but another suggestion is that it arises from a days at sea measure. Working at Rockall keeps days at sea elsewhere intact (the years in question do correspond to the introduction of the days at sea legislation) and it is possible that vessels are either working extra days in VIb or they are simply reporting extra days from VIb. It is difficult to conclude which of these scenarios is more likely. The Irish otter-trawl effort-series indicated low values between 2002 and 2005 with the lowest value in In , the effort increased considerably. The WG decided that the commercial cpue data, which do not include discards and have not been corrected for changes in fishing power despite known changes in vessel size, engine power, fish-finding technology and net design, were unsuitable for catch-at-age tuning. B.5. Other relevant data C. Historical stock development Model used: The assessment is based on catch-at-age data and one survey index (Scottish Groundfish Survey) and conducted using the XSA method. Software used: XSA from Lowestoft suite of VPA programs

89 ICES WGCSE REPORT Model Options chosen: Settings for the final XSA assessment in the recent years are shown in the Table below. Assessment year Assessment model XSA XSA XSA XSA XSA Time-series weights none none none none none Model power power power power power Catchability dependent for ages < Regression type C C C C C Q plateau Shk se Shk age-yr 4 yrs 3 ages 4 yrs 3 ages 4 yrs 3 ages 4 yrs 3 ages 4 yrs 3 ages Min se Plus group Fbar Input data types and characteristics: Type Name Year range Age range Variable from year to year Yes/No Caton Catch in tonnes Yes Canum Catch-at-age in Yes numbers Weca Weight-at-age in Yes the commercial catch West Weight-at-age of Yes the spawning stock at spawning time. Mprop Proportion of natural mortality before spawning No, set to 0 for all ages in all years Fprop Matprop Proportion of fishing mortality before spawning Proportion matureat-age No, set to 0 for all ages in all years No, the same ogive for all years Natmor Natural mortality No, set to 0.2 for all ages in all years Tuning data: Type Name Year range Age range Tuning fleet 1 SCOGFS

90 1460 ICES WGCSE REPORT 2013 D. Short-term projection Model used: Age-structured Software used: MFDP prediction with management option table and yield-perrecruit routines. MLA used for probability profiles and sensitivity analysis. Initial stock size: Taken from XSA for age 1 and older. The recruitment-at-age 1 in 2009 is estimated using RCT3. For forecasting recruitment in 2010 and thereafter, a geometric mean was used for Natural mortality: Set to 0.2 for all ages in all years. Maturity: The same ogive as in the assessment is used for all years. F and M before spawning: Set to 0 for all ages in all years. Weight-at-age in the stock: Three-year means (mean weights in the stock are assumed to be the same as catch weights, see below). Weight-at-age in the catch: Three-year means. Exploitation pattern: Average of the three last years. Landings F are varied in the management option table. Intermediate year assumptions: Status quo F. Stock recruitment model used: XSA estimate of recruits at-age 1 for intermediate year. RCT3 model used for intermediate year +1 in 2009 and the long-term geometric mean recruitment-at-age 1 is used for forecasting recruitment in 2010 and thereafter. Procedures used for splitting projected catches: F vectors in each of the last three years of the assessment are multiplied by the proportion landed at age to give partial F for landings. The vectors of partial F are then averaged over the last three years to give the forecast values. E. Medium-term projections Model used: Age structured Software used: MLA used for Medium-term projections. Initial stock size: Taken from the XSA for age 1 and older. The recruitment-at-age 1 in 2009 is estimated using RCT3. For forecasting recruitment in 2010 and thereafter, a geometric mean was used for Natural mortality: Set to 0.2 for all ages in all years. Maturity: The same ogive as in the assessment is used for all years. F and M before spawning: Set to 0 for all ages in all years. Weight-at-age in the stock: Three-year means (mean weights in the stock are assumed to be the same as catch weights, see below). Weight-at-age in the catch: Three-year means. Exploitation pattern: Average of the three last years. Intermediate year assumptions: Stock recruitment model used: RCT3 model used for intermediate year +1 in Uncertainty models used:

91 ICES WGCSE REPORT ) Initial stock size: 2 ) Natural mortality: 3 ) Maturity: 4 ) F and M before spawning: 5 ) Weight-at-age in the stock: 6 ) Weight-at-age in the catch: 7 ) Exploitation pattern: 8 ) Intermediate year assumptions: 9 ) Stock recruitment model used: F. Yield and biomass-per-recruit/long-term projections Model used: Yield and biomass-per-recruit over a range of F values. Software used: MLA and st graf. Maturity: Fixed maturity ogive as used in the assessment. F and M before spawning: Set to 0 for all ages in all years. Weight-at-age in the stock: Three-year means (mean weights in the stock are assumed to be the same as catch weights, see below). Weight-at-age in the catch: Three-year means. G. Biological reference points Biological reference points for this stock are given below: Blim: 6000 t (lowest observed SSB) BPA: 9000 t (Bloss 1.4) FPA: 0.4 (by analogy with other haddock stocks). H. Other issues None. I. References Blacker R.W Rockall and its fishery. Laboratory Leaflet, Lowestoft. No pp. Chuksin, Yu.V. and E.M.Gerber Soviet fishery in the Rockall and Porqupine areas. Zaprybpromrazvedka. Kaliningrad, 8 pp. ICES. WGFTFB Report of the ICES-FAO Working Group on Fish Technology and Fish Behaviour (WGFTFB). ICES CM 2007/FTC:06. ICES. WGNSDS Report of the Working Group on the Assessment of Northern Shelf Demersal Stock. ICES CM 2001/ACFM:01. ICES. WGNSDS Report of the Working Group on the Assessment of Northern Shelf Demersal Stock. ICES CM 2002/ACFM:02. NEAFC AM Report of the 25th Annual Meeting of the North-East Atlantic Fisheries Commission. NEAFC AM Report of the 26th Annual Meeting of the North-East Atlantic Fisheries Commission.

92 1462 ICES WGCSE REPORT 2013 Newton, A. W., Peach, K. J., Coull, K. A., Gault, M. and Needle, C. L Rockall and the Scottish haddock fishery. Fisheries research, 94: Pálsson, Ó. K A length-based analysis of haddock discards in Icelandic fisheries. Fisheries Research, 59: Pálsson, Ó. K., Karlsson, G., Arason, A., Gíslason, G. S., Jóhannesson, G. and Aðalsteinsson, S Analysis of cod and haddock discards Reports of the Marine Research Institute No. 90, pp. 18. [Hafrannsóknastofnun Fjölrit Nr. 90, in Icelandic]. Shestov, V.P Rockall haddock. Fishery biological resources of the North Atlantic and adjacent seas of the Arctic Ocean. Moscow, p Sokolov, K. M Estimation of cod discards in the Barents Sea and adjacent waters in Comparison of results obtained using different methods. Working Document 9 to the Arctic Fisheries Working Group, Pasaia, 23 April 02 May Stratoudakis, Y., Fryer, R. J., Cook, R. M., Pierce, G. J Fish discarded from Scottish demersal vessels: estimators of total discards and annual estimates for targeted gadoids. IC- ES Journal of Marine Science, 56:

93 ICES WGCSE REPORT Table Details of Scottish discard trips in the Rockall area (Newton et al., 2003). TRIP NO. DATE GEAR NO. OF HOURS % (BY WEIGHT) % (BY WEIGHT) HAULS FISHED HADDOCK DISCARDED OF LANDED OF HADDOCK CATCH 1 May 85 Heavy Trawl 2 Jun 85 Heavy Trawl 3 Jun 99 Heavy Trawl 4 Apr 01 Heavy Trawl 5 Jun 01 Heavy Trawl 6 Aug 01 Heavy Trawl Table Landings and Discards haddock estimates at Rockall from discard observer trips conducted aboard Irish vessels between 1995 and 2001, and from an observer trip aboard the MFV (February March 2000). (ICES CM 2004/ACFM:33). FAT/ KBG/ FAT/ KBG/ FAT/ KBG/ FAT/ KBG/ FAT/ KBG/ FAT/ KBG/ FAT/ KBG/ FEB 2000 DISCARD RATE 00/4 01/12 95/1 95/2 97/7 97/8 98/4 Landing Discards % discarded %

94 1464 ICES WGCSE REPORT 2013 Table Scottish landings and raised discards of haddock in 1999 estimates at Rockall from discard observer trips conducted on Scottish vessels. AGE Total Landing, N (*1000) Landing, tonnes Discards, N (*1000) 1 Discards, tonnes 1 Discards, N (*1000) 2 Discards, tonnes raised estimates from discard observer trips at Rockall. 2 estimates obtained from a logistic discard curve for Table Scottish landings and raised discards of haddock in 2001 estimates at Rockall from discard observer trips conducted aboard Scottish commercial vessels. AGE Total Landing, N (*1000) Landing, tonnes Discards, N (*1000) 1 Discards, tonnes 1 Discards, N (*1000) 2 Discards, tonnes raised estimates from discard observer trips at Rockall. 2 estimates from a logistic discard curve for 2001.

95 ICES WGCSE REPORT Table Values of DL50 by Scottish discard trips in the Rockall area. YEAR DL50 B Theoretical: Selectivity Length, cm Figure Theoretical haddock selectivity curve used to estimate the proportion of haddock lifted onboard Russian trawlers a b 3 % % Length, cm Length, cm Figure Length distribution of haddock in 2003: 1 by Scottish groundfish survey, 2a by commercial Russian trawlers in June, 2b by commercial Russian trawlers in July, 3 theoretically-derived.

96 1466 ICES WGCSE REPORT 2013 N, speciment Landing+Discards Landing Discarded 85,4% ( ) N, speciment Landing+Discards Landing Discarded 74,8% Length, cm Length, cm Figure Length distribution and quantity of haddock lifted onboard and landings by Scottish trawlers in 1999 and 2001 (unpublished data, Newton, 2004) Selectivy Length, cm Figure Theoretical haddock selectivity curve used to estimate the proportion of haddock lifted onboard Scottish trawlers % 15 % Length, cm Length, cm Figure Length distribution of haddock in 1999 and 2001: 1 by Scottish groundfish survey, 2 by commercial Scottish trawlers, 3 theoretically-derived.

97 ICES WGCSE REPORT Discards ratio Length, cm Figure Selectivity curve used to estimate the proportion of discarded haddock in catches Scottish trawlers. N, speciment Length, cm 1 2 N, speciment Length, cm Figure Length distribution of discarded haddock in catches Scottish trawlers in 1999 and 2001: 1 research data; 2 theoretically-derived.

98 1468 ICES WGCSE REPORT Irish M=38.47 cm Scottish M=35.51 cm Irish M=39.71 cm Scottish M=36.80 cm Irish M=35.58 cm Scottish M=37.82 cm Irish M=39.29 cm Scottish M=35.42 cm Irish M=38.47 cm Scottish M=36.93 cm Figure Length distribution of haddock landings in VI b (Scottish and Irish data).

99 ICES WGCSE REPORT Stock Annex 4.4: Whiting in VIb Stock Working Group whi-rock WGCSE Date May 2013 Revised by WGCSE / Helen Dobby A. General A.1. Stock definition There is an absence of information on whiting stock structure in this region and whiting caught at Rockall may potentially be part of the adjacent VIa stock. A.2. Fishery No specific information is available for Whiting at Rockall are taken as a bycatch in fisheries for other species such as haddock and anglerfish. A.3. Ecosystem aspects B. Data B.1. Commercial catch Only official landings data are available for whiting in VIb. There are no discard data. Reported landings are currently negligible (9 tonnes in 2011). In the past official landings have shown very high inter-annual variation and it is not known whether these are a true reflection of removals. There are doubts on the accuracy of the reported landings as these are reported by vessels operating in both Divisions VIa and VIb. B.2. Biological None. B.3. Surveys Survey catch rates of whiting at Rockall are extremely low and are therefore unlikely to provide a reliable index of abundance. B.4. Commercial cpue None. B.5. Other relevant data None. C. Assessment: data and method No assessment was conducted.

100 1470 ICES WGCSE REPORT 2013 Advice for 2013 was provided (in 2012) according to the data-limited approach for stocks with negligible landings and no information on abundance or exploitation. This consists of a precautionary reduction in catches. This corresponds to a decrease in catch by 20% in relation to the last three years average landings (14 t), corresponding to catches of no more than 11 t. In cases where this approach has been used, the expectation (according to ICES data-limited approach) is for multiannual constant catch advice. D. Short-term projection Model used: None. E. Medium-term projections None. F. Long-term projections None. Procedures used for splitting projected catches: G. Biological reference points None. H. Other issues H.1. Historical overview of previous assessment methods No assessments have ever been conducted for whiting at Rockall. Advice was first provided for this stock in 2013 (?). I. References

101 ICES WGCSE REPORT Stock Annex 5.2: Northern Shelf Anglerfish Stock IIIa, Working Group Anglerfish (Lophius piscatorius and L. budegassa) in Division and Subareas IV and VI Working Group on the Celtic Seas Ecoregion Date 1 April 2013 Revised by WKROUND 2013 Liz Clarke. A. General A.1. Stock definition Evidence from Anon (2001) and Hislop (2001) indicate that anglerfish Divisions IVa, Division VIa and Rockall should be treated as a single stock. The stock might in fact extend into VII, V or IIa, although there is currently insufficient evidence to support an extension of the stock area. A.2. Fishery UK vessels account for more than 50% of the total reported anglerfish landings from the Northern Shelf area. The Danish and Norwegian fleets are the next most important exploiters of this stock in the North Sea while Irish and French vessels take a significant proportion of the landings to the West of Scotland. The fishery for anglerfish in Subarea VI occurs largely in Division VIa with the UK and France being the most important exploiters, followed by Ireland. Landings from Rockall (Division VIb) are generally less than 1000 t with the UK taking on average around 50% of the total. In the North Sea, the majority of landings are reported in Division IVa which reflects the northerly distribution of the species within the North Sea (Knijn et al., 1993). A general description of the anglerfish fisheries of the most important nations taking part in this fishery is given below: Scottish (UK) fishery The Scottish fishery for anglerfish in Division VIa comprises two main fleets targeting mixed roundfish. The Scottish Light Trawl Fleet takes around 60% of landings and the Scottish Heavy Trawl Fleet over 20%. Around 10% of landings are bycatch from the Nephrops trawlers. The development of a directed fishery for anglerfish has led to considerable changes in the way the Scottish fleet operates. Part of this is a change in the distribution of fishing effort; the development of a directed fishery having led to effort shifting away from traditional roundfish fisheries in inshore areas to more offshore areas and deeper waters. The expansion in area and depth range fished has been accompanied by the development of specific trawls and vessels to exploit the stock. There has been an almost linear increase in landings from Division VIa since the start of the directed fishery until 1996 which has been followed more recently by a very severe decline, indicating the previous increase was almost certainly due only to the expansion and increase in efficiency of the fishery. More recent declines in landings ( ) may have been due to restrictive TACs and the decline is not necessarily representative of the actual landings.

102 1472 ICES WGCSE REPORT 2013 The Scottish fleet operating in VIb consists mainly of large otter trawlers targeting haddock and anglerfish at Rockall. Their activity is dependent on weather and the availability of haddock quota in VIb. The Scottish fishery for anglerfish in the North Sea is located in two main areas: on the Shelf Edge to the north and west of Shetland and at the Fladen Ground. It expanded in a similar manner since the 1980s to that operating in Division VIa. The fishery to the north and west of Shetland operates as an extension to that in Division VIa and consists mainly of light trawlers targeting mixed round-fish. The highest reported landings in recent years come from the statistical rectangles around Shetland. The landings from the fishery at Fladen are lower but still significant (around 15% of the total) with anglerfish caught as a bycatch in the Nephrops fishery which consists of approximately 200 vessels. A small component of the landings (~10% in recent years) comes from the gillnet fishery which operates on the shelf edge in the far northwest of Division IVa. A large proportion of the landings in the gillnet fishery are taken by Spanish owned, UK registered vessels. Irish fishery The Irish fleet which takes around 15 20% of the total Division VIa landings is a light trawl fleet targeting anglerfish, hake, megrim and other gadoids on the Stanton Bank and on the slope northwest of Ireland. This fleet uses a mesh size of 80 mm or greater. Irish Division VIa landings come mainly from the Stanton bank with some landings from Donegal Bay and the slope northwest of Ireland. Since 1996 there has been an increase in the number of vessels using twin rigs in this fleet. There have also been changes to the fleet composition since 2000, with around ten vessels decommissioned and four new vessels joining the fleet. The activity of this fleet is not thought to have been significantly affected by the recent hake and cod recovery plans. The Irish fleet otter trawl in Division VIb take anglerfish as a bycatch in the haddock fishery on the Rockall Bank. The fleet targeting haddock uses 100 mm mesh and twin rig trawls. Occasionally Irish-Spanish flag vessels target anglerfish, witch and megrim with 80 mm mesh on the slope in VIb. Discarding practices of these vessels are not known although discarding of anglerfish from the fleet targeting haddock in Division VIb is not thought to be significant (Anon, 2001). The fleet composition changed in In 2006 and 2007, the effort of the Irish fleet operating at Rockall has increased with the increase in Rockall haddock TAC. Danish fishery According to logbook records, the majority of Danish anglerfish landings are taken in the northeastern North Sea, in the part constituting the Norwegian Deeps, situated in the Norwegian EEZ of the North Sea. Other important fishing areas for anglerfish are the Fladen Ground (also in IVa) and in the Skagerrak (IIIa). More than 80% of the Danish landings come from ICES Divisions IVa and IIIa. The remaining part is from the most northern part of Division IVb. The majority of the Danish vessels are taking anglerfish with demersal trawls with over 90% of these vessels in the size range m. Fishery definitions by gear type and mesh size as currently used by Danish Fisheries Directorate for the North Sea are given in the following text table:

103 ICES WGCSE REPORT Fishery/gear Dem. Trawl Nephrops trawl Shrimp trawl Industrial trawl Beam trawl Mesh size, mm >= 100 mm mm mm < =32 mm >=80 mm Note that in the North Sea demersal trawls account for more than 90% of total Danish landings. However, it is necessary to further specify that at present the majority of the Danish catches of anglerfish are taken by fisheries in the Norwegian zone of IVa applying demersal trawls with mesh size >=120 mm. In 2006, the fishery with demersal trawl in the Norwegian Deeps (in the Norwegian zone) accounted for around 75% of total Danish landings by all gears from the entire North Sea. In the Skagerrak (IIIa) the two main fisheries taking anglerfish are the (mixed) Nephrops fishery and the demersal trawl fishery. In both areas minor landings are taken in gillnets and as bycatch in fisheries for shrimp (Pandalus). Typically anglerfish constitutes less than 15% by weight of the landings from demersal trawlers fishing in the Norwegian Deeps. Norwegian fisheries A Norwegian directed gillnet fishery (360 mm mesh size), targeting large anglerfish, carried out by small vessels in coastal waters in the eastern part of the northern North Sea started in the early 1990s. These vessels are responsible for around 60 70% of the total Norwegian landings from this area and they comprise around 6% of the total landings from Division IVa since The remaining Norwegian landings in IVa are mostly bycatch in various trawl fisheries. A similar pattern of fishing is found in the Skagerrak (IIIa). The third quarter has in recent years been the most important season for the directed fishery, while the second quarter is apparently most important for other gears. Other fisheries French demersal trawlers also take a considerable proportion of the total landings from this area. The vessels catching anglerfish may be targeting saithe and other demersal species or fishing in deep water for roundnose grenadier, blue ling or orange roughy. Since the mid-1990s, a deepwater gillnet fishery targeting anglerfish has been conducting a fishery on the continental slopes to the west of the British Isles, north of Shetland, at Rockall and the Hatton Bank. These vessels, though mostly based in Spain are registered in the UK, Germany and other countries outside the EU such as Panama. Gear loss and discarding of damaged catch are thought to be substantial in this fishery. Until now these fisheries have not been well documented or understood and they seem to be largely unregulated, with little or no information on catch composition, discards and a high degree of suspected misreporting. A.3. Ecosystem aspects Sea temperature limits the distribution of anglerfish to the north of the Northern Shelf particularly at depths where cold-water currents of polar descent occur. Lophius piscatorius is predominant throughout the area, with Lophius budegassa occurring in

104 1474 ICES WGCSE REPORT 2013 greater density towards the southern part of the area as befits the more general distribution of these two species (Fariña et al., 2008). B. Data B.1. Commercial catch B.1.1. Fishery landings data Fishery landings data for this stock are compiled from official statistics supplied by individual countries for ICES Divisions IIIa, IVa, IVb, IVc, VIa and VIb. Historical UK landings based on official statistics have been adjusted to correct for area misreporting for reasons described below. The TACs for both the west of Scotland and North Sea areas were reduced substantially in 2003 and 2004, and at previous WGs it has been highlighted that these reductions would likely imply an increased incentive to misreport landings and increase discarding unless fishing effort was reduced accordingly (Section 6.4.6, ICES WGNSDS 2003). Anecdotal information from the fishery in 2003 to 2005 appeared to suggest that the TACs were particularly restrictive in these years. The official statistics for these years are, therefore, likely to be particularly unrepresentative of actual landings. The introduction of UK & Irish legislation requiring registration of all fish buyers and sellers may mean that the total reported landings from 2006 onwards are more representative of actual total landings in the UK & Ireland. The absence of a TAC for Subarea IV prior to 1999 means that before 1999, landings in excess of the TAC in other areas were likely to be misreported into the North Sea. In 1999, a precautionary TAC was introduced for North Sea anglerfish, but was set in accord with recent catch levels from the North Sea which included a substantial amount misreported from Subarea VI. The area misreporting practices thus became institutionalised and the statistical rectangles immediately east of the 4 W boundary (E6 squares) accounted for a disproportionate part of the combined VIa/North Sea catches of anglerfish. The Working Group historically (prior to 2005) applied the following method to correct for area misreporting: 1 ) Estimate a value for the true catch in each E6 square and then allocate the remainder of the catch into VIa squares in proportion to the reported catches in those squares. 2 ) Estimate the true catches in the E6 squares by replacing the reported values by the mean of the catches in the adjacent squares to the east and west. This mean is calculated iteratively to account for increases in catches in the VIa squares resulting from reallocation from the E6 squares. Such a reallocation of catches may still inadvertently include some landings taken legally in Division IVa on the shelf-edge to the west of Shetland, but these are likely to comprise fish within the distribution of the Division VIa stock component. From WGCSE 2010, this procedure was adjusted to reallocate data to the whole of Area VI: i.e. not just VIa but including Rockall (VIb). This was based on information received from Marine Scotland Compliance indicating that some vessels fishing for anglerfish at Rockall are reporting large catches in the E6 squares from the same voyage. The distribution of landings this new scheme produced was more in keeping with the distribution of the stock as indicated from the anglerfish surveys.

105 ICES WGCSE REPORT B.1.2. Split of landings data by species (L. piscatorius and budegassa) The landings data are not currently split by species. B Fishery discards estimates Sampling schemes are in place through DCF to estimate discards in European fleets, and with no MLS discard rates appear to be very low. Scottish discard rates estimated from Scottish observer scheme (see next section for details) have been low in recent years. The Irish fleet otter trawl in Division VIb take anglerfish as a bycatch in the haddock fishery on the Rockall Bank. The fleet targeting haddock uses 100 mm mesh and twin rig trawls. Occasionally Spanish vessels target anglerfish, witch and megrim with 80 mm mesh on the slope in VIb. Discarding practices of these vessels are not known. Discarding of anglerfish from the fleet targeting haddock in Division VIb is not thought to be significant (Anon, 2001). Since the mid-1990s, a deep-water gillnet fishery targeting anglerfish has operated on the continental slopes to the West of the British Isles, North of Shetland, at Rockall and the Hatton Bank. These vessels, though mostly based in Spain are registered in the UK, Germany and other countries outside the EU such as Panama. Gear loss and discarding of damaged catch are thought to be substantial in this fishery. Until now these fisheries have not been well documented or understood and they seem to be largely unregulated, with little or no information on catch composition, discards and a high degree of suspected misreporting. B.1.3. Fishery length compositions Scotland Scottish anglerfish catch compositions are estimated from data collected under the Scottish demersal market and observer sampling programmes. Data are currently recorded for both anglerfish species but reported for both species combined. Anglerfish discards have been very low in recent years. The demersal market sampling scheme is stratified by market, aiming to visit each of the four major Scottish markets, which account for around 80% of demersal species landings by weight, at least 36 weeks per year. Random selection sheets are used to pick which vessel s landings to sample, and the sampling team check there are no missing categories for the relevant species before sampling. Boxes are selected from each sale size category and all the fish from a selected box are measured on a cm measuring board. Otoliths are taken for eight major Scottish species, including anglerfish. Annual sample size by species for the market sampling scheme is around 200 fishing trips from area IV, and around 40 fishing trips from Area VI. The demersal observer programme comprises 68 trips a year, allocated in very rough proportion to landings across Areas IV and VI, and covering the three main vessel groups that land the majority of Scottish demersal species: demersal trawlers, small Nephrops trawlers which mainly work inshore, and large Nephrops trawlers which mainly work offshore. Each trip in each area and vessel group combination is selected by contacting consecutive vessels on a randomly ordered vessel list until a vessel is found that intends to fish in the correct area, will accept the observer on board, and the observer considers meets their logistical and safety criteria. Annual sample sizes are approximately 50 fishing trips for Area IV and 20 for Area VI.

106 1476 ICES WGCSE REPORT 2013 Estimates for landed numbers-at-age and length are obtained by raising market samples to the six vessel group and area combinations by quarter, by means of poststratification, including appropriate sampling probabilities, and using species landed weight as an auxiliary variable. Estimates for discarded numbers-at-age and length are obtained by raising discard samples to the six vessel group and area combinations by year, including appropriate sampling probabilities and using gadoid landed weight as an auxiliary variable. Mean weights-at-age and length are obtained using survey-based weight length relationships. A total international catch-at-length distribution for Division VIa was obtained by summing national raised catch-at-length distributions and then raising this distribution to the WG estimates of total international catch from this area. Landings officially reported to ICES were used for countries not supplying estimates directly to the WG. Since 2001, the Scottish market sampling length weight relationships (given below) have been used to raise the sampled catch-at-length distribution data working group estimates of total landings for Division VIa. Length frequency data availability for VIb has been limited to Scottish and Irish samples. Year Range Formula (L length in cm, W weight in g) Source W= L2.988 Coull et. al., onwards W=0.0232L2.828 Scottish Market Sampling As a first step in assembling assessment data for the North Sea component of the stock, length compositions from Scottish market sampling have been raised to working group estimates of total landings in the past. The working group estimate of total landings was assumed equal to the landings obtained by national scientists plus official landings as reported to ICES for those countries not providing landings data to the working group. Total international catch-at-length distribution data for the whole Northern Shelf (Division IIIa, Subarea IV and Subarea VI) have previously been obtained by summing the length distributions from the individual areas and assuming that this distribution is representative of the whole Northern Shelf. This was then raised to working group estimates of total landings for the Northern Shelf. B.2. Biological B2.1. Growth and ageing An international ageing exchange in 2011 (ICES 2012) found little agreement between age estimation from both otoliths and illicia and concluded that anglerfish ages could not determine accurately enough for the purposes of producing an international catch-at-age dataset for stock assessment purposes. WKFLAT (ICES 2012) concluded that for Lophius piscatorius the studies of growth of Landa et al. (2012) should be used as the basis for length-based assessments, and this growth rate is supported by data in Laurenson et al. (2005). However the available growth curves assume asymptotic growth whereas anglerfish data from the Sco-IV- VI-AMISS-Q2 survey described below show linear growth. WKFLAT (ICES 2012) also concluded that for anglerfish in Divisions IIa and IIIa, Subarea IV and Subarea VI, ageing based on otoliths exists and age-based assessments could be considered for this stock if the internal consistency of the age compo-

107 ICES WGCSE REPORT sition of the data were examined in more detail and sensitivity to growth assumptions considered. Further growth and (ageing) age validation studies taking sex into account are required. B2.2. Maturity Historically, the catch-at-age analysis of anglerfish in Division VIa used the same maturity ogive as that applied to anglerfish in Subareas VII and VIII by the Working Group on the Assessment of Southern Shelf Demersal Stocks. However, a number of more recent maturity studies based on the VIa stock indicate that maturity does not occur until much later than previously estimated. Afonso-Dias and Hislop, 1996 give a length maturity ogive for this stock, 50% maturity at approximately 74 cm in females, and 50 cm in males. However, this study was based on few samples. New information has become available from the EU-funded project (Anon, 2001) which indicates female 50% maturity at approximately 94 cm and males at 57 cm. The corresponding age-based ogives indicate 50% maturity at approximately age 9 in females and age 5 in males. This has also been supported by more recent studies by Laurenson et al., B2.3. Natural mortality Previous assessments of this stock used the natural mortality rate applied to anglerfish in Division VI adopted by an earlier Hake Assessment Working Group of 0.15 yr -1. B.3. Surveys B 3.1. Sco-IV-VI-AMISS-Q2 survey In previous length-based assessments of this stock, a recruitment index was used which had been obtained from the Scottish March West Coast survey. The index consisted of numbers of anglerfish less than 30 cm caught per hour. However, meetings of the WG around 2003/2004, it was concluded that the IBTS groundfish surveys are ineffective at catching anglerfish and do not provide a reliable indication of stock size. As a result of this conclusion, and the urgent requirement for fisheryindependent data, Marine Scotland Science began a new joint science/industry survey in 2005 (current survey name is Sco-IV-VI-AMISS-Q2). B Design of survey This is a targeted anglerfish survey with a scientific design using commercial gear, currently covering part of IVa and the whole of VIa and VIb. In 2005, 2008 and 2010, the survey covered VIa down to 56. In 2006, 2007 and 2009, Ireland also participated, extending the anglerfish survey to cover the remaining part of VIa (from to ). In 2011 and 2012, the Scottish survey covered the whole of VIa. The survey area is stratified into four main areas, East East of Orkney and Shetland, North North of Scotland, West West of Scotland and Rockall, as shown in Figure 5.2.1, which are stratified by depth (0 200, , ), and with the stratum being further stratified in all except the east stratum, based on industry perceptions and the results of the 2005 survey. Within a stratum, the expected densities in substrata are defined as very high, high, medium or low. The sampling effort within each substratum is allocated according to its expected density, and the sample locations are chosen at random from grids of points within strips of equal area. This is to ensure equal probability of selection and even coverage within a stratum. Ap-

108 1478 ICES WGCSE REPORT 2013 proximately tows are taken each year. Tow duration is one hour. Each stratum is surveyed by one vessel, with at least one stratum being surveyed by MSS RV Scotia and with the other 2 4 strata being surveyed by industry charters. Each vessel on the survey employs exactly the same gear, the specification of which was drawn up in partnership with industry. Every anglerfish caught is measured for length, sex, maturity, total weight and gutted weights, and otoliths and illicia taken. In , surveys took place towards the end of October and beginning of November. However, bad weather affected survey coverage to the extent that the survey was moved to April from 2008 onwards. The stratification and station locations used in 2012 are presented in Figure A more detailed description of the survey including information on design, sampling, gear and vessel can be found in Fernandes et al. (2007). However, estimation has been further developed since then and is described below. B Estimation of abundance and catchability components a ) The estimation of abundance and biomass from these surveys is described below. The estimates represent the take into account the following factors: herding of anglerfish by the trawl doors and sweeps; b ) escapes of fish under the trawl footrope (details given below); c ) anglerfish abundance and biomass in the southern part of Area VI not covered in 2005, 2008 and 2010; d ) visual counts of anglerfish in areas closed to trawling at Rockall; e ) variability due to: i ) sampling; ii ) missing ages; iii ) herding (based on experimental data); iv ) footrope escapes (based on experimental data). The estimates currently do not take account of the following: areas in the central and southern North Sea (eastern part of ICES Division IVa and all of IVb and IVc); areas inaccessible to the trawl in Division VIa. To estimate the total number of anglerfish, N, from the survey, S, carried out in year y, Nˆ y, a Horvitz-Thompson estimator is used. This requires the inclusion probability, π f, of each fish, f, captured in the survey (i.e. the probability of that fish being captured on the survey), which we also need to estimate: ˆ N y = f S 1 πˆ f It is assumed that the inclusion probability of a fish depends on its length l, the haul i, and the stratum s, of the survey the haul is in, as we shall see below, so that π =. Since the inclusion probability is the same for fish of the same length on f π ils the same haul, the estimation equation can be written as Nˆ y = n πˆ ils s S i s l ils

109 ICES WGCSE REPORT where nils is the number of fish of length l captured in haul i of stratum s. Millar and Fryer (1999) partition the probability of capture for a fish in the population into three parts, potentially all being a function of length l: the probability of being retained in the net given the fish has encountered the net, r l, the probability that the fish encounters the gear given that it is available to the gear, and the probability of the fish being available to the gear. It is assumed that all fish in a stratum are randomly distributed in that stratum so that they have equal probability of encountering the gear, so that the probability that a fish is in the path of haul i towed in the stratum s is given by v i As, where vi gives the area swept by the doors (and the net) on tow i and A gives the area of stratum s. Somerton et al. (1999) show that the probability that a fish contacts the fishing net given that it was in the area swept by the haul i is given by s v1 i + hv v + v 1i 2i 2i, where v 1 i is v the area swept by the net in haul i (the area swept by the wings), 2 i is the sweep area in trawl i i.e. the area swept by the doors minus that swept by the wings, and h is the herding coefficient, which gives the proportion in v 2 i herded into v 1 i. Thus the inclusion probability of a fish of length l is given by the following equation: π il where v1 i + hv = rˆ l v1 i + v2 2i i vi As rˆ l is the estimated probability that a fish which encounters the net is retained in the codend of the fishing net, i.e. does not escape under the footropes, and vi v1 i + v2i =, N y can be estimated by: Nˆ y = n ils nils 1 nils = = A s π s S i s l v + hv v + v s S s ( v + hv ) rˆ s S i s l ils rˆ l v 1i 1i + v 2i 2i 1i A s 2i i 1i 2i l l Similarly, the biomass B in year y is estimated by: y B y = nils 1 1 = A s wilsj s S i s v1i + hv2i l rˆ l j 1 where wilsj is the weight of fish j of length l on haul i of stratum s of survey S. a ) Estimation of herding of anglerfish by the trawl doors and sweeps The estimation of the herding coefficient ( h ˆ = ) is described in Reid et al. (2007a). An individual-based particle-tracking model at to simulate the capture process was constructed using behavioural observations of 54 anglerfish

110 1480 ICES WGCSE REPORT 2013 captured from TV footage from the wingends and along the sweeps. Detailed analysis of the recordings showed that the fish did not appear to herd and many of the encounters with the wires were passive. All fish in the path of the net were captured, whereas more than half of the fish between the wings and the doors were not. b ) Estimation of escapes of fish under the trawl footrope The proportion of fish at length that escaped below the gear was estimated from a series of experimental trawls using bags below the fishing line. These trials are described in Reid et al. (2007b). Selectivity functions were fitted to these data in a GLM framework assuming a Binomial distribution. A comparison of three selectivity functions was made: simple logistic, asymmetric logistic (estimating an asymptote parameter), however the simple logistic curve was found to be adequate for these data: exp( β 0 + β1l) r( l) = 1+ exp( β + β l) 0 1 This model was then applied to the length data from each survey to correct for those fish that were likely to escape under the net as described above. c ) anglerfish abundance and biomass in the southern part of Area VI not covered in 2005, 2008 and 2010; Estimates of the proportion of anglerfish in the southern part of ICES Division VIa were derived from 2006, 2007 and 2009 when Ireland contributed to the survey and covered this area completely. The proportions of abundance in this area relative to the whole Northern shelf were 8.6% in 2006, 13.6% in 2007 and 4% in 2009; the proportions of biomass were 5.5% in 2006, 7.4% in 2007 and 2.5% in The averages of these proportions (i.e. 8.7% for abundance and 5.1% for biomass) were used to raise the estimates of the surveys in 2005, 2008 and 2010 when Scotland did not survey this area and the Irish did not participate. d ) Estimation of anglerfish abundance in areas closed to trawling at Rockall Visual counts of anglerfish in areas closed to trawling at Rockall have been carried out in all years since 2007, as described in McIntyre et al. (in press). A deep-towed vehicle was developed, equipped with video, lights and other sensors, to enable visual surveying at depths of over 300 m, at speeds of up to 3 knots and altitudes of up to 10 m (from the seabed). This vehicle was used to survey large areas of the seabed around Rockall, in areas which are closed to trawling due to the presence of the deep-water coral Lophelia pertusa. Counts were made from visual inspection of the TV footage. The total area surveyed by video was calculated by summing the area surveyed every second over the entire transect, the latter being calculated from the trapezoid between the midpoints of one image frame and the next image frame a second later. The number of anglerfish N c, y in closed area c in year y was estimated by: ˆ N c y A ˆ, cρc, y where =, Ac is the surface area of the closed area c (km 2 )

111 ICES WGCSE REPORT and ˆρ c, y is the mean density of closed area c in year y, estimated from the survey in year y as follows: c, y = 1 n ˆρ (11) where: i, c, y t c, y i c Ai, y tc,y is the number of tows in closed area c in year y ni,c,y is number of anglerfish recorded in tow i in closed area c in year y Ai is the area of the seabed (km 2 ) surveyed by the visual tow i in year y. Biomass in these areas was calculated by multiplying the abundance by the average weight of anglerfish in the adjacent trawl strata of each survey. The abundance and biomass of anglerfish in the northwest Rockall closure was added to the estimates as additional strata The proportions of abundance and biomass in the northwest Rockall closed areas relative to that in the adjacent two strata were then used to estimate the abundance and biomass in the Empress of Britain bank closure. Abundances in the closed areas surveyed between have been estimated by this method as thousand fish, approximately 1% of anglerfish in numbers on the Northern Shelf. B Estimation of length and age compositions Each fish caught on the survey is measured and aged and so the number-at-length, Nly and the number-at-age N ay are estimated from the survey data in a similar way to the total numbers and biomass: Nˆ ly = A s rˆ ( v + hv ) s S i s l 1i n ils 2i and Nˆ ay where S. n 1 nilsa = As + s S i s v1 i hv2i l rˆ l ilsa is the number of fish of age a and length l on haul i of stratum s of survey The ages are estimated from otoliths read for all surveys by the same experienced reader. B.4. Commercial cpue B.4.1. Official logbook data Previous length-based assessments attempted to use effort data to constrain the temporal trend in fishing mortality. Scottish Light Trawl data, disaggregated into an inshore and offshore component, the latter of which is associated with the anglerfish fishery, for both West of Scotland and Shetland (N Sea) were provided to the working group. However, these data are no longer considered to be reliable due to non-

112 1482 ICES WGCSE REPORT 2013 mandatory recording of hours fished in the logbook data. Further details of the Scottish fleet effort recording problem can be found in the report of the 2000 WGNSSK (ICES, 2001). Since these data are considered unreliable, they are not presented here. Irish lpue data in terms of hours fished have been presented to the WG for Division VIa and Division VIb for all fleets up to The measure of kwdays is believed to be a more reliable proxy for effort than hours fished due to reporting issues and these data are now presented in the WG report. Danish landings and effort data (hours fished) from logbook data are also available to the WG for Division IIIa and Division IVa. Although these data are considered to be reliable (in terms of accuracy of reporting), it is not know to what extent they are useful in providing an indicator of stock size due to management regulations in the Norwegian zone (TAC constraints) and technological creep. B.4.2. Tallybook data Analysis of skippers personal diary information collected in 2004 and 2005 in an attempt to improve knowledge of the state of the stock and of the Scottish anglerfish fishery provided valuable information to ICES (Bailey et al., 2004) on temporal and spatial trends in catch rate. Following the success of this data collation exercise, ICES advised the process to continue and a more formal scheme was proposed by FRS. Extensive discussions with the fishing industry during 2005 resulted in FRS implementing the monkfish tallybook project at the start off The project is part of a long-term approach to providing better information on the monkfish fishery and the state of the stock, and is being operated in conjunction with fishers organisations (Scottish Fishermen s Federation, Fishermen s Association Limited and Pecheurs de Manche et Atlantique) and the North Atlantic Fisheries College (NAFC) Marine Centre, Shetland. These organisations have been responsible for distributing the tallybooks, coordinating the returns and allocating a vessel code before the anonymised tallybook sheets are forwarded to FRS. The tallybooks are filled in on a haul-by-haul basis to give weight caught by size category and information on haul location, duration and depth in a standardized format as well as gear and mesh being used. Additionally information on mature females has been requested. Data are stored in a database at FRS. The time-series is relatively short, with the first returns from fishing trips at the end of December 2005 and the most recent from March Initial participation in the scheme was high with returns received from up to 37 vessels with a wide spatial coverage (across Subarea VI, Division IVa, IIa and Vb) and different target species. Of the 37 vessels which supplied information, two were French and these operated towards the southern end of the shelf edge in Division VIa northwest of Ireland. The haul depth information collated indicates that most of the hauls were taken in depths between 100 and 400 m although there are a significant number of hauls from depths between 600 and 800 m. The records from the deeper water were largely from the French vessels although a number of the Scottish vessels make occasional trips into deeper water. Average catch rates are similar to those previously seen in the diary data and observer data (presented in previous WG reports) and range from around 10 kg/hr for boats targeting Nephrops to over 100 kg/hr for some whitefish boats. Analysis of the catch rate data is presented in the WG report and in Dobby et al., 2007.

113 ICES WGCSE REPORT B.5. Other relevant data C. Assessment: data and method Since 2012, this stock has been assessed in accordance with the approach suggested by ICES (2012) for data-limited stocks, category 3: survey only assessments, method for if there are survey data on abundance but there is no survey-based proxy for MSY Btrigger and F values or proxies are not known. 1 ) Determine the catch advice from the survey adjusted status quo catch: = C 1 2 y 1 SASQ C y + 1 y 1 y y 2 y 5 where Ii is the biomass index from the survey in year i. 2 ) Apply the 20% Uncertainty Cap to the catch advice: C 0.8C if C SASQ y 1 y+ 1 y 1 UC SASQ SASQ = C if 0.8C y C C y y y+ 1 y 1 SASQ 1.2C y 1 if C 1.2C y+ 1 y 1 I I i i 0.8C Apply this cap to the catch advice to address uncertainty or noise in the data and its potential influence on the catch advice. This cap should be applied to all quantitative advice for data-limited stocks, regardless of category; however if the advice is simply based on last year s advice, no uncertainty cap is applied. 3 ) Then apply the Precautionary Buffer to the catch advice: C PB y + 1 UC 0.8C unless exceptions apply y + 1 y 1 1 = I UC 2 y 2 C given exceptions e.g. y + 1 y 3 1 I 3 y 5 i i 1.5, or decreasing effort Apply this buffer when reference points are unknown; how if substantial increases in abundance indices are consistently observer or there are substantial reductions in fishing mortality or effort in the target fishery, this precautionary buffer may not apply to catch advice. When the precautionary buffer is applied, the catch advice should apply for at least three years unless new information or analyses indicate a new situation (e.g. there is a clearly marked increase in stock indicators.) D. Short-term projection Not applicable.

114 1484 ICES WGCSE REPORT 2013 E. Medium-term projections Not applicable. F. Long-term projections Not applicable. G. Biological reference points Not known. H. Other issues In previous ( catch-at-length ) assessments of this stock, the SSB was always estimated to be at a very low level. The length data have been based on the U.K. landings only (in Subdivisions. IVa and VIa), where very few individuals over 80 cm appear in the catch and therefore the model predicts very few in the population. Since females do not mature until they are over 90 cm in length the SSB is estimated to be very low. The length data from the eastern part of the North Sea (Danish and Norwegian fisheries) for the recent years indicate a higher amount of larger individuals in the catches. Although the Danish and Norwegian landings are small in comparison to the UK landings, the inclusion of the Danish and Norwegian length frequencies in the data used for any future assessment may change the concept of the magnitude of the SSB. The fact that mature female anglerfish are rarely observed either on scientific surveys or by observers on board commercial vessels supports a very low estimate of spawning stock biomass, yet there is little evidence of reduction in spatial distribution as fish are still recruiting to relatively inshore areas. It has been hypothesized that females may become pelagic when spawning as they produce a buoyant, gelatinous ribbon of eggs, and would therefore not appear in the catch of trawlers. (Anglerfish have been caught near the surface, Hislop et al., 2000). This would imply different exploitation patterns for males and females: a dome-shaped pattern (decreased exploitation at larger sizes) for females and a logistic pattern for males. It is also not known whether anglerfish are an iteroparous or semelparous species. The latter would also account for the almost complete absence of spawning females in commercial catches or research vessel surveys. The key features of the species life history in relation to its exploitation are the location of the main spawning areas, and whether or not there is any systematic migration of younger fish back into the deeper waters to spawn. At present, despite the large increase in catches during the mid-1990s, there is no apparent contraction in distribution; fish are still recruiting to relatively inshore areas such as the Moray Firth in the northern North Sea. The fact that spawning may occur largely in deep water off the edge of the continental shelf may offer the stock some degree of refuge. However, this assumes that the spawning component of the stock is resident in the deep water, and is thus not subject to exploitation. It is not known to what extent this is true, but if such a reservoir exists then the currently used assessment methods which make dynamic pool assumptions about the population are likely to be inappropriate. Nevertheless, it is clear that further expansion of the fishery into deeper water is likely to have a negative effect on the SSB and given the spatial development of the fishery, it cannot be ruled out that the serial depletion of fishing grounds has been occurring. In addition, some life-history characteristics of anglerfish suggest that it may be particularly vulnerable to high exploitation. A detailed discussion of the fishery develop-

115 ICES WGCSE REPORT ment and biology can be found in Sections and of the 2000 report of this working group (ICES, 2001). H.1. Historical overview of previous assessment methods A length-based model was used up to 2003 (Dobby, 2002) but was subsequently abandoned due to lack of confidence in the landings data. Since then WGs have presented trends from the Sco-IV-VI-AMISS-Q2 survey. I. References Afonso-Dias, I. P. and J. R. G. Hislop The population of anglerfish (Lophius piscatorius) from the northwest coast of Scotland. J. Fish. Biol. 49 (Suppl A): Anon The distribution and biology of anglerfish and megrim in waters to the west of Scotland. EC Study Contract 98/096 Final Report August Coull, K.A., A.S. Jermyn, A.W. Newton, G.I. Henderson and W.B. Hall Length/Weight relationships for 88 species of fish encountered in the North East Atlantic. Scottish Fisheries Research Report No. 43. Dobby, H A length-based assessment of anglerfish in Division VIa: developments in growth modelling. Working Document for the Working Group on the Assessment of Northern Shelf Demersal Stocks, Dobby, H., L. Allan, M. Harding, C. Laurenson and H. A. McLay Improving the quality of information on Scottish anglerfish fisheries making use of fishers data. ICES CM 2007/K:16. Fernandes, P. G., Armstrong, F., Burns, F., Copland, P., Davis, C., Graham, N., Harlay, X., O Cuaig, M., Penny, I., Pout, A. C. and Clarke, E. D Progress in estimating the absolute abundance of anglerfish on the European northern shelf from a trawl survey. ICES CM 2007/K:12. Hislop, J. R. G., A. Gallego, M. R. Heath, F. M. Kennedy, S. A. Reeves and P. J. Wright A synthesis of the early life history of anglerfish, Lophius Piscatorius (Linnaeus, 1756) in northern British waters. ICES Journal of Marine Science, 58, ICES Report of the Working Group on the Assessment of Demersal Stocks in the North Sea and Skagerrak, ICES CM 2001/ACFM:07. ICES ICES implementation of RGLIFE advice on Data Limited Stocks (DLS), ICES CM 2012/ACOM:68. ICES Report of the Anglerfish (Lophius piscatorius) illicia and otoliths exchange ICES Report of the Benchmark Workshop on Flat-fish Species and Anglerfish (WKFLAT), ICES CM 2012/ACOM:46. Knijn, R. J., T. W. Boon, H. J. L. Heesen and J. R. G. Hislop Atlas of North Sea Fishes ICES Cooperative Research Report No Landa, J., Duarte, R. and Quincoces, I Growth of white anglerfish (Lophius piscatorius) tagged in the Northeast Atlantic, and a review of age studies on anglerfish. ICES J. Mar. Sci. 65: Laurenson, C. H., A. Johnson and I. G. Priede Movements and growth of monkfish Lophius piscatorius tagged at the Shetland Islands, Northeastern Atlantic. Fisheries Research 71: McIntyre, F.D., Collie, N., Stewart, M., Scala, L., and Fernandes, P.G. In press. A visual survey technique for deep water fish: estimating anglerfish abundance in closed areas. Journal of Fish Biology.

116 1486 ICES WGCSE REPORT 2013 Millar, R.B and Fryer, R.J. Estimating the size-selection curves of towed gears, traps, nets and hooks Reviews in Fish Biology and Fisheries 9: , Reid, D.G., R.J. Kynoch, I. Penny and K. Peach Estimation of catch efficiency in a new angler fish survey trawl ICES CM2007/Q:22. Reid, D. G., Allen, V. J., Bova, D. J., Jones, E. G., Kynoch, R. J., Peach, K. J., Fernandes, P. G., and Turrell, W. R Anglerfish catchability for swept-area abundance estimates in a new survey trawl. ICES Journal of Marine Science, 64: Sullivan, P. J., H-L Lai and V. F. Gallucci A catch-at-length analysis that incorporates a stochastic model of growth. Can. J. Fish. Aquat. Sci. 47, Figure Map of the northern continental shelf around the British Isles showing the areas surveyed during the anglerfish surveys in 2011, shaded according to the survey strata as indicated in the legend. Sample positions (n=153) are indicated by the black crosses (FRV Scotia, n=104) and black circles (MFV Ocean Venture, n=49).

117 ICES WGCSE REPORT Stock Annex 5.3: Megrim in Divisions VIa and IVa Stock Working Group Megrim in Divisions VIa and IVa WGCSE Date May 2012 Revised by IBP-MEG/2012; Norman Graham A. General A.1. Stock definition Since the end of the 1970s ICES has assumed three different stocks for assessment and management purposes: megrim in ICES Subarea VI, megrim in Divisions VIIb k and VIIIa,b,d and megrim in Divisions VIIIc and IXa. Megrim stock structure is uncertain and historically the Working Group has considered megrim populations in VIa and VIb as separate stocks. The Review Group questioned the basis for this in Data collected during an EC study contract (98/096) on the Distribution and biology of anglerfish and megrim in the waters to the West of Scotland demonstrated significantly different growth parameters and significant population structure difference between megrim sampled in VIa and VIb (Anon, 2001). Spawning fish occur in both areas but whether these populations are reproductively isolated is not clear. As noted by WGNSDS 2008, megrim in IVa has historically not been considered by ICES and WGNSDS 2008 recommended that VIa megrim should be considered by WGCSE. Landings data from IV and IIa are now included in this Report and work is underway to collect international catch and weight-at-age data for IV as well as VI. However, the availability of aggregated and age-disaggregated is sporadic. Data from both the commercial fishery (using VMS and catches by statistical rectangle) and from fishery-independent surveys provide little evidence to support the view that megrim in VIa and IVa are indeed separate stocks. Based on the recommendations from WKFLAT (2011), megrim in VIa and IVa are considered a single unit stock and assessed accordingly. Megrim in VIb is considered a separate stock unit for assessment purposes. A.2. Fishery Megrim are predominately taken in otter trawl and to a lesser extent by Scottish seine. Analysis of VMS data indicates that megrim is taken in spatially discrete shelf fisheries and also in trawl fisheries conducted along the 200 m shelf break. Historically, ICES has assumed that megrim catches are closely linked to those of monkfish. Area misreporting of monkfish from VIa into IVa as a result of restrictive TACs in VIa is known to have occurred historically and catches have been redistributed into VIa using an algorithm developed by the Marine Science Scotland (see stock annex for Monkfish). Due to the assumed linkage between megrim and monkfish, megrim caught in VIa are also considered to have been area misreported and therefore the Working Group has historically applied the same redistribution method as used for monkfish. It remains unclear whether this pattern has continued in recent years, in 2009 the Working Group did not redistribute megrim catches in VIa as the historic pattern, higher catches in the statistical rectangles immediately east of the 4 line, was not observed in 2009, indeed the 2009 pattern may indicate a reversal of the process

118 1488 ICES WGCSE REPORT 2013 due to a more restrictive TAC in IVa. However, treating megrim in VIa and IVa as a single unit stock has mitigated this problem. The introduction of the Cod Long-Term Management Plan (EC Regulation 1342/2008) and additional emergency measures applicable to VIa in 2009 (EC Regulation 43/2009, annex III 6) has impacted on the amount of effort deployed and increased the gear selectivity pattern of the main otter-trawl fleets. Additionally, EC regulation 43/2009 has effectively prohibited the use of mesh sizes <120 mm for vessels targeting fish, which had been used particularly by the Irish fleet up to that point. Effort associated with the French fleet has continued to decline while the decline in both the Irish and Scottish TR1 fleets (120 mm mesh) appears to have stabilized. The increase in mesh size (from 100 to 120 mm) has also impacted on the retention length of megrim, increasing L50 from 28 cm to 42 cm, an increase of almost 50%. Fishing effort in IV for the main Scottish otter fleet (TR1) have stabilized since the large effort reductions observed in previous years, effort levels associated with this mesh band have fallen by 64% since Following the increases in Irish effort in Subdivision VIb from , effort in 2009 has declined significantly. These reductions in effort in Scotland and Ireland are considered to have contributed to the decline of landings in Subarea VI. Landings in VI are well below the TAC. Uptake by France, who account for 44% of the TAC, is very low (~11%). Official landings in Subarea IV and Division IIa in recent years are close to the TAC. There is anecdotal information from the Scottish industry that since the introduction of the Conservation Credits Scheme in Area IV, those vessels have responded with increasing focus on anglerfish and megrim in both IVa and VIa. Based on landings data presented to the Working Group, only 53% of the overall TAC for VI, EC waters of Vb and international waters of XII and XIV was used. The TAC in IV was fully utilized. Commercial catches are dominated by female megrim, typically 90% of the total catch. Analysis of Irish logbook data by Anon (2002) showed that cpue trends varied throughout the year, showing a maximum in late spring/early summer following the spawning period and at their lowest in late autumn. A.3. Ecosystem aspects None considered. B. Data B.1. Commercial catch Commercial landings by country are available since The UK accounts for ~80% of the total landings. Over 50% of the landings are taken in the North Sea (IVa) with the remainder taken in VIa (~40%) and VIb, there are also landing reported from other areas (IVb and IVc), but these are negligible. International landings-at-age data based on quarterly market sampling are available from 1990 for V.I. Note that up until 2000, catch-at-age data from VIa and VIb was aggregated, only partial landings-at-age are available for VIb (post 2000). Landings numbers-at-age are available for IVa (post 2005), depending on year and country. Ireland provides landings numbers-at-age by quarter, age disaggregated discard numbers-at-age by annum for both VIa and VIb. Scotland provides annual catch numbers-at-age by Divisions VI and VI and discards estimates by weight and number

119 ICES WGCSE REPORT with associated length distribution. Since 2011, France has provided landings and effort (hours fished) by statistical rectangle with quarterly length distributions of landings and discards with associated sampling effort (hours fished). The general paucity of both landings and discard data covering the assessment area has prevented the construction of a full-time and spatial series for megrim separately in VIa, VIb and IVa. The available data are not separated by sex. Females make up approximately 90% of the landings, but survey data show that the relative proportion of males increases with depth. The quality of the available landings data (unknown area misreporting), discard information, lack of effort data and cpue data for the main fleet in the fishery, and disaggregated landings-at-age data at an appropriate area level severely hampers the ability of ICES to carry out an assessment for this stock. Prior to 2000, discard data for VIa were combined together with data from VIb and no data fom IVa are available prior to The available data shows that discarding is variable and given the increases in mesh sizes introduced in 2000 (North Sea) and 2009 (West of Scotland) it is expected that discard rates have declined. Laurenson and MacDonlad (2008) note that while discarding of megrim below minimum landing size is low (<1%), discarding of legal sized fish was much higher at 22%. This is attributed to low market price for small grades and bruised fish, resulting in high grading of catches on length/quality reasons to maximise the value of a restrictive quota. B.2. Biological Megrim exhibit a strong negative growth relationship with increasing depth. Fish found in deep water (>200 m) are commonly the same size as fish one year younger found in shallower areas (Gerritsen et al., 2010). Analysis of age-at-length data shows a wide length distribution within ages and that age precision deteriorates when sampling levels fall below ~500 per annum. Poor age precision in recent years prevents the development of an age based assessment. C. Assessment: data and method The assessment method: Bayesian state space biomass dynamic model. C.1. Input data C.1.1. Catch International landings data collated by the ICES Working Group on the Celtic Seas Ecoregion (WGCSE) is used as an estimate of catch. However, it is recognised that discarding is a feature of the fishery but note that discard data are not available for the entire time-series and the availability or raised discard data is highly variable across fleets and areas therefore if catch data is to be used, then some assumptions regarding the historic discard pattern must be made. To assess the sensitivity of the model outputs to this assumption, two alternative model runs with (i) a fixed 20% discard proportion over the full landings time-series and (ii) a linear decline in proportion from 30% at the start of the time-series to 10% at the end. It is probable that the proportion of megrim discarded in IVa has declined since 2000 and in VIa since 2009 the mesh size in the North Sea increased from 100 to 110 mm and was further increased to 120 mm in 2001, while in Division VIa, the mesh size was increased from 100 to 120 mm in It is therefore likely that the discard-

120 1490 ICES WGCSE REPORT 2013 ing profiles have probably changed significantly in line with these mesh size increases and this option is used for the final run. For catch data from 2011 onwards, discard estimates provided to the working group are used. C.1.2. Survey indices Indices from six fishery-independent surveys are used (Table ) for the assessment, four associated with the International Bottom Trawl Survey (IBTS) and two associated with the relatively recent (2005) dedicated anglerfish survey. Survey trends in cpue are shown in Figure and tabulated in Table Table Survey indices used for surplus production model. NUMBER SURVEY NATIONALITY AREA TIME-SERIES 1 Sco-IBTS-Q3 Scotland IVa Sco-IBTS-Q1 Scotland IVa ScoGFS-WIBTS- Scotland VIa Q1 4 ScoGFS-WIBTS- Scotland VIa Q4 5 SAMISS-Q2 Scotland VIa*/IVa IAMISS-Q2 Ireland VIa* *VIa data from IAMISS-Q2 and SAMISS-Q2 combined into a single cpue estimate with variance.

121 ICES WGCSE REPORT Table Input parameters, individual survey cpue indices, landings and modelled discards for the final assessment run. Year Sco-WIBTS-Q1 Sco-WIBTS-Q4 Sco-IBTS-Q1 Sco-IBTS-Q3 SAMISS-Q2 SAMISS-Q2/ IAMISS-Q2 Landings (t) Discards (t) 1985 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA 432

122 1492 ICES WGCSE REPORT 2013 Figure Trends in survey cpue indices used in the assessment of megrim in VIa and IVa. The asterix shown in Sco-WIBTS Q1 and Sco-WIBTS Q4 relates to the survey cpue in 2011/2012 but is not used due to changes in survey gear and design. C IBTS survey indices IBTS survey data from Scottish groundfish survey data (Surveys 1 4, Table 2.2.1) are available for quarters 1 and 4 in ICES Area VIa and quarters 1 and 3 in ICES Area IVa north. The survey design is based on ICES statistical rectangles. One tow is selected per rectangle based on a library of clean tows. The tow location is largely the same every year and as such the design may be considered fixed station although minor changes to tow locations can occur. In 2010 both the groundgear and the survey design associated with the ScoGFS-WIBTS Q1 and Q4 surveys were changed. Rather than relying on fixed trawling locations moved to a new random-stratified survey

123 ICES WGCSE REPORT design with trawl locations randomly distributed within ten a priori sampling strata. While there were rationale reasons for these changes, it has resulted in a breach in the time-series and it will not be possible to use these indices until a reasonable timeseries ca. five years has been built up. Catch weights are not routinely collected on all IBTS surveys so the length data were converted to weight using the length weight relationship. where is the weight in grams and is the length in centimeters. This relationship was estimated using all available megrim length weight measurements from the dedicated monkfish survey. The weights were then raised by the numbers-at-length per tow and summed to provide a catch in kilograms per tow. This was divided by the duration of the tow in decimal hours to provide a cpue measured in units of kg.hr -1. The data from all four surveys exhibit a relatively large proportion of zeros; therefore the delta method of Stefánsson (1996) was used to extract indices. The uncertainty surrounding each survey index (observation error) can be estimated within the assessment model or estimated externally and entered into the assessment model as a fixed quantity. For the present analysis we used the mean delta-gamma cpue estimates (for the IBTS surveys only) and allowed the model to estimate the measurement error of each survey. [1] C Anglerfish survey indices Scottish (SAMISS) and Irish (IAMISS) dedicated anglerfish surveys (Surveys 5 6, Table 1.2.1) have been undertaken in VIa and IVa (SAMISS only) since The survey design is stratified based on expected densities of anglerfish (not megrim), within each strata, the location of individual tows are randomly selected. The modelling approach of Stefánsson (1996) is mainly applicable to a fixed station design and therefore for the anglerfish indices we used the weighted cpue estimates and allow the observation error to be estimated within the model. The anglerfish survey provides absolute estimates of abundance and biomass. The average fish density at age a in stratum s, ρ as, is estimated from the weighted mean of fish densities corrected for the catchability of each trawl, as follows: ˆ ρ as where: = n lai nlai wi = wi v Qˆ ê v + v ĥ) i s l a 1i li i s l a l ( 1i 2i n lai is the number of fish of age a and length l caught in trawl i, v1i + v2i wi = ( v1i + v2i ) i, v 1 i is the area swept by gear in trawl i (the area swept by the wing), i v 2 is the sweep area of gear in trawl i i.e. the area swept by the door minus that swept by the wing,

124 1494 ICES WGCSE REPORT 2013 Qˆ v 2i li = êl + êlĥ is the catchability estimate for a fish of length l in trawl i, following v1i the definition by Somerton et al. (2007), ê l is the estimated footrope selectivity-at-length l, is the proportion of fish of length l originally in the area swept by the wing which are caught by the net and do not escape under the footrope, ĥ is the estimated herding coefficient. ( ĥ =0.017). It should be noted that the methods outlined above were specifically designed for anglerfish. The most significant issue for megrim is that as there is no estimates of footrope selectivity, êl is assumed to be 1. While this is not an issue when the survey indices are treated in a relative sense as presented here for megrim, Fernandes (2010) does use this approach to provide a raised absolute biomass based but notes that due to the full retention assumption for groundgear selectivity, the estimates are considered as a minimum estimate. C.2. Method Surplus production methods (Schaefer, 1954; Pella Tomlisson, 1969) offers a potential modelling approach in the absence of reliable catch-at-age data. Surplus production pools the overall positive contributory effects (growth and recruitment) with removals due to mortality into a single production function, thus the stock is considered solely in terms of biomass without regard for differences in age, size of sex structure. Surplus production models are commonly used when only relative biomass indices, either from survey or from commercial fisheries, and landings data are available. For computational simplicity, earlier methods assumed that the yield from the fishery is in equilibrium, where each year s catch and effort data represent an equilibrium (steady-state) situation where the catch is assumed to equal the surplus production. This can result in overly optimistic estimates of MSY, particularly problematic when a stock is in decline. Process error methods also use catch and effort data, but do not make the assumption that the population is in equilibrium. Process error methods make the assumption that the measurement of catch and effort are measured without error. Conversely, observation error methods assume that the biomass response is correct and that all error is associated with measurement error. Polacheck et al. (1993) compared the performance of all three approaches and found that observation methods performed best, with the process method proving very imprecise. However, it would be preferable to consider both process error associated with the inherent population dynamics and observation error which describes the inherent variance in catch and effort observations. The development of state space models has the ability to separately model and incorporate both process and observation error (Meyer and Millar, 1999). Due to ageing issues with megrim in VIa and IVa associated with low sampling size and depth-dependent growth issues, a surplus production process model is used (Schaefer, 1954) to describe the current exploitation of megrim relative to FMSY and stock biomass relative to BMSY. The biomass dynamics are given by a difference form of a Schaefer biomass dynamic model:

125 ICES WGCSE REPORT where is the biomass at time, is the intrinsic rate of population growth, is the carrying capacity, and is the catch, assumed known exactly. To assist the estimation the biomass is scaled by the carrying capacity, denoting the scaled biomass. Log-normal error structure is assumed giving the scaled biomass dynamics (process) model: where the logarithm of process deviations are assumed normal ; is the process error variance. The starting year biomass is given by, where is the proportion of the carrying capacity in 1980.The biomass dynamics process is related to the observations on the indices through the measurement error equation: where is the value of abundance index in year, is index-specific catchability,, and the measurement errors are assumed log-normally distributed with ; is the index-specific measurement error variance. C.2.1. Estimation prior distributions Estimation is undertaken in a Bayesian framework with Markov Chain Monte Carlo (MCMC) sampling using WinBUGS (Spiegelhalter et al., 1999). Prior distributions are given in Table Note that prior distribution assumptions are important. In these preliminary runs we have assumed largely uninformative priors to see what information is present in the data to update these priors. Sensitivities to K, assuming uniform normal or log-normal, distributions have been tested and although the fitted and posterior parameters are quite similar. The major difference being in the parameter K, which has an extremely long tail when a uniform prior is assumed. Most of the density of K is similarly distributed (good overlap when the distributions are overlayed). As the uniform prior distribution on the logarithm of K avoided long tails (which may have a very large effect on the mean), this was chosen in subsequent runs (e.g. retrospective and final). Catchability sensitivity Assigning a prior distribution that is uniform on the logarithmic scale is recommended for catchability in biomass dynamics models (Punt and Hilborn, 1997). A corresponding fit allowing for catchability to range over resulted in a poorly converged model with unrealistic estimated absolute abundances (order of 500 thousand tonnes). The range of the catchability parameter was thus scaled to have a lower limit of -11 on the logarithmic scale, this corresponds to a lower limit on q of exp(-11)= 1.67e-05, which allowed for biomass to range over 100 thousand tonnes from each series.

126 1496 ICES WGCSE REPORT 2013 Table Lepidorhombus whiffiagonis in ICES Areas VIa and IVa. Prior distributions on parameters. Parameter Symbol Prior distribution Notes Intrinsic rate of population growth Carrying capacity Catchabilities Process error variance Measurement error variances Proportion of K in 1985 From the maximum catch to ten times the cumulative catch across all years assuming uniform distribution on the logarithmic scale Uniformly distributed on log-scale. See catchability sensitivity in Section Gamma distributed on inverse variance (precision) scale Gamma distributed on inverse variance (precision) scale D. Short-term projection Model used: Risk based forecast Software used: R The lack of recruitment data and age data precludes the provision of a short-term forecast based on spawning stock and recruitment relationships. Instead, using the historic dynamics of the stock, the likelihood of the stock exceeding FMSY under a range of catch options is presented. Advice is based on maintaining the risk of FMSY exceeding 5%. E. Medium-term projection Not presented. F. Long-term projection Not presented.

127 ICES WGCSE REPORT G. Biological reference points Type Value Technical basis MSY Btrigger % of BMSY Approach FMSY 0.31 Estimated from model Blim % of BMSY Precautionary Approach BPA Not defined Flim 1.7 FMSY. Fishing mortality that drives the stock to Blim FPA Not defined H. References Anon Report of the NAFO Study Group on Limit Reference Points Lorient, France, April, NAFO SCS Doc. 04/12, Serial No. N4980. Dahm and Wienbeck Escapement of fish underneath the groundrope of a standard bottom trawl used for stock assessment purposes in the North Sea. ICES CM 1992/B:20. Fernandes Indices of the abundance and biomass of megrim (Lepidorhombus spp.) on the European Northern Shelf: Working Document presented to WGCSE (2011). Hilborn, R. and Walters, C. J Quantitative Fisheries Stock Assessment: Choice, Dynamics,and Uncertainty. Chapman and Hall. ICES Report of the Workshop on Implementation in DATRAS of Confidence Limits Estimation May 2006, Copenhagen, Denmark. 16 pp. ICES Report of the Workshop on Anglerfish and Megrim (WKAGME). ACOM: February Aberdeen, UK. 112 pp. ICES Report of the Joint NAFO/ICES Pandalus Assessment Working Group (NIPAG), October ICES. CM 2010/ACOM:14. Schaefer, M.B Some aspects of the dynamics of populations important to the management of the commercial marine fisheries. Bulletin of the Inter- American Tropical Tuna Commission, 1: Pella, J.J., and Tomplinson, P.K., (1969). A generalized stock production model. Bulletin of the Inter-American Tropical Tuna Comission, 13: ICES Report of the Inter Benchmark Protocol for Megrim in Subarea IV and Division IVa (IBPMeg), 2 6 April By correspondence. ICES CM 2012/ACOM: pp. Polacheck, T, Hilborn, R. and Punt, A.E Fitting surplus production models: comparing methods and measuring uncertainty. Can. Jou. Fish.Aqu.Sci. 49: Prager A suite of extensions to a nonequilibrium surplus-production model Fish. Bul. 92: Punt, A.E., and Hilborn, R Fisheries stock assessment and decision analysis: the Bayesian approach. Reviews in Fish Biology and Fisheries 7: Reid, D.G., Allen, V.J., Bova, D.J., Jones, E.G., Kynoch, R.J., Peach, K.J., Fernandes, P.G. and Turrell, W.R Angler fish catchability for swept area abundance estimates in a new survey trawl. ICES J. Mar. Sci. 64. Spiegelhalter, D.J., Thomas, A., Best, N., Lunn, D WinBUGS User Manual version 1.4 January MRC Biostatistics Unit, Inst. of Public Health, Cambridge, England. Stefánsson, G Analysis of groundfish survey abundance data: combining the GLM and delta approaches. ICES Journal of Marine Science, 53,

128 1498 ICES WGCSE REPORT 2013 Stock Annex 5.3b: VIb (Rockall) Megrim (Lepidorhombus spp) in ICES Division To be updated.

129 ICES WGCSE REPORT Stock Annex 6.2: VIIa Cod Stock Working Group Irish Sea Cod (Division VIIa) Celtic Seas Ecoregion (WGCSE) Date May 2013 Revised by Alessandro Ligas A. General A.1. Stock definition All catches and survey data from within ICES Division VIIa are assumed to come from a unit stock. Stock structure of cod in European waters has been the subject of increasing study in recent years. Current information is summarized in Crozier et al., Recent egg surveys in 2006 and 2008, using DNA probes to distinguish early stage eggs of cod from other gadoids, confirm the location of distinct cod spawning grounds in the western and eastern Irish Sea (Goodsir et al., 2008). Historical tagging studies indicated spawning site fidelity but varying degrees of mixing of cod between the Irish Sea, Celtic Sea and west of Scotland/north of Ireland (see Lordan et al., 2011). Studies based on meristic characteristics, allele frequencies and microsatellite markers genetics and population structure have not provided unequivocal evidence of genetically isolated stocks in the Irish Sea and surrounding waters. A recent tagging programme run from , in which over 2200 cod were tagged using external and data storage tags revealed that although there was some movement of cod between the Irish and Celtic Seas, the component of Irish Sea cod in the Celtic Sea was low. Furthermore, no cod tagged in the Celtic Sea were recovered from the Irish Sea (Connolly and Officer, 2001). One problem with interpreting this evidence is that the overall stock sizes in both areas have declined significantly in recent years. There may therefore have been changes in geographic range and movement patterns making comparison of recent results with earlier studies problematic. More recent tagging of cod off Greencastle on the north coast of Ireland (Ó Cuaig and Officer, 2007), and limited tagging on UK Fisheries Science Partnership surveys, have demonstrated movements of cod between Division VIa and VIIa. Most recaptures in VIIa from cod tagged in VIa have come from the North Channel and in or near the deep basin in the western Irish Sea that is a southward extension of the North Channel. The research surveys used for tuning the VIIa cod assessment cover only the western and eastern Irish Sea, and do not extend into the deeper water of the North Channel, where large catches of cod were made by midwater trawlers in the 1980s and 1990s. Immature cod may disperse over a wide area as demonstrated by fish tagged and released from various parts of the Irish Sea (including Belfast Lough). These demonstrated a substantial migration into the Celtic Sea and round the north and west of Ireland. Once these fish mature however they appear to return to the Irish Sea spawning grounds. Extensive tagging off the West of Scotland produced no recaptures from the Irish Sea.

130 1500 ICES WGCSE REPORT 2013 Further tagging and genetics studies are required to investigate stock structure, seasonal movements and mixing in VIIA and neighbouring areas. New tagging studies of cod in the Irish Sea, Celtic Sea and to the north of Ireland have taken place in A.2. Fishery Irish Sea fisheries for cod have changed considerably over the last four decades: A brief description is given below. 1960s and 1970s. UK and Irish single otter trawlers targeted spawning cod in spring in both the western and eastern Irish Sea. Fisheries for young cod (codling) took place in autumn and winter. The growing single rig Nephrops fleet took bycatches of cod. Several strong year classes of cod were formed resulting in good catches. Fleets were catching around 40 50% of the stock of adult fish each year. 1980s. Development of midwater trawls and bottom trawls capable of fishing on rough grounds opened up opportunities to fish in difficult areas such as the North Channel. Dual purpose trawls were developed to optimize catches of Nephrops and whitefish. The English beam trawl fleet grew rapidly in the 1980s, taking a bycatch of cod. The percentage of the stock of adult cod caught each year increased from 50% to 60%. Throughout the 1980s, TACs remained well above scientific advice to avoid triggering of The Hague Preference agreement which would have given Irish fleets a relatively bigger fraction of the TAC. 1990s. Mid water trawlers developed a summer and autumn fishery for cod. The English otter-trawl fleet declined and was reduced to inshore vessels taking mixed demersal fish, including codling. Fishing effort of the English beam trawl fleet peaked in 1990 then declined. Twin rig trawling for Nephrops and whitefish grew rapidly in the 1990s. This fleet also took a bycatch of cod. The Irish whitefish fleet moved increasingly to grounds off the south and west coasts, leaving mainly a Nephrops fleet and a number of vessels fishing rays, cod and haddock in the Irish Sea. A major change in the 1990s was the growth of the haddock stock. Vessels that would have fished for cod also targeted haddock in the western Irish Sea, although still taking a bycatch of cod in certain areas and time periods onwards. Emergency measures were introduced in 2000 to allow the maximum number of cod to spawn. These measures included a closure of the western and eastern Irish Sea spawning grounds from mid-february to the end of April, and modifications to trawl gear to improve selectivity. The closure was retained in , but only in the western Irish Sea. Derogations were allowed for Nephrops fishing in the closure, and experimental fisheries for haddock, flatfish and rays were permitted in some years with observers. Irish scientists successfully tested inclined separator panels in Nephrops trawlers, demonstrating large reductions in bycatch of cod. Vessels using such panels have been allowed to fish over a wider area of the closure since Vessels displaced from the closed area either switched to twin rigging for Nephrops, fished for cod in the North Channel and Clyde, or tied up. From 2001, the Clyde fishing grounds were also closed in spring as part of emergency measures to protect west of Scotland cod. TACs for Irish Sea cod from 2000 onwards were reduced substantially. Technical measures The fishery is managed by a combination of TAC, area closures, technical measures, and effort restrictions.

131 ICES WGCSE REPORT In 2000, the cod spawning grounds were closed for ten weeks from mid February to maximize the reproductive output of the stock (EU Regulations 304/2000 and 2549/2000). Revisions in 2001, 2002, and 2003 reduced the closure to the western Irish Sea only, coupled with changes in net design to improve selectivity. There are various derogations for gears not targeting cod, notably Nephrops trawls. During the closure, whitefish trawlers have been displaced to other fishing grounds or have switched to Nephrops trawling using mm mesh nets. The cod recovery plan introduced a system for limiting fishing effort by adjusting the number of fishing days allowed for various vessel categories deploying gears with various mesh sizes. STECF 2008 reported that the fishing effort of trawlers using mm mesh declined by 83% between 2003 and 2007, and by 86% for vessels with a track record of <5% cod in their landings. This was as a consequence of a combination of factors restricting the activities of these vessels. A number of whitefish vessels switched to Nephrops gears to take advantage of the additional days at sea and the high value of Nephrops and some were removed from the fleet. Despite vessels switching to Nephrops fishing, the fishing effort of trawlers with mm mesh declined by 2% between 2003 and 2007 and by 16% for vessels with a track record of <5% cod. This partly reflects vessels being decommissioned or increasingly fishing in the North Sea. Effort of beam trawlers declined by 48% and effort of fixed nets declined by 31% between 2003 and New technical regulations for EU waters came into force on 1 January 2000 (Council Regulation (EC) 850/98 and its amendments). The regulation prescribes the minimum target species composition for different mesh size ranges. Since 2001, cod in Division VIIa have been a legitimate target species for towed gears with a minimum codend mesh size of 100 mm. The minimum landing size for cod in the Irish Sea is 35 cm. Management plans Regulation (EC) No 423/2004 introduced cod recovery measures aimed at progressive rebuilding of spawning stock biomass. However in 2008 the EU adopted a long term plan for cod stocks and the fisheries exploiting those stocks (Council Regulation (EC) 1342/2008) that repeals Regulation (EC) No 423/2004, and has the objective of ensuring the sustainable exploitation of the cod stocks on the basis of maximum sustainable yield while maintaining a target fishing mortality of 0.4 on specified age groups. The scientific evaluation of the revised cod Management Plan (Council Regulation (EC) 1342/2008) indicates that it may not be sufficiently precautionary to allow rebuilding of the Irish Sea cod stock to a level where it can regain historical productivity by 2015 (see WGCSE 2009 Report, Section 9.2). The probability of recovery of the cod stock will be increased by measures to eliminate discards of cod which historically have mainly comprised undersized fish. Effort control Direct control of fishing effort has been a key aspect of cod recovery plans. Monthly effort limitation was extended to the Irish Sea (and other cod recovery areas) under Annex V to Council Regulation (EC) No 2287/2003. This Regulation and subsequent amendments (e.g. Council Regulation (EC) No 27/2005) restrict the number of allowable days fishing per month according to gear type, mesh size band and various derogations.

132 1502 ICES WGCSE REPORT 2013 The effort regulations have provided an incentive for some vessels previously using >100 mm mesh in otter trawls to switch to smaller mesh gears, thus claiming a larger number of days at sea. A.3. Ecosystem aspects Recruitment and the environment There is evidence that the reduction in cod recruitment observed in the Irish Sea since the 1990s may be as a consequence of a combination of small spawning stock biomass and poor environmental conditions, coinciding with a shift towards above average sea temperatures (ICES, 2006). B. Data B.1. Commercial catch B1.1. Landings Quarterly and annual landings data have been supplied by the UK (N. Ireland), UK (E&W), UK (Scotland), Ireland, Belgium, and the IOM from databases maintained by national Government Departments and research agencies. The landings figures may be adjusted by national administrations or scientists to correct for known or estimated misreporting by area or species. To avoid double counting of landings data, each UK region supplies data for UK landings into its regional ports, and landings by its fleet into non UK ports. Member States that have collected length and age-composition data for VIIa cod as required by the EU data Collection Framework entered quarterly and annual landings at age data on InterCatch. Quarterly and annual estimates of landings at age are provided by the UK (E&W), UK (NI), Belgium and Ireland. These have been raised to include landings by the other countries, then summed over quarters to produce the annual figures for input to stock assessment. In addition, the stock coordinator compiles the international landings and catch at age data and maintains a time series of such data with any amendments (Cod 7a CNAA XLS). The Excel spreadsheet files used for age distribution, adjustments and aggregations can be found with the stock co ordinator. Adjustments to official landings data The input data on fishery landings and age compositions are split into four periods (Figure ): 1 ) Landings in this period, provided to ICES by stock coordinators from all countries, are assumed to be un-biased and are used directly as the input data to stock assessments. 2 ) TAC reductions in this period caused substantial misreporting of cod landings into several major ports in one country, mainly species misreporting. Landings into these ports were estimated based on observations of cod landings by different fleet sectors during regular port visits. For other national landings, the WG figures provided to ICES stock coordinators were used. 3 ) Cod recovery measures were considered to have caused significant problems with estimation of landings. The ICES WG landings data

133 ICES WGCSE REPORT provided by stock coordinators for all countries are considered uncertain and estimated within an assessment model. Observations of misreported landings were available for 2000, 2001, 2002 and However, they have generally not been used to correct the reported landings but have been used to evaluate model estimates in those years. 4 ) The introduction of the UK buyers and sellers legislation is considered to have reduced the bias in the landings data but the level to which this has occurred is unknown. Consequently comparisons were made between the fit of the model to recorded landings under an assumption of bias and unbiased information. In addition to the above, Irish landings of cod reported from ICES rectangles immediately north of the Irish Sea/Celtic Sea boundary (ICES rectangles 33E2 and 33E3) have been reallocated into the Celtic Sea as they represent a combination of inaccurate area reporting and catches of cod considered by ICES to be part of the Celtic Sea stock (ICES, 2009). The amount of Irish landings transferred from VIIa to VIIe k by year is shown below: Year Tonnes The higher level in 2007 and 2008 was a consequence of limited quota in VIIe k and available quota in VIIa. Since 2009 more restrictive monthly quotas have been set for VIIa during periods of high cod abundance close to the VIIa VIIg boundary. There are no long-term trends in catch weights-at-age from 1982 onwards. Weightsat-age prior to 1982 are fixed at constant values lower than estimated for subsequent years, leading to sum-of-products errors, and weights-at-ages 6+ are becoming noisy for the last few years. Given these problems, and the likelihood of further deterioration in the quality of the data on older aged fish, WGCSE and WKROUND2 considered that future revision of historical catch-at-age data and associated weights is considered appropriate. However, WKROUND2 established that revising the weight-at-age would only represent only minor refinements to the model estimates of mortality and SSB trends and the reference point which are dependent on them compared to the sensitivities associated with the estimation of unallocated mortality. Consequently the revision of the weights-at-age should be conducted following the determination of the reasons for the current high mortality rates on the stock. Total mortality rates for the stock have been high throughout the time period for which information is available. Even when the stock was considered abundant and recruitment levels supported high levels of catch the gradient of the catch curve was in the range Year classes rapidly disappeared from the commercial landings data. The increase in the negative slope indicates that total "mortality" rates have increased over time and now are double that recorded in the historic data during the period when the stock was abundant. There is currently no evidence from the age compositions from surveys or commercial fishery operations of any improvement in age structure that would result from a reduction in total mortality.

134 1504 ICES WGCSE REPORT Landings (t) Official (WG) landings for all countries and Landings based on Landings considered ports used in the assessment quayside surveys biased but likely to plus reported to WG including have improved since B1.2. Discards No discards data are currently included in the assessment. Suitable discards estimates are not available prior to the mid-1990s and are not complete for many subsequent years. Available data indicates that discarding was historically mainly a function of MLS (35 cm) and therefore mainly restricted to catches of <=1-gp cod. EU countries are now required under the EU Data Collection Framework to collect data on discards of cod and other species. Consequently at WKROUND 2012 collation of recent discard information provided by Member States for the stock was carried out as a scoping exercise ready for future modelling and the provision of advice. Up to 2003, estimates of discards are available only from limited observer schemes and a self-sampling scheme. Observer data are collected using standard at-sea sampling schemes. Results have been reported to ICES. Discards data (numbers-at-age and/or length frequencies) are have been supplied for VIIa cod by Ireland, UK (Northern Ireland) and UK (E&W) and Belgium. The data were supplied raised to the appropriate fleet/métier level by the member states. Discards raising Ireland: Length frequencies from Irish (Marine Institute) observer trips in specified fleet métiers are raised to the trip level, averaged across trips during each year (not by quarter) then multiplied by the annual number of trips per year in the Irish fleet in VIIa to give raised annual LFDs for discards. An age length key from discards trips is then applied to give annual discards by age class and métier. Northern Ireland self-sampling scheme: The quantity of cod discarded from the UK (NI) Nephrops fishery from 1996 to 2002 was estimated on a quarterly basis from samples of discards and total catch provided by skippers. The discards samples contain

135 ICES WGCSE REPORT the heads of Nephrops tailed at sea. Using a length weight relationship, the live weight of Nephrops that would have been landed as tails only is calculated from the carapace lengths of the discarded heads. The number of cod in the discard samples is summed over all samples in a quarter and expressed as a ratio of the summed live weight of Nephrops in the discard samples (i.e. those represented as heads only in the samples). The reported live weight of Nephrops landed as tails only is then used to estimate the quantity of cod discarded using the cod:nephrops ratio in the discard samples. The length frequency of cod in the discard samples is then raised to the fleet estimate. Northern Ireland observer trips: Length frequencies from NI (AFBI) observer trips in specified fleet métiers are raised to the trip level, summed across trips during each year or by quarter then raised to the annual number of trips per year in the NI fleet in VIIa to give raised annual LFDs for discards. An age length key from discards trips is then applied to give annual discards by age class and métier. UK(E&W) observer trips: Trips are arranged on vessels selected using a vessel randomisation scheme. Discard numbers are raised to sampled hauls then to the trip. The trip-raised length frequencies from Cefas observer trips in specified fleet métiers are then raised to the trip level, summed across trips during each quarter. Sampled quarters are then raised to total discards by quarter from the landings to discards ratios at age. As recorded in the data sent annually to ICES catches and discards of cod within the Irish Sea by UK(E&W) vessels have been extremely low for a number of years. For instance in 2010, 63 hours fishing were observed distributed across quarters 1 4 with three cod caught and one discarded in quarter 1 (six hours trawling), 21 caught and 20 discarded in quarter 2 (32 hours) and 0 (zero) cod caught and discarded in quarters 3 (twelve hours) and 4 (13 hours). Belgium observer trips: Several Belgian métiers are operating in the Irish Sea. The beam-trawl fleet targeting sole and plaice (TBB_DEF_70-99_0_0) is the most important fleet, but, it should be noted that the OTB_DEF_70-99_0_0 métier (otter trawls) is becoming more important each year. Part of the landings and effort that could not be allocated to the main métiers, are referred to as: no allocated métier. Since the observers only collect information from the commercial beam trawlers, the data can only be raised to the TBB_DEF_70-99_0_0 fleet and not to all Belgian métiers operating in the Irish Sea. In order to find the most suitable raising procedure for the Belgian discard (and landing) data, the tools developed by the COST project were used. Having considered the different raising procedures, raising by hauls was found to be the most appropriate method for the Belgian cod VIIa data. The results of the raising procedure were scaled relative to the official landings. The time stratification for the Belgian data is by year, as sampling was insufficient to provide quarterly figures. It should be noted that due to the lack of Belgian individual length weight information, the length weight keys used in the analyses, are based on Irish sampling data. Note also that the Belgian minimum landing size has changed a couple of times over the last years, which is reflected in the differences in length frequency distributions between years of the retained and discarded part of the catch. From the beginning of 2004 until the 30th of June 2008: 40 cm; From the 1st of July 2008 until 30th of September 2011: 50 cm; From the 1st of October 2011 up to today: 35 cm.

136 1506 ICES WGCSE REPORT 2013 Raising to total international discards National, raised to fleet discard numbers-at-age from Ireland, Belgium, UK(E&W) and NI were added to give the international numbers (with no additional weighting). The data represents the main fleets discarding cod, i.e. Nephrops and beam trawlers. The current discard information is considered representative of the information for the main fleets highlighting strong differences between national, quarterly and potentially regional discard rates as the national fleets tend to fish differing areas with differing gears. The time-series are still too short to include the data within an assessment and at the youngest ages discard raising still needs some development, however that also applies to landings numbers-at-age, which have deteriorated significantly in quality in recent years in terms of sampling levels due to low levels of landings. B.2. Biological Natural mortality The current assessment uses constant values of M=0.2 (all ages). Maturity Maturity-at-age is summarized in the text table below. Age Proportion Mature ( ) Proportion Mature (2001-to present) A transition from the historic value of 0.38, developed at WKROUND2, is adopted for the period between 1996 and B.3. Surveys Nine research vessel survey series for cod in VIIa were used by WGCSE in In all surveys listed the highest age represents a true age not a plus group. UK(England and Wales) Beam Trawl Survey (E/W BTS-Q3): ages 0 (brought forward to obtain indices of age 1), years 1993 (1994) 2012(2013). The survey covers the entire Irish Sea excluding the North Channel and is conducted in September on the R.V. Corystes. The survey uses a 4 m beam trawl targeted at flatfish. The survey is stratified by area and depth band, although the survey indices are calculated from the total survey catch in the eastern Irish Sea, and without accounting for stratification except for ALKs. Numbers of 0-gp and 1-gp cod at-age per 100 km towed are provided for prime stations only (i.e. those fished in most surveys). An automated data extraction and analysis routine in r is now used, and the series was revised in 2008 using this routine. The 2009 assessment used data for years 1993 onwards. UK(Northern Ireland) October Groundfish Survey (NIGFS-WIBTS Q4): ages 1 3, years The survey series commenced in its present form in It comprises 45 three mile tows at fixed station positions in the northern Irish Sea, with an additional 12 one

137 ICES WGCSE REPORT mile tows at fixed station positions in the St George s channel from October 2001 (the latter are not included in the tuning data). The surveys are carried out using a rock hopper otter trawl deployed from the R.V. Lough Foyle. The survey designs are stratified by depth and seabed type. Virtually all cod are aged apart from 0-gp and 1- gp fish when particularly abundant. An ALK for the whole survey is used for filling in for any length groups with no ages at a station. Mean numbers at age per three mile tow are calculated separately by stratum, and weighted by surface area of the strata to give a weighted mean for the survey or group of strata. From 2002 onwards, all stations in the survey have been reduced to one nautical mile. A number of comparative 1-mile and 3-mile tows are done during each survey to build up calibration data. Since 2005, the RV Lough Foyle used for all surveys since 1992 has been replaced by the larger RV Corystes. The trawl gear and towing practices have remained the same. UK(Northern Ireland) March Groundfish Survey (NIGFS-WIBTS Q1): ages 1 5, years General description as for October Surveys above, except that 3-mile stations have been retained in all strata other than in the St Georges Channel. Since 2005, the RV Lough Foyle used for all surveys since 1993 has been replaced by the larger RV Corystes. The trawl gear and towing practices have remained the same. The 1992 survey had only partial coverage of the western Irish Sea and is no longer used in the assessment. UK(Northern Ireland) Methot Isaacs Kidd Survey (NIMIK): age 0 (brought forward to obtain indices of age 1), years 1994 (1995) 2012 (2013). The survey uses a Methot Isaacs Kidd frame trawl to target pelagic juvenile gadoids in the western Irish Sea at stations. The survey is stratified and takes place in June during the period prior to settlement of gadoid juveniles. Indices are calculated as the arithmetic mean of the numbers per unit sea area. UK(Scotland) groundfish survey in Spring (ScoGFS-WIBTS Q1): ages 1 5, years This survey represented an extension of the Scottish West Coast groundfish survey (Area VI), using the research vessel Scotia. The survey gear is a GOV trawl, and the design is two fixed position stations per ICES rectangle from 1997 onwards (17 stations) and one station per rectangle in 1996 (nine stations). The survey extends from the Northern limit of the Irish Sea to around UK (Scotland) groundfish survey in Autumn (ScoGFS-WIBTS Q4): ages 1 2, years This survey represented an extension of the Scottish West Coast groundfish survey (Area VI), using the research vessel Scotia. The survey gear is a GOV trawl, and the design is two fixed position stations per ICES rectangle from 1997 onwards (17 stations) and one station per rectangle in 1996 (nine stations). The survey extends from the Northern limit of the Irish Sea to around UK Fishery Partnership Surveys (UK-FSP), Western Irish Sea, in March: ages 1 5, years The Irish Sea roundfish survey was initiated in 2003 as a fully collaborative project between the fishing industry and Cefas scientists. It forms part of the UK Fisheries Science Partnership funded by the UK s Department for Environment, Food and Rural Affairs (Defra). The main objective of the Irish Sea roundfish survey is to develop

138 1508 ICES WGCSE REPORT 2013 a time-series of data to track year-on-year changes in abundance, population structure and distribution of the target species (cod, haddock and whiting). The results of the surveys provide information supporting the scientific assessment of the stocks and the management of the fisheries in the Irish Sea. The surveys were designed to achieve full coverage of potential cod, haddock and whiting habitats within the area of the main roundfish fisheries of the Irish Sea, using a stratified design to allow additional trawling effort in areas expected to have the greatest densities of cod, haddock or whiting. UK Fishery Partnership Surveys (UK-FSP), Eastern Irish Sea, in March: ages 1 5, years The Irish Sea roundfish survey was initiated in 2003 as a fully collaborative project between the fishing industry and Cefas scientists. It forms part of the UK Fisheries Science Partnership funded by the UK s Department for Environment, Food and Rural Affairs (Defra). The main objective of the Irish Sea roundfish survey is to develop a time-series of data to track year-on-year changes in abundance, population structure and distribution of the target species (cod, haddock and whiting). The results of the surveys provide information supporting the scientific assessment of the stocks and the management of the fisheries in the Irish Sea. The surveys were designed to achieve full coverage of potential cod, haddock and whiting habitats within the area of the main roundfish fisheries of the Irish Sea, using a stratified design to allow additional trawling effort in areas expected to have the greatest densities of cod, haddock or whiting. UK IS-AEMP, Irish Sea, in March: SSB index, years 1995, 2000, 2006, 2008, The Annual Egg Production Survey has been used to obtain estimates of cod spawning stock biomass in the Irish Sea. B.4. Commercial cpue Commercial cpue for several national fleets are available to ICES, but these are no longer used in the assessment. B.5. Other relevant data No other relevant data. C. Assessment: data and method Model used: SAM (State space assessment model) Software used: FLSAM The SAM model allows for the inclusion of a spawning biomass index which the B- ADAPT would need recoding for. Consequently and in view of the current trend within ICES to move assessments towards statistical models, and the consistency with B-ADAPT in the previous WGCSE evaluation of the stock, the SAM model was chosen for the assessment of the Irish Sea cod. Model configuration: # Auto generated file # Datetime : :44:57 # Min, max age represented internally in model 1 6

139 ICES WGCSE REPORT # Max age considered a plus group? (0 = No, 1= Yes) 1 # Coupling of fishing mortality STATES (ctrl@states) # # # catch # NIGfsMar # ScoGfsQ # ScoGfsQ # NIGfsOct # UKFspW # UKFspE # EngBtsSep # NIMikNet # EggSurvey # Coupling of catchability PARAMETERS (ctrl@catchabilities) # # # catch # NIGfsMar # ScoGfsQ # ScoGfsQ # NIGfsOct # UKFspW # UKFspE # EngBtsSep # NIMikNet # EggSurvey # Coupling of power law model EXPONENTS (ctrl@power.law.exps) # # # catch # NIGfsMar # ScoGfsQ # ScoGfsQ # NIGfsOct # UKFspW # UKFspE # EngBtsSep # NIMikNet # EggSurvey # Coupling of fishing mortality RW VARIANCES (ctrl@f.vars) # # # catch # NIGfsMar # ScoGfsQ # ScoGfsQ # NIGfsOct # UKFspW # UKFspE # EngBtsSep # NIMikNet # EggSurvey # Coupling of log N RW VARIANCES (ctrl@logn.vars) # Coupling of OBSERVATION VARIANCES (ctrl@obs.vars) # # # catch # NIGfsMar # ScoGfsQ # ScoGfsQ # NIGfsOct # UKFspW

140 1510 ICES WGCSE REPORT # UKFspE # EngBtsSep # NIMikNet # EggSurvey # Stock recruitment model code (0=RW, 1=Ricker, 2=BH,... more in time 0 # Years in which catch data are to be scaled by an estimated parameter (mainly cod related) 9 # Years #Ages # Fbar range 2 4 # Checksums to ensure correct reading of input data D. Short-term projection Not carried out. E. Medium-term projections Not carried out. F. Long-term projections Not carried out. G. Biological reference points TYPE VALUE TECHNICAL BASIS MSY MSY Btrigger t Bpa Approach FMSY 0.4 Provisional proxy. Fishing mortalities in the range of are consistent with FMSY. Blim 6000 t Blim= Bloss, lowest observed level. Precautionary BPA t Bpa = MBAL; this level affords a high probability of maintaining the SSB above Blim. Below this value the probability of below-average recruitment increases. Approach Flim 1.00 Flim= Fmed (unchanged since: 2010) FPA 0.72 Fpa: Fmed* This F is considered to have a high probability of avoiding Flim. Fishing mortalities above Fpa have been associated with the observed stock decline.

141 ICES WGCSE REPORT H. Other issues H.1. Historical overview of previous assessment methods Until Model used: B Adapt. Software used: B Adapt F.exe (13/5/06) Model Options chosen: SETTING Plus group VALUES 5 plus FBAR range 2 4 (arithmetic mean) Year range for tuning VPA 1992 onwards Surveys after final year of catch data used. FMULT = 0.75 VPA model or cohort analysis used First age with constant catchability q plateau Tapered time weighting applied Number of missing catch multipliers No. ages for terminal F mean, and scaling factor for mean v (exact) 0 for all tuning fleets Entered as 3 for all tuning fleets No 9 (from 2000 onwards) ages = 1; scaling factor = 1.0; arithmetic mean (i.e. F(4) = F(3)) Constraint on F or catch? Stiffness weight (λ) Constrain F; λ = 1.0 Prior weighting of fleets Output tables None VPA output I. References Connolly, P. and Officer R The use of tagging data in the formulation of the Irish Sea cod recovery plan. ICES CM 2001/O:05. Crozier, W., Heath, M. and Righton, D Spatial structure of cod populations: What are the implications for the assessment and management of cod stocks? A Mini Symposium, May, 2007, The Agri Food and Biosciences Institute (AFBI), Newforge Lane, Belfast, United Kingdom BT9 5PX. Goodsir, F., Armstrong, M.J., Witthames, P.R., Maxwell, D.L. and Fox, C.J The use of species-specific TaqMan probes for identifying early stage gadoid eggs following formaldehyde fixation. ICES J. Mar. Sci., 65, ICES Report of the Working Group on the Assessment of Northern Shelf Demersal Stocks, 9 18 May 2006 (ICES CM 2006/ACFM:30). ICES Report of the Working Group on Celtic Seas Ecosystems, May 2009 (ICES CM 2009/ACOM:09).

142 1512 ICES WGCSE REPORT 2013 Lordan, C., Ó Cuaig, M., Graham, N., and Rihan, D The ups and downs of working with industry to collect fishery-dependent data: the Irish experience. ICES Journal of Marine Science, 68: Ó Cuaig, M. and Officer, R Evaluation of the benefits to sustainable management of the seasonal closure of the Greencastle codling (Gadus morhua) fishery. Fisheries Bulletin No. 27/2007. STECF Evaluation of closed area schemes (SGMOS 07 03).

143 ICES WGCSE REPORT Stock Annex 6.3: Irish Sea haddock (VIIa) Stock Working Group Haddock in VIIa (Irish Sea) Working Group on the Celtic Sea Ecoregion (WGCSE) Date 8 February 2013) Revised by WKROUND 2013 (Pieter Jan Schön) A. General A.1. Stock definition The independence of the haddock population in Division VIIa from populations off the west of Scotland (Division VIa) and the Celtic Sea (Divisions VIIb k) is currently inferred only from the very different trends in year-class strength in each area. A.2. Fishery Historically, fishing for haddock in the Irish Sea was mainly carried out by UK(NI) semi-pelagic (midwater) trawlers using 100 mm mesh codends, particularly targeting aggregations that can be detected acoustically. These conditions prevail mainly during winter and spring when the hours of darkness are longest, and the fish are aggregating on the spawning grounds in the western Irish Sea. However, due to the decline in cod abundance in the Irish Sea and the introduction of effort limitations, the directed fishery has declined substantially to the point that haddock are now only reported as bycatches associated with the UK (NI) and Irish Nephrops fisheries using single nets with 70 mm codends or twin trawls with 80 mm codends. The haddock stock is mainly distributed in the western Irish Sea and south of the Isle of Man, preferring the coarser seabed sediments around the periphery of the muddy Nephrops grounds. Juveniles are taken extensively in the otter trawl fisheries in these areas, leading to substantial discarding (see Section B1.2). The nature of the fishery has been modified by the cod closure since 2000 (Council Regulation (EC) No 304/2000) Targeted fishing with whitefish trawls was prohibited inside the closure from mid-february to the end of April. Derogations for Nephrops fishing were allowed. Irish Nephrops trawlers were involved in an experiment to test inclined separator panels in 2000 and 2001, the object being to minimize the bycatch of cod. Fishing inside a small area of the western Irish Sea closed to all fishing in spring 2000 and 2001 was permitted if separator panels were used. These panels would also have allowed escapement of part of the haddock catch. Closure of the main whitefish fishing grounds in spring 2000 resulted in a shift in fishing activities of midwater trawlers and other UK(NI) whitefish vessels into the North Channel (Area VIIa) and Firth of Clyde (VIa south). A subsequent closure of the Firth of Clyde in spring 2001 under the VIa cod recovery programme (Council Regulation (EC) No 456/2001) resulted in a reduction in reported fishing activity in this region. Several rounds of decommissioning in , 2001 and 2003 have reduced the size of the commercial fleets. UK vessels decommissioned at the beginning of 2002 accounted for 17% of the haddock landings from the Irish Sea in A further round of decommissioning in 2003 removed 19 out of 237 UK vessels that operated in the Irish Sea at the beginning of 2004, representing a loss of 8% of the fleet by number and 9.3% by tonnage. The decline in fishing effort associated with this fleet has continued due to effort restrictions associated with the long-term management plan for cod (EC

144 1514 ICES WGCSE REPORT 2013 Regualtion, 1342/2008) which has resulted in sequential and ongoing effort reductions particularly with the larger mesh gadoid fleet, effort levels are now to all intents and purposes negligible, with only the Nephrops fishery contributing to catches of haddock, many of which are discarded. A.3. Ecosystem aspects Decadal changes in environmental conditions have occurred in the Irish Sea. An increase in sea surface temperature occurred during the 1990s, during which period the biomass of haddock in the Irish Sea increased dramatically, although there may be no causal link. For North Sea haddock, no link could be found between temperature and recruitment (Cook and Heath, 2005). The northwards shift of both cold and warm-water zooplankton Calanus into the Irish Sea is expected to impact on the distribution of many species (Richardson, 2008). Reductions in cod recruitment in the Irish Sea since the 1990s may be due to a combination of small spawning stock biomass and poor environmental condition (Drinkwater, 2005). Lynam et al. (2011) observed an abrupt step change ( regime shift ) in the production of copepod biomass in the Irish Sea in the late 1980s (Figure A3.1). Copepods are particularly important prey for larval and pelagic post-larval haddock, although it is unknown what impact the reduced productivity of copepods may have had on the haddock population which increased during the 1990s. Figure A.3.1. Copepod biomass regime shift (left hand) in the Irish Sea from the Continuous Plankton Recorder (From Lynam et al., 2011) and the associated changes in annual production (right hand). The diet of haddock varies seasonally and according to location and body size. In winter, haddock of all sizes feed mainly on benthic invertebrates, for example, polychaetes, small crustaceans and echinoderms. In spring and summer, fish prey, especially sandeels, are important particularly for larger haddock. Norway pout is also important prey for haddock. During herring spawning seasons, haddock will feed heavily on herring eggs. B. Data B.1. Commercial catch B1.1 Landings series The following table gives the source of fishery data for Irish Sea haddock:

145 ICES WGCSE REPORT Kind of data Country Caton (catchin-weight) Canum (catch-atage/innumbers) X* Weca (weight-atage in the catch) X* Matprop (proportion mature-byage) Length composition in catch UK(NI)* UK(E&W) UK(Scotland) UK (IOM) Ireland France Belgium X X X X X X X X X X* X* X *Complete data is not available for 2003 for UK (NI). Official landings data are census data provided from databases maintained by national Government Departments according to EU logbook data and/or sales slips. Quarterly landings are provided by the UK(E&W), UK(Scotland), Belgium and France, and annual landings are provided by UK(IOM). In some cases the official landings estimates are biased due to known or suspected misreporting. Two forms of misreporting have been corrected for in landings-series from UK(NI) and Ireland: 1 ) Reporting of landings by Irish fleets from the rectangles immediately south of VIIa into the rectangles immediately north. These landings are removed from the VIIa data and included in the data for the VIIb k haddock assessment. 2 ) Landings by UK(NI) vessels from 1993 onwards were subject to inaccurate reporting due to quota restrictions, and were estimated from a survey approach rather than from official statistics. Trip landings by individual vessels, stratified by port and fishing method, were recorded during weekly market sampling trips to the main NI ports. Total fleet landings were estimated by raising from sampled vessel trips to the total recorded number of fishing trips in each stratum. From 2006 onwards, market sampling observations of landings were sufficiently similar to official statistics to allow use of the official data. No observations were possible in B1.2. Discards estimates Discards estimates for haddock are available from four main sources: Observer scheme in Northern Ireland (2006 onwards, plus some additional trips from ). The scheme from 2006 onwards is based on a more robust and extensive sampling design than the sampling in Observer scheme in Ireland (1996 onwards for fleets catching haddock). Self-sampling scheme on Northern Ireland Nephrops trawlers ( and 2008 onwards). Haddock are currently mainly caught as a bycatch in the Irish and Northern Irish Nephrops fishery. Historically, the Northern Irish fleet also targeted gadoids with semi-pelagic gear but this fishery has ceased. Observer coverage of the semi-pelagic fleet was insufficient to allow estimates for this fleet to be calculated. Therefore total

146 1516 ICES WGCSE REPORT 2013 fleet discard estimates are only produced for the Irish and Northern Irish OTB fleets (which mainly target Nephrops), based on the at-sea sampling schemes. Gerritsen and Schön (2013) estimated total fleet discards from the UK(NI) observer scheme (2006 onwards) and the Irish observer scheme (1996 onwards). Various raising variables and stratifications were explored. Auxiliary variables tested included effort (in hours trawled, days-at-sea and number of trips), haddock landings, gadoid landings and total landings of all species. Stratification by target species was also explored based on the species composition of the landings. Days fished was found to be the most appropriate auxiliary variable. On this basis, discard numbers and length frequencies are estimated from representative samples per haul, then raised to the haul and then to all hauls in the trip. The discard rates of sampled vessels (number and weight of discarded haddock per calendar day) are then averaged over all trips sampled and raised to the total census estimates of days fished in that stratum. Discards estimates from the UK(NI) self-sampling scheme have been estimated using the landings live weight of Nephrops that are landed as tails as the auxiliary variable for raising. The live weight of tailed Nephrops was calculated by applying a carapace length live weight relationship to the length frequency of carapaces in the discards samples. The raising factors were obtained using the fleet census data on total live weight of Nephrops landed as tails by the fleet. B1.3. Fishery length and age compositions B Landings Length compositions of landed haddock are estimated from regular market sampling in Ireland and Northern Ireland. Length compositions for sampled trips are recorded within sampling strata defined by quarter and gear, after post-stratification to include only VIIa trips. Gear strata for Northern Ireland are Nephrops gears and whitefish gears (mainly semi-pelagic gears historically). Gear strata for Ireland are otter trawls (OTB) and beam trawls (TBB). Age compositions are estimated for these quarter/gear strata from national age length keys built up from samples collected during the market sampling trips (Gerritsen, 2013). International length and age compositions are derived by raising the combined UK(NI) and Ireland data to total international landings. No data are available for UK(NI) for B Discards The length compositions of discarded haddock from the at-sea sampling schemes are obtained from the raising procedure described in B.1.2. Because of the rapid growth of young haddock, age length keys should be constructed on a quarterly basis. However, age sampling during the at-sea sampling schemes in Northern Ireland and Ireland was considered to be insufficient even at an annual level. Therefore knife-edge age splits to quarterly length compositions were applied each year and verified by age data from surveys and observer trips. The following criteria were used: LENGTH (CM) AGE 0 AGE 1 AGE 2 AGE 3 Q Q Q Q

147 ICES WGCSE REPORT Weights-at-length were estimated by applying a fixed length weight relationship with the parameters a = and b = B.2. Biological B2.1. Growth and stock weights-at-age Since 2001 WGNSDS and WGCSE calculated stock weights of Irish Sea haddock by fitting a von Bertalanffy growth curve to all available survey estimates of mean length-at-age in March, with an additional vector of parameters estimated to allow for year-class effects in asymptotic length. To increase the number of observations for older age classes, the mean lengths-at-age in UK (NI) first-quarter landings were included for age classes three and over. The growth of haddock was modelled at WKROUND (2013) from the same dataset (Ligas et al., 2013) using a Multiple Linear Regression model, and a Mixed Effect model, using the package R ( (R Core Team, 2012). The results were compared to fitting a von Bertalanffy growth function (VBGF) with year-class effects. The predicted VBGF lengthsat-age were considered most appropriate, and are used to estimate stock weights-atage for each year (as was done in previous years). B Spawning and maturity Egg surveys show that spawning of Irish Sea haddock occurs mainly in the western region and commences in January, peaks in mid to late March, and is complete by the end of April (Figure B2.1). The majority of 0-gp haddock caught in trawl surveys are in or near the area of spawning. Patterns of maturity have been investigated from length-stratified samples of fish taken from all trawl catches of haddock on AFBI trawl surveys in March from 1992 onwards, i.e. fish collected near the peak of spawning. At this time, maturity is easily identified as females in the batch spawning cycle or recently spent, or males with milt production or recently spent. A GLM analysis carried out by the UK(NI) to evaluate the effects of year, region, age, and length on the probability of being mature showed that maturity is determined differently for male and female haddock. Maturity was found to be predominantly a function of length in male haddock, whereas age was the main factor in females. Interannual variation in the proportion mature was mostly confined to the age 2 group, whereas other age groups were either fully immature or fully mature. Annual maturity ogives are relatively noisy, so the assessment currently uses a longer term average proportion mature for combined sexes: Age 0 and 1: 0.00; age 2: 0.72; age 3: 0.97; age 4+: 1.00 Fecundity data collected during the Irish Sea egg production surveys in 2006, 2008 and 2010 indicate that relative fecundity (eggs.g -1 ) increases with fish size, so that older mature haddock contribute more to spawning than indicated by their biomass.

148 1518 ICES WGCSE REPORT 2013 Figure B2.1. Spatial distribution of stage-1 haddock eggs in the Irish Sea in 2006, 2008 and 2010, at peak spawning time, from a GAM model fit to data collected on Cefas (England), AFBI (N.Ireland) and Marine Institute (Ireland) research vessels. Right hand plots show the GAM estimates of seasonal pattern of egg production in each year. B3.1. Natural mortality Natural mortality rates have not been estimated directly for Irish Sea haddock. A range of life-history based methods to infer possible age-invariant values for M, using the methods of. Alverson and Carney (1975); Hoenig (1983), Pauly (1980), Ralston, 1987), give vaues between 0.3 and 0.4. Age-dependent M values can be inferred from the methods of Gislason et al. (2010) and Lorenzen (1996). The Lorenzen method was used by WKROUND (2013) for exploring age-based assessments.

149 ICES WGCSE REPORT AGE 0 AGE 1 AGE 2 AGE 3 AGE 4 AGE 5 Lorenzen M predictions B.3. Surveys Eight research vessel survey-series for haddock in VIIa are available. (In all surveys listed the highest age represents a true age not a plus group, data ranges indicate data available for the current assessment). UK(NI) groundfish survey (NIGFS) in March (age classes 1 to 5, years 1992 to 2012). Acronym changed from NIGFS-Mar to NIGFS-WIBTS-Q1. The survey-series commenced in its present form in It comprises 45 three mile tows at fixed station positions in the northern Irish Sea, with an additional 12 one mile tows at fixed station positions in the St George s channel from October 2001 (the latter are not included in the tuning data). The surveys are carried out using a rockhopper otter trawl deployed from the R.V. Lough Foyle ( ) and the R.V. Corystes since The survey designs are stratified by depth and seabed type. The mean numbers-at-length per three mile tow are calculated separately by stratum, and weighted by surface area of the strata to give a weighted mean for the survey or group of strata. The survey design and time-series of results including distribution patterns of whiting are described in detail in Armstrong et al., UK(NI) groundfish survey (NIGFS) in October (age classes 0 to 3; years 1991 to 2011). Acronym changed from NIGFS-Oct to NIGFS-WIBTS-Q4. Description as for UKNI-GFS-March above. UK(NI) Methot Isaacs Kidd (MIK) net survey in June (age 0; years 1994 to 2011). The survey uses a Methot Isaacs Kidd frame trawl to target pelagic juvenile gadoids in the western Irish Sea at stations. The survey is stratified and takes place end of May/early June during the period prior to settlement of gadoid juveniles. Indices are calculated as the arithmetic mean of the numbers-per-unit sea area. UK Fishery Science Partnership (FSP) Irish Sea roundfish survey (age 1 to 7; 2005 to 2012). The Fisheries Science Partnership (FSP) is a government-funded (UK) initiative which began in This survey is conducted by CEFAS using two commercial fishing vessels. A stratified design to allow additional trawling effort in areas expected to have the greatest densities of cod, haddock or whiting. Republic of Ireland Irish Sea-Celtic Sea groundfish survey (IR-ISCSGFS) in November (ages 0 to 5; years 1997 to 2002). This survey commenced in 1997 and is conducted in October November on the R.V. Celtic Voyager. The survey operates mainly in the western Irish Sea but has included some stations in the eastern Irish Sea. The survey design has evolved over time and has different spatial coverage in different years. Indices are calculated as arithmetic means of all stations, without stratification by area. The survey was terminated in 2002 as a consequence of a vessel change. Republic of Ireland groundfish survey (IR-GFS) in autumn (age classes 0 to 6, years ).

150 1520 ICES WGCSE REPORT 2013 This survey commenced in 2003 and is an IBTS-coordinated survey, conducted in October November on the R.V. Celtic Explorer. The survey is an extension of a survey covering Divisions VI and VIIb k. A GOV otter trawl with standard groundgear and a 20 mm codend liner is used. Indices are calculated as arithmetic means of all stations, without stratification by area. The survey operated for only two years within the Irish Sea. UK(Scotland) groundfish survey (SCOGFS) in spring (age classes 1 to 6, years ). This survey represents an extension of the Scottish West Coast groundfish survey (Area VI), using the research vessel Scotia. The survey gear is a GOV trawl, and the design is two fixed-position stations per ICES rectangle from 1997 onwards (17 stations) and one station per rectangle in 1996 (nine stations). The survey extends from the Northern limit of the Irish Sea to around 53 o 30. The survey was terminated in UK(Scotland) groundfish survey (SCOGFS) in autumn (age classes 0 to 6, years ). The survey covers a similar area to the ScoGFS in Spring, but has only stations. The survey was terminated in B.4. Commercial cpue Only one historical fishery lpue dataseries (Irish otter trawl) has been provided to the WG for VIIa haddock, but is not used in any assessment. IRE OTB [Irish Otter trawl - Effort in hours numbers-at-age in 1000 s] B.5. Other relevant data None. C. Assessment: data and method Model used: SURBA Software used: SURBA version 3.0 Model Options chosen: Year range: Age range: 1 5 Catchability: 1.0 at all ages Age weighting 1.0 at all ages Smoothing (Lambda): 1.0 Cohort weighting: not applied Reference age 2 Survey used NIGFS-WIBTS-Q1

151 ICES WGCSE REPORT The 2005 WG performed an extensive analysis of survey data for Irish Sea haddock. The effect of smoothing (lambda=1.0 and 0), fitting constant catchability (1.0 for all ages) or variable catchability-at-age and the choice of reference age were explored. The results indicated that the choice of catchability-at-age and using different values for the smoothing parameter had very little effect on the temporal trends in SSB or recruitment, and a lambda value of 1.0 reduces the noise in Z without oversmoothing the trends. Changing the reference age had very little effect on the results. The SURBA parameters used in the 2012 assessment have remained unchanged from assessments made during The following input data for catch-at-age assessments was developed by WKROUND 2013: Type Name Year range Caton Catch-in-tonnes 1993 last data year Canum Catch-at-age in numbers 1993 last data year Weca Weight-at-age in the 1993 last commercial catch data year West Mprop Fprop Weight-at-age of the stock at spawning time. Proportion of natural mortality before spawning Proportion of fishing mortality before spawning 1993 last data year 1993 last data year 1993 last data year Matprop Proportion mature-at-age 1993 last data year Natmor Natural mortality 1993 last data year Age Variable from year to year range Yes/No 0 5+ Yes 0 5+ Yes 0 5+ Yes 0 5+ Yes: uses growth model from UK (NI) March GFS data 0 5+ No set to 0 for all ages in all years 0 5+ No set to 0 for all ages in all years 0 5+ No the same ogive for all years 0 5+ No Lorenzen (1996) M-atage vector, plus age-invariant M ( ) for sensitivity analysis. D. Short-term projection Short-term trends-based projections have been carried out using the results of the Surba model fitted to NIGFS-WIBTS-Q1 survey data. Final-year model estimates of numbers-at-age are projected forwards using the recent Z at-age estimates, with unobserved recruitments predicted as long-term GM of observed values, and stock weights assumed to be mean of last three years. E. Medium-term projections No medium-term projections are done for this stock as the short time-series of stock and recruitment estimates precluded any meaningful prediction of the medium-term dynamics of the stock. F. Yield and biomass per recruit/long-term projections Last calculations of yield-per-recruit reference points was by WGNSDS 2004 based on the exploitation patterns from XSA fitted to data out to a 5+ group. This is no longer valid, in the absence of a catch-based assessment providing fishery selectivity patterns.

152 1522 ICES WGCSE REPORT 2013 G. Biological reference points No biological reference points are currently established directly for this stock. ICES proposes that FPA be set at 0.5 by association with other haddock stocks; however in the absence of a catch-based analytical assessment, the absolute value of F in this stock at present is poorly known. H. Previous assessments of the stock Working groups in explored the use of TSA, ICA and B-Adapt (which allows for years with missing catch data) with estimates of landings-at-age. The results of these models were unsatisfactory. Because the assessment suffers from poor data quality with a relatively short time-series, from 2004 onwards the WG presented assessments of recent stock trends based on survey data only. Since 2005 a Survey Based Assessment (SURBA) was used; which is considered to give a reliable picture of the status of the stock at least for SSB and recruitment. The use of multiple surveys within single SURBAR models was compared to the historic stock trends derived from single surveys. This revealed low confidence in a SURBAR model with multiple surveys, with a high signal to noise ratio in some selected surveys resulting in poor convergence of the model (Benchmark Workshop 2013). The NIGFS-WIBTS-Q1 survey data was considered to provide the most reliable model outputs. I. References Alverson, D. L., and M. J. Carney A graphic review of the growth and decay of population cohorts. J. Cons. Int. Explor. Mer 36: Armstrong, M.J., Peel, J., McAliskey, M., McCurdy, W., McCorriston, P. and Briggs, R Survey indices of abundance for cod, haddock and whiting in the Irish Sea (Area VIIaN): Working Document No. 3 submitted to 2003 meeting of the ICES Working Group on Assessment of Northern Shelf Demersal Stocks. 33pp. Beverton, R. J. H Patterns of reproductive strategy parameters in some marine teleost fishes. J. Fish Biol. 41(Supplement B): Borges, L., Zuur, A.F., Rogan, E. and Officer, R Choosing the best sampling unit and auxiliary variable for discards estimations. Working Document No. 3 submitted to 2005 meeting of the ICES Working Group on Assessment of Northern Shelf Demersal Stocks. 25pp. Gerritsen, H Haddock landings in VIIa: Revision of the historic Irish and Northern Irish landings numbers at age. Working Document: ICES WKROUND Gerritsen, H. and Schön, P.-J Haddock discards in VIIa. Evaluation of raising methods and derivation of a revised series. Working Document: ICES WKROUND Gislason H., Daan N., Rice J.C., Pope J.G Size, growth, temperature and the natural mortality of marine fish. Fish and Fisheries 11, Hoenig, J. M Empirical use of longevity data to estimate mortality rates. Fish.Bull. 82: Ligas, A., Schön, P.-J. and Lundy, M Modelling haddock growth in ICES Division VIIa using alternative methods to the von Bertalanffy Growth Model. Working Document: IC- ES WKROUND Lorenzen, K The relationship between body weight and natural mortality in juvenile and adult fish: a comparison of natural ecosystems and aquaculture. Journal of Fish Biology 49,

153 ICES WGCSE REPORT Lynan, C.P., Lilley M.K.S., Bastin, T., Doyle, T.K., Beggs, S.E. and Hays, G.C Have jellyfish in the Irish Sea benefited from climate change and over fishing. Global change biology. 17, Pauly, D On the interrelationships between natural mortality, growth parameters, and mean environmental temperature in 175 fish stocks. J. Cons. Int. Explor.Mer 39: Ralston, S Mortality rates of snappers and groupers. In J. J. Polovina, S. Ralston (eds.), Tropical Snappers and Groupers: Biology and Fisheries Management. Westview Press: Boulder, CO. pp

154 1524 ICES WGCSE REPORT 2013 Stock Annex 6.4: Nephrops in FU 14 Irish Sea east Stock Working Group Date updated Revised by Irish Sea east Nephrops (FU14) May 2010, Assessment of Northern Shelf Demersal Stocks May 2013, Working Group on the Celtic Seas Ecoregion WGCSE /Ana Leocádio A. General A.1. Stock definition Throughout its distribution, Nephrops is limited to muddy habitat, and requires sediment with a silt & clay content of between % to excavate its burrows, and this means that the distribution of suitable sediment defines the species distribution. Adult Nephrops only undertake very small scale movements (a few 100 m) but larval transfer may occur between separate mud patches in some areas. In the eastern Irish Sea the Nephrops stock inhabits an area of muddy sediment extending along the Cumbria coast and its fishery contributes to less than 10% of overall Irish Sea landings. There is little evidence of mixing between the east and west Irish Sea stocks due to the nature of water current movements in the Irish Sea. The two are treated as separate populations since they have differing population characteristics. FU NAME ICES DIVISIONS STATISTICAL RECTANGLES 14 Irish Sea east VIIa 35 38E6; 38E5 A.2. Fishery Between 1999 and 2003 the number of vessels fishing for Nephrops in FU14 declined by 40% to a fleet of around 50 vessels. This was largely due to the reduction in the number of visiting UK vessels and the decommissioning of part of the Northern Irish and local English fleets. Since then the number of vessels fishing the area has returned to around 70 vessels mainly from England and Northern Ireland. Currently, around 30 of these vessels, between six and 23 m in length, have their home ports in Whitehaven, Maryport and Fleetwood, England. The rest of the fleet is generally made up of larger vessels from Kilkeel or Portavogie, Northern Ireland. Between 1987 and 2006, landings from FU14 appeared relatively stable, fluctuating around a long-term average of about 550 t. Landings in 2007 rose significantly (959 t) reaching maximum values since 1978, this is after landings dropped in 2003 to their lowest apparent level since The following years 2008 and 2009 figures of 676 and 708 t respectively kept high comparing with any other figures recorded since Recently ( ), landings went back to historical values, averaging 558 t. Although, the introduction of the buyers and sellers legislation in 2006 really precludes direct comparison with previous years as reporting levels are considered to have significantly improved since. Over the last ten years ( ) UK vessels have landed, on average, 92% of the reported annual international landings. Republic of Ireland vessels increased their share of the landings to 35% in 2002 but it has since declined to 6% (average 2008 to 2012). In 2009, most of the landings were made into England with a high proportion

155 ICES WGCSE REPORT of these landings (67% of the directed landings and 62% of the total landings) being made by visiting Northern Irish vessels. UK Nephrops directed effort has fluctuated around a downward trend since From , fewer of the Northern Irish fleet were landing in England. The differences between lpue figures for individual vessels suggest that earlier years may have included less truly directed effort. Reductions in quota between 2002 and 2006 for VIIa cod and plaice may have restricted total effort in FU14 thereby reducing the more casual effort on Nephrops. From 2003 the main fleets targeting Nephrops include Nephrops directed single-rig and twin-rig otter trawlers operating out of ports in UK(NI), UK(E&W) and Ireland. Regulations Regulations introduced as part of a revised package of EC Fisheries Technical Conservation measures in 2000 remain in place. This legislation incorporates a system of 'mesh size ranges' for each of which has been identified a list of target species. In effect, nets in the mm mesh size range must have at least 35% of the list of target species (which includes Nephrops) and the mm mesh size range requires at least 30% of the list of target species. A square mesh panel (SMP) of 80 mm is required for mm nets in the Irish Sea. Vessels using twin-rig gear in the Irish Sea must comply with a minimum mesh size of 80 mm (no SMP is required for nets with 80 mm meshes and above). Other regulations restricting trawling in other fisheries within the Irish Sea will affect effort on these and other stocks. This could either attract local effort or even relocate effort to fisheries in other areas. Although unrestrictive the result of better catch information through the buyers and sellers legislation introduced to the UK from 2006 will have the same effect as quota uptake of stocks which used to be misreported will be quicker. As well as an Area VII TAC other Nephrops conservation measures in the Irish Sea are a minimum landing size of 20 mm CL length (equivalent to 37 mm tail length or 70 mm total length). In addition to Nephrops measures the cod spawning areas of the Irish Sea are closed to whitefish directed vessels between 14th February to 30th April part of the Irish Sea cod recovery plan. There is derogation for Nephrops vessels during this closure. A.3. Ecosystem aspects The working group has collated no information on the ecosystem aspects of this stock. B. Data B.1. Commercial catch Length and sex compositions of Nephrops landed from the Irish Sea east are estimated from port sampling by England and Wales. Length data from this sampling are applied to catch samples collected at sea and raised to total international landings. Catch length samples are collected independently of landings length samples but both are considered representative. The independent raising process means that the final annual catch length frequency distribution still requires scaling to the reported landings. Using a discard ogive derived from samples collected in the early 1990s an initial estimate of discards is taken from the catch distribution. These are then added to the landings distribution to create a dummy catch distribution. The difference be-

156 1526 ICES WGCSE REPORT 2013 tween the numbers-at-length for both the raised sampled and dummy catch distribution was then used to tune a raising factor by minimizing the sums of squares. Once the raising factor is derived, the final discard length distribution is the difference between the raised catch distribution and the landings distribution and a final catch distribution is a sum of the landings and discard distributions. In 2008 a new discard ogive was calculated from the discard samples collected from 2003 until March 2008 and applied to the 2003 data to date. The lack of discard and catch data between 1995 and 1999 is likely to adversely affect the quality of any analytical assessments. Apparent differences between catch LFDs and discard practices in 1992 to 1994 and 1999 to 2000 are discussed in the Section 5.12 of the 2001 WGNEPH report (ICES, 2001). In 2001 and 2002 catch and landings sampling provided catch compositions to help estimate the LFDs for the missing years. Quarterly discard distributions for the years 1995 to 1999 were estimated by using the discard LFDs for the two preceding and the two following years. Trial XSAs using these data were attempted at the 2003 WGNEPH. In the absence of routine methods of direct age determination in Nephrops, age compositions of removals were inferred from length compositions by means of slicing. This procedure, introduced at the 1991 WG, uses von Bertalanffy growth parameters to determine length boundaries between age classes. All animals in length classes between boundaries are assigned deterministically to the same age class. The method was implemented in the L2AGE programme which automatically generated the VPA input files. The programme was modified in 1992 to accommodate the two-stage growth pattern of female Nephrops (ICES, 1992) and again in 2001 to separate true as opposed to nominal age classes (ICES, 2001). The age classes are true to the extent that the first slicing boundary, i.e. lower length boundary for age 0, is the length-atage zero rather than the lowest length in the data. This was to ensure comparability of age classes across stocks. B.2. Biological Mean weights-at-age for this stock are estimated from studies by Bailey and Chapman, A natural mortality rate of 0.3 was assumed for all age classes and years for males and immature females, with a value of 0.2 for mature females. The lower value for mature females reflects the reduced burrow emergence while ovigerous and hence an assumed reduction in predation. The time-invariant values used for proportion mature-at-age are: males age 1+: 100%; females age 1: 0%; age 2+: 100%. The source of these values is not known. Proportion of F and M prior to spawning was specified as zero to give estimates of spawning stock biomass at January 1. In the absence of independent estimates, the mean weights-at-age in the total catch were assumed to represent the mean weights in the stock. B.3. Surveys Abundance indices are available since 2008 onwards. The first ever trials of the UWTV surveys in this ground were done in the 1990s (1997 and 1998) but with limited success due to weather conditions. Was only in 2007 where the UWTV surveys started to be conducted regularly. These surveys occur annually in August in a combined survey with the input of Northern Ireland, Ireland

157 ICES WGCSE REPORT and England. This survey covers both West and East Irish Sea (FU15 and FU14) and is done onboard the northern Irish research vessel Corystes. In the east Irish Sea (FU14) the survey is of a fixed grid design (3.4 x 3.4 nm) and is carried out using the same protocols used in UWTV surveys in the western Irish Sea. This FU is composed by a main ground, east of the Isle of Man, and another stretch in Wigtown Bay (North of Isle of Man). The survey design is documented at WKNEPHTV (ICES, 2007). In 2007 poor visibility hampered the survey and despite repeated attempts at over 15 stations, turbidity scores precluded the use of some of the counts. Following reanalysis in 2010 these data were considered too unreliable and no abundance estimates were derived for this year. The algorithm used to determine the distance towed on each station changed in the WG of GPS measurements are recorded at one second intervals during each tow. Prior to 2011 the distance towed was determined by summing up the distance travelled between each positional record. As the GPS transceiver is mounted high up on the research vessel, the positional data generated will be influenced by the seastate far more than the sledge. Close examination of the GPS points showed that rolling of the vessel was recorded and this motion is not transmitted to the sledge. In order to reduce the influence of ship-motion on the sledge distance, a smooth spline model of position was fitted to each tow with sufficient flexibility to capture large, slow movements whilst capable of smoothing through the short frequency movement cause by wave action. The previous practice of determining distance travelled by summing up the distance between each recorded ping appears to have significantly over-estimate the distance travelled (typically +30%) which translated into a reduced density of burrows. In 2012 there was an historical revision of burrow density estimates from the TV survey ( series) using a more accurate field of view (0.75 m) and a bias of 1.2. The new estimates show a decrease of around 10% in abundance compared with last year s estimations for the dataseries. Since 2011, the procedure to estimate Nephrops abundance uses a geostatistical approach and runs using R statistical package with the gstat, maptools, and spatstat libraries. The former approach used calculated the mean density of non-zero counts which was raised to the total fished area. This approach ignored the spatial distribution of the counts and was highly sensitive to the total area used for raising. The geostatistical procedure takes the spatial position of the burrow density estimates and fits a semi-variogram model to describe the how variance changes with distance. The process is described below. A boundary file is created using the VMS and BGS sediment data on the MapInfo GIS system and is used to delimit the boundaries of the kriged map ( km 2 ). Mean density per station and the geographical coordinates (transformed from latitude and longitude into metres) are first fitted with a variogram model. The following commands are used to fit the variogram (the data is held in dataframe recounts7 ). gstat.recount <- gstat(id="burrowdensity",formula=burrowdensity~1, locations=~lon.m+lat.m, data=recounts7) vario.recount <- variogram(burrowdensity~1, locations=~lon.m+lat.m, data=recounts7) fit.vario.recount <- fit.variogram(vario.recount, model=vgm(0.1, "Exp", 15000, 0.03))

158 1528 ICES WGCSE REPORT 2013 plot(vario.recount, fit.vario.recount) A Kriged estimate of density is then produced for a 500*500m grid of points lying inside the boundary with the following code. coordinates(recounts7)=~lon.m+lat.m #and the grid we're going to produce pred.lat <- seq(from=y.range[1], to=y.range[2], by=500) pred.lon <- seq(from=x.range[1], to=x.range[2], by=500) recount.grid <- data.frame(lat.m=rep(pred.lat, each=length(pred.lon)), lon.m=rep(pred.lon, times=length(pred.lat))) pos <- point.in.polygon(recount.grid$lon.m, recount.grid$lat.m, boundary$dist.lon, boundary$dist.lat) recount.grid <- recount.grid[pos>0,] gridded(recount.grid)=~lon.m+lat.m coordinates(boundary)=~dist.lon+dist.lat #krig it krige.recount <- krige(burrowdensity~1, recounts7, recount.grid, model=fit.vario.recount) res <- (sum(krige.recount$var1.pred*250000)/ ) /bias# each cell represents a 500m*500m block = sq m, divide by 1million to get the index in millions. By bootstrapping the recount data with replacement it is possible to estimate the uncertainty on the survey abundance estimate. Typically this comes out at a ~10% confidence interval. UWTV relative to absolute conversion factors A number of factors are suspected to influence the ability of the surveys to map directly to absolute abundance (Table 1). Table 1. Absolute abundance conversion factors. TIME PERIOD EDGE DETECTION SPECIES OCCUPANCY CUMULATIVE ABSOLUTE EFFECT RATE IDENTIFICATION CONVERSION FACTOR <= Edge effect: Same sledge and set up as Western Irish Sea. Larger burrows systems increase the edge effect. Detection rate: Same sledge and set up as Western Irish Sea and same staff so detection rate maintained. Species identification: Factor kept the same as Wastern Irish Sea; less Calocaris spp but Goneplax spp. are prevalent across the ground.

159 ICES WGCSE REPORT B.4. Commercial cpue B.5. Other relevant data C. Assessment: data and method Advice Generation Protocol 1 ) Survey indices are worked up annually resulting in the TV index. 2 ) Apply the Absolute Conversion Factors (see Section B3, Table 1). The combined effect of these factors is to be applied to the new survey index. 3 ) Generate mean weight in landings. Check the time-series of mean landing weights for evidence of a trend in the most recent period. If there is no firm evidence of a recent trend in mean weight use an average taken over an appropriate time scale. If, however, there is strong evidence of a recent trend then apply most recent value (don t attempt to extrapolate the trend further in the future). 4 ) The catch option table will include the harvest ratios associated with fishing at F0.1, F35%SpR and FMAX. These values were estimated using a LCA model at WGCSE 2010 but may be revised if there indications of changes to fishery or biological factors (Table 2). 5 ) Create catch option table on the basis of a range of harvest ratios ranging from 0 to the maximum observed ratio or the ratio equating to FMAX, whichever is the larger. Insert the harvest ratios from step 4 and also the current harvest ratio. 6 ) Multiply the survey index by the harvest ratios to give the number of total removals. 7 ) Create a landings number by applying the discard ratio. 8 ) Produce landings biomass by applying mean weight. D. Short-term projection Not available. E. Medium-term projections Not available. F. Long-term projections Not available. G. Biological reference points FMSY proxies have been developed for FU14. However, MSY Btrigger is not defined. The current sampling levels are considered too low for reliable length frequency determination and the time-series of abundance estimates is too short. Sampling needs to be intensified for this stock to improve assessment and management advice.

160 1530 ICES WGCSE REPORT 2013 TYPE VALUE TECHNICAL BASIS MSY MSY Btrigger Not defined Short time-series of abundance estimates Approach FMSY 9.8% harvest ratio Equivalent to F0.1 for combined sexes Table 2. LCA model outputs calculated at WGCSE These are used in the catch option table. FBAR 20-40MM HARVEST RATES SPR Female Male Female Male Combined % 44.60% 42.60% F0.1 Female % 43.50% 41.40% Male % 45.30% 43.30% Combined % 35.90% 33.40% F35%Spr Female % 34.70% 32.20% Male % 37.10% 34.60% Combined % 28.90% 26.20% FMAX Female % 27.30% 24.50% Male % 30.00% 27.20% Biological input parameters PARAMETER VALUE SOURCE Discard Survival 0.00 MALES Growth - K Irish Sea West data ; Bailey and Chapman (1983) Growth - L(inf) 60 " Natural mortality - M 0.3 Brander and Bennett (1986, 1989) Length/weight - a Hossein et al. (1987) Length/weight - b " FEMALES Immature Growth Growth - K Irish Sea West data ; Bailey and Chapman (1983) Growth - L(inf) 60 " Natural mortality - M 0.3 Brander and Bennett (1986, 1989) Size at maturity 24 Briggs (1988) Mature Growth Growth - K Irish Sea West data ; Bailey and Chapman (1983) Growth - L(inf) 56 " Natural mortality - M 0.2 Brander and Bennett (1986, 1989) Length/weight - a Hossein et al. (1987) Length/weight - b "

161 ICES WGCSE REPORT H. Other issues H.1. Historical overview of previous assessment methods Not relevant for this stock. I. References ICES Report of the Working Group on Nephrops Stocks. ICES CM 2001/ACFM:16. ICES Report of the Working Group on Nephrops Stocks. ICES CM 2003/ACFM:18. ICES Workshop on the Use of UWTV Surveys for Determining Abundance in Nephrops Stocks throughout European Waters April 2007, Heraklion, Crete, Greece. ICES CM 2007/ACFM: pp. ICES Report of the Working Group on the Celtic Seas Ecoregion (WGCSE), May 2010, Copenhagen, Denmark. ICES CM 2010/ACOM: pp.

162 1532 ICES WGCSE REPORT 2013 Stock Annex 6.5: Irish Sea West Nephrops (FU15) Stock Working Group Irish Sea West Nephrops (FU15) WKNEPH 2009 (WKNEPH2009) Date 6 March 2009 A. General A.1. Stock definition Throughout its distribution, Nephrops is limited to muddy habitat, and requires sediment with a silt & clay content of between % to excavate its burrows, and this means that the distribution of suitable sediment defines the species distribution. Adult Nephrops only undertake very small scale movements (a few 100 m) but larval transfer may occur between separate mud patches in some areas. In the western Irish Sea the Nephrops stock inhabits an extensive area of muddy sediment between the Isle of Man and Northern Ireland and its fishery contributes to more than 90% of overall Irish Sea landings. There is little evidence of mixing between the east and west Irish Sea stocks due to the nature of water current movements, which is characterised in the west by a gyre, which has a retention effect on both sediment and larvae. The eastern and western Nephrops stocks are treated as separate populations as they have different population characteristics. A.3. Ecosystem aspects A number of studies have examined Nephrops larvae distribution in order to examine how recruitment may impinge upon the distribution of a catchable (adult) Nephrops population and the maintenance of the population. Hillis (1968) found that although generally the larvae occupied the same areas as the adults, there was some evidence of advective losses to the southeastern part of their range, most probably due to tidal currents (White et al., 1988). More recent studies in the western Irish Sea have uncovered the existence of a seasonal cyclonic gyre which appears to facilitate retention of larvae over the mud patch (Dickey-Collas et al., 1996; Hill et al., 1996; Horsburgh et al., 2000). B. Data B.1. Commercial catch Length and sex compositions of Nephrops landed from the Irish Sea West are estimated from port sampling by Ireland and Northern Ireland and Ireland. A lack of cooperation by the Northern Ireland industry prevented sampling commercial catches over the period The Irish LFDs are therefore raised to the international catch for these years. Northern Ireland sampling resumed in 2008 and these data are combined with those from Ireland for that year. Sample data is used to compute international removals (Landings + dead discards). Landings per unit of effort time-series are available from the following fleets: Northern Ireland Nephrops trawl gears. Landings-at-age and effort data from this fishery since 1986 are used to generate a cpue index. There is also a cpue series since 1995 for a subset of Republic of Ireland Nephrops vessels. Catch-at-age are estimated

163 ICES WGCSE REPORT by raising length sampling of discards and landings to officially recorded landings and slicing into ages (knife-edge slicing using growth parameters). Cpue is estimated using officially recorded effort (hours fished). Discard sampling commenced in the mid-1980s by Northern Ireland and the Republic of Ireland. There is no account taken of any technological creep in the fleet. B.2. Biological Mean weights-at-length for this stock are estimated from studies by Pope and Thomas (1955). A natural mortality rate of 0.3 was assumed for males and immature females, with a value of 0.2 for mature females. The lower value for mature females reflects the reduced burrow emergence while ovigerous and hence an assumed reduction in predation. Maturity for females is taken as 22.1 mm carapace length (McQuaid et al., 2006). Proportion of F and M prior to spawning was specified as zero to give estimates of spawning stock biomass at January 1. In the absence of independent estimates, the mean weights-at-age in the total catch were assumed to represent the mean weights in the stock. B.3. Surveys Ireland and Northern Ireland jointly carry out underwater television (UWTV) surveys on the main Nephrops grounds in the western Irish Sea (Figure 1) since These surveys are based on a randomised fixed grid design. The methods used during the survey are similar to those employed for UWTV surveys of Nephrops stocks elsewhere and are detailed in WKNEPHTV, 2007 and WKNEPHBID, Northern Ireland have carried out a spring (April) and summer (August) Nephrops trawl surveys since These surveys provide data on catch rates and length frequency distributions from of stations throughout in the western Irish Sea. These surveys generate data on Nephrops size composition, mean size, maturity and sex ratio. A number of factors are suspected to contribute bias to the UWTV surveys. In order to use the survey abundance estimate as an absolute it is necessary to correct for these potential biases. The history of bias estimates are given in the following table and are based on simulation models, preliminary experimentation and expert opinion, the biases associated with the estimates of Nephrops abundance in the Irish Sea West are: Edge detection species Cumulative Time period effect rate identification occupancy bias FU15: Irish Sea West <= B.4. Commercial cpue B.5. Other relevant data Table 1 is a summary of available data along with an assessment of its reliability. Table 2 is a summary of assessment parameters. C. Historical stock development 1 ) Survey indices are worked up annually resulting in the TV index.

164 1534 ICES WGCSE REPORT ) Adjust index for bias (see Section B3). The combined effect of these biases is to be applied to the new survey index. 3 ) Generate mean weight in landings. Check the time-series of mean landing weights for evidence of a trend in the most recent period. If there is no firm evidence of a recent trend in mean weight use the average of the three most recent years. If, however, there is strong evidence of a recent trend then apply most recent value (don t attempt to extrapolate the trend further in the future). D. Short-term projection 1 ) The catch option table will include the harvest ratios associated with fishing at F0.1 and FMAX. These values have been estimated by the benchmark workshop (see Section 9.2) and are to be revisited by subsequent benchmark groups. The values are FU specific and have been put in the stock annexes. 2 ) Create catch option table on the basis of a range of harvest ratios ranging from 0 to the maximum observed ratio or the ratio equating to FMAX, whichever is the larger. Insert the harvest ratios from step 4 and also the current harvest ratio. 3 ) Multiply the survey index by the harvest ratios to give the number of total removals. 4 ) Create a landings number by applying a discard factor. This conversion factor has been estimated by the benchmark workshop and is to be revisited at subsequent benchmark groups. The value is FU specific and has been put in the stock annex. 5 ) Produce landings biomass by applying mean weight. The suggested catch option table format is as follows. Implied fishery Harvest rate Survey Index Retained number Landings (tonnes) 0% % " % " % " % " F % " % " % " Fmax 13.50% " % " % " % " % " % " Fcurrent 21.5% "

165 ICES WGCSE REPORT E. Medium-term projections None presented. F. Long-term projections None presented. G. Biological reference points Harvest ratios equating to fishing at F0.1 and FMAX were calculated in WKNeph (2009). These calculations assume that the TV survey has a knife-edge selectivity at 17 mm and that the supplied length frequencies represented the population in equilibrium. F0.1 = 10.9% FMAX= 20.2% I. References Dickey-Collas, M., Gowen, R.J. and Fox, C.J Distribution of larval and juvenile fish in the western Irish Sea: Relationship to phytoplankton, zooplankton biomass and recurrent physical features. Marine and Freshwater Research 47: Horsburgh, K.J., Hill, A.E., Brown, J., Fernand, L., Garvine, R.W., Angelico, M.M.P Seasonal evolution of the cold pool gyre in the western Irish Sea. Progress in Oceanography 46: Hill, A.E., Brown, J. and Fernand, L The western Irish Sea gyre: a retention mechanism for the Norway Lobster (Nephrops norvegicus) Oceanologica Acta 19: Hillis, J.P Larval distribution of Nephrops norvegicus (L.) in the Irish Sea and North Channel. ICRES C.M Doc. No. K6. (Mimeo). McQuaid, N., Briggs, R.P.and Roberts, D Estimation of the size of onset of sexual maturity in Nephrops norvegicus (L.). Fisheries Research. White, R.G., Hill, A.E. and Jones, D.A Distribution of Nephrops norvegicus (L.) larvae in the western Irish Sea: an example of advective control on recruitment. Journal of Plankton Research 10(4):

166 1536 ICES WGCSE REPORT 2013 Table 1. Summary table of available data. FU15 Irish Sea West: Data Available Data Commercial Data pre Landings Effort cpue/lpue Mean size Sex ratio LFDs Catch Landings Discards Survey Data Trawl surveys Catch rate mean size LFDs Sex ratio Camera Surveys Density estimate Data Quality Poor Acceptable Reliable

167 ICES WGCSE REPORT Table 2. Biological input parameters. Parameter Value Source Discard Survival 0.10 ICES (1991a) Discard rate 40.2% 2007 discard sampling. MALES Growth - K Hillis (1979) ; ICES (1991a) Growth - L(inf) 60 " Natural mortality - M 0.3 Brander and Bennett (1986, 1989) Length/weight - a After Pope and Thomas (1955) (data for Scottish stocks) Length/weight - b " FEMALES Immature Growth Growth - K Hillis (1979) ; ICES (1991a) Growth - L(inf) 60 " Natural mortality - M 0.3 Brander and Bennett (1986, 1989) Size at maturity 22.1 McQuaid et al., 2006 Mature Growth Growth - K Hillis (1979) ; ICES (1991a) Growth - L(inf) 56 " Natural mortality - M 0.2 Brander and Bennett (1986, 1989) Length/weight - a After Pope and Thomas (1955) (data for Scottish stocks) Length/weight - b "

168 1538 ICES WGCSE REPORT N E3 E4 E5 54.6N N 54.2N 54.0N 53.8N N N W 6.2W 6.0W 5.8W 5.6W 5.4W 5.2W 5.0W 4.8W 4.6W Figure 1: Western Irish Sea Nephrops stations

169 ICES WGCSE REPORT Stock Annex 6.6: Whiting VIIa Stock Working Group Irish Sea Whiting (Division VIIa) Assessment of Northern Shelf Demersal Stocks Last updated WGCSE 2011 Updates Inclusion of Fishery Data from Ireland A. General A.1. Stock definition Whiting in Division VIIa are considered a single stock for management purposes. In 2004 an informal meeting was established to review current knowledge of the distribution, movements and stock structure of whiting in the Irish Sea, and linkages between whiting in the Irish Sea and surrounding management areas. Information on egg and larval, tagging, survey studies was presented as a working document (WD10) in WGNSDS, The results of this are synopsized below: UK egg and larva surveys have shown that whiting spawn in spring throughout the eastern Irish Sea and in the coastal waters of the western Irish Sea. This is supported by the distribution of actively spawning fish caught during trawl surveys in March. Transport of whiting eggs, larvae or pelagic pre-recruits from Celtic Sea spawning grounds into the Irish Sea is likely to be impeded by the Celtic Sea thermal front that becomes increasingly established from spring onwards. Whiting recruitment grounds are in the same general area as the spawning grounds, and young whiting are widespread in the coastal bights of the Irish Sea. The gyre system that becomes established from late spring onwards in the western Irish Sea appears important in retaining larvae and pelagic pre-recruits of whiting, as shown by the results of frametrawl surveys of pelagic pre-recruits in the western Irish Sea. As the whiting become demersal from late summer onwards, they are found throughout the western Irish Sea although densities appear highest around the periphery of the mud patch in coastal waters and along the southern boundary between Ireland and the Isle of Man. This pattern is also noted by fishermen operating in this area. Densities of young whiting in the eastern Irish Sea appear highest off Cumbria and the Solway Firth in autumn, but are more widespread in spring. Tagging studies in the late 1950s show some seasonal dispersal of whiting from the Irish Coast to as far as the Clyde, Liverpool Bay and the Celtic Sea, with evidence of return migrations. Whiting tagged in these studies ranged from about cm, averaging around 30 cm. Whiting recaptured well away from the tagging sites off County Down in the western Irish Sea tended to be several cm larger, on average, than the tagged whiting. Both the western Irish Sea and the Clyde have historically been characterised bycatches of immature and first-maturing whiting, whilst the eastern Irish Sea has a broader agerange of whiting. This pattern persists to the present day.

170 1540 ICES WGCSE REPORT 2013 The evidence for interchange of whiting between the western Irish Sea and other areas within the Irish Sea precludes treating different areas within the Irish Sea as containing functionally separate stocks. Spatial modelling of the populations would require information on rates of dispersal between areas. Trawl surveys continue to show that juvenile whiting are very abundant in the coastal waters of the Irish Sea, and that whiting are one of the most abundant fish species taken in the surveys. Hence, there have been no indications of depressed recruitment associated with the apparent steep decline in abundance of large whiting. Length at 50% maturity in female whiting is only cm in the Irish Sea and neighbouring management areas, and spawning appears predominantly by young whiting of 1 3 years old. A.2. The fishery Most landings by the Irish and UK (NI) fleet, which take the bulk of the Division VIIa whiting catch, are from the western Irish Sea (ICES CM 2003/ACFM:04) and are made predominately by single- and twin-rig trawlers. A small number of UK pair trawlers also fish for whiting. The UK (E&W) fleet has declined substantially over time, and the bulk of its landings are from inshore otter trawlers targeting mixed flatfish and roundfish in the eastern Irish Sea. Discarding in this stock is thought to be high in all fleets, particularly in the Nephrops fishery. The Nephrops directed fishery operates on the main whiting nursery areas in the western Irish Sea, and is particularly intensive in the summer months. The mesh size mainly in use in the fishery is 70 mm in single trawls and 80 mm in twin trawls targeting Nephrops. The western Irish Sea fishery for whiting has declined substantially in recent years, and the increase in abundance of haddock has resulted in few vessels targeting whiting. Vessels operating with 70 mm and 80 mm mesh are required to use square mesh panels. Square mesh panels were introduced as a technical measure to reduce fishing mortality on whiting. Square mesh panels have been mandatory for all UK trawlers (excluding beam trawlers) in the Irish Sea since 1993 and for Irish trawlers since While the effects of this technical measure have not been formally evaluated, the Nephrops fishery still generates substantial quantities of whiting discards. Effort by Irish Nephrops trawlers in the main areas of whiting bycatch has shown some reduction during the period of the Irish Sea cod recovery plan closures. However, the summer peak in activity of the Nephrops fishery was not affected by the recovery plans. As the activities of the Nephrops fleet were not restricted by the cod recovery plan, it is unlikely that the recovery plan was effective in reducing levels of discarding in this stock. There has been some recent decommissioning of vessels in the Irish Sea. Ireland introduced a further decommissioning scheme in 2008 with the aim of removing GT from the fleet register. This is targeted at vessels over 10 years of age and >18 m in length. The reported landings of whiting in by UK vessels decommissioned in 2002 amounted to about 7% of the total international landings of whiting in those years. Whilst few new Irish vessels have joined the fishery, some vessels from County Donegal have reported catches of whiting in VIIa. These vessels have been attracted into the Celtic Sea fishery in recent years in response to poor catches in other areas. Irish landings of whiting in the southwestern part of VIIa now contribute the bulk of the total Irish landings in the Division (ICES CM 2003/ACFM:04). The difference in grounds in the southern part of

171 ICES WGCSE REPORT VIIa means that whiting in the area are more likely to function as part of the Celtic Sea stock rather than the Irish Sea stock. Irish otter board trawlers fishing ICES area VIIa generally use twin-rig gear to fish for Nephrops. However there are also localized mixed fisheries both in the north and south ends of VIIa. The Irish Sea Nephrops fleet is highly opportunistic and of this fleet, there are only a handful of boats that fish the Irish Sea Prawn Grounds 100% of the time. The rest of the fleet divides its time between the Irish Sea, Smalls, Aran and Porcupine Grounds dependant on tides, weather and market forces. In late 2009 and 2010, a number of Irish vessels operating within the Irish Sea Nephrops fishery incorporated a Swedish grid into otter trawls, as part of the cod long-term management plan. It is expected that this will reduce the whiting catches of these vessels by ~60% in weight. Furthermore, a small number of vessels began utilizing an inclined separator panel expected to reduce whiting catch by ~75% in weight (STECF, 2010). Preliminary Irish discard data shows a reduction in 45% by number of whiting on boats using these gear technology mitigation measures. In recent times, Nephrops landings from the Smalls grounds (VIIg) have surpassed those from the Irish Sea grounds. This reflects the increasing amount of effort by east coast vessels in 7g where in general, better prices are obtained for their catch. The main species targeted by the otter trawl fleet are Nephrops, cod, ray, haddock, anglerfish and whiting. The Irish beam-trawl fleet predominantly targets black sole and other high-quality flatfish and divides its effort between VIIa and VIIg depending on weather, tides and market forces. For the UK NI fleet decommissioning at the end of 2003 removed 19 out of 237 UK vessels that operated in the Irish Sea, representing a loss of 8% of the fleet by number and 9.3% by tonnage. Of these vessels, 13 were vessels that used demersal trawls with mesh size >=100 mm. The previous round of decommissioning in 2001 removed 29 UK(NI) Nephrops and whitefish vessels and four UK(E&W) vessels registered in Irish Sea ports at the end of Of these, 13 were vessels that used demersal trawls with mesh size >=100 mm. A.3. Ecosystem aspects Recruitment in Irish Sea whiting appears less variable than in cod and haddock, although there is some similarity in the timing of strong and weak year classes that may indicate a similar response to changes in environmental conditions affecting spawning or earlystage survival. The diet of Irish Sea whiting has been examined in some detail since the 1970s using samples collected from research vessels. Cannibalism occurs in adult whiting; however the effect of this on the assessment of the stock has not yet been investigated. Young whiting are common in the diets of larger predators such as cod and anglerfish.

172 1542 ICES WGCSE REPORT 2013 B. Data B.1. Commercial catch B1.1. Landings The following table gives the source of landings data for Irish Sea whiting: Kind of data Country Caton (catchin-weight) Canum (catchat-age in numbers) Weca (weight-atage in the catch) Matprop (proportion mature-by-age) Length composition-incatch UK(NI) UK(E&W) UK(Scotland) UK (IOM) Ireland France Belgium Netherlands X X X X X X X X X X X X X X X X X X X Quarterly landings and length/age composition data are supplied from databases maintained by national Government Departments and research agencies. These figures may be adjusted by national scientists to correct for known or estimated misreporting by area or species. Data are supplied on paper or Excel files to a stock coordinator nominated by the ICES Northern Shelf Demersal Working Group, who compiles the international landings and catch-at-age data, and maintains a time series of such data with any amendments. To avoid double counting of landings data, each UK region supplies data for UK landings into its regional ports, and landings by its fleet into non-uk ports. The UK(E&W) currently supplies raised quarterly length frequencies of landings but only sporadic age data. The catch and mean weight-at-age are estimated using combined UK(NI) and Irish quarterly length weight relationships and age length keys. Quarterly landings are provided by the UK(Scotland), Belgium and France and annual landings are provided by UK(IOM). The quarterly estimates of landings-at-age into UK(E&W), UK(NI) and Ireland are raised to include landings by France, Belgium, UK(Scotland), UK(IOM) (distributed proportionately over quarters), and then summed over quarters to produce the annual landings-at-age. The Excel spreadsheet files used for age distribution, adjustments and aggregations can be found with the stock co-ordinator and for the current and previous year in the ICES computer system under w:\acfm\wgnsds\year\personal\name (of stock co-ordinator). The result files (FAD data) can be found at ICES and with the stock coordinator, as ASCII files on the Lowestoft format, under w:\acfm\wgnsds\year\data\whg_7a. B1.2. Discards The Irish Sea Nephrops fishery takes place on the whiting nursery grounds of the northwestern Irish Sea and has traditionally produced high whiting discarding. The quantity

173 ICES WGCSE REPORT of whiting discarded from the UK (NI) Nephrops fishery in 2002 was estimated on a quarterly basis from samples of discards and total catch provided by skippers. The discards samples contain the heads of Nephrops tailed at sea. Using a length weight relationship, the live weight of Nephrops that would have been landed as tails only is calculated from the carapace lengths of the discarded heads. The number of whiting in the discard samples is summed over all samples in a quarter and expressed as a ratio of the summed live weight of Nephrops in the discard samples (i.e. those represented as heads only in the samples). The reported live weight of Nephrops landed as tails only is then used to estimate the quantity of whiting discarded using the whiting:nephrops ratio in the discard samples. The length frequency of whiting in the discard samples is then raised to the fleet estimate, and numbers and mean weight-at-age of discarded whiting is computed from the age length key and length weight parameters for whiting. The UK(NI) estimates are available since 1980 but the reliability of these estimates has not been determined. Roughly 40 discard samples are collected annually. There are several limitations to these data: only a small subset of single-rig trawlers is sampled; the method of raising to the fleet discards will be affected by any inaccuracies in the reported landings of Nephrops; and there are no estimates of landings of whiting from these vessels with which to calculate proportions discarded-at-age. However, the WG has used these data in past assessments because removal of discards data would remove a large fraction of catch from the assessment. A re-analysis of the Irish discard data raised to the Nephrops landings produced estimates of discards from the Irish Nephrops fleet that were more consistent with those of the UK(NI) Nephrops fleet. However, this method of raising could not be used to recalculate an entire time-series of discard estimates from the Irish Nephrops fleet. The quarterly UK(NI) discard ratios were therefore used by the working group to estimate the tonnage discarded from the Irish Nephrops fishery. Length frequencies and age length keys from the whiting discarded by the Irish Nephrops fleet are used to estimate the numbers discarded-at-age from the Irish Nephrops fleet. At the WGCSE 2011 Irish discard estimates ( ) raised according to the methods described in Borges et al., 2005 were available to the working group See Table 1.0. These are available in the ICES files. Discard rates in this series were variable compared with previous estimates based on the UK NI self-sampling scheme. Given the differences in raising procedure applied to the NI Discard estimates and the Irish discard estimates further examination of the discard data is needed before international estimates of discard numbers-at-age can be made. The working group therefore has not estimate international discard volumes and numbers-at-age since B.2. Biological Natural mortality was assumed to be constant (M=0.2, applied annually) for the whole range of ages and years. A combined sex maturity is assumed, knife-edged at age 2. The use of a knife edged maturity ogive has been a source of criticism in previous assessments. However, recent research on gadoid maturity conducted by the UK(NI) gives no evidence for substantial change in whiting maturity since the 1950s, although there has been an increase in the incidence of precocious maturity at-age 1, particularly in males, since 1998.

174 1544 ICES WGCSE REPORT 2013 As in previous years, SSB is computed at the start of each year, and the proportions of M and F before spawning were set to zero. Stock weights are calculated using a procedure first described in the 1998 working group report. To derive representative stock weights for the start of the year for year i and age j the following formula is adopted: (CW i,j + CW i+1,j+1)/2 = SW at start of year. These values are then smoothed using a three year moving average. Recent investigations into the biological parameters (maturity, sex and growth parameters) of whiting in VIIa (funded under the Data Directive Regulation (1639/2001)) took place during a Biological Sampling survey (BBS) in March Parameter estimates of maturity-at-length indicate the L50 for whiting in VIIa for males and females is cm and cm, respectively. Maturity-at-age for both sexes are similar for most stock area (VIIa, b, j and g) with the notable exception of age 1 males in the Celtic Sea where the estimates are outside the 95% CI bounds for VIIa and considerably lower than VIa. In most areas whiting were mature by age three and most were mature at age 2. The sex ratio for whiting tended to increase with length for nearly all the age classes in all areas indicating that females tend to have larger length-at-age than males (Gerritsen, 2005). Gerritsen et al., 2002 describes the relationships between maturity, length and age of whiting sampled on a length-stratified basis from NI groundfish surveys of the Irish Sea during spawning in spring Findings show that most one year old females were immature whilst most two year old females were mature; almost all 3 year olds of both sexes were mature. Length at 50 maturity average around 19 cm in males and 22 cm in females. B.3. Surveys Seven research vessel survey-series for whiting in VIIa were available to the working group in In all surveys listed the highest age represents a true age not a plus group. UK(England and Wales) Beam Trawl Survey (UK (E&W)-BTS-Q3): ages 0 and 1, years : The survey covers the entire Irish Sea and is conducted in September on the R.V. Corystes. The survey uses a 4 m beam trawl targeted at flatfish. The survey is stratified by area and depth band, although the survey indices are calculated from the total survey catch without accounting for stratification. Numbers of whiting at-age per km towed are provided for prime stations only (i.e. those fished in most surveys). UK(Northern Ireland) October Groundfish Survey (NIGFS-WIBTS-Q4-EAST & WEST): ages 0 5, years : The survey-series commenced in its present form in It comprises 45 three mile tows at fixed station positions in the northern Irish Sea, with an additional 12 one mile tows at fixed station positions in the St George s channel from October 2001 (the latter are not included in the tuning data). The surveys are carried out using a rock-hopper otter trawl deployed from the R.V. Lough Foyle. The survey designs are stratified by depth and seabed type. The mean numbers-at-length per three mile tow are calculated separately by stratum, and weighted by surface area of the strata to give a weighted mean for the survey or group of strata. The strata are grouped

175 ICES WGCSE REPORT into western Irish Sea and eastern Irish Sea, and a separate age length key is derived for each area to calculate abundance indices by age class. The survey design and time-series of results including distribution patterns of whiting are described in detail in Armstrong et al., UK(Northern Ireland) March Groundfish Survey (NIGFS-WIBTS-Q1-EAST & WEST): ages 1 5, years : Description as for NIGFS-WIBTS-Q4-EAST & WEST above. UK(Northern Ireland) Methot Isaacs Kidd Survey (NIMIK): age 0, years : The survey uses a Methot Isaacs Kidd frame trawl to target pelagic juvenile gadoids in the western Irish Sea at stations. The survey is stratified and takes place in June during the period prior to settlement of gadoid juveniles. Indices are calculated as the arithmetic mean of the numbers-perunit sea area. Ireland s Irish Sea Celtic Sea Groundfish Survey (IR-ISCSGFS): ages 0 5, years : This survey commenced in 1997 and is conducted in October November on the R.V. Celtic Voyager. The α and β of the series are set to account for the variable timing of this survey within the fourth quarter. The survey uses a GOV otter trawl with standard ground gear and a 20 mm codend liner. The survey operates mainly in the western Irish Sea but has included some stations in the eastern Irish Sea. The survey design has evolved over time and has different spatial coverage in different years. Indices are calculated as arithmetic means of all stations, without stratification by area. UK(Scotland) groundfish survey in Spring (ScoGFS-spring): ages 1 8, years : This survey represents an extension of the Scottish West Coast groundfish survey (Area VI), using the research vessel Scotia. The survey gear is a GOV trawl, and the design is two fixed-position stations per ICES rectangle from 1997 onwards (17 stations) and one station per rectangle in 1996 (nine stations). The survey extends from the Northern limit of the Irish Sea to around 53 o 30. UK(Scotland) groundfish survey in Autumn (ScoGFS-autumn): ages 0 5, years : The survey covers a similar area to the ScoGFS in spring, but has only stations. IRGFS (Ireland) (IR-Q4 IBTS): This survey commenced in 2003 aboard the R.V. Celtic Explorer. It is a depth stratified survey using a GOV trawl with a 20 mm meshliner on the codend. The survey currently covers VIIb, j, g and VIa. Prototcols for the survey are governed by the International Bottom Trawl Survey Working Group (IBTS). To allow the inclusion of the IR-Q4 IBTS and ScoGFS-Spring surveys for the year after the last year with commercial catch data in an XSA, the surveys may be treated as if they took place at the end of the previous year, and the age range and year range of the surveys may be shifted back accordingly in the data files. The following research surveys were available to the 2011 Working group: UK(NI) groundfish survey: March

176 1546 ICES WGCSE REPORT 2013 UK(NI) groundfish survey: October UK(Scotland) groundfish survey: March UK(Scotland) groundfish survey: autumn Irish groundfish survey: autumn 2003 and UK(NI) MIK net surveys of pelagic-stage 0-group cod, western Irish Sea UK(E&W) beam trawl survey: 0-1 gp cod, FSP surveys of Irish Sea round fish: Further details of the tuning data are given in Appendix 1 and 2 of the 1999 WG Report. B.4. Commercial cpue No cpue data have been provided for the French (Lorient) trawl fleet since Four commercial catch effort dataseries were available to the WG: Irish otter trawl (IR-OTB): ages 1 6, years : Effort and cpue data provided for the Irish fleet comprise total annual effort (hours fished, not corrected for fishing power) and total numbers-at-age in landings from otter trawlers. The data were revised to take account of updated logbook information. This fleet operates mainly in the western Irish Sea, targeting Nephrops and/or whitefish. The distribution of fishing is concentrated in the western part of the range of the whiting stock in the Irish Sea. Hence the catch rates will represent changes in abundance of whiting in the western part of VIIa. The use of this fleet as a tuning index therefore relies on the assumption that trends in abundance in the west of VIIa reflect those of the entire stock. The catch-at-age data comprise a large proportion of the total international catch. Hence, some correlation of errors can be expected between the tuning dataset and the catch-atage data. The effect of such correlations has not been evaluated. The ottertrawl catch-at-age data contained data for landings only. Hence the reliability of the tuning fleet will be limited for age groups which are heavily discarded. UK(Northern Ireland) pelagic trawl: ages 2 6, years : The pelagic trawl catch-at-age data contained data for landings only. Hence the reliability of the tuning fleet will be limited for age groups which are heavily discarded. This fleet currently targets haddock and cod in the deeper waters of the western Irish Sea and the North Channel. Bycatches of whiting are currently very small and are heavily discarded due to their low value. The fleet is considered unsuitable for indexing whiting abundance. UK(Northern Ireland) single rig otter trawl: ages 0 6, years : This fleet operates mainly in the western Irish Sea. The distribution of fishing does not encompass the entire range of the whiting stock (which surveys suggest is distributed across the Irish Sea). Whiting discards from single-rig trawlers (estimated from fisher self-sampling scheme) are included. UK(England and Wales) otter trawl: ages 2 6, years : Estimates up to and including 2000 of commercial lpue from UK(E&W) otter trawlers contain data for landings only. Hence the reliability of the tuning fleet will be limited for age groups which are heavily discarded. This fleet operates mainly in the

177 ICES WGCSE REPORT eastern Irish Sea. The distribution of fishing does not encompass the entire range of the whiting stock (which surveys suggest is distributed across the Irish Sea) or the main whiting nursery grounds (in the western Irish Sea). Age compositions in most years have been estimated from length frequencies using ALKs that were obtained from sampling of fleets operating mainly in the western Irish Sea. This has introduced additional uncertainties into the data. B.5. Other relevant data None. C. Historical stock development No assessment was carried out for this stock in The last assessment for this stock was a survey based assessment in Model used: XSA (up to 2002) SURBA SURBA SURBA Software used: Lowestoft VPA suite SURBA 2.2 XSA Model Options chosen: Tapered time weighting not applied Catchability independent of stock size for all ages Catchability independent of age for ages >= 4 Survivor estimates shrunk towards the mean F of the final five years or the two oldest ages S.E. of the mean to which the estimate are shrunk = Minimum standard error for population estimates derived from each fleet = Prior weighting not applied

178 1548 ICES WGCSE REPORT 2013 Input data types and characteristics for XSA: Type Name Year range Age range Variable from year to year Yes/No Caton Catch in tonnes 1980 last data year Canum Weca West Mprop Fprop Matprop Catch-at-age in numbers Weight-at-age in the commercial catch Weight-at-age of the stock at spawning time. Proportion of natural mortality before spawning Proportion of fishing mortality before spawning Proportion mature-at-age 1980 last data year 1980 last data year 1980 last data year 1980 last data year 1980 last data year 1980 last data year Natmor Natural mortality 1980 last data year 0 6+ Yes 0 6+ Yes 0 6+ Yes 0 6+ Yes: uses smoothed catch weights adjusted to start of year 0 6+ No: set to 0 for all ages in all years 0 6+ No: set to 0 for all ages in all years 0 6+ No: the same ogive for all years 0 6+ No: set to 0.2 for all ages in all years Tuning data: Type Name Year range Age range Tuning fleet 1 NIGFS-Oct 1992 last data year 0 5 Tuning fleet 2 NIGFS-Mar (adjusted) 1991 (last data year-1) 0 4 Tuning fleet 3 ScoGFS-Spring 1996 last data year 1 5 Tuning fleet 4 UK(E&W) BTS 1988 last data year 0 1 Input data types and characteristics for SURBA 2.2 Default settings used Tuning data: Type Name Year range Age range Tuning fleet 1 Tuning fleet 2 NIGFS-Oct NIGFS- WIBTS-Q4-EAST & WEST NIGFS-Oct NIGFS- WIBTS-Q1-EAST & WEST ALL ALL

179 ICES WGCSE REPORT D. Short-term projection Currently, there is no analytical assessment for this stock. Model used: Age structured Software used: MFDP prediction with management option table and yield per recruit routines. MLA suite (WGFRANSW) used for sensitivity analysis and probability profiles. Initial stock size. Taken from the XSA for age 1 and older. The recruitment at-age 0 in the last data year is estimated as a short-term GM (1992 onwards) because of a reduction in mean recruitment since then. Natural mortality: Set to 0.2 for all ages in all years. Maturity: The same ogive as in the assessment is used for all years. F and M before spawning: Set to 0 for all ages in all years. Weight-at-age in the stock: average stock weights for last three years. Weight-at-age in the catch: Average weight of the three last years. Exploitation pattern: Average of the three last years. Discard F s, which are generated by the Nephrops fleet as there are no discard estimates for other fleets, are held constant while landings F s are varied in the management option table. Intermediate year assumptions: status quo F Stock recruitment model used: None, the short-term geometric mean recruitment-at-age 0 is used. Procedures used for splitting projected catches: F vectors in each of the last three years of the assessment are multiplied by the proportion landed or discarded at-age to give partial Fs for landings and discards. The vectors of partial Fs are then averaged over the last three years to give the forecast values. E. Medium-term projections No medium-term projections are done for this stock due to problems with estimating current F. F. Yield and biomass per recruit/long-term projections Model used: yield and biomass per recruit over a range of F values that may reflect fixed or variable discard F s. Software used: MFY or MLA Selectivity pattern: mean F array from last three years of assessment (to reflect recent selection patterns).

180 1550 ICES WGCSE REPORT 2013 Stock and catch weights-at-age: mean of last three years (weights-at-age have declined as the stock has declined since the 1980s; it is not known if this is an environmental effect on growth that is independent of stock size). Proportion discarded: partial F vectors are the recent average. Maturity: Fixed maturity ogive as used in assessment. G. Biological reference points Precautionary approach reference points have remained unchanged since Bpa is set at 7000 t and is defined as Blim*1.4. Blim is defined as the lowest observed SSB (ACFM, 1999), considered to be 5000 t. There is not considered to be clear evidence of reduced recruitment at the lowest observed SSBs. FPA is set at 0.65 on the technical basis of high probabilities of avoiding Flim and of SSB remaining above BPA in the long term. Flim is defined as 0.95, the fishing mortality estimated to lead to a potential stock collapse. H. Other issues None. I. References Armstrong, M.J., Peel, J., McAliskey, M., McCurdy, W., McCorriston, P. and Briggs, R Survey indices of abundance for cod, haddock and whiting in the Irish Sea (Area VIIaN): Working Document No. 3 submitted to 2003 meeting of the ICES Working Group on Assessement of Northern Shelf Demersal Stocks. 33pp. Borges, L.; Rogan, E. and Officer, R Discarding by the demersal fishery in the waters around Ireland, Fish. Res. (in press). Gerritsen, H Biological parameters for Irish Demersal Stocks in WD5 (WGNSDS, 2005)

181 ICES WGCSE REPORT Table 1.0. Discard estimates raisesd according to the method oulined in Borges et al., Numbers Weight Numbers Weight Numbers Weight Numbers Weight Numbers Weight Numbers Weight Numbers Weight Numbers Weight Age ('000) (kg) ('000) (kg) ('000) (kg) ('000) (kg) ('000) (kg) ('000) (kg) ('000) (kg) ('000) (kg) Total weight (t) Sampling Information Number of Trips Number of Hauls Numbers Weight Numbers Weight Numbers Weight Numbers Weight Numbers Weight Numbers Weight Numbers Weight Numbers Weight Age ('000) (kg) ('000) (kg) ('000) (kg) ('000) (kg) ('000) (kg) ('000) (kg) ('000) (kg) ('000) (kg) Total weight (t) Sampling Information Number of Trips Number of Hauls

182 1552 ICES WGCSE REPORT 2013 Stock Annex 6.7: Irish Sea Plaice Stock Plaice (Division VIIa) Working Group Celtic Seas Ecoregion Date 17th May 2013 By Tim Earl A. General A.1. Stock definition There are three principle spawning areas of plaice in the Irish Sea: one off the Irish coast, another northeast of the Isle of Man towards the Cumbrian coast, and the third off the north Wales coast (Nichols et al., 1993; Fox et al., 1997; Figure A1). Cardigan Bay has also been identified as a spawning ground for plaice in the Irish Sea (Simpson, 1959). The level of mixing between the east and west components of the Irish Sea stock appears small (Dunn and Pawson, 2002). Length-at-age measurements from research surveys as well as anecdotal information from the fishing industry suggests that plaice in the western Irish Sea grow at a much slower rate than those in the eastern Irish Sea. Earlier studies have suggested that the east and west components of the stock are distinct (Brander, 1975; Sideek, 1981). Morphometric differences have been observed between the east and west components of the stock; the 2004 WG indicated that the UK(E&W) beam-trawl survey in September (from 1989) catches plaice off the Irish coast that are smaller at-age than those caught in the eastern Irish Sea. Although considered separate stocks, the stocks of plaice in the Irish Sea and the Celtic Sea do mix during spawning. Tagging studies have indicated a southerly movement of mature fish (or fish maturing for the first time) from the southeast Irish Sea, off North Wales, into the Bristol Channel and Celtic Sea during the spawning season, such that 43% of the new recruits are likely to recruit outside of the Irish Sea (Figure A1). While some of these migrant spawning fish will remain in the Bristol Channel and Celtic Sea, the majority ( 70%) are expected to return to summer feeding grounds in the Irish Sea (Dunn and Pawson, 2002). Very little mixing is considered to occur between the Irish Sea and Channel stocks or between the Irish Sea and North Sea (Pawson, 1995). Nevertheless, time-series of recruitment estimates for all stocks in waters around the UK (Irish Sea, Celtic Sea, western and eastern Channel, North Sea) show a significant level of synchrony (Fox et al., 2000). This could indicate that the stocks are subject to similar large-scale environmental forces and respond similarly to them, or alternatively that there are subpopulations that share a common spawning.

183 ICES WGCSE REPORT Figure A1. Principal substock areas and movements of plaice on the west coast of England and Wales. Percentages are the recaptures rates of tagged plaice <25 cm total length when released, and >26 cm when recaptured in English and Welsh commercial fisheries. Tagging exercises in and were combined based on the assumption that the dispersal patterns of plaice were consistent over time. For each substock, the main feeding area (derived from tag recaptures during April December; light shading), and the main spawning area (derived from tag recaptures during January March, and ichthyoplankton surveys; dark shading) are indicated. The substocks tagged have been coloured green, red and blue. The substocks coloured orange are less well determined, with the feeding area around south-east Ireland unknown. Letters represent return migrations, where A 6%, and B+C 46%. Reproduced from Dunn and Pawson (2002). A.2. Fishery The status and activities of the fishing fleets operating in ICES Subdivision VIIa are described by Pawson et al. (2002). Following the massive decline in effort (hours fished) by otter trawlers targeting demersal fish in the early1990s, the majority of fisheries effort in the Irish Sea is now exerted by otter trawlers fishing for Nephrops in the western Irish Sea followed by beam trawlers targeting sole in the eastern Irish Sea. Only a small proportion of otter trawlers still target cod, haddock, whiting and plaice with bycatch of angler-fish, hake and sole. From 2001, trawlers for demersal fish adopted mesh sizes of mm and other gear modifications depending on the requirements of recent EU technical conservation regulations and national legislation. However, in 2004 the effort exerted by UK trawlers with mesh mm declined to low levels. In 2006, the effort by UK trawlers targeting demersal fish with mesh mm also declined to low levels. Concomitantly, the effort by UK trawlers targeting Nephrops with mesh mm increased to record highs. Square mesh panels have been mandatory for UK otter trawlers since 1993 and for Irish trawlers since 1994, but this will have little effect on plaice catches. Four Irish trawlers for Nephrops have made use of grids since 2009 and reported 75% drop in fish bycatch. Fishing effort in 2009 by the Irish and UK(E&W) otter fleets targeting demersal fish reached historic lows.

184 1554 ICES WGCSE REPORT 2013 Beam trawling increased in the Irish Sea during the late 1980s, with vessels from England and Belgium exploiting sole. This fishery has important bycatch of plaice, rays, brill, turbot and anglerfish. The fishing effort of the Belgium beam-trawl fleet varies according to the catch rates of sole in the Irish Sea relative to the other areas in which the fleet operates. In 2009, effort (hours fished) by the UK(E&W) beam trawl fleet fell to the lowest observed level. A fleet of vessels primarily from Ireland and Northern Ireland take part in a targeted Nephrops fishery using 70 mm mesh nets with 75 mm square mesh panels. This fishery takes a substantial bycatch of whiting, most of which is discarded. Some inshore shrimp beam trawlers occasionally switch to flatfish when shrimp become temporarily unavailable. Other gear types employed in the Irish Sea to catch demersal species are gillnets and tanglenets, notably by inshore boats targeting cod, bass, grey mullet, sole and plaice. The minimum landing size for plaice in the Irish Sea was set in 1980 to 25 cm (Council Regulation (EEC) No 2527/80). This was increased in 1998 to 27 cm (Annex XII of Council Regulation 850/98). Since 2000, a recovery program has been implemented to reduce exploitation of the cod spawning stock in the Irish Sea. In 2002 the European Commission regulations included a prohibition on the use of demersal trawl, enmeshing nets or lines within the main cod spawning area in the northwest Irish Sea between the 14th February and 30th April. Some derogations were permitted for Nephrops trawls and beam trawlers targeting flatfish. A.3. Ecosystem aspects Plaice are preyed upon and consume a variety of species through their life history. However, plaice have not as yet been included in an interactive role in multispecies assessment methods (e.g. ICES WGSAM 2008). Among other prey items, plaice typically consume high proportions of polychaetes and molluscs. Other than statistical correlations between recruitment and temperature (Fox et al., 2000) little is known about the effects of the environment on the stock dynamics of plaice in the Irish Sea. Negative correlations between year-class strength of plaice (in either the Irish Sea, Celtic Sea, Channel and North Sea) and sea surface temperature are generally strongest for the period February June. However, western (North Sea and Channel) and eastern (Irish Sea and Celtic Sea) stocks have been found to respond to different time-scales of temperature variability, which might imply that different mechanisms are operating in these stocks and/or that the Irish Sea and Celtic Sea share common spawning (Fox et al., 2000). B. Data B.1. Commercial catch Landings International landings-at-age data based on quarterly market sampling and annual landings figures are available from Since 1978, quarterly age compositions have typically represented around 80 90% of the total international landings. Table B1 details the derivation of international landings for the period since 1978.

185 ICES WGCSE REPORT Prior to 1983 the stock was assessed on a separate sex basis: the catch numbers of males and females were worked up separately and the numbers of males and females in the stock as estimated from each assessment combined to give a total biomass estimate. Since 1983 a combined sex assessment of the stock has been conducted and the numbers of males and females in the catch have been combined at the international data aggregation level prior to running a single assessment. Data exploration Data exploration for commercial landings data for Irish Sea plaice has involved: 1 ) expressing the total landings-at-age matrix as proportions-at-age, normalized over time, so that year classes making above-average contributions to the landings are shown as large positive residuals (and vice-versa for below average contributions); 2 ) applying a separable VPA model in order to examine the structure of the landed numbers-at-age before they are used in catch-at-age analyses, in particular whether there are large and irregular residuals patterns that would lead to concerns about the way the recorded catch has been processed. Given that discards now represent a larger proportion of the catch than the landings method 1 should be applied to the discard-at-age matrix in addition to the discard-atage matrix and method 2 is unnecessary. Discards In 1986, the UK fleet was restricted to a 10% bycatch of plaice for almost the entire year. Estimates were made of the increased quantity of plaice that would have been discarded based on comparisons of lpue values for with those for The estimated quantity of 250 tonnes was added to the catch. A similar situation arose the following year and 250 tonnes was added to the catch for The 10% plaice bycatch restriction was enforced again in 1988 to all UK(E&W) vessels in the 1st quarter and to beam trawlers in the 2nd and 3rd quarters. However, this time the landings were not corrected for discard estimates. Discard information was not routinely incorporated into the assessment prior to benchmarking by WKFLAT in B.2. Biological Weights-at-age A number of different methodologies have been employed to determine weights-atage for this stock. Stock weights and catch weights-at-age were determined on a separate sex basis and remained unchanged from 1978 until Catch weights were derived from a von Bertalanffy length-at-age fit to Belgian (70 74), UK(E&W) (64 74) and Irish (62 66) catch samples. The estimated lengths-at-age were converted to weights-at-age using a Belgian length weight dataset (ages 2 15 females; 3 9 males). Stock weights were calculated as the mean of adjacent ages from the catch weights, where catch weights represented 1st July values and stock weights 1st January. From 1983 weights-at-age have been calculated on a combined sex basis. Catch weights were taken from market sampling measurements combined on a sex weighted basis and smoothed. For the period 1983 to 1987 catch weights were

186 1556 ICES WGCSE REPORT 2013 smoothed by eye, from 1988 onwards a smooth curve was fitted using a numerical minimisation routine. Stock weights were derived from the smoothed international catch weights-at-age curve with values representing 1st January. In 1985 the stock weights-at-age were adjusted for ages 1 to 4. The difference between the smoothed catch weights and survey (F.V. Silver Star) observations were adjusted using the maturity ogive to give "best estimate" stock weights "for ages where growth and maturity differences can bias sampling procedures". The same procedure was adopted in 1986 (when stock weights in 1982 and 1983 were also revised so as to be consistent with this methodology) and In 1988 however, the Silver Star survey was discontinued and stock weights-at-ages 1 to 3 were calculated as means of the three previous years. Correction of the estimated stock weights of the younger age groups did not occur in 1989 or in subsequent years which explains the sudden increase in weight of the younger age groups for this stock from 1988 onwards. WKFLAT 2011 rejected the use of the polynomial smoother for weights-at-age and suggested that raw annual catch weights are used in future. Raw data back to 1995 was obtained by WKFLAT and used to update the catch weights and stock weights files. Discard weight-at-age were also calculated back to 2004 from UK(E&W) and Belgian data. However, given that the discard weight prior to 2004 were unknown the stock weights file was not updated to include the discard component. This requires further work. Males are smaller than females and mean weight-at-age and mean length-at-age of both sexes has generally declined since the mid-1990s. Commercial data indicate declines in mean weight-at-age of fish age 4 and older since 1995, particularly since 2004 (Figure A2). Survey data indicate that males of ages 1 5 and females of age 1 3 are generally below minimum landing size (MLS, Figure A3). Figure A2. Commercial weight at age data from 1995 (raw, left and standardized, right).

187 ICES WGCSE REPORT Figure A3. Mean length (cm) at-age data from 1993 by sex and area within the Irish Sea: Irish Sea North (ISN), Irish Sea East (ISE), Irish Sea West (ISW), St. George s Channel (SGC). Natural mortality and maturity ogives As for the weights-at-age, natural mortality and maturity was initially determined on a separate sex basis. Natural mortality was taken as 0.15 for males and 0.1 for females. In 1983 when a combined sex assessment was undertaken a sex weighted average value of 0.12 was used as an estimate of natural mortality. This estimate of natural mortality has remained unchanged since The maturity estimates used prior to 1982 are not specified. A new separate sex maturity ogive (Sideek, 1981) was implemented in This ogive was recalculated as sex weighted mean values in 1983 when the assessment was conducted on a combined sex basis. The maturity ogive was revised again in 1992 based on the results of an EU project. Maturity ogives are applied as vectors to all years in the assessment. WKFLAT 2011 was unable to update the maturity ogive due to time restraints. However, preliminary analysis indicated that the ogive may have changed over time, in each sector of the Irish Sea, such that plaice mature at a smaller size and age than previously.

188 1558 ICES WGCSE REPORT 2013 Table A1. Maturity ogives for Irish Sea plaice used in ICES WGs. AGE WG WG WG M F The proportion of fishing mortality and natural mortality before spawning was originally set to 0. It was changed in 1983 to a value of 0.2 on the grounds that approximately 20% of the catch was taken prior to March (considered to be the time of peak spawning activity). As for Celtic Sea plaice the proportion of F and M before spawning was reset to 0, as it was considered that these settings were more robust to changes in the fishing pattern, especially with respect to the medium-term projections. B.3. Surveys In 1993, the UK(E&W) beam-trawl survey-series that began in 1988 was considered to be of sufficient length for inclusion in the assessment. Since 1991, tow duration has been 30 minutes but prior to this it was 15 minutes. In 1997, values for 1988 to 1990 were raised to 30 minute tows. However, data for 1988 and 1989 were of poor quality and gave spurious results: thus, the series was truncated to A similar March beam-trawl survey began in 1993 and was made available to the WG in The March beam-trawl survey ended in 1999 but continued to be used as a tuning index in the assessment until In 2011, the UK(E&W) beam-trawl survey was re-examined and additional stations sampled in the western Irish Sea and St Georges Channel (Cardigan and Caernarfon Bays) since 1993 were included in the index. The extended index replaced the earlier prime stations index since it was considered more representative of the entire stock (WKFLAT 2011). An Irish juvenile plaice survey index was presented to the WG in 2002 ( , ages 2 8). Between 1976 and 1990 this survey had used an average ALK for that period. Serious concerns were expressed regarding the quality of the data for this period and the series was truncated to The stations for this survey are located along the coast of southeast Ireland between Dundalk Bay and Carnsore Point and there was some concern that this localised survey-series would not be representative of the plaice population over the whole of the Irish Sea. Numerous tests were conducted at the 2002 WG to determine the validity of this and other tuning indices and it was concluded that this survey could be used as an index of the plaice population over the whole of the Irish Sea. This survey is no longer used in the assessment. The SSB of plaice can be estimated using the Annual Egg Production Method (AEPM) (Armstrong et al., 2002 and WD9, WGCSE 2011). This method uses a series of ichthyoplankton surveys to quantify the spatial extent and seasonal pattern of egg production, from which the total annual egg production can be derived. The average fecundity (number of eggs spawned per unit body weight) of mature fish is estimated by sampling adult females immediately prior to the spawning season. Dividing the

189 ICES WGCSE REPORT annual egg production by average fecundity gives an estimate of the biomass of mature females. Total SSB can be estimated if the sex ratio is known. Although substantial discrepancies between absolute estimates of SSB from the Annual Egg Production method (AEPM) and the ICES catch-based assessments were observed, they do confirm that SSB of plaice in the Irish Sea is currently at high levels. AEPM estimates of SSB for plaice (RSE = relative standard error, as %), based on production of Stage 1 eggs) are shown below (note 1995 and 2000 estimates were revised in 2010 and 2006 and 2008 estimates revised in 2011 see WD9, WGCSE 2011): Table A3. AEPM estimates of SSB for Irish Sea plaice. All estimates from stratified mean (designbased) estimates. TOTAL WEST EAST Year SSB(t) RSE SSB(t) RSE SSB(t) RSE Splitting the SSB estimate by substrata (Figure A4 below) suggests that the perceived increase in plaice SSB is limited to the eastern Irish Sea. This finding agrees with an analysis of NIGFS-WIBTS data and UK(E&W)-BTS-Q3 by substrata, which also indicate increases in biomass limited to the eastern Irish Sea. SSB estimates from AEPM 18,000 16,000 14,000 12,000 10,000 8,000 6,000 4,000 2, y = x R² = y = x R² = West East Total Figure A4. AEPM estimates by year and substrata. B.4. Commercial lpue Prior to 1981 tuning data were not used in the assessment of this stock. A separable assessment method was used and estimates of terminal S and F were derived iteratively based on an understanding of the recent dynamics of the fishery. In 1981 the choice of terminal F was determined from a regression of exploited stock biomass on cpue. Catch and effort series were available for the UK(E&W) trawl fleet and the Belgian beam-trawl fleet for the period 1964 to In 1994 the Belgian and

190 1560 ICES WGCSE REPORT 2013 UK cpue series were combined to provide one mean standardised international index. The UK(E&W) trawl series was revised in 1986 (details not recorded) and in 1987 was recalculated as an age-based cpue index enabling the use of the hybrid method of tuning an ad hoc VPA. The UK(E&W) trawl tuning-series was revised in 1999 and separate otter-trawl and beam-trawl tuning-series were produced using length samples from each gear type and an all gears ALK. Since the data could only be separated for 1988 onwards the two new tuning-series were slightly reduced in length. In 1996 UK(E&W) commercial effort data were re-scaled to thousands of hours so as to avoid numerical problems associated with low cpue values and in 2000 the UK(E&W) otter-trawl series was recalculated using otter-trawl age compositions only rather than combined fleet age compositions as previously. Two revised survey indices for the Lough Beltra were presented to the WG in 1996 though they were considered too noisy for inclusion in the assessment. They were revised again for the following year and found to be much improved but were again not included because they ended in 1996 and the WG felt that they would add little to the assessment. An Irish otter trawl tuning index was made available in 2001 ( , age 0 to 15). While this fleet mainly targets Nephrops, vessels do on occasion move into areas where plaice are abundant. Landings of plaice by this fleet were approximately 15% of total international landings in 2000 and the WG considered that this fleet could provide a useful index of abundance for plaice. The effects of vessel characteristics on lpue for UK(E&W) commercial tuning-series was investigated in 2001 to investigate the requirement for fishing power corrections due to MAGP IV re-measurement requirements. It was found that vessel characteristics had less effect on lpue than geographic factors and unexplained noise and concluded that corrections were not necessary. However, vessels of certain size tended to fish in certain rectangles. This confounding may have resulted in the underestimation of vessel effects. Currently, age-based tuning data available for this assessment comprise three commercial fleets; the UK(E&W) otter-trawl fleet (UK(E&W)OTB, from 1987), the UK(E&W) beam-trawl fleet (UK(E&W)BT, from 1989) and the Irish otter-trawl fleet (IR OTB, from 1995). However, as a consequence of inconsistencies in these commercial tuning fleets and surveys in the Irish Sea no commercial tuning information is used in the assessment. The area and HP-correction employed to calculate the UK(E&W) commercial effort indices require re-evaluation since vessels have changed greatly since the relationship was modelled. Commercial lpue data are no longer used in the assessment. B.5. Other relevant data Model used: Aarts and Poos (2009) (AP) Software used: R version Model Options chosen:

191 ICES WGCSE REPORT Input data types and characteristics: ASSESSMENT YEAR 2011 WKFLAT Assessment model Tuning fleets UK-BTS Sept (Trad) Series omitted Selectivity model Discard fraction Landings num-at-age, range: AP UK(E&W)-BTS-Q present, ages 1 6 UK(E&W)-BTS-Q1 UK(E&W) OTB UK(E&W) BT IR-OTB NIGFS-WIBTS-Q1 NIGFS-WIBTS-Q4 Survey omitted Series omitted Series omitted Series omitted 1993-present 1993-present Linear Time Varying Spline atage (TVS) Polynomial Time Varying Spline at age (PTVS) 1 9+ Discards num-at-age, year range, age range , ages 1 5 C. Historical stock development The stock of plaice in the Irish Sea has been assessed by ICES since Assessment methods and settings In 1987 the stock was assessed using a Laurec Shepherd (hybrid) tuned VPA. Concerns about deteriorating data quality prompted the use in 1994 of XSA. A subsequent divergence in commercial cpue and survey data, and the wish to include biomass indices, prompted the use of ICA. The settings for each of the assessments between 1991 and 2009 are detailed in Table B.2. Since 2006, the assessment has been an update ICA assessment with the separable period increased by one year at each assessment working group. In 2009 and 2010, FLICA was used to run the assessment: the R and FLR packages have been documented within the WG report. In 2011, WKFLAT estimated discards-at-age and proposed that the AP model is used to model the stock. Over the years, trial runs have explored many of the options with regards XSA settings, including: The applicability of the power model on the younger ages was explored in: 1994; 1996; 1998; 1999; 2000 and Different levels of F shrinkage were explored in 1994; 1995; The effect of different time tapers was investigated in The S.E. threshold on fleets was examined in The level of the catchability plateau was investigated in ICA settings explored since 2005 have included: The length of the separable period. The reference age.

192 1562 ICES WGCSE REPORT 2013 The age range of the landings data. The effect of including hypothetical discard reconstructions in the catch. AP model settings were trialled in 2011: The various combinations of time-variance for selectivity and discard fraction. The suitable age range of the discards was investigated. The suitable starting year of the model was investigated with values from 1990 to 1993 trialled. D. Short-term projection Short-term projections are not made for Irish Sea plaice at present. However, the methodology last employed follows for reference by future working groups. Software: Multi Fleet Deterministic Projection (MFDP) Age-based short-term projections were conducted for a three year period using initial stock numbers derived from ICA analyses. Numbers-at-age 2 were considered poorly estimated and generally overwritten using a geometric mean (GM) of past recruitment values. Population numbers-at-age 3 in the intermediate year (terminal year +1) were also overwritten with the GM estimate depreciated for Fsq and natural mortality. Recruitments since 1990 have been estimated to be at a lower level and to be less variable than those earlier in the time-series. Consequently a short-term geometric mean (from 1990 to two years before the terminal year) was used. Previously, the exploitation pattern is an unscaled three year arithmetic mean. However, alternative options may be used depending on recent F trajectories and the working group s perception of the fishery. Catch and stock weights-at-age were generally taken as the mean of the last three years and the maturity ogive and natural mortality estimates are those used in the assessment method. E. Medium-term projections Medium term projections are not carried out for this stock. Previous Software: MLA miscellany Input values to the medium-term forecast were the same as those used in the shortterm forecast. Although a Beverton Holt stock recruit relationship has been assumed previously, a simple geometric mean may now be more appropriate. F. Yield and biomass per recruit/long-term projections Software: Multi Fleet Yield per Recruit (MFYPR) Yield-per-recruit calculations are conducted using the same input values as those used for the short-term forecasts. Currently the YPR calculations are used as a basis for determining the catch option for advice. G. Biological reference points WKFLAT have rejected the use of reference points given the current trends only assessment and indicated that these will need to be revised. Biological reference points, last used by WGCSE in 2010, were proposed for this stock by the 1998 working group; see below:

193 ICES WGCSE REPORT Type Value Technical basis Blim Not defined. There is no biological basis for defining Blim as the stock recruitment data are uninformative. Bpa 3100 t Bpa = Bloss. Precautionary approach Flim Not defined. There is no biological basis for defining Flim as Floss is poorly defined. Fpa 0.45 Fpa = Fmed in a previous assessment, and in long-term considerations. This is considered to provide a high probability of maintaining SSB above Bloss in the long term. Targets Fy Not defined. Yield and spawning biomass per Recruit F-reference points: FISH MORT YIELD/R SSB/R Ages 3 6 Average last 3 years F Fmed Estimated by the WG in MSY reference points were explored by WGCSE 2010 using the Cefas ADMB code presented to WKFRAME (ICES 2010). However, due to the high level of discards in the stock and unreliable estimates of recruitment, MSY reference points were rejected by the working group. H. Other issues None. I. References Aarts, G. and Poos, J. J Comprehensive discard reconstruction and abundance estimation using flexible selectivity functions. ICES Journal of Marine Science, 66: Armstrong, M.J., Dickey-Collas, M., Gerritsen, H., Bromley, P., Dunn, M., Fox, C., Milligan, S., O'Brien, C., Pawson, M., Stewart, C., Witthames, P., Warr, K., Woolner, L., Connolly, P., Whitmore, J., Hoey, S., O'Brien, D., Danilowicz, B., Heffernan, O., Nash, R., Geffen, A. and Blythe, R Development and validation of egg-production based biomass estimates, using cod and plaice stocks in the Irish Sea. Final Report of EU contract 98/090 to Commission of the European Communities. April pp. Brander, K The population dynamics and biology of cod (Gadus morhua L.) in the Irish Sea 104pp. Dunn and Pawson The stock structure and migrations of plaice populations on the west coast of England and Wales Journal of Fish Biology

194 1564 ICES WGCSE REPORT 2013 Fox, O'Brien and Dickey-Collas Modelling fish larval distributions in the western Irish Sea Ichthyoplankton Ecology 16. Fox, C.J., Planque, B.P., and Darby, C.D Synchrony in the recruitment time-series of plaice (Pleuronectes platessa L) around the United Kingdom and the influence of sea temperature. Journal of Sea Research 44: ICES Report of the Workshop on Implementing the ICES FMSY Framework, March 2010, Copenhagen, Denmark. ICES CM 2010/ACOM: pp. ICES Report of the Working Group on Multispecies Assessment Methods (WGSAM), 6 10 October 2008, ICES Headquarters, Copenhagen. ICES CM 2008/RMC: pp. Nichols, Haynes, Fox, Milligan, Brander and Chapman Spring plankton surveys of the Irish Sea in 1982, 1985, 1987, 1988 and 1989; hydrography and the distribution of fish eggs and larvae 111pp. Pawson, M.G Biogeographical identification of English Channel fish and shellfish stocks. Fisheries Research Technical Report No. 99. MAFF Directorate of Fisheries Research, Lowestoft. Pawson, Pickett and Walker The coastal fisheries of England and Wales, Part IV: A review of their status pp. Sideek The estimation of natural mortality in Irish Sea plaice (Pleuronectes platessa L.) using tagging methods 206pp. Simpson The spawning of the plaice (Pleuronectes platessa L.) in the Irish Sea Fishery Investigations Series II pp.

195 ICES WGCSE REPORT Table B.1. Data sources and derivation of international landings and, from 2011, discards; where % sampled indicates the percentage of the total landings represented by sampling. YEAR SOURCE OF WG 1978 Len. comp. DATA UK BELGIUM IRELAND NETHER LANDS DERIVATION OF INTERNATIONAL LANDINGS AND DISCARDS quarterly 1 quarterly 1 quarterly 1 Irish raised to Irish and N.Irish; UK raised to UK (E&W) and Scotland ALK quarterly 1 quarterly 1 quarterly 1 Belgian raised to Belgian, Dutch and French Age comp Len. comp. quarterly 1 quarterly 1 quarterly 1 UK + Bel + IR combined to total int. separate sex quarterly 1 quarterly 1 quarterly 1 Irish raised to Irish and N.Irish; UK raised to UK (E&W), Sco and IOM. ALK quarterly 1 quarterly 1 quarterly 1 Belgian raised to Belgian, Dutch and French Age comp As for As for Len. comp. quarterly 1 quarterly 1 quarterly 1 UK + Bel + IR combined to total int. separate sex As for 1980 As for 1980 As for 1980 As for 1980 As for 1980, separate sex As for 1980; sexes combined quarterly 2nd qtr quarterly Irish raised to Irish and N.Irish ALK quarterly 2nd qtr quarterly UK raised to UK (E&W), Scotland, I.O.M., French, Dutch and Belgian Age comp Len. comp. quarterly 2nd qtr quarterly UK + IR combined to total int. sexes combined quarterly quarterly quarterly Irish raised to Irish and N.Irish; UK raised to UK (E&W), Sco and IOM % SAMPLED

196 1566 ICES WGCSE REPORT 2013 YEAR SOURCE OF WG DATA UK BELGIUM IRELAND NETHER DERIVATION OF LANDS INTERNATIONAL LANDINGS AND DISCARDS ALK quarterly quarterly quarterly Belgian raised to Belgian, Dutch and French Age comp Len. comp. quarterly quarterly quarterly UK + Bel + IR combined to total int. sexes combined quarterly quarterly quarterly Irish raised to Irish.,N.Irish and French ALK quarterly quarterly quarterly UK raised to UK (E&W), Scotland and I.O.M.; Belgian used alone Age comp As for As for As for As for As for As for As for As for As for Len. comp. quarterly quarterly quarterly UK + Bel + IR combined to total int. As for 1986 As for 1986 As for 1986 As for 1986 As for 1986 As for 1986 As for 1986 As for 1986 As for 1998 As for 1986 As for 1986 As for 1986 As for 1986 As for 1986 As for 1986 As for 1986 As for 1986 As for 1998 As for 1998 % SAMPLED 91 As for As for 1986 except Irish beam trawl raised using UK age comps As for 1986 (Irish beam trawl now sampled) As for As for As for As for 1986 (Belgian samples supplemented with UK data) 90 As for As for quarterly quarterly quarterly quarterly Irish raised to Irish., N.Irish and French; Belgian and Dutch used alone 87

197 ICES WGCSE REPORT YEAR SOURCE OF WG DATA UK BELGIUM IRELAND NETHER DERIVATION OF LANDS INTERNATIONAL LANDINGS AND DISCARDS ALK quarterly quarterly quarterly quarterly UK raised to UK (E&W), Scotland and I.O.M. Age comp As for As for As for As for Len. comp. quarterly quarterly quarterly quarterly UK + Bel + IR + NL combined to total int. As for 1986 As for 1999 As for 1998 As for 1986 As for 1986 As for 1999 As for 1998 As for 1986 As for 1998 As for (except UK raised to include NL landings) As for As for As for quarterly 1st qtr quarterly Belgium raised using 1st qtr values ALK quarterly 1st qtr quarterly UK raised to Sco and France; Irish raised to Irish and N.Irish Age comp Len. comp. quarterly 1st qtr quarterly UK + Bel + IR combined to total int. quarterly quarterly quarterly 52 ALK quarterly - quarterly UK raised to Sco and France; Irish raised to Irish, N.Irish and Bel Age comp Len. comp. quarterly - quarterly UK + IR combined to total int. quarterly quarterly quarterly 81 ALK quarterly qrts 1,2 quarterly UK raised to Sco and France; Irish raised to Irish, N.Irish and Bel Age comp Len. comp. quarterly qrts 1,2 quarterly UK + IR combined to total int. quarterly quarterly quarterly 92 3 % SAMPLED 70

198 1568 ICES WGCSE REPORT 2013 YEAR SOURCE OF WG DATA UK BELGIUM IRELAND NETHER DERIVATION OF LANDS INTERNATIONAL LANDINGS AND DISCARDS ALK quarterly quarterly quarterly UK raised to Sco and France; Irish raised to Irish, N.Irish and Bel Age comp Len. comp. quarterly quarterly quarterly UK + IR combined to total int. quarterly quarterly quarterly 90 3 ALK quarterly quarterly quarterly UK raised to Sco and France; Irish raised to Irish and N.Irish Age comp Len. comp. quarterly quarterly quarterly UK + Bel + IR combined to total int. quarterly annual quarterly 94 ALK quarterly annual quarterly UK raised to Sco and France; Irish raised to Irish and N.Irish Age comp Len. comp. quarterly annual quarterly UK + Bel + IR combined to total int. quarterly quarterly quarterly 89 ALK quarterly quarterly quarterly UK raised to Sco and France; Irish raised to Irish and N.Irish Age comp Len. comp. quarterly quarterly quarterly UK + Bel + IR combined to total int. quarterly quarterly quarterly 94 ALK quarterly quarterly quarterly UK raised to Sco and France; Irish raised to Irish and N.Irish Age comp Len. comp. quarterly quarterly quarterly UK + Bel + IR combined to total int. quarterly quarterly quarterly 100 ALK quarterly quarterly quarterly UK raised to Sco and France; Irish raised to Irish and N.Irish % SAMPLED

199 ICES WGCSE REPORT YEAR SOURCE OF WG DATA UK BELGIUM IRELAND NETHER Age comp. Discard len comp Discard age comp 2012 Len. comp. LANDS DERIVATION OF INTERNATIONAL LANDINGS AND DISCARDS quarterly quarterly - UK + Bel + IR combined to total int. quarterly quarterly - UK(raised) + Bel combined to total int. quarterly - - UK(raised) + Bel combined to total int. quarterly quarterly quarterly 100 ALK quarterly quarterly quarterly UK raised to Sco and France; Irish raised to Irish and N.Irish Age comp. Discard len comp Discard age comp 2013 Len. comp. quarterly quarterly - UK + Bel + IR combined to total int. quarterly quarterly - UK(raised) + Bel combined to total int. quarterly - - UK(raised) + Bel combined to total int. quarterly quarterly quarterly 100 ALK quarterly quarterly quarterly UK raised to Sco and France; Irish raised to Irish and N.Irish Age comp. Discard len comp quarterly quarterly - UK + Bel + IR combined to total int. quarterly quarterly - UK(raised) + Bel combined to total int. % SAMPLED Discard age comp quarterly - - UK(raised) + Bel combined to total int. 1 Assumed (not explicitly stated in report). 2 Revised Revised 2008.

200 1570 ICES WGCSE REPORT 2013 Table B.2. Assessment model settings since ASSESSMENT YEAR Assessment Age Range Fbar Age Range Assessment Method L.S. L.S. XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA Tuning Fleets UK trawl, years: ages: UK otter, years: ages: UK beam, years: ages: Bel Beam, years: ages: IR otter, years: ages: UKBTS Sept, years: ages: UKBTS Mar, years: ages: IR-JPS, years: ages: Time taper 20yr tri 20yr tri 20yr tri No No No No No No No No No Power model ages P shrinkage True False True True True True True False False False False False Q plateau age F shrinkage S.E Number of years Number of ages Fleet S.E

201 ICES WGCSE REPORT ASSESSMENT YEAR Assessment model ICA ICA ICA ICA ICA ICA Tuning fleets UK(E&W)OTB UK(E&W)-BTS-Q ages: UK(E&W)-BTS-Q UK(E&W)BT IR-OTB NIGFS-WIBTS-Q Biomass index NIGFS-WIBTS-Q Biomass index Time series weights Full time series Full time series Full time series Full time series Full time series Full time series Num years for separable Reference age Terminal S Catchability model fitted linear linear linear Linear linear linear SRR fitted No No No No No No Landings number-at -age, range:

202 1572 ICES WGCSE REPORT 2013 ASSESSMENT YEAR Assessment model AP AP AP Tuning fleets UK-BTS Sept (Trad) Series omitted Series omitted Series omitted UK(E&W)-BTS-Q , ages , ages , ages 1 6 UK(E&W)-BTS-Q1 Survey omitted Survey omitted Survey omitted UK(E&W) OTB Series omitted Series omitted Series omitted UK(E&W) BT Series omitted Series omitted Series omitted IR-OTB Series omitted Series omitted Series omitted NIGFS-WIBTS-Q NIGFS-WIBTS-Q Time series weights n/a n/a n/a Num yrs for separable n/a n/a n/a Reference age n/a n/a n/a Terminal S n/a n/a n/a Catchability model fitted n/a n/a n/a SRR fitted n/a n/a n/a Selectivity model Discard fraction Linear Time Varying Spline at age (TVS) Polynomial Time Varying Spline at age (PTVS) Linear Time Varying Spline at age (TVS) Polynomial Time Varying Spline at age (PTVS) Landings num at age, range: Linear Time Varying Spline at age (TVS) Polynomial Time Varying Spline at age (PTVS) Discards N at age, yrs ages r , ages , ages , ages 1 5

203 ICES WGCSE REPORT Stock Annex 6.8: Irish Sea Sole Stock Working Group Irish Sea Sole (Division VIIa) WGCSE Date 6 Feb 2011 (WKFLAT 2011) Last updated 12 May 2013 by Sofie Nimmegeers A. General A.1. Stock definition Sole occur throughout the Irish Sea, but are found more abundant in depth less than 60 m. Recent information on stock identity, distribution and migration issues is included in the report of WKFLAT Cuveliers et al. (2011) combined the results obtained from ten microsatellite markers (long-term estimate of population structure) with results from otolith microchemistry analyses (short-term estimate of connectivity) on adult sole populations in the North-east Atlantic area. Major large scale differentiation was detected between three distinct regions (Baltic transition area, North Sea, Irish/Celtic Seas) with both types of markers. The assignment success of individuals to their collection location was much higher based on otolith edge microchemistry compared to the genetic assignments at all sampling locations, except for the Irish Sea. Only 28.6% of individuals (n = 30) caught in the Irish Sea could be assigned to their catch location based on otolith edge microchemistry, whereas this region showed high genetic self-assignment scores (ca. 60% of 91 individuals) suggesting a spawning population that is genetically distinct. 32% of the misclassifications based on otolith microchemistry were allocated to the neighbouring Celtic Sea. These results are consistent with tagging studies of sole in the Irish Sea and Bristol Channel, showing mainly local recruitment and limited movement of sole outside the management areas (Horwood, 1993; Williams, 1965). Therefore, the management unit is considered to correspond to the stock unit for Irish Sea sole. A.2. Fishery There are three main countries fishing for sole in the Irish Sea; Belgium, taking the bulk of the landings (70 80% in recent years). Ireland and UK taking about 17% and 5% respectively of the sole landings. Northern Ireland, Scotland, Isle of Man and France take the remainder. Approximately 19 Belgian beam trawlers are operating in the eastern part (Liverpool Bay and Morecambe Bay) and southwestern part of the Irish Sea. The UK trawl fleet operates predominantly in the eastern part of the Irish Sea. Sole catches from Ireland are mainly coming from bycatches in the Nephrops fishery (operation in the northwest of the Irish Sea).

204 1574 ICES WGCSE REPORT 2013 Effort of the Belgian beam trawl fleet in When fishing in VIIa it is prohibited to use any beam trawl of mesh size range mm unless the entire upper half of the anterior part of such a net consists of a panel of netting material attached directly to the headline of the net, extending towards the posterior of the net for at least 30 meshes and constructed of diamond-meshed netting material of which no individual mesh is of mesh size less than 180 mm. The Irish otter-trawl fleet employs either a 70 mm mesh with square mesh panels or more commonly an 80 mm mesh. Similarly the Belgian and UK(E&W) beam trawls use 80 mm mesh gear. Otter trawlers targeting roundfish have, since 2000, used 100 mm mesh gear. It was concluded at the 2000 working group and confirmed in 2001 that the cod recovery measures first enacted (EU Regulations 304/2000 and 2549/ revisions in ) in 2000 would have had little impact on the sole fishery. The closed area in 2001 covered a reduced area confined to the west of the Irish Sea and therefore is also expected to have had little effect on the level of fishing effort for sole The spawning closure for cod in 2002 is also unlikely to have had an impact on the sole fishery. The effort regulations and maximum daily uptake, implemented in 2003 will delay the uptake of the quota but is also unlikely to be restrictive for the total uptake. It is unlikely that any measures concerning the cod management plan in the Irish Sea had restrictions on the sole fishery after Discard are estimated to be minor. Preliminary data indicate ranges from 0 to 8% by weight discarded. Although no data are available on the extent of misreporting of landings from this stock, it is not considered to be a problem for this stock, given the partial uptake of the agreed TAC in recent years.

205 ICES WGCSE REPORT A.3. Ecosystem aspects

206 1576 ICES WGCSE REPORT 2013 (Connolly, P.L. et al., 2009)

207 ICES WGCSE REPORT B. Data B.1. Commercial catch Quarterly age compositions are available from UK(E&W), Belgium and Ireland, as well as quarterly landings from France, Northern Ireland, Isle of man and Scotland. The total international age composition is obtained using a combined ALK from UK(E&W), Belgium and Ireland raw data, responsible for 97% of the total international sole landings. The combined ALK is applied to the length distributions of the separate countries to obtain an aggregated age composition. Catch weights were obtained from the combined AWK (UK(E&W), Belgium and Ireland raw data). Stock weights were obtained using the Rivard weight calculator ( that conducts a cohort interpolation of the catch weights. B.2. Biological Currently there are no direct (from tagging) or independent (from survey information) estimates of natural mortality. Therefore, annual natural mortality (M) is assumed to be constant over ages and years, at 0.1 yr -1. The maturity ogive used in this and previous assessments is based on survey information for this stock.: Age and older Mat Proportions of M and F before spawning were set to zero, as in previous years. Males and females of this stock are strongly dimorphic, with males showing much reduced rates of growth after reaching maturity, whilst females continue to grow. Given the minimum landing size of 24 cm the majority of landings represent mature females. B.3. Surveys One survey is used in the assessment of VIIa sole: the UK beam-trawl survey (UK (BTS- 3Q)). Area covered Irish Sea; 52 0 N to 55 0 N; 3 0 W to W.

208 1578 ICES WGCSE REPORT 2013 UK - Irish Sea survey E3 E4 E5 E6 E Target species Flatfish species, particularly juvenile plaice and sole. Length data recorded for all finfish species caught; samples for age analysis taken from selected species. Time period : September (continuing) Gear used Commercially-rigged 4 m steel beam trawl; chain matrix; 40 mm codend liner. Mean towing speed: 4 knots over the ground. Tow duration: 30 minutes. Tow duration for trips in was 15 minutes; in 1992 comparative tows of 15 and 30 minutes length were carried out, and subsequent cruises used a standard 30 minute tow. The data from earlier years were converted to 30 minutes tow equivalent using relationships for each species derived from the comparative work in Vessel used: R.V. Endeavour (CEFAS). Survey design Survey design is stratified by depth band and sector (Depth bands are 0 20, 20 40, 40+). Station positions are fixed. Number of stations = 35 in the eastern Irish Sea, 15 in the western Irish Sea, and 16 in St George s Channel (primary stations). Sampling intensity highest in the eastern Irish Sea, in the main flatfish nursery and fishery areas. Method of analysis Raised, standardized length frequencies for each station combined to give total length distribution for a stratum (depth band/sector). Sector age length keys applied to stratum

209 ICES WGCSE REPORT length distributions ; stratum age length keys applied 1995 onwards. Mean stratum cpue (kg per 100 km and numbers-at-age per 100 km) are calculated. Overall mean cpue values are simple totals divided by distance in metres (or hours fished). Population number estimates derived using stratum areas as weighting factors. The September beam-trawl survey has proven to estimate year-class strength well, and providing 50% to over 90% of the weighting to the total estimates of the incoming year classes. B.4. Commercial lpue Lpue and effort-series were available from the Belgium beam trawlers, UK(E&W) beam and otter trawlers, the Irish otter trawlers and from two UK beam-trawl surveys (September and March). Lpue for both UK and Belgian beam trawlers has declined since the beginning of the time-series, but has remained relatively constant over the last decade, with a renewed increase over for Belgium. After an increase in , the UK beam-trawl lpue dropped to a remarkable low level. Effort from both commercial beam-trawl fleets increased from the early seventies until the late eighties. Since then Belgian beam trawl effort has declined over the nineties, increased again in the period and subsequently dropped to much lower values in (the lowest values since 1984). In the nineties, the UK beam-trawl effort fluctuated around a lower level than the late eighties, and dropped during the 21st century to a lower value of the time-series. Indices of abundance derived from the UK September survey (UK (BTS-3Q)) (data from 1988 onwards) are shown in WGNSDS 2002 (Table ). High abundance indices for the UK September survey (UK (BTS-3Q)) can be seen for year classes 1989, 1995 and The dataseries from the UK March beam-trawl survey (UK (BTS-1Q)) is rather short (from 1993 to 1999), and therefore difficult to interpret. B.5. Other relevant data No information. C. Assessment: data and method Model used: XSA Software used: IFAP/Lowestoft VPA suite

210 1580 ICES WGCSE REPORT 2013 Model Options chosen since 2004: ASSESSMENT YEAR Assessment Model XSA SURBA XSA XSA Fleets BEL-CBT omitted omitted UK-CBT omitted omitted UK rec yr rec yr 2 7 (BTS-3Q) 2 9 UK (BTS-1Q) 2 9 Time-Ser. Wts tricubic 20yrs linear 20 yrs linear 20 yrs Power Model None None none Q plateau Shk se Shk Age-yr 5 yrs 5 ages 5 yrs 3 ages 5 yrs 3 ages Pop Shk se Prior Wting None None none Plusgroup FBAR ASSESSMENT YEAR 2011 WKFLAT Assessment Model XSA XSA XSA Fleets BEL-CBT Omitted omitted omitted UK-CBT Omitted omitted omitted UK 1988 rec yr rec yr rec yr 2 7 (BTS-3Q) UK omitted Omitted omitted (BTS-1Q) Time-Ser. Wts uniform Uniform uniform Power Model none None none Q plateau Shk se Shk Age-yr 5 yrs 3 ages 5 yrs 3 ages 5 yrs 3 ages Pop Shk se

211 ICES WGCSE REPORT Prior Wting None None None Plusgroup FBAR Input data types and characteristics: TYPE NAME YEAR RANGE AGE RANGE VARIABLE FROM YEAR TO YEAR YES/NO Caton Catch in tonnes 1970 last data year Canum Weca West Mprop Fprop Matprop Catch-at-age in numbers Weight-at-age in the commercial catch Weight-at-age of the spawning stock at spawning time Proportion of natural mortality before spawning Proportion of fishing mortality before spawning Proportion mature-at-age 1970 last data year 1970 last data year 1970 last data year 1970 last data year 1970 last data year 1970 last data year Natmor Natural mortality 1970 last data year 2 8+ Yes 2 8+ Yes 2 8+ Yes 2 8+ Yes-but based on back calculated catch weights 2 8+ No-set to 0 for all ages in all years 2 8+ No-set to 0 for all ages in all years 2 8+ No-the same ogive for all years 2 8+ No-set to 0.1 for all ages in all years Tuning data: TYPE NAME YEAR RANGE AGE RANGE Tuning fleet 1 UK (BTS-3Q) 1988 last data year 2 7 Note: several other commercial tuning fleets BEL-CBT (Belgian beam-trawl fleet), UK- CBT (UK beam-trawl fleet), UK-COT (UK otter-trawl fleet), IRL-COT (Irish otter-trawl fleet) and two other surveys (UK (BTS-1Q) and Irish Juvenile Plaice Survey) have been used or made available in the past. A thorough investigation of the utility of these tuning indices was conducted at the 2002 working group. The results are summarized in the stock annexes of the reports of WGNSDS and WGCSE D. Short-term projection Model used: Age structured deterministic projection Software used: MFDP

212 1582 ICES WGCSE REPORT 2013 Initial stock size: Taken from the XSA for ages 3 and older. The recruitment-at-age 2 in the last data year is estimated using RCT3. The short-term geometric mean recruitment ( ) is used for age 2 in all projection years. Maturity: the same ogive as in the assessment is used for all years (see table above) F and M before spawning: set to 0 for all ages in all years Weight-at-age in the stock: average weight of the last three years Weight-at-age in the catch: average weight of the three last years Exploitation pattern: average of the three last years, not scaled to the last year s FBAR (4 7) if no trend in F was detected (scaled to the last year s FBAR (4 7) if a trend in F was detected). Intermediate year assumptions: status quo F in and TAC constraint in 2013 Stock recruitment model used: none Procedures used for splitting projected catches: not relevant E. Medium-term projections Medium-term projections were not conducted at WKFLAT The last medium-term projections were carried out in The settings used are described below. Model used: Age structured Software used: IFAP single option prediction Initial stock size: Same as in the short-term projections. Natural mortality: Set to 0.2 for all ages in all years Maturity: The same ogive as in the assessment is used for all years F and M before spawning: Set to 0 for all ages in all years Weight-at-age in the stock: Assumed to be the same as weight-at-age in the catch Weight-at-age in the catch: Average weight of the three last years Exploitation pattern: Average of the three last years, scaled by the FBAR (3 6) to the level of the last year Intermediate year assumptions: F-factor from the management option table corresponding to the TAC Stock recruitment model used: None, the long-term geometric mean recruitment at-age 2 is used Uncertainty models for excel, Latin Hypercubed, 500 iterations, fixed random number generator Initial stock size: Lognormal distribution, LOGNORM(mean, standard deviation), with mean as in the short-term projections and standard deviation calculated by multiplying the mean by the external standard error from the XSA diagnostics (except for age 2, see recruitment below).

213 ICES WGCSE REPORT Natural mortality: Set to 0.2 for all ages in all years. Maturity: The same ogive as in the assessment is used for all years. F and M before spawning: Set to 0.2 for all ages in all years. Weight-at-age in the stock: Assumed to be the same as weight-at-age in the catch. Weight-at-age in the catch: Average weight of the three last years. Exploitation pattern: Average of the three last years, scaled by the FBAR (3 6) to the level of the last year. Intermediate year assumptions: F-factor from the management option table corresponding to the TAC. Stock recruitment model used: Truncated lognormal distribution, TLOGNORM(mean, standard deviation, minimum, maximum), is used for recruitment age 2, also in the initial year. The long-term geometric mean, standard deviation, minimum, maximum are taken from the XSA for the period th last year. F. Long-term projections Model used: age structured deterministic projection Software used: MFYPR Inputs as for short-term projection. G. Biological reference points Type Value Technical basis MSY MSY Btrigger 3100 t Default to value of Bpa. Approach FMSY 0.16 Provisional proxy based on stochastic simulations assuming a Ricker S/R relationship (range ). Blim 2200 t Blim = Bloss. The lowest observed spawning stock (ACFM 1999), followed by an increase in SSB. Precautionary Bpa 3100 t Bpa Blim * 1.4. The minimum SSB required ensuring a high probability of maintaining SSB above its lowest observed value, taking into account the uncertainty of assessments. Approach Flim 0.4 Flim = Floss. Although poorly defined, there is evidence that fishing mortality in excess of 0.4 has led to a general stock decline and is only sustainable during periods of aboveaverage recruitment. Fpa 0.3 This F is considered to have a high probability of avoiding Flim. Precautionary approach reference points have not been changed during In this period, FPA was set at 0.45 on the technical basis of high probabilities of avoiding Flim

214 1584 ICES WGCSE REPORT 2013 and of SSB remaining above BPA. In 2007, FPA was changed to 0.3 due to the rescaling of SSB estimates. In 2010, MSY reference points were added by WGCSE. H. Other issues A management plan for Irish Sea sole could be developed, also taking into account the dynamics of the plaice stock in that area. I. References ICES Report of the Working Group on the Assessment of Northern Shelf Demersal Stocks (WGNSDS). ICES CM 2002/ACFM: pp. ICES Report of the Working Group on the Celtic Seas Ecoregion (WGCSE), May 2009, Copenhagen, Denmark. ICES CM 2009/ACOM: pp. ICES Report of the Working Group on Celtic Seas Ecosystems, May 2010, Copenhagen, Denmark. ICES CM 2010/ACOM: pp. ICES Report of the Working Group on Celtic Seas Ecosystems, May 2011, Copenhagen, Denmark. ICES CM 2011/ACOM: pp. Connolly, P.L., Kelly, E., Dransfeld, L., Slattery, N., Paramor, O.A.L., and Frid, C.L.J MEFEPO North Western Waters Atlas. Marine Institute. Horwood, J. W. 1993a. The Bristol Channel Sole (Solea solea (L.)): A Fisheries Case Study. Advances in Marine Biology, 29: Williams, T Movements of tagged soles in the Irish Sea and Bristol Channel. ICES CM 1965, Near Northern Seas Committee, No. 87 (mimeo).

215 ICES WGCSE REPORT Stock Annex 7.2: Celtic Sea cod Stock Expert Group Cod in VIIe k (Celtic Sea cod) Celtic Sea Ecoregion Working Group Date March 2012 Revised by Colm Lordan, Lionel Pawlowski A. General A.1. Stock definition Since 1997, this assessment has related to the cod in Divisions VIIe k, covering the Western Channel and the Celtic Sea. Tagging information presented at WKROUND 2012 (WDs 9 and 11) confirms minimal movement of cod from VIIe k to other areas. In fact even within VIIe k there seems to be limited mixing between fish tagged in VIIg or VIIa South and those tagged in VIIf and VIIe. Up to 2008, the management area was set in Divisions VIIb k,viii, IX, X, and CECAF which does not correspond to the area assessed. The management area was revised in 2009 to exclude VIId. The new TAC covers ICES Areas VIIb c, VIIe k, VIII, IX, X, and CECAF (1). This is more representative of the stock area in recent years and landings from VIIbc, VIII, IX and X have been minimal. The area assessed has gradually increased from VIIfg before 1994 to VIIfgh, to VIIefgh in 1996 and finally to VIIe k. In 1994, at the request of ACFM, the ICES Working Group on Southern Shelf Demersal Stocks (WGSSDS) studied the possible extension of the area assessed from VIIfg to VIIfgh. Examination of data from surveys and logbooks indicated a continuity of the distribution of VIIg cod into VIIh. Depending on the year, catches in Division VIIh represented 9 15% of the catches in VIIfg, with a coincidence of years of peak or low catches in both areas. Therefore, catches from VIIh were included in the assessment. In 1996, at the request of ACFM, WGSSDS studied the possible extension of the area assessed from VIIfgh to VIIefgh. The population dynamics parameters for VIIfgh and VIIe cod were examined and compared for the period , when independent tuning fleets, international catch-at-age, mean weights-at-age in the landings and in the stocks were available for both areas. Patterns of F were consistent between VIIe and VII fgh in earlier years ( ), and SSBs trends were similar in the period The patterns of recruitments (age 1) were found to be fairly consistent through this period , though it cannot be assumed that this consistency was also valid in earlier years when catch at age were only available in Divisions VIIf, g, h. It was therefore decided to combine Western Channel Cod with the Celtic Sea Cod assessment for the years , but an independent assessment of Celtic Sea cod in VIIfgh was maintained for the longer period available This was to allow scaling of the historic ( ) SSBs and recruitments values from VIIfgh to VIIe h. At WGSSDS 1997, due to the lack of a long independent series of catch-at-age in Divisions VIIj,k, the estimate of landings from Divisions VIIjk was discussed and it was decided to combine the data of Divisions VIIe,f,g,h and Divisions VIIjk for the period 1993

216 1586 ICES WGCSE REPORT and to raise the data in Divisions VIIe h to landings in Divisions VIIe k for the period The results of an XSA assessment of this series in Divisions VIIe k for had been compared with the results of the assessment in Divisions VIIe h in terms of trends of F, SSB and recruitment. Patterns of these parameters were found very similar and the merging of Divisions VIIjk with Divisions VIIe h mainly resulted in a scaling upwards of SSB and recruitment. The new assessment areas comprised cod in Divisions VIIe k. At the 1999 WGSSDS meeting, an alternative procedure to the tedious re-scaling of SSB and recruitment of the earlier series in VIIfgh to VIIe k every year was proposed (Bellail, 1999, WD3). A long series of landings data from was reconstructed. An average raising factor (1.24) from VIIfgh to VIIe k in the period was applied to VIIfgh landings of the series Results of assessment in terms of SSB and R were very close to those obtained when these parameters were scaled. ACFM accepted this procedure. In the past few biological criteria have been used to justify the widening the stock area. However, recent tagging work by Ireland and the UK supports the idea that there is a resident stock in the Celtic Sea and Western Channel (VIIe k) and mixing with other areas appears to be minimal. The Irish Sea front, running from SE Ireland (Carnsore point) to the Welsh Coast, appears to act as boundary between the Irish Sea and Celtic Sea stock. Juveniles found close to the SE Irish Coast (south of VIIa) are considered part of the Celtic Sea stock. Some migrations and mixing are known to occur in this cod stock. Both conventional and DST tagging information for VIIg (where the majority of landings are made) shows that distribution remained fairly constrained within VIIg. There was some preference to central areas within VIIg during January March. Between April and June the cod appeared to be more widely dispersed within VIIg during Q1 & Q2. Fish tagged in VIIf tended to mix with those off shore in VIIg and h. Whereas some fish tagged in the western English Channel VIIe migrated into VIId for at least part of the year. A.2. Fishery The majority of the landings are made by demersal trawls targeting roundfish (i.e. cod, haddock and whiting), although, in recent years an increasing component has been from gillnets and otter trawls targeting Nephrops and benthic species. Landings are made throughout the year but are generally more abundant during the first semester. Constraining TACs set since 2003 and the impact of the Trevose Head Closure applied since 2005 have reduced landings in Q1 somewhat and spread landing more throughout the year. WGCSE should routinely monitor spatial and temporal changes in landings, effort and lpue for the main fleets catching cod in VIIe k. This has previously been done using maps of landings and lpue by ICES rectangle. A.3. Ecosystem aspects Cod recruitment success has generally shown an increase over the period during which time sea surface temperature in the Celtic Sea has increased (Lynam et al., 2009). Notably the highest recruitment success was for cod spawned in 1986, a year with

217 ICES WGCSE REPORT an exceptionally cold spring. Lynam et al. (2009) also found that SST in spring (MAM) and Calanus helgolandicus, abundance in the Celtic Sea, did prove to be significant predictors of recruitment in Celtic Sea cod in a GAM model. The time lag between availability of this SST and zooplankton information means that their model cannot be readily used in forecasting recruitment in advance of what groundfish surveys might detect. Nevertheless this research should be pursued further, particularly in the context ecosystem determinants of the strong 2009 and 2010 year classes. B. Data B.1. Commercial catch Landings On a quarterly basis, France, Ireland and UK(E+W) have provided catch numbers-at-age and catch weights-at-age for their landings. The Irish landings in VIIg are augmented with some landings made or reported off the southeast coast of Ireland in ICES rectangles 33E2 and 33E3. These rectangles are in the very south of VIIa. Landings only are available for Belgium. France, UK and Ireland data are added quarterly and raised to international landings taking into account Belgian data. Then the quarterly datasets are summed up to the annual values. As a consequence of an update to the French database of landings statistics, some minor revisions (downward) have been applied since 2002 and the updated datasets for international landings. There is no information on the absolute level of misreporting for this stock but there is evidence that misreporting has increased from 2002 when quotas became restrictive with a maximum in Misreporting has decreased since then. Discards Discards data sampled under EU/DCR since 2003 have been generally presented in previous WGCSE but not used in the assessments as they do not cover all the main fleets and quarters yet. Due to the annual management system adopted by the French POs since 2003 in response to the quota restrictions, highgrading has occurred in the French fishery, mainly in VIIfgh. A procedure using both the UK and French landings length data enabled estimation of the French high grading for the years (WD 1 WGSSDS 2006). The adjustments were reapplied to improved estimates of French landings from 2006 at the ICES WKROUND French landings have been corrected with high grading estimates from 2003 to The method used to estimate the highgraded component is described in WD#1 of the WGSSDS For smaller length classes, a scaling of French numbers-at-length based on UK length frequencies or UK number-at-length has been used to estimate length compositions of the French component of highgrading. The accuracy of this method is unknown but it probably underestimates the highgrading levels for those years. Unfortunately, the sampling level of total catch at sea in that period was too poor to get an estimate of the

218 1588 ICES WGCSE REPORT 2013 level of bias. This method was not applied from 2006 onward because highgrading was also observed in the UK landings. Instead, self-sampling data obtained in have been used to estimate the French highgrading level, assuming that the discarding practices in were the same as those observed in 2008 for the main self-sampled fleet. Applying this method back to 2003 was considered inappropriate. The representatives of Fishermen Organisations at WKROUND 2009 indicated that the discarding level was probably not the same in earlier years as highgrading practices are linked to the level of the TAC. The whole method has been described in the WD#17 of WKROUND In 2009 and 2010, the low estimate of highgrading is likely to be related to the French vessels not being restricted by quota because of the decommissioning plan and the reports of effort directed towards more profitable species. Highgrading has also occurred in the UK catches in but given the low level of landings it was not estimated prior this year as well as for Belgian landings. The MLS of the Belgian landings is currently set at 50 cm since The estimates of highgrading by year are slightly revised when annual landings statistics are updated. In 2008 the French-self sampling programme on Celtic Sea cod has produced datasets enabling estimation of discarding and highgrading rates. Assuming the same pattern of discarding in recent years, estimates of French discarding and highgrading back to 2003 were also computed. Estimates of highgrading were also calculated for the French tuning fleets used in the analysis (ICES WKROUND, 2009, WD 17). In 2009 and 2010, the low estimate of highgrading is likely to be related to the French vessels not being restricted by quota because of the decommissioning plan and the reports of effort directed towards more profitable species. Discard estimates are available from Ireland since 1995 (see Marine Institute & Bord Iascaigh Mhara, 2011). For now the assumption is that the discards are mainly at age 1 and the estimates are very uncertain. There are indications that Irish discard rates have increased since 2005 this is something that WGCSE should monitor and discards should be included in the assessment if there are major changes or it is found to have a large impact on the assessment. Lpue Landings and effort data are available for all the main fleets operating in the area and catching cod. The table below summarizes the available data. WGCSE should monitor changes in these fleets over time. NAME AREA SERIES FR gadoid fleet 1 VIIfgh 1983 onwards FR Nephrops fleet 1 VIIfgh 1983 onwards FR otter trawlers 2 VIIe 1983 onwards FR otter trawlers 2 VIIfgh 1983 onwards FR otter trawlers 2 VIIe-k 1983 onwards UK otter trawlers VIIe 1972 onwards UK otter trawlers VIIe-k 1972 onwards UK beam trawlers VIIe-k 1978 onwards IR otter trawlers VIIg 1995 onwards

219 ICES WGCSE REPORT IR beam trawlers VIIg 1995 onwards IR Scottish seiners VIIg 1995 onwards IR otter trawlers VIIj 1995 onwards IR beam trawlers VIIj 1995 onwards IR Scottish seiners VIIj 1995 onwards 1 For Q2+3+4 for consistency with the Trevose Head Closure since 2005 during the first quarter. 2 Annual values, including the Fr gadoid and Nephrops fleets. B.2. Biological Weights-at-age At the 1999 WGSSDS, data for the years were set to the average A revision was carried out at 2001 WGSSDS where the values for the period were set to the average values Depending on the annual datasets available by country for the period , catch weights-at-age data were calculated as the weighted means from French, Irish and UK datasets. Since 2002, VIIe k catch weights-atage have been calculated as the annual weighted means of French, Irish and UK datasets. WKROUND 2012 reviewed the data and concluded that there is a downward trend in mean weights-at-age during the 1980s but they have been relatively stable since then at about 10% lower mean weights than observed in the 1980s. There is some evidence of year effects (e.g and 2005) and cohort effects (e.g. 1999). Stock weights-at-age are the catch weight-at-age data from the 1st quarter. Maturity The maturity ogive applied since 1999, was estimated from the datasets of the UK- WCGFS survey (1st quarter) has been used for the overall series. It replaced an assumed ogive used for the year prior to 1999, derived from Irish sea cod data, when both stocks (VIIa and VIIfg) were assessed in the Irish Sea & Bristol Channel WG up to The table below summarizes the maturity ogives used. AGE Before Current Natural mortality In the assessments, natural mortality is assumed to be constant for the whole range of years and is age-dependant The table below summarizes the values of M accordingly to age. AGE M B.3. Surveys Three survey-series are available.

220 1590 ICES WGCSE REPORT 2013 The discontinued UK-WCGFS-Q1 ( ), conducted during the first quarter, is generally truncated into a shorter series ( ) as it showed a strong trend (domeshaped) when using the full series. This pattern is related to the progressive extension of the studied area of this survey from VIIe to VIIefgh over the years. This time-series only contributes to the estimates at older ages (4 and older). Due to the lack of new data, the series is no longer used in the assessment. The FR-EVHOE (EVHOE-WIBTS-Q4) survey (1997 ), during the 4th quarter, covers the Divisions VIIfghj. The IrGFS (IGFS-WIBTS-Q4) survey (2003 ), during the 4th quarter, in VIIg and VIIj is also used in the assessment. The absolute numbers of cods caught in all of these surveys are extremely low. Attempts to combine survey data have been done at WKROUND 2009 and 2012 to overcome that problem. WKROUND 2012 tested two combinations: mixing data for the whole area and just those in the overlapping area. WKROUND concluded that the overlap area combined index was an improvement on using the two surveys independently or using the full area index. This conclusion was based on the good cohort tracking and fairly consistent catch curves in the combined index Ages 1 4. B.4. Commercial cpue FR-OTDEF: a new time-series of tuning indices has been introduced at WKROUND 2012 upon French datasets considering landings and fishing efforts from otter trawlers (OTDEF métier) which catch per trip are at least 40% made of gadoids in divisions VIIb k during quarters 2 to 4. FR-OTDEF is a substitute for the discontinued FR-Gadoid and FR- Nephrops fleet. B.5. Other relevant data Input from industry No new datasets. There are several industry science partnerships regarding cod: French industry self-sampling programme. Ireland-UK tagging programme in the Irish and Celtic Seas. Irish industry science partnership quarter 1 cod survey At the moment only the data from the French self-sampling programme are integrated into the observation-at-sea dataset used at the assessment working group. Information on tagging are however reviewed each year at the WG and by WKROUND. An Irish industry science partnership survey was carried out in Q This survey has not been repeated due to resource constraints. Any new information provided by the industry is also reviewed each year. C. Historical stock development Model to be used: XSA Software: R with FLR packages FLCore 2.2, FLAssess 2.0.1, FLXSA 2.0, FLEDA 2.0.

221 ICES WGCSE REPORT Model Options agreed at WKROUND 2012: Taper : no Age s catch dep. Stock size : none q plateau : 3 F shrinkage se : 1 F shrinkage year range : 5 F shrinkage age range : 3 F shrinkage age range of mean F : 2 5 Fleet SE threshold : 0.3 Prior weights : No Input data types and characteristics: TYPE NAME YEAR RANGE AGE RANGE VARIABLE FROM YEAR Caton Catch in tonnes Yes Canum Weca West Mprop Fprop Matprop Landings-at-age in numbers Weight-at-age in the commercial catch Weight-at-age of the spawning stock at spawning time. Proportion of natural mortality before spawning Proportion of fishing mortality before spawning Proportion matureat-age Yes Yes Yes No No No Natmor Natural mortality No TO YEAR YES/NO Tuning data: TYPE NAME YEAR RANGE AGE RANGE French Otter Trawler in VIIek Q2 Q4 Combined EVHOE- WIBTS, IGFS-WIBTS FR-OTDEF FR-IR-WIBTS D. Short-term projection Model used: Age structured

222 1592 ICES WGCSE REPORT 2013 Software used: MFDP Initial stock size: 1 ) the survivors at age 2 and greater from the XSA assessment 2 ) N at-age 1 = long-term geometric mean omitting the last two years. Maturity: same ogive as in the assessment F and M before spawning: 0 (for all ages and years) Weight-at-age in the stock: average stock and catch weights over the preceding three years. Exploitation pattern: The F vector used is the average F-at-age in the last three years, scaled by FBAR (2 5) to the level of last year unless there is strong indication of a significant trend in F. In the latter case the average selectivity pattern will be rescaled to the final F in the series. E. Medium-term projections Medium-term forecasts are not provided for this stock. F. Yield and biomass per recruit No stock recruit relationship exists for this stock. Software used : YPR 3.0 (NOAA fisheries toolbox) Stock/catch-at-age/spawning stock weights-at-age: Average last five years; Selectivity on fishing mortality: Rescaled F Average last five years; Selectivity on natural mortality: Rescaled M-at-age (Lorenzen), M-at-age 1=1; Fraction mature: same as maturity ogive; Proportion of fishing/natural mortality before spawning: 0.0. G. Biological reference points TYPE VALUE TECHNICAL BASIS MSY MSY Btrigger 8800 t Provisionally set at Bpa. Approach FMSY 0.40 Provisional proxy based on FMAX (ICES, 2010). Blim 6300 t Blim=Bloss (B76), the lowest observed spawning stock biomass. Precautionary Bpa 8800 t Bpa=Blim*1.4. Biomass above this value affords a high probability of maintaining SSB above Blim, taking into account the variability in the stock dynamics and the uncertainty in assessments. Approach Flim 0.90 The fishing mortality estimated to lead to potential collapse Fpa 0.68 Fpa=5th percentile of Floss. This F is considered to have a high probability of avoiding Flim and maintaining SSB above Bpa in the medium term (assuming normal recruitment), taking into accounts the uncertainty

223 ICES WGCSE REPORT assessments. H. Other issues None. I. References Armstrong, M Results of the UK Fisheries Science Partnership surveys of Celtic Sea cod, haddock and whiting and Western Channel sole and plaice: WD10 contributing to the 2006 WGSSDS 4p. Bellail, R Derivation of Cod landings in Divisions VIIe-k and application for an assessment using a long series WD3 contributing to the 1999 WGSSDS. 4p. Bellail, R Revision of weight at age and catch numbers at age in series of cod in Areas VIIe k. WD4 contributing to the 2001 WGSSDS. 4p.( datasets of CW@A replaced by mean values of the series ). Bellail, R An attempt to estimate and integrate French highgrading of cod in Celtic Sea into annual length and age compositions of landings from WD1 contributing to the 2006 WGSSDS. 7p. Bellail,R.; Peronnet, I.; Rochet, MJ; Lamothe J Self sampling of cod in the Celtic Sea by French trawlers. WD1 contributing to the 2008 WKSC. 13pp. Biseau, A Effect of the cod closure in the Celtic Sea on the fishing behaviour. WD2 contributing to the 2005 SSDSWG. 2p. Biseau, A Trends in French fishing effort in ICES Division VIIg / Impact of the Trevose closure? WD2 contributing to the 2007 SSDSWG. 10p. Biseau, A.; Bellail, R Celtic Sea Cod: Catch limitation and LPUE series for French trawlers. WD1 contributing to the 2004 SSDSWG. Biseau, A.; Bellail, R Trends in French Gadoids effort in the Celtic Sea. Impact of the closure of zones on the fishing strategy of the French vessels? WD2 contributing to the 2006 SSDSWG. 11p. Brander, K. M The location and timing of cod spawning around the British Isles. ICES Journal of Marine Science, 51: Darby, C B-Adapt. WD 15 contributing to the ICES 2004 WGNSDSK. Demaré, W On the CPUE and effort of the Belgian beam trawl fleet operating in the Celtic Sea. WD 14 contributing to the 2006 WGSSDS. Gerritsen, H Maturity at age estimates for Irish Demersal Stocks in VIIb k WD 1 contributing to the 2008 WGSSDS. ICES-ISBCWG Report of the Irish Sea and Bristol Channel Working Group.ICES C.M.1992/Assess:1. (Assessed area= VIIfg). ICES-PGCCDBS Report of the Planning Group on Commercial Catch, Discards and Biological Sampling. ICES CM 2007/ACFM: pp. ICES-PGCCDBS Report of the Planning Group on Commercial Catch, Discards and Biological Sampling. ICES CM 2008/ACOM:29. 91pp. ICES-WGCSE Report of the Working Group on the Celtic Seas Ecoregion (WGCSE), May 2011, Copenhagen, Denmark. ICES CM 2011/ACOM:12.

224 1594 ICES WGCSE REPORT 2013 ICES-WGQAF Report of the Working Group on Quantifying all Fishing Mortality. ICES CM 20008/FTC:03. 24pp. ICES-WGSSDS Report of the Working Group on the Assessment of Southern Shelf Demersal Stocks. ICES C.M. 1995/Assess:6.(Assessed area= VIIfgh). ICES-WGSSDS Report of the Working Group on the Assessment of Southern Shelf Demersal Stocks. ICES C.M. 1997/Assess:5.(Assessed area= VIIefgh). ICES-WGSSDS Report of the Working Group on the Assessment of Southern Shelf Demersal Stocks. ICES C.M. 1998/Assess:4. (Assessed area= VIIe k). ICES-WGSSDS Report of the Working Group on the Assessment of Southern Shelf Demersal Stocks. ICES C.M. 2000/ACFM:4. (Estimating landings VIIe k for years =VIIfgh landings raised by 1.24, new maturity ogive). ICES-WGSSDS Report of the Working Group on the Assessment of Southern Shelf Demersal Stocks. ICES C.M. 2002/ACFM:5. (revision of datasets ). ICES-WGSSDS Report of the Working Group on the Assessment of Southern Shelf Demersal Stocks. ICES C.M. 2005/ACFM:03. ICES-WGSSDS Report of the Working Group on the Assessment of Southern Shelf Demersal Stocks. ICES C.M. 2005/ACFM:04. ICES-WGSSDS Report of the Working Group on the Assessment of Southern Shelf Demersal Stocks. ICES C.M. 2006/ACFM:33. ICES-WGSSDS Report of the Working Group on the Assessment of Southern Shelf Demersal Stocks. ICES C.M. 2007/ACFM:28. ICES-WGSSDS Report of the Working Group on the Assessment of Southern Shelf Demersal Stocks. ICES C.M. 2008/ACOM:12. ICES-WGFTFB Report of the Working Group on Fish Technology and Fish Behaviour. ICES CM 2006/FTC:06. ICES-WGFTFB Report of the ICES-FAO Working Group on Fish Technology and Fish Behaviour (WGFTFB), April 2007, Dublin, Ireland. ICES CM 2007/FTC: pp. ICES Report of the ICES FAO Working Group on Fish Technology and Fish Behaviour (WGFTFB), April 2008, Tórshavn, Faroe Islands. ICES CM 2008/FTC: pp. ICES-WKSC Report of the Workshop on Fishers Sampling of Catches 55p. Lordan, C.; Gerritsen, H The GM assumption in the Short Term Forecast of Cod in VIIe k. WD 7 contributing to the 2008 WGSSDS. Nielsen, A State space fish stock assessment model: SAM model. WD contributing to the ICES 2008 WGNSSK. Righton, D., Quayle, V.A., Hetherington, S. and Burt, G Movements and distribution of cod (Gadus morhua) in the southern North Sea and English Channel: results from conventional and electronic tagging experiments. J. Mar. Biol. Ass. U.K. 87: Scott, R.; Houghton, C.; Woods, T Impacts of the 2005 and 2006 Celtic Sea closed area on the catches of cod, plaice and sole. WD11 contributing to the 2006 WGSSDS. 29p. Warnes,S.;Pawson, M.; Bellail, R Revision of Maturity ogive for Cod in Celtic Sea (ICES Divisions VIIe k. WD 11 contributing to 1999 SSDSWG. 4p.

225 ICES WGCSE REPORT Stock Annex 7.4: Haddock VIIb,c,e k Stock Working Group Haddock VIIb,c,e k WGCSE Date last revision 11/05/13 Revised by Hans Gerritsen A. General A.1. Stock definition For assessment purposes, the stock is defined as VIIb k excluding VIId and including rectangles 33E2 and 33E3. The had7a Benchmark in 2013 decided to exclude these rectangles from VIIa (because they are geographically part of the Celtic Sea stock), therefore they have been retrospectively included in the stock area. The TAC for haddock is set for VIIb k, VIII, IX and X. However, official international landings from VIII, IX and X have been less than 2% of all landings in the TAC area in most years since Adult haddock appear to be continuously distributed from the north of Biscay along the Irish coasts and the west of Scotland into the North Sea. It is not clear from their distribution if the VIIb,c,e k stock is distinct from the surrounding areas. Irish otter-trawl lpue in the northernmost rectangles of VIIb is relatively high and similar lpue continues into VIa, suggesting that the haddock in the north of VIIb might belong to the same stock as those in VIa (Gerritsen, 2009). The pattern of lpue in the Irish Sea appears to be relatively distinct from VIIb,c,e k with relatively high otter and beam-trawl lpue in VIIg, low lpue in VIIa-South and high lpue in VIIa north (Gerritsen, 2009). Results from the French EVHOE-WIBTS-Q4 survey suggest that relatively low densities of haddock continue from VIIh into VIIIa. Irish Groundfish Survey (IGFS-WIBTS-Q4) data indicate two distinct nursery areas with high catches of 0-group haddock: one area off the southwest coast of Ireland (VIIb south and VIIj north) and one area off the southeast coast (VIIg north). Catches of older haddock in VIIb are generally low and it is not clear whether the young fish from VIIb move north to VIa or south to VIIj stock (Gerritsen and Stokes, 2006). A.2. Fishery Haddock in Divisions VIIb,c,e k are taken as a component of catches in mixed trawl fisheries. France usually takes about 50 80% of the landings. French landings are made mainly by gadoid trawlers, which prior to 1980 were mainly fishing for hake in the Celtic Sea. Ireland has historically taken about 25 40% of the landings. Fleets from Belgium, Norway, the Netherlands, Spain, and the UK take the remainder of the landings. Landings reported between 1984 and 1995 varied between 2600 t and 4900 t, then increased sharply to t in Since then the landings have varied between 5000 t and t. The vast majority of the landings are taken by otter trawls, most of the remainder of the landings are taken by seines and beam trawls.

226 1596 ICES WGCSE REPORT 2013 A.3. Ecosystem aspects Haddock are widely distributed throughout the stock area across a range of habitats. They have a varied diet but do not appear to be cannibalistic (Needle et al., 2003). The mixed trawl fisheries impacts on benthic communities through bottom contact. Other ecosystem impacts result from discarding of non-target, undersize, over-quota or lowvalue fish. Recruitment of haddock is highly variable. For North Sea haddock, no link could be found between temperature and recruitment (Cook and Heath, 2005). But parental condition has been linked to recruitment success in northwest Atlantic haddock (e.g. Friedland et al., 2003; Marshall and Frank, 1999). B. Data B.1. Commercial catch Sampling and data raising Data on landings-at-age and mean weight-at-age-are available for fleets landing into Ireland since 1993, and from France and the UK since Irish age compositions from VIIgj were used to estimate the age compositions of the international landings. Note that Irish landings contributed around 30% of the international landings so there is considerable uncertainty about the age composition of the landings before The UK landings numbers-at-age are supplied for the combined VIIe k area and the landings data from each Division are used to scale the catch numbers to each Division. French VIIfgh landings numbers are combined with Irish VIIg data to estimate VIIfgh landings numbers. Since 2009, the French landings numbers-at-age are supplied for the whole stock area (VIIb,c,e k). The table below shows the data available and the procedures used to derive quarterly length compositions, age compositions and mean weightsat-age.

227 ICES WGCSE REPORT Data source: Division Data UK France Ireland Belgium Derivation of international landings VII b,c Length composition VII b ALK VII b Age Composition VII b IRL raised Mean weight at age VII b IRL VIIb Landings VIIb,c VIIb,c VII e Length composition VIIe-k Derived from UK VIIe-k ALK VIIe-k Raised to international Landings Age Composition VIIe-k Mean weight at age VIIe-k Landings VIIe VIIe VIIe VII f,g,h Length composition VII f,g,h VII g ALK VII f,g,h VII g Age Composition VII f,g,h VII g IRL & FRA raised Mean weight at age VII f,g,h VII g IRL & FRA raised Landings VIIf,g VIIf,g,h VIIf,g,h,j,k VIIe-h Length composition VIIf,g,h & VIIe ALK Age Composition Mean weight at age Landings VII j-k Length composition VII j IRL raised ALK VII j Age Composition VII j IRL raised Mean weight at age VII j IRL VIIj Landings VIIj,k VIIj,k VIIk VII b,c,e,f,g,h,j,k Length composition ALK Age Composition VIIb,c + VIIe + VIIfgh + VIIjk Mean weight at age Weighted mean by numbers caught Landings Weights-at-age Discard weights were estimated from a fixed length weight relationship (a = ; b = 3.069). This was applied to the discard length distributions-at-age. For the landings weights, length weight relationships were estimated for each year and quarter from the individual weights of the fish that were aged. Landings and discard weights are combined to estimate catch weights. The values are weighted by the numbers-at-age. Quarter-1 catch weights were used as stock weights. If no data were available, quarter 2 weights were used. Previous to the WGSSDS 2004, a three year running average was applied to the stock weights-at-age. In 2004, the working group estimation of stock weights was done using a quadratic function fitted through cohorts to the first-quarter catch weight data. In 2005 the stock weights were modelled using a von Bertalanfy growth equation. The raw stock weight data show significant year effects and although these might be due to changes in sampling or ageing errors, it is also possible that weights-at-age are subject to interannual variation in condition. As the modelled stock weight did not fit the data very well and because it is not clear whether stock weights-atage are more influenced by cohort- or year effects, it was decided in 2007 to revert to using a three year running average to smooth the data, and constraining the weights in older ages to at least those of the preceding age in the cohort. B.2. Biological Natural mortality estimates were derived from mean catch weights-at-age using the approach proposed by Lorenzen (1996). Parameter values were obtained from Table 1 in the Lorenzen paper (ocean ecosystems: α = 3.69; β = -3.05).

228 1598 ICES WGCSE REPORT 2013 Age 0 Age 1 Age 2 Age 3 Age 4 Age 5 Age 6 Age 7 Age Maturity was assumed to be knife-edged at age 2. Recent Irish Survey data is generally in agreement with this maturity ogive, although males occasionally mature at-age one. F and M before spawning were set to 0 for all ages in all years. B.3. Surveys and commercial tuning fleets Description The surveys described below are coordinated by the IBTSWG (International Bottom Trawl Survey Working Group). The French 7fghj EVHOE-WIBTS-Q4 annual groundfish has been carried out since 1997 on the RV Thalassa. Age data are available from 2001 onwards. ALK data from Irish surveys were applied to the EVHOE data for the years to estimate numbers-atage for these years. The sampling design is a stratified random allocation. The number of hauls per stratum is optimised by a Neyman allocation taking into account the most important commercial species in the area (hake, monkfish and megrim). The fishing gear used is a GOV with an average vertical opening of 4 m and a horizontal opening of 20 m. The Irish Groundfish Survey (IGFS-WIBTS-Q4) has been carried out since 2003 and covers VIaS, VIIbgj. This survey is carried out on RV Celtic Explorer. The IGFS has a random stratified design and uses a GOV (with rockhopper in VIa) with a 20 mm codend liner. The two surveys were combined to provide a single index that covers nearly the full stock area. Gerritsen (2012a) describes the justification and for combining the surveys. The two indices are directly combined, weighted by the surface area covered by each survey ( nm 2 for the IGFS and nm 2 for the EVHOE). The combined survey starts in The EVHOE data before 2003 are not used. A French commercial OTB DEF tuning fleet is available but this fleet takes the majority of the landings and is therefore not included as tuning fleet. An Irish commercial OTB fleet is available from 1995 onwards. This fleet is based on the landings and effort from ICES Rectangles 32D9, 31D9, 31E0, 31E1, 31E2, 32E1 and 32E2. These rectangles were selected in order to avoid changes in lpue due to shifts in targeting behaviour. The selected rectangles do not include any major Nephrops or hake, monkfish or megrim fishing grounds or areas with seasonal closures. Consistency The survey shows good internal consistency for ages 0 to 4. The Irish tuning shows good consistency from the age of 2 to 7. However discards are not included in this index and it is not known if discarding patterns have been consistent over time, therefore ages 2 and 3 were not included.

229 ICES WGCSE REPORT B.4. Commercial cpue Effort and lpue data are available from the Irish otter trawl fleets operating in Divisions VIIb, VIIj and VIIg since 1995, French demeral trawlers in VIIfgh since 2004 and effort data are available for the UK beam trawl fleet in VIIe k and all other trawl gears in VIIe k since The effort in the French gadoid fleet has decreased in recent years and is now at a similar level to the Irish and UK fleets. Effort in the Irish OTB VIIg fleet has increased in recent years, while the Irish OTB effort in VIIb and VIIj appears to have levelled off in recent years. The lpue of the French gadoid fleet is still much higher than that of the other fleets. The Irish and UK fleets have seen a minor increasing trend in lpue in recent years. B.5. Other relevant data Discard data Discard data are available from the Irish fleet since Data were raised using effort (hours fished) as auxiliary variable and stratified by ICES division. The number of trips in some years is quite low, leading to concerns about the precision of the data. French discard data are available since These data were also raised using effort (hours fished) as auxiliary variable. Data before 2008 are considered unreliable. Therefore French discards were estimated from the mean discard rate-at-age for the period It was assumed that 90% of one-year-olds, 50% of two-year-olds and 10% of threeyear-olds were discarded. These proportions were applied to the French catch numbersat-age to estimate historic discards. For the period , no French age composition data were available, therefore Irish age composition data were raised to French landings and the discard numbers were estimated from these. French and Irish discard data were combined and a further raising factor was applied to account for discards from other countries. This raising factor was estimated from the total landings of all countries as a proportion of the combined French and Irish landings. This raising factor did not exceed 1.15 in any year. No French age data are available for the discards. Irish age data are available but there are some concerns about the reliability of these data. For this reason, a quarterly length split is applied to the smallest length classes (where the cohorts are quite distinct). For larger fish, quarterly ALKs from the French and Irish landings are used. Length splits applied to the discard data. For lengths where landings ALKs were available, these were used. Country Area Quarter Age 0 Age 1 Age 2 Age 3 Ireland VIIb Ireland VIIgj

230 1600 ICES WGCSE REPORT France VIIbk C. Historical stock development Model used: ASAP; (XSA is also used for quality control purposes; if the two models disagree the differences will need to be explained.) Software used: ASAP V2.0 NOAA Fisheries toolbox ( VPA95 ( FLR with R version with packages FLCore 2.2, FLAssess 2.0.1, FLXSA 2.0 and FLE- DA 2.0 ( ASAP is proposed as the main assessment model. However, due to the short time-series and noisy catch data, it is uncertain whether the separable assumption holds. Therefore it is proposed to also use XSA to monitor if the two models continue to provide similar trends and absolute estimates of SSB and F. C.1. Input data types and characteristics A plusgroup of 8+ was used. Age group 0 was included in the assessment data to allow inclusion of 0-group indices. However, catch numbers and selectivity-at-age 0 were set to zero in all years because catches at this age were very low or zero. Discard estimates are included in the catch numbers and weights, therefore catch is explicitly defined here as landings + discards. Data Year range Catch (tonnes) 1993 current Catch-at-age in numbers (thousands) 1993 current Weight-at-age in the commercial catch (kg) 1993 current Weight-at-age of the stock at spawning time (kg) current Weight-at-age of the stock at January 1 (same as stock weights) 1993 current Proportion of natural mortality before spawning (Lorenzen M) 1993 current Proportion of fishing mortality before spawning (XSA only) 1993 current Variable Age range from year to year 0 8+ Yes 0 8+ Yes 0 8+ Yes 0 8+ Yes 0 8+ Yes 0 8+ No 0 8+ No

231 ICES WGCSE REPORT Proportion mature-at-age 1993 current Natural mortality 1993 current 0 8+ No 0 8+ No C.2. Model Options ASAP Note that ASAP does not accommodate inclusion of data for age 0. Therefore the ages in ASAP are offset by 1 year. All age settings above refer to the real age, not the age group used by ASAP.

232 1602 ICES WGCSE REPORT 2013 Option Include discards separately Use likelihood constant Setting No Yes Mean F (Fbar) age range 3 5 Number of selectivity blocks 1 Fleet selectivity Fixed at 0 for age 0; freely estimated for age 1 and 2, fixed at 1 for ages 3 8+ Discards Index units Index month Index selectivity linked to fleet Included in catch (not specified separately from landings) 2 (numbers) Index age range FR_IR_IBTS: 0-5; IR_GAD: 3 7 Index Selectivity FR_IR_IBTS Index Selectivity - IR_GAD Index CV & ESS FR_IR_IBTS Index CV & ESS IR_GAD Phase for F-Mult in 1st year 1 Phase for F-Mult deviations 2 Phase for recruitment deviations 3 Phase for N in 1st Year 1 Phase for catchability in 1st Year 3 Phase for catchability deviations Phase for unexploited stock size 1 Phase for steepness Catch total CV Input effective sample size Lambda for recruit deviations Lambda for total catch 1 Lambda for total discards Lambda for F-Mult in 1st year Lambda for F-Mult deviations Lambda for index Lambda for index catchability Lambda for catchability devs Lambda N in 1st year deviations Lambda devs initial steepness Lambda devs unexpl stock size FR_IR_IBTS: 11; IR_GAD: 7 (7 = July 1st, the middle of the year) -1 (not linked; the commercial index does not include discards) Fixed at 1 for all ages Freely estimated at age 3, fixed at 1 for all other ages CV 0.3 all years, estimated sample size 40 for all years CV 0.2 all years, estimated sample size 40 for all years -5 (Assume constant catchability in indices) -5 (Do not fit stock-recruitment curve) 0.3 for ; 0.2 for 2008-present (reliable discard data available) 25 for ; 50 for 2002-present (only Irish age comp before 2002) 0 (freely estimated) NA (discards included in catch) 0 (freely estimated) 0 (freely estimated) 1 for both indices in the model 0 for all indices (freely estimated) NA (phase is negative) 0 (freely estimated) NA (phase is negative) 0 (freely estimated) Discards were not included separately because this resulted in undesirable residual patterns. Only one selectivity block was used due to the short time-series, as the time-series gets longer it may be appropriate to allow a separate block for the time period where

233 ICES WGCSE REPORT observed discard data are available. Fleet selectivity was forced to be flat-topped to reduce the number of parameters to be estimated. The F-pattern from XSA indicated flattopped selectivity. XSA Option Setting Ages catch dep stock size None Q plateau 4 Taper No F shrinkage SE 1.5 F shrinkage year range 5 F shrinkage age range 3 Fleet SE threshold 0.3 Prior weights No There is no evidence to suggest that catchability is dependent on stock size; the linear regression fits the data well. The effect of releasing the q-plateau was investigated and catchability appeared to level off at age 4. There is no evidence to suggest that the tuning fleets have changed over time, therefore no tapered time weighting was applied. In recent years there has not been a clear retrospective pattern, therefore a relatively high F shrinkage SE was used with a short year and age range. The fleets are relatively well behaved so an SE threshold of 0.3 was applied. Tuning data: Type Name Year range Age range Survey FR_IR_IBTS 2003 present 0 5 Commercial IR_GAD 1995 present 3 7 D. Short-term projection Model used: Multifleet Deterministic Projection. Landings and discards are modelled as separate fleets. Software used: MFDP1a (

234 1604 ICES WGCSE REPORT 2013 Option Initial stock size Natural mortality Setting Long-term GM (omitting last two years) Stock numbers-at-age 1 and older from model Lorenzen M, as in model Maturity Knife-edged at age 2 F and M before spawning Stock / catch weights-at-age Exploitation pattern Intermediate year assumptions Stock recruit model Fbar range 5 5* Rescale to last year 0 for all ages in all years Average last 3 years Average last 3 years F in the last year check retrospective pattern for evidence of bias None, long-term GM recruitment (omitting last two years) No * The FBAR age range used in the assessment model outputs is 3 5 this F refers to the catch (including discards). Ages 3-5 are fully selected in the catch (but not landings). MFYPR output supplies YPR based on landings F. In order to compare (landings) F reference points with the (catch) FBAR it was decided to calculate FBAR only for age 5 because at this age the catch and landings are both fully selected and because a flattopped selection pattern was applied in ASAP the result will be correct. So, in this context FMAX refers to the catch F where the landings per recruit are maximised. E. Medium-term projections None. F. Yield and biomass per recruit No stock recruit relationship exists for this stock; recruitment is characterised by sporadic extreme recruitment events. Software used: NOAA fisheries toolbox YPR V3.0. Option Stock / catch weights-at-age Selectivity Natural mortality Setting Average last three years Average last three years Lorenzen M, as in model Maturity Knife-edged at age 2 G. Biological reference points No reference points have been defined for this stock. The following results from the analyses by WKROUND could be informative: FMAX (landings) = 0.28 F0.1 (landings) = 0.19 FMSY = FMAX = 0.28 Bloss = 7500 tonnes H. Other issues None.

235 ICES WGCSE REPORT I. References Cook, R.M. and Heath, M.R The implications of warming climate for the management of North Sea demersal fisheries. ICES JMS 62: doi: /j.icesjms Friedland K D, Hare JA. Wood GB. Col LA, Buckley LJ, Mountain DG, Kane J, Brodziak J, Lough RG, Pilskaln CH Does the fall phytoplankton bloom control recruitment of Georges Bank haddock, Melanogrammus aeglefinus, through parental condition? Canadian Journal of Fisheries and Aquatic Sciences, Volume 65, Number 6, pp Gerritsen, H Spatial distribution of Irish Landings, Effort and LPUE of Demersal Stocks in Divisions VI and VII in Working Document 1 to ICES WGCSE, Copenhagen May Gerritsen H. 2012a. Haddock 7b k Combined IGFS-WIBTS-Q4 and EVHOE_WIBTS-Q4 surveys. Working document 12 ICES WKROUND, Aberdeen Febuary Gerritsen H. 2012b. Irish haddock discards in VIIb k. Evaluation of raising methods. Working document 13 ICES WKROUND, Aberdeen Febuary Gerritsen H and Stokes D Stock structure of Haddock in VIIb k: Information from the Irish Groundfish Surveys Working document 7. ICES WGSSDS, Copenhagen, 27 June 6 July Lorenzen K The relationship between body weight and natural mortality in juvenile and adult fish: a comparison of natural ecosystems and aquaculture. Journal of Fish Biology 49, Marshall CT, Frank, KT The effect of interannual variation in growth and condition on haddock recruitment. Canadian Journal of Fisheries and Aquatic Sciences, 56: Needle CL, O Brien CM, Darby CD, and Smith MT Incorporating time-series structure in medium-term recruitment projections. Sci Mar 67 (suppl 1):

236 1606 ICES WGCSE REPORT 2013 Stock Annex 7.5: Nephrops FU17, Aran Grounds Stock Aran Grounds Nephrops (FU17) Date 06 March 2009 (WKNEPH 2009) Revised by Colm Lordan (WGCSE, 2011 to address RGCSE 2010 comments) A. General A.1. Stock definition Nephrops is limited to muddy habitat, and requires sediment with a silt and clay content of between % to excavate its burrows, and this means that the distribution of suitable sediment defines the species distribution. Adult Nephrops probably only undertake very small scale movements (a few 100 m) but larval transfer may occur between separate mud patches in some areas. In FU17, the main Nephrops stock inhabits an extensive area of muddy sediment known as the Aran Grounds which lie to the west and southwest of the Aran Islands; there are also smaller discrete mud patches in Galway Bay and Slyne Head. A.2. Fishery In recent years the Nephrops stock in FU17 are almost exclusively exploited by Irish vessels. Figure A.2.1 shows the spatial distribution of landings and lpue for Irish otter trawl vessels in 2005 using logbook and VMS data linked together to give finer spatial resolution. The Aran ground fishery is clearly highlighted. The Nephrops fishery at the back of the Aran Islands can be considered the mainstay of the Ros a Mhíl fleet. Without this Nephrops fishery the majority of vessels in the fleet would cease being economically viable (Meredith, 1999). The Irish fishery consists of entirely of otter trawl vessels. The majority of vessels use twin-rigs and 80 mm. Smaller vessels do use 70 mm with a SMP. Some vessels have using 90 mm. Vessels from Ros a Mhíl, Dingle, Union Hall, Dunmore East, Clogherhead and Kinsale mainly exploit the fishery. The number of Irish vessels reporting Nephrops landings from FU17 has fluctuated around 50/yr (Figure A.2.2). Around 18 vessels report landings in excess of10 t. These are the main vessels in the fishery accounting for around 85% of the total landings. The majority of these vessels are between m overall length (Figure A.2.3). There has been a slight shift to lager vessels over time. The majority of vessels are in the power range of KW (Figure A.2.4). There has also been a shift to more powerful vessels over time with the introduction of twin-rigs to the fishery in the early 2000s. Most of the larger boats move freely between the Nephrops fisheries in FUs 15, 16, and other areas depending on the tides and weather. The fishery shows a distinctive seasonal pattern with highest landings, catches, lpue and cpue in April June and October November. The monthly landings time-series with the average pattern is shown in Figure A.2.5. The first period of elevated landings is associated with the emergence of females from their burrows post hatching of their eggs. The

237 ICES WGCSE REPORT sex ratio during this period is biased towards females (Figure A.2.6). Females mature quickly during the early summer and spawning occurs in July and August. This is coincident with a decline in landings and cpue in the fishery. The Ros a Mhíl fleet traditionally tie up in August each year for maintenance and refurbishment. The following TCMs are in place for Nephrops in VII (excluding VIIa) after EC 850/98: Minimum Landing Sizes (MLS); total length >85 mm, carapace length >25 mm, tail length >46 mm. Mesh Size Restrictions; Vessels targeting Nephrops using towed gears having at least 35% by weight of this species on board will require 70 mm diamond mesh plus an 80 mm square mesh panel as a minimum or having at least 30% by weight of Nephrops on board will require mm diamond mesh. A.3. Ecosystem aspects Physical oceanography The Aran Ground is coincident with a pool of oceanic water, which is rich in nutrients and low in dissolved oxygen. The currents throughout the water column over the ground are generally weak although there is a well-documented bottom density front on the eastern flank of the ground (Nolan and Lyons, 2006). This is a seasonal feature, which establishes in May and persists until autumn. The front causes a persistent jet like flow from south to north close to the seabed through the Nephrops ground. The mean position of jet varies from year to year by up to 30 km. Timing and position of the jet may influence recruitment and settlement success of post-larval Nephrops since it could potentially advect larval from the area. Salinity differences, due to overwinter freshwater input, are thought to heavily influence the density structure and location of this front. Until a time-series of recruitment and jet dynamics is established it is not possible to draw any firm conclusions about the impact of this ecosystem feature on the stock and fishery. Potential sinks for advected larvae include Slyne head and possibly Galway Bay. Temperature and salinity time-series An emerging time-series of temperature and salinity data are available for a transect through the Aran Grounds (Nolan and Lyons, 2006). In all years since 1999 (except 2001) the 53 N section has exhibited positive anomalies in temperature of between 0.2 C and 2 C (Figure A.3.1). In 2001, the temperature anomaly from the long-term climatology was zero. Years with lower temperature anomalies seem to coincide with years of strongly negative salinity anomalies (e.g and 2005, 2006) perhaps reflecting the limited influence of ENAW on the section in those years as the section is dominated by coastal discharges from the Loire and Shannon. Salinity anomalies along 53 N range from -0.3 to +0.1 psu over the period. The freshest years were 2001, 2005 and In 2000, 2003 and 2004 ENAW has a stronger influence on the salinity structure and positive anomalies in salinity from the long-term climatology are the result. The higher UWTV abundance in 2003 and 2004 is coincident with the warmest anomaly but the time-series remains too short to draw definitive conclusions. Sediment distribution There is a growing body of information on the spatial extent of the sediment suitable for Nephrops from UWTV surveys, seabed mapping programmes and the fishing industry.

238 1608 ICES WGCSE REPORT 2013 Figure A.3.1 depicts contour and post plots of the a) mean size (phi) and classification based on the Friedman and Sanders (1978) scales and b) sorting (σg) of the sediments on the Aran Grounds based on PSA results from samples collected from UWTV surveys. The majority of the ground has similar mean particle size at around 4 5 µm. There are some patches of softer silt towards the middle of the ground. Figure A.3.2 is bathymetry of the Aran grounds obtained from seabed mapping programmes. The eastern flank of the ground shallows up quickly but the majority of the ground is gradually deepening from around 100 m to 110 m with the deepest parts to the southwest. B. Data The table below summarises the available data for this stock and attempts to quantify the quality subjectively. Data Source Fishery Dependent Survey Units Landings Data Tonnes Effort Data Hrs (uncorrected) Capacity Number & Power of Vessels Standardised Effort Data Effective effort (Hrs& Capacity) Commercial LPUE Kg/Hrs Commercial CPUE Kg/Hrs Landings Size distributions (mm) Catch size distributions (mm) Sex Ratio in Landings % Sex Ratio in Catch % Maturity Data % IBTS Trawl survey catch size distributions (mm) Commercial Trawl survey CPUE & size Kg/Hrs & (mm) UWTV survey Abundance numbers UWTV -Beam size distributions (mm) Unreliable Potentially poor quality Good B.1. Commercial catch Prior to 1988 landings data for this fishery are only available to the WG for France. Since 1988 reported landings data for the Irish fleet were obtained from EU logbooks. The quality of landings data is not well known. In earlier, years there are no landings from Ireland although there was probably some catch. The Irish landings have been close to quota for this TAC area since around 1997 (Figure B.1.1). In more recent years ( and 2008) there are a few observations of both under and over reporting but it is not possible to correct landings using these as it is not known how representative they might be. Landings length and sex compositions were estimated from port sampling by Ireland (between ). There was a perception during this period that that discarding was not significant. In 2002 a new catch self-sampling programme was put in place. This involves unsorted catch and discard samples being provided by vessels or collected by observers at sea on discard trips. The catch sample is partitioned into landings and discards using an onboard discard selection ogive derived for the discard samples (Table B.1.1). Sampling effort is stratified monthly but quarterly aggregations are used to derive length distributions and selection ogives. The length weight regression parameters given in Table B.2.1 are used to calculate sampled weights and appropriate quarterly raising factors. The sampling intensity and coverage has varied over the time-series (Table B.1.1). The quality of the sampling has not yet been qualitatively assessed in terms of precision and accuracy.

239 ICES WGCSE REPORT Nephrops landings and discards from the Aran Grounds have not been sampled for the majority of 2006 and all 2007 due to a lack of co-operation by the industry. However, sampling resumed in 2008 and the intensity and coverage is considered the best to date. Fish and other bycatches in the fishery have been collected by onboard observers since The number of trips is variable over time with a gap in the series in 2006 and B.2. Biological Biological parameters for this stock are outlined in Table B.2.1. Length weight Mean weights-at-age for this stock are estimated from studies on Scottish stocks by Pope and Thomas (1955). This relationship was examined in 2003 and it seemed appropriate. Given the variability in length weight parameters found in Allan et al., 2009 it would be worth monitoring these more closely in the future. Natural mortality A natural mortality rate of 0.3 was assumed for all age classes and years for males and immature females, with a value of 0.2 for mature females. The lower value for mature females reflects the reduced burrow emergence while ovigerous and hence an assumed reduction in predation. The accuracy of these assumptions is unknown. Cod are not common on the Aran Grounds but other potential predators include dogfish, monkfish megrim and gurnards. Stomach contents data on the Irish GFS could be used to examine this in the future. Maturity The L50 of females using a macroscopic visual maturity scale is known to vary depending sampling month (Lordan and Gerritsen, 2006). The L50 in July was chosen as the most appropriate estimate given the maturity schedules observed (Figure B.2.1). It is worth mentioning that commercial vessel surveys in November 2001 and in June 2002 demonstrated considerable differences between the maturity schedules of female Nephrops sampled in shallower waters of Galway Bay compared with the Aran Grounds. Proportion of F and M prior to spawning was specified as zero to give estimates of spawning stock biomass at January 1. In the absence of independent estimates, the mean weights-at-age in the total catch were assumed to represent the mean weights in the stock. Discard survival Given the trip durations (~five days average) and behaviour of the fleet the majority of discards on the Aran Grounds are returned to the sea over suitable sediment. The proportion scavenged by birds is probably quite low. Tow durations, volume of catches, prolonged sorting on deck and relatively high density of Nephrops on the seabed probably results in relatively low discard survival. This is estimated to be around 10%.

240 1610 ICES WGCSE REPORT 2013 B.3. Surveys Since 2002 Ireland has conducted underwater television survey (UWTV) annually on the main Nephrops grounds - Aran grounds. Indicator camera stations are also carried out on the adjacent grounds of Galway Bay and Slyne Head weather and time permitting. The surveys were based on a randomised fixed grid design. The methods used during the survey were similar to those employed for UWTV surveys of Nephrops stocks around Scotland and elsewhere and are documented by WKNEPHTV (ICES, 2007). A number of factors are suspected to contribute bias to the surveys. In order to use the survey abundance estimate as an absolute it is necessary to correct for these potential biases. The history of bias estimates are given in the following table and are based on simulation models, preliminary experimentation and expert opinion, the biases associated with the estimates of Nephrops abundance in the Aran Grounds are: species Time period Edge effect detection rate identification occupancy Cumulative bias FU17: Aran <= B.4. Commercial cpue Prior to 1988 landings data for this fishery are only available to the WG for France. Since 1988 reported landings data for the Irish fleet were obtained from EU logbooks (Table B.4.1). Effort data for FU17 is available from 1995 for the Irish otter trawl Nephrops directed fleet (Table B.4.2). A threshold of 30% of Nephrops in reported landings by trip is used to identify the catches and effort of this fleet. This threshold was based on an analysis of the trip-by-trip catch compositions. In 2007 this fleet accounted for ~90% of the landings and compared with an average of 70% over the time period. These data have not been standardised to take into account vessel or efficiency changes during the time period. Landings per unit of effort (lpues) have been fluctuating around an average of 39 kg/hr with an increasing trend since 2004, to the highest observed (59 kg/hr) in the time-series in 2007 (Figure B.4.1). B.5. Other relevant data C. Historical stock development Age structured XSA assessment for this stock was carried Nephrops WG in 2003 (ICES, 2003). The results were considered unreliable for several reasons most importantly; inadequate historical sampling of catch, growth and natural mortality assumptions and concern about accuracy of tuning data. Since then the focus has been on developing a time-series of UWTV survey data as the basis of assessment and advice for this stock. The 2009 Benchmark decided on the following procedure: 1 ) Survey indices are worked up annually resulting in the TV index. 2 ) Adjust index for bias (see Section B.3). The combined effect of these biases is to be applied to the new survey index.

241 ICES WGCSE REPORT ) Generate mean weight in landings. Check the time-series of mean landing weights for evidence of a trend in the most recent period. If there is no firm evidence of a recent trend in mean weight use the average of the three most recent years. If, however, there is strong evidence of a recent trend then apply most recent value (don t attempt to extrapolate the trend further in the future). D. Short-term projection 1 ) The catch option table will include the harvest ratios associated with fishing at F0.1 and FMAX. These values have been estimated by the benchmark workshop (see Section 9.2) and are to be revisited by subsequent benchmark groups. The values are FU specific and have been put in the stock annexes. 2 ) Create catch option table on the basis of a range of harvest ratios ranging from 0 to the maximum observed ratio or the ratio equating to FMAX, whichever is the larger. Insert the harvest ratios from step 4 and also the current harvest ratio. 3 ) Multiply the survey index by the harvest ratios to give the number of total removals. 4 ) Create a landings number by applying a discard factor. This conversion factor has been estimated by the benchmark workshop and is to be revisited at subsequent benchmark groups. The value is FU specific and has been put in the stock annex. 5 ) Produce landings biomass by applying mean weight. The suggested catch option table format is as follows. Implied fishery Harvest rate Survey Index Retained number Landings (tonnes) 0% % " % " % " % " F % " % " % " Fmax 13.50% " % " % " % " % " % " Fcurrent 21.5% "

242 1612 ICES WGCSE REPORT 2013 E. Medium-term projections None presented. F. Long-term projections None presented. G. Biological reference points The time-series of available length frequencies were insufficient to generate reliable estimates of F0.1 and FMAX. H. Other issues None. I. References Fernand, L., Nolan, G.D., Raine, R., Chambers, C.E., Dye, S.R., White, M. and Brown, J The Irish coastal current: A seasonal jet-like circulation. Continental Shelf Research, Vol. 26, Issue 15, ICES Report of the Workshop on Nephrops Stocks. Annex 6: Working Document by Lordan and Gerritsen. ICES CM 2006/ACFM:12. ICES Report of the Workshop on the use of UWTV surveys for determining abundance in Nephrops stocks throughout European waters (WKNEPHTV). ICES CM: 2007/ACFM: 14 Ref: LRC, PGCCDBS. ICES Report of the Workshop and training course on Nephrops Burrow Identification (WKNEPHBID). ICES CM: 2008/LRC: 03 Ref: ACOM. ICES Report of the Working Group on the Assessment of Southern Shelf Stocks of Hake, Monk andmegrim (WGHMM). ICES CM: 2008/ ACOM:07. ICES Using UWTV surveys to develop a conceptual ecosystem model of Aran Grounds Nephrops population distribution. ICES CM 2005/L:30 Annual Science Conference. Colm Lordan and Hans Gerritsen The accuracy and precision of maturity parameters from sampling of female Nephrops from stocks around Ireland. WD6 in the Report of the Workshop on Nephrops stocks. ICES CM 2006/ACFM:12. Nolan,G.D. and Lyons Ocean climate variability on the western Irish Shelf, an emerging time series., K., Proceedings of the ICES Annual Science Conference, Theme Session C, C:28.

243 ICES WGCSE REPORT Table B.1.1. Nephrops in FU17 (Aran Grounds) Landings and discard numbers by year and sex. Female Numbers '000s Male Numbers '000s Both sexes Year Landings Discards Landings Discards % Discard ,665 12,161 29,949 13,250 34% ,105 9,374 31,256 8,326 25% ,530 9,577 29,538 8,744 29% ,109 7,068 12,930 4,282 28% ,280 11,383 21,828 8,967 33% No Sampling Table B.2.2. Numbers of samples and numbers measured for the FU17 Nephrops Stock by year. Number of Samples Total numbers of Nephrops measured Year Graded Landings Catch Discards Year Graded Landings Catch Discards

244 1614 ICES WGCSE REPORT 2013 Table B.2.1. Biological Input Parameters for FU17 Nephrops Stock. Parameter Value Source Discard Survival 10% WKNEPH 2009 MALES Growth - K 0.16 based on FU15 Growth - L(inf) 60 based on FU15 Natural mortality - M 0.3 assumed, in line with other stocks Length/weight - a based on Scottish data (Pope and Thomas, 1955) Length/weight - b " FEMALES Immature Growth Growth - K 0.16 based on FU15 Growth - L(inf) 60 based on FU15 Natural mortality - M 0.3 assumed, in line with other stocks Size at maturity (L50) 22 ICES 2006 (Lordan and Gerritsen) Mature Growth Growth - K 0.1 based on FU15 and FU16 Growth - L(inf) 56 based on FU15 Natural mortality - M 0.2 assumed, in line with other stocks Length/weight - a based on Scottish data (Pope and Thomas, 1955) Length/weight - b "

245 ICES WGCSE REPORT

246 1616 ICES WGCSE REPORT 2013 Figure A.2.1. Effort, catch and catch per unit of effort for Nephrops, Irish otter trawlers in The boxed and zoomed in plots show a zoomed in view of landings and lpue from the fishery on the Aran Ground Numbers of vessels All Vessels Vessels Landing >10t\year Years Figure A.2.2. Time-series of the number of Irish vessels reporting landings of Neprhrops from FU17. The vessels with annual landings >10 t/yr can be considered the main participants in the fishery these general account for ~85% of the total landings.

247 ICES WGCSE REPORT Frequency Frequency Frequency Vessel Length Vessel Length Vessel Length Frequency Frequency Frequency Vessel Length Vessel Length Vessel Length Frequency Frequency Frequency Vessel Length Vessel Length Vessel Length Frequency Frequency Frequency Vessel Length Vessel Length Vessel Length Frequency Vessel Length Figure A.2.3. The time-series of length distributions of Irish vessels landing >10 t of Nephrops from FU17.

248 1618 ICES WGCSE REPORT 2013 Figure A.2.4. Box plot of the time-series of vessel power in KW of Irish vessels landing >10 t of Nephrops from FU17. Figure A.2.5. Monthly landings of Nephrops from FU17 from The inset shows the average pattern for all years.

249 ICES WGCSE REPORT % of Catch Males 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Month Fem Pale Fem Med Fem Dark Fem Eggs % of Fem at each maturity stage 100% 80% 60% 40% 20% 0% Month Figure A.2.6. The upper panel shows the sex ratio in sampled catches (error bars = 95% confidence intervals). The low panel shows the female maturity schedule i.e. percentage at each maturity stage by month.

250 1620 ICES WGCSE REPORT 2013 Figure A.3.1. Anomalies in temperature (upper panel) and salinity (lower panel) for the 53 N section running through the Aran Grounds ( ).

251 ICES WGCSE REPORT a) Coarse Silt Very coarse Silt Very fine Sand Fine Sand Medium Sand Coarse Sand b) SORTING LEGEND 0-1 WELL TO MODERATELY SORTED 1-2 POORLY SORTED >2 VERY POORLY SORTED Figure A.3.1. Contour and post plots of the a) mean size (phi) and classification based on the Friedman and Sanders (1978) scales and b) sorting (σg) of the sediments on the Aran Grounds based on PSA results from samples collected from

252 1622 ICES WGCSE REPORT Decimal Degrees North Decimal Dergees West Figure A.3.2. The bathymetry of the Aran grounds Ireland Landings Quota Landings (Tonnes) Figure B.1.1. Nephrops landings and quota for Ireland since the introduction of TACs in 1987.

253 ICES WGCSE REPORT P mature Jan L50= () 2 samples; n=19 P mature Functional Unit 17 Feb Mar L50=36.9 (5.5) 2 samples; n=40 P mature P mature Apr Length (mm) Length (mm) Length (mm) Length (mm) P mature May L50=19.8 (0.7) 4 samples; n=13 P mature Jun L50=20.4 (0.3) 2 samples; n=62 P mature Jul L50=22.3 (0.1) 8 samples; n=31 P mature Aug Length (mm) Length (mm) Length (mm) Length (mm) P mature Sep P mature Oct P mature Nov L50=29.2 (0.5) 2 samples; n=33 P mature Dec L50=30 (1.1) 5 samples; n= Length (mm) Length (mm) Length (mm) Length (mm) Figure B.2.1. Female proportions mature-at-length for FU17. The 95% confidence limits of the proportions mature-at-length are indicated by the vertical bars. The black curve indicates the model and its standard errors are given by the blue lines. The L50 is the estimated length at 50% maturity and its standard error is given between brackets. Blank plots indicate no sampling took place.

254 1624 ICES WGCSE REPORT 2013 Stock Annex 7.6: Nephrops in FU 16 Porcupine Bank Working Group WGCSE 2010 Date Version 1 04/05/2010 Revised by Jennifer Doyle A. General A.1. Stock definition The Functional Unit for assessment includes some parts of the following ICES Divisions VIIb,c,j,k. The exact stock area is shown on the map below includes the following ICES Statistical rectangles: D5 D6; D7 D J A.2. Fishery France The French fleet fishing Nephrops in FU 16 also fishes in Division VIIg h and was described in detail in the 1999 WGNEPH report (ICES, 1999a). The French fleet only lands large Nephrops from this FU. Investigation of the landings data by statistical rectangle carried out by WGNEPH in These indicated that the majority of the French landings between were from the south of the Porcupine Bank. Ireland The fishery is mainly seasonal taking place mainly between April and July; landings for the remainder of the year are minimal. Most of the Irish vessels are multipurpose trawlers and are relatively large (between 20 and 35 m in total length). Irish vessels land both

255 ICES WGCSE REPORT whole prawns and tails depending on markets from this FU and the sizes of the Irish landings are significantly smaller than those for the French and Spanish fleets. The Irish vessels are mainly using twin-rig trawls. Fishing is often weather dependent (particularly for the smaller vessels), with trip duration varying between seven and ten days. Investigation of the landings data by statistical rectangle provided to the WGNEPH in 2002 indicates that the majority of the Irish landings between 1995 and 2001 were from the south central area of the Porcupine Bank. The recent spatial distribution of the fishery is shown in Figure 1. Spain The Spanish fishery in the Porcupine area is a typical multispecies fishery, targeting different demersal species, amongst which Nephrops. The fleet, which consists of about 35 vessels, is composed of side-trawlers and is part of the so-called 300 fleet in the Adhesion Treaty of Spain to the EEC in Within the Porcupine fleet, two components can be distinguished: one consisting of vessels fishing with finfish trawls (average engine power 980 hp), and the other fishing with Nephrops trawls (average engine power 680 hp). The average duration of their trips is 15 days, of which are actual fishing days. The major landing port is La Coruña. The target species for the finfish directed fleet are hake, megrim and anglerfish, with Nephrops as a valued bycatch. Vessels fishing with Nephrops trawls are much more directed towards Nephrops (especially in spring and summer), and fish is a bycatch. These two fleets not are currently disaggregated in the time-series. A.3. Ecosystem aspects Productivity of deep-water Nephrops stocks is generally lower that those on the shelf although individual Nephrops grow to relatively large sizes. A persistent Taylor column circulation around Porcupine Bank provides an important mechanism for the retention of pelagic eggs and larvae of the various marine species spawning in the area. (Mohn et al., 2002). The Nephrops stock on the Porcupine Bank are distributed on mud patches in relatively deep waters m. It is not known how larvae are retained over these grounds but the Taylor column may help with larval retention.

256 1626 ICES WGCSE REPORT 2013 Figure 1. The spatial distribution of LPUE of Nephrops caught by Irish otter trawlers between derived using integrated VMS and logbook records.

257 ICES WGCSE REPORT B. Data B.1. Commercial catch Commercial catch and effort data is supplied by Ireland, France, Spain and the UK. These are the countries exploiting the stock. B.2. Biological BIOLOGICAL PARAMETERS Parameter Value Source Discard Survival Discards considered negligible MALES Growth - K based on values in other areas (Anon., 1991) Growth - L(inf) 75 based on maximum sizes observed in samples Natural mortality - M 0.2 Anon.1990 (estimated) Length/weight - a based on Celtic Sea (FU 20 22) Length/weight - b " FEMALES Immature Growth Growth - K Not applicable Growth - L(inf) 75 Natural mortality - M 0.2 Size at maturity 26.2 Fariña and González Herraiz (2001) Mature Growth Growth - K Anon., 1991 Growth - L(inf) 60 based on maximum sizes observed in samples Natural mortality - M 0.2 As for males Length/weight - a " Length/weight - b " B.3. Surveys The only fishery-independent source of data the Spanish Porcupine trawl survey which commenced in Further information on thus survey is provided in the IBTS report (ICES, 2010) and in previous IBTS reports.

258 1628 ICES WGCSE REPORT 2013 Abertura vertical Abertura puertas Tiempo llegada fondo Profundidad (m) P04 P05 P06 P07 P08 Profundidad (m) P04 P05 P06 P07 P08 Profundidad (m) P04 P05 P06 P07 P m m min m m min P04 P05 P06 P07 P08 P04 P05 P06 P07 P08 P04 P05 P06 P07 P08 Campañas Campañas Campañas Figure 2. Door spread, vertical opening and time to settle on the ground between 2004 and 2008.

259 ICES WGCSE REPORT Nephrops norvegicus P01 P02 P kg 8 kg 8 kg P04 P05 P kg 8 kg 8 kg P07 P08 P kg 8 kg 8 kg Figure 3. Distribution of Nephrops norvegicus catches in biomass in Porcupine surveys between 2001 and B.4. Commercial lpue The Nephrops fishery on the Porcupine Bank is both seasonal and opportunistic with increased targeting during periods of high Nephrops emergence and good weather. Effort and lpue data are not standardised, and hence do not take into account vessel capabilities, efficiency, seasonality or other factors that may bias perception of lpue abundance trend over the longer term. The available effort time-series are summarized below:

260 1630 ICES WGCSE REPORT 2013 COUNTRY FIRST YEAR OF UNITS COMMENT EFFORT DATA France 1983 Hours For trips where Nephrops constituted 10% of the landed value Ireland 2005 Hours For trips where Nephrops constituted 30% of the landings in weight Spain 1971 ay*bhp/100 (x1000) Only commercial landings data are available for all countries involved in the fishery. B.5. Other relevant data C. Historical stock development An experimental age-structured assessment for this stock was carried out by the Nephrops WG in 1993 (ICES, 1993), in 2003 (ICES, 2003) and by the WGHMM (ICES, 2005) in all cases the assessments being considered inadequate. This conclusion was based on poor quality, and unexplainable inconsistencies in the input data. Unknown growth rates and concern about the utility of age-based assessment models impeded progress to an accepted assessment. In additional the lack of a time-series of reliable standardised cpue data was also perceived as a problem. This problem has been solved with the developing Porcupine trawl survey-series. Model used: XSA, LCA Software used: n/r Model Options chosen: No Final model was accepted G. Biological reference points No reference points have been proposed or used for this stock. H. Other issues None. I. References Gerritsen, H Working Document 1-ICES Working Group for the Celtic Seas Ecoregion13 19 May 2009.

261 ICES WGCSE REPORT Stock Annex 7.7a: Nephrops FU 20 (Labadie, Baltimore and Galley), FU 21 (Jones and Cockburn) To be updated.

262 1632 ICES WGCSE REPORT 2013 Stock Annex 7.7b: Nephrops FU 22 (the Small) To be updated.

263 ICES WGCSE REPORT Stock Annex 7.8: Nephrops FU19 (South and Southwest Ireland) To be updated.

264 1634 ICES WGCSE REPORT 2013 Stock Annex 7.9: Plaice in Division VIIb,c (West of Ireland) Stock Working Group Plaice in Division VIIb,c (West of Ireland) WGCSE Date April 2013 Revised by WGCSE / Hans Gerritsen A. General A.1. Stock definition There are two distinct areas in which plaice are caught by Irish vessels in VIIb: an area to the west of the Aran Islands and an area in the north of VIIb which extends into VIa (the Stags and Broadhaven Ground). Plaice in the north of VIIb appear to be more linked with VIa than populations further south. It is not known how much exchange there is between plaice on the Aran grounds and those on the Stags ground. A.2. Fishery Plaice in VIIb are mainly caught by Irish vessels on sandy grounds in coastal areas. Plaice catches in VIIc are negligible. During a large proportion of the VIIbc plaice landings were taken from the Stags Grounds (Rectangles 37D8, 37D9, 37E0 and 37E1). The landings and lpue in this area have dropped sharply since 2000, in line with a general decrease of lpue of plaice in Division VIa. The landings and lpue on the Aran grounds appear to have been more or less stable since the start of the logbooks timeseries in 1995 (WD 1, WGCSE 2009). A.3. Ecosystem aspects B. Data B.1. Commercial catch Landings data are available. B.2. Biological Due to the negligible volume of landings, no sampling of this stock takes place. B.3. Surveys The IGFS-WIBTS-Q4 covers VIIb but the gear is not particularly suited for flatfish and catchability is very variable. B.4. Commercial cpue The commercial lpue time-series may not be reflective of overall stock abundance due to changing fishing practices.

265 ICES WGCSE REPORT B.5. Other relevant data C. Assessment: data and method Model used: Depletion-Corrected Average Catch (DCAC; MacCall, 2009) Software used: NOAA Fisheries Toolbox: DCAC.exe Model Options chosen: Because the value of the depletion delta parameter is unknown, a range of values were used (10%, 50% and 90%; delta is the difference in biomass in the first year and biomass in the last year as a proportion of the virgin biomass (unfished vulnerable abundance). All other settings are based on default values and recommendations from MacCall (2009). LAND CV M STDEV FMSY/M STDEV BMSY/B0 STDEV DELTA STDEV lognormal bounded 0 1 beta 0.1, 0.5, bounded 0 1 beta Input data types and characteristics: Because average catch is analysed, the year range chosen can have a large influence on the results. Two year ranges were tested: 1950 present (the time period after WWII when the stock was heavily exploited) and 1995 present (the time period when the landings showed a declining trend). D. Short-term projection None. E. Medium-term projections None. F. Long-term projections None. G. Biological reference points Not set. H. Other issues None. I. References MacCall, AD Depletion-corrected average catch: a simple formula for estimating sustainable yields in data-poor situations. ICES J Mar Sci 66:10 p

266 1636 ICES WGCSE REPORT 2013 Stock Annex 7.10: Stock Working Group Celtic Sea Plaice Plaice (Division VIIf&g) Celtic Seas Ecoregion Date May 2013 By Jonathan Gillson A. General A.1. Stock definition The degree of separation between the stocks of plaice in the Celtic Sea and the Irish Sea is unclear. Historic tagging studies indicate a southerly movement of mature fish (or fish maturing for the first time) from the southeast Irish Sea, off North Wales, into the Bristol Channel and Celtic Sea during the spawning season (Figure A1). While some of these migrant spawning fish will remain in the Bristol Channel and Celtic Sea, the majority are expected to return to summer feeding grounds in the Irish Sea (Dunn and Pawson, 2002). Very little mixing is considered to occur between the stocks (Pawson, 1995). Nevertheless, time-series of recruitment estimates for all stocks in waters around the UK (Irish Sea, Celtic Sea, western and eastern Channel, North Sea) show a significant level of synchrony (Fox et al., 2000). This could indicate that the stocks are subject to similar large-scale environmental forces and respond similarly to them.

267 ICES WGCSE REPORT Figure A1. Principal substock areas and movements of plaice on the west coast of England and Wales. Percentages are the recaptures rates of tagged plaice <25 cm total length when released, and >26 cm when recaptured in English and Welsh commercial fisheries. Tagging exercises in and were combined based on the assumption that the dispersal patterns of plaice were consistent over time. For each substock, the main feeding area (derived from tag recaptures during April December; light shading), and the main spawning area (derived from tag recaptures during January March, and ichthyoplankton surveys; dark shading) are indicated. The substocks tagged have been coloured green, red and blue. The substocks coloured orange are less well determined, with the feeding area around south-east Ireland unknown. Letters represent return migrations, where A 6%, and B+C 46%. Reproduced from Dunn and Pawson (2002). A.2. Fishery The main fishery is concentrated on the Trevose Head ground off the north Cornish coast and around Land s End. Although plaice are taken throughout the year, the bulk of landings are taken in March, after the peak of spawning, with a second peak in September. The fisheries harvesting plaice in the Celtic Sea mainly involve vessels from Belgium, France, England and Wales. A.3. Ecosystem aspects Plaice are preyed upon and consume a variety of species through their life history. However, plaice have not as yet been included in an interactive role in multispecies assessment methods (e.g. ICES, WGSAM 2008). Among other prey items, plaice typically consume high proportions of polychaetes and molluscs. Other than statistical correlations between recruitment and temperature (Fox et al., 2000), little is known about the effects of the environment on the stock dynamics of plaice in the Irish Sea. Negative correlations between year-class strength of plaice (in either the Irish

268 1638 ICES WGCSE REPORT 2013 Sea, Celtic Sea, Channel and North Sea) and sea surface temperature are generally strongest for the period February June. However, western (North Sea and Channel) and eastern (Irish Sea and Celtic Sea) stocks have been found to respond to different timescales of temperature variability, which might imply that different mechanisms are operating in these stocks and/or that the Irish Sea and Celtic Sea share common spawning (Fox et al., 2000). B. Data B.1. Commercial catch Landings International landings-at-age data based on quarterly market sampling and annual landings figures are available from Landings rose to a maximum in the late 1980s, declined during the early 1990s, and then fluctuated around 1000 t. The decline reach a low at 390 t in 2005 following which there has been a gradual increase. Estimates of the level of discarding have been collected since 2004 and have shown a consistent increase, apart from 2007 when a substantial increase occurred by all fleets, followed by a return to the previously lower levels. For the period 1991 to 2005 quarterly age compositions have typically represented around 70% of the total international landings, though in 2002 this fell to around 25% when age compositions were not available for the Belgian fleet. Belgian age sampling in 1993 was at a reduced level and was augmented with UK data. There was no UK sampling in the 4th quarter of 1994 and landings of 1 year olds by the UK otter-trawl fleet may be underestimated in this year. Sampling levels during the earlier years in the timeseries are considered to be low for all fleets and the quality of the catch data, particularly for older ages, up until around 1992 is believed to be poor. In 1995, UK age compositions for the period were revised using new ALKs which used data from adjacent time periods where necessary. In the 2005 benchmark assessment, it was noted that numbers-at-age 1 in the landings data were very sparse and variable, reflecting the selection on this age (and especially considering the probable substantial discarding), so the values were replaced by zero to avoid fitting to noise. Keeping age 1 in the assessment allows the survey data at age 1 to contribute. Discards Discard information was not routinely incorporated into the assessment prior to Working group estimates of the combined, raised, level of discards are available from 2004, they have shown a consistent increase apart from 2007 when a substantial increase occurred in the discarding by all fleets followed by a return to the previously lower levels. Recent discard rates, although variable, are substantial in some fleets/periods. Total raised discard information is available for some fleets, and data raised to sampled vessels for others.

269 ICES WGCSE REPORT B.2. Biological Weights-at-age Landings Historically, landings weights-at-age were constructed by fitting a quadratic smoother through the aggregated catch weights for each year. WKFLAT (2011) decided not to continue this approach, following concerns raised by WGCSE that poor fits of the quadratic smoothing curve were resulting in the youngest ages being estimated to have heavier weights than adjacent older ages. WKFLAT (2011) rejected the use of the polynomial smoother for weights-at-age and suggested that raw landings weights are used in future. Raw data back to 1995 was obtained by WKFLAT (2011) and used to update the catch weights and stock weights files. Discards Discard weight-at-age data was available for Belgium, the UK(E+W) and Ireland. Previously, Irish discard data was available but not used in the assessment. The UK weight-atage data was derived from data collected by Cefas for each year ( ). The Belgian weight-at-age data was derived using estimates of total catch biomass and total numbersat-age for years These values were used to derive a weight-at-age matrix in grams for an individual fish. Prior to 2012, the UK(E&W) estimates of discard weights-atage were used to derive the Northern Irish and the Irish discard component. The three national weight-at-age matrices were combined to a total international matrix by weighting the individual weights-at-age for each year, by the total discard tonnages from the two countries for that year. Where only one estimate of weight was available for an age/year, then that estimate was used. The above processes also produced estimates of discard numbers-at-age for the three countries. The UK estimates were raised to incorporate equivalent levels of discards for the un-sampled countries of France and Northern Ireland (on the basis of similar gear types). A raising factor based on tonnages landed for these countries was calculated and applied to the UK(E+W) estimates of discard numbers. Finally, these estimates were added to those calculated for Belgium to give total international discard numbers-at-age estimates. Stock weights For the years where discard estimates were available, a revised set of stock weights-at-age were calculated. The stock weights-at-age based on landings; with SOP correction but no fitting were combined with the international discard weights-at-age data. These were weighted by the relative landed or discarded international annual tonnages. The international annual discard tonnage was not readily available, as the unsampled countries did not have estimates. These were derived using the ratio of UK(E+W) tonnages of landings and discards and this ratio was applied to these unsampled nations landings to produce an estimate of total discard biomass for each of these countries. For the years prior to 2004, a revised set of stock weights-at-age data based on the international landings only was produced. These new values were based on

270 1640 ICES WGCSE REPORT 2013 the observed weight data, but were SOP corrected. For this series of data, the smoothing of the data by fitting a curve through the observed data was removed. Natural mortality and maturity ogives Initial estimates of natural mortality (0.12 for all years and ages from tagging studies) and maturity were based on values estimated for Irish Sea plaice. A new maturity ogive based on UK(E&W) VIIfg survey data for March 1993 and March 1994 (Pawson and Harley, 1997) was produced in 1997 and is applied to all years in the assessment. Age Historic maturity Revised maturity The proportion of mortality before spawning was originally set at 0.2 since approximately 20% of the total catch was taken prior to late February early March, considered to be the time of peak spawning activity. The proportion of F and M before spawning was changed to zero at the request of ACFM in 1996 given that it was considered that these settings were more robust to seasonal changes in fishing patterns, especially with respect to the medium-term projections. No updated information was provided to WKFLAT and the estimates were retained. B.3. Surveys Indices of abundance are available from the UK(E&W)-BTS-Q3 beam-trawl survey in VIIf and the Irish Celtic Explorer IBTS survey (IGFS-WIBTS-Q4). The UK(E&W) beam-trawl survey series that began in 1988; since 1991, tow duration has been 30 minutes but prior to this it was 15 minutes. In 1997, values for 1988 to 1990 were raised to 30 minute tows. However, data for 1988 and 1989 were of poor quality and gave spurious results: thus, the series was truncated to A similar March beam-trawl survey began in 1993 and was made available to the working group in The March beam-trawl survey ended in 1999 but continued to be used as a tuning index in the assessment until Recent data have shown less correlation between ages than the historic time-series which should be monitored in case it is a developing problem. The log catch curves show good consistency over time and the reduction through time of the negative slope indicates that mortality rates have been declining. The IGFS is a demersal trawl survey which started in It is coordinated through the ICES International Bottom Trawl (IBTS) Working Group, providing annual indices of abundance for commercially exploited groundfish stocks on the Irish continental shelf (ICES VIa, VIIb,g&j) for Q3-4. Plaice are caught by the survey off the SE coast up to, and just over, the border of VIIg with VIIa (ICES rectangles (32E2, 32E3). Year effects in the survey catch rates dominate the abundance indices. The year-class and catch-curve plots illustrates that the consistency of plaice year-class abundance estimates at each age is relatively poor. The survey was not fitted within the assessment model, but will be monitored as the time-series progresses.

271 ICES WGCSE REPORT B.4. Commercial lpue Commercial tuning indices of abundance from the UK(E&W) beam-trawl and otter-trawl data are used in the assessment to provide information on the oldest ages in the population. Historically, only ages 4 8 have been used to calibrate the assessment because of concerns about the level of discarding at the youngest ages. The data show good historical consistency of year-class estimates throughout the time-series, especially for the beam trawls, with more noise resulting from two major year effects in the otter-trawl data. C. Stock assessment Historically the stock was assessed using XSA, under the assumption that discarding had a minimal effect on the estimates. Recent increases in the level of discarding led to this assumption being untenable and so at WKFLAT (2011) discard estimates were introduced to the assessment fitted using the AP model. The settings and data for the model fits are set out in the table below: ASSESSMENT YEAR 2012 Assessment model AP Catch data Including discards Tuning fleets UK(E&W)-BTSurvey ages 1 5 UK commercial beam trawl ages 4 8 UK commercial otter trawl ages 4 8 Ire GFS Q3/4 Series omitted Selectivity model Linear Time Varying Spline atage (TV) Discard fraction Polynomial Time Varying Spline at-age (PTVS) Landings number-at-age, range 1 9+ Discards number-at-age, year range, age range , ages 1 7 The three AP model variants could not be distinguished in terms of the AIC, similar residual patterns and fits to the dataseries; the TI_PTVS, TI_TVS and TV_PTVS models. WKFLAT (2011) concluded that the TV_PTVS model, which allows for variation in time in the selection patterns of both landings and discards, was the most plausible model; given the known changes in gear types and discarding. However, it was not statistically distinguishable from the models which maintain the landings selection pattern as constant throughout the time-series. Comparison of the management and stock metrics from the three model fits showed very similar time-series trends in the estimates of fishing mortality, SSB and total estimated discards. WKFLAT therefore concluded that: 1 ) Due to the change in estimated fishing mortality when discards are included within the model fit, that discards should be retained within the assessment model structure.

272 1642 ICES WGCSE REPORT ) Given that the time-series of discard data to which the models are fitted is short and that, consequently, there are likely to be changes in the management estimates as discard data are added in subsequent years, no definitive model structure can be recommended at this stage in the development process. 3 ) The most flexible of the models TVS_PTVS should be used as the basis for advice; in terms of relative changes in estimated total fishing mortality and biomass. 4 ) The other two models which provide similar structures should continue to be fitted at the working group to provide sensitivity comparisons. 5 ) As the dataseries are extended a final model selection can be then determined. In 2013, no assessment was presented for this stock given that the preferred Aarts and Poos (2009) model failed to converge and other model variants could not provide realistic representations of observed landings and discards. Consequently, WGCSE 2013 decided to avoid the use of the preferred TV_PTVS AP model variant and instead focus on assessing the stock using trends derived from the fishery-independent UK(E&W) beamtrawl survey. Trends derived from the UK(E&W) beam-trawl survey were selected for the basis of advice given that this survey most appropriately covered the spatial extent of the stock and well represented the mean age (2 5) landed in the fishery. The UK(E&W) beam-trawl survey was used to infer trends in recruitment, stock size (spawning stock biomass) and fishing mortality. D. Short-term projection For short-term forecasts based on the revised assessment it is recommended that the current methods be applied to the populations and fishing mortalities (separated into discard and landings mortalities) derived from the PV_TVS model (assuming that the previously discussed sensitivity analyses do not indicate a change of model), in order to provide indications of the expected trends in discards, landings and spawning biomass. No short-term projections were presented for this stock at WGCSE E. Medium-term projections Medium term projections are not carried out for this stock. F. Yield and biomass per recruit/long-term projections Yield-per-recruit calculations are conducted using the same input values as those used for the short term-forecasts. Currently the YPR calculations are used as a basis for determining the catch option for advice. G. Biological reference points The addition of discards increases the estimates of spawning biomass in the most recent years following the increased estimates of discards in time. Similarly fishing mortality averaged across ages 3 6, which include ages that are discarded also increases. Previous BRPs may therefore not be consistent with new assessment methodology and should not be used until the assessment methodology is considered sufficiently stable (a longer timeseries of discard data) to evaluate new reference levels.

273 ICES WGCSE REPORT H. References Dunn, M.R., and Pawson, M.G The stock structure and migrations of plaice populations on the west coast of England and Wales. Journal of Fish Biology 61: Fox, C.J., Planque, B.P., and Darby, C.D Synchrony in the recruitment time-series of plaice (Pleuronectes platessa L) around the United Kingdom and the influence of sea temperature. Journal of Sea Research 44: Pawson, M.G Biogeographical identification of English Channel fish and shellfish stocks. Fisheries Research Technical Report No. 99. MAFF Directorate of Fisheries Research, Lowestoft. Pawson, M.G. and Harley, B.F.M Revision of Maturity Ogives for plaice in the Irish Sea (ICES Division VIIa) and Celtic Sea (ICES Division VIIf+g). Working document for ICES Northern Shelf Demersal Working Group. Sideek, M.S.N The estimation of natural mortality in Irish Sea plaice (Pleuronectes platessa L.) using tagging methods, 206pp.

274 1644 ICES WGCSE REPORT 2013 Stock Annex 7.11: Stock Working Group Plaice in Divisions VIIh k (Southwest of Ireland) Plaice in Divisions VIIh k (Southwest of Ireland) WGCSE Date May 2013 Revised by WGCSE / Hans Gerritsen A. General A.1. Stock definition Plaice landings occur near the Irish coast (VIIj) and in the north of VIIh. There is no evidence to suggest that fish from these two areas belong to the same stock. The assessment is performed for VIIjk only. Plaice landings by rectangle, all gears, all nations (STECF data) on a linear scale (left) and a log scale (right). The assessment area is outlined in blue. Landings in VIIh k are marginal and mainly occur in inshore areas. A.2. Fishery Plaice in VIIj are mainly caught by Irish vessels on sandy grounds off counties Kerry and west Cork. Plaice catches in VIIk are negligible.

275 ICES WGCSE REPORT A.3. Ecosystem aspects B. Data B.1. Commercial catch Landings data are available for VIIh,j and k. Some discard data are available but these are not included in the model because the precision of these data is low. B.2. Biological CNAA and CWAA are sampled from the Irish landings which are nearly exclusively from VIIj. No sampling data are available for VIIh. B.3. Surveys The IGFS-WIBTS-Q4 covers VIIj but the gear is not well suited for quantifying plaice abundance because catchability of flatfish is very variable. B.4. Commercial cpue The commercial lpue time-series may not be reflective of overall stock abundance due to changing fishing practices. B.5. Other relevant data C. Assessment: data and method A trends-only assessment is performed on the VIIjk part of the stock only (because no biological data are available for VIIh and because it is not clear that VIIh is part of the same stock). Model used: separable vpa Software used: FLR with R version and packages FLCore 2.5.0; FLEDA 2.5 and FLAssess Model Options chosen: Age classes 4 to 8+ were included in the model. Younger ages were omitted because significant discarding is expected to take place at these ages. A terminal S of 1.0 was used because the catch curves and catch ratio plots suggest a flat selection pattern after age 4. A terminal F was of 0.5 chosen because effort has been fairly constant in recent years so one would not expect a strong trend in F in those years. The separable model was applied to the last six years only because the fishery appeared to have remained stable in this period Note that the parameters for terminal F and terminal S are set as follows in FLR: sep.sel = 1/terminal_S and ref.harvest = terminal_f * terminal_s

276 1646 ICES WGCSE REPORT 2013 Input data types and characteristics: TYPE NAME YEAR RANGE AGE RANGE VARIABLE FROM YEAR Caton Landings in tonnes 1993 present Yes Canum Weca West Landings-at-age in numbers Weight-at-age in the commercial catch Weight-at-age of the spawning stock at spawning time. Weca is used 1993 present 4 8+ Yes 1993 present 4 8+ Yes 1993 present 4 8+ Yes TO YEAR Mprop Proportion of natural mortality before spawning all 4 8+ No Maturity ogive: (based on WG maturity ogive for plaice in VIIfg): AGE PropMat D. Short-term projection Recruitment appears to have been quite stable since Before that, recruitment might have been higher or the discard rate could have been lower. There are some indications that discard patterns may have been different in that period because relatively small plaice were more marketable. For this reason the recruitment assumption for the shortterm forecast was the geometric mean from 2003 and omitting the last two years. Three year averages were used for F and weights-at-age. E. Medium-term projections None. F. Long-term projections None. G. Biological reference points TYPE VALUE TECHNICAL BASIS MSY MSY Btrigger Not defined Approach FMSY 0.43 Provisional proxy based on WGCSE 2013 estimate of FMAX Blim Not defined Precautionary BPA Not defined

277 ICES WGCSE REPORT Approach Flim Not defined FPA Not defined H. Other issues None. I. References None.

278 1648 ICES WGCSE REPORT 2013 Stock Annex 7.12: Stock Working Group Sole in Division VIIb,c (West of Ireland) Sole in Division VIIb, c (West of Ireland) WGCSE Date April 2013 Revised by WGCSE / Hans Gerritsen A. General A.1. Stock definition In VIIb there are two distinct areas where sole are caught: an area to the west of the Aran Islands and an area in the north of VIIb which extends into VIa (the Stags and Broadhaven Ground). It is not known how much exchange there is between sole on the Aran Grounds and those on the Stags Ground. A.2. Fishery Sole in VIIb are mainly caught by Irish vessels on sandy grounds in coastal areas. Sole catches in VIIc are negligible. The landings and lpue of sole in VIIbc appear to have been more or less stable since the start of the logbooks time-series in 1995 (WD1, WGCSE 2009; Figure ). A.3. Ecosystem aspects B. Data B.1. Commercial catch Landings data are available. B.2. Biological Due to the low volume of landings, no sampling of this stock takes place. B.3. Surveys The IGFS-WIBTS-Q4 covers VIIb but the gear is not particularly suited for flatfish and catchability is very variable. B.4. Commercial cpue The commercial lpue time-series may not be reflective of overall stock abundance due to changing fishing practices. B.5. Other relevant data C. Assessment: data and method Model used: Depletion-Corrected Average Catch (DCAC; MacCall, 2009)

279 ICES WGCSE REPORT Software used: NOAA Fisheries Toolbox: DCAC.exe Model Options chosen: Because the value of the depletion delta parameter is unknown, a range of values was used (10%, 50% and 90%; delta is the difference in biomass in the first year and biomass in the last year as a proportion of the virgin biomass (unfished vulnerable abundance). All other settings are based on default values and recommendations from MacCall (2009). LAND CV M STDEV FMSY/M STDEV BMSY/B0 STDEV DELTA STDEV lognormal bounded 0 1 beta 0.1, 0.5, bounded 0 1 beta Input data types and characteristics: Because average catch is analysed, the year range chosen can have a large influence on the results. Two year ranges were tested: 1950 present (the time period after WWII when the stock was heavily exploited) and 1995 present (the time period when the landings showed a declining trend). D. Short-term projection None. E. Medium-term projections None. F. Long-term projections None. G. Biological reference points Not set. H. Other issues None. I. References MacCall, AD Depletion-corrected average catch: a simple formula for estimating sustainable yields in data-poor situations. ICES J Mar Sci 66:10 p

280 1650 ICES WGCSE REPORT 2013 Stock Annex 7.13: Stock Working Group Celtic Sea Sole Sole (division VIIf,g) Assessment of Southern Shelf Demersal Stocks Date 29 July 2004 Last updated 17 May 2013 by Willy Vanhee A. General A.1. Stock definition A description of the stock definition of sole in the Celtic Sea was given in the leaflet Fisheries information - cod, sole, plaice and whiting in the south west of the British Isles published by Cefas under a EU funded project (SAMFISH: EU Study Contract , Improving sampling of western and southern European Atlantic Fisheries) and is taken over here. In the coastal waters of western England and Wales, sole are found in greatest abundance in the northeastern Irish Sea and the eastern Celtic Sea. The main spawning areas for sole in the Celtic Sea are in deep waters (40 75 m) off Trevose Head, where spawning usually takes place between March and May. Sole nursery grounds are generally located in shallow waters such as estuaries, tidal inlets and sandy bays. Juvenile sole (0 and 1 year old fish) are found chiefly in depths up to 40 m, and adult sole (fish aged 3 plus) are generally found in deeper water. Spawning and nursery grounds are well defined. Over 6000 sole were tagged on the nursery grounds of the Bristol Channel and the Irish Sea between 1977 and The majority of fish tagged in Swansea Bay and Carmarthen Bay were between 15 and 24 cm in length. Most of the recaptures of these tagged fish occurred two or more years after release, which meant that many fish tagged as juveniles were recaptured as adults. The majority of returned fish were reported off the north coasts of Devon and Cornwall, and over a wide area in the eastern Celtic Sea and St George's Channel. These results suggest that once an adult sole has recruited to an area, it tends to remain there, and that there is only limited movement of sole between the Celtic Sea and adjoining areas. Figure A.1. Nursery and spawning areas of sole in the Celtic Sea (After Coull, K.A., Johnstone, R., and S.I. Rogers Fisheries Sensitivity Maps in British Waters. Published and distributed by UKOOA Ltd.).

281 ICES WGCSE REPORT A.2. Fishery Fisheries for sole in VIIf,g involve vessels from Belgium, taking approximately two thirds, the UK taking approximately one quarter, and France and Ireland taking minimal amounts of the total landings. Nominal landings are available from 1986 onwards. Sole are mainly targeted by beam trawlers and the fishery is concentrated on the north Cornish coast off Trevose Head and around Land s End. There is an average landing of 1000 tonnes throughout its history (See also Figures A.2 and A.3). Discard information is being collated since 2004 and it seems to be minor. Discarding of sole in the UK(E&W) fleet was estimated to fluctuate between 1% and 9% in numbers. Discard rates of sole in the Belgian beam trawl fleet (responsible for the main uptake of this stock) account for about 2 5% in weight. Figure A.2. Effort distribution of the Belgian beam trawl fleet operating in the Celtic Sea. (VMS data 2002). # # # # # # # # # # # # # # # # # # # Figure A.3. Effort distribution of the English beam trawl fleet operating in the Celtic Sea. Data based on total demersal landings in A.3. Management Celtic Sea sole has been managed by TAC since??? Other management measures are technical measures including minimum landing size (24 cm since???) and minimum mesh sizes (80 mm for beam trawlers since???).

282 1652 ICES WGCSE REPORT 2013 Besides national authorities can impose additional management measures, such as temporal closures, trip catch controls and monthly catch controls. Council Regulation (EC) No 27/2005, Annex III, part A 12 (b) prohibited fishing in ICES rectangles 30E4, 31E4 and 32E3 during January-March This prohibition did not apply to beam trawlers during March. Council Regulation (EC) No 51/2006, Annex III, part A 4.2 prohibited fishing in ICES rectangles 30E4, 31E4 and 32E3 during February and March 2006 with derogations for vessels using pots, creels or nets with less than 55 mm mesh size. The prohibition does not apply within 6 nautical miles from the baseline. Council Regulation (EC) No 41/2007, Annex III, part A 7.2 prohibited fishing in ICES rectangles 30E4, 31E4 and 32E3 during February and March 2007 with derogations for vessels using pots, creels or nets with less than 55 mm mesh size. The prohibition does not apply within 6 nautical miles from the baseline. Council Regulation (EC) No 40/2008, Annex III, part A 6.2 prohibited fishing in ICES rectangles 30E4, 31E4 and 32E3 during February and March The prohibition does not apply within 6 nautical miles from the baseline. Council Regulation (EC) No 43/2009, Annex III, part A 6.2 prohibited fishing in ICES rectangles 30E4, 31E4 and 32E3 during February and March The prohibition does not apply within 6 nautical miles from the baseline. Council Regulation (EC) No 1288/2009, Article 1 stipulates that the prohibited fishing in ICES rectangles 30E4, 31E4 and 32E3 during February and March referred to in Council Regulation (EC) No 43/2009, Annex III, part A 6.2 shall be applicable until 30 June Council Regulation (EC) No 579/2011, Article 2 stipulates that the prohibited fishing in ICES rectangles 30E4, 31E4 and 32E3 during February and March stipulated in Council Regulation (EC) No 43/2009, Annex III, part A 6.2, and prolonged in Council Regulation (EC) No 1288/2009, Article 1, shall be applicable until 31 December Council Regulation (EC) No 227/2013, Article 29c of the European Parliament and of the Council of 13 March 2013 amending Council Regulation (EC) No 850/98 for the conservation of fishery resources through technical measures for the protection of juveniles of marine organisms and Council Regulation (EC) No 1434/98 specifying conditions under which herring may be landed for industrial purposes other than direct human consumption. A.4. Ecosystem aspects Physics Bathymetry: Shelf sea south of Ireland, limited to the west by the slope of the Porcupine Seabight and the Goban Spur. Circulation: Along the shelf edge, there is a poleward flowing slope current ; on the shelf a weaker current flows north from Brittany across the mouth of the English Channel. Thermal stratification and tidal mixing generates the Irish coastal current which runs westwards in the Celtic Sea and northwards along the west coast of Ireland. Several riv-

283 ICES WGCSE REPORT ers discharge freshwater into the ecoregion and influence the circulation patterns. These are notably the River Loire, the Severn and the Irish rivers Lee and Blackwater. Fronts: The Irish Shelf Front is located to the south and west of Ireland (at ca. 11 W), and consists of a tidal mixing front existing all year-round. On the shelf, there are the Ushant Front in the English Channel and the Celtic Sea front at the southern entrance to the Irish Sea. Temperature: Sea surface temperatures measured in coastal stations northwest of Ireland since the 1960s show a trend of sustained positive temperature anomalies from An offshore weather buoy maintained off the southwest coast of Ireland (51.22 N W) since mid-2002, indicated that 2003 and 2005 had the warmest summer temperatures of the record while 2007 saw the warmest winter temperatures. Temperatures in 2008 started above the time-series mean ( ) until April and from July onwards, temperatures remained well below the time-series mean (WGOH 2009). Biology Phytoplankton: Productivity is reasonably high on the shelf with a rapid decrease west of the shelf break. Continuous Plankton Recorder (CPR) data suggests a steady increase in phytoplankton over at least the last 20 years. Toxic algal blooms occur around Irish coasts esp. along the southwest of Ireland. Zooplankton: CPR data suggest an overall decline in the abundance of zooplankton in recent years. Calanus abundance is now below the long-term mean. Benthos, larger invertebrate, biogenic habitats: The major commercial invertebrate species is Norway lobster (Nephrops norvegicus). Two epibenthic assemblages predominate in the Celtic Sea: one along the shelf edge and the slope, dominated by the anemone Actinauge richardi and a more widely distributed assemblage on the continental shelf, dominated by Pagurus prideaux and other mobile invertebrates (shrimps and echinoderms). Fish Community: The area is a spawning area for key migratory fish species, notably mackerel Scomber scombrus and horse mackerel Trachurus trachurus. On the continental shelf the main pelagic species are herring Clupea harengus, sardine Sardina pilchardus and sprat Sprattus sprattus. The groundfish community consists of over a hundred species with the most abundant 25 making up 99% of the total biomass. Surveys revealed a downward trend in the biomass and abundance of cod, whiting and hake. Birds, Mammals and Elasmobranchs: Basking shark (Cetorhinus maximus) is seen throughout the area but the stock seems to be severely depleted. Blue sharks (Prionace glauca) are found during the summer. The Harbour porpoise hocoena phocoena is the most numerous cetacean in the region. Bottlenosed dolphins (Tursiops truncates) occur in large numbers while the common dolphin (Delphinus delphis) is also widely distributed in the area. White-beaked dolphin and white-sided dolphin (Lagenorhynchus albirostris and L. acutus) occur over much of the shelf area. Grey seals (Halichoerus grypus) are common in many parts of the area. Petrels (fulmar and storm-petrel) dominate the seabird populations in the west of Ireland and Celtic Sea region but there are also large breeding colonies of kittiwake, guillemot and gannet. Environmental signals and implications: Increasing temperature and changes in zooplankton communities are likely to have an impact on the life histories of many species.

284 1654 ICES WGCSE REPORT 2013 Cod in the Celtic Sea are at the southern limit of the range of the species in the Northeast Atlantic. It is known that at the southern limits of their range, recruitment tends to decrease in warmer waters (above 8.5 C), and that cod are not found in waters warmer than 12 C. Celtic Sea cod has higher growth rates and mature earlier than other cod stocks. Although it is uncertain, Drinkwater (2005) has predicted that a sustained 1 C rise in sea bottom temperature, over the course of this century, could result in the disappearance of cod stocks from the Celtic Sea and the English Channel. Already there has been a northward shift in the distribution of some fish with an increase of sea bass Dicentrarchus labrax and red mullet Mullus surmuletus populations around British coasts. The region also recently experienced an unprecedented increase in the numbers of snake pipefish, Entelurus aequoreus. Abundance of herring Clupea harengus and pilchard Sardina pilchardus occurring off the southwest of England, has been shown to correspond closely with fluctuations in water temperature. Sardines were generally more abundant and their distribution extended further to the east when the climate was warmer, whilst herring were generally more abundant in cooler times. The migration timing of squid (Loligo forbesi) and flounder (Platichthys flesus) off the southwest of England has also been linked to temperature (Sims et al., 2001; 2004). Zooplankton abundance has declined in the region in recent years and the overall substantial decline in Calanus abundance, which is currently below the long-term mean, may have longer term consequences given the fish community shift towards smaller pelagic species feeding on zooplankton. Fishery effects on benthos and fish communities: Temporal analyses of the effects of fishing and climate variation suggest that fishing has had a stronger effect on sizestructure than changes in temperature. A marked decline in the mean trophic level of the fish community over time has been documented and this has resulted from a reduction in the abundance of large piscivorous fishes such as cod and hake, and an increase in Nephrops and smaller pelagic species such as boarfish (Capros aper) which feed at a lower trophic level. In the Celtic Seas, discarding levels differ between the different fleets but can be as high as two thirds of the total catch with increasing trends in recent years. Discarding of undersized fish is a problem in several fisheries (e.g. cod, haddock, Nephrops and megrim). Improving the selection pattern should benefit the stocks and result in a higher long-term yield. Sole and plaice are predominantly caught by beam-trawl fisheries. Beam trawling, especially using chain-mat gear, is known to have a significant impact on the benthic communities, although less so on soft substrates and in areas which have been historically exploited by this fishing method. Benthic drop-out panels have been shown to release around 75% of benthic invertebrates from the catches. Information from the UK industry (Trebilcock and Rozarieux, 2009) suggests that uptake in 2008 was minimal. The high mud content and soft nature of Nephrops grounds means that trawling readily marks the seabed, trawl marks remaining visible for some time. Despite the high intensity of fishing (some areas are impacted >seven times/year) burrowing fauna can be seen re-emerging from freshly trawled grounds, implying that there is some resilience to trawling. Cetacean bycatch has been noted in some fisheries, including the pelagic trawl fishery for mackerel and horse mackerel in the SW of Ireland, although the numbers caught were low.

285 ICES WGCSE REPORT B. Data B.1. Commercial catch Quarterly data are available for catch numbers for the Belgian, the Irish and UK fleets. These comprise around 95% of the international landings. Derivation of the age composition is shown in the table below. Quarterly total landings are available from France and also from Northern Ireland. Data source: VIIFG BE IR* UK DERIVATION OF INTERNATIONAL LANDINGS IN VIIFG Length composition VIIfg VIIfg VIIfg ALK VIIfg VIIfg VIIfg Age Composition VIIfg VIIfg VIIfg B, IRE + UK, raised to total international landings* * From 2005 to 2009 no Irish length compositions or ALK s therefore from 2005 to 2009, BE + UK age composition raised to total international landing. Numbers at-age 1 in the catch are low in most years, therefore these were not considered to add useful information and are replaced by zeros. Historical compilation of the commercial catch data not included yet. B.2. Biological Weights-at-age The total international catch weights-at-age are calculated as the weighted mean of the annual weight-at-age data supplied by Belgium, UK(E&W) and Ireland, which account for 95% of the total international landings (weighted by landed numbers), and smoothed using a quadratic fit where catch weights-at-age are mid-year values (age = 1.5, 2.5, etc.). These quadratic fits through these points differ from year to year. Therefore they are provided for each year separately. The method is always the same, just the fit differ from year to year as the catch weights differ from year to year. Catch weights-at-age have been scaled to give a SOP of 100%. This technique has been used for many years (at least since stock has been assessed by the Southern Shelf Demersal WG. The same technique has been used in other stocks in the WGCSE (e.g. plaice VIIe). The text table below shows the quadratic fit of the data, the R² of the fit, the periodicity of the data being collected and the countries that delivered the data to calculate the fit. YEAR QUADRATIC FIT W(T) = R² PERIODICITY DATA CONTRIBUTING COUNTRIES *(AGE+0.5) *(AGE+0.5)² 0.93 Quarterly B, IRE, UK *(AGE+0.5) *(AGE+0.5)² 0.91 Quarterly B, IRE, UK *(AGE+0.5) *(AGE+0.5)² 0.96 Quarterly B, IRE, UK *(AGE+0.5) *(AGE+0.5)² 0.96 Quarterly B, UK *(AGE+0.5) *(AGE+0.5)² 0.95 Quarterly B, UK

286 1656 ICES WGCSE REPORT *(AGE+0.5) *(AGE+0.5)² 0.95 Quarterly B, UK *(AGE+0.5) *(AGE+0.5)² 0.88 Quarterly B, UK *(AGE+0.5) *(AGE+0.5)² 0.90 Quarterly B, UK 2010 Not available for the moment due to a hard disk crash Quarterly B, IRE, UK *(AGE+0.5) *(AGE+0.5)² 0.97 Quarterly B, IRE, UK *(AGE+0.5) *(AGE+0.5)² 0.92 Quarterly B, IRE, UK *(AGE+0.5) *(AGE+0.5)² 0.97 Quarterly B, IRE, UK For the period the stock weights-at-age are the catch weights of the Belgian beam-trawl fleet (BEL-BEAM) in the first quarter, smoothed by fitting a Gompertz function. The text table gives a historical overview of the parameters that have been used in the Gompertz function [ W ( t) = a exp( b (1 exp( c t))) ]. YEAR PARAMETER A PARAMETER B PARAMETER C R² For the period , the stock weights were calculated as the weighted mean of the 1st quarter weights-at-age data supplied by Belgium and UK(E&W) (weighted by landed numbers) and soothed using a quadratic fit through these points. Since 2010 Ireland also provides this information and this data was included in the calculations. The text table below shows the quadratic fit of the data, the R² of the fit, the periodicity of the data being collected and the countries that delivered the data to calculate the fit. Stock weights-at-age have been scaled to give a SOP of 100%. YEAR QUADRATIC FIT W(T) = R² PERIODICITY DATA CONTRIBUTING COUNTRIES *(AGE) *(AGE)² st quarter B, UK *(AGE) *(AGE)² st quarter B, UK *(AGE) *(AGE)² st quarter B, UK *(AGE) *(AGE)² st quarter B, UK *(AGE) *(AGE)² st quarter B, UK 2010 Not available for the moment due to a hard disk crash Quarterly B, IRE, UK *(AGE+0.5) *(AGE+0.5)² 0.92 Quarterly B, IRE, UK *(AGE+0.5) *(AGE+0.5)² 0.98 Quarterly B, IRE, UK *(AGE+0.5) *(AGE+0.5)² 0.92 Quarterly B, IRE, UK Stock and catch weights have no explicit trends. The values for 2001 showed a strange convergence and were replaced by the mean of the 2000 and the 2002 weights. At some ages, the weights in the stock are higher than the weights in the catch. This is because sole caught from spawning concentrations in the 1st quarter are heavier (10 to 15%) than after spawning.

287 ICES WGCSE REPORT Historical compilation of the weight-at-age data not fully included yet. Natural mortality and maturity ogives Natural mortality was assumed to be 0.1 for all ages and years. This is consistent with the natural mortality estimates used for sole by other ICES working groups (WGNSSK: IV, VIId, WGNSDS: VIIa, WGSSDS: VIIfg, VIIIa,b) and consistent with estimates of M reported in Horwood (1993). The maturity ogive applied to all years is, a combined sex maturity ogive taken from Area VIIfg attributed to Pawson and Harley, WD presented to WGSSDS in Age and older The proportion of M and F before spawning was set to zero. B.3. Surveys Abundance indices for Celtic Sea sole are available for one survey, the UK beam-trawl survey (UK(E&W)-BTS-Q3). The survey has been conducted in September for approximately 24 days annually since There are 101 core fishing and hydrographic stations distributed around the Irish Sea, Bristol Channel and Celtic Sea between 50 to 55 degrees N and between the English, Welsh and Irish coasts. The survey is coordinated by the ICES BTS WG. Need to include map with stations,. Abundance indices for all ages used in the assessment (standardised to the mean of the respective ages) are given in the figure below. The figure shows that the survey is able to track the strength of the year classes reasonably well.

288 1658 ICES WGCSE REPORT 2013 B.4. Commercial cpue Commercial cpue data are available from the Belgian, the UK(E&W) and the Irish beamtrawl fleets, as well as the UK(E&W) and Irish otter-trawl fleets. There is also information on the cpue of the hardly significant Scottish seine fleet for the sole fisheries. B.5. Other relevant data No other relevant data included so far. C. Historical stock development During the eighties fishing mortality increased for this stock. In the following decades fishing mortality fluctuated around this higher level. However fishing mortality has decreased since the late 1990s and is estimated to be below FMSY (0.31) from 2005 until Fishing mortality in 2012 is estimated to be 0.45, above FPA (0.37). Recruitment has fluctuated around 5 million recruits with occasional strong year classes. The 1998 year class is estimated to be the strongest in the time-series and the 2007 year class to be the second highest for this stock. The 2009 year class is by far the lowest in the time-series. The incoming recruitment (year class 2011) is estimated to be above average. SSB has declined almost continuously from the highest value of 8000 t in 1971 to the lowest observed in the time-series in The exceptional year class of 1998 has increased SSB to above the long-term average. The good recruitment in 2008 and above average recruitment in 2009 and 2012 is predicted to keep SSB well above BPA/Btrigger.

289 ICES WGCSE REPORT Tuning data XSA tuning data that have been used in recent assessments are those from Belgium beam trawlers (BEL-CBT), 1971 onwards; from the UK beam-trawl fleet (UK-CBT), Division VIIf, 1991 onwards; and from the UK Corystes September beam-trawl survey (UK(E&W)- BTS-Q3 survey), 1988 onwards. The Belgian beam-trawl fleet is temporally discontinued in This is due to a change in the calculation of the effort statistics from the official logbooks and sale slip notes in the most recent years. Before the next benchmark assessment, a new derivation of these data should become available. There do exist other tuning data for this stock (e.g. UK otter-trawl fleet), but these have not been included in the assessment as they were not considered to be representative for this stock. The Irish Groundfish survey, held in the 4th quarter is available since 2003 but is not yet used in the XSA as the time-series is too short. Assessment methods and settings Celtic Sea sole has been assessed with XSA since??? An overview of the changes in parameter settings of the XSA are given.below:

290 1660 ICES WGCSE REPORT 2013 assessment assessment assessment Fleets Years Ages α-β Years Ages α-β Years Ages α-β BEL-CBT commercial 71-asses-year asses-year asses-year UK-CBT commercial 91-asses-year asses-year asses-year UK(E&W)-BTS-Q3 survey 88-asses-year asses-year asses-year First data year Last data year assessment year-1 assessment year-1 assessment year-1 -First age Last age Time series weights None None None -Model Mean q model all ages Power model (ages 1 & 2) Power model (ages 1 & 2) -Q plateau set at age Survivors estimates shrunk towards mean F 5 years / 5 ages 5 years / 5 ages 5 years / 5 ages -s.e. of the means Min s.e. for pop. Estimates Prior weighting None None None Fbar (4-8) 2003 assessment assessment assessment 2006-Current Fleets Years Ages α-β Years Ages α-β Years Ages α-β BEL-CBT commercial 87-asses-year asses-year asses-year UK-CBT commercial 91-asses-year asses-year asses-year UK(E&W)-BTS-Q3 survey 88-asses-year asses-year asses-year First data year Last data year assessment year First age Last age Time series weights None None None -Model Power model (ages 1 & 2) Power model (ages 1 & 2) Mean q model all ages -Q plateau set at age Survivors estimates shrunk towards mean F 5 years / 5 ages 5 years / 5 ages 5 years / 5 ages -s.e. of the means Min s.e. for pop. Estimates Prior weighting None None None Fbar (4-8) D. Short-term projection Population numbers for ages 2 and older are taken from the XSA output (estimates of the year = the assessment year minus 1). The long-term geometric mean (starting year up to assessment year minus 3) is assumed for age 1 in the forecast. Fishing mortality is set at the mean over the last three years, not rescaled. If a trend occurs in fishing mortality (three consecutive higher or lower estimates), the working group may use a scaled F to the last year. In the 2007 assessment, the mean fishing mortality was rescaled to F Weights-at-age in the catch and in the stock are averaged over the last three years. E. Medium-term projections Population numbers for ages 2 and older are taken from the prediction output (estimates of the year = the assessment year). The long-term geometric mean (starting year up to assessment year minus 3) is assumed for age 1. Fishing mortality is set at the mean over the last three years, not rescaled. If a trend occurs in fishing mortality (three consecutive higher or lower estimates), the working group may use a scaled F to the last year. Weights-at-age in the catch and in the stock are averaged over the last three years. Since 2007 no medium-term projections were done.

291 ICES WGCSE REPORT F. Yield and biomass per recruit / long-term projections Population numbers for ages 2 and older are taken from the prediction output (estimates of the year = the assessment year). The long-term geometric mean (starting year up to assessment year minus 3) is assumed for age 1. Fishing mortality is set at the mean over the last three years, not rescaled. If a trend occurs in fishing mortality (three consecutive higher or lower estimates), the working group may use a scaled F to the last year. In the 2007 assessment, the mean fishing mortality was rescaled to F Weights-at-age in the catch and in the stock are averaged over the last three years. G. Biological reference points Biological reference point values are given in the text table below: TYPE VALUE TECHNICAL BASIS MSY MSY Btrigger 2200 t BPA Approach FMSY 0.31 Provisional proxy based on stochastic simulations Precautionary Approach Blim Not defined BPA 2200 t There is no evidence of reduced recruitment at the lowest biomass observed and BPA can therefore be set equal to the lowest observed SSB. Flim 0.52 Flim: Floss. FPA 0.37 This F is considered to have a high probability of avoiding Flim and maintaining SSB above Bpa in ten years, taking into account the uncertainty of assessments. FPA: Flim 0.72 implies a less than 5% probability that (SSBMT<BPA). H. Other issues No other issues so far. I. References Connolly, P.L., Kelly, E., Dransfeld, L., Slattery, N., Paramor, O.A.L., and Frid, C.L.J MEFEPO North Western Waters Atlas. Marine Institute. ISBN To be completed intersessional.

292 1662 ICES WGCSE REPORT 2013 Stock Annex 7.14: Stock Working Group Sole in Divisions VIIh k (Southwest of Ireland) Sole in Divisions VIIh k (Southwest of Ireland) WGCSE Date May 2013 Revised by WGCSE / Hans Gerritsen A. General A.1. Stock definition Sole landings occur near the Irish coast (VIIj) and in the north of VIIh. There is no evidence to suggest that fish from these two areas belong to the same stock. The assessment is performed for VIIjk only. Sole landings by rectangle, all gears, all nations (STECF data) on a linear scale (left) and a log scale (right). The assessment area is outlined in blue. Landings in VIIh k are marginal and mainly occur in inshore areas. A.2. Fishery Sole in VIIj are mainly caught by Irish vessels on sandy grounds off counties Kerry and west Cork. Sole catches in VIIk are negligible.

293 ICES WGCSE REPORT A.3. Ecosystem aspects B. Data B.1. Commercial catch Landings data are available for VIIh,j and k. B.2. Biological CNAA and CWAA are sampled from the Irish landings which are nearly exclusively from VIIj. No sampling data are available for VIIh. B.3. Surveys The IGFS-WIBTS-Q4 covers VIIj but the gear is not well suited for flatfish and catchability is very variable. B.4. Commercial cpue The commercial lpue time-series may not be reflective of overall stock abundance due to changing fishing practices. B.5. Other relevant data C. Assessment: data and method A trends-only assessment is performed on the VIIjk part of the stock only (because no biological data are available for VIIh and because it is not clear that VIIh is part of the same stock). Model used: separable vpa Software used: FLR with R version and packages FLCore 2.5.0; FLEDA 2.5 and FLAssess Model Options chosen: Age classes 2 to 10+ were included in the model. A terminal S of 1.0 was used because the catch curves and catch ratio plots suggest a flat selection pattern after age 5. A terminal F of 0.15 was chosen because effort has been fairly constant in recent years so one would not expect a strong trend in F in those years. The separable model was applied to the last ten years only because the fishery appeared to have remained stable in this period. Note that the parameters for terminal F and terminal S are set as follows in FLR: sep.sel = 1/terminal_S and ref.harvest = terminal_f * terminal_s

294 1664 ICES WGCSE REPORT 2013 Input data types and characteristics: TYPE NAME YEAR RANGE AGE RANGE VARIABLE FROM YEAR Caton Catch in tonnes 1993 present Yes Canum Weca West Catch-at-age in numbers Weight-at-age in the commercial catch Weight-at-age of the spawning stock at spawning time. Weca is used 1993 present Yes 1993 present Yes 1993 present Yes TO YEAR Mprop Proportion of natural mortality before spawning all No Maturity ogive: (based on WG maturityogive for sole in VIIfg) AGE PropMat D. Short-term projection Recruitment appears to have been quite stable throughout the time-series. Therefore the recruitment assumption for the short-term forecast was the geometric mean from 1993 and omitting the last two years. Three year averages were used for F and weights-at-age. E. Medium-term projections None. F. Long-term projections None. G. Biological reference points TYPE VALUE TECHNICAL BASIS MSY MSY Btrigger Not defined Approach FMSY 0.25 WKFRAME (2011) performed a meta-analysis on sole MSY reference points and concluded that for most stocks an F target (ages 3 8) of 0.25 is a good choice. Blim Not Precautionary BPA Not Approach Flim Not FPA Not

295 ICES WGCSE REPORT H. Other issues None. I. References None.

296 1666 ICES WGCSE REPORT 2013 Stock Annex 7.15: Stock Working Group Whiting VIIe k Whiting VIIe k Celtic Sea Ecoregion Date 12 May 2013 Revised by David Stokes A. General A.1. Stock definition The degree of separation of whiting stocks between the Irish Sea, and ICES Divisions VIIb c from the Celtic Sea, is currently unclear. SAMFISH (EU Study Contract , Improving sampling of western and southern European Atlantic Fisheries) described the stock unit as follows: The main spawning areas of whiting in the Western Channel and Celtic Sea are off Start Point, off Trevose Head and southeast of Ireland. The spawning season is from February to May, and the larvae are found in mid-water before moving to live near the seabed by September. For the next two years, juvenile whiting are found in shallow coastal and estuarine areas, being particularly abundant around Start Point. Nearly 4000 adult whiting were tagged and released off Start Point during August 1958 and Most returns were within three months of release and demonstrated little indication of movement. Subsequent recaptures indicated more movement of whiting into the Celtic Sea than between the western and eastern Channel. Whiting released in summer between 1957 and 1961 near Carmarthen Bay moved south and west towards the two spawning grounds off Trevose and southeast of Ireland. There was no evidence of emigration out of the Celtic Sea area. Returns of whiting tagged and released in the County Down spawning area in the Irish Sea demonstrate more movement south into the Celtic Sea than north to the west of Scotland. A.2. Fishery Whiting in Divisions VIIe k are taken as a component of catches in mixed trawl fisheries. Whiting landings through the mid 1980s totalled between t and t, through the mid to late 1990s landings were elevated to around t. Since the turn of the century, landings have been in decline and are now below t. Through the 1980s and early 1990s France accounted for around 60 85% of landings. While Ireland accounted for between 10% and 20% of landings, the UK 10%, and Belgium had minimal contribution (1 2%). Landings from both the UK and Belgium have remained at similar levels over time. Since the early 1990s Ireland has accounted for a greater proportion of landings. Proportions since 2004 have been similar to France whose landings have been falling since the turn of the century. French landings are made mainly by gadoid trawlers, which prior to 1980 were mainly fishing for hake in the Celtic Sea. Irish demersal trawlers from Dunmore East and Castletownbere and other ports in southwest Ireland have traditionally targeted Celtic Sea

297 ICES WGCSE REPORT whiting in a mixed trawl fishery. In response to poor catches in other areas vessels have been attracted into this fishery in recent years from County Donegal. A detailed description of the Irish fishery is given in the annual WD to WGSSDS: A summary of the Irish fishery and sampling of whiting in VIIe k. A.3. Ecosystem aspects No relevant information has been made available to the working group. B. Data B.1. Commercial catch Data on international landings-at-age and mean weight-at-age are available for Irish, French and UK fleets from 1999 to present. Data made available through InterCatch from the fishing year 2012 is already raised to VIIe k. The table below presents the data available and the procedures used to derive quarterly length compositions, age compositions and mean weights-at-age. DATA SOURCE: Division Data UK France Ireland Belgium /Other VII ek Length composition ALK Age Composition Mean weightat-age VIIe k VIIe k VIIe k VIIe k VIIe k VIIe k VIIe k VIIe k VIIe k VIIe k VIIe k VIIe k VIIe k Derivation of international landings: VIIe k VIIe k VIIe k Weighted by numbers caught Landings VIIe k VIIe k VIIe k VIIek VIIe k B.2. Biological Age group 0 is included in the assessment data to allow inclusion of 0-group indices in the XSA, although in most years, no landings are recorded. Very small landings of 0- group whiting were not included in the catch-at-age datafile to avoid spurious F- shrinkage effects at this age. Mean weights-at-age in the catch were derived by combining French, Irish and English data, weighted by the numbers landed at-age. Mean weight-at-age in the stock are taken as mean weights-at-age in the quarter 1 catch. Where age 1 was poorly represented in quarter 1 landings, quarter 2 values were used as estimates of mean weight-at-age 1 in the stock. Stock weights-at-age are smoothed using a three year rolling average across ages to dampen the noise exhibited by the stock weight dataset. This approach is also used in Irish Sea whiting and Celtic Sea haddock. Natural mortality is assumed to be 0.2 over all age groups and years.

298 1668 ICES WGCSE REPORT 2013 Maturity data collected in the Celtic Sea in November 2002 during the French EVHOE survey were presented to the WG (Working Document 1: WGSSDS 2003). Results indicated 13% of age 1 fish are mature, 97% at-age 2, and 100% at-age 3 and older. These results are similar to previous assumptions of knife-edged maturity at-age 2. Exploratory analyses indicated that use of the French maturity ogive made little impact on the assessment. The WG therefore retained the assumptions of knife-edged maturity at-age 2. Since 2006 the knife edge maturity ogive has been replaced with indices calculated based on data from the UK WCGFS (Working Document 3: WGSSDS 2006) but a fixed vector is still used. Maturity sampling by Ireland and the UK on dedicated surveys confirms the use of this ogive but is insufficient to provide annual data. The proportions of F and M before spawning were both set to zero to reflect the SSB calculation date of 1 January. The knife edge maturity ogive was replaced with new indices calculated based on data from the UK WCGFS as detailed in WD 3, WGSSDS, Age Maturity B.3. Surveys The following surveys are available as survey tuning data input for the assessment of whiting VIIe k: UK-WCGFS, The March UK groundfish survey was extended in 1992 to provide better coverage for gadoids in VIIf,g. The whiting tuning data calculated from this survey is for VIIf,g. The survey was carried out on the RV Cirolana until In 2004 it was carried out on the RV Endeavour and discontinued thereafter. The survey fished fixed station positions allocated by area and depth strata. The survey used a modified Portuguese High- Headline trawl (PHHT) with 350 mm rubber bobbins, a bunt tickler chain and a 20 mm codend liner. The mean log standardized index by year demonstrated some evidence of positive catchability in the last three years of the survey ( ) and cohort tracking in the mean standardized index up to then was very noisy in the last three years. These years were not included in the final assessment. UK-BCCSBTS-S, The Autumn UK Bristol Channel beam trawl survey (VIIf) is commercially rigged (1989 style) with 4 m beam trawl fitted with a chain mat, flip-up ropes, and a 40 mm codend liner. The gear is towed at 4 knots (ground speed) for 30 minutes. This survey provides information for age 0 and age 1 whiting. FR-EVHOE, 1997 present This fourth-quarter annual groundfish is carried out on the RV Thalassa. Age data are available from 2001 onwards. The sampling design is a stratified random allocation. The number of hauls per stratum is optimized by a Neyman allocation taking into account the most important commercial species in the area (hake, monkfish and megrim). The fishing

299 ICES WGCSE REPORT gear used is a GOV with an average vertical opening of 4 m and a horizontal opening of 20 m. IR-WCGFS, The fourth-quarter Irish west-coast groundfish survey (WCGFS) was carried out in VIaS and VIIbj on chartered commercial vessels. The sampling design attempted to allocate at least two stations per rectangle. Stations were selected randomly within each rectangle from known clear tow positions. A Rockhopper GOV with 12 inch discs was used. The nets were fitted with a 20 mm codend liner. This survey was discontinued after the 2002 survey, giving way to a new Irish groundfish survey on board the RV Celtic Explorer. IR-ISCSGFS, Ireland commenced a Celtic Sea research vessel survey on board the RV Celtic Voyager in 1997 carried out in VIIa and VIIg. The survey used a GOV Trawl with a mean vertical opening is 6 m and door spread 48 m. Data from this survey (IR-ISCSGFS) were presented for the first time to the 2003 WG. The data made available were from prime stations only in a limited area of Division VIIg. The survey was discontinued after the 2002 survey, giving way to a new Irish groundfish survey on board the RV Celtic Explorer. IR-GFS 7g and j, 2003 present Ireland commenced a new fourth quarter survey in 2003 on board the RV Celtic Explorer which covers VIaS, VIIbgj as part of the internationally coordinated, Quarter 4 IBTS survey program. The IGFS has a random stratified design and uses a GOV (with rockhopper in VIa) with a 20 mm codend liner. This is a substantially different design to the Irish Sea/Celtic Sea groundfish survey (IR-ISCSGFS) it replaces. Data from this survey (IR-GFS) were presented for the first time to the 2004 WG. IR-IGFS Swept Area, 1999 present This survey index constitutes a combination of the IR-ISCSGFS and IR-GFS surveys in the area of overlap between them (VIIg). The two surveys were standardized using a sweptarea estimate of catches, described in WD 5 (WGSSDS 2006). This survey was presented for the first time to the 2006 WG. The mean standardized index by year demonstrated good tracking of the strong 1999 year class to age 7 with the exception of age 4 in Although the source data were checked, this is probably an anomaly of the year effect in This point has been removed from recent assessments to ensure the survey gets higher scaled weight in further runs. This compromise is not ideal but given the short time-series of the survey and apparently good performance otherwise the WG considered that the survey should be a good index for this stock. B.4. Commercial cpue Information on effort, and whiting landings and lpue are available from a number of commercial fleets. This includes two French (gadoid and Nephrops directed) since 1983, four Irish (VIIj, and VIIg otter trawlers, and Scottish seines) since 1995, in addition to effort only from UK England and Wales VIIe k beam trawlers and VIIe k otter trawlers since Across the majority of commercial fleets lpue has fallen over time, as is the case with landings. In the mid 1990s at the start of the Irish Scottish seine dataseries lpue was high,

300 1670 ICES WGCSE REPORT 2013 falling steeply over several years. Lpue continues to remain at these lower levels with some annual fluctuation. In relation to otter trawlers, the French gadoid directed fleet consistently revealed the highest lpue. This too has declined over the period of data available to levels half those of the early 1980s. The Irish VIIg otter-trawl fleet is the only one to demonstrate an overall increasing lpue trend although the increase has been relatively small. B.5. Other relevant data No other relevant data to report. C. Historical stock development Data screening: Exploratory data analysis carried out using FLR. A separable VPA was performed using the Lowestoft VPA95 software to screen for outliers in the catch numbers. Model used: XSA Software used: FLR under R version in conjunction with FLCore 1.4 3, FLAssess 1.4.1, FLXSA and FLEDA Lowestoft VPA95 software also for XSA and separable VPA Model Options: Option Setting Ages catch dep stock size None Q plateau 5 Taper No F shrinkage SE 1.00 F shrinkage year range 5 F shrinkage age range 3 Fleet SE threshold 0.50 Prior weights No Input data types and characteristics: Type Name Year range Age range Variable year to year Caton Catch in tonnes 1982 current 0 7+ Yes Canum Catch-at-age in numbers 1982 current 0 7+ Yes Weca Weight-at-age in the commercial catch 1982 current 0 7+ Yes West Weight-at-age of the stock at spawning time 1982 current 0 7+ Yes: Mprop Proportion of natural mortality before spawning 1982 current 0 7+ No Fprop Proportion of fishing mortality before spawning 1982 current 0 7+ No Matprop Proportion mature-at-age 1982 current 0 7+ No Natmor Natural mortality 1982 current 0 7+ No Tuning data:

301 ICES WGCSE REPORT Type Name Year range Age range Tuning fleet 1 FR-Gadoid Late 1993 current 3 6 Tuning fleet 2 FR-Nephrops 1993 current 3 6 Tuning fleet 3 FR-EVHOE 1997 current 0 4 Tuning fleet 4 UK-WCGFS 1987 current 1 6 Tuning fleet 5 IR-IGFS Swept area 1999 current 0 6 Settings for each assessment since 1999 are detailed in Table 1. Trial runs have, over the years, explored most of the options with regards XSA settings. This stock has not had a benchmark assessment; however exploratory assessments have been carried out within the WGSSDS up until D. Short-term projection Model used: Multi Fleet Deterministic Projection Software used: MFDP1a Initial stock size: initial stock numbers derived from XSA analyses. Numbers-at-age 0 are not considered to be well estimated and are replaced with a geometric mean of the full time-series ( ). Recruitment has been at a low level since 1995 with the exception of the 1999 year class. The two most recent years have displayed good recruitment, with last year s being revised downward. Recruitment is solely estimated from the FR-EVHOE and IR-GFS7gSweptArea surveys, in recent years the French survey estimates have been far higher than those of the Irish survey. Because of these reasons the geometric mean is used. Natural mortality: That used in the assessment Maturity: Maturity ogive used in the assessment F and M before spawning: Those used in the assessment method Weight-at-age in the stock: Unscaled three year arithmetic mean Weight-at-age in the catch: Unscaled three year arithmetic mean Exploitation pattern: Unscaled three year arithmetic mean (though alternative options may be used depending on recent F trajectories and the working group s perception of the fishery). Intermediate year assumptions: Status quo F Stock recruitment model used: Geometric mean of full time-series (1982 to present-1) for age 0 recruitment FBAR: That used in the assessment E. Medium-term projections None.

302 1672 ICES WGCSE REPORT 2013 F. Long-term projections Model used: Multi Fleet Yield-per-recruit Software used: MFYPR2a Yield-per-recruit calculations are conducted using the same input values as those used for the short-term forecasts. G. Biological reference points A summary of reference point proposals to date, their technical basis and currently adopted reference points is given in the text Table below: WG 1998 ACFM 1998 WG 2000 ACFM 2000 F lim No Proposal No Proposal 1.18 (Flim=Floss) No Proposal F pa No Proposal No Proposal 0.72 (F pa =F lim x e x 0.3 ) No Proposal B lim 15,000 t 15,000 t t (Blim=Bloss) 15,000 t (Blim=Bloss) B pa 18,000 t 21,000 t t (Bpa=Bloss x 1.4) 21,000 t (Bpa=Bloss x 1.4) The technical basis of ACFM s 1998 BPA proposal is given below (1999 WG text): BPA = Blim x 1.4 = t. In the past the WG have selected MBAL as t based on evidence of reduced recruitment at SSB s < t. However this MBAL is driven by a period of low recruitments at low SSB in the earlier years of the time-series ( ) when the data are probably not reliable. Examination of the stock recruit plot provides no compelling evidence of reduced recruitment below SSB of t. The technical basis of the WG s 2000 Flim and FPA proposals are given below: On the basis of results obtained from a LOWESS fitted non-parametric stock and recruitment relationship and the derived equilibrium SSB and yield curves with the original data trajectories the 2000 working group considered that FPA and Flim could be defined because Floss appeared reasonably estimated. However, taking into account the uncertainties in the data the 2000 working group decided to use 0.3 as the SE in calculation of FPA from Floss. The technical basis for the proposed reference points are defined below: Flim = Floss (1.18 in this year s assessment) FPA = Flim x e-1.645*0.3 = 0.72 The currently adopted reference points are as follows: F lim F pa Current Reference Points No Proposal No Proposal B lim 15,000 t (B LIM = B LOSS 1983, ACFM 1998 ) B pa 21,000 t (B PA = B LOSS 1983 x 1.4) H. Other issues No other issues.

303 ICES WGCSE REPORT I. References

304 1674 ICES WGCSE REPORT 2013 Table 1. Model settings/input data/tuning data. Catch date range: Assmnt Method: Fbar Age Range: Time taper: Q plateau age: F shrinkage S.E: Fleet S.E: Commercial Tuning Fleets: FR-Gadoid FR-Gadoid Late FR-Nephrops IR-7g&j-OT Survey Tuning series: FR-EVHOE UK-WCGFS UK-BCCSBTS IR WCGFS IR-IGFS Swept area Years Ages XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA No No No No No No No No No No No Num yrs Num ages Yrs Ages Yrs Ages Yrs Ages Yrs Ages Yrs Ages Yrs Ages Yrs Ages Yrs Ages 1-1 Yrs Ages

305 ICES WGCSE REPORT Stock Annex 8.2: Western Channel plaice Stock Date Revised by Western Channel Plaice (VIIe) 4th March 2010 (last revised at WKFLAT 2010); updated times-series Ian Holmes May 2011 Ian Holmes, S Kupschus and C Lynam (Cefas-Lowestoft). A. General A.1. Stock definition The management area for this stock is strictly that for ICES Area VIIe called the western Channel, although the TAC area includes the larger component of VIId (eastern Channel). Between 1965 and 1976, more than 5500 plaice were tagged and released around Start Point. Previous analysis of the recaptures from plaice tagged whilst spawning in the Channel (eastern and western areas) during January and February showed that 20% spent the summer in the western Channel, 24% in the eastern Channel, and approximately 56% migrated to the North Sea after spawning (Pawson, 1995). Few of the plaice tagged in the western Channel during April and May were recaptured outside the Channel however, suggesting that there is a resident stock that does not migrate to the North Sea after spawning in the Channel. The main spawning areas are south of Start Point and south of Portland Bill. Spawning takes place between December and March with a peak in January and February. Figure A shows the spawning areas for VIIe plaice. The spawning habitat in VIIe is much smaller than that in VIId and tagging studies have estimated that 87% of the recruits to the western Channel (VIIe) come from outside the area (34% from the eastern Channel VIId and 53% from the North Sea; Pawson, 1995). Similarly, 38% of recruits to the eastern Channel are estimated to have come from the North Sea. The historic tagging data on which these studies were based also show that there is substantial mixing of adult plaice between the western and eastern Channel and between the English Channel and the North Sea, but very limited exchange between the Channel and the Celtic and Irish Seas (Burt et al., 2006). The stocks of plaice in the Channel and North Sea are known to mix greatly during the spawning season (January February). At this time many western Channel and North Sea plaice may be found in the eastern Channel (Pawson, 1995). The comparable lack of spawning habitat in the western Channel alone suggests that this migration from VIIe to VIId during the first quarter may be of considerable importance. North Sea (IV) plaice have been shown to spawn in VIId during January February and subsequently return to the North Sea (Hunter et al., 2004). This migration is tracked by the international fleets fishing in the area: landings peak in January over the spawning grounds, when migrant fish are present, and track the movement towards the North Sea in February and March. A similar migration of plaice from the smaller VIIe stock into VIId during quarter 1 is believed to take place. Once fish have moved into VIId to spawn they are then subject to fishing, largely by the Belgian and French trawlers that take the majority of their annual catch in January and February. Conventional tags inform the recapture position and date of a tagged fish (with known release point) and such data has been investigated to estimate the likely

306 1676 ICES WGCSE REPORT 2013 movement rates of fish from VIId in quarter 1 into VIIe and IV. The movement rates can then be used to determine the proportion of the catch in VIId during quarter 1 that is due to immigrant spawning fish. The resulting estimates of the catch of fish from VIIe and IV that are caught in VIId can then be reallocated to the appropriate catch-at-age matrix. WKFLAT re-analysed data from historical tagging experiments on plaice, which were archived in the Cefas Tagfish database (Burt et al., 2006). The tags were captured through the fisheries and most are returned to Cefas within a few months of release; however these fish have had little chance to migrate. Therefore data from tagged fish with <six months at liberty were excluded from further analysis. In order to focus on movement rates of fish that are available to the fishery only fish greater than the minimum landing size were considered for further analysis. Since tags are returned via the fishery the probability that a tag will be caught depends on the catch of plaice in an area: the greater the catch taken the more likely the tag to be caught. However, the more fish that are present within an area the less likely a tag is to be caught. Therefore the probability that a tag is caught in an area (Number recaptured / Number released) in a particular period must be weighted by the ratio of biomass/catch in that area and year so that probabilities can be comparable between areas and years. The resulting weighted proportions of tags returned from each area provide estimates of the movement probabilities between areas (Table below). Release Information period WEIGHTED BY INTN CATCH AND SSB pr(recap) after 6 or more months at liberty DIV Sex Release Recapture N 7A 7E 7D 4 VIIe B ALL M Jan-Mar F M Apr_Dec F M Jan-Mar Apr_Dec F M Apr_Dec Jan-Mar F VIId B ALL M Jan-Mar F M Apr_Dec F M Jan-Mar Apr_Dec F M Apr_Dec Jan-Mar F IVc B ALL M Jan-Mar F M Apr_Dec F M Jan-Mar Apr_Dec F M Apr_Dec Jan-Mar F Summary of estimated movement probabilities for plaice ( 270 mm) recaptured after six or more months at liberty, for data collected between 1960 and The best estimates of the proportion of fish in quarter 1 in VIId that would return, if not caught by the fishery, to VIIe and IV are circled in red in the table above. So 14% of males and 9% of females would migrate to VIIe, while 52% of males and 58% of females would migrate to IV. To the nearest 5%, this suggests that 10 to 15% of the

307 ICES WGCSE REPORT catch in Q1 in VIId should be allocated to VIIe, while between 50 and 60% of the catch in Q1 in VIId should be allocated to IV. These estimates are in agreement with previous analyses (based on the same data) reported by Pawson (1995), which suggest that 20% of the plaice spawning in VIIe and VIId spend the summer in VIIe, while 56% migrate to the North Sea. Given the assumptions involved in these calculations and the relatively small numbers of adult tags returned the estimates of movement rates are subject to great variability. The limitations of the data do not permit an estimate of annual movement probabilities. Recent studies based on data storage tags suggest that the retention rate of spawning plaice tagged in the eastern Channel is 28%, while 62% of spawning fish tagged were recaptured in the North Sea (Kell et al., 2004). WKFLAT 2010 adopted a 15% movement of catches from VIId into VIIe in Q1 and similarly an additional 50% movement in Q1 from VIId to IV. A.2. Fishery In the western Channel, plaice are taken largely as a bycatch in beam trawls directed at sole and anglerfish. The main plaice fishery is concentrated to the south and west of Start Point. Although plaice are taken throughout the year, landings are usually heaviest during February/March and October/November. The fisheries taking plaice in the western Channel mainly involve vessels from the bordering countries: UK, France and Belgium. Main métiers There are ten main métiers that exploit important fish and shellfish stocks in the Channel. Otter trawling accounts for a wide range of target species in season - cuttlefish, anglerfish, gurnard, rays, cod, whiting, plaice, sole, squid and lemon sole - and involves boats from France (600), England (470), Belgium (15) and the Channel Islands (eleven). Beam trawling is also important for boats from the three former nations (26, 83 and 65 respectively), targeting sole, anglerfish and plaice, with up to 25 of the Belgian boats extending this fishery into the Bay of Biscay. Many boats from France (626) and England (80) join two Channel Islands vessels dredging for scallops and taking a valuable bycatch of sole and anglerfish. The other main towed gear is midwater trawls, used either for the small pelagic species - mackerel, sprat, pilchard and herring - or for bass and black bream with a bycatch of gadoids by French (40) and English (25) boats. Purse-seines are used by eight UK vessels to take mainly mackerel and pilchard in the western Channel. The fixed netting métier in the Channel is really composed of several métiers using specific net gears and mesh sizes depending on target species, the most important being with gillnets and trammelnets (580 French and 380 English boats) for sole, cod, ling, pollack, hake, plaice, bass and spider crab. Rays, anglerfish, turbot, crabs, lobster and crawfish are also taken in tanglenets (305 French, 300 English and seven Channel Islands). Similarly, potting (960 French, 275 English and 560 Channel Islands) uses several distinct gears to catch brown (edible) crabs, spider crabs, cuttlefish, lobsters and whelk, both inshore and offshore, and there are zones in the western Channel partitioning potting and towed gears for alternating periods. Longlining has been replaced by fixed net in many cases, but conger eel, sharks, rays and bass are still taken (260 French, 60 English and 13 Channel Islands). Handlines are used for mackerel, bass, pollack and ling by small boats working along both the English (390) and

308 1678 ICES WGCSE REPORT 2013 French (120 French and 90 Channel Islands) coasts of the Channel. This information is accurate as at WG07. A.3. Ecosystem aspects Other than statistical correlations between recruitment and temperature (Fox et al., 2000) little is known about the effects of the environment on the stock dynamics of VIIe plaice. Environment influences were considered by WKFLAT by incorporating sea surface temperature into the XSA model as a tuning fleet for age 1 catch numbers i.e. as an index of recruitment (ICES Working Document 4.3). Although the large recruitment signal in the late 1980s was partly tracked by the temperature time-series little information was gained, other than a mean recruitment level, for the recent period. There is some anecdotal evidence of changes in the range of some species such as langoustine, triggerfish, and black sea bream from warmer parts of the Atlantic. B. Data B.1. Commercial catch Landings The fisheries that take plaice in the western Channel mainly involve vessels from the bordering countries: UK vessels report about 68%, France 24% and Belgium 8% of the total plaice landings from ICES Division VIIe (based on 2007/08). Although plaice are taken throughout the year, landings are usually heaviest during February/March and October /November. Landings reached a peak of around 2600 tonnes in 1990 after a series of good recruitments in the late 1980s. Landing levels then declined rapidly once recruitment levels returned to average levels. Since 1994, landings have been stable at around 1200 tonnes; however, in 2007 and 2008 landings have been below this level. Most of the landings are made by beam trawlers with around 70% of the UK landings being reported by these vessels and another 25% being landed by otter trawlers. The unallocated landings reported in the WG landings table in recent years are generally additional French landings derived from sales note information. Sampling and data raising Quarterly age compositions were available only from UK(E&W) landings for the years (and 1989), which accounted for approximately 68% of total international landings. The total international age composition was obtained by raising the combined gears quarterly UK(E&W) age compositions to include the landings of the Channel Isles, France and Belgium, and summing to give an annual total. For the earlier years of , French age compositions were also available. For these years, the UK(E&W) age compositions were raised to UK(Total) by including landings from the Channel Islands. Finally, UK(Total) and French age compositions were combined and raised to include Belgian landings. For the years Prior to this, the stock data was aggregated for area of VIId+VIIe. For these years, Belgium also provided age compositions data and this was combined with UK(Total) and French age compositions. French age compositions were based on age data provided by the UK.

309 ICES WGCSE REPORT WKFLAT 2010 recommended a migration model; this model reassigns 15% of the first quarter Belgian, French and UK catch in VIId to the VIIe catch-at-age matrix and similarly raises the landings by including 15% of the first quarter landings in VIId for each country. During the meeting, quarterly data for Belgium and France were available back to 1998 and UK data to In order to extend the time-series back to 1980 the first quarter landings and catch-at-age matrix for each country were inferred from the total annual international landings and catch-at-age data (which begin in 1980 for VIId). Total annual international catch-at-age at-age data ( for France and Belgium and for UK) were down-raised using the average proportion of catch at each age in the first quarter by each country over the period in which quarterly data were available. Similarly, SOP corrected Q1 landings for each country were calculated back to 1980 using the mean (calculated over the period in which quarterly data were available) proportion of the annual landings that were landed in Q1. Age data representing French landings were available for 2002 and 2003, but were not used in the assessment. Table A shows the national data availability for VIIe plaice stock for the time period Table B shows a time-series of CVs of numbers-at-age for sampling UK(E&W) all fleets combined. Weights-at-age Total international catch and stock weights-at-age were calculated as the weighted mean of the annual weight-at-age data supplied (weighted by landed numbers), and smoothed using a quadratic fit: [e.g.: Wt = (0.1109*Age) - (0.0004*(Age 2 )) ; R 2 = 0.98] where catch weights-at-age are mid-year values (age = 1.5, 2.5, etc.), and stock weights-at-age are 1st January values (age = 1.0, 2.0, etc.). Catch weights-at-age have been scaled to give a SOP of 100%, and the same scaling has been applied to stock weights-at-age. This technique has been used for many years (at least since stock has been assessed by the Southern Shelf Demersal WG. In early years in the time-series, weights-at-age were averaged over a period of years, and derived from separate-sex mean weightsat-age. WKFLAT 2010 recommended a migration model that alters the catch-at-age data. However, this model does not alter the weight-at-age matrix since it is not possible to distinguish which weight measurements in VIId are from VIIe migratory spawners. B.2. Biological The main spawning areas for plaice in the western Channel are south of Start Point and Portland Bill. Spawning takes place from December to March, with a peak in January and February. On average, about a quarter of plaice in the western Channel are mature at-age 2, half are mature at-age 3 and all are mature at-age 5. The majority of plaice landed in the western Channel in 2001, for example, were at ages 2 5, and therefore 73% of those landed were mature.

310 1680 ICES WGCSE REPORT 2013 Natural mortality and maturity ogives Initial estimates of natural mortality (0.12 yr all years and all ages) and maturity were based on values estimated for Irish Sea plaice (Siddeek, 1981). A new maturity ogive based on UK(E&W) VIIfg survey data for March 1993 and March 1994 (Pawson and Harley, 1997) was produced in 1997 and is applied to all years in the assessment. Age Old Maturity New Maturity The proportion of mortality before spawning was originally set at 0.2 since approximately 20% of the total catch was taken prior to late February early March, considered to be the time of peak spawning activity. The proportion of F and M before spawning was changed to zero prior to the 1994 Southern Shelf Demersal Working Group as it was considered that these settings were more robust to seasonal changes in fishing patterns, especially with respect to the medium-term projections. B.3. Surveys and survey tuning data An annual 4 m beam trawl survey has taken place in the Lyme Bay area of the western Channel since 1984, initially aboard chartered fishing Vessels (MV BOGEY 1 and latterly MV CARHELMAR) and more recently aboard the Cefas research vessel CORYSTES, coming back to MV CARHELMAR in Appendix 1 provides a history of the survey and details the survey methodology and objectives. The western Channel beam-trawl survey data is used to calculate assessment tuning data for both VIIe plaice and sole. Indices of abundance-at-age for years 1986 to the present, and for ages 1 5 have been used. Since 2007, this age range has been extended to include data for ages 1 8. Appendix 1 also describes how these indices of abundance-at-age are derived. Since 2003 a Fisheries Science Partnership (FSP: Cefas-UK industry cooperative project) has been conducting a survey using commercial vessels with scientific observers and following a standard grid of stations extending from the Scilly Isles to Lyme Bay. The survey covers a substantially larger area than the current survey (UK-WECBTS) and is thought to be more representative of the stock in UK waters. This dataset was first included in the 2007 assessment, and the exploratory analysis can be seen in that report (ICES, 2007; Section 3.2.5). However, recently the vessel(s) used for the survey have changed from the FV Nellie and the FV Lady T, to the FV Carhelmar. In 2008, in addition to the vessel changes there have been other sample protocol changes, notably the change to using 4m survey beam trawls from the commercial 12 m beam trawls previously used by the other vessels. The working group, WGCSE 2009, decided to leave out the 2008 data from the FSP survey since it had an undue influence on estimates of SSB and F. B.4. Commercial lpue The UK(E&W) commercial lpue data is calculated for two gear groups (beam trawl, and otter trawlers both over 40 ft) and for three sectors within VIIe (VIIe north,vii south and VIIe west) made up of collections of ICES rectangles. The lpue values are corrected for fishing power using a given relationship between fishing power and

311 ICES WGCSE REPORT gross tonnage and are calculated using the total effort for a month/sector not speciesdirected effort. This relationship is FP=0.0072*GRT and this is standardised fit to pass through the mean GRT of Irish Sea trawlers in 1979 (Brander, unpublished). Beam-trawl lpue in the North of VIIe reached a peak in 1990, fell sharply to 1994 and now fluctuates at low levels. The south and west sectors both peaked in the early 1990s but have steadily declined since. Otter-trawl lpue in north of VIIe peaked in 1988 before falling sharply until Since then it has remained at these much lower levels. Lpue in the south is generally lower, but fluctuates to high peaks throughout the time-series, whereas in the west it has remained stable at a lower level for the duration of the time-series. UK beam-trawl effort has increased rapidly over the time-series, reaching record high levels in 2003 and has remained at this high level since. UK trawl effort has slowly decreased over the time-series, reaching a record low level in Effort is calculated as fishing power corrected using GRT. Figures B and C show plots of UK effort for by ICES rectangle for ottertrawl and beam-trawl gears, respectively. Commercial tuning data Commercial tuning information for this stock comprises of the UK(E&W) otter-trawl fleet and the UK(E+W) beam-trawl fleet. These fleets have been used by working groups for a number of years, and initially contained data for years back to 1976 (otter) and 1978 (beam). However in the most recent assessments carried out for this stock, otter-trawl fleet data is currently used only for years 1988 to the present and for ages 3 9 and Beam-trawl fleet is currently used for years 1989 to the present, and ages 3 9. Since 2004, an historic otter trawl fleet ( ) has been re-introduced using ages 2 9 only and this is calculated differently from the later data. WKFLAT proposed a migration model for western Channel plaice. If this is not acceptable and the truncated model is taken forward then the commercial beam-trawl and commercial otter-trawl fleets should be truncated so that the first year of the time-series is 1998 and the last year is the most recent year. The truncated model does not use the historic commercial otter-trawl fleet, but has F-shrinkage increased from 2.5 to 1.0 to compensate for the increased variability in estimates of F. B.5. Other relevant data Discarding Discard length summary data from the UK(E&W) and French discard sampling programmes has been made available to ICES working groups for the period In addition, in 2010, Belgian quarterly discard length compositions were also available. All data indicate that discarding is at its highest in quarters 1 and 2 in this fishery, but is still low compared to other plaice stocks. No attempt has previously been made to raise these estimates to total landings. For the 2010 benchmark meeting (WKFLAT), an analysis was carried out to determine the true level of discarding including trends in sampling effort, discarding patterns and an attempt to raise the sampling to an estimate of total discards. This work was presented to the meeting as ICES WKFLAT 2010, Working Document 4.4 western Channel (VIIe) plaice discard data availability, trends and raising estimates to total landings, and comparisons with the trends of adjacent plaice stocks. The summary points made were as follows:

312 1682 ICES WGCSE REPORT 2013 Previous assumptions made by the working group that discarding is small compared to other plaice stocks, and that most discarding takes place in Quarter 1 and 2 appear robust. VIIe discard rates range from 9% in 2003 to 24% in 2008 with an average of 16%. Discarding is at its heaviest in quarters 1 and 2 with 26% and 19% discarded in these quarters and around 5% discarded in the remainder of the year. The discard rate appears to be increasing over time but is still at relatively low levels. Discard rates for VIIe plaice stock (16%) are much less than those for adjacent plaice stocks in VIId (57%) and VIIfg (73%). Sampling effort on discards is very good for the VIIe plaice stock and discard sampling effort is increasing. Most of the sampling effort has been carried out on beam and otter trawlers. Most discard sampling was carried out on vessels of length 10<20 m and with engine power between 100<300 Kw. Around 10% by weight are discarded and this measure is increasing. The proportion discarded by weight has increased steadily from 5% in 2002 to around 13% in This compares favourably with the adjacent stocks that have rates of around 40% in VIId and around 60% in VIIfg (in 2008). There is no evidence of seasonal differences in the proportions discarded at length. The proportions of fish discarded at length for this stock shows good levels of consistency over the time period and in addition the L50 values for each year are very close. This is not the case for the VIId and VIIfg stocks but for these stocks, the inconsistencies may be a feature of lower sample numbers. Around 60 70% of fish discarded are regarded as immature. Raising the discard sample data is possible by using either landings or effort but neither method is perfect. The main problem encountered was the limited availability of age data at the smaller/larger lengths. Most discards are at age 2 and age 3, where an estimated 28% and 5% respectively would be added to the landings age composition. For 2008, the resulting age compositions from both raising methods were almost identical although this may not be the case for other years. The total weight of the discarded catch in 2008 was estimated to be approximately 55 t amounting to around 6% of the commercial landings. On reflection, the workshop considered the possible effects of the lack of discards included in this assessment and recommended that further investigations are conducted to include discard information in future assessments, but not to include the preliminary information available as it may reduce the management of the exploited portion of the stock. The data suggests discarding is minor in the years it has been raised to the fleet level. It was therefore concluded that the effect of including this data in the assessment would at best change the level of F and SSB over the whole time-series and at worst obscure the trends now seen because of the short and variable time-series of discard data available. Potential discard raising methods Two methods were used to raise the discard sample data to total discards. 1 ) Using landings. Sample data for the two main gear groups of beam trawl (gear 1) and otter trawl (gears 2,3,7) and the remaining gears (other) were

313 ICES WGCSE REPORT extracted by quarter. For each gear group and quarter, the weight of the total catch from the sampled trips was calculated by quarter using the formula (W=aL^b * N) where a and b were quarterly condition factors for the stock in use within Cefas stock processing. The discarded Length Distributions (LD s) were then raised to total catches using the ratio of total reported catch/weight of discard trip catches. An Age Length Key (ALK) was applied to each raised quarterly LD to produce quarterly Age Compositions (AC) for each gear group/quarter. The ALK data used was taken from the age samples from the discard programme. Due to the small quantity of discard age data available, the ALK used was at the annual level. However even the ALK at this level only had small numbers of fish and did not cover the full length range of the discard LDs. In these instances, the discard ALK was supplemented by supplements by annual ALK data from the relevant commercial landings samples. At the smallest lengths without age data, an assumption about the age structure was made, but these were generally considered to be age 1. These discarded ACs were then combined across gears and then across quarters to give an annual estimate of discarded catches. 2 ) Using effort data. Given the recognised difficulties is assessing the true effort levels of gears such as gillnetters and longlines, discard sample data only for the two main gear groups of beam trawl (gear 1) and otter trawl (gears 2,3,7) were extracted by quarter. The discarded LDs were raised to total catches using the ratio total reported effort (hours fished) catch/hours fished on sampled trips. The same ALK as constructed above was applied to the quarterly raised LDs to give quarterly age compositions by gear/quarter. At the quarterly level, the two age compositions were combined and then raised to include the catches form the other gears. These ACs were then combined across gears and then across quarters to give an annual estimate of discarded catches. C. Historical stock development This stock was assessed by the ICES Southern Shelf Demersal WG from 1992 to For years 2009 present, this stock was assessed at the ICES Celtic Seas Ecoregion Working Group. The stock has been managed by a TAC since The TAC is applicable to VIId (Eastern Channel) and VIIe combined, although in 1997 there was a separate limit for landings from VIIe. This was unpopular with the industry due to the national split being based on VIId+VIIe combined reported landings for the reference period, and has not been repeated since. Benchmark 2010 This stock was benchmarked at the WKFLAT 2010 meeting where the main issue under review was to overcome the problematic retrospective pattern that meant that forecasts had not been possible for some years. Solutions explored included making an allowance for migration patterns between the two channel plaice stocks, termed the migration model ; this clearly had a knock-on effect on the eastern Channel stock and the North Sea where there was also migration issues. Another option considered (the truncate model ) involves truncating the commercial otter and commercial beam fleets back to 1998 but this was thought to only temporarily hide the underlying problem. Additionally, the truncate model excludes the commercial historic otter-

314 1684 ICES WGCSE REPORT 2013 trawl time-series and increases F-shrinkage from 2.5 to 1.0. WKFLAT 2010 recommends that the FBAR range is altered to 3 6 since very few age 7 fish are caught by the fishery (<4% of the catch numbers). The age range of the FSP survey was reduced to 2 8 since very few age 9 are caught by the survey and that age created positive residuals in catchability for every year. Outcome: The workshop considered making an allowance for migration between the two Channel plaice stocks. Having further examined tagging evidence available it was agreed that an allowance of 15% of quarter 1 catches (both landings and the catch numbers-at-age) from VIId needed to be added into quarter 1 of the VIIe. This was required from all contributing nations. The combination of the two Channel plaice stocks was examined. It was agreed that this would require further investigation as the inclusion of the North Sea stock would also need to be considered. Any combining of stocks would a have a wide ranging impact on the assessment and any subsequent management. The issue of including discard estimates was also considered, but based on the short time-series of data available and the limited impact on the assessment outcome, this inclusion was deferred until a longer time-series of data was available. Technical measures in force Technical measures currently in force in the western Channel are a minimum mesh size of 80 mm for otter and beam trawlers and 70 mm for Nephrops trawlers. Panels of 75 mm square mesh are compulsory in all Nephrops fisheries in ICES Subarea VII. There is also a minimum landing size (MLS) on 27 cm in force. Model used: XSA Software used: Lowestoft VPA suite Model Options chosen: Input data types and characteristics: TYPE NAME YEAR RANGE AGE RANGE VARIABLE Caton Catch in tonnes Yes Canum Catch-at-age in numbers Yes Weca Weight-at-age in the Yes commercial catch West Weight-at-age of the Yes spawning stock at spawning time. Mprop Proportion of natural No mortality before spawning Fprop Proportion of fishing No mortality before spawning Matprop Proportion mature at-age Age1 0%; Age2 26% Age3 52%, Age4 86% Age % No Natmor Natural mortality (0.12) No FROM YEAR TO YEAR YES/NO

315 ICES WGCSE REPORT Tuning data: migration model TYPE NAME YEAR RANGE AGE RANGE Survey fleet 1 Commercial fleet 1 Commercial fleet 2 Commercial fleet 3 UK Western beam-trawl survey (UK-WEC- OT) UK Western Channel Otter Trawl (UK- WECOT) UK Western Channel Beam Trawl (UK- WECBT) UK Western Channel Otter Trawl - Historic (UK-WECOT historic) Survey fleet 2 UK FSP Survey (UK(E&W) FSP) Tuning data: truncated model TYPE NAME YEAR RANGE AGE RANGE Survey fleet 1 Commercial fleet 1 Commercial fleet 2 Commercial fleet 3 UK Western beam-trawl survey (UK-WEC- OT) UK Western Channel Otter Trawl (UK- WECOT) UK Western Channel Beam Trawl (UK- WECBT) UK Western Channel Otter Trawl - Historic (UK-WECOT historic) excluded Survey fleet 2 UK FSP Survey (UK(E+W) FSP) History of assessment methods and settings investigations The standard settings for a catch data screening run using a separable VPA are reference age of 4; F set to 0.7 and S set to 0.8. In 1991 the stock was assessed using a Laurec Shepherd tuned VPA. Concerns about deteriorating data quality prompted the use in 1992 of XSA. Trial runs have, over the years, explored most of the options with regards XSA settings: The effect of the power model on the younger ages was explored in 1994; 1995; 1996; 1998, 2004 and The use of P shrinkage was investigated in 2001; Different levels of F shrinkage were explored in 1994; 1995; 2000; 2002; 2004 and The level of the + group was examined in 1995, 2004 and The effect of different time tapers was investigated in The S.E. threshold on fleets was examined in 1996; 2001 and The level of the catchability plateau was investigated in 1994; 1995; 2002; 2004 and Table C shows the history of VIIe plaice assessments and details the parameters used.

316 1686 ICES WGCSE REPORT 2013 D. Short-term projection Standard ICES software is used for the short-term projections-mfdp. No short-term forecast has been provided since 2006 as the review group deemed it unhelpful in the management of the stock given the strong retrospective bias in F. However WKFLAT was able to carry out a forecast following the removal of the strong retrospective bias in F. The diagnostics suggest that estimation of the recruiting year class (age 1) is poorly estimated in the assessment, both because catchability is very low in the commercial fisheries and because the surveys are very noisy at this age. Consequently, estimation of survivors from the recruiting age is poorly estimated and should not be used in the forecast. It was deemed more appropriate to estimate survivors at-age 2 on the basis of the geometric mean abundance of historic recruitment. The time period chosen should be consistent with that chosen for estimating future recruitment. Currently this could be formulated as. The short-term forecast uses: 1 ) the survivors at-age 3 and greater from the XSA assessment; 2 ) N at-age 2 = mean(ln(recruitment (1998 current year-1))*exp ( mean(f(age 1))); 3 ) Stock and Catch weights = average stock and catch weights over the preceding three years, unless there is an indication that there are strong trends in these, in which case they will be need to be dealt with appropriately by WGCSE; 4 ) The F vector used will be the average F-at-age in the last three years, unless there is strong indication of a significant trend in F. In the latter case the average selectivity pattern will be rescaled to the final F in the series. This procedure is in line with the convention used at WGCSE and the historic treatment of the short-term forecast for this stock. E. Medium-term projections F. Yield and biomass per recruit / long-term projections Standard ICES software is used for the long-term projections-mfypr. As with most plaice stocks, there is no clear stock recruitment relationship evident. Not carried for this stock between YPR projections run for G. Biological reference points WGCSE 2010-FMSY evaluation To derive an FMSY estimate the SRMSYMC package was employed and FMSY was calculated based on the three common stock recruit relationships; Ricker, Beverton Holt and smooth hockey-stick. Models were fitted using 1000 MCMC re-samples. For all three stock recruit relationships (SRR), all re-samples allowed FMSY and Fcrash values to be determined. All three models show that there is little evidence of a stock recruitment relationship with only limited information as to the trends at extreme levels of SSB.

317 ICES WGCSE REPORT The smooth hockey-stick model showed a break-off point in the SRR that was inconsistent with the data and as such was rejected. The yield per recruit estimates were highly uncertain with high CVs. Therefore these estimates were also rejected. The two SRR models have very different levels of estimated FMSY. Full diagnostics for all model fits can be found in the WGCSE 2010 report. STOCK RECRUIT RELATIONSHIP MODEL FMSY FCRASH Ricker Beverton Holt Therefore, the suggested level of FMSY for this stock is F s within the range of 0.14 and FMSY (and PA) reference points in use after the WGCSE2010 TYPE VALUE TECHNICAL BASIS MSY Approach Precautionary Approach MSY Btrigger 2500 t BPA FMSY 0.19 Provisional proxy by analogy with plaice in the Celtic Sea. Fishing mortalities in the range are consistent with FMSY Blim 1300 t Blim=Bloss The lowest observed spawning stock biomass. BPA 2500 t MBAL, biomass above this affords a high probability of maintaining SSB above Blim, taking into account the uncertainty in assessments. Flim Not defined. Fpa 0.45 This F affords low probability that (SSBMT< BPA). However the working groups since 2004 had considered the precautionary reference points for this stock as unreliable for the following reasons: The stock recruitment relation shows no evidence of reduced recruitment at low stock levels; The basis for BPA is weak, and heavily dependent on two consecutive points (1985 and 1986); FPA is based on BPA, and then this reference point is also rejected. In 2010, WKFLAT examined the stock dynamics provided by the new preferred XSA model based on migration at length to determine appropriate biological reference points for this stock on the basis of the new assessment. It concluded that the historic reference points for this stock were no longer appropriate as the new assessment indicated significant changes to the historical perspective of the stock caused by the inclusion of catches from VIId in the VIIe plaice stock. In the event that alternate assessment models be used, these reference point discussions will need to be repeated on the basis of the alternative model, as our understanding of stock dynamics are likely to be different for such a model. Examination of the biomass reference points indicated with some certainty that recruitment to the stock was not negatively impacted by SSB levels greater than 2200 t

318 1688 ICES WGCSE REPORT 2013 (Bloss (1996) following which a significant recovery in SSB of the stock had been observed, MBAL.), but there was little or no evidence of stock collapse at lower SSB levels Consequently, the group had difficulty in deciding whether this should be considered a limit reference point or a precautionary reference point. Dependent on this choice BPA would either be 2200 t (with a commensurate Blim set at 1600 t), or 3100 t (Blim = 2200 t) on the basis that there should be a 40% buffer between the two reference points (procedure consistent with the development of reference points in WGCSE). F reference points consistent with these biomass reference points based on a shortterm recruitment series were calculated on the basis of the yield per recruit calculations and shown in the table below as option 1 and 2. Bold numbers indicate the basis of the reference points for each option. OPTION 1 OPTION 2 OPTION 3 Blim BPA Flim FPA Option 1 indicates that Blim is lower than the observed spawning stock biomass for this stock, whilst option 2 suggests that Flim is higher than levels of F observed in the stock, therefore both sets of reference points would move to areas of stock dynamics not previously observed which the group considered risky. The new assessment indicates that the trend in F has been relatively flat since the late 1980s at levels around 0.6. Over this period SSB has increased and declined in response to recruitment, but without causing a collapse in the stock. It might therefore be considered as a limit reference point (Flim), option (3). The problem with this stock is that we have an insufficient understanding of the stock dynamics outside the relatively small range of Fs and little or no response in recruitment to the range of SSBs observed. Consequently, each of the choices made in considering the calculation of the other reference points is also precautionary so that the final set of reference points invariably is ultra-precautionary. The group could not come to a consensus with regards to suitable precautionary reference points but clearly stated that Fsq is currently too high and should be reduced, whilst biomass dynamics below the reasonably well estimated SSB levels of 2200 t are poorly understood. The group felt more confident in using the 2200 t as a Btrigger in the new advisory framework based on MSY based management targets, provided that the management intervention at this level of SSB was sufficient to move the stock away from this level of SSB with considerable certainty. It is deemed unlikely that low levels of SSB near Btrigger would be reached if long-term management aimed to attain F levels near an appropriate proxy of FMSY. No appropriate proxy was developed for FMSY given the current uncertainty over the basis for such advice, however the WKFLAT 2010 commented that because plaice are taken largely in conjunction with sole in Area VIIe it is important that the target levels between the stocks are consistent especially because a management plan has been agreed for sole VIIe.

319 ICES WGCSE REPORT Previous biological reference points proposed for this stock by the 1998 working group have been in use until 2009 (as below). Flim Not defined Fpa 0.45 (low probability that SSBMT<BPA) Blim 1300 t (equal to Bloss) BPA 2500 t (equal to MBAL) The recent working group s view of these reference points had been that they were considered unreliable. H. Other Issues I. References Burt, G., Goldsmith, D., Armstrong, M A summary of demersal fish tagging data maintained and published by Cefas. Science Series Technical Report No Cefas, Lowestoft. DC.J. Fox, B.P. Planque, C.D. Darby Synchrony in the recruitment time-series of plaice (Pleuronectes platessa L) around the United Kingdom and the influence of sea temperature. Journal of Sea Research 44: Hunter, E., Metcalfe, J.D., Arnold, G.P. and Reynolds, J.D Impacts of migratory behaviour on population structure in North Sea plaice Journal of Animal Ecology 73, Kell, L.T., Scott, R., and Hunter E Implications for current management advice for North Sea plaice: Part I. Migration between the North Sea and English Channel. Journal of Sea Research 51: Pawson, M.G Biogeographical identification of English Channel fish and shellfish stocks. Fisheries Research Technical Report No. 99. MAFF Directorate of Fisheries Research, Lowestoft.

320 1690 ICES WGCSE REPORT 2013 Spawning Areas Figure A. Map of spawning areas for VIIe plaice.

321 ICES WGCSE REPORT Table A. VIIe plaice - Catch Derivation table for assessment years YEAR OF WG 1981* length composition SOURCE DATA UK BELGIUM FRANCE DERIVATION OF INTERNATIONAL LANDINGS quarterly quarterly quarterly UK ALK used with French LDs ALK quarterly quarterly - UK+Belgium+France combined to total international Age composition 1982* As for * As for * As for * As for * As for * As for * As for * length composition quarterly quarterly - No analytical assessment carried out As for 1981 As for 1981 As for 1981 As for 1981 As for 1981 As for 1981 As for 1981 As for 1981 As for 1981 As for 1981 As for 1981 As for 1981 As for 1981 As for 1981 % SAMPLED 100 As for As for As for As for As for As for As for quarterly - - UK raised to total international ALK quarterly - - Age quarterly - - composition 1990 length composition quarterly - quarterly UK+France raised to total international ALK quarterly - quarterly Age quarterly - quarterly composition 1991 As for As for As for length composition - As for As for As for As for As for As for quarterly - quarterly UK ALKs applied to French LDs ALK quarterly - - UK+France raised to total international Age composition quarterly As for As for

322 1692 ICES WGCSE REPORT 2013 YEAR SOURCE DATA UK BELGIUM FRANCE DERIVATION OF INTERNATIONAL LANDINGS OF WG 1996 As for As for As for As for As for As for As for As for As for As for As for As for As for As for As for As for As for As for As for As for As for As for As for As for As for As for As for As for Migration correction added equal to 15% of Q1 VIId Landings from UK, Belgium and France. In addition, 15% Of Q1 Age comps added to the VIIe international AC. Also back calculated for years As 2009 with Netherlands VIId Q1 component added % SAMPLED * stock assessed as VIId,e plaice.

323 ICES WGCSE REPORT Figure B. UK(E+W) Otter-trawl fleet effort (hours fished) based on demersal landings.

324 1694 ICES WGCSE REPORT 2013 ALL Beam.Trawl.1 Calculated Effort Figure C. UK(E+W) Beam-trawl fleet effort (hours fished) based on demersal landings.

325 ICES WGCSE REPORT Table B. CV of numbers-at-age for commercial sampling. CV BY AGE YEAR COUNTRY UK(E+W) 18% 3% 3% 3% 6% 7% 11% 10% 9% 2006 UK(E+W) 21% 4% 3% 5% 5% 8% 10% 15% 14% 2007 UK(E+W) 42% 5% 3% 4% 6% 6% 9% 13% 20% 2008 UK(E+W) 42% 4% 4% 5% 6% 8% 8% 10% 14% 2009 UK(E+W) 39% 5% 3% 6% 7% 9% 11% 11% 16% 2010 UK(E+W) 17% 4% 3% 3% 7% 9% 14% 26% 23%

326 1696 ICES WGCSE REPORT 2013 Table C. History of VIIe plaice assessments. VIIE PLAICE - ASSESSMENT PARAMETERS USED ( ) BENCHMARK 1991* 1992* 1993* Assessment Age Range Fbar Age Range Assessment Method Tuning Fleets : LS/Trad VPA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA UK trawl yrs Ages UK trawl (historic) yrs Ages UK beam yrs Ages UK b/trawl survey yrs Ages UK FSP survey yrs Ages Time taper 20yr tri 20yr tri 20yr tri 20yr tri None None None None None None None None None None None None None None None None None Power model ages P shrinkage TRUE TRUE TRUE TRUE TRUE TRUE FALSE TRUE TRUE TRUE TRUE TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE Q plateau age F shrinkage S.E Num yrs Num ages Fleet S.E * Early version of XSA/VPA and tuning fleet age/year ranges used not specified. Assumed all years used but age range used uncertain.

327 ICES WGCSE REPORT 2013ICES WGCSE REPORT Appendix A. Beam-Trawl surveys in the western Channel (VIIe) 1. History of the survey Complaints from the fishing industry in the southwest about the lack of scientific investigation and knowledge of the local sole stock provided the catalyst for the survey in VIIe. Following enquiries of the local fishery officers and normal tendering procedures, a skipper-owned 300 hp beam trawler, the Bogey 1, was selected. The first year (1984) the survey consisted of a collection of tows on the main sole grounds. In 1989 the Bogey 1 was replaced with the Carhelmar and the survey continued unchanged until 2002 when R.V. Corystes took over the survey as an extension to its near-west groundfish survey. Due to the changes occurring through the time-series, the surveys completed on R.V. Corystes (2002 onwards) will be described separately to the previous surveys (pre 2002). 2.a. Survey objectives (1984 to 2001, and 2005 onwards) To provide independent (of commercial) indices of abundance of all age groups of sole and plaice on the west channel grounds, and an index of recruitment of young (1 3 year old) sole prior to full recruitment to the fishery. 2.b. Survey objectives (2002 to 2004) The primary objectives of the Irish Sea beam-trawl survey are to (a) carry out a 4 m beam-trawl survey of groundfish to i) obtain fisheries-independent data on the distribution and abundance of commercial flatfish species, and ii) derive age compositions of sole and plaice for use in the assessment of stock size; and (b) to collect biological data, including maturity and weight-at-age, for sole, plaice, lemon sole and other commercially important species. The epibenthic bycatch from these catches has been quantified, and these surveys are also used to collect biological samples in support of other Cefas projects and training courses. 3.a. Survey methods (1984 to 2001, and 2005 onwards) For the years the vessel was unchanged and was equipped with two 6 m chain-mat beam trawls with 75 mm codends. For the survey hauls one of the codends was fitted with a 60 mm liner. In 1989 the Bogey 1 was replaced by the latest design 24 m 300 hp(220 kw) beam trawler Carhelmar. In 1988 two commercial chain-mat 4 m-beam trawls (measured inside the shoe plates) were purchased by MAFF as dedicated survey gear. Both beams were fitted with the standard flip-up ropes and 75 mm codend. For years 1989 and 1990 only one codend was fished with a 40 mm liner but from 1991 with the introduction of 80 mm codends both were fitted with 40 mm liners. The vessel and gear has remained unchanged since Between 1989 and 2001 the survey remained relatively unchanged apart from small adjustments to the position of individual hauls to provide an improved spacing. In 1995 two inshore tows in shallow water (8 15 m) were introduced. The survey now consists of 58 tows of 30 minutes duration, with a towing speed of 4 knots in an area within 35 miles radius of Start Point. The survey design is stratified by distance from the coast bands, in contrast to the VIIa,f+g survey that is stratified by depth bands. The reason for this is that the coastal shelf with a depth of water less than 40 m is

328 1698 ICES WGCSE REPORT 2013ICES WGCSE REPORT 2013 relatively narrow and in addition is often fished with fixed gear. The survey bands (in miles) are 0 3, 3 6, 6 12, 12+ inshore, and 12+offshore. 3.b. Survey methods (2002 to 2004) The standard gear used is a single 4 m beam trawl with chain-mat, flip-up rope, and a 40 mm codend liner to retain small fish. The gear is towed at 4 knots (over the ground) for 30 minutes, averaging 2 nautical miles per tow. Fishing is only carried out in daylight, shooting after sunrise and hauling no later than sunset, as the distribution of some species is known to vary diurnally. Once onboard the catch is sorted to species level, with the exception of small gobies and sandeels, which are identified to genus. Plaice, sole, dab, and elasmobranches are sorted by sex, all fish categories weighed, and total lengths are measured to the full centimetre below, or half centimetre if the species is pelagic. Area stratified samples of selected species are sampled for weight, length, sex, maturity, and otoliths or scales removed for ageing. The standard grid of 58 stations was fished in 2002 and 2003 (see map), and although other stations have been fished in this period, they were for exploratory purposes and were not included in the assessment. 4. Abundance index calculation Plaice and sole abundance indices are calculated by allocating the appropriate ages to the fish that are caught. This gives the age composition (AC) of the catch, and this is used in the appropriate working group analysis. The ACs are calculated by proportioning a length distribution (LD) to an appropriate age length key (ALK). To account for possible population differences within ICES Division VIIe, biological samples are taken from sectors stratified by distance from shore (see map). The survey bands (in miles) are 0 3, 3 12, 12+ inshore, and 12+ offshore. Where appropriate the ALKs are separated by sex, and this allows a particular sector, depth-band and sex ALK to be raised to the corresponding LD to give an accurate AC for that particular habitat. The AC s can then be combined as required to give results in the form of numbers-at-age, per distance or time. Between 1984 and 1990 a total survey age length key was applied to the grid length distribution, but from 1990 onwards stratum stratified age length keys were used. The table below show the stratifications currently used to calculate the near-west ground fish survey abundance indices. 5. Map of survey grid Additional stations have been fished throughout the time period, but as these stations are not consistently fished, they are excluded from this map.

329 ICES WGCSE REPORT 2013ICES WGCSE REPORT = Stations 0-3 miles from shore = Stations 3-12 mils from shore = Stations 12+ miles inshore = Stations 12+ miles offshore 6. Summary AREA COVERED ICES DIVISION VIIE Target species Flatfish, particularly pre-recruit plaice and sole Time period September October to present. Gear used * 6m beam trawls * 4m beam trawls 1* 4m beam trawl 2005 Present 2 * 4m beam trawls Mean towing speed 4 knots over the ground Tow duration 30 minutes Vessel used F.V. Bogey F.V. Carhelmar R.V. Corystes 2005 Present - F.V. Carhelmar

330 1700 ICES WGCSE REPORT 2013ICES WGCSE REPORT 2013 Stock Annex 8.3: Sole in Division VIIe Stock Sole in Division VIIe (Western Channel) Date 13/03/2009 Revised by Sven Kupschus (revised at WKFLAT ICES, 2012) A. General A.1. Stock definition The management area for this stock is strictly that for Division VIIe. Biologically speaking however the picture is much less clear. Sole in general are relatively sedentary, once settled they perform a seasonal inshore offshore movements during their spawning migration with a random longshore component. Therefore the management unit of the stock is well defined for mature fish. There is good evidence to suggest that the stock is split into two biological stocks on either side of the Hurd Deep. If this prevents complete mixing of the stock it an assessment methodology capable of taking account of this should be applied. This could explain differences in the trends representative of stock dynamics in the different fisheries. The two main fisheries on the UK coast around Lyme Bay and the Start as well as the fishery on the coast in the eastern part of the management area are clearly separated by the deeper waters of the channel, so that the fishery covers only about half of the management area so that incomplete mixing may be a problem in this stock. With respect to the stock as observed by the fishery there seem to relatively few issues regarding stock identity and once recruited the stock appears to represent a closed population. Spawning migrations by sole tend to be in a seasonal onshore offshore pattern with a small random movement alongshore described for the species in other areas. Given the layout of the stock and the apparent breaks in the distribution of sole at the edges of the management unit there appears to be little concern for significant leakage across stocks. However the biological stock unit for Division VIIe is much less certain at the larval and prerecruit stage. The proportion of the area that represents nursery grounds is much smaller than those for other sole stocks of equal size, with only two small regions (the inner part of Lyme Bay and the Bay de Mount St Michelle) known to regularly produce 1-groups sole. Tagging information of juvenile sole, mostly 1 3 year olds show that there is significant ingress of recruits from the adjacent stock in ICES Division VIId from both the French and the UK coast that appear in the region of out Lyme Bay. Unfortunately, very little tagging data are available to examine if there is an equal or greater reciprocal movement in the opposite direction, but given the limited nursery habitat and the abundance of sole recruits in Division VIIe it seems reasonable to assume that there is a net inwards migration of prerecruits that remain in the area following maturation. Spawning is known to occur in the division from survey evidence in a relatively small concentration on the Bank de Langustine and intermittently in very low concentrations in the western part of the UK coastal region and around the edges of the Hurd Deep. Little is known about the fate of the spawning products, but given the relatively long egg and larval stage as well as the significant net eastward movement of waters in the channel it is plausible that the stock utilises nursery habitat in the eastern half of the channel. The degree of stock isolation in terms of these recruits has

331 ICES WGCSE REPORT 2013ICES WGCSE REPORT not been investigated, as it is possible that the recruits contribute to a common pool of recruits with the eastern stock. Isolation from the Celtic Sea (both the Bristol Channel and the Bay of Biscay) appears to be more rigorous according to tagging information, with few individual traversing the strong environmental and habitat gradients found in the rocky areas around lands end. However, the 1998 year class is indicated to be above average from all tuning information with the exception of the UK-BTS survey. The fact that this cohort is not well represented in what is thought to be the best indicator of recruitment, yet is readily observed in information from the more westerly and offshore parts of the stock area may indicate that there are other, as yet poorly understood recruit sources within the region. From a stock assessment point of view and in the absence of a modelled stock recruitment relationship there appears then relatively little concern over a lack of a closed population given the low movement rates post maturation. The low movement and its seasonality in conjunction with the high concentration of fishing effort around Start Point may produce effects of local depletion that may imply higher rates of fishing mortality for the UK-CBT fleet when compared to mortality rates from other indices covering a wider area. Such conjecture is potentially supported by the fact that when the new Q1SWBeam survey is viewed as an absolute index of abundance it produces higher estimates of stock size than the assessment. While stock size remains relatively stable and the behaviour of the fishery remains stable this is likely to have little impact on the assessment as the difference is absorbed in the estimates of catchability. If the fishery expands spatially with a commensurate reduction in the per-unit-area effort, or as migration rates change in response to stock size such affects may become more apparent in the assessment so that it is important to consider/examine such changes in future. The assessment method agreed by WKFLAT 2012 (ICES, 2012), and described in this stock annex, does not specifically deal with the uncertainty regarding stock boundaries, nor the issue of incomplete mixing and spatial dynamics in the stock and fishers. However, for advisory purposes the assessment methodology agreed at WKFLAT 2012 is able to provide robust advice despite these slight omissions. Part of the problem is that such process error is apparent in this stock only because of the high degree of precision and certainty in the data. Spatial issues are known to occur in other stocks, but the results of this process error are not apparent from the assessments because overall variability is much greater. A.2. Fishery The principal gears used for sole in the Western Channel are beam and otter trawls, for the UK fleet and entangling nets and otter trawls for the French fleet. In recent years, UK vessels have accounted for around three quarters of the total international landings, with France taking approximately a quarter and Belgian vessels the remainder. UK landings were low and stable between 1950 and the mid-1970s, but increased rapidly after 1978 as a consequence of the replacement of otter trawlers by beam trawlers. Because the UK fleet is the major component of the international landings, they follow a similar trend. Sole is the target species of an offshore beam-trawl fleet, which is concentrated off the south Devon and Cornish coasts, and also catches plaice and anglerfish. In recent years a winter fishery targeting cuttlefish has developed for the English beam-trawl fleet in the Western Channel, lasting from November till the end of March. This has taken some of the reliance of the fleet away from sole, but sole still represents a substantial portion of the catch during this time so it is

332 1702 ICES WGCSE REPORT 2013ICES WGCSE REPORT 2013 not clear to what degree the switch to cuttle-fishing has reduced fishing mortality on sole. Discarding of sole in this fishery is thought to be minor, supported by the time-series ( ) of discard information for the UK fleet shown in Figure A.2.1. Landings of sole reached a high level above 1400 t in the 1980s, boosted initially by high recruitment in the late 1970s, followed by an increase in exploitation. Landings declined between 1988 and 1991, following the recruitment of three below-average year classes ( ); since 1991 they have fluctuated between 800 t and 1100 t. Substantial quantities of sole caught in VIIe have been reported to two rectangles in VIId in order to avoid quota restrictions. Corrections for this misreporting were first made during the 2002 WG, but misreporting to other areas has been more difficult to identify. In addition, black landings are likely to have occurred to various degrees since quotas became restrictive in the late 1980s. No estimates of the scale of the problem exist so that this uncertainty has not been incorporated into the assessment process. Since the development of the beam-trawl fleet in the Western Channel in the early 1980s there has been a consolidation to larger more powerful vessels, particularly in the late 1990s and early 2000s. However, the severe quota restrictions at that time have lead to a reversal of this trend and a lesser emphasis on sole as the major income for the fleet. Undoubtedly sole still form the backbone of this fishery, due to the steady availability over the ground. However in recent years the fishery has adapted with smaller more flexible vessels and an overall reduction in kwh as well as a further small decrease in the number of boats due to a decommissioning scheme, to make the most of other resources such as scallops, cuttlefish, gurnards, etc. foregoing possible higher catch rates of sole. This is reflected in the offshore movement of the fishery around Start Point. At the lower catch rates described above the fleet is at an appropriate capacity to take the available quota and appears to have sufficient financial stability and certainty to allow for continued investment in the fishery. Were the industry to return to previous patterns of exploitation targeting the younger and more abundant sole in Lyme Bay it would almost certainly be able to increase the fishing mortality to levels greater than that assumed to be sustainable. The current enforcement regulations with a change in the attitude of the industry have meant that the TAC is an appropriate management tool in at least the UK fishery. Limiting days at sea further will have a perverse tendency to reverse this trend and focus effort grounds in Lyme Bay because of their proximity and the higher catch rates. A.3. Ecosystem aspects Little is known with regards of the effect of the environment on the stock dynamics of VIIe sole. Certainly the division is on the convergence between the Celtic Sea proper and the Channel/North Sea ecosystem. If predicted increases in temperature were to materialize changes to the stock dynamics of this and other species in the Division would be expected. To date there is good evidence of a sizeable increase in the abundance of bass in the area, a species with a similar pan European distribution as sole. In addition there is some anecdotal evidence of changes in the range of some species such as langoustine, triggerfish, and black sea bream from warmer parts of the Atlantic. In the North Sea it has also been suggested that cold periods immediately prior to spawning have a tendency to increase year-class strength and there is some indication of this for this stock, but no statistical analysis has been carried out to date.

333 ICES WGCSE REPORT 2013ICES WGCSE REPORT Beam trawling is known to have a significant impact on the seabed. It is understood though that those areas affected continue to be productive in terms of the target species. After the initial degradation of the habitat usually associated with the loss of sessile macro fauna, continued use of beam trawls seems to have few further impacts. B. Data B.1. Commercial catch UK (>60%) and France (>30%) together provide almost all the catches for this stock. UK Landings data are based on EU logbook data for VIIe catches. In 2002 the UK industry indicated that there had been substantial misreporting of landings to two rectangles in Division VIId. It was possible to identify the misreported landings spatially and by reported lpue. Having identified misreported landings, data were corrected back to 1985 by the 2002 WG. This method of correction is ongoing. French official landings statistics have been poor since 1997, but since 1997 landings data have been calculated much more accurately using buyer and sellers notes. France has provided corrected landings information to the working group since Numbers-at-age prior to 1994 are calculated by raising the UK age composition to UK and Channel Island Catches, adding the French age composition data, and finally raising the resulting age composition to the total international landings. From 1995 WG to 2005 WG the international landings for the stock were based entirely on English quarterly sampling effort then raised to quarterly international landings. Since 2006 WG French age data from 2003 onwards have been included. Numbers-at-age 1 in the catch are low or zero in most years and most likely reflect variation in the sampling, rather than variation in the stock itself. Therefore, these were not considered to add useful information and are replaced by zeros. Table A demonstrates the history of the derivation of catch numbers-at-age. B.2. Biological Weights-at-age Total international catch and stock weights-at-age for each year s catch data are calculated as the weighted mean of the annual weight-at-age data (weighted by catch numbers), and smoothed in-year using a quadratic fit so that: Wt = a + b*age +c* Age 2 where catch weights-at-age are mid-year values, and stock weights-at-age are 1 January values. Following the estimation of the weights-at-age catch-numbers are adjusted to so that the sum of products of the weights and catches sum to the estimated Landings (SOP correction). Catch numbers-at-age 1 are replaced by zeros, but the catch weights-at-age 1 were retained because they are part of the smoothing procedure and do not affect the assessment. They are also essential if a medium-term forecast is performed. A smoother is applied to sampled catch weights-at-age to adjust for variation in the weight-at-age that may result from low levels of sampling rather than differences in growth rate between cohorts. It also allows estimation of the stock weights-at-age by extrapolation of the curve rather than by using quarter 1 samples, which may be sparse. However this smoother is applied through the plus group and the age range

334 1704 ICES WGCSE REPORT 2013ICES WGCSE REPORT 2013 in the plus group is such that this will tend to overestimate the weights at the younger ages. This needs to be corrected as soon as possible. Natural mortality and maturity-at-age Natural mortality is assumed constant over ages and years at 0.1. This is consistent with the natural mortality estimates used for sole by other ICES working groups (WGNSSK: IV, VIId, WGCSE: VIIa, VIIfg, VIIIa,b) and consistent with estimates of M reported in Horwood, 1993 for VIIfg sole as well as other stocks and papers cited therein. Assessments prior to 1997 had use knife edge maturity-at-age 3. This was changed in 1997 to a maturity ogive from area VIIf and g according to Pawson and Harley (WD presented to WGSSDS in 1997), which is applied in all years, 1969 to present, since the 1997 WG. Age ,7, 12+ Prop. Mature Proportions of F and M before spawning are both set to zero to reflect the SSB calculation date of 1 January. B.3. Surveys UK-BTS The longest survey time-series available for this stock is the Western Channel Beam- Trawl Survey conducted by the UK in late September, early October (UK-BTS). The survey covers a relatively small area of VIIe from StartPoint through to the middle of Lyme Bay and out to the edges of the Hurd Deep covering the immediate area of fishing for the Brixham and Plymouth fleets. Sampling started originally in 1984 on the chartered commercial fishing vessel Bogey One, replaced in 1988 by the Carhelmar and moved to the research vessel Corystes in 2002 to Concerns were raised regarding differences in catchability between the Carhelmar and Corystes, and in 2003 the survey was carried out on both vessels. The results of the comparison convinced Cefas to return the survey to the long-serving Carhelmar and to replace the 2003 data with the data from the comparison trials in order to improve consistency. Consequently, the time-series has been largely recovered, with only 2002 and 2004 data coming from the RV Corystes. The survey cpue demonstrates a decline from 1986 to 1995 in line with the commercial data, after which SSB seems to have largely stabilized at lower levels. The abundance indices at-ages 1 and 2 demonstrate little overall trend, but ages 3 to 6 indicate a decline over the middle part of the series, despite intermittent peaks and troughs. More recently survey cpue has increased to the highest level over the consistent timeseries (starting in 1988 as used in the assessment) with the majority of the increase coming from the younger ages and only a marginal increase at the older ages. The age information is internally consistent to the survey, with 1989 year class is indicated to be strong at all ages and this year class can also be traced through the catch-at-age matrix. More recently the 1998 year class can be tracked reasonably consistently. UK-FSP A shorter, but more spatially extensive survey-series has been developed and managed by Cefas since 2003 in the UK in conjunction with the industry. Age sampling

335 ICES WGCSE REPORT 2013ICES WGCSE REPORT issues preclude the use of the data in the first year and the time-series is used here since The survey vessels (two separate trips are carried out annually see Annex 1 of this stock annex) are subject to a three yearly tendering procedure and vessels characteristics and gears used have changed over the time period, which is why the index has been standardised by meter beam and hour fished. The survey covers the extent of the UK fishery for the species including the less frequently exploited western part of the stock, which is why it is principally to be preferred over the more limited UK-BTS survey but is expected to be more variable due to the inconsistency of vessels used. Age information from this survey has shows evidence of some internal consistency in the medium age range but the series is too short to evaluate this at the older or younger ages at present. However the survey appears to show consistency with other survey indices and is therefore included in the present assessment for the entire age range available (ages 2 11). Data from this survey has been used in the plaice assessment since Q1SWBeam This survey was included in the assessment for the first time in WKFLAT The survey-series starts in Important considerations for WKFLAT 2012 (ICES, 2012) were that the survey is based on a stratified random survey approach and covers the entire region of the management area and some adjacent waters which may not fully conform to the delineation. The survey shows strong gradients in species composition within the western channel (justifying the stratification approach), although there is some indication that more appropriate post stratification could potentially provide an increase in precision of single species abundance estimates. Given sampling effort, fundamentally this survey is more variable than fixed stations survey designs of equal effort, but also inherently is less biased when there are potential changes in the distribution of the species within the area. Although estimates of survey variance of the limited dataseries are available, these are unlikely to reflect the full range of the variance that would be encountered in a longer time-series as variance estimates are unlikely to have reached their asymptote, particularly since the range of SSBs observed by the survey is very restricted. The survey-series was started in 2006 and surveys have been conducted consistently since then. To include as much information as is available at the time of the assessment working group the survey that is conducted in the first quarter has been shifted to back by one year and one age. This practical, because it adds further available information on the abundance of recruitment into the assessment, particularly important since there is uncertainty regarding the estimation of recruits from the UK- BTS which otherwise is the sole source of information of this parameter. The benefits of shifting the series were thought to out-weight the potential error that may be introduced by this procedure if the seasonal pattern of true F were to change in future. Age information provides estimates of abundance for all ages in the assessment, despite the fact that the survey only catches between 250 and 300 sole in a given year. Theoretically this removes the necessity of retaining the commercial lpue (at age) series required as the UK-BTS survey does not cover the full age range in the assessment. Internal consistency estimation is very difficult given the short time-series, and relatively small contrast in cohort strength observed (based on other series). Despite this some cohort tracking is apparent and the signal matches the cohort signal from other survey series, particularly the FSP survey.

336 1706 ICES WGCSE REPORT 2013ICES WGCSE REPORT 2013 Given these uncertainties regarding true survey variance and concerns regarding future funding for the survey it seemed unreasonable to put the entire weight on this survey, so at this stage it is not sensible to remove the commercial fleets from the assessment as they provide a high degree of precision at the cost of introducing some bias into the assessment. B.4. Commercial cpue The commercial tuning-series available for the assessment are the same as in previous assessments. Two historic surveys had been included in previous versions of the assessment because historically reference points in the stock had been based on historic development of the fishery and variance in the early time-series indicated considerable uncertainty with respect to these historic estimates as a response to the choice of plus group in the assessment. The new assessment is less susceptible to these variable estimates of catch-at-age, and the group decided to not base reference points on the historic development of the stock so that the historic indices are no longer required in the assessment and are not discussed further here. UK-COT The UK otter-trawl index is the same as presented in previous assessments. As previously observed the index suffers from two distinct negative year effects in and These inconsistencies were observed in previous assessments and the WG concluded that given the length of the period the effects of these in the historic period were minor on the current estimates of F and SSB as they are modelled mainly as residuals in the XSA model. For the new assessment there were no indications to presume that these effects were detrimental to the accuracy of the assessment so that the information is included as in previous years. Currently this fleet contributes only a small proportion of the overall landings, but it is sampled much more heavily than its representation in the landings so continues to provide a good independent time-series from the main commercial catches. It is uncertain whether the new DFC sampling will continue to provide such accurate data as the intent is to sample catches more proportional to landings. Despite the year effects the series is characterised by high internal consistency and is also consistent with other series in identifying strong cohorts. UK-CBT The time-series of commercial beam-trawl information has always formed the backbone of this assessment, but investigations at WKFLAT 2009 (ICES, 2009) indicated that this series showed declining lpue, particularly at the younger ages, in contrast to other information in the surveys and to a lesser degree to the catch-at-age despite the fact that the fleet accounts for around 60% of the landings in the stock. It was assumed that it was largely this fleet that was responsible for the persistent bias in the assessment. Historic area misreporting by the fleet prior to 2010 had been an issue, but after discussions with the industry in 2002 landings information and lpue data have been corrected for this, and the incidence of this practice had been decreasing. Increased scrutiny by enforcement, and lpue limits imposed by the producer organisation contributed to the reduction. The operation of the fleet was examined at this WK using VMS data from The conclusions from this analysis were that since 2006 the fleet has been increasingly shifting its effort southwards more into the central regions of the channel. Effort in

337 ICES WGCSE REPORT 2013ICES WGCSE REPORT Lyme Bay, the region where catch data and survey information indicate the majority of younger fish are found are now much lower than previously and have ceased almost entirely in 2010 and This shift in the selectivity towards older ages is very apparent also from the catch-at-age information for the fleet from market sampling records suggesting that it would be appropriate to split the fleet on the basis of inconsistent operation. It was not possible from independent information to discern when the majority of the contrast in this information occurred, and hence to decide on appropriate time to split the series, because VMS data are not available prior to Information from the industry also confirmed that there had been changes in the operation of the fleet, but again suggested that these changes had been gradual, rather than abrupt making the choice of the year for a split of this fleet difficult. The WK determined that 2002, the period when the area misreporting was officially acknowledged, would be an appropriate point for splitting and would also be suitable for the assessment, as this would retain a sufficiently long time-series over which to estimate the new catchabilities for the fleet. This new methodology was adapted and the UK-CBT fleet is used in the assessment as two fleets UK-CBT-early ( ) and UK-BTS-late ( ). B.5. Other relevant data None. C. Assessment: data and method Model used: extended survivor analysis. Software used: FLXSA (version 1.4-2)

338 1708 ICES WGCSE REPORT 2013ICES WGCSE REPORT 2013 Model Options chosen: and after Assmnt Age Range Fbar Age Range F(3 9) F(3 9) Assmnt Method XSA XSA Tuning Fleets Q1SWBeam (offset by 1y 1a) UK-FSP UK combined beam Ages (early) UK combined beam Ages (late) UK otter trawl Ages UK BTS yrs Ages Time taper No No Power model ages No No P shrinkage No No Q plateau age 6 6 F shrinkage S.E or 0.5* Num yrs 3 3 Num ages 5 5 Fleet S.E *Final decision on F shrinkage S.E will be made in 2013 at the WGCSE based on retrospective pattern. D. Short-term projection ICES has provided advice for this stock on the basis of a short-term forecast with the exception of 2009 when the advice was based on a trends only assessment. The assessment methodology developed at this benchmark meeting is determined to be appropriate to such projections and advice. This conclusion is largely based on the diagnostics of the assessment. The forecast methodology described below has not been specifically been evaluated at the benchmark, but given the biology of the species, the understanding of fleet dynamics and the similarity to previous assessment the previous procedure as described below is considered suitable. Input data Short-term forecasts require the input of a selection pattern, which is taken from the average of the last three years. In cases where a Fsq forecast is appropriate (i.e. where there is no documented trend in the level of F in the final three years) the selection pattern is scaled to the average F over the final three years. When there are significant changes in F over the last three years the selectivity pattern is rescaled to the final year to estimate catches in the interim year. When catches have been constrained at the level of the TAC a TAC constraint is implemented and the selectivity pattern is rescaled by the value of F that produces landings equal to the TAC for the interim year.

339 ICES WGCSE REPORT 2013ICES WGCSE REPORT Survivor estimates for fish greater than age three in the interim year are used in the projections. Recruits, including the last cohort in the assessment (age one, given as survivors at-age 2) are not thought to be particularly reliably estimated as they are poorly selected even in the inshore survey so their values is replaced by geometric mean recruitment determined as in the paragraph below depreciated for natural mortality. Recruitment in subsequent years is determined as geometric mean recruitment over the appropriate time-series. For this stock in recent year this is currently the entire time-series excluding the last two years (i.e for the 2011 assessment). Historically there have been periods where recruitment was thought to be lower or higher, in which case GM is calculated over a shorter recruitment-series, minus one year). E. Medium-term projections No longer applicable. F. Long-term projections Long-term projections are no longer carried out as part of the stock assessment procedure at working groups. However, STECF (SGMOS 9-02, SGMOS 10-06a) carried out long-term simulations as part of the management plan evaluations. The methodology examined the effects of different types of biases and uncertainty on the management of the stock running stochastic simulations under similar assumptions to the short-term forecast. This method was also employed to derive the level of MSY Btrigger by WKFLAT 2012 (ICES, 2012). G. Biological reference points Biological reference points in this stock were originally set in 1998 as described in the table below along with the reasoning and amended in 2001 to take account of a change to the assessment methodology. WG(1998)/ACFM(1998) since WG(2001)/ACFM (2001) Age range extended from to Flim 0.36 (Floss WG98) 0.28 (Floss WG01) FPA 0.26 (Flim*0.72) 0.20 (Flim*0.72) Blim 1800 t (Bloss= B73 WG98) 2000 t (Bloss= B00 WG01) BPA 2500 t (Blim*1.4) 2800 t (Historical development) The assessment methodology that formed the basis for these precautionary reference points was rejected by WKFLAT 2009 (ICES, 2009) and resulted in rejection of the reference points. ICES has adopted a provisional MSY Btrigger based on the former BPA as the technical basis. Having developed a new assessment methodology during WKFLAT 2012 (ICES, 2012) appropriate values for the assessment, given a sound technical basis, were determined as shown below.

340 1710 ICES WGCSE REPORT 2013ICES WGCSE REPORT 2013 Precautionary approach MSY approach Type Value Technical basis Blim 1300 t WKFRAME 2 meta-analysis (ICES, 2011) BPA 1800 t WKFRAME 2 meta-analysis (ICES, 2011) Flim FPA (unchanged since 2012). Undefined Undefined FMSY 0.27 Based on a suitably defined Fmax and stochastic LT simulations MSY Btrigger 2800 t Based on the lower 95% confidence limits of exploitation at Fmax from LT simulations. H. Other issues H.1. Sole in Division VIIe management plan A management plan was agreed for VIIe sole in 2007: Council Regulation (EC) No 509/2007 establishes a multiannual plan for the sustainable exploitation of sole in Division VIIe. Years were deemed a recovery plan, with subsequent years being deemed management plan. For the TAC was required to be at a value whose application will result in a 20% reduction in F compared with Fbar (03 05). If this value exceeded a 15% change in TAC, a 15% change in TAC was to be implemented. Fishing mortality <0.27 was reached in 2009, although the average fishing mortality over three years as prescribed by the management plan was only reached in After reaching FMSY=0.27 the stock is to be maintained at this level of fishing mortality. H.2 Historical overview of previous assessment methods Although this stock has been exploited historically for a long time at low levels, official landing statistics and catch-at-age data are available from 1969 onwards. At this time landings were 353 t mainly attributable to otter trawlers and netters. The development of a beam-trawl fleet in UK waters lead to rapid increases in landings from the stock in the late 1970s which resulted in a commensurate decline in SSB after an initial increase in stock size to its maximum in 1980 as a consequence of particularly good recruitment in The decline as assessed by XSA occurred despite some subsequent good recruitment in 1980, 1984, 1986 until 1990 where the SSB appears to have level out near 3000 t. More recent estimates of recruitment are estimated to be high again and SSB has started to increase in response to this recruitment and reduced fishing mortality since the introduction of the single area licence since the end of Fishing mortality appears to have been stable in the fishery since the early 1980s at around 0.3 before declining to near FMAX since Key uncertainties with regards to the data quality/assessment quality of this stock are the uncertainty regarding the degree of mixing between this and adjacent stock, particularly with regards to recruitments, the fact that the survey covers only a small portion of the stock the lack of a discernible stock recruit relationship which does not allow us to determine reference points with any certainty. Table B demonstrates the history of Division VIIe sole assessments and details the assessment model used (XSA) and the parameters and settings used in each year s assessment until 2008.

341 ICES WGCSE REPORT 2013ICES WGCSE REPORT I. References ICES Report of the Benchmark and Data Compilation Workshop for Flatfish (WKFLAT 2009), 6 13 February 2009, Copenhagen, Denmark. ICES CM 2009/ACOM: pp. ICES Report of the Workshop on Implementing the ICES FMSY Framework (WKFRAME- 2), January 2011, ICES, Denmark. ICES CM 2011/ACOM: pp. ICES Report of the Benchmark Workshop on Flatfish Species and Anglerfish (WKFLAT 2012), 1 8 March 2012, Bilbao, Spain. ICES CM 2012/ACOM:47. Horwood, J The Bristol Channel sole (Solea solea (L.)): a fisheries case study. Advances in marine biology. 1993, vol. 29, pp Pawson, M.G and Harley, B.F.M Unpublished. Cefas internal report).

342 1712 ICES WGCSE REPORT 2013ICES WGCSE REPORT 2013 Table A. VIIe Sole. Catch derivation table for assessment years Year of WG source Data UK France derivation of international 1981 length composition landings quarterly quarterly UK ALKs applied to French LDs ALK quarterly - UK+France raised to total international Age composition quarterly As for As for As for As for As for length composition As for 1981 As for 1981 As for 1981 As for 1981 As for 1981 As for As for As for As for As for quarterly quarterly UK+France raised to total international ALK quarterly quarterly Age quarterly quarterly composition 1988 As for As for As for As for As for As for length composition As for 1987 As for 1987 As for 1987 As for 1987 As for 1987 As for 1987 % sampled As for As for As for As for As for As for quarterly quarterly UK ALKs applied to French LDs ALK quarterly - UK+France raised to total international Age quarterly - composition 1995 length quarterly - UK raised to total international 81 composition ALK quarterly - Age quarterly - composition 1996 As for As for

343 ICES WGCSE REPORT 2013ICES WGCSE REPORT Year of WG source Data UK France derivation of international 1997 As for As for As for As for As for As for length composition ALK As for 1995 As for As for As for As for As for As for As for As for 1995 landings sampled - As for As for As for As for As for As for quarterly UK and French raised to total ~95% international biannually ~95% As for 2003 As for 2003 As for 2003 As for 2003 As for 2003 As for 2003 As for 2003 As for 2003 ~95% As for 2003 ~95% As for 2003 ~95% As for 2003 ~95% As for 2003 ~95% As for 2003 ~95% As for 2003 ~95% %

344 1714 ICES WGCSE REPORT 2013ICES WGCSE REPORT 2013 Table B. History of VIIe sole assessments Assmnt Age Range Fbar Age Range F(3-8) F(3-7) F(3-7) F(3-7) F(3-7) F(3-7) F(3-7) F(3-7) F(3-7) F(3-7) F(3-7) F(3-7) F(3-7) F(3-7) F(3-7) F(3-7) F(3-7) F(3-7) Assmnt Method L.S. XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA XSA Tuning Fleets UK Inshore beam Ages UK Offshore beam Ages UK < 24m beamtr Ages UK > 24m beamtr Ages UK combined beam Ages UK otter trawl Ages UK BTS yrs Ages Time taper yr tri yr tri yr tri yr tri No No No No No No No No No No No No No Power model ages No No No No P shrinkage Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No No No No Q plateau age F shrinkage S.E Num yrs Num ages

345 ICES WGCSE REPORT 2013ICES WGCSE REPORT Fleet S.E Assmnt Age Range Fbar Age Range F(3-9) F(3-9) F(3-9) Assmnt Method Trends XSA XSA XSA Tuning Fleets UK Inshore beam Ages UK Offshore beam Ages Q1SWBeam (offset by 1y 1a) UK-FSP UK combined beam Ages (early) UK combined beam Ages (late) UK otter trawl Ages UK BTS yrs Ages Time taper No No No Power model ages No No No P shrinkage No No No Q plateau age F shrinkage S.E Num yrs Num ages Fleet S E

346 1716 ICES WGCSE REPORT UK discards in 2002 Dis Q1 Dis Q2 Dis Q3 Dis Q4 Ret Q1 Ret Q2 Ret Q3 Ret Q UK discards in 2003 Dis Q1 Dis Q2 Dis Q3 Dis Q4 Ret Q1 Ret Q2 Ret Q3 Ret Q UK discards in 2004 Dis Q1 Dis Q2 Dis Q3 Dis Q4 Ret Q1 Ret Q2 Ret Q3 Ret Q UK discards in 2005 Dis Q1 Dis Q2 Dis Q3 Dis Q4 Ret Q1 Ret Q2 Ret Q3 Ret Q UK discards in 2006 Dis Q1 Dis Q2 Dis Q3 Dis Q4 Ret Q1 Ret Q2 Ret Q3 Ret Q UK discards in 2007 Dis Q1 Dis Q2 Dis Q3 Dis Q4 Ret Q1 Ret Q2 Ret Q3 Ret Q UK discards in 2008 Dis Q1 Dis Q2 Dis Q3 Dis Q4 Ret Q1 Ret Q2 Ret Q3 Ret Q Figure A2.1. Time-series of UK discard data raised to trip information. Figure B4.1. Areas used for the calculation of lpue time-series exploring temporal changes in the distribution of stock and effort.

347 ICES WGCSE REPORT Stock Annex 9.2: Pollack in Subareas VI and VII (Celtic Sea and West of Scotland) To be updated.

348 1718 ICES WGCSE REPORT 2013 Stock Annex 9.3: Grey gurnard in Subarea VI and Divisions VIIa c and e k (Celtic Sea and West of Scotland) To be updated.

349 ICES WGCSE REPORT Stock Annex 10.1: Stock Working Group European sea in Subarea VIb,c and VIIa,d h European sea bass (Dicentrarchus labrax) in Subarea IVb,c and VIIa, d h (BSS-47) WGCSE Date May 2013 Revised by Mike Armstrong, May 2013 (stock annex developed by IBPNEW 2012, retaining only information for BSS-47) A. General A.1. Stock definition Bass Dicentrarchus labrax is a widely distributed species in northeast Atlantic shelf waters with a range from southern Norway, through the North Sea, the Irish Sea, the Bay of Biscay, the Mediterranean and the Black Sea to North-west Africa. The species is at the northern limits of its range around the British Isles and southern Scandinavia. Stock structure of sea bass in the Northeast Atlantic has been reviewed by WGNEW 2012 and IBP-NEW 2012 based on evidence from genetics studies, tagging studies, distribution of commercial catches and similarities in stock trends between areas, drawing also on extensive information contained in previous WGNEW and ICES SGBASS reports. IBP-NEW considers that stock structure remains uncertain, and recommends further studies on sea bass stock identity, using conventional and electronic tagging, genetics and other individual and population markers (e.g. otolith microchemistry and shape), together with data on spawning distribution, larval transport and VMS data for vessels tracking migrating bass shoals, to confirm and quantify the exchange rate of sea bass between sea areas that could form management units for this stock. Such information is critical to support development of models to describe the spatial dynamic of the species under environmental drivers (eg. temperature and food). Such a modelling work is being carried out in France in the framework of a PhD study (R. Lopez). The pragmatic view of IBP-NEW 2012 is to structure the baseline stock assessments into four units: Assessment area 1. Sea bass in ICES Areas IVbc, VIId, VIIe,h and VIa,f&g (lack of clear genetic evidence; concentration of Area IV bass fisheries in the southern North Sea; seasonal movements of bass across ICES Divisions). Relatively data-rich area with data on fishery landings and length/age composition; discards estimates and lengths; growth and maturity parameters; juvenile surveys, fishery lpue trends. Assessment area 2. Sea bass in Biscay (ICES Subarea VIIIa,b). Available data are fishery landings, with length compositions from 2000; discards from 2009; some fishery lpue. Assessment area 3. Sea bass in VIIIc and IXa (landings, effort). Assessment area 4. Sea bass in Irish coastal waters (VIa, VIIb, VIIj). Available data: Recreational fishery catch rates; no commercial fishery operating.

350 1720 ICES WGCSE REPORT 2013 Fishery landings of sea bass are extremely small in Irish coastal waters of VIIa and VIIg and the stock assessment for assessment rea 1will not reflect the sea bass populations around the Irish coast, which may be more strongly affiliated to the population in area 4 off southern, western and northern Ireland. A.2. Fishery General description The commercial sea bass fisheries in Areas IV and VII have two distinct components: an offshore fishery on pre-spawning and spawning bass during November to April, predominantly by pelagic trawlers from France and the UK, and small-scale fisheries catching mature fish returning to coastal areas following spawning and in some cases immature sea bass. The inshore fisheries include many small (10 m and under) vessels using a variety of fishing methods (e.g. trawl, handline, longline, nets, rod and line). The fishery may be either targeting sea bass or taking sea bass as a bycatch with other species. Historical landings data for the small-scale fisheries have often been poorly recorded. Although sea bass can occur as target or bycatch of many vessels, the bulk of the catch can be taken by relatively few vessels. For example in the UK in 2010, sea bass landings were reported by 1480 vessels (including 1207 of 10 m and under), 10% of which were responsible for over 70% of the total landings of 719 t (Walmsley and Armstrong, 2012). For France, in 2009 sea bass landings were reported by 2226 vessels including 976 of 10 m and under. Three main métiers were responsible for over 83% of the total landings. Pelagic trawlers (31.5% of total landings, for 58 vessels and 276 seamen) and "liners+handliners (21.7% of total landings for 416 vessels and 634 seamen) are very economically dependent of this species (Drogou et al., 2011). French bottom trawlers often do not target sea bass, but this gear does represent 30.1% of the total landings (for 832 vessels and 2769 seamen). (Drogou et al., 2011). According to the CHARM 3 Atlas of the Channel Fisheries, sea bass production in value represented in It s the third most valuable species caught in the Channel (source: Agrimer) in 2008 behind sole and monkfish (tuna is not included in statistics). The market value sea bass depends greatly on how its caught, giving added value to certain métiers: according to CHARM3 Atlas of the Channel Fisheries, mean price of sea bass sold in the Channel (7EH+7D) by liners was per kg in 2007 compared with 6.52 per kg for pelagic trawl, reflecting differences in volume and fish condition. Sea bass are a popular target for recreational fishing in Europe, particularly for angling in the UK, Ireland and France, and increasingly in parts of southern Norway, the Netherlands and Belgium. Relatively little historical data are available on recreational fisheries although several European countries are now carrying out surveys to meet the requirements of the EU Data Collection Framework and for other purposes (ICES WKSMRF 2009; PGRFS 2010 & 2011; WGRFS 2012; Herfault et al., 2010; Rocklin et al., 2012 in prep; Van der Hammen and De Graaf, 2012). More detailed descriptions of national fisheries can be found in ICES SGBASS (ICES 2004a). Fishery management regulations Sea bass are not subject to EU TACs and quotas. Commercial vessels catching bass within cod recovery zones are subject to days-at-sea limits according to gear, mesh

351 ICES WGCSE REPORT and species composition. Under EU regulation, the MLS of bass in the Northeast Atlantic is 36 cm total length, and there is effectively a banned range for enmeshing nets of mm stretched mesh in Regions 1 and 2 of Community waters1. A variety of national restrictions on commercial bass fishing are also in place. These include: a landings limit of 5 t/boat/week for all French and UK trawlers landing bass; closure of 37 bass nursery areas in England and Wales to specified fishing methods; UK regional byelaws in Cornwall and South Wales stipulating a 37.5 cm MLS; a minimum gillnet mesh size of 100 mm in South Wales; a variety of control measures in Ireland that effectively ban commercial fishing for bass in Irish waters; plus MLS of 40 cm;. a licensing system from 2012 in France for commercial gears targeting sea bass. voluntary closed season from February to mid-march for longline and handline bass fisheries in Brittany; Depending on country, measures affecting recreational fisheries include minimum landing sizes, restrictions on sale of catch, bag limits (Ireland), and gear restrictions (France; Netherlands). A.3. Ecosystem aspects Temperature appears to be a major driver for bass production and distribution (Pawson, 1992). Reynolds et al. (2003) observed a positive relationship between annual seawater temperature during the development phases of eggs and larvae of sea bass and the timing and (possibly) abundance of post-larval recruitment to nursery areas. In addition, early growth is related to summer temperature and survival of 0-groups through the first winter is affected by body size (and fat reserves) and water temperature (Lancaster, 1991; Pawson, 1992). Prolonged periods of temperatures below 5 6 C may lead to high levels of mortality in 0-groups in estuaries during cold winters. As a result, any SSB recruit relationships may be obscured by temperature effects (Pawson et al., 2007a). B. Data B.1. Commercial catch B1.1 Landings data Data available Landings series for use in the assessment are available from three sources: 1 Region 1: All waters which lie to the north and west of a line running from a point at latitude 48 ºN, longitude 18 W; thence due north to latitude 60 N; thence due east to longitude 5 W; thence due north to latitude 60 30'N; thence due east to longitude 4 W; thence due north to latitude 64 N; thence due east to the coast of Norway. Region 2: All waters situated north of latitude 48 N, but excluding the waters in Region 1 and ICES Divisions IIIb, IIIc and IIId.

352 1722 ICES WGCSE REPORT 2013 i ) ii ) iii ) Official statistics recorded in the FishStat database since around the mid- 1970s. French landings for from a separate analysis by Ifremer of logbook and auction data. Survey estimates of landings from the UK fleet of 10 m and under vessels (which are not obliged to provide EU logbooks), carried out by Cefas Total international landings of sea bass in European waters from sources (i) and (ii) combined increased from around 2000 t in the late 1970s to over 8000 t by 2006, the bulk coming from Areas IVb,c, VIIe and VIII. An important driver of the increase in landings since the 1990s was the increased landings in Divisions IVb,c, VIId and VIIe,h, coinciding with the large 1989 year class and a northward expansion of the sea bass population in the North Sea during a period of increasing sea temperatures. Quality of official landings data From 1999 onwards, French landings data from FishStat are replaced by more accurate figures from a separate analysis of logbook and auction data carried out by Ifremer, in which landings have been correctly allocated to fishing ground. The timeseries for each component fishing ground therefore has a step change around To create input landings data from the 1980s for the North Sea, Channel and Celtic and Irish Seas, it has been necessary to assume that the ratio of FishStat to Ifremer landings figures for 2000 onwards can be applied to Fishstat figures for earlier years. The sensitivity of the assessment to this adjustment should be evaluated. Factors for adjustments have been calculated for each area and applied to the French landings data from FishStat: for IVbc+VIId area 1.04 has been used; for VIIeh area 1.6 has been used, and for VIIafg 0.62 has been used. The accuracy of total landings statistics for Subareas IV and VII are expected to have improved further since 2006 since the introduction of the registration of Buyers and Sellers in the UK, particularly for small vessels that do not have to supply EU logbooks. The accuracy of Dutch landings data from Area IV and VII has improved since the recreational line fishers were registered as commercial fishers when they want to sell their landings. Cefas bass logbook estimates of landings The UK has previously attempted to estimate the sea bass landings of inshore commercial and recreational fishing boats between 1984 and 2006 using a voluntary logbook scheme in conjunction with a biennial census of vessels catching sea bass that covers different segments of coast in different years (Pickett, 1990). Estimates of annual catch and effort are obtained from a stratified selection of vessels issued with a bass logbook, and raised to the census counts of vessels in the same survey strata. The landings tables in previous WGNEW and ACOM advice (up to 2011) included unallocated landings which were the difference between the voluntary logbook estimates and the official UK statistics in each ICES area. The coverage of the logbook scheme has declined substantially. A review of the scheme in 2012 (Armstrong and Walmsley, 2012a) showed that the previous estimates provided to WGNEW included recreational charter boats, which have now been removed from the estimates for the years when they occur. Coverage of trawls has been extremely low. The Cefas logbook estimates for nets and lines still show substantial differences with official estimates, even for recent years since 2006 when the Registration of Buyers and Sellers

353 ICES WGCSE REPORT has vastly improved recording of landings by 10 m and under vessels. The utility of the logbook data are reviewed in the IBP-NEW 2012 benchmark assessment report. Further information on availability and quality of landings data by country is provided by SGBASS (ICES, 2004). B1.2. Discards estimates UK data Survey design and analysis The UK sampling scheme involves a vessel list sampling frame and random selection of vessels within strata defined by quarter, area and fleet métier. The vessel list for each quarter is stratified by area and predominant métier. A random, ranked draw list is generated each quarter for each vessel stratum, and observers work down the list to board the next available vessel on completion of a previous trip, according to targets for numbers of trips per stratum. Numbers and length compositions of discards and retained fish for each sampled trip are estimated by random sampling of the catch from a minimum of 60% of hauls during a trip. Estimation of annual bass discards and length compositions of landed and discarded fish at the fleet level are obtained by raising from sampled trips within each stratum then combining over strata. Discards estimates for IBP-NEW 2012 were obtained by use of a ratio estimator (auxiliary variable = landings), as there was evidence for a linear relationship between landings and discards of sea bass at the trip level. Data coverage and quality UK discards data are available for métiers associated with trawls and fixed/driftnets only. Discards from commercial line boats are expected to be relatively low and have high survival, so this fleet sector is excluded from the scheme for sea bass. As sampling is targeted at all species, annual coverage of the bass fisheries is relatively limited. Sample numbers by gear type and area are highest for otter trawls and nets (see benchmark assessment report), but of these, a variable and often low number of trips have bass catches. Only length frequency data are available for discarded sea bass. French data Survey design and analysis The French sampling schemes also utilise vessel list sampling frames and random selection of vessels within strata defined by area and fleet sector. From the activity calendars of French vessels for year n-1, vessels are grouped by the métiers practiced. Thus, a vessel may belong to multiple groups if practicing several métiers in the period. If the métier has to be sampled in priority No. 1, the vessel to be boarded is chosen randomly within this group of vessels. The observer then chooses to go onboard for a trip. During the trip, the fishing operations corresponding to métier no. 1 are sampled. Optionally, if the vessel practices several métier during the trip, fishing operation of the métier no.2 will also be sampled if the métier no. 2 is included in the annual sampling plan. If the métier is not part of the plan, it is requested to sample at least one fishing operation of this métier in the trip. (complete document on sampling protocol in French : pdf)

354 1724 ICES WGCSE REPORT 2013 Data coverage and quality Discards data are only available for French fleets from 2009 onwards, and only as length frequencies. Spain No bass discards were observed for any métier in the period. Number of sampled hauls per métier and area were presented to IBP-NEW 2012 (see assessment report). Discards data from other European countries Discards data for Dutch beam trawlers were presented to ICES IBP-NEW 2012, as annual mean numbers discarded per hour in No commercial fisheries for sea bass exist in Ireland. B1.3. Recreational catches Recreational marine fishery surveys in Europe are still at an early stage in development (ICES WKSMRF 2009; PGRFS 2010 & 2011; WGRFS 2012). The following information was available to WGNEW Data from France The first national survey of recreational fishing in France (2006 to 2008) revealed that sea bass was the main target species for recreational fishermen, and that people had fished recreationally for bass. Survey method and analysis A new study targeting sea bass was conducted between 2009 and In 2009, households were phoned in the targeted districts using random digit dialling (RDD). The main goal was to estimate the population of sea bass recreational fishers and their socio-demographic profiles in the Bay of Biscay and in the Channel. In a panel of 258 recreational fishermen was recruited during the RDD screening survey and kept diaries of their catches for one year for a total of 1170 trips. The main goal was to obtain a detailed description of fishing trips (travel, area of fishing, gears ) and the description of their catches (species, weight, length ) to be used for assessment. Sea bass catches were estimated by raising the mean annual estimates from the diaries to the total population of recreational fishers from the RDD survey, by survey stratum. The estimated recreational catch of bass in the Bay of Biscay and in the Channel was 3170 t of which 2350 t was kept and 830 t released. The main gears used, in order of total catch, were fishing rod with artificial lure, fishing rod with bait, handline, longline, net and spear fishing. Approximately 80% of the recreational catch was taken by sea angling (rod and line or handline) t total catch and 1840 t kept (29% release rate). Around 60% of the recreational catch estimate was from Bay of Biscay. Data quality The precision of the estimate is relatively low (CV =-51%). Increasing the panel from 121 to 500 fishermen would be expected to improve precision to 25% and increasing this panel to 1000 would improve precision to 18%.

355 ICES WGCSE REPORT UK(E&W) Several attempts have been made in the past to estimate recreational sea angling catches of sea bass in England and Wales or more restricted areas of the UK (Dunn et al., 1989; Dunn and Potten, 1994). A new survey programme based on a statisticallysound survey design commenced in 2012 to estimate fishing effort, catches (kept and released) and fish sizes for shore based and boat angling in England. The survey does not cover other forms of recreational fishing. Estimates will not be available until late Netherlands Sea bass are taken by recreational sea anglers in the Netherlands. A recent survey investigated the amount of sea bass caught (Van der Hammen and De Graaf, 2012; ICES WGRFS, 2012). Survey method and analysis Research was conducted from 2009 onwards. First a screening survey was carried out to identify fishing households, profile fishing households and select participants for a follow-up. In the screening survey, people were approached by phone in order to estimate the number of inland and marine recreational fishers. Questions were asked to assess the age, gender and avidity of the fishers. The screening survey was followed by a diary survey, in which 1043 marine recreational fishers returned their diary at least once. Fishing (and economic) activity through regular contact (monthly) by survey interviewers was monitored. In addition, a small-scale on-site sampling programme was implemented to provide additional independent data on catch, size and species composition of recreational fishers along the coast and charter boats. Preliminary results of these surveys show that in total about 360 thousand individual sea bass were caught in Of these, 218 ± 130 (95% CI) thousands were retained, which is about 61%. In weight, 161 tonnes of sea bass were caught in total. Of this, 96 ± 60 (95% CI) tonnes were retained, which is 60%. These results are mainly applicable to Subarea IV. Other countries Sea bass are a popular angling species in Ireland and are also caught in Belgium. Time-series of sea angling catch rates of sea bass in southern Ireland were presented at IBP-NEW2012 by a stakeholder representative. B.2. Biological sampling B2.1. Length and age compositions of landed and discarded fish in commercial fisheries Length and age compositions of sea bass landings were available to WGNEW and IBP-NEW 2012 from sampling in the UK and France. UK Sampling methods and analysis The UK(E&W) sampling programme for length compositions of sea bass covers sampling at sea and on shore. The sampling design for at-sea sampling is described above. The onshore sampling programme uses an area list frame comprising port

356 1726 ICES WGCSE REPORT 2013 days, currently stratified by quarter, ICES division and an index of port size. Large ports are sampled more intensively than small ports. Separate list frames of ports are established for pelagic trawlers, beam trawlers and demersal trawl, nets and lines. Sampling targets are set to achieve a specified number of port visits by stratum, taking account the need for fleet based as well as stock-based data specified by the EU Data Collection Framework, although other diagnostics are monitored such as numbers of fish measures and otoliths/scales collected by species. This scheme has only been in development and operation since around 2010 when Cefas took over the sampling from the Marine and Fisheries Agency. Prior to then, the sampling targets were mainly set as numbers of fish of each species to measure or age by quarter, district, and gear groupings, with minimum numbers of sampling trips also specified to spread the sampling out. Length compositions are first vessel-raised using ratios of landed live weight to predicted live weight of the length frequency calculated from a length weight relationship: W (kg) = (L+0.5 )2.969 Raised LFDs are then summed over vessels within a sampling stratum and raised to give total raised fleet LFDs per stratum, which are then combined. This procedure ensures sums-of-products ratios of 1.0, but will lead to some bias in numbers-atlength due to discrepancies between true fish weights and calculated fish weights from the length weight relationship. Data coverage and quality Length and age compositions are supplied by the UK since 1985 for IVb&c, VIId, VIIe,h and VIIa,f&g, disaggregated by five gear types: otter trawl, pelagic pair trawl, drift and gillnets, lines, and other gears. Although separate ALKs are derived for the five areas, the same ALK is applied to all gear groups meaning that the age composition estimates for the different gears are not independent.uk sampling rates for length compositions have been very variable between area, gear and year strata. Most strata have some sampling coverage with the exception of pair trawls which have had zero or very low coverage in many years despite large catches, although sampling has improved recently (see assessment report). The sampling rate (trips sampled per tonne landed) has declined for all gears since the mid 2000s. France Sampling methods and analysis The French sampling programme for length compositions of sea bass covers sampling at sea and on shore. Since 2009, both sampling types are first based on métiers composition and their relative importance per fishing harbours and month. Both are also designed to sample the whole catch following a concurrent sampling of species, potentially leading to low sea bass sample size. In order to complement this effort, specific sampling for sea bass at the market is added at times and harbours when higher landings are occurring, especially from métiers targeting sea bass. The sampling frame is based on the main harbours, gear types (or grouping of métiers) and month and is available to all samplers on a dedicated website. Real-time follow-up of the plan, refusal rates and their reasons, time taken to sample, all this information is also available from the website, together with sampling protocol (in French):

357 ICES WGCSE REPORT pdf). Before 2009, only market specific sampling was in place, and the sampling plan was designed and followed by the stock coordinator. The French sampling programme for age compositions of sea bass is based on age length keys with fixed allocation. For the VIIeh area, quarterly French landings at auctions are sampled in order to collect five scales (from 2000 to 2008) or three scales (from 2009) by length class (cm). For the VIIIab area the information is available only from For other areas the information is not available. All length samples are populated in a central data base (Harmonie) and regular extracts are available in the COST format. Raising the data to the population is done using COST tools and a special forum for discussing the outcomes of the analysis is held every year in March, in order to gather all stock coordinators and prepare the datasets for the assessment working groups. Data coverage and quality Sampling has been very variable between areas and gears, with greatest consistency between years in the neighbouring stock unit in VIIIa,b. There has been a general increase in numbers of trips sampled for length since 2009 (see assessment report). Quarterly landings length and age compositions are available for all métiers in VIIeh area from For all other areas, only length compositions are available. For VIId, quarterly length distributions are available from 2003 for bottom trawl and pelagic trawl. For IVbc length distributions are available from 2009 for various gears. The statistical design of fishery sampling schemes has undergone change in recent years in the UK and France, following recommendations from ICES workshops on sampling survey design, with a move towards more representative sampling across trips within fleet segments. This can result in sampling more trips that have small catches of bass, and is one reason for the increase in numbers of sampled trips with bass since 2009 in France which does not imply an increase of the proportion in numbers of fish measured per trip. Other countries Fishery landings length or age compositions from other countries catching bass were not available to WGNEW or IBP-NEW The Netherlands did collect age samples of sea bass every year from 2005 to From 2010 onwards, age samples are collected only once every three year. Otoliths and scales that are retrieved from the fish are sent to Cefas in the UK for age reading. Length samples are collected every year. All samples are collected in the auctions where most sea bass is landed, in the south of the Netherlands. The quality of the data is good enough to use them in assessments. However, both the length and age data need processing before they can be inserted in an assessment. Effective sample sizes for length and age compositions The effective sample size for annual estimates of length or age composition lie between the number of trips sampled and the number of fish measured or aged, due to cluster sampling effects. Effective sample sizes have not been computed yet for UK and French sampling data for sea bass. In the meantime, numbers of fishing trips sampled for length or age could be used as an annual measure of relative precision of datasets.

358 1728 ICES WGCSE REPORT 2013 Accuracy and validation of age estimates Age reading consistency Consistency in age reading of sea bass between four operators in Cefas and Ifremer was examined during a limited exchange of otolith and scale images between laboratories in 2011, organised by the ICES Planning Group on Commercial Catches, Discards and Biological Sampling (Mahé et al., 2012). A total of 155 fish of cm was sampled onboard French research vessels during two international surveys. The precision of ageing was similar for scales and otoliths. The coefficient of variation of age readings for individual fish was around 12% implying a standard deviation of +/- 1 year for a ten year-old fish, with relatively few fish having identical readings by all four operators. However it was noted by the operators that photographic images were more difficult to evaluate than original age material, which was likely to have a negative effect on the consistency of ageing. These results provide no indication of the validity of ages, only the consistency between operators, and cannot indicate data quality in earlier years when different operators provided the age data. A more extensive age exchange is to be carried out in Age validation WGNEW was not aware of specific studies to validate absolute ages of sea bass derived from otolith or scale readings. Strong and weak year classes can be followed clearly to over 20 years of age in UK sample data although it is not known to what extent the elevated numbers of sampled fish in immediately adjacent year classes is a true reflection of year-class strength or a consequence of age errors discussed in the previous section. Year-class tracking is less clear in the younger ages 3 5 although this will be affected by gear selectivity and changes in fish behaviour. Sea bass show relatively broad length-at-age distributions, and it has been noted in French data (Laurec et al., 2012 WD to IBP-NEW) that the length-at-age distributions can have unusual patterns including some multiple modes that could indicate age errors. This will result in some smoothing of age data across neighbouring year classes. In the UK data, unusual patterns in length-at-age distributions for some younger ages appear related more to effects of minimum landing size on data from the fishery. Inclusion of age error parameters in Stock Synthesis model CVs for ageing error by age class can be input to Stock Synthesis. Based on the ICES sea bass scale exchange in 2002, the CVs of ~12% can be specified as increasing values per age class to give a standard error of ~1 year per age class. B2.2. Growth parameters Pickett and Pawson (1994) provide plots of growth curves for female and male bass based on samples collected in the 1980s in Areas IV and VII. The samples used by Pickett and Pawson (1994) for growth and maturity analysis were obtained from a range of fishery and other sources. A re-analysis of UK historical age length data including more recent samples was conducted in 2012, using data for the full UK sampling-series from 1985 to 2010 (Armstrong and Walmsley, 2012b). The data are derived from sampling of UK fishery catches around England and Wales as well as from trawls surveys of young bass in the Solent and Thames estuary. More than sea bass have been aged since 1985.

359 ICES WGCSE REPORT The inshore surveys are mainly young sea bass up to 3 5 years of age, whereas the fishery samples include fish up to 28 years of age. All ageing is done from scales, excluding scales considered to be re-grown. On surveys, scales are collected in a length-stratified manner from individual hauls with a view to building age length keys. A similar approach has historically been adopted for catch sampling. This may lead to non-random sampling of individual age groups when the catch numbers are well in excess of numbers sampled from an individual catch. It will also lead to some overestimation of the standard deviation of lengths-atage. All ages for fitting growth curves are referred to a nominal January 1 birthdate, according to month of capture. Parameters of the von Bertalanffy growth curve were fitted in Excel Solver using non-linear minimisation of (obs-exp)^2 for lengths-atage of individual fish, by area and for all data combined. Von Bertalanffy model parameters were as follows: AREA IVBC VIID VIIE VIIAFG ALL AREAS Linf (cm) K t0 (years) Standard deviation of length-at-age distributions increases linearly with age according to: SD (age) = *age B2.3. Maturity Spawning grounds and season Ripe adult bass have been caught by pelagic trawling in the south of Division VIIIa and in the north of Division VIIIb in the Bay of Biscay during January March (Morizur, unpublished data), and planktonic egg surveys (Thompson and Harrop, 1987; Jennings and Pawson, 1992) have shown that bass spawn offshore in the English Channel and eastern Celtic Sea from February to May. Spawning started in the Midwestern Channel when the temperature range associated with bass egg distributions was C, and appeared to spread east through the Channel as the surface water temperature exceeded 9 C. Seasonal patterns of occurrence of advanced maturity stages in UK samples also indicate spawning mainly January to May in ICES Areas IV and VII (Armstrong and Walmsley, 2012c). Spawning and ripe bass are also found in the southern North Sea (information from commercial fisheries and angler reports in Netherlands supplied to IBP-NEW 2012 by F. Quirijns). Previous estimates of maturity at length/age, and data available for re-analysis SGBASS (ICES 2004) reported that around Britain and Ireland, male bass mature at a length of cm, aged 4 7 years, and females at cm, aged 5 8 years, (Kennedy and Fitzmaurice, 1972; Pawson and Pickett, 1996), and data from the southern part of the Bay of Biscay (Lam Hoai, 1970; Stequert, 1972) indicate that male matures at a length of 35 cm (age 4) and females at 42 cm (age 6). Data provided by Masski (1998) from samples taken from VIIe bottom trawlers (41 females) indicate that 40% and

360 1730 ICES WGCSE REPORT % of females were mature at age 6 and 7 respectively, with a very small percentage mature at-age 5. Collection of maturity data are difficult as few adult bass are caught in surveys and bass are typically landed whole and are extremely expensive to purchase. Samples collected by the UK (Cefas) during and 2009 in ICES Areas IV and VII were re-analysed for ICES IBP-NEW 2012 (Armstrong and Walmsley, 2012c). Samples have come from all around the coast of England and Wales, though few fish have been sampled in the Irish Sea (VIIa). Defining a maturity marker for sea bass Sea bass are multiple batch spawners, as indicated by size distributions of oocytes (eggs) in ovaries (Mayer et al., 1990). This means that the ovary will start to mature oocytes through to vitellogenic stages during the months immediately prior to the spawning season. Historical maturity stageing of sea bass by the UK has used the maturity key given in Pawson and Pickett (1996; Table B2-1). In their analyses, they treated stage 2 as mature, and stage 3 as immature. Their reasoning was that stage 3 ovaries (early maturing) were found in smaller bass than later stages (4+) indicating that many of these fish may not proceed to spawning. Sea bass migrate offshore to spawning grounds, and it is likely that early maturing fish could be over-represented, and advanced maturing fish under-represented in inshore catches sampled during the period of spawning migrations. An additional spent stage (VIII) has been occasionally recorded. The identification of a suitable marker to identify maturity has to take into account the probability of finding a fish at any maturity stage in different months, the duration of a stage, and the availability/catchability of fish at that stage of maturity. When the majority of mature sea bass have entered the batch spawning cycle in spring, all stages represented in batch spawning (III to VII) will be evident and should be distinct from immature fish. Hence, the best markers for maturity are the maturity stages representing different stages in the batch spawning cycle, sampled at a time when spawning is taking place (or immediately before), provided fish in all stages are equally catchable. This is the conclusion of recent ICES workshops on maturity staging of gadoids and flatfish, which recommends sampling within a month or so of the beginning and end of the spawning season. Experience with other roundfish and flatfish stocks is that it can be very difficult to distinguish between virgin females and fish that have spawned previously, when sampled in the non-spawning period. The UK data were therefore re-analysed using samples from December to April, treating all fish of maturity stages 3 to 7 as mature. Re-estimation of maturity ogives from UK data Maturity was modelled using a binomial error structure and logit link function, fitted in R to individual observations. The logistic model describing proportion mature by 1 cm length class L was formulated as: Pmat(L) = 1/(1+e -(a+bl) ) defined by the parameters slope b and length intecept a. Parameters were estimated separately for females and males. This can also be expressed as Pmat(L) = 1/(1+e -b(l+c) ) where c = a/b. For Stock Synthesis 3 model inputs, the parameters required are the slope (b: entered as a negative value) and the length inflection, which is the estimated length at 50% maturity (L50%).

361 ICES WGCSE REPORT The 2009 data come from a large sample of sea bass taken in spring from a few trips specifically to revisit bass maturity, but this sample dominates the time-series of sampling which is spread over very many more trips and months than in 2009 and therefore has better coverage. Maturity ogives were therefore fitted including and excluding 2009 data. The inclusion of 2009 data, which was for a relatively restricted length range of fish around 40 cm, has the effect of improving the fit of the model near the top of the ascending limb of the maturity ogive for females (Figure B2-1). However the very high weighting for these lengths compared to the data for lengths <35 cm, results in the model fitting very poorly to the smaller length classes. Excluding the 2009 data allows the length classes <35 cm to carry more weight, and the ogive appears to fit the data for cm sea bass more closely, although the fit for lengths >40 cm is poorer. Addition of the 2009 data effectively shifts the L50% from around 41 cm to 35 cm. In contrast, inclusion or exclusion of the 2009 data has less effect on the model fit for males (Figure B2-1). On balance, it was considered undesirable for a few large hauls n a recent year to have excessive leverage in the model fit, and the model excluding 2009 was considered preferable as a long-term maturity ogive for use in assessments. Table B2-1. Macroscopic characteristics of the maturity stages of the gonads of bass. (Pawson and Pickett, 1996) MATURITY STAGE OVARY TESTIS I Immature Small thread-like ovary, reddish-pink Small, colourless, thread-like; testis not practical to differentiate macroscopically <TL 20 cm II III IV V Recovering spent Developing (early) Developing (late) Gravid (ripe) Ovaries one-third length of ventral cavity, opaque, pink with thickened walls and may have atretic eggs Ovaries up to one-half length of ventral cavity, orange-red, slight granular appearance, thin, translucent walls Ovaries greater than one-third length of ventral cavity, orange-red; eggs clearly visible, but none hyaline Swollen ovaries two-thirds length of ventral cavity, pale yellow-orange; opaque eggs clearly visible with some hyaline VI Running Ovaries very swollen; both opaque and larger hyaline eggs clearly visible beneath thin almost transparent ovary wall, and expressed freely with light pressure Testis one-third length of ventral cavity, often bloodshot with parts dark grey Testes thickness 10 20% of length, dirty white, tinged grey or pink Testes flat-oval in cross-section and thickness >20% of length, half to twothirds of ventral cavity. White colour and milt expressed from vent if pressure applied to abdomen Testes bright white and more roundedoval in cross-section. Only light pressure required to cause milt to flow from vent Testes becoming grey-white and less turgid. Milt extruded spontaneously VII Spent Ovary flaccid but not empty, deep red; very thick ovary wall; dense yellow atretic eggs may be visible Testes flattened and grey, flushed with red or pink, larger than those at stage II or III

362 1732 ICES WGCSE REPORT 2013 Figure B2.1. Logistic maturity ogives (with 95% confidence intervals) fitted to individual maturity records for sea bass during December April. Top plot: excluding 2009 data (top); bottom plot: including 2009 data. Points are proportion mature in the raw data. Dotted line is the number of observations per length class. The parameters of the model Pmat(L) = 1/(1+e -b(l+c) ) are given below: (a) Females (b) Males INTERCEPT (A) Slope (b) c = a/b L25% L50% L75%

363 ICES WGCSE REPORT The logistic model for females and males is: Pmat(L) = 1/(1+e (L ) ) Pmat(L) = 1/(1+e (L ) ) (females) (males) The maturation range for females occurs at ages 4 to 7, and for males at ages 3 6, as shown by the proportion mature at-age in the same samples used for estimation of length-based maturity ogives (Table B2-2). Table B2-2. Raw proportion mature at-age in UK samples from all areas. FEMALES MALES age age age age age age age age age Data on sea bass maturity have also been collected in the Netherlands since Methods and data are described by Quirijns and Bierman (2012). For male fish, too few specimens were measured to estimate maturity. Maturity-at-age and length is plotted in Figure B2-2. Note that only few fish were measured in the lowest age and length groups. At age 4, 50% of the females are mature. This is substantially lower than the age at 50% maturity in the Cefas samples (Table B2-2), and closer to the ogive from Cefas data including the large 2009 sample (Figure B2-1), for which L50 was around 35 cm (~four years old). This may confirm that sea bass could now be maturing earlier than in the 1980s early 2000s, at least for the North Sea. The plot showing maturity-at-length for Netherlands samples is not based on enough measurements to show a reliable maturity ogive.

364 1734 ICES WGCSE REPORT 2013 Figure B2-2. Proportion of mature at-age and length (length in m) for female sea bass sampled in the southern North Sea by the Netherlands during 2005 (thick line). The thin line shows the number of fish measured on which the proportion of maturity is based. B2.4. Larval dispersal, nursery grounds and recruitment Bass larvae resulting from offshore spawning move steadily inshore towards the coast as they grow and, when they reach a specific developmental stage at around mm in length (at days old), it is thought that they respond to an environmental cue and actively swim into estuarine nursery habitats (Jennings and Pawson, 1992). From June onwards, 0-group bass in excess of 15 mm long are found almost exclusively in creeks, estuaries, backwaters, and shallow bays all along the southeast, south, and west coasts of England and Wales, where they remain through their first and second years, after which they migrate to over-wintering areas in deeper water, returning to the larger estuaries in summer. Several studies indicate the existence of similar bass nursery areas in bays and estuaries on the French coasts of the Channel and Bay of Biscay and southern Ireland. During the winter, juvenile bass move into deeper channels or into open water, and return in spring to the larger estuaries and shallow bays on the open coast, where they remain for the next 2 3 years. On the south and west coasts of the UK, juvenile bass emigrate from these nursery areas at around 36 cm TL (age 3 6 years, depending on growth rate), often dispersing well outside the home range, and not necessarily recruiting to their specific parent

365 ICES WGCSE REPORT spawning stock (Pawson et al., 1987; Pickett and Pawson., 2004). It appears that there is substantial mixing of bass at this stage throughout large parts of the populations distribution range. When they reach four or five years of age their movements become more wide-ranging and they eventually adopt the adult feeding/spawning migration patterns (Pawson et al., 1994). B2.5. Natural mortality M There are no direct estimates of natural mortality available for Northeast Atlantic sea bass. Predation up to around age 4 will be in and near estuaries and bays. As with other fish species it is expected that M will be relatively high at the youngest ages, particularly given the slow growth rate in bass. A variety of methods are given in the literature relating natural mortality rate M to life-history parameters such as von Bertalanffy growth parameters k and Linf (asymptotic length), length or age at 50% maturity and apparent longevity particularly in an unexploited or very lightly exploited population. The probability of encountering very old bass is partly a function of the interaction of year-class strength and sampling rates, as well as mortality, however the occurrence of sea bass to almost 30 years of age suggests low rates of mortality. The observed maximum age of 28 years in sea bass samples in the UK was recorded in the early 1980s, following a period of relatively low fishery landings. Age compositions of recreational fishery caught bass in southern Ireland, presented by stakeholders at IBP-NEW 2012, also show ages up to 26 years (Figure B2-3). This stock has been subject to a commercial fishery ban for many years. Age at capture of 1,145 bass by anglers in Ireland Percentage frequency Age, years Figure B2-3. Age composition of bass from samples collected from recreational catches in southern Ireland (data courtesy Ed Fahy, IBP-NEW 2012 meeting). Inferences on sea bass natural mortality based on some life-history models in the literature are given in IBP-NEW 2012 benchmark assessment section. The inferred values of M, with the exception of the Beverton method, are in the range (Armstrong, 2012). Hooking mortality of discarded / returned bass The NMFS in the US has in the past used an average hooking mortality of 9% for striped bass, estimated by Diodati and Richards Striped bass are very similar to European sea bass in terms of morphology, habitats and angling methods. A literature review of hooking mortality for a range of species compiled by the Massachusetts Division of Marine Fisheries included a total of 40 different experiments by 16 different authors where striped bass hooking mortality was estimated over two or

366 1736 ICES WGCSE REPORT 2013 more days (Gary A. Nelson, Massachusetts Division of Marine Fisheries, pers. comm.) The mean hooking mortality rate was 0.19 (standard deviation 0.19). Direct experiments are needed on European sea bass to estimate hooking mortality for conditions and angling methods typical of European fisheries. B.3. Surveys B3.1. UK Solent and Thames pre-recruit surveys The UK has conducted pre-recruit trawl surveys in the Solent and the Thames Estuary since 1981 and 1997 respectively. These surveys all ended in 2009 although the Solent survey was repeated as a one-off survey in autumn 2011 to help provide recruitment indices for the bass benchmark assessment. The location of the surveys and the tow positions are shown in Figure B3-1. Both surveys use a high headline bass trawl, although in the Thames it is deployed as a twin rig and in the Solent as a single rig. Figure B3-1. Location and tow positions for UK(England) Solent and Thames sea bass surveys. The Solent survey has previously been presented to WGNEW as a combined index across ages in each year class. The index was derived by firstly rescaling the annual mean catch rate per age class to the mean for that age in the survey series, then taking the average of the rescaled values for ages 2 4 in each year class from surveys in May July and September (i.e. up to six values represented in the annual combined index). The Thames survey data were worked up in the same way, although using a different age range for the combined index (ages 0 3). WGNEW 2012 provided the survey data in the more conventional tuning-file format, giving the standardised catch rates (arithmetic mean numbers per ten minute tow) by year and age, separately for the two surveys (data in assessment report). These surveys have now been discontinued and will not be updated by future working groups unless new resources are allocated.

367 ICES WGCSE REPORT B3.2. Other 0-gp and 1-gp surveys The UK has undertaken a seinenet survey in the Tamar Estuary, since Additional data are available from the Camel estuary and power stations in the Thames and Severn Estuary. These surveys are used as supporting information and not included in the assessment. Abundance indices for these surveys are given in Table B3.1. The Tamar survey abundance indices need to be updated to include more recent surveys. Seinenet surveys in the UK estuaries Fal and Helford also have data on 0-gp and 1-gp bass. Table B3-1. Abundance indices for 0-gp and 1-gp bass. ( discontinued) ESTUARY SEINE SURVEYS POWER STATION SCREEN TAMAR (0-GROUP) TAMAR (1-GROUP) CAMEL SEVERN THAMES VIIE VIIE VIIF VIIF IVC

368 1738 ICES WGCSE REPORT 2013 ESTUARY SEINE SURVEYS POWER STATION SCREEN TAMAR (0-GROUP) TAMAR (1-GROUP) CAMEL SEVERN THAMES VIIE VIIE VIIF VIIF IVC B3.3. Evhoe survey: France Sea bass are caught in small numbers in the French Evhoe trawl survey, which extends to the shelf edge in Subareas VII and VIII but also extends into coastal areas of the Bay of Biscay and the Celtic Sea where bass may be caught (cf the station map). Less than 10% of the stations have bass catches in most years. A mean of 0.5 sea bass per trawl has been recorded from Abundance indices are calculated as stratified means..figure B3-2. Station positions for French Evhoe bottom-trawl survey. B.4. Commercial lpue B4.1. UK bass logbook scheme The UK bass logbook scheme is described in Section B1.1. Although the survey has severe limitations for estimation of total bass landings for UK vessels, individual