report of the inter-benchmark protocol for sea bass in the ...ices.dk/sites/pub/publication...

ICES IBP BASS REPORT 2014 ICES ADVISORY COMMITTEE

ICES CM 2014/ACOM:45

REF. ACOM, WGCSE

Report of the Inter-Benchmark Protocol for Sea Bass in the Irish Sea, Celtic Sea, English Channel and Southern North Sea (IBP Bass)

1 January–30 April 2014

By correspondence

International Council for the Exploration of the Sea Conseil International pour l’Exploration de la Mer

H. C. Andersens Boulevard 44–46 DK-1553 Copenhagen V Denmark Telephone (+45) 33 38 67 00 Telefax (+45) 33 93 42 15 www.ices.dk [email protected]

Recommended format for purposes of citation:

ICES. 2014. Report of the Inter-Benchmark Protocol for Sea Bass in the Irish Sea, Celtic Sea, English Channel and Southern North Sea (IBP Bass), 1 January–30 April 2014, By correspondence. ICES CM 2014/ACOM:45. 218 pp.

For permission to reproduce material from this publication, please apply to the Gen-eral Secretary.

The document is a report of an Expert Group under the auspices of the International Council for the Exploration of the Sea and does not necessarily represent the views of the Council.

© 2014 International Council for the Exploration of the Sea

ICES IBP Bass REPORT 2014 | i

Contents

1 Executive Summary ....................................................................................................... 3

2 Landings, discards and recreational catches ............................................................. 4

2.1 Description of the fleets ....................................................................................... 4 2.1.1 UK fleets .................................................................................................... 4 2.1.2 French fleets .............................................................................................. 8 2.1.3 Other fleets .............................................................................................. 12

2.2 Landings .............................................................................................................. 13 2.2.1 UK landings data ................................................................................... 13 2.2.2 French landings data ............................................................................. 19 2.2.3 Dutch data ............................................................................................... 23

2.3 Discards ................................................................................................................ 28 2.3.1 UK data ................................................................................................... 28 2.3.2 French data ............................................................................................. 33

2.4 Recreational catches ........................................................................................... 34 2.5 Cpue data ............................................................................................................. 37

2.5.1 Dutch data ............................................................................................... 37

3 Survey data ................................................................................................................... 43

3.1 UK data ................................................................................................................ 43 3.2 French data .......................................................................................................... 47

3.3 Dutch data ........................................................................................................... 52 3.3.1 DFS (Dutch Demersal Fish Survey) ..................................................... 52

4 Stock assessment .......................................................................................................... 58

4.1 Base case WGCSE 2013 assessment-starting point for IBPBass .................... 58

4.2 Development of the SS3 model at IBPBass ...................................................... 62 4.2.1 New data available ................................................................................ 62 4.2.2 Stages in model development .............................................................. 62

4.3 Conclusions from model development process ............................................. 86 4.4 Final assessment model and diagnostics ......................................................... 89

4.5 IBPBass revised stock assessment inputs and model structure/ parameters ......................................................................................................... 100

5 Forecast ........................................................................................................................ 104

5.1 Estimating year-class abundance ................................................................... 104

6 Biological reference points ...................................................................................... 110

6.1 Background........................................................................................................ 110 6.2 Potential MSY BRPs .......................................................................................... 111

6.2.1 Yield-per-recruit reference points ...................................................... 111

ii | ICES IBP Bass REPORT 2014

6.2.2 BMSY trigger point ................................................................................. 114 6.2.3 Limit and precautionary reference points ........................................ 116 6.2.4 Candidate biological reference points ............................................... 117

7 References ................................................................................................................... 119

Annex 1: Stock annex .......................................................................................... 121

Appendix 1 .......................................................................................................................... 167

Appendix 2 .......................................................................................................................... 175

Appendix 3 .......................................................................................................................... 188

Appendix 4 .......................................................................................................................... 189

Annex 2: IBPBass Working documents ........................................................... 191

Annex 3.1: Summary of likelihood values from SS3 runs presented ........... 205

Annex 3.2: Summary of SSB plots from SS3 runs presented .......................... 209

Annex 4: Participants list .................................................................................... 216

ICES IBP Bass REPORT 2014 | 3

1 Executive Summary

The ICES Inter-Benchmark Protocol for Sea Bass in the Irish Sea, Celtic Sea, English Channel and Southern North Sea (IBP Bass) worked by correspondence and WebEx from 1 January to 30 April 2014 to: i) refine the fleet structure, length-age compositions and selectivity models used in the Stock Synthesis assessment and evaluate other po-tential tuning data; ii) evaluate sensitivity of the Stock Synthesis assessment to the scale of geographic aggregation of data, to plausible scenarios for pre-1985 commercial fish-ery landings series, and to other input parameters; iii) develop catch forecasts; iv) de-scribe the choice of preferred method and settings for data analysis and assessment in a concise report, including recommendations on progress to be made in cases where work is not yet finalized; and v) describe the resulting data analysis procedure and assessment methodology in an updated Stock Annex.

IBP Bass was unable to address the issues of scale of geographic aggregation of data or pre-1985 commercial fishery landings, but made a number of improvements to the Stock Synthesis model that had been developed and benchmarked in 2012 for use by the ICES Working Group on the Celtic Seas Ecoregion (WGCSE) in 2013. These in-cluded: increasing the plus-group from 12+ to 16+; combining the UK trawl, net and line fleets into a single fleet with double-normal selectivity; use of age-based rather than length-based selectivity for UK fleets; inclusion of the French Channel Groundfish survey abundance indices and length compositions; exclusion of the Thames and So-lent spring juvenile survey data (retaining the Solent autumn survey); inclusion of a fixed vector of recreational fishing mortality-at-age (averaging 0.09 at ages 5-11) that was consistent with recreational fishery removals of around 1500t in 2012 (the sum of annual removals estimates from recreational fishery surveys in France, England, Neth-erlands and Belgium since 2010); and reducing the natural mortality rate from 0.20 to 0.15. These changes led to an overall increase in biomass and recruitment estimates over the series compared with the final assessment from WGCSE 2013, which was caused by the added recreational mortality, the adoption of double-normal selectivity for UK trawls, nets and lines, and the fitting of age-based selectivity rather than length-based selectivity. The relative trend in recruitment estimates was largely unchanged historically. The average total F for the combined commercial and recreational fishery was around the same as estimated for the commercial fishery in the previous assess-ment, but the new model gave a less steep trend over time. It was not possible to de-termine the shape of the stock–recruit relationship, and a proxy for FMSY was developed as F35%SPR – the F giving SSB per recruit equivalent to 35% of the value in the absence of any fishing. The FMSY proxy is relatively low (F=0.13 at ages 5 - 11) reflecting the slow growth, delayed maturity and the assumption of a relatively low value of M of 0.15 which is consistent with the observed longevity to nearly 30 years. A procedure for short-term forecasts was developed. The Stock Annex was updated later at WGCSE 2014 to include the revised data inputs and model settings although it is given in the present report for completeness.

4 | ICES IBP Bass REPORT 2014

2 Landings, discards and recreational catches

2.1 Description of the fleets

This section provides a general overview of the fleets from the UK, France, and other countries. Length or age composition data are available for several UK and French fleets for estimation of selectivity parameters, while for other UK and French fleets and fleets from other countries no composition data are included in the assessment. The commercial sea bass fisheries in Subareas IV and VII have two distinct compo-nents: an offshore fishery on prespawning and spawning bass during November to April, predominantly by pelagic trawlers from France and the UK, and small-scale fish-eries catching mature fish returning to coastal areas following spawning and immature sea bass large enough to be selected by the fishing gear. 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.

2.1.1 UK fleets

Four general métiers groups have been adopted for modelling annual sea bass landings and age compositions of UK vessels: bottom trawls; midwater pair trawls; drift and gillnets; and lines (longline and handline). A fifth group comprising all other gears does not have an adequate time-series of catch composition and is combined with land-ings from countries other than UK and France for which composition data are also not available or adequate.

Although sea bass can occur as target or bycatch of many vessels, the bulk of the catch is taken by relatively few vessels targeting bass seasonally, according to availability. For example in the UK in 2010, sea bass landings were reported by 1480 vessels (in-cluding 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). The small-scale under-10 m fleet includes many vessels catching sea bass using fixed or driftnets, and handlines or longlines (Table 2.1.1.1). Most of the over-10 m fleet catches of bass are using otter trawls or midwater pair trawls.

UK otter trawls

Otter trawl catches of bass are widespread around the coasts of England and Wales, and predominantly from the Celtic Sea round to the southern North Sea (Figure 2.1.1.1). Bass are taken all year-round, and catches tend to become more concentrated inshore during quarters 2 and 3 where a mixture of juvenile and adult bass are caught. In the Channel and Celtic Sea, most bass are taken using 80 mm mesh codends, and in some areas especially the eastern Channel, many bass under the 36 cm MLS are dis-carded. The size/age distribution of bass taken by this fleet is skewed towards younger bass due to the inshore trawling activities of the under-10 m vessels in particular, and the selectivity is expected to be domed.

UK midwater pair trawls

This fleet including vessels registered in England and Scotland predominantly targets adult bass on or near offshore spawning grounds in the Channel and Celtic Sea from late winter through spring (Figure 2.1.1.1) although some trawling for smaller bass


takes place in the eastern Channel. It is effectively the same métier as the offshore pe-lagic fleet operation from French ports. Pair trawling for bass is currently prohibited inside 6 nautical miles to reduce incidental bycatches of cetaceans. The catches have a larger average size of bass than in the otter trawl fleet. Targeting adult bass with very large nets means that the selectivity is expected to be asymptotic or at least less domed than for inshore trawls and nets.

UK fixed and driftnets, and trammelnets

This is mainly an inshore activity of under-10 m vessels all around the coasts of Eng-land and Wales from the Irish Sea to the Northeast coast (Figure 2.1.1.1). The minimum mesh size for bass is 100 mm, and it is expected that the selectivity of the gear will be domed. Observer data show that discard rates are relatively low.

UK lines

Sea bass are taken commercially by handlines, rod and line and to a lesser extent by longlines, mainly by under-10 m vessels operating inshore from the Celtic Sea around to the southern North Sea (Figure 2.1.1.1). It is a highly selective form of fishing target-ing tide races and other areas where bass are known to congregate. Selectivity will be a function of where fishing takes place in relation to the distribution of different sizes of bass, and factors such as hook size. Outside the spawning season, it is expected that all sizes of bass will be available inshore and vulnerable to capture using hooks, and therefore for the assessment it is assumed that the selectivity of this fleet is asymptotic. There are few data on discarding, however it is expected to be low in commercial fish-eries predominantly targeting areas with marketable bass, and survival of discarded bass caught by handline or rod and line in shallow water may be relatively high.

UK other gears

Catches of other gears are relatively low although there appears to be an increasing use of seinenets in some areas where bass aggregations can be located. There are few data on sizes of fish caught, or rates of discarding.


Table 2.1.1.1. Number of UK vessels recording a bass catch in 2010 per gear type (from Walmsley and Armstrong, 2012). An individual polyvalent vessel may have catches using more than one gear.

UNDER-10 M VESSELS OVER-10 M VESSELS

Gear type Vessels Landings (t) Vessels Landings (t)

Demersal fish otter trawl 139 55 107 92

Nephrops otter trawl 30 <0.5 29 1

Beam trawl 2 <0.5 61 4

Midwater pair trawl 0 0 9 54

Fixed / drift/ trammelnets 721 300 39 17

Handlines/hooks 250 133 4 11

Longlines 37 8 7 <0.5

Scottish seine 0 0 6 40

Other 9 <0.5 7 <0.5


Figure 2.1.1.1. Distribution of average sea bass landings (2005–2011) of the four UK fleet métier groups used in the assessment, and combined “other” métiers.


2.1.2 French fleets

Four general groups of métiers have been adopted for modelling annual sea bass land-ings and age compositions of French vessels: bottom trawls; midwater pair trawls; gill-nets; and longlines. Midwater trawls and otter trawls are the main métier of the area.

FR midwater pair trawls

Few vessels (around 30) target sea bass during winter on spawning grounds in the Channel when sea bass are aggregated. This is an offshore fishery, usually>12 miles. Mesh size used is 100 mm or sometimes more. Figure 2.1.2.1 presents the French monthly landings of sea bass in 2011 of the midwater pair trawl fleet.

Figure 2.1.2.1. Sea bass landings in 2011 from the French midwater pair trawl fleet from the SA-CROIS DPMA information system (Ifremer, 2011).


FR otter trawls

Many vessels in the fisheries with many characteristics (including Great Vertical Open-ing) and a large fishing area (less than 12 miles and >12 miles). Sea bass is often not a target but can be by some vessels. Catches occur all the year. Mesh size used is gener-ally 80 mm. Figure 2.1.2.2 presents the French monthly landings of sea bass in 2011 of the otter trawl fleet.

Figure 2.1.2.2. Sea bass landings in 2011 from the French otter trawl fleet from the SACROIS DPMA information system (Ifremer, 2011).


FR Longlines

This is mainly an inshore fishery (<12 miles) from April to November on feeding grounds. Sea bass is a target species. Figure 2.1.2.3 presents the French monthly land-ings of sea bass in 2011 for lines (handlines and longlines). In the studied area (Channel, North Sea) this is manly longlines which are used (except in the area North of Nor-mandy).

Figure 2.1.2.3. Sea bass landings in 2011 from the French handlines and longlines fleet from the SACROIS DPMA information system (Ifremer, 2011).


FR gillnets

Many vessels can catch sea bass, with many characteristics. In this north area French fleets usually don't target this species unlike in the Bay of Biscay. Figure 2.1.2.4.presents the French monthly sea bass landings in 2011 from gillnets.

Figure 2.1.2.4. Sea bass landings in 2011 from the French gillnets fleet from the SACROIS DPMA information system (Ifremer, 2011).


FR other gears

Catches of other gears are relatively low although there appears to be an increasing use of seinenets in some areas where bass aggregations are located. There are few data on sizes of fish caught, or rates of discarding. Figure 2.1.2.5 presents French monthly land-ings in 2011 of other gears not covered above.

Figure 2.1.2.5. French sea bass landings in 2011 from other gears (from the SACROIS DPMA infor-mation system (Ifremer, 2011).

2.1.3 Other fleets

Landings of sea bass from Subareas IV and VII, other than France and the UK, are mainly from Belgium and the Netherlands. In Ireland there is a ban on commercial fishing for sea bass.

Belgium

Landings into Belgium were not available by métier to IBPBass.


Netherlands

Catches of bass by vessels from the Netherlands are mainly in the southern North Sea and English Channel. The main gears used are beam trawls, flyshoot, seines, lines and gillnets. The total landings by the beam trawl fleet have decreased, whereas landings from lines and flyshoot have increased over the period 2002 to 2012. The main reason for this is that the total effort of beam trawlers has decreased, whereas the effort of the other fleets have increased. From the end of winter sea bass migrate to the north and move southwards again in autumn, resulting in relatively more sea bass being caught in the south in quarter 3.Almost all landings in quarter 1 are from beam trawlers and flyshoots, whereas catches by lines are mainly landed in quarters 2, 3 and 4. Most of the catches from gillnets and seines come from the Dutch coastal area in Q3.

2.2 Landings

The landings by gear groups used for development of the Stock Synthesis model (Table 2.2.1) are for the years 1985 to 2012 and are as given in the ICES WGCSE 2013 report (ICES, 2013a).

Table 2.2.1. Fishery landings series (tonnes) used in the sea bass assessment. (French landings from 2000 on are from official reported landings processed annually to reallocate trips to the correct fishing ground; pre-2000 total all-gears French landings are the official figures for IVb,cVIIa,d,e,f,g,h multiplied by the average adjustment for 2000 onwards landings, and disaggre-gated by métier groups using data from Y. Morizur (Ifremer) supplied to ICES SGBASS 2003). All data are as used by ICES WGCSE 2013.

2.2.1 UK landings data

2.2.1.1 Time-series of landings

The proportion of the total officially reported landings attributed to each gear group has varied over time as the total landings have increased (Figure 2.2.1.1). The midwater trawl fishery took place mainly over a decade from the mid-1990s.

YearBottom trawls

Pair trawl

Drift and gil l nets Lines

All other gears

Total all gears

Bottom trawl

Pelagic trawl Nets

Hand- l ines

Long- l ines

all other gears

Total all gears Belgium Denmark

Nether lands

Channel Is.

TOTAL all countries

1985 15 1 30 15 1 61 447 393 6 0 24 0 870 0 0 0 18 9491986 21 2 61 34 2 120 538 588 42 0 13 0 1180 0 0 0 15 13151987 45 0 55 18 7 126 194 1514 41 28 63 0 1840 0 0 0 14 19791988 70 8 64 30 1 172 315 479 100 70 64 0 1028 0 18 8 12 12381989 91 9 61 29 1 192 468 214 130 31 74 0 917 0 2 2 48 11611990 75 23 47 14 0 159 486 177 104 31 51 0 849 0 0 0 25 10331991 49 14 113 61 1 238 478 371 44 33 44 0 971 0 0 0 16 12251992 51 8 64 24 1 148 447 423 48 36 46 0 1001 0 0 0 36 11851993 95 1 65 62 2 226 602 261 44 22 50 0 979 0 1 0 45 12511994 140 0 229 155 11 535 546 111 49 24 56 0 786 0 0 0 49 13701995 179 1 262 169 40 651 616 291 59 26 64 0 1057 0 1 0 69 17771996 144 87 186 129 17 563 875 1417 30 24 49 0 2395 0 1 8 56 30231997 158 71 195 120 15 560 692 1138 54 22 78 0 1984 0 1 1 74 26201998 157 85 108 121 14 486 677 1015 71 2 9 0 1773 0 2 48 79 23881999 150 220 136 148 23 677 812 958 51 0 21 0 1843 0 1 32 108 26602000 156 52 103 53 32 396 692 681 108 201 104 20 1806 0 5 60 130 23972001 161 95 121 58 3 439 713 659 110 211 163 26 1883 0 2 77 80 24822002 187 109 233 75 23 627 911 415 128 204 145 22 1824 0 1 96 73 26222003 230 127 146 65 2 569 1087 773 152 277 161 23 2471 154 1 163 84 34422004 202 131 206 72 2 613 1236 820 150 208 173 17 2604 159 1 191 159 37272005 164 78 172 59 1 474 1239 1319 148 238 201 17 3161 206 1 327 220 43902006 201 33 198 107 3 542 1110 1420 140 296 258 35 3259 211 2 308 193 45152007 202 64 239 167 4 675 1187 841 158 323 237 24 2770 178 1 376 160 41602008 231 20 322 162 7 741 1145 1012 128 263 162 41 2750 188 0 380 143 42022009 185 11 312 146 14 669 1052 1098 94 173 78 154 2649 173 0 395 103 39892010 155 42 299 180 21 697 819 1828 160 182 96 152 3236 215 4 399 144 46962011 141 98 327 143 23 732 790 1143 129 242 116 106 2526 152 2 395 0 38062012 163 49 411 186 72 881 825 1143 142 211 82 207 2610 149 3 372 46 4060

UK fleets French fleets Other countries


Figure 2.2.1.1. Composition of UK landings of sea bass in Subareas IV and VII by métier groupings used in the assessment.

2.2.1.2 Quality of UK landings data

The official reported landings of sea bass in the UK are known to underestimate the true total landings, particularly for small-scale inshore fisheries where there has been no requirement to submit EC logbooks. Prior to the introduction of Buyers and Sellers regulations in 2006 requiring sales documentation, local fishery inspectors estimated landings of the under-10 m fleet using whatever information they had available from auctions, and frequently entered aggregated estimates for multiple vessels into the fishery landings database. Unfortunately the Buyers and Sellers regulations do not cover all landings. Where a single landing has less than 25 kg of fish, it is not mandatory to have a record of the landing from sales slips or other documents. This is ostensibly to reduce the administrative burden for a skipper disposing of small quantities for per-sonal use. However, for small-scale fisheries where there are very large numbers of small vessels often catching small quantities, the cumulative catch of unrecorded small landings can be relatively high. This is likely to be an issue over the full time-series.

Due to the known inaccuracies in reported landings of such vessels, Cefas (UK) imple-mented an independent logbook scheme and port survey in England and Wales in 1985 to estimate mean cpue (annual landings per vessel, based on a logbook kept by selected skippers) and total number of vessels catching sea bass (from an annual port survey covering different stretches of coastline in successive years). Total bass landings were estimated as the product of cpue and vessel numbers. The scheme was stratified by area, gear and vessel characteristics. Selected vessels from the strata kept logbooks for


periods ranging from 1 to 25 years, and comprised what could be described as a “ref-erence” fleet as opposed to a randomised selection of vessels each year. The scheme is described by Armstrong and Walmsley (2012), who identified some issues with the survey related to overstratification and extensive imputation needed due to small and declining numbers of logbooks, and the non-random nature of vessel selection for log-books. Sufficient logbooks were available only for inshore vessels using fixed/driftnets or lines. The scheme was terminated in 2007 and 2008, and reinstated for a further two years (2009 and 2010) before being terminated again. The scheme has now been sus-pended permanently.

Despite the potential biases, the survey results for commercial vessels confirm that the historical official reported landings of sea bass are likely to be underestimates (Figure 2.2.1.2.1). For fixed/driftnets, the landings including the Cefas logbook estimates for under-10 m vessels results in a landings series that is on average around three times higher than the official statistics. For lines, the ratio fluctuates around 3.0 for a large part of the series but was larger from 2000–2005.

Figure 2.2.1.2.1. Top: estimates of sea bass landings for under-10 m UK netters and liners based on the Cefas logbook and port survey scheme. Bottom: ratio of landings of these gears including the Cefas logbook estimates for <10 m commercial vessels, and the total official reported landings of all UK vessels using these gears.

The historical differences between official landings statistics and Cefas logbook land-ings estimates for sea bass are hard to interpret because there are no diagnostics pro-vided on how representative the Cefas logbooks are of the overall fleets in each survey stratum, and because extensive imputations have been needed due to over-stratifica-tion. Precision is likely to be low due to the small number of logbooks, although this


cannot be quantified. Although the logbook estimates are subject to bias and low pre-cision, a bias factor of ~3.0 (Figure 2.2.1.2.1) could be applied to the input landings in Stock Synthesis to evaluate its effect on the assessment trends and advice.

2.2.1.3 Length/age structure

Age compositions for UK commercial fishery landings of sea bass are derived from biennial (January–June and July–December) age–length keys (ALK) constructed for four areas: IVb,c; VIId, VIIe,h and VIIa,f,g. These are applied to fleet-raised landings length frequencies for each of four gear groups (bottom trawls; midwater trawls; fixed/driftnets and lines) in each area. Further details are given in the ICES IBP-NEW (ICES, 2012a) and WGCSE (ICES, 2013a) reports and in the stock annex along with ta-bles giving numbers of trips sampled for length and age and numbers of fish measured and aged.

A recommendation of WGCSE 2013 was to expand the UK age frequencies to the full recorded age range and to re-evaluate the plus-group definition (previously at 12+). Sea bass have been recorded to almost 30 years of age, and it was thought that having more true ages in the Stock Synthesis input data could allow better estimates of early recruit deviations. The necessary extractions were done for IBPBass, and the data were examined in detail by Armstrong and Readdy (Annex 2 - Working Document_01). Bub-ble plots and catch curves showed that coherent information on year classes was pre-sent well beyond the last true age (11) previously adopted. The exploratory SS3 runs show that the different choices of plus-group (12+; 16+; 18+; 20+) have relatively little impact on the results, other than (as hoped) a slightly better estimation of early recruit deviations. Expanding the age compositions may help fit domed selection curves for fleets where this is appropriate, but risks an increasing number of zero catch entries for older ages as recent weak year classes feed into future catches and become depleted. A plus-gp of 16+ was recommended for further model development and agreed by IBPBass. The estimates for each gear group are given in Tables 2.2.1.3.1–2.2.1.3.4. Sam-pling of midwater trawls prior to 1996, and in 1997, was considered too poor to develop age compositions. All datasets show a very strong 1989 year class and very weak 1985–1987 year classes.

For a number of year–gear combinations for French fleets since 2009, length composi-tions of retained fish were supplied to IBPBass. These were for input to Stock Synthesis runs with fleet-data disaggregated into the same four groupings used for UK fleets. IBPBass considered that there were too few years of data to allow a separate estimation of selectivity parameters for the four French fleets, and it may be more appropriate to combine UK and French landings and composition data for these four groups. How-ever, the UK data are included in Stock Synthesis as age compositions, and French data were only supplied as length compositions. To allow the fleets to be combined, UK ALKs for IVbc, VIId, VIIe,h and VIIa,f,g, were applied to the French length composi-tions. The ALKs are available on the IBP-Bass SharePoint site (folder: data/landings/UK landings).


Table 2.2.1.3.1. Estimated numbers-at-age in UK bottom-trawl landings of sea bass from IVb,c and VIIa,d,e,f,g,h.

Table 2.2.1.3.2. Estimated numbers-at-age in UK midwater pair trawl landings of sea bass from IVb,c and VIIa,d,e,f,g,h.

(a) Bottom trawlYear 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16+1985 0 337 1137 949 3226 455 1229 3733 950 461 518 133 278 25 2711986 0 244 3504 1549 572 1866 430 1325 4914 672 577 619 239 299 4261987 0 357 15559 24879 7749 835 1366 190 408 2417 1187 653 294 397 14231988 0 1949 20375 44317 16538 4686 511 1162 357 240 4423 1335 496 140 4261989 35205 4714 282 3349 15900 13205 4847 2083 2321 1131 2133 7106 1139 1131 31041990 0 1047 1564 722 4889 19561 13248 3650 1096 726 502 441 3556 353 15511991 0 1390 19502 744 412 1653 5054 3738 1399 96 1212 743 651 1709 8761992 0 6005 24306 15529 1387 560 946 2687 1508 383 228 193 29 233 19061993 0 631 60748 47526 14394 575 440 1371 3323 2177 723 335 351 181 11561994 0 920 16392 166192 15152 5328 468 60 572 2162 1096 945 205 311 5871995 0 1469 11036 26876 170029 13062 4679 350 168 151 1389 1055 369 88 10641996 0 1184 9448 9977 32547 90577 5666 2429 123 208 246 1089 651 196 6431997 0 637 3882 33109 23308 21338 68531 4077 1586 192 256 158 840 1150 11291998 0 351 23377 30135 59712 19962 11422 29310 1470 522 71 23 277 486 5951999 247 97 50631 80956 21144 18712 5704 3855 11645 766 476 46 198 36 3612000 0 6227 1868 94742 46856 8763 8924 4145 4733 7811 816 180 6 1 4982001 306 5909 56085 27594 66949 13492 4391 5200 3642 5509 2112 399 578 47 5722002 18 4123 20123 118151 9450 42831 10313 4786 4719 1545 2256 6583 153 154 6042003 0 4175 44667 36380 86804 6201 30867 12486 3768 2669 2027 1375 3436 814 7192004 0 1207 15586 116336 48334 56381 1369 6078 2818 1314 662 629 177 463 1562005 0 8788 80929 40490 45629 18450 18304 2172 5194 864 1100 158 59 640 7392006 0 12111 75122 77812 30523 35432 13327 18831 1604 1608 881 655 0 0 6482007 0 369 24466 108745 59765 22244 17333 6531 4130 1224 1851 147 16 8 1102008 0 3917 51035 133436 72814 28101 10391 9063 3204 3297 186 255 475 105 2102009 0 1099 33713 76895 67771 24893 7124 8109 2194 1636 1686 1963 722 0 02010 0 149 17245 58897 52931 37918 10982 2909 1333 1375 443 391 258 116 112011 0 522 17337 61301 39159 26496 14262 7399 2203 1970 1147 418 304 115 2352012 0 672 5116 81107 102396 18353 7651 5290 2431 1557 550 159 70 46 43

(b) Midwater pair trawlYear 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16+

1996 0 0 289 796 3892 71665 5583 1648 21 334 154 622 485 199 5601998 0 0 264 6405 12691 9161 8714 26925 2696 370 100 57 128 957 6141999 0 0 2988 18438 15167 27342 13892 18263 43646 4481 1695 324 387 308 27622000 0 15 60 2475 7585 3269 4496 1459 2829 7075 633 174 39 96 4212001 0 0 176 884 19449 19953 6925 5181 3072 2797 9505 843 625 121 2582002 0 3 37 2349 1558 23776 9568 6215 5901 444 5591 9096 0 0 5242003 0 0 2503 9885 36543 7420 37748 9582 2943 3029 576 157 3764 72 2622004 0 7 1310 13054 15006 50232 3340 21608 8363 369 1876 1192 95 1011 202005 0 0 130 2404 17514 16457 19899 2180 5924 0 2093 113 0 45 6862006 0 0 105 263 282 1892 660 1432 118 82 52 0 0 0 262007 0 0 659 4305 12037 9213 11685 4780 3249 1079 1380 21 64 32 2072008 0 52 513 1775 3779 2060 1627 1794 870 1106 35 211 565 0 472009 0 0 101 712 2439 2911 945 880 189 334 194 12 190 0 12010 0 9 36 1741 5545 8261 6677 4755 403 3786 152 294 313 551 512011 0 0 255 4397 10231 13639 15908 13641 4424 4232 2773 1688 1003 264 4242012 0 0 391 4456 10762 10003 8746 5782 2738 1133 289 433 143 127 226


Table 2.2.1.3.3. Estimated numbers-at-age in UK fixed and driftnet landings of sea bass from IVb,c and VIIa,d,e,f,g,h.

(c) Fixed / drift netsYear 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16+

1985 39 7382 18413 2615 1875 215 820 2407 566 1318 464 122 291 182 3631986 0 4197 15123 10292 1928 5015 546 2881 6301 776 1168 1737 293 605 18051987 0 43 20164 30309 6291 1046 1250 392 438 3043 840 795 250 289 7511988 0 0 678 10436 19191 6702 1128 1242 519 195 2193 1050 882 733 26171989 846 735 516 2523 18808 16067 6006 2018 1194 1155 1323 2101 336 183 4091990 0 1066 860 1548 3606 9555 6882 1541 1124 741 369 670 1540 68 13321991 3297 12578 20667 1287 1207 5251 15095 12644 5225 551 1229 1517 286 4542 24991992 0 10422 29796 17866 417 213 1106 2464 3446 854 207 212 111 227 17051993 0 232 32451 25641 9567 651 163 472 1153 1217 459 163 373 223 11071994 2 285 12922 176117 25082 12277 773 213 1687 4150 2878 773 413 300 26891995 0 3478 22316 47981 180152 6314 2833 383 250 840 2235 2322 1500 3315 21071996 145 7977 25959 14290 39275 104809 2554 911 59 113 196 735 895 255 9641997 0 2209 4284 27827 20255 20214 77283 3896 2969 334 388 114 1698 585 10341998 0 764 20593 15179 24737 7687 8101 18406 1336 315 18 105 127 263 5261999 11 0 29417 52990 23456 18153 5183 4129 9056 453 220 3 81 37 3912000 0 2906 757 60052 28238 6480 6492 2122 1714 2577 175 33 4 43 1862001 161 3286 20439 4323 41916 10945 4216 5141 3243 4375 6745 498 255 46 4102002 487 10454 33273 155106 9245 32629 8064 3457 4055 1333 1938 4710 443 93 982003 0 1847 23694 26865 69781 2934 12995 4282 1298 1214 936 674 1838 168 682004 0 3509 27534 97330 28102 44945 1626 5603 1700 772 340 403 178 567 2172005 0 4708 24222 24735 57927 16893 34204 3729 6640 689 384 1149 598 17 632006 0 6605 48039 57081 22822 41163 9618 17053 1384 2899 1445 182 364 341 13712007 0 560 11218 72349 43624 24680 19076 7143 10213 9805 2752 1408 0 231 3232008 0 6035 64178 169746 74269 23102 10878 9145 5448 3072 676 586 589 202 6402009 0 1445 37102 89160 100353 38821 11348 6468 6371 6423 1888 286 594 0 2252010 0 172 54111 87458 59240 36789 16550 8233 5252 4310 2064 1628 1928 386 14962011 0 307 10895 59507 50386 41929 37223 22801 7622 8642 4832 2340 1789 528 13822012 0 938 9290 89836 121080 43500 31849 25975 12592 7708 3462 3601 2395 483 867


Table 2.2.1.3.4. Estimated numbers-at-age in UK lines landings of sea bass from IVb,c and VIIa,d,e,f,g,h.

2.2.2 French landings data

Landings for the years 1985 to 2012 are as given in the ICES WGCSE 2013 report (ICES, 2013a).


The proportion of the total officially reported landings attributed to each gear group has varied over time as the total landings have increased (Figure 2.2.2.1). Midwater trawlers and bottom trawlers are the main fisheries of the area.

(d) LinesYear 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16+

1985 0 4136 5179 1407 2739 445 415 1977 441 219 292 133 371 83 2471986 0 443 9260 3720 1002 2341 492 999 5091 974 751 838 430 435 9691987 0 138 917 3559 1962 517 433 339 503 1560 617 213 136 168 9701988 0 53 1359 8675 5406 2249 550 979 152 538 1920 823 520 450 9191989 47 0 1142 7803 18240 1535 805 251 503 24 297 1528 184 138 3801990 0 98 32 173 1580 2893 1915 695 187 259 188 277 720 30 4441991 0 193 6497 410 861 2175 7485 5927 2361 265 940 1076 462 3069 14991992 0 1479 6110 5553 217 212 487 1289 1581 522 111 124 118 121 10651993 20 253 9450 10471 6261 398 257 1008 3810 3221 1134 462 729 827 32811994 0 91 4245 99667 16980 9393 837 159 1585 4644 4919 1112 723 372 32121995 0 792 5453 16556 81447 8648 5547 678 622 1145 4917 3781 1682 911 41011996 0 179 6469 8090 17433 56415 3831 3110 170 473 528 2231 1106 787 18501997 330 1181 2227 12662 7395 9522 40539 4023 1398 401 311 317 1924 1473 20991998 0 12840 14548 10186 20898 7663 7048 24119 2554 779 113 201 763 1182 14211999 88 43 17564 26749 11963 14121 4937 5051 22706 2058 1457 211 528 475 26792000 0 256 179 14000 9220 2541 3246 1542 1822 5432 419 166 16 46 4212001 21 215 3720 1006 14353 6047 2123 2631 1573 2158 4502 567 324 19 2142002 17 491 1791 11133 2385 16881 6499 2555 4187 1151 1503 3562 227 101 1742003 0 32 3477 5492 25393 1914 9806 2582 1221 770 907 949 1411 89 1242004 0 152 1447 9796 8285 16422 1535 5117 3515 936 1014 1163 373 1717 372005 0 446 1975 4418 13781 4877 6813 689 3830 1267 115 1072 46 19 1312006 0 26 10642 23236 8097 11518 2846 6085 901 1609 1063 301 227 589 19582007 0 18 5679 27108 24376 12464 15320 7298 9408 4437 2681 4397 71 296 4162008 0 171 4313 23524 36230 18567 10563 10821 4700 4272 1181 889 2955 171 8532009 0 272 7306 17308 29503 22587 10565 8200 4413 4731 4191 713 952 0 02010 0 746 6436 41672 50140 36359 14497 4711 2192 2664 646 478 329 75 2542011 0 12 4178 11123 21603 16322 15828 11181 5419 5426 2409 2071 1993 569 4752012 0 80 1496 16212 25243 20749 20254 15810 9005 6820 2643 2360 1906 527 724


Figure 2.2.2.1. Composition of FR landings of sea bass in Subareas IV and VII by métier groupings used in the assessment.

2.2.2.2 Quality of French landings data

From 1999 onwards Ifremer has provided revised French landings from a separate analysis of logbook and auction data and VMS which allocates landings correctly by fishing ground (SACROIS methodology used at present as Official Landings). Quality of data from 2000 can be considered as high. From 1985 to 2000, quality of data is poorer. To generate a consistent series of French landings from 1985 onwards for the Subarea IV and VII assessment, IBPNew 2012 (ICES, 2012a) adjusted pre-1999 official FishStat landings by the average of the Ifremer correction factors by area from 1999–2010. To split landings by métier for IBPBass 2014, the ratio of landings among métiers was used to convert total French landings estimated by ICES IBPNEW 2012 to French landings by métier. This was the best way to get consistent time-series of French land-ings. Moreover some uncertainties remain from 1985 to 2000 because the French land-ings of sea bass are probably underestimated particularly for small-scale inshore fisheries.


Length compositions of French sea bass commercial landings are constructed for four areas from 2009 to 2012 (IVb,c; VIId, VIIe,h and VIIa,f,g), for four métiers : GTR_DEF; LLS_DEF; OTB_DEF and OTM_DEF (Tables 2.2.2.3.1 and 2.2.2.3.2). It was not possible to disaggregate total length compositions, from 2000 to 2008. Only numbers-at-length for all gears are available from 2000 to 2009 (Table 2.2.2.3.3.).

0

500

1000

1500

2000

2500

3000

3500

1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

2005

2007

2009

2011Re

port

ed la

ndin

gs (t

)

Annual landings of FR fleets

FRDanishseine

FRPurseSeine

FRLonglines

FRHandlines

FRNets

FRPelagictrawl

FRBottomTrawl

Frother

0%10%20%30%40%50%60%70%80%90%

100%

1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

2005

2007

2009

2011Re

port

ed la

ndin

gs (t

)

Composition of landings of FR fleets

FRDanishseine

FRPurseSeine

FRLonglines

FRHandlines

FRNets

FRPelagictrawl

FRBottomTrawl

Frother


For a number of year–gear combinations for French fleets since 2009, length composi-tions of retained fish were supplied to IBPBass. These were for input to Stock Synthesis runs with fleet-data disaggregated into the same four groupings used for UK fleets. IBPBass considered that there were too few years of data to allow a separate estimation of selectivity parameters for the four French fleets, and it may be more appropriate to combine UK and French landings and composition data for these four groups. How-ever, the UK data are included in Stock Synthesis as age compositions, and French data were only supplied as length compositions. To allow the fleets to be combined, UK ALKs from Divisions IVbc, VIId, VIIe,h and VIIa,f,g, were applied to the French length compositions. The ALKs are documented in Appendix 5.

Table 2.2.2.3.1.Estimated numbers-at-length in French gillnets (GTR_DEF) and French longlines (LLS_DEF) landings of sea bass from Divisions IVb,c and VIIa,d,e,f,g,h.

mm 2010 2012 2009 2010 2011 2012340 2790 798 320 614 0 0360 8730 1912 2293 3361 1163 1775380 11729 1431 4452 5525 6850 7245400 7627 789 5129 7126 7209 6359420 4245 1241 4254 16830 7164 5944440 2775 1844 2016 11250 13009 4254460 1663 1984 4029 11885 9862 9757480 711 893 5383 8066 12905 5339500 863 1364 2663 6671 9373 5548520 570 444 3943 3770 6406 8014540 835 1339 5406 9299 8489 5311560 334 2392 3290 4005 6841 5711580 0 2060 3679 4638 7675 3037600 0 2194 4648 4638 9067 5332620 0 2355 3429 3789 5868 2938640 79 1438 2888 2044 4850 4945660 0 1045 3868 2200 3772 3578680 0 1826 2093 0 1977 2327700 79 1040 1787 1700 1339 1223720 0 769 2047 602 596 1536740 0 442 1585 0 620 0760 0 93 507 0 0 0780 0 93 1066 0 0 0800 0 0 507 549 0 305

GTR_DEF LLS_DEF


Table 2.2.2.3.2. Estimated numbers-at-length in French bottom trawlers (OTB_DEF) and French midwater trawlers (OTM_DEF) landings of sea bass from Divisions IVb,c and VIIa,d,e,f,g,h.

mm 2009 2010 2011 2012 2010 2011 2012300 0 0 18325 0 0 0 0320 0 0 47121 0 0 0 0340 59635 3189 146948 6417 0 0 3095360 71769 7973 135391 67580 14152 9215 7398380 164889 6736 122733 142915 133927 20664 22603400 119334 11777 69455 98398 215482 74831 36861420 92468 11652 36441 86860 385709 158226 44924440 68746 14134 51158 52909 234024 148981 70524460 60264 14095 31427 81004 133425 161433 84057480 42989 10401 35598 48315 171834 141641 97325500 39911 11534 26699 45317 90103 97454 77995520 23952 9977 19451 43345 117640 58884 82843540 22704 11849 15866 19918 148162 50947 58996560 16550 8468 12756 17217 13058 39390 42773580 19414 4570 13048 12679 3574 11061 29576600 14017 3856 10290 12295 2648 4578 22218620 13521 5196 3873 8274 2184 3794 16551640 10643 3478 2680 5748 2522 1062 9029660 7519 981 1737 3381 610 1701 8116680 3992 1381 1157 4210 1824 2498 4597700 2858 1240 721 2707 1265 2255 2800720 3007 559 478 1350 2007 2133 1216740 725 124 0 906 1618 2197 845760 690 62 70 0 1125 1137 398780 0 0 94 316 434 843 121800 1450 0 0 0 88 349 117820 0 0 0 0 346 249 19840 0 0 0 0 0 0 21860 0 595 0 0 0 0 0880 0 0 0 0 88 0 0

OTB_DEF OTM_DEF


Table 2.2.2.3.3. Estimated numbers-at-length of all gears of French landings of sea bass from Divi-sions IVb,c and VIIa,d,e,f,g,h, 2000–2012.

2.2.3 Dutch data

Landings and effort data from the commercial fleet are available from the EU logbooks; market category composition of landings is available from the auction data (sale slips); and size and age data are available through market sampling. These data have been summarized in van der Hammen et al. (2013).

EU logbook data

Official EU logbook data of the entire Dutch fleet are maintained by the NVWA (for-merly known as the General Inspection Service, AID). IMARES has access to these log-books and stores the data in a database (VISSTAT). EU logbook data contain information on:

• landings (kg): by vessel, trip, ICES statistical rectangle and species; • effort (days absent from port): by vessel, trip and ICES statistical rectangle; • vessel information: length, engine power and gear used.

Logbook data are available from the entire Dutch fishing fleet and from foreign vessels landing their catches in the Netherlands.

Length 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 201220 0 0 0 0 0 0 0 0 0 0 717 0 022 0 0 0 0 0 0 0 0 0 0 0 0 024 0 0 0 0 0 0 0 0 0 0 0 0 026 0 0 0 0 0 0 0 0 0 0 0 0 028 0 0 0 3455 0 0 0 0 0 292 0 0 121930 0 0 1015 13054 14 0 15689 0 0 473 0 0 032 0 0 0 58717 13057 9903 32459 181 8250 2239 9811 1976 158334 9931 17962 12469 105655 78811 29872 179130 4715 28986 10714 28290 13885 651836 34932 19809 38249 125326 127801 97890 285704 39335 229758 124925 169311 57121 8576038 85866 68920 46427 180475 124051 128022 217657 102714 263071 211881 177571 87842 17251040 126730 76594 62503 119495 227214 231750 178250 146272 266408 225545 182105 128838 14027342 102836 98008 82461 145456 282390 266905 196868 145122 237160 193030 283064 187586 14789544 80478 109595 91064 104545 243107 344681 289998 164011 270810 222613 251956 201447 16233346 93344 106857 86723 130023 188494 270532 285451 130859 228996 238849 230227 199487 18075248 80934 77694 62163 115806 126685 239265 263272 100043 142650 155222 188149 194697 15849050 55399 57055 55905 91915 72581 169478 200874 99210 112385 159658 186310 145447 13075952 52948 51658 46180 93878 82331 115269 119836 75929 74336 114530 109212 124239 10721454 42094 36737 35998 48742 50633 62106 99509 74405 66260 84649 120550 92526 9063856 26460 35839 26001 60839 60284 67741 99674 55147 48853 96257 71590 72471 7893458 27357 22762 19019 31614 31334 61132 54522 46087 39689 51578 62211 46869 5486960 23581 25834 14210 33688 19126 43591 45908 28056 29840 36547 31544 31690 3538762 14295 18773 11129 30691 23996 35774 23763 23057 28335 57472 19076 19998 3308564 18044 13532 16771 18823 14799 25788 20607 18091 14420 24016 62005 17624 1771466 10773 11068 11011 13230 10650 12456 14969 8715 12694 21415 26388 14720 1517068 9903 9120 5447 7960 8569 13360 13976 8793 9039 27466 9340 7906 937470 5709 11771 4795 5374 4880 8908 9653 4835 6821 20198 8541 6114 811472 5721 5733 4559 5617 2974 8053 4521 2707 4714 12083 29128 2082 414774 2345 5345 1825 3275 2675 9811 3424 1962 1623 7551 1884 1163 231376 2595 2782 1260 1356 2567 5020 2883 1010 1257 979 2114 1096 154078 2102 1691 357 297 548 2378 731 399 534 1765 182 476 113480 888 583 155 783 425 1365 201 158 261 264 5525 148 28282 1021 296 109 112 149 107 261 37 8 1004 6097 104 45184 548 204 0 148 295 0 30 59 0 0 863 0 2986 123 0 0 0 0 0 0 0 0 0 0 0 2788 0 61 0 0 149 0 0 0 0 0 1207 0 090 0 0 0 0 0 0 0 0 0 0 0 0 0


Auction data: landings by market category

Auction data cover both the total Dutch fishing fleet and foreign vessels landing their catches on Dutch auctions. These data are also stored in VISSTAT and contain infor-mation on:

• landings by market category (kg): by vessel, trip (landing date) and species.


Dutch sea bass landings have increased substantially from ~50 tonnes a year, to 300–400 tonnes a year since 2005. Most catches are from ICES Divisions IVc and VIId (Figure 2.2.3.1.1). Sea bass is landed in all quarters, but is mostly landed in quarters 1 and 2 (Figure 2.2.3.1.2).

Most sea bass is caught by beam trawl, flyshoot, lines, seines and gillnets. The total landings by the beam trawl fleet have decreased, whereas landings from lines and flyshoot have increased during the timespan of the series (Figure 2.2.3.1.3 and Table 2.2.3.1.1). The main reason for this is that the total effort of the beam trawl has de-creased, whereas the effort of the other fleets has increased.

Figure 2.2.31.1. Dutch sea bass landings by ICES area and year in IVbc, VIId, VIIa–h.

year

wei

ght (

tonn

es)

100

200

300

400

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

IVcVIIdVIIaghe, IVb


Figure 2.2.3.1.2. Dutch sea bass landings by quarter.

Figure 2.2.3.1.3. Annual Dutch landings (tonnes) per gear and year.

year

wei

ght (

tonn

es)

050

100

150

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

3

050

100

150

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

4

050

100

150

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

1

050

100

150

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

2

Dutch Seabass landings by gear

land

ings

(ton

nes)

100

200

300

400

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

beamtrawlflyshootgillnets-seineslinesother


Table 2.2.3.1.1. Sea bass Dutch landings (in tonnes) for different gear categories in ICES Subarea IV and Division VIId.

GEAR

area YEAR beam trawl flyshoot gillnets-seines lines other

IV 2000 33 0 3 0 5

IV 2001 34 0 12 3 12

IV 2002 60 0 8 6 7

IV 2003 105 0 15 7 22

IV 2004 115 0 17 8 23

IV 2005 161 3 21 52 27

IV 2006 161 4 29 62 27

IV 2007 160 1 31 88 40

IV 2008 137 8 31 96 30

IV 2009 176 19 34 87 30

IV 2010 143 14 37 119 14

IV 2011 128 6 39 149 11

IV 2012 80 5 29 127 7

IV 2013 64 22 19 112 5

VIId 2000 0 1 0 0 9

VIId 2001 0 2 0 0 8

VIId 2002 0 2 0 0 15

VIId 2003 0 3 0 0 11

VIId 2004 0 4 0 0 12

VIId 2005 0 8 0 0 36

VIId 2006 0 10 0 0 5

VIId 2007 0 30 0 0 22

VIId 2008 0 55 0 1 15

VIId 2009 0 36 0 1 7

VIId 2010 0 54 0 0 6

VIId 2011 0 49 0 0 7

VIId 2012 0 100 0 0 5

VIId 2013 0 103 0 0 1

Almost all sea bass landed in Dutch harbours is caught in the southern North Sea and in the English Channel (Figure 2.2.3.1.4). From the end of winter sea bass migrates to the north and migrates to the south again in autumn. This is reflected in the quarterly catches, when relatively more sea bass is caught in the south in quarter 3.

Almost all landings in quarter 1 are from beam trawlers and flyshoots, whereas catches by lines are mainly landed in quarters 2, 3 and 4. Most of the catches from gillnets and seines come from the Dutch coastal area in quarter 3 (Figure 2.2.3.1.4).


Figure 2.2.3.1.4. Quarterly Dutch catches (tonnes) per ICES rectangle and gear (average 2008–2012).


In the IMARES market sampling data on length, age, sex and weight are collected for several commercially important species. For sea bass this is done on an irregular basis and data are only available for some years. Market sampling is done since 2005 (Table 2.2.3.2.1). The age sampling frequency is now set triennially (2010, 2013, etc.). Every three years four samples of 15 fish (60 fish in total) are aged and every year the lengths of 24 samples of 15 fish (360 fish in total) are taken.

-4 -2 0 2 4 6

4950

5152

5354

Q1-flyshoot

longitude

latit

ude

0.7

0.2

0.1

0.5

3.9 0.1

1.9

12.5

1.1

1

6.7

4.4 0.5

0.1 0.2

0.2

0.1

0.2

-4 -2 0 2 4 6

4950

5152

5354

Q2-flyshoot

longitude

latit

ude

0.1

0.2

0.2

0.3 0.1

0.1

0.9

6

0

0.5

4.9

2.5

0

0.2

0

0.4 0.6

0.8

0.1

0.2

0

0

0

0.1

0

0

0

0.1

0.1

-4 -2 0 2 4 6

4950

5152

5354

Q3-flyshoot

longitude

latit

ude

0.2

0

0

0

0.1

0.1

0

0.1 1.4

0.2

0

0.1

0

0

0

-4 -2 0 2 4 6

4950

5152

5354

Q4-flyshoot

longitude

latit

ude

0

0.6

0.1

1.4

5.2

0.1

1.1

0.2

7.6

2

0.1

0.3

1.5

1

4.6

0.1

0

-4 -2 0 2 4 6

4950

5152

5354

Q1-gillnets-seines

longitude

latit

ude

0.1

0

0.1

0

0

0

0.4

0

0

-4 -2 0 2 4 6

4950

5152

5354

Q2-gillnets-seines

longitude

latit

ude

0

0.1

0.1

0

0.5

0

0

0.3

0.7

0.9

0.2

0

0.2

0.5

1.6

1.1

0.8

0

0.1

1.1

0

-4 -2 0 2 4 6

4950

5152

5354

Q3-gillnets-seines

longitude

latit

ude

0

0.1

0

0.6

2.1

1.1

0

0.1

0.6

2.4

3.9

2.4

2

0.2

0.1

0.1

3.5 0.6

-4 -2 0 2 4 6

4950

5152

5354

Q4-gillnets-seines

longitude

latit

ude

0.2

0.1

0

0

1

0.6

0.1

0.2

0.1

0.9

2.6

1

0.2

0.1

0.2

-4 -2 0 2 4 6

4950

5152

5354

Q1-lines

longitude

latit

ude

0.1

1 0.4

0.5

0.4

0

0.1

2

0.1

0.9

-4 -2 0 2 4 6

4950

5152

5354

Q2-lines

longitude

latit

ude

0.1

0.2

0.2

0.1

0.1

7.6

7.4

0.1

0.1

0.4

27.2

0.5

0.2

0.4

0.2

0.1

0

0

0

-4 -2 0 2 4 6

4950

5152

5354

Q3-lines

longitude

latit

ude

0.7

0.3

0.2

3.7

6.6

0.1

2.2

0.1

21.1

1.3

0

0.1

0.2

0.9

0.3

0.4

0.3 0

-4 -2 0 2 4 6

4950

5152

5354

Q4-lines

longitude

latit

ude

0.4

0.2 0.1

0.3

0.4

0.2

3.2

6.8

0.1

0.6

0.1

18.4

1.5

0.2

1.1

0.5

0.3

0

0

0

-4 -2 0 2 4 6

4950

5152

5354

Q1-beamtrawl

longitude

latit

ude

0.9

0.2

13.7

30.4

6.9

0.9

0

0

0.1

1

1.1

0.3

0

-4 -2 0 2 4 6

4950

5152

5354

Q2-beamtrawl

longitude

latit

ude

1.4

0.5

0.1

10.3

17.8

2.5

0.3

0

0.2

2.1

1.8

0.2

0.2

0

0.1

0.1

0

0

0

-4 -2 0 2 4 6

4950

5152

5354

Q3-beamtrawl

longitude

latit

ude

0.7

1.1

4.8

4.4

1.8

0.2

0

0

0.1

1

0.6

0.2

0

0

0.1

0.1

0

-4 -2 0 2 4 6

4950

5152

5354

Q4-beamtrawl

longitude

latit

ude

0.3

0.1

8.4

11.9

2.7

0.4

0

0.1

0.1

0.8

1.4

0.5

0

0

0

0


Table 2.2.3.2.1. Netherlands sea bass age and length samples from market sampling.

YEAR NR AGE SAMPLES NR LENGTH SAMPLES

2005 44 46

2006 55 57 2007 110 125 2008 202 202 2009 0 609 2010 340 838 2011 0 704

2012 0 421

2.3 Discards

2.3.1 UK data

2.3.1.1 Methods

An observer scheme to estimate discards has been in place in England and Wales rou-tinely since 2002 to meet DCF requirements, and sporadically prior to then. The Scot-tish observer scheme covers fleets that predominantly cover areas at or above the northern limit of sea bass, and therefore only the UK(E&W) data are considered here. The design of the scheme is based around random selection of vessels within regional and gear-related strata. The sampling scheme covers all types of vessels for which a derogation under DCF is not present. Significant fleet sectors with no sampling include lines and mollusc dredges. Discarding in hook and line fisheries is expected to be low in the commercial fisheries predominantly targeting areas with marketable bass, and survival of discarded bass caught by handline or rod and line in shallow water may be relatively high. However the absence of data for line fisheries will lead to some under-estimation of discards and an inability to apply release mortality rates.

The numbers of trips sampled in the main gear groupings that could catch bass and are covered by the scheme, otter trawls, beam trawls and lines, has varied widely from year to year and between areas (Table 2.3.1.1.1.1). Sporadic trips have also been sam-pled on board longliners, midwater trawlers and seines, but there are too few data to allow estimation of discard quantities and retention curves for sea bass.

Estimation of bass discards is problematic because the observer scheme covers all ves-sel trips in a stratum without reference to target species, and overall it samples less than 1% of all fishing trips. As bass are absent or at very small numbers in a large fraction of fishing trips throughout the year, particularly in winter, the amount of sam-ple data on bass is very low and the estimates are likely to have poor precision and variable biases related to inclusion of under-10 m vessels. Vessels under 10 m, which are responsible for a large fraction of bass catches, were excluded until recent years on health and safety and other logistical grounds, and although under-10 m vessels are now included, the fleet of vessels under 7 m remains excluded.

Two approaches to raising the UK discards data were considered in previous assess-ments: 1) a “design-based” estimator calculating sampling probability as the ratio of numbers of trips in the fleet to numbers of trips sampled within a stratum; 2) a ratio estimator using the ratio of reported landings in a stratum to the computed retained weight of sea bass in sampled trips. Neither of these approaches is unbiased, as the


variable treatment of under-10 m vessels precludes reliable estimation of sampling probabilities for method (1) for many of the years, and also affects the assumption of method (2) that the retention curve for sampled vessels is representative of all vessels in the stratum. The non-reporting of potentially large quantities of sea bass landings in the under-10 m fleet, due to the dispensation from reporting trip landings under 25 kg (see Section 2.2.1.2), also means that applying method (1) may lead to raised estimates of retained catch that are much higher on average than the reported landings for the fleet. Therefore the inclusion of absolute estimates of discards from method (1) may lead to incorrect estimation of retention parameters and fishing mortality in Stock Syn-thesis. To correct for this would mean re-scaling the discards estimates according to the ratio of estimated to reported retained catch, which is equivalent to method (2). Method (2) raising using ratio of reported to observed landings, was therefore used for sea bass to ensure internal consistency in the handling of the data within Stock Synthe-sis, but many caveats remain concerning the accuracy of the estimates.

[Note: UK discards estimates provided to IBPBass contained a raising error detected at the subsequent ICES WGCSE 2014 meeting. Correct figures were provided to WGCSE, giving lower discard rates, and the revised figures are included in the IBPBass report].


Table 2.3.1.1.1.1. Numbers of trips sampled for discards by Cefas (UK): 2002–2012, by gear group and area.

2.3.1.2 Time-series of discards and length compositions

Numerically, the largest numbers of bass discarded from UK fisheries are from the bottom-trawl fleet (Table 2.3.1.2.1), with much smaller numbers from nets (Table 2.3.1.2.2) and even less from beam trawls (Figure 2.3.1.2.1). Only eleven midwater pair trawl trips have been observed, and discarding of bass was negligible as the fishery targets mainly adult bass. No bass discards were observed in the eight longline trips observed. The raised length frequencies of discarded and retained bass, aggregated over all years, are plotted in Figure 2.3.1.2.1 along with the retention ogives. It is clear that discarding is driven primarily by the minimum landing size of 36 cm.

(a) otter trawlersDivision 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Grand TotaIV 16 34 56 37 41 85 58 49 46 42 54 30 548VIIafg 8 15 23 8 11 43 50 28 22 22 22 12 264VIId 1 2 4 3 1 2 1 6 7 9 4 5 45VIIeh 9 24 37 31 49 90 87 38 29 32 29 45 500total 34 75 120 79 102 220 196 121 104 105 109 92 1357

(b) NettersDivision 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Grand TotaIV 2 1 11 31 15 20 15 11 13 18 137VIIafg 3 7 5 3 7 8 9 10 7 16 22 16 113VIId 1 17 6 4 1 7 10 42 88VIIeh 1 5 9 2 3 16 10 14 19 17 25 24 145total 4 12 17 6 21 72 40 48 42 51 70 100 483

(c) Beam trawlersDivision 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Grand TotaIV 8 5 10 1 2 1 7 6 8 4 9 61VIIafg 3 11 7 4 8 1 2 3 3 1 4 8 55VIId 1 5 2 3 1 1 2 4 1 2 3 25VIIeh 10 17 27 16 24 32 18 13 17 27 22 21 244total 21 34 49 23 37 34 22 25 30 37 32 41 385

(d) Long linersDivision 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Grand TotaIV 1 1 2 4VIIafg 1 1 2VIIdVIIeh 1 1 2total 0 1 1 0 0 1 2 0 0 0 2 1 8

(e) Midwater trawlsDivision 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Grand TotaIV 1 1 2VIIafgVIId 1 1VIIeh 1 1 1 2 1 2 8total 1 1 1 3 2 1 0 0 0 0 0 2 11

(f) SeineDivision 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Grand TotaIV 1 1 2VIIafg 1 1 1 3VIIdVIIehtotal 1 0 1 0 1 1 0 0 0 0 0 1 5


Figure 2.3.1.2.1. Numbers of bass retained and discarded, summed over the period 2002–2013, for three UK fleet groupings. Plot (d) shows the retention ogive for the three gears.


Table 2.3.1.2.1. Estimated annual numbers and weight of sea bass retained and discarded by UK otter trawl fleets in Areas IV, VIId, VIIeh and VIIafg, based on at-sea sampling, and raised from landings in sampled strata to landings in all strata. Numbers of sampled trips are shown.

2002 2002 2003 2003 2004 2004 2005 2005 2006 2006 2007 2007Length cm Discarded Retained Discarded Retained Discarded Retained Discarded Retained Discarded Retained Discarded Retained

14 0 0 0 0 0 0 0 0 0 0 0 016 0 0 0 0 0 0 0 0 0 0 0 018 0 0 0 0 0 0 0 0 0 0 0 020 0 0 0 0 55 0 0 0 0 0 0 022 0 0 0 443 0 0 0 0 0 0 0 024 0 0 557 0 590 0 0 0 1293 0 20307 026 0 0 1670 0 3170 0 0 0 140 0 49045 028 0 0 43 25 27805 0 1941 0 7966 0 43833 030 34518 0 1338 0 74804 0 4010 0 7876 0 66122 032 12952 0 22651 125 62348 1627 3168 0 79717 0 19281 78034 6476 45702 16641 36414 16348 29584 6492 18021 13024 150931 6306 2550436 0 16350 0 70184 0 55920 0 31911 176 47651 0 5843438 0 9874 0 42931 91 46143 0 53634 0 50527 0 4959040 0 39592 0 40345 0 37867 0 40399 0 26373 0 5125042 0 22826 0 29675 0 30534 0 12694 0 18928 0 3477144 0 6795 0 14886 0 14326 0 12655 0 9631 0 2024546 0 21200 0 10948 0 10852 0 6700 0 4608 0 839648 0 39911 0 6657 0 4372 0 1213 0 3589 0 511850 0 4530 0 10981 0 3726 0 34856 0 1264 0 659752 0 0 0 3129 0 2197 0 4751 0 3169 0 517954 0 3398 0 3604 0 2062 0 3639 0 594 0 97356 0 3398 0 1435 0 1385 0 0 0 646 0 74758 0 0 0 2708 0 823 0 0 0 896 0 69960 0 0 0 523 0 748 0 0 0 491 0 062 0 0 0 1157 0 367 0 0 0 0 0 064 0 0 0 2227 0 937 0 0 0 0 0 066 0 0 0 0 0 367 0 0 0 140 0 6568 0 0 0 0 0 306 0 0 0 0 0 47770 0 0 0 0 0 762 0 0 0 0 0 072 0 0 0 0 0 122 0 0 0 646 0 19574 0 0 0 0 0 61 0 0 0 0 0 076 0 0 0 443 0 61 0 0 0 0 0 0

Total 53947 213575 42899 278841 185212 245150 15611 220473 110191 320084 204893 269021Trips sampled 34 75 120 79 102 220

Catch weight (t) 21 199 19 238 69 203 6 199 45 213 62 219

2008 2008 2009 2009 2010 2010 2011 2011 2012 2012 2013 2013Length cm Discarded Retained Discarded Retained Discarded Retained Discarded Retained Discarded Retained Discarded Retained

14 0 0 13412 0 0 0 0 0 0 0 0 016 0 0 13412 0 0 0 0 0 0 0 0 018 0 0 147535 0 0 0 0 0 0 0 0 020 0 0 228009 0 0 0 0 0 0 0 0 022 0 0 162898 0 0 0 0 0 427 0 0 024 0 0 40434 0 11020 0 0 0 577 0 0 026 0 0 28735 0 25098 0 241 0 1331 0 0 1523128 718 0 3103 0 24678 0 2717 0 1903 0 347 1719630 6066 645 7572 0 31413 83 5985 0 19237 0 2879 1115632 5429 18 32708 891 51214 72 10449 0 26924 3015 3845 575234 3567 20288 2402 5823 10166 11836 2984 11229 19352 53149 2297 1005936 0 29509 32 31223 281 28919 75 36715 181 70375 0 3676038 0 46158 0 38411 0 32474 0 42188 0 25638 0 3556540 0 54494 0 23533 0 26718 0 11499 0 29408 0 2485442 0 43448 0 31421 0 24799 0 17061 0 8260 0 1278944 0 34317 0 13676 0 13494 0 11330 0 8099 0 603346 0 10650 0 6419 0 8565 0 6257 0 11913 0 319048 0 11372 0 6989 0 4526 0 6972 0 4795 0 438250 0 5190 0 15994 0 3746 0 5712 0 330 0 179952 0 3713 0 4776 0 5099 0 1080 0 2567 0 49154 30 4157 0 1853 0 3000 0 3127 0 346 0 175256 0 969 0 2624 0 90 0 1618 0 2442 0 259258 0 451 0 444 0 221 0 776 0 797 0 69460 0 1984 0 862 0 2278 0 2895 0 356 0 49162 0 645 0 0 0 2057 0 1189 0 401 0 064 0 0 0 85 0 1590 0 705 0 632 0 066 0 153 0 42 0 85 0 60 0 0 0 068 0 0 0 0 0 0 0 0 0 0 0 070 0 0 0 891 0 0 0 0 0 0 0 072 0 17 0 778 0 686 0 0 0 0 0 074 0 0 0 891 0 0 0 0 0 0 0 42276 0 0 0 1781 0 0 0 0 0 0 0 0

Total 15809 268180 680253 189406 153868 170340 22451 160414 69931 222525 9368 191210Trips sampled 196 121 104 105 109 92

Catch weight (t) 6 239 94 193 52 161 9 145 29 163 4 131


Table 2.3.1.2.2. Estimated annual numbers and weight of sea bass retained and discarded by UK vessels using fixed or driftnets in Areas IV, VIId, VIIeh and VIIafg, based on at-sea sampling, and raised from landings in sampled strata to landings in all strata. Numbers of sampled trips are shown. Results for 2002–2006 are omitted due to insufficient coverage of area strata.

2.3.2 French data

2.3.2.1 Time-series of discards and length structure

French data on discards of sea bass were not used in IBPBass. The French OBSMER project currently estimates discards to meet DCF requirements. The methodology and the results are described in the annual French report on: http://ar-chimer.ifremer.fr/doc/00167/27787/25978.pdf.

In WGCSE 2013 some information on precision is given: "Discards estimates for UK and France are from vessel selections that for some areas and gears include relatively limited numbers of observed trips where sea bass is caught and discarded. Precision is therefore very low at current sampling rates".

In France the low sampling rate observed can be explain by the low discarding rate. For information in 2013, only 26 tons would have been discarded in the studied area. Length frequencies were not available at IBPBass. [Detailed French data on discards of sea bass were not available to IBPBass, but were provided, by weight, to WGCSE 2014.]

2007 2007 2008 2008 2009 2009 2010 2010 2011 2011 2012 2012 2013 2013Length cm Discarded Retained Discarded Retained Discarded Retained Discarded Retained Discarded Retained Discarded Retained Discarded Retained

14 0 0 0 0 0 0 0 0 0 0 0 0 0 016 0 0 0 0 0 0 0 0 0 0 0 0 0 018 0 0 0 0 0 0 0 0 0 0 0 0 1657 020 0 0 0 0 0 0 8810 0 0 0 0 0 1657 022 139 0 0 0 360 0 0 0 0 0 0 0 0 024 139 0 0 0 0 0 0 0 0 0 0 0 0 026 139 0 0 0 360 0 0 0 0 0 0 0 0 028 139 0 0 0 0 0 0 0 0 0 591 0 0 030 312 0 0 0 360 0 0 0 0 0 0 0 0 032 1108 554 0 0 0 0 0 0 33990 1073 0 791 3363 034 416 3741 4590 1468 360 360 0 0 0 242 3560 17404 0 765636 0 5139 0 4895 0 720 0 0 0 1194 0 29666 0 2203938 0 9006 0 17434 0 104506 0 8810 0 1073 0 31444 0 4921840 0 12495 0 43330 0 2879 0 0 0 1194 0 16613 0 2633242 0 9296 0 3549 0 57920 0 0 0 1209 0 24830 0 3357844 0 6902 0 21531 0 56031 0 17619 0 410 0 23248 0 1877346 0 4538 0 27221 0 6657 0 0 0 7277 0 3545 0 1585848 0 1148 0 36097 0 5487 0 17619 0 52698 0 30924 0 1146750 0 3628 0 16676 0 17802 0 23160 0 20486 0 17340 0 3512752 0 10739 0 16112 0 10745 0 39571 0 13661 0 14771 0 3718454 0 5653 0 5326 0 9666 0 19255 0 45211 0 17134 0 2559656 0 7340 0 19989 0 1799 0 14350 0 40899 0 23294 0 1186658 0 11332 208 5693 0 3238 0 18706 0 3281 0 25406 0 541960 0 10809 0 7213 0 2159 0 0 0 1230 0 9113 0 775462 0 9307 0 2921 0 7777 0 4448 0 1346 0 8863 0 439164 0 17565 0 697 0 0 0 0 0 137 0 2954 0 066 0 655 0 830 0 360 0 8810 0 0 0 0 0 102868 0 277 0 415 0 0 0 0 0 137 0 0 0 102870 0 450 0 489 0 360 0 0 0 0 0 1182 0 072 0 757 0 489 0 360 0 0 0 0 0 0 0 074 0 0 0 0 0 360 0 0 0 0 0 0 0 076 0 139 0 979 0 0 0 0 0 0 0 0 0 102878 0 0 0 0 0 0 0 0 0 0 0 0 0 080 0 0 0 0 0 0 0 0 0 0 0 0 0 082 0 0 0 0 0 0 0 0 0 0 0 0 0 084 0 0 0 0 0 360 0 0 0 0 0 0 0 0

Total 2390 131469 4798 233355 1439 289544 8810 172347 33990 192756 4151 298523 6677 315341Trips sampled 72 40 48 42 51 70 100

Weight (t) 0.9 239.3 2.8 318.7 0.4 313.4 1.0 304.9 14.2 327.2 1.9 410.4 1.7 405.2

http://archimer.ifremer.fr/doc/00167/27787/25978.pdf



2.4 Recreational catches

Recreational marine fishery surveys in Europe are still at an early stage in development (ICES WGRFS 2012b, 2013b). Survey design and methods of recreational catch estima-tion are described in the stock annex (Annex 1).

France

A survey of recreational fishers, focusing mainly on bass, was conducted between 2009 and 2011. Estimates of sea bass recreational catches were obtained from a panel of 121 recreational fishermen recruited during a random digit dialling screening survey of 15 000 households in the targeted districts. 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 (Table 2.4.1).The estimates for Subarea IV and VII were 940 t kept and 332 t released.

The precision of the combined Biscay and Channel estimate was relatively low (CV =26%; note that the figure of 51% given in IBPNEW 2012 (ICES, 2012a) was incorrect). This gives an average and 95% confidence intervals of 3170 t [1554 t; 4786 t] for the whole Subareas IV, VII and VIII. Increasing the panel from 121 to 210 fishermen would be expected to improve precision to 20% and increasing this panel to 500 would im-prove precision to 13%.

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 rec-reational catch was taken by sea angling (rod and line or handline).

Taking into account a potential hooking mortality of 20%, the estimate of annual French recreational fishery removals from Areas IV and VII in 2009–2011 is increased to just over 1000 t.

A new survey was conducted from July 2011 to December 2012, based on a similar methodology to the previous study (not only on sea bass this time, but also on other marine species including crustaceans and cephalopods). A random digit dialling screening survey of 16 130 households led to the recruitment of a panel of 183 fisher-men to keep logbooks. In parallel, 151 fishermen were recruited on site by the Promo-peche association, and 30 more via the sea bass fishermen panel set up in 2009. This resulted in 364 panel members keeping logbooks describing their catches (species, weight, size, etc.) The focus of the survey on sea bass shows that in Atlantic (Bay of Biscay and Channel), the estimated recreational catch of bass in 2012 was 3922 t of which 3146 t was kept and 776 t released. At this time results have to be considered as provisional, (results split between Bay and Biscay and Channel are not available yet with relative standard error).

Netherlands

A recent survey investigated the amount of sea bass caught by recreational fishers (van der Hammen and de Graaf, 2012; ICES, 2012b) from March 2010 to February 2011. Es-timates of sea bass catches recreational were obtained from a panel of 1043 recreational fishermen recruited during a telephone survey of 109 293 people. Revised estimates were provided to WGCSE 2013 (ICES, 2013a). The catch weights are estimated with a limited amount of length–frequency data, and are therefore less reliable than the esti-mates in numbers (and may also be adjusted if more data are available). For the same reason, there are no ‘returned’ estimates by weight (yet).


The estimated total recreational catch of sea bass was 366 000 fish (RSE 30%), of which 234 000 were retained, equivalent to 138 t (Table 2.4.2). These results are mainly appli-cable to Subarea IV.

UK (E&W)

A new survey programme Sea Angling 2012, based on a statistically sound survey de-sign started 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. Results are available at http://www.marinemanage-ment.org.uk/seaangling/.

The surveys adopted, where possible, statistically sound, probability-based survey de-signs, building on knowledge gained through participation in the ICES Working Group on Recreational Fishery Surveys (WGRFS). Two survey approaches were adopted: first a stratified random survey of charter boats from a list frame covering ports in England, and secondly an on-site stratified random survey of shore anglers and private boat anglers to estimate mean catch per day, combined with annual effort estimates derived from questions added to a monthly Office of National Statistics household survey covering Great Britain.

A list of almost 400 charter boats was compiled for the charter boat survey, and 166 skippers agreed to participate. Each month over a twelve month period in 2012 and 2013, 34 randomly selected skippers completed a diary documenting their activities, catches and sizes of fish. A diary was completed whether or not any fishing took place. Data from 5300 anglers were collected. Total annual catches were estimated by raising the monthly catches per vessel from the diaries to all vessel-month combinations in the frame, and raising this to all vessels including refusals. The estimated total annual catch of sea bass for the entire coast of England was 44 t (RSE 31%) of which 31 t was kept. The release rate by number was 37%. The charter boat survey has potential bias due to the large non-response rate, if non respondents have different catch rates to respond-ents.

The Office of National Statistics (ONS) household survey covered 12 000 households during 2012, and from this it was estimated that 2.2% of adults over 16 years old went sea angling at least once in the previous year. The surveys estimated there are 884 000 sea anglers in England. Estimation of fishing effort by shore and private boat anglers proved very difficult due to the overall small number of households with sea anglers in the survey. A range of methods was explored to estimate annual and seasonal effort using the ONS data alone, and combining it with observations from on-site and online surveys. It has not been possible yet to agree on a best estimate of effort, and for that reason the estimates of total catch (cpue × effort) for shore and private boat angling are given as a range of plausible values.

The survey of anglers fishing from the shore and private boats to estimate cpue was carried out throughout 2012 using on-site interviews. A stratified random design was adopted to select shore sites and boat landing sites on a weekly basis from site lists stratified into low-activity and high-activity sites. The shore survey used roving-creel methods (collecting data from partial angling trips), and the private boat survey a rov-ing access-point survey (data from completed trips). Visits were made to 1475 shore sites and 425 private boat sites, and 2440 anglers were interviewed. The mean daily catch rate of kept and released fish of each species was estimated based on the survey design, and sizes of caught fish were recorded. Cpue for shore angling was estimated

http://www.marinemanagement.org.uk/seaangling/



using catches for the observed trip duration and estimates of expected total trip dura-tion for that day. A length-of-stay bias correction was applied based on expected total trip duration. The catch-per-day estimates were combined with estimates of total an-nual fishing effort (days fished) obtained from the ONS survey to estimate total annual catches. Release rates, by number, were 82% for shore angling and 57% for private boats. Non-response rates were very low (<10%) in this survey. The range of point es-timates for shore-caught bass was 98–143 t (total) and 38–56 t (kept), and for private and rented boats was 194–546 t (total) and 142–367 t (kept). The relative standard errors for the individual shore and private boat estimates were relatively high at 40–50%.

Combining the catch estimates for charter boats, private boats and shore angling, the point estimates of annual kept weights of sea bass ranged from 230 t–440 t (Table 2.4.1c), compared with total UK commercial landings of almost 900 t in 2012. The com-bined estimates of bass catches had precision (relative standard error) estimates of 26%–38% for the different effort estimation methods. The relatively large standard er-rors combined with the range of plausible methods of estimating effort for shore and private boats.

Hooking mortality rates

The US National Marine Fisheries Service has in the past used an average hooking mortality of 9% for striped bass, estimated by Diodati and Richards (1996). 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 experi-ments by 16 different authors where striped bass hooking mortality was estimated over two or 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 condi-tions and angling methods typical of European fisheries.

Total recreational catch

The recent estimates of total recreational removals of sea bass for France, Netherlands and England in Subareas IV and VII amount to 1300–1500 t. Assuming a 20% hooking mortality rate, an additional quantity of around 110–130 t of releases will have died, assuming the same release rate in the Netherlands as in England (release rates by num-ber in England and the Netherlands were similar). The total recreational removals were therefore around 1400 t–1600 t compared with total reported commercial fishery land-ings of 4100 t on average during 2009–2012. The reported commercial fishery landings are an underestimate of total commercial removals due to exclusion of discards and unreported landings associated with the exemption to report trip landings under 25 kg in the UK (30 kg elsewhere). However, the recreational catch estimates exclude figures for Belgium, Wales or any other European countries. It must therefore be concluded that recreational fishing may account for almost 30% of total fishery removals, and this represents a significant missing catch from the assessment. The impact of this on the assessment is unknown, because there are no historical estimates to determine the trend in recreational catches which may differ from commercial fishery. It is possible that, before the large growth in biomass of the stock in the 1990s, recreational fishing may have been a much larger proportion of total fishery removals.


Table 2.4.1. Estimates of annual recreational fishery removals of sea bass in France (catches by weight, in tonnes).

KEPT RSE RELEASED RSE TOTAL RSE RELEASE

RATE

2009–2011

NE Atlantic

2343 t 830t 3173 t 26% 26%

ICES Areas IV & VII

940 t 332t 1272 t >26% 26%

2011–2012

NE Atlantic

3146 t 776t 3922 t 20%

1 RSE was 26% for Area VII and VIII combined; Area VII represented 40% of total.

(~80% by weight in 2009/2011 was recreational sea angling).

Table 2.4.2. Estimates of annual recreational fishery removals of sea bass in the Netherlands (North Sea).


RATE

March 2010–February 2011

By number 234 000 38% 131 000 27% 366 000 30% 64%

By weight 138 t 37%

(98% by weight is recreational sea angling).

Table 2.4.3. Estimates of annual recreational sea angling removals of sea bass in England (ranges of values are for different methods of estimating angling effort): North Sea; Channel; Celtic & Irish Seas.


RATE

2012 By weight 230–440 t 150–250 t 380–690 t

26–38% 36–39%

2.5 Cpue data

2.5.1 Dutch data

A problem with commercial lpues (landings per unit of effort) for sea bass is that the fishing effort is distributed across many areas where sea bass have low probability of capture. Dutch lpue series were created in van der Hammen et al. (2013) by selecting gears with the highest sea bass landings and by selecting the rectangles where sea bass was caught in substantial amounts lpues were calculated for five gear groups (gillnets and seines, lines, flyshoot and beam trawl) from 2000 to 2013 (Table 2..5.1.1). In addi-tion, the following selection was made:

• Only those rectangles were selected which were sampled in eleven out of 13 years (85%);

• Only those rectangles were selected in which sea bass was caught at least once;


• For lines, the time-series was limited to 2005–2012, because permits were obligated since 2005 for commercial line fishing;

• Specific selections per gear are listed in:

Table 2.5.1.1. Dutch lpue series (gears and ICES rectangles selection).

GEARS GEARCODES DATABASE ICES

AREA ICES RECTANGLES OTHER SELECTIONS

lines "LH","LHM","LHP", "LL" and "LLS"

IVc 34F4, 35F4, 33F3, 33F4, 32F4, 33F2, 32F2, 32F3, 31F2, 31F3

The time-series was restricted to years after 2005, due to obligation of permits for commercial line fishing since 2005.

Gillnets/seines "GN", "GND", "GNS", "GTR" and "PS"

IVb,c 35F6, 36F6, 35F4, 35F5, 34F4, 34F5, 33F4, 34F3, 32F4,33F3, 32F2,32F3,31F3,31F4

Years >2000 (catches registered since 2000).

Beam trawl "TBB" IVb,c 31F1, 31F2, 31F3,

32F1, 32F2, 32F3, 33F2, 33F3, 33F4, 34F2, 34F3, 34F4, 35F2, 35F3, 35F4, 36F2, 36F3, 37F1, 37F2, 37F3, 38F2, 39F6

Only vessels with kW >221 were included in the analysis. Years >2000 (catches registered since 2000).

flyshoot "SDN" and "SSC" VIId 29E7, 29E9, 29F0 Only ICES rectangles in Area VII were included. In Area IV three rectanges had enough sampled years, but only very little catches per rectangle.

twinrig "OTT" - - Not enough landings for a time-series

2.5.1.1 Lpue methods

Landings per Unit of Effort (lpue) data were corrected for targeting behaviour as de-scribed below. The methods are similar to those used to analysed commercial lpue data for North Sea plaice, described in van der Hammen et al. (2011). Landing rates (lpue) were calculated for the period 2002–2012.

2.5.1.2 Correction for targeting behaviour

Fishers target fishing areas with high concentrations of fish. Dividing total landings by total effort without taking in account targeting behaviour may result in bias of com-mercial lpue because of possible changes in the spatial distribution of fishing effort. Therefore, a correction was carried out using EU logbook data. Lpue was first calcu-lated per ICES rectangle, per year. Next, a selection was made in which only those rectangles visited by at least eleven out of 13 years (85% of the years). This ensures that the lpues are valid for the core area of the fleet, and not influenced much by many missing values. Subsequently, the lpues by ICES rectangles were averaged to calculate


the lpue by year for the core fishing area of the Dutch beam trawlers in the North Sea. This removes the major effects of changes in spatial effort allocation due to, for in-stance, changing targeting behaviour.

2.5.1.3 Lines

The selection of ICES rectangles resulted in ten rectangles, which all lay in ICES Divi-sion IVc. The effort in days at sea for the commercial fisheries with lines has increased from less than 500 days at sea in 2005 to almost 1000 days at sea in 2011 (Figure 2.5.1.3.1). The catches also increased, from 52 tonnes to 147 tonnes in 2011. The lpue fluctuates without clear trend from 2005–2009, and had somewhat higher lpues in 2010–2012, the last three years of the time-series (Figure 2.5.13.1).

Figure 2.5.1.3.1. Effort (in days at sea) and lpue (in kg per day at sea) of lines, for the selected rec-tangles. Source van der Hammen et al. (2013).

2.5.1.4 Gillnets-seines

The selection of ICES rectangles resulted in 14 rectangles in ICES Divisions IVbc. The effort in days at sea for the fisheries with gillnets and seines increased from less than 358 days at sea in 2000 to 2241 days at sea in 2012 (Figure 2.5.2). The catches also in-creased, from 2.6 tonnes in 2000 to 28 tonnes in 2012. The lpue increased from 3 kg in 2000 to 29 kg per day at sea in 2012 (Figure 2.5.1.4.1). The highest lpue is in 2011.

2005 2006 2007 2008 2009 2010 2011 2012

050

010

0015

00

effort: lines

year

Effo

rt (d

ays

at s

ea)

2005 2006 2007 2008 2009 2010 2011 2012

050

100

150

200

LPUE (kg/day at sea): lines

year

LPU

E (k

g/da

y at

sea

)


Figure 2.5.1.4.1. Effort (in days at sea) and lpue (in kg per day at sea) of gillnets-seines, for the se-lected rectangles. Source van der Hammen et al. (2013).

2.5.1.5 Beam trawl

The selection of ICES rectangles resulted in 22 rectangles in ICES Divisions IVb and c. The effort in days at sea for the large beam trawlers in these rectangles has nearly halved from almost 30 000 days at sea in 2000 to less than 15 000 days at sea in 2012 (Figure 2.5.1.5.1). The catches increased from 31 tonnes at the beginning of the time-series to around 150 tonnes between 2005 and 2009, but in recent years the catches de-creased again. The lpue has increased between 2001 and 2010 and decreased again in 2011 and 2012 (Figure 2.5.3. On average, the highest lpues are found in the most south-ern part of the North Sea. The trend fluctuates between ICES rectangles, although most rectangles in the southern North Sea follow the trend of higher lpue at the beginning of the time-series and lower at the end of the time-series.

2000 2002 2004 2006 2008 2010 2012

050

015

0025

00

effort: gillnets-seines

year

Effo

rt (d

ays

at s

ea)

2000 2002 2004 2006 2008 2010 2012

010

2030

4050

LPUE (kg/day at sea): gillnets-seines

year

LPU

E (k

g/da

y at

sea

)


Figure 2.5.1.5.1. Effort (in days at sea) and lpue (in kg per day at sea) of beam trawl, for the selected rectangles. Source van der Hammen et al. (2013).

2.5.1.6 Flyshoot

The selection of ICES rectangles resulted in six rectangles, three in ICES Subarea IV and three in Division VIId. The amount of catches in Subareas IV and VII differed about a factor 30–50, with high catches in Subarea VII and very low catches in Subarea IV (close to 0). It was therefore decided that the analysis was done for the three ICES rectangles in Subarea VII only. The effort in days at sea for the flyshoot in these rectangles has increased from less than 100 days at sea in 2000 to almost 1000 days at sea in 2011 (Figure 2.5.1.6.1). The catches also increased, from 930 tonnes to 86 484 tonnes in 2012. The lpue increases during the time-series from less than 10 tonnes per day at sea in 2000 to almost 1000 tonnes per day at sea in 2012, with a small dip between 2009 and 2011 (Figure 2.5.1.6.1).

2000 2002 2004 2006 2008 2010 2012

010

000

2000

030

000 effort: beamtrawl

year

Effo

rt (d

ays

at s

ea)

2000 2002 2004 2006 2008 2010 2012

05

1015

2025

30

LPUE (kg/day at sea): beamtrawl

year

LPU

E (k

g/da

y at

sea

)


Figure 2.5.1.6.1. Effort (in days at sea) and lpue (in kg per day at sea) of flyshoot, for the selected rectangles. Source van der Hammen et al. (2013).

2000 2002 2004 2006 2008 2010 2012

020

060

010

00

effort: flyshoot

year

Effo

rt (d

ays

at s

ea)

2000 2002 2004 2006 2008 2010 2012

020

4060

8010

0

LPUE (kg/day at sea): flyshoot

year

LPU

E (k

g/da

y at

sea

)


3 Survey data

3.1 UK data

The UK has conducted prerecruit 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 in autumn 2011 and again in autumn 2013 to maintain a series for the assessment. The location of the surveys and the tow positions are shown in Figure 3.1.1. Both surveys use a high headline bass trawl, although in the Thames it is de-ployed as a twin rig and in the Solent as a single rig.

Figure 3.1.1. Location and tow positions for UK (England) Solent and Thames sea bass surveys.

The spring and autumn Solent survey index series are given in Table 3.1.1, updated to include the autumn 2013 survey and to amend an error in the autumn survey indices in 2000. The surveys do not show major year effects, but as noted in previous assess-ments the autumn (September) survey shows a general increase in recruitment during the 1990s up to the mid-2000s, with very low indices for the 2008 onwards year classes, while the spring survey shows poor recruitment from around 2002 onwards (Figure 3.1.2). Previous Stock Synthesis runs show that the autumn survey is much better fitted than the spring survey. The spring survey is likely to be more strongly affected by weather and by temperature effects on distribution.

The Thames survey series is given in Table 3.1.2. The series indicates an increase in recruitment from the mid-1990s to early 2000s followed by some poor year classes, possibly a strong 2007 year class; then weak year classes in 2008 and 2009 (Figure 3.1.3). A problem with the use of the Thames survey is that it may reflect recruitment from spawning that became established in the southern North Sea as the stock expanded in the 1990s under warmer sea conditions, and may therefore not reflect recruitment trends that influence the larger stock components in the Channel and Celtic Sea.


A justification of using the Solent survey as an index of recruitment over the full range of the stock was the results of a statistical, UK-only fleet-based separable model devel-oped by Pawson et al. (2007) and updated by ICES WGNEW (Kupschus et al., 2008). The Pawson et al. model was fitted only using UK age compositions for trawls, mid-water trawls, nets and lines, separately for ICES Divisions IVbc, VIId, VIIeh and VIIafg, and was intended mainly to estimate fleet selection patterns. Although it excluded any tuning data, the recruitment-series for each sea area closely resembled the Solent sur-vey indices and to an extent the shorter Thames series, and was able to provide coher-ent selection patterns by fleet.

The full Solent survey series was subject to a change in gear design in 1993. Some com-parative trawling was carried out to develop age-varying calibration factors, but these are poorly documented and the original raw data and calibration results are currently being sought at Cefas. Pending an evaluation of this, the benchmark Stock Synthesis runs included a sensitivity run with the series split into two periods around the gear change. Some additional issues with calibration factors applied to the spring survey were detected during the benchmark, and this is considered later in the sections on model development.

A precision estimate was calculated for the Solent and Thames surveys based on the between-tow variations in catch rate of all the age classes used in the index. For the Solent spring, Solent autumn and Thames surveys, the relative standard errors were 0.42, 0.25 and 0.43 respectively.


Table 3.1.1. Time-series of relative abundance indices for sea bass age groups 2, 3 and 4 from the UK Solent spring and autumn trawl surveys. A change in trawl design took place in 1993.

MAY–JULY SEPTEMBER

Year age 2 age 3 age 4 age 2 age 3 age 4

1981 0.00 0.30 0.25 No survey

1982 0.51 2.17 0.16 3.25 10.10 0.38

1983 No survey 9.87 0.91 1.88

1984 0.95 2.66 0.43 1.38 0.65 0.09

1985 0.00 10.33 2.56 No survey

1986 No survey 0.27 4.26 1.31

1987 0.00 0.42 3.18 0.05 0.28 2.27

1988 0.00 0.02 0.47 No survey

1989 No survey 6.68 0.37 0.00

1990 2.84 2.48 0.00 2.81 1.15 0.02

1991 5.78 0.62 0.09 3.08 0.21 0.03

1992 0.11 7.04 0.35 0.95 18.59 0.16

1993 0.05 7.33 14.02 6.65 3.59 4.39

1994 0.04 1.63 1.14 3.33 1.84 0.29

1995 0.05 1.57 0.97 4.83 4.69 0.72

1996 1.43 4.09 3.36 5.52 0.43 0.11

1997 0.27 1.94 0.11 33.62 4.52 0.06

1998 0.00 6.75 5.79 1.22 5.50 0.61

1999 0.61 0.95 12.30 19.37 0.67 0.87

2000 0.49 37.03 1.06 6.07 11.35 0.03

2001 1.71 6.33 3.43 34.42 3.92 1.57

2002 0.63 1.62 0.29 7.42 3.87 0.40

2003 0.06 0.32 0.38 8.37 4.60 0.59

2004 0.17 0.28 0.16 No survey

2005 0.05 0.42 0.35 13.12 7.98 0.84

2006 0.44 2.47 1.03 9.51 9.21 1.02

2007 0.33 0.50 0.50 3.42 1.78 0.30

2008 No survey 18.52 6.66 0.34

2009 0.72 1.03 0.13 13.25 6.25 0.33

2010 No survey No survey

2011 No survey 2.25 1.39 0.42

2012 No survey No survey

2013 No survey 1.34 0.08 0.10

Note: September 2000 data amended.


Table 3.1.2. Time-series of relative abundance indices for sea bass age groups 0–3 from the UK Thames trawl survey.

YEAR AGE 0 AGE 1 AGE 2 AGE 3

1997 7.737 0 0.048 0.41

1998 No survey

1999 19.54 6.033 0.764 0

2000 4.015 14.74 0.832 0.089

2001 121.5 11.47 5.108 0.171

2002 469 20.71 2.716 1.093

2003 225.6 35.76 4.429 0.159

2004 238.92 44.99 7.32 1.03

2005 37.04 14.49 6.86 0.75

2006 245.54 11.26 3.46 0.94

2007 No survey

2008 107.55 50.69 1.86 0.2

2009 95.43 7.79 13.59 0.91

Figure 3.1.2. Year and year-class effects in the Solent spring and autumn surveys, updated to include the September 2013 survey.


Figure 3.1.3. Year and year-class effects in the Thames survey.

3.2 French data

Raw data on sea bass from the French scientific trawl survey "Channel Ground Fish Survey - CGFS" were not available for the previous benchmark in 2012 (IBPNEW, ICES, 2012a). Details of the survey are given in Coppin et al. (2002), which includes a full description of the GOV trawl used in October each year at the 82 stations in ICES Di-vision VIId shown in Figure 3.2.1. The majority of sea bass are caught in the coastal waters of England and France (Figure 3.2.1). The abundance indices from all the sta-tions give similar trends as from a subset of stations in the main coastal areas, and trial runs with SS3 gave similar trends. Therefore, for further SS3 development, the indices calculated from all the area are used.

The abundance indices are calculated applying a stratified random swept-area based estimator. Strata correspond to ICES statistical rectangles. Swept-area is calculated us-ing wingspread. As this is a stratified swept-are based indicator, uncertainty is based on between haul variance within a strata and summation of variances across strata. Full methodology is presented in the WD_01, available in Annex 2.

The swept-area indices are given in Table 3.2.1. The trends in both the index and in the proportion of stations with sea bass show some similarities to the trend in total biomass estimates from the ICES WGCSE 2013 update assessment using Stock Synthesis, lend-ing a-priori support to the use of the index in the assessment (Figure 3.2.2).

The swept-area index length frequencies are given in Table 3.2.2.

In order to visualize the CGFS data more easily, the annual survey length frequencies were converted to age compositions using the July–December age–length keys (ALKs) from the UK commercial fishery sampling in Division VIId, supplemented by ALK data from the Solent autumn survey for the length classes not present in the commer-cial fishery (Figure 3.3.3). As would be expected due to the large overlap in length-at-age distributions between neighbouring ages, the resultant age compositions share similar features to the UK fishery age compositions using the same ALKs, including the reduced numbers of young bass in recent years.

The precision of the swept-area indices appears unrealistically high in some years (e.g. 0.025 in 1991), which may indicate that the index trends are driven largely by the inci-dence of positive catches. Modelling of the data using delta lognormal models may provide more realistic precision. During trial Stock Synthesis runs, the use of the CVs


in Table 3.2.1 resulted in an inability to fit the selection curve for the survey due to individual years being given far too much weight. Relaxing the CVs to 0.30 for all years except the first three years (set to 0.6 given the very low incidence of positive stations) allowed the model to fit the length compositions more closely over the series.

Figure 3.2.1. Left: stations fished during the Channel Groundfish Survey carried out annually by France. Right: distribution of total catches of sea bass over the survey series.

Figure 3.2.2. Mean-standardised time-series of (a) swept-area abundance indices and (b) percentage of stations with sea bass from the Channel Groundfish Survey, plotted together with the time-series of total biomass estimates from the WGCSE 2013 Stock Synthesis run.


Figure 3.3.3. French CGFS length compositions decomposed to age compositions using UK age–length keys from the VIId July–December fishery and Solent autumn survey. The 1989 year class is indicated. Size of bubbles is proportional to the square root of the index.


Table 3.2.1. Sea bass indices of abundance (swept-area) from the Channel Groundfish Survey. The relative standard error CV is the log transformed value used in SS3 (sqrt(loge (1+CV^2)).

YEAR TOTAL

HAULS NO. HAULS

WITH SEA

BASS

PROPORTION OF

HAULS WITH SEA

BASS

MEAN

NUMBERS OF

SEA BASS PER

HAUL

SWEPT-AREA

ABUNDANCE

INDEX

CV

1988 68 6 8.8 9 245 776 0.149

1989 61 3 4.9 4 77 716 0.579

1990 75 8 10.7 63 1 129 914 0.115

1991 79 19 24.1 163 4 250 636 0.025

1992 60 23 38.3 305 2 617 986 0.110

1993 65 21 32.3 177 2 299 919 0.101

1994 86 19 22.1 103 1 097 828 0.111

1995 166 17 10.2 78 1 021 741 0.089

1996 134 26 19.4 82 1 224 238 0.125

1997 169 31 18.3 181 1 817 599 0.121

1998 82 38 46.3 317 2 531 043 0.078

1999 102 37 36.3 286 1 642 271 0.118

2000 100 36 36.0 325 2 570 994 0.081

2001 109 39 35.8 363 3 150 674 0.142

2002 100 44 44.0 506 3 872 427 0.108

2003 94 41 43.6 813 8 739 056 0.111

2004 94 44 46.8 344 3 598 436 0.096

2005 105 40 38.1 297 3 005 315 0.084

2006 110 36 32.7 521 5 518 000 0.124

2007 103 33 32.0 261 3 661 314 0.136

2008 105 40 38.1 401 6 468 839 0.150

2009 102 26 25.5 192 2 564 694 0.090

2010 101 30 29.7 123 1 804 538 0.099

2011 108 27 25.0 108 1 513 742 0.123

2012 96 25 26.0 135 2 034 552 0.109


Table 3.2.2. Swept-area abundance indices by length class for the Channel Groundfish Survey.

Year/Length Bins (cm) 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 201214 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 016 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 018 0 0 32733 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 020 0 0 98199 0 0 0 18458 0 0 0 0 0 18644 0 0 0 0 0 0 0 0 0 0 0 022 0 0 120021 0 0 0 0 0 10096 0 0 18433 65257 0 0 22663 0 15565 0 0 10729 0 0 0 024 0 0 96615 304624 0 5551 5108 65628 71187 52420 0 15128 62040 59740 10739 13218 2692 0 9814 0 52876 0 0 0 026 0 0 222797 1719409 0 38106 71094 182886 198585 172277 10254 83225 125296 413187 67655 161174 0 210994 197122 47391 128066 21413 20661 13881 2992228 0 0 345962 1119677 63079 121556 30647 191483 162449 157267 42156 167646 417455 767358 197986 795498 122879 418270 866753 407308 266412 77225 79020 0 1500930 0 0 107244 531390 287513 170213 64428 118285 209883 364320 199875 146928 440978 518235 381355 720874 403673 315562 893641 386452 229414 253598 106240 45451 16136932 0 22235 0 110365 510612 347228 59303 68606 102756 311146 248028 90595 309620 300493 832988 802826 714884 267295 788190 394375 731914 386804 232673 95055 40045834 0 0 0 162253 638459 237317 79319 94174 137006 265276 445903 177558 338993 248669 623508 1375758 730822 354515 1120577 609358 1661470 279160 118288 204438 24718236 58885 22235 18463 126980 530944 321637 77545 58735 50619 102658 469290 153662 204879 158653 439729 1218315 416792 215506 531053 479157 931172 234313 117886 227355 20088738 0 0 0 45607 309090 247396 139680 62595 60715 136668 406992 195992 181687 247952 380196 1251975 287359 225666 311464 314000 880632 217371 316567 208472 18331540 76121 11011 6106 16723 78328 266632 69887 72147 0 71024 271169 214407 95559 66992 270697 623889 324966 181642 137430 219897 758996 238780 207944 118040 8436142 20143 0 0 16703 79289 204261 157500 8650 46710 62096 77307 105858 108703 125042 189153 464021 129278 106404 190156 169994 224765 238652 97593 120428 19728044 0 0 6106 10625 26490 133186 113756 15487 0 42836 56211 79810 70962 43191 102893 274147 137035 138927 96298 170481 208707 168707 110810 68027 10303146 0 0 23906 0 42201 158039 52799 10698 25805 3198 109682 47256 23940 37193 112164 207776 115126 146271 71441 111148 112521 110777 191255 88143 8644248 0 0 0 16885 0 17036 63325 0 38406 27070 49306 78075 21506 10236 50660 203176 31042 109108 84223 80056 79753 82775 62479 90627 7738250 20143 22235 17254 0 0 0 37992 20034 67089 0 13620 7863 14766 24597 44892 72670 39375 98370 86273 57131 13259 99010 14580 60054 3195752 0 0 17254 5352 15138 0 18996 15487 0 36775 13620 17639 0 0 16171 136756 29156 34142 31873 123245 72807 56555 62090 47411 4319554 39616 0 17254 30272 5247 14798 18996 0 0 0 0 0 24695 14837 27189 185043 18125 9487 22996 25737 14967 38480 0 9869 1995656 0 0 0 16885 0 0 0 0 0 0 25037 9059 0 24648 21612 89875 9347 52772 25308 18311 28587 0 22711 0 3978158 15434 0 0 0 10494 0 0 17299 21710 0 27240 0 14306 13083 15932 52508 0 51553 19628 0 19628 21413 0 33403 1185160 15434 0 0 0 0 16962 0 0 0 0 13620 9059 0 52462 12167 26254 61811 21543 9814 11929 0 12158 14580 36907 4061062 0 0 0 0 0 0 0 0 0 0 13620 0 9307 0 0 27514 2692 10751 3032 0 0 0 0 46184 1995664 0 0 0 0 0 0 18996 0 0 0 0 15018 13096 0 27484 0 2692 10751 0 10662 0 0 14580 0 2030566 0 0 0 0 15853 0 0 19546 0 0 0 0 0 12053 12167 13127 9347 0 9814 5331 42166 0 0 0 2030568 0 0 0 16885 0 0 0 0 0 0 0 0 9307 0 25511 0 0 0 0 19351 0 12249 0 0 070 0 0 0 0 0 0 0 0 0 12568 11417 0 0 0 0 0 0 10223 0 0 0 0 0 0 072 0 0 0 0 5247 0 0 0 0 0 0 0 0 12053 0 0 0 0 11099 0 0 0 0 0 074 0 0 0 0 0 0 0 0 21222 0 26697 0 0 0 9579 0 0 0 0 0 0 15256 14580 0 076 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 078 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 080 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 9347 0 0 0 0 0 0 0 082 0 0 0 0 0 0 0 0 0 0 0 9059 0 0 0 0 0 0 0 0 0 0 0 0 084 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 086 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 088 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 090 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 092 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 094 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0


3.3 Dutch data

3.3.1 DFS (Dutch Demersal Fish Survey)

The Dutch Demersal Fish Survey (DFS) is part of an international inshore survey car-ried out by the Netherlands, England, Belgium and Germany (van Beek et al., 1989). The Dutch survey covers the coastal waters from the southern border of the Nether-lands to Germany, including the Wadden Sea, the outer part of the Eems-Dollard estu-ary, the Westerschelde and the Oosterschelde, Figure 3.3.1. This survey has been carried out in September–October since 1970 (van der Hammen et al., 2013). All anal-yses and results presented below are taken from van der Hammen et al. (2013).

3.3.1.1 Analyses

Data from six distinct areas were analysed: the Dutch Western and Eastern Wadden Sea, the Dutch Wadden Sea coastal zone, the Southern coastal zone, the Oosterschelde and the Westerschelde. Sampling effort has been relatively constant over the years. For each haul, the position, date, time of day, depth and surface water temperature were recorded. The Westerschelde and Wadden Sea are sampled with a 3 m beam trawl, while along the Dutch coast a 6 m beam is used. The beam trawls were rigged with one tickler chain, a bobbin rope, and a fine-meshed codend (20 mm). Fishing is restricted to the tidal channels and gullies deeper than 2 m because of the draught of the research vessel. The combination of low fishing speed (2–3 knots) and fine mesh size results in selection of mainly the smaller species and younger year classes. Sample locations are stratified by depth. The mean abundance per area was calculated weighted by surface area for each depth stratum.

Lengths of sea bass are also measured in the DFS. For the Westerschelde area, the anal-ysis is also done by length class.

3.3.1.2 Results

The analyses resulted in six time-series (Figure 3.3.2), showing that sea bass abundance increased in the last 10–15 years in all areas (Tulp, 2008). However, in most areas, sea bass is not caught frequently (~1/10 000 m2), only in the Westerschelde the DFS catches sea bass more frequently. The DFS abundance index is listed in Table 3.3.1.

The analysis by length class shows high variation in the length distribution per year (Figure 3.3.3), but does not show a clear trend over the years. Sea bass matures at around 41 cm (females) or 34 cm (males), the survey thus catches juveniles (Figure 3.3.2).

The survey relative trend in the Westerschelde (which appears to comprise a mixture of 0-gp, 1-gp and some 2-gp) is similar to the trend in the indices by age in the UK Thames survey (Figure 3.3.4) indicating a growth in recruitment from the mid-1990s to the early 2000s, followed by a decline. The large catch rates in the Dutch survey in 2010 are not reflected in any UK indices. The Westerschelde (the area with the highest den-sities) indices from 2011 to 2013 are close to zero, mirroring the recent trend of very poor recruitment in the UK Solent survey.

3.3.1.3 Conclusions

In the past, sea bass was caught only occasionally by the DFS. Since the beginning of the 2000s sea bass is caught more frequently, especially in the Westerschelde (Figure 3.3.1). Most if not all of the sea bass caught is juvenile. The trend in the Westerschelde


time-series of the DFS survey, which has the largest catch rates, is similar to the trends in the UK Thames survey suggesting growth in recruitment as the sea bass stock ex-panded into the North Sea, and possibly representing recruitment from spawning in the southern North Sea. Recent recruitment in this area appears very weak.

Table 3.3.1. DFS time-series of sea bass abundance (n/10 000 m2).

YEAR WESTERN DUTCH WADDEN SEA

EASTERN DUTCH WADDEN SEA

DUTCH WADDEN COAST

SOUTHERN DUTCH COAST

OOSTERSCHELDE WESTERSCHELDE

1970 0.000 0.000 0.000 0.000 0.077 0.000

1971 0.000 0.000 0.000 0.000 0.000 0.000

1972 0.000 0.000 0.000 0.000 0.000 0.092

1973 0.000 0.000 0.000 0.000 0.000 0.000

1974 0.000 0.000 0.000 0.000 0.000 0.000

1975 0.000 0.000 0.000 0.000 0.000 0.000

1976 0.000 0.000 NA NA 0.000 0.066

1977 0.000 0.000 0.000 0.000 0.000 0.000

1978 0.000 0.000 0.000 0.000 0.000 0.000

1979 0.000 0.000 0.000 0.045 0.000 0.063

1980 0.000 0.000 0.000 0.228 0.000 0.064

1981 0.000 0.000 0.000 0.085 0.000 0.000

1982 0.000 0.000 0.000 0.031 0.000 0.000

1983 0.000 0.000 0.000 0.093 0.000 0.000

1984 0.000 0.000 0.000 0.027 0.000 0.386

1985 0.000 0.000 0.040 0.117 0.000 0.000

1986 0.000 0.000 0.000 0.000 0.000 0.000

1987 0.000 0.000 0.000 0.000 0.523 0.000

1988 0.000 0.000 0.000 0.000 0.000 0.075

1989 0.000 0.000 0.000 0.000 0.000 0.000

1990 0.000 0.000 0.000 0.038 0.000 0.323

1991 0.000 0.000 0.000 0.095 0.000 0.129

1992 0.000 0.000 0.000 0.000 0.143 3.205

1993 0.000 0.000 0.000 0.000 0.000 0.290

1994 0.000 0.000 0.000 0.000 0.000 36.397

1995 0.000 0.000 0.000 0.079 0.000 4.322

1996 0.000 0.000 0.000 0.188 0.129 0.330

1997 0.041 0.000 NA NA 0.000 1.238

1998 0.230 0.000 0.000 0.000 0.048 2.197

1999 0.000 0.000 0.000 0.055 0.185 11.139

2000 0.199 0.000 0.000 0.595 0.805 7.503

2001 0.044 0.074 0.000 0.132 0.551 9.561

2002 1.233 0.661 0.050 0.000 0.572 20.933

2003 0.000 0.000 0.000 1.222 0.000 17.253

2004 5.842 0.449 0.000 0.996 0.000 56.001

2005 0.000 0.000 0.000 0.040 0.608 6.031

2006 0.107 0.137 0.000 0.167 0.135 7.313

2007 0.033 0.000 0.000 0.123 0.117 15.655


YEAR WESTERN DUTCH WADDEN SEA

EASTERN DUTCH WADDEN SEA

DUTCH WADDEN COAST

SOUTHERN DUTCH COAST

OOSTERSCHELDE WESTERSCHELDE

2008 0.041 0.000 0.303 0.065 0.122 6.980

2009 1.170 0.160 0.041 0.275 0.171 7.989

2010 0.075 0.000 0.000 0.000 0.267 48.489

2011 0.246 0.633 0.000 0.000 3.347 0.809

2012 0.189 0.000 0.000 0.000 0.000 0.994

2013 0.000 0.061 0.000 0.000 0.000 1.086


Figure 3.3.1. Overview of sampling locations in the coastal area of the Dutch “DFS survey”. Taken from van Damme et al. (2013).


Figure 3.3.2. Time-series for the density of sea bass from 1970 to 2012. Above: Wadden Sea, middle: coastal zone and below: Oosterschelde and Westerschelde. The black dots are the average densities in numbers per hectare. The blue line shows the trend. The grey areas indicate the upper and lower limit of the 95% confidence intervals (Tulp et al., 2008 and Tulp: personal communication).

0.0

0.2

0.4

0.6

(n/1

0.00

0m2)

1970 1980 1990 2000 2010

01

23

45

6

(n/1

0.00

0m2)

1970 1980 1990 2000 2010

0.00

0.05

0.10

0.15

0.20

0.25

0.30

(n/1

0.00

0m2)

1970 1980 1990 2000 2010

0.0

0.2

0.4

0.6

0.8

1.0

1.2

(n/1

0.00

0m2)

1970 1980 1990 2000 2010

01

23

(n/1

0.00

0m2)

1970 1980 1990 2000 2010

010

2030

4050

60

(n/1

0.00

0m2)

1970 1980 1990 2000 2010


Figure 3.3.3. Number per hectare (nha) per length class and year of sea bass in the DFS survey in Westerschelde area.

Figure 3.3.4. Comparison between the relative abundance trends in the Dutch Westerschelde survey with the indices from the UK Thames survey at ages 0–3. Data are standardised to the mean for 1994–2009. The Dutch survey includes several early year classes but is referenced to the year of the survey.

length (cm)

nha

051015

4 5 6 7 8 9 10 11 12 1314 15 16 17 18 1920 21 22 23 2425

2001

4 5 6 7 8 9 10 11 12 13 14 1516 17 18 19 2021 22 23 24 25

2012

2000

051015

2011051015

1999 2010

1998

051015

2009051015

1997 2008

1996

051015

2007051015

1995 2006

1994

051015

2005051015

1993 2004

1992

051015

2003051015

1991 2002


4 Stock assessment

4.1 Base case WGCSE 2013 assessment-starting point for IBPBass

WGCSE 2013 conducted an assessment using Stock Synthesis 3 (SS3) (Methot, 2010). The software used was Stock synthesis v3.23b (Methot, 2011), according to the Stock Annex developed by ICES IBPNEW 2012 with inclusion of fishery data for 2012. The IBPNEW 2012 benchmark assessment required a modelling framework capable of han-dling a mixture of age and length data for fisheries and surveys (fleet-based landings; landings age or length compositions, age-based survey indices for young bass) and biological information on growth rates and maturity. Landings-at-age were available for four UK fleets from 1985 onwards, whereas French fleets had length composition data that were available only since the 2000s. The Stock Synthesis assessment model was chosen, primarily for its highly flexible statistical model framework allowing the building of simple to complex models using a mix of data compositions available. The model is written in ADMB (www.admb-project.org), is forward simulating and avail-able at the NOAA toolbox: http://nft.nefsc.noaa.gov/SS3.html.

A mixed age–length model was fitted as the base case, with a length-only model for comparison. Some adjustments were made by WGCSE 2013 to the model: i) UK fishery compositions for 2012 were input to the age/length model as length compositions be-cause age compositions were not available; ii) the UK midwater trawl series was re-duced to 1996 onwards and was input as length compositions because unusual length-based selection curve parameters were obtained when inputting the data as age com-positions; iii) recruit deviations were estimated back to 1965.

IBPBass addressed the following recommendations of WGCSE 2013 for developing the assessment during the inter-benchmark meeting. Work completed is indicated in pa-rentheses:

• Source and review information on historical catches and develop plausible scenarios including over the 20+ year burn-in period for the assessment [some investigations were pursued in France but yielded no clear information on pre-1985 landings].

• Review the derivation and quality of historical fishery length/age composi-tion data [not done beyond the information on sampling intensity and coverage already available].

• Expand UK fishery age compositions to all true ages [done; see below]. • Rationalise the fleet definitions, and reduce to the minimum sufficient to

provide robust SS3 stock trends [done; see below]. • Source and evaluate candidate lpue or effort series for tuning abundance or

fishing mortality on older ages [fishery dependent abundance indices were not considered other than some information presented in Section 2 on lpue of fleets in the Netherlands].

• Collate and evaluate other survey data on bass abundance that could be in-corporated in the model [French Channel Groundfish Survey was evaluated and incorporated in the assessment].

• Determine the most robust approach to incorporating mean length-at-age and length-at-age distributions in SS3 [Not done].

• Investigate potential biases in using combined-sex growth parameters [Not done].

http://nft.nefsc.noaa.gov/SS3.html


• Further explore the sensitivity of the assessment to decisions on model structure and inputs [See model development and sensitivity analyses carried out below].

• Consider if simpler assessment approaches are warranted [IBPBass focused exclusively on Stock Synthesis to try to make best use of all available data].

The structure and input data / parameters of the WGCSE base case age and length model are summarized below:

Model structure

• Temporal unit: annual based data (landings, survey indices, age frequency and length frequency);

• Spatial structure: One area; • Sex: Both sexes combined.

Fleet definition

Six fleets were defined: 1. UK bottom trawls, 2. UK midwater pair trawls; 3. UK fixed and driftnets; 4. UK lines; 5. French fleets (combined); 6. Other (other countries and other UK fleets combined).

Landed catches

Annual landings in tonnes from 1985 to 2012 for the six fleets from ICES Subdivisions IVb and c, VIIa, d–h were used in the assessment. French data were as provided by Ifremer.

Abundance indices

Ten independent abundance index series were defined, each being a single age group (up to four years old) from one of the three available trawl surveys on juvenile bass. They are treated as ten independent surveys (following a recommendation from R. Methot) to circumvent difficulties in estimating selectivity parameters for each survey series comprising only 3–4 young age groups, although this approach loses covariance information due to year-effects in each survey.

• Spring Solent survey in ICES Subdivision VIId covering ages 2 to 4 for years 1985 to 2009;

• Autumn Solent survey in ICES Subdivision VIId covering ages 2 to 4 for years 1986 to 2011;

• Autumn/Winter Thames survey ICES Subdivision IVc covering ages 0 to 3 for years 1997 to 2009.

Fishery landings age composition data for age–length model

Age bins for this model were set at 0 to 11 with a plus group for ages 12 and over, because historical UK data had been compiled that way. Age compositions for UK trawlers, netters and liners for 1985–2011 were expressed as fleet-raised numbers-at-age, although they are treated as relative compositions in SS3.

As UK age data for 2012 were not available in time for WGCSE, the length composi-tions for the three fleets in 2012 were included in the model.


Length composition data for age–length model

The length bin was set from 4 to 100 cm by 2 cm intervals. Length compositions for the following fishing fleets were used:

• UK otter trawl, nets and lines (three fleets): 2012 only (age data for previous years);

• UK midwater pair trawls: 1996–2012 data. Sampling of midwater pair trawl landings has been low and patchy over time, and some years have no sam-ples. Some of the midwater trawl data during the 1985–1995 period, before the fishery became established, were from very small sample sizes and ex-hibited unusual length and age compositions. All length and age data for this fleet during 1985–1995 were excluded from the model. This was a change from the IBPNEW approach, but resulted in <1% change to annual SSB estimates.

• French all fleets combined: 2000 to 2012 data.

Model assumptions and parameters

The following table summarises key model assumptions and parameters. Other pa-rameter values and input data characteristics are defined in the SS3 control file BassIVVII.ctl, the forecast file Forecast.SS and the data file BassIVVII.dat as used by WGCSE 2013. Changes from IBPNEW settings are indicated.


Table 4.1. Key model assumptions and parameters from the WGCSE 2013 final run.

CHARACTERISTIC SETTINGS

Starting year 1985

Ending year 2012 (IBPNEW: 2010)

Equilibrium catch for starting year 0.82* 1985 landings

Number of areas 1

Number of seasons 1

Number of fishing fleets 6

Number of surveys (recruit surveys) three surveys, modelled as 10 single-age fleets at ages 0–4

Individual growth von Bertalanffy, parameters fixed, combined sex

Number of estimated parameters 62 (includes additional years and main recruit deviations back to 1965)

Population characteristics

Maximum age 30

Genders 1

Population length bins 4–100, 2 cm bins

Ages for summary total biomass 0–12+

Data characteristics

Data length bins (for length structured fleets) 14–94, 2 cm bins

Data age bins (for age structured fleets) 0–12+

Minimum age for growth model 0 [age 2 for age–length model]

Maximum age for growth model 30

Maturity Logistic 2-parameter – females; L50 = 40.65cm

Fishery characteristics

Fishery timing -1 (whole year)

Fishing mortality method Hybrid

Maximum F 2.9

Fleet 1: UK Trawl selectivity Asymptotic length based

Fleet 2: UK Midwater trawl selectivity Asymptotic length based

Fleet 3: UK Nets selectivity Asymptotic length based

Fleet 4: UK Lines selectivity Asymptotic length based

Fleet 5: Combined French fleet selectivity Asymptotic length based

Survey characteristics

Solent spring survey timing (yr) 0.42

Solent autumn survey timing (yr) 0.83

Thames survey timing (yr) 0.75

Catchabilities (all surveys) Analytical solution

Survey selectivities [all survey data entered as single ages; sel = 1]

Fixed biological characteristics

Natural mortality 0.2

Beverton–Holt steepness 0.999

Recruitment variability (σR) 0.9

Weight–length coefficient 0.00001296

Weight–length exponent 2.969



Maturity inflection (L50%) 40.649 cm

Maturity slope -0.33349

Length-at-age Amin 19.6 cm at Amin=21

Length-at-Amax 80.26 cm

von Bertalanffy k 0.09699

von Bertalanffy Linf 84.55 cm

von Bertalanffy t0 -0.730 yr

Std. Deviation length-at-age (cm) SD = 0.1166 * age + 3.5609

Other model settings

First year for main recruitment deviations for burn-in period

1965 (IBPNEW: 1980)

1 as recommended by R. Methot after scrutinizing earlier SS3 runs during IBPNEW 2012, and used by IBPNEW and WGCSE. The WGCSE 2012 tabulated the original value of 5.78 cm at-age 0 in error.

4.2 Development of the SS3 model at IBPBass

4.2.1 New data available

The following new data were available for use in model development:

• UK fishery age composition data for 2012; • Historical French landings data by métier groups; • French fleet length compositions for the four métier groups: otter trawls,

midwater pair trawls, nets and lines, by area (IVbc; VIId; VIIeh; VIIafg) for 2009–2012 (not all area-year combinations available for each gear).

• Solent survey indices for 2013 (no survey in 2012). • Length-based abundance indices for the Channel Ground Fish survey in Di-

vision VIId (1988–2012).

4.2.2 Stages in model development

The WGCSE base-case SS3 model (Section 4.1) was developed in incremental stages adding new data and exploring different model configurations and fleet groupings. At each stage the changes in model diagnostics were tabulated, plotted and evaluated. Results were progressively presented over four WebEx meetings at which agreement was reached on models to take forward for the next stage of development. Annex 3.1 describes the successive runs carried out, and gives the likelihood components, num-ber of active parameters and convergence level for each run. Annex 3.2 shows the SSB plots (with +/- asymptotic confidence intervals) for all runs.

4.2.2.1 Expansion of input age compositions (runs 1–3)

Expanding the UK fleet age–frequency distributions from a 12 plus group to 16+, 18+ or 20+ provides information on more year classes in the fishery data during the 1980s and early 1990s, and hence improves the estimates of recruit deviations prior to 1985 (Figure 4.2.1). However, expanding to a higher plus group can result in more zeroes in the age structure for the oldest true ages, noticeable mainly for weak year classes. IBPBass decided to expand only to a 16+ group to benefit the estimation of early recruit deviations while reducing the likelihood of incorporating zeroes for the older age groups.


Figure 4.2.1. Plus-group sensitivity analysis: log recruitment deviations with 95% confidence inter-val for 12+, 16+, 18+ and 20+ groups. Right hand plots show stock trends for the four plus groups (lowest recent F is from the run with 20+ gp).

4.2.2.2 Incorporating UK fleet age compositions for 2012 and UK midwater trawl age compositions for 1996–2012 (runs 4–5)

The WGCSE 2013 assessment had only length compositions for UK fleets in 2012, and had replaced the UK midwater trawl age compositions with length frequencies due to an unusual length-based selectivity pattern. The main effect of adding in the 2012 age compositions for otter trawl, nets and lines (run 4), using the run 1 model with 16+ group, is to increase the SSB in the final years, and, with the additional age information, generate a less smooth trend in recruitment (Figure 4.2.2).

The model is insensitive to changing the input data from length compositions to age compositions for UK midwater trawls, while fitting a length-based selection model for both (Figure 4.2.2). However the fit to the age composition for the midwater trawl was better than the fit to the length compositions. The low influence on stock trends prob-ably reflects the small contribution of this fleet to total landings.


Figure 4.2.2. Left plots: Runs 4 and 5: Effect on assessment of adding the 2012 UK age frequencies (run 4) and replacing UK midwater trawl length compositions with age compositions for 1996–2012 (run 5). Dashed line is the equivalent run without these changes (run 1). Right plots: Runs 6 and 7: effect of changing the effective sample sizes (ESS) for UK fleets to 50 in all years (run 6), and re-ducing the ESS for 1985–1990 UK data to 10 and changing the ESS for French fleet to 200 (run 7). Dashed line is run 5 which had the original ESS from the WGCSE 2013 run.

4.2.2.3 Changing the input effective sample sizes for fleet age compositions (runs 6–7)

The IBPNEW (2012) and WGCSE (2013) assessments included input effective sample sizes (ESS) that were based on numbers of length and age samples collected each year for the component fleets (these are documented in the reports). Following some initial runs, these had been adjusted to reduce input ESS for years where the model fits were clearly poor while sample sizes were low, a process that inevitably resulted in an im-proved correlation between input and output ESS. At IBPBass, the stock assessors were concerned that this iterative approach may have been inappropriate, and that the widely divergent input ESS that resulted within and between UK fleets was unreason-able given poor knowledge of how representative the sampling has been over the years. The ESS for IBPBass were therefore set to a constant value of 50 for each of the four UK fleets (Run 6). The effect of this was to increase the age composition likelihoods from 313 to 649 for the UK fleets, and SSB in the early part of the series was decreased and F increased (Figure 4.2.2, right hand plots). In view of the low sample numbers at times for the UK fleets up to 1990, a further run (7) was carried out reducing the ESS to


10 for those years, while also adjusting the ESS for the combined French fleet from values that varied between 82 and 300 to a constant 200. This caused the estimates of SSB to be reduced slightly from the 1990s onwards. The UK fleet likelihoods were re-duced from 649 to 424, while the French length composition likelihoods were almost unchanged. IBPBass concluded that the model is sensitive to the information provided to the model about the precision of the UK age compositions, but the overall relative trend in estimates of biomass and F from the early 1990s was relatively unchanged. Run 7 was taken forward for further development.

4.2.2.4 Selectivity of UK fleets (run 9)

An empirical analysis of the UK fishery age compositions shows that the landings-at-age for UK trawls and nets are more heavily weighted towards younger bass than in the lines and midwater pair trawl fishery. The comparative selectivity of the fleets was investigated outside of SS3 by treating the combined age compositions for 1985 on-wards in each fleet as a “pseudo cohort”, and carrying out a simple cohort analysis with a range of terminal F’s on the oldest true age (11). Figure 4.2. 3 shows the results of tuning the terminal F so that the partial F’s for the lines and midwater trawl approx-imate an asymptotic pattern. In this configuration, the partial F’s for trawls and nets show a very similar, domed pattern. Note that this reflects landings only; the larger discard quantities in the trawl fishery would broaden the dome towards younger fish.

Figure 4.2.3. Empirical selectivity curves for UK fleets based on a cohort analysis on a “pseudo co-hort” comprising the combined age compositions for 1985–2011 for each fleet, tuned to give approx-imately asymptotic selectivity for lines and midwater trawl.

An attempt to replicate this in SS3 using double-normal length-based selectivity ap-plied to UK trawls and lines, using Run 7 for other inputs and settings, led to an im-proved fit to the UK trawl and lines age compositions (Run 9). However, the model fitted selection curves for these two fleets that appeared to have unrealistically narrow dome-width, falling to low selection probability by 45–50 cm. Domed selection curves for these fleets were therefore set aside (temporarily) and were reintroduced later in the model building process as age-based selection curves, which appeared more real-istic.

A change in MLS from 32 cm to 36 cm occurred in 1990 and was considered by Pickett et al. (1995) to have been followed by an increase in discarding. This is expected alt-hough observations are confounded by the impact of the very strong 1989 year class on fishery length compositions. IBPBass did not attempt to estimate separate selectivity


parameters for the 1985–1989 period, but the much lower input effective sample sizes for 1985–1990 composition data reduced the weighting of data from this period in the estimation of selectivity parameters applicable to the 1990–2012 period.

4.2.2.5 Combining French and UK fleets (run 8)

An important objective of IBPBass was to try to derive a more parsimonious model with fewer commercial fleets and their associated selectivity models, particularly since the previous model had four individual UK fleets (some contributing only a small amount to total international landings) while similar métier groups were lumped to-gether into a single combined French fleet which has most of the total landings.

The first approach adopted was to set up four combined UK-French fleets comprising: i) bottom otter trawlers (gear code OTB); ii) midwater pairtrawlers (PTB); iii) fixed/driftnets; and iv) hooks and lines. IBPBass was provided with a time-series of French landings in these gear groupings (see Section 2). A major problem was however encountered in developing a consistent and sufficient series of age-compositions for the four combined fleets. While age compositions were available for UK fleets in these gear groupings, France was only able to provide length compositions for individual gears from 2009 onwards, and incompletely covering all the fleet-year combinations. Two options were explored:

i ) Include only the UK age compositions to characterize the composition of the combined UK-France landings;

ii ) Convert the available French LFDs to AFDs using UK age–length keys, and combine these with the UK age compositions.

A range of models was explored, based on the four combined UK-French fleets and the “other” fleet which included other UK gears (minor landings) and other countries, pri-marily the Belgian and Dutch fleets whose landings increased in the 2000s following the expansion of the bass stock into the North Sea. The IBPBass stock assessors were however concerned that age compositions from UK bottom trawl, net and lines fleets operating mostly along the coasts of England and Wales were being used to infer se-lectivity of combined UK-French fleets, particularly trawls, for which the bulk of catch was taken by French vessels. To address this issue, a comparison was made between the raised fleet length compositions of landings provided by France for the years 2009 onwards, and the equivalent UK length compositions for the same gears in the same ICES Divisions and years (Annex 2: Working Document 02). The mean length of sea bass in otter trawlers was consistently lower in the UK fleet than in the French fleet, whereas the compositions were very similar for midwater pair trawls. Both national fleets showed the broadest length composition in Divisions VIIe,h, which is closest to the main sea bass spawning sites in the western Channel and Approaches. This com-parison is weak due to small sample sizes. IBPBass decided that, on balance, any ben-efits of combining the UK and French fleets were outweighed by the loss of length composition data for the combined fleets from 2000 onwards, the uncertainties regard-ing the comparative selectivity of UK and French fleets (other than midwater pair trawl vessels targeting spawning aggregations), and the greater uncertainty in deciding which fleet selectivity to apply to the landings from Belgium and the Netherlands, the landings of which have increased rapidly as the stock expanded further into the North Sea in the 2000s. Further model development therefore focused on combining UK fleets to reduce the number of selectivity models being fitted.


4.2.2.6 Inclusion of Channel Groundfish Survey (run 11b)

The French Channel Groundfish Survey (CGFS) was included in the model as a series of combined-length swept-area indices of abundance (with relative standard error), and length composition data for each year from 1988–2012 (with information on num-bers of stations with sea bass, which is used to set the input effective sample sizes for the composition data) (see Section 3, Table 3.2.1). The selectivity curve for the survey was investigated initially by examining the ratio of survey numbers-at-length to pop-ulation numbers-at-length from the model runs excluding this series. The selectivity appeared to be domed, and input parameters for a 6-parameter double normal rela-tionship were derived using the SS Excel spreadsheet developed to guide the estima-tion of input selectivity parameters. The survey was added to the model configuration for Run 7 (all commercial fleets with asymptotic selectivity).

Initial runs indicated that the length frequencies were very poorly fitted. This appeared to be due to using the original relative standard errors supplied with the data, some of which are unusually small. The effect of this was to give very high weighting to some individual years, causing the model to converge on selectivity patterns that fitted those years. The relative standard errors were therefore input as a value of 0.30 for all years, around double the average for the series. This was justified by: a) there are many sta-tions with zero sea bass catch which will lead to some bias in variance estimation (a delta-lognormal approach may be more useful in this case), and b) the accuracy of a survey series should also reflect unpredictable year-effects due to factors such as weather affecting gear performance. The model with input relative standard error of 0.30 provided a much better fit to the data. An RSE of 0.6 was applied to the first three years due to the very small number of stations with sea bass.

Initial runs showed a poor fit to the 1988–1990 length compositions, these years having much smaller proportion of stations with sea bass than in subsequent years. The com-position data for these three years were therefore omitted although the overall abun-dance indices were retained. Some diagnostics of the model fit to the CGFS data are given in Table 4.2.1 and Figure 4.2.4. It is first of all noted that the trends in the raw survey data, both in terms of the abundance index and also in the proportion of stations that have a sea bass catch, follow loosely the trends in total biomass given by the WGCSE 2013 assessment (see Figure 3.2.2). The fitted selectivity parameters are mostly close to the input parameters (Table 4.2.1), which may reflect the empirical method of deriving the inputs as explained above, or may indicate that the length compositions are not informative for estimating selectivity. The fitted selectivity curve is strongly domed, peaking at 33 cm and declining to a selection probability of 0.3 by around 55 cm (Figure 4.2.4a).

Table 4.2.1. Input and output values for selectivity parameters for the French Channel Groundfish Survey. “Value” = fitted parameter.

The overall CGFS abundance index has some strong apparent year effects such as in 2003 when all sizes of fish increased in abundance in the survey. The abundance indices

Label Value Active_Cnt Phase Min Max Init Status Parm_StDev PR_typeSizeSel_17P_1_CGFS1 32.7578 64 2 20 94.1 32 OK 0.727918 Sym_BetaSizeSel_17P_2_CGFS1 -4.78793 65 3 -6 4 -6 OK 3.2102 Sym_BetaSizeSel_17P_3_CGFS1 3.21198 66 3 1 5 3.3 OK 0.215352 Sym_BetaSizeSel_17P_4_CGFS1 4.5596 67 3 3 6 4.4 OK 0.481568 Sym_BetaSizeSel_17P_5_CGFS1 -7.77741 68 2 -8 9 -8 OK 0.891579 Sym_BetaSizeSel_17P_6_CGFS1 -0.8706 69 2 -5 9 -1 OK 0.294689 Sym_Beta


track the SS3 trend of increasing abundance from the mid-1990s and the decline after around 2007, but fall below the SS3 fitted trend in 1994 and 1995 (Figure 4.2.4b). The fit to the annual length compositions appears reasonable (Figure 4.2.4c). The input and output effective sample sizes are loosely correlated suggesting the model is correctly reflecting the effect of numbers of samples on the precision of the length compositions. The residuals have some tendency for diagonal patterns suggesting year-class effects (Figure 4.2.4d), for example around the time of the strong 1989 year class.

The inclusion of the survey has a surprisingly negligible effect on the estimated trends in SSB, recruitment and fishing mortality (Figure 4.2.5). However, despite the addi-tional six parameters to be estimated, the precision of the SSB estimates for recent years is improved (Table 4.2.2).

Table 4.2.2. SS3 estimates of SSB (and relative standard error, RSE) from Run 7 (excluding the CGFS) and run 11b (including the CGFS).

RUN 7 RUN 11B

Year SSB RSE SSB RSE

2009 8787 0.064 8943 0.056

2010 8855 0.084 9029 0.071

2011 8146 0.122 8275 0.099

2012 7480 0.168 7520 0.134

IBPBass concluded that the CGFS is a valuable addition to the assessment, for the fol-lowing reasons: a) general agreement in overall stock trends between the CGFS raw data and the SS3 fit; b) reasonably good fit to the length data and c) coverage of the full size range of the stock (although with domed selectivity), and coverage of a larger part of the range of the stock than the Solent survey of juvenile sea bass (though still only covering VIId).


(a) (b)

(c) (d)

Figure 4.2.4. Run 11b: Results of including the Channel Groundfish Survey into the SS3 Run 7 configuration. (a) fitted length-selectivity curve; (b) fit to total abundance indices; (c) fit to annual survey length compositions; (d) residuals of length composition fit. All commercial fleet selectivi-ties are length-based asymptotic.


Figure 4.2.5. SS3 estimates from Run 11b, a replication of Run 7, but including the Channel Ground-fish Survey.

4.2.2.7 Comparison of length-based and age-based selectivity for UK fleets

All runs reported so far have modelled the selectivity of UK fleets as a function of length, despite the use of age–frequency distributions to characterize the composition of the fleet landings. This approach was used because selectivity was assumed to be principally a function of fish length. However, the length-based selectivity for the UK trawls, nets and lines tend to be very steep, almost knife-edged (Figure 4.2.6 from Run 11b). A run was made (Run 11c) replicating Run 11b, but fitting age-based selectivity curves. The resultant age-based selectivity curves are shown in Figure 4.2.7 compared with the empirical estimates given in Section 4.2.2.4, and the age-based selectivity curves derived by the SS3 algorithms from the length-based selectivity curves in Run 11b (values given as “Asel2” where Asel2_is_sizesel*size_at_age(ALK). The fitted age-based selectivity curves in Run 11c are close to the empirical ones for the ascending limb in otter trawls and nets, and overall for lines and midwater trawls. The inferred age selectivity curves in Run 11b differ substantially from these. IBPBass concluded that for the UK fleets, selectivity should be estimated directly from the age composition data. All subsequent runs fitted age-based selectivity. This model configuration results in larger SSB and recruitment estimates, and correspondingly smaller F estimates throughout the series, but does not change the relative trends (Figure 4.2.8).


Figure 4.2.6. Length-based selectivity curves estimated for UK and French commercial fleets in Run 11b.

Figure 4.2.7. Age-based selectivity curves estimated for UK commercial fleets in Run 11c (asymp-totic age-based selectivity fitted), compared with the age-based selectivity estimated empirically from the raw catch-at-age data, and the age-based selectivity derived by SS3 in Run 11b by combin-ing the length-based selectivity ogives and length-at-age.


Figure 4.2.8. Comparison of stock trends from Run 11b (length-based asymptotic selectivity for UK fleets) and Run 11c (age-based asymptotic selectivity curves estimated UK fleets).

4.2.2.8 Removal of UK Spring Solent survey and Thames survey data

Previous runs have shown that the Thames survey and Solent spring survey fit poorly in the model.

The Thames survey provides an indicator of recruitment in the southern North Sea, and gives trends similar to those from beam trawl surveys in Dutch coastal waters (see Section 3). Recruitment in these areas appears to have increased as the stock expanded into the North Sea during the 1990s and 2000s, possibly in response to establishment of spawning in the southern North Sea. Hence, the surveys in the southern North Sea may not be a good indicator of the overall trends in recruitment in the combined Area IV and VII. These series would be more useful in the event of an area-disaggregated assessment.

The Solent Spring survey indicates a longer period of very low recruitment in recent years than given either by the autumn survey or the fishery composition data. Also,


some unusual survey calibration factors were noted in the early part of the spring So-lent survey series that have not been resolved and require investigation.

Both surveys were removed from the assessment on a priori grounds. Comparative runs of SS3 with and without the surveys indicated negligible differences in stock trends.

4.2.2.9 Combining UK fleets (runs 16a,b,c; 18a,b,c;)

Earlier SS3 runs kept UK otter trawls, midwater trawls, nets and lines as separate fleets for estimation of selectivity, mainly to allow a direct evaluation of the selectivity pa-rameters in relation to possible technical conservation measures. However IBPBass noted that within a year and region, a common age–length key was applied to length compositions from each of the fleets. Given the large overlap in length-at-age distribu-tions, this is likely to introduce a large correlation in errors between the fleets leading to a spurious increase in precision of stock estimates. This effect is likely to be greatest in trawls, nets and lines which are used throughout the stock area, and less so for mid-water pair trawls which operate mainly in one area (VIIe,h) and in one season. IBPBass therefore decided on a priori grounds that the age compositions and landings of UK otter trawls, nets and lines should be combined in a single fleet.

Two runs were carried out (16a, b) in which the combined UK otter trawl / nets / lines fleet age compositions were fitted assuming a six-parameter double-normal selectivity curve (Run 16a) and an asymptotic selectivity curve (Run 16b). A third run, 16c, had double-normal selectivity for the UK combined fleet and the French fleet, leaving only the UK midwater trawl fleet with asymptotic selectivity.

Note that, at this point, the additional survey index data from the UK Solent autumn survey in 2013 were incorporated (no survey was done in 2012). This has no effect on the fitted biomass and fishing mortality trends, but does affect the most recent recruit-ment which has yet to feed into the fishery landings and spawning stock.

Runs 11c and 16b directly compare two assessments in which the UK otter trawls, nets and lines are treated either as separate (Run 11a) or combined (16b) and with asymp-totic age-based selectivity. Run 16a is equivalent to 16b, but with double-normal selec-tivity for the combined UK otter trawl, nets and lines fleet. Run 11c adds a double-normal selectivity to the French combined fleet (also mirrored in the “other” fleet). These different model configurations have little impact on the relative trends over time, but double-normal selectivity scales the biomass up and fishing mortality down, with little effect on recruitment (Figure 4.2.9). The fitted length-based selection for the French fleet in Run 16c has a very broad plateau (Figure 4.2.9).

IBPBass concluded that the choice of asymptotic or double-normal selectivity curve for the combined UK fleets makes little difference to relative stock trends, but more realis-tically represents the selectivity derived empirically from the raw catch-at-age data. Fitting a double-normal selectivity to the French fleet data is less defensible without investigation of individual catch-at-age data for the component fleets, and also consid-ering the large component of the fishery comprising midwater pair trawling and lines.

All further runs were based on Run 16a, with double-normal age selectivity for UK combined otter trawl, nets and lines; asymptotic age-based selectivity for UK midwater trawls, and asymptotic length-based selectivity for French fleets.


Figure 4.2.9. Comparison of stock trends from Runs 16a–c (combined UK otter trawl, nets and lines, with age-based selectivity) and Run 11c (separate UK fleets with asymptotic age-based selectivity). Run 16a: UK otter/nets/lines fleet with double-normal selectivity (shown top right); 16b: UK otter, nets/lines with asymptotic selectivity; 16c: as 16a but with double-normal length-based selectivity fitted to the combined French fleet (shown bottom right).

4.2.2.10 Increasing the Effective Sample Size for UK trawls, nets and lines (Run 18a)

Runs so far have allocated an Effective sample size of 50 (10 up to 1990) for the com-bined UK trawls, nets and lines. This was the same value as for the individual fleets in earlier runs. However, although the same ALK has been used for each gear type and region, the length compositions are independent. Therefore, a run was conducted dou-bling the ESS for this combined gear group to 100 (20 up to 1990). This resulted in a slight change in the SSB and F trends (Figure 4.2.10) and an increase in total negative log likelihood from 471 to 537. However the relative standard error of SSB in 2012 was reduced from 0.173 to 0.164. This change was retained for subsequent runs.


Figure 4.2.10. Comparison of runs 16a and 18a. In 18a the effective sample sizes for UK combined trawls, nets and lines are doubled from 50 (10 up to 1990) to 100 (20 up to 1990).

4.2.2.11 Parameter correlations: Run 18a

The parameter correlation matrix for Run 18a was examined to see if any selectivity parameters in particular were correlated in a way that may allow some to be fixed, reducing the number of parameters. The following correlations with R >0.5 were noted (“Initial F 4” is the initial equilibrium fishing mortality for fleet 4 and is not a selectivity parameter):

Table 4.2.3. Parameter correlations >0.5 involving selectivity parameters; Run 18a.

PARAMETER 1 PARAMETER 2 R

Initial F 4 France 2 0.76

France 2 CGFS 3 0.93

France 1 CGFS 4 -0.69

CGFS 3 CGFS 6 -0.57

CGFS 6 UKOTB-Nets-Lines 3 0.89

UK MWT 2 UKOTB-Nets-Lines 6 0.7


The only within-survey correlation is for the Channel Groundfish Survey parameters 3 (width of ascending limb) and 6 (the selectivity that the descending limb reaches at the oldest ages), with an R of -0.57. IBPBass considered that there was no clear evidence for fixing either of these parameters.

It is noted in Section 4.2.2.13 that a run with M=0.3 and steepness = 0.8 caused the OTB/nets/lines selectivity parameter 6 to change suddenly to give a more-or-less as-ymptotic pattern. To avoid this issue occurring in sensitivity runs around the final as-sessment model, this parameter was fixed for those runs at a value equivalent to the one estimated for the final run.

4.2.2.12 Setting priors or soft bounds for all parameters (runs 18b, 18c)

In all runs so far, the input selectivity parameters were input based on trials using the selectivity spreadsheet supplied for use with SS3. The inputs included priors (hard bounds). This resulted in the model fitting a mixture of hard and soft bounds for dif-ferent parameters. IBPBass considered that the use of soft bounds throughout was to be preferred, unless it can be proven that the posterior values differ from the priors in a way that suggests the data are informative. Two runs were conducted using Run 16a as the base case, one setting soft bounds for all parameters (Run 18b) and one setting priors for all the parameters (Run 18c). These changes made negligible differences to the time-series of SSB, recruitment and F. On the basis of the recommendations from the IBPBass group members and external reviewer, the use of soft bounds throughout (Run 18a) was carried forward for the next stage in model development.

4.2.2.13 Sensitivity to choice of natural mortality M and stock–recruit steepness (Runs 20a-x)

The value (or vector) of natural mortality used in an analytical assessment is a key parameter determining the estimated productivity, abundance and MSY or other ref-erence points (RPs). The assumed shape of the stock–recruit curve acts with the as-sumed M to constrain any possible biological reference points (Mangel et al., 2013). A key parameter defining a stock–recruit curve is steepness, defined as the ratio of re-cruitment from an unfished population to recruitment when the spawning–stock bio-mass is at 20% of the unfished level. Mangel et al. concluded that “there is much work to be done to resolve the difficulties caused by the linkage among steepness, life-history parameters, and RPs.”

All model runs presented so far have fixed steepness at 0.999, and have fixed the nat-ural mortality rate at 0.2 for all years and ages, based on a range of published methods for inferring M from life-history parameters and oldest observed ages. The different values of M derived for sea bass by IBPNEW (2012), based on a Working Document by Armstrong and Walmsley (2012), are given in Table 4.2.4. The inferred values of M, with the exception of the Beverton method, are in the range 0.15–0.22. Armstrong and Walmsley (2012) recommended that the sea bass assessment should use M=0.18 as the baseline M, and to examine the sensitivity of the assessment to M= 0.15 and M = 0.22. The final WGNEW assessment adopted a value of 0.2.

The influence of M and steepness were investigated using Run 16a as the baseline. Runs were carried out for M values of 0.15, 0.20 and 0.25, and for a range of steepness values from 0.5 to 0.999 (Runs 20a-l). The total of negative log likelihood was compiled for the range of combinations, along with the SSB depletion in 2013 from the virgin SSB, and the relative standard error of the SSB in that year.


Table 4.2.4. Inferences on sea bass M based on some life-history models in the literature (Armstrong and Walmsley, 2012).

Total negative log likelihood tended to be lowest at steepness values approaching unity, with the greatest tendency at low values of M, and likelihood also decreased with increasing M (Figure 4.2.11; left). Despite the lower value of negative log likeli-hood, the relative standard errors of the SSB estimates for 2012 (from the inverse Hes-sian) increased with increasing M but were almost unaffected by steepness (RSE values at steepness 0.999 were 0.158 at M=0.15; 0.164 at M=0.20; 0.175 at M=0.25 and 0.199 at M=0.30). The unusual value at steepness 0.8, M=0.3 in Figure 4.2.10 was a result of the selectivity curve for the combined UK trawls, nets and lines flipping from a domed to an asymptotic pattern, suggesting some instability in allowing all six parameters to be estimated (see later). Otherwise, the values in Figure 4.2.10 show smooth relationships with input M and steepness.

The depletion of SSB in 2013 compared with the virgin SSB was progressively lower as M was increased (Figure 4.2.11, right), but was far less sensitive to steepness.

Recruitment in sea bass has varied widely in response to environmental factors includ-ing conditions in the estuarine and other inshore nursery habitats. There is almost no information to indicate declining recruitment at lower SSB and to discern the true value of steepness. A wide range of values appears plausible (Figure 4.2.12).


Figure 4. 2. 11. (left): Total negative log likelihood from variants of Run 16a with different fixed input values of natural mortality M and stock–recruit steepness; (right): relationship between SSB depletion in 2013 and input values of M and steepness.

Figure 4. 2. 12. Stock–recruit estimates from Run 16a variants with M=0.20 and steepness values fixed at 0.5, 0.8 or 0.999.

4.2.2.14 Incorporating information on recreational fishery catches (run 22)

Recent surveys in France, England and the Netherlands have provided estimates of recreational fishery removals that are around a third of the commercial fishery land-ings in each country (see Section 2). Taking possible hooking mortality of released fish into account, the total recreational removals from the stock may be as high as 1500 t. The relative standard error of this estimate is likely to be in the range 0.2–0.3 based on the information available on the component surveys. With only one combined survey


estimate available, IBPBass could see no easy way to incorporate this directly into Stock Synthesis.

However, by not representing recreational fishery removals in the assessment, the es-timates of total mortality derived from the commercial fishery age and length compo-sition data are attributed only to natural mortality and to commercial fishing mortality. In reality, part of the observed total mortality is attributable to the unaccounted for recreational fishery removals. This becomes a problem when forecasting stock size and yields based on multipliers applied to the apparent commercial fishing mortality, when in fact only part of the apparent F is due to the commercial fleet.

IBPBass considered that one way to evaluate the recreational fishery F is to include a fixed vector of F-at-age, representing the selectivity of the recreational fishery, and to add this to the fixed natural mortality vector. The recreational F vector would then be scaled iteratively until the expected recreational removals in 2012 are 1500 t; the recent total recreational removals estimate from the surveys in France, England and the Neth-erlands. This is equivalent to treating recreational fishers as a predator whose historical abundance and predation is largely unknown, but are assumed to impose a constant mortality in exactly the same was as all other predators are subsumed into a constant M vector.

There is some evidence from a series of surveys in England and Wales since 1970 that the number of people going sea angling has fluctuated without obvious increasing or decreasing trend (Table 4.2.5; Figure 4.2.13). Part of the observed variability will relate to differences in survey methodology, but all are based on some form of representative sampling of the population. Sea bass has been a prized target for recreational sea an-glers in England and Wales (and southern Ireland) over a much longer period than the current assessment, and sea bass angling was developed to a high level of technical skill and knowledge of the species as far back as the 1970s. There is no information on the actual effort expended by the angling population on sea bass as the stock has changed in abundance, or on changes in efficiency, but an assumption of a constant recreational fishing mortality is a reasonable first approximation for evaluating recre-ational F.

Table 4.2.5. Estimated numbers of sea anglers in the UK (England and Wales) from a number of different population surveys

REPORT YEAR OF SURVEY ESTIMATED NUMBER OF SEA

ANGLERS

National Angling Survey 1970 1 280 000

National Angling Survey 1980 1 791 000

National Rivers Authority, 1995 1994 1 104 000

Drew Associates, 2004 2003 1 450 000

Simpson and Mawle, 2005 2005 2 035 705 (small sample)

Sea Angling, 2012. 2012 960 000


Figure 4.2.13. Plot of sea angler numbers from Table 4.2.5, by year of survey.

Recreational fishery selectivity

It is first necessary to identify an appropriate selectivity function to characterize the selectivity of recreational fishing. The recreational fishery in England and Wales is pre-dominantly rod-and-line, but it is known that other gears (especially fixed or driftnets, or seines) are deployed recreationally throughout Europe. However, in France around 80% of the recreational fishing catch is from sea angling. The length frequency of re-tained sea bass recorded during the recent Sea Angling 2012 survey in England is very noisy due to small samples (many bass are released) and a clear tendency in some cases for lengths to have been reported to the nearest 5 cm (Figure 4.2.14). However, the distribution is clearly much more similar to that of the commercial line fishery than to the other commercial fishery gear groups (Figure 4.2.14). It was therefore decided to use the UK commercial line fishery selectivity (age-based) to represent the selectivity of the recreational fishery for the whole of the bass stock. The line fishery selectivity-at-age was obtained from Run 15b (combined UK/French métiers with asymptotic se-lectivity estimated for lines based on UK age composition inputs). The assumption re-garding selectivity may not hold beyond the UK, but was adopted in the absence of other information.

Figure 4.2.14. Raised length–frequency distribution of sea bass caught and kept by recreational sea anglers in England during an onsite survey of shore and private boat anglers, and a diary survey of charter boats, in 2012. The average length compositions of bass landed by commercial otter trawls (OTB), nets and lines in UK (England and Wales) are shown for comparison.

0

5

10

15

20

25

14 18 22 26 30 34 38 42 46 50 54 58 62 66 70 74 78 82 86 90 94

Perc

enta

ge

Sea angling bass length composition: England 2012

Kept: sea angling

OTB

Nets

Lines


Natural mortality

A range of natural mortality values from 0.15 to over 0.20 can be inferred from life-history parameters and observed maximum age. (Section 4.2.2.10). IBPNEW 2012 adopted a value of 0.20 for all ages. The SS3 runs carried out by IBPBass at different values of M (Section 4.2.2.10) indicated that likelihoods improved as M was increased, but the precision of recent biomass estimates declined, and at M=0.30 some instability in fitting selectivity parameters was evident. IBPBass therefore set up a run using the lower bound of M inferences (M=0.15) in order to avoid the combined M and recrea-tional F going too high, and scaled the recreational fishery selectivity so that the calcu-lated recreational landings in 2012 were 1500 t. The selectivity and recreational F vector achieving this are given in Table 4.2.6. The Recreational F(5–11) is 0.07 derived from a multiplier of 0.076 applied to the selectivity curve. The combination of M=0.15 and rec-reational F = 0.07 leads to a combined mortality vector increasing from 0.15 at the youngest ages to 0.226 at-age 8, i.e. close to the total M value of 0.20 used in the previ-ous sea bass runs.

Table 4.2.6. Age-based selectivity for UK line fishery (Run 15a) scaled to a recreational fishery F of 0.07, and added to a natural mortality M of 0.15 to give a combined mortality vector that results in a total recreational fishery removals of 1500 t in 2012. This is the final IBPBass run specification.

4.2.2.15 SS3 run with recreational F vector included (Run 22)

Run 22 was configured as:

• Four commercial fleets (1: UK OTB/nets/lines; 2: UK midwater; 3: French combined; 4: other). Fleet:selectivity - 1: double normal age-based; 2: asymp-totic age-based; 3: asymptotic length-based; 4: mirrored with fleet 3.

• Surveys: CGFS (length-based asymptotic selectivity); Solent autumn (ages 2–4).

• Recreational fishing mortality vector with F(5–11) = 0.07; asymptotic selec-tivity by age.

• Natural mortality M=0.15.

Recreational F multiplier: 0.076Recreational F(5-11) 0.070

Age selectivity Rec. F M total1 0.0003 0.000 0.15 0.1502 0.0023 0.000 0.15 0.1503 0.0180 0.001 0.15 0.1514 0.1289 0.010 0.15 0.1605 0.5445 0.041 0.15 0.1916 0.9062 0.069 0.15 0.2197 0.9873 0.075 0.15 0.2258 0.9984 0.076 0.15 0.2269 0.9998 0.076 0.15 0.226

10 1.0000 0.076 0.15 0.22611 1.0000 0.076 0.15 0.22612 1.0000 0.076 0.15 0.22613 1.0000 0.076 0.15 0.22614 1.0000 0.076 0.15 0.22615 1.0000 0.076 0.15 0.22616+ 1.000 0.076 0.15 0.226


• All parameters fitted with soft bounds.

The stock trends for this run are given in Figure 4.2.15, and show a lack of any retro-spective pattern.

Sensitivity to different recreational catch values for 2012

Increasing the input values of recreational F causes higher estimates of biomass and recruitment, and lower estimates of commercial fishing mortality (Figure 4.2.16). The combined commercial and recreational F is relatively constant across the different runs. This is because the total mortality rate is largely defined by the age and length compo-sitions for the commercial fleets, so that an increase in recreational F has to be matched by a corresponding reduction in commercial F. The relative trends in biomass, recruit-ment and commercial fishery F are largely unaffected by the additional recreational fishing mortality because it is held constant over time.

Figure 4.2.15. Stock trends from Run 22 including a time-invariant recreational fishing mortality vector. Retrospective runs peeling back five years are shown. Note that SSB, total biomass and commercial fishery F from run ending 2013 should be ignored as there is no fishery catch included for 2013; only the recruitment series is relevant due the additional Solent juvenile survey in 2013 (has no noticeable effect on 2012 and earlier estimates of biomass and F). Recreational F(5–11) is constant at F=0.07.


Figure 4.2.16. Stock trends from Run 22 with recreational F = 0.07 (dashed line) and equivalent runs (23 a–f) with recreational F(5–11) ranging from zero to 0.092 (SSB and recruitment increase with increasing recreational F while commercial F declines).

Figure 4.2.17. SS3 model Final run 22 and sensitivity runs 23 a–f (effects of recreational F multiplier at M=0.15) and 24a–d (effects of recreational F multiplier at M=0.20): left: estimates of recreational fishery removals in 2012 plotted against input recreational F; right: SSB 2013vs.recreational F.

Sensitivity of final Run 22 to input value of M

Increasing the input natural mortality value from 0.15 to 0.20, and scaling the recrea-tional fishery F(5–11) from 0.07 down to 0.055 to give 1500 t removals in 2012 (see Fig-ure 4.2.17, right hand plot), has the effect of scaling up the SSB and recruitment over the full time-series and reducing the commercial fleet fishing mortality (Figure 4.2.18). The combined commercial and recreational F values show a similar pattern (differing in absolute values by the difference between the recreational Fs of 0.07 and 0.055), while the total mortality (F+M) is effectively the same over the series.


The total likelihood for the run with M=0.20 is marginally lower than for the run with M=0.15, and the relative standard error of the recent SSB estimates is also slightly lower (Table 4.2.7). However, at M=0.20, the SSB in 2013 is estimated to be 62% of the unex-ploited (virgin) SSB, while at M=0.15 it is 49% of virgin SSB.

As there is little to choose between these two options, and the relative stock trends are the same, IBPBass decided to retain an M of 0.15 for the final run. This accommodates the observation that fish up to 28 years of age were recorded historically during periods when the recreational fishery and natural mortality were dominant. The life-history methods of inferring M that use oldest observed age as a parameter (Table 4.2.3) sug-gest M values of 0.15–0.16 for a maximum age of 28. Keeping the combination of M and recreational F as low as possible therefore is in accordance with these methods.

Table 4.2.7. Diagnostics of final run 22 with input M of 0.15 (Final run) and 0.20 (sensitivity). Both runs have recreational F multipliers giving recreational removals in 2012 of ~1500 t. RSE = relative standard error of SSB estimates.

Sensitivity to uncertainty regarding UK inshore nets and lines landings (run 25b)

As described in Section 2.2.1.2 (Figure 2.2.2), not all commercial landings of sea bass are recorded, particularly for inshore small-scale fisheries. Vessels landing less than 25 kg from a single fishing trip (due to become 30 kg) are not required to provide sales slips. This is to reduce administrative burden. As many sea bass in the UK are taken by small inshore vessels, many of which use nets and lines and bring in small catches, the potential for underestimation of landings is substantial. An independent logbook scheme run by Cefas, in conjunction with a port survey to quantify numbers of such vessels catching sea bass, suggested that true landings of sea bass could be three times higher than reported. This survey has a number of potential biases, and the figure of 3 is very approximate. A sensitivity run of the Final model Run 22 was carried out ex-panding the annual landings of UK nets and lines by a factor of three and adjusting the total trawls/nets/lines landings and age compositions accordingly (Run 25b). The rec-reational F(5–11) giving 1500 t recreational removals in 2012 was 0.055 compared with 0.07 in the final Run 22. The effect on the assessment is simply to scale up the biomass, while relative stock trends are unaffected (Figure 4.2.19).

TOTAL likelihoodCatchEquil_catchSurveyLength_compAge_compRecruitmentForecast_RecruitmentParm_priorsParm_softboundsCrash_PenConvergenceSPB_2011; RSE 13836 0.144 17138 0.144SPB_2012; RSE 12969 0.169 16193 0.164SPB_2013; RSE 11114 0.208 14138 0.193SSB 2013/virgin SSB 0.49 0.62

8.36E-05 2.24E-05

0 00.0312498 0.0322878

0 0

181.455 181.43427.3061 26.3039

0 0

0.0192335 0.0022012864.8834 64.3149261.071 261.423

Run 22 M=0.15 Run 22 M=0.20534.766 533.511.25E-07 1.25E-07


Figure 4.2.18. Sensitivity of final Run 22 to input natural mortality of 0.15 and 0.20 (both runs have recreational F multipliers giving ~1500 t recreational removals in 2012.


Figure 4.2.19. Comparison of stock trends from final Run 22, and an equivalent run (25b) in which the annual landings of UK nets and lines have been increased by a factor of three to examine sen-sitivity to underreporting due to the 25 kg per trip exemption from supplying sales slips.

4.3 Conclusions from model development process

The model run 22 considered by IBPBass as suitable for future update assessments by ICES differs in several important respects from the model presented by ICES WGCSE in 2013. Key changes include:

• Addition of the Channel Groundfish survey; • Removal of UK spring Solent and autumn Thames surveys, leaving the au-

tumn Solent Survey; • Combination of the UK trawls, nets and lines fleets in a single fleet; • Use of age-based selectivity for fleets with age compositions, instead of

length-based selectivity; • Adoption of a double-normal selectivity for the combined UK fleets; • Revision of input effective sample sizes; • Addition of a constant recreational fishing mortality (F5–11 = 0.07); • Reduction of the base natural mortality rate from 0.2 to 0.15.

Exploratory runs indicated that the changes to the surveys, adding the CGFS and re-moving the Thames and Solent spring survey, had relatively little impact on the stock trends and absolute values of biomass, recruitment and F. However, the general trends in the CGFS raw data support the general trends in the current and previous assess-ments, and the survey includes a wide size range of bass. This addressed a key concern of WGCSE that the only survey data in the assessment were for young juveniles.


The combination of the UK trawls, nets and lines in a single fleet addressed a concern that the individual fleets may have strongly correlated errors due to the use of a com-mon age–length key for all fleets in each area and year. This will have led to more realistic estimates of variance from the inverse hessian.

Two factors affecting the absolute values for biomass and fishing mortality (causing a general increase in biomass and decrease in F) were the adoption of age-based selec-tivity for the UK fleets rather than length-based selectivity, and the implementation of a domed (double-normal) selectivity for the UK fleets. Runs using length-based selec-tivity resulted in some unusually steep ascending limbs, and the selectivity-at-age de-rived by SS3 from the length based selectivity and the size-at-age data were not in accordance with empirically derived selectivity-at-age. Switching to age-based selec-tivity produced selectivity curves very similar to the empirical ones, but this resulted in an increase in absolute biomass. The use of a double-normal selectivity for UK fleets was supported by empirical observations from the raw data, combining domed selec-tivity for nets and trawls with asymptotic selectivity for lines, to produce a composite pattern. This form of selectivity pattern can cause an overall reduction in F and increase in biomass because some of the steepness in the age profiles is now explained by de-creasing fishery selectivity with increasing age. The assumption of asymptotic selectiv-ity for the French combined fleet is based on a much larger proportion of the catch coming from midwater trawls and lines than is the case for UK fleets, and relatively less inshore fishing by bottom trawlers. The asymptotic selectivity on fleets taking a large part of the international catch helps prevent the model from creating a large “cryptic” biomass of old bass that are alive in the sea but unavailable to any sampling gear.

The current assessment does not include discards, mainly from the trawl fleets, and therefore cannot replicate the true fishery selectivity pattern. The IBPBass estimates of selectivity are in fact a composite of selectivity and retention curves. Available discards data indicate that the retention curve for each UK gear is steep and centred on the min-imum landing size. Most discarding in the UK fleets is by otter trawlers fishing inshore. The 1985–2013 average trawl discard rate was 16% by weight and 37% by number. This fleet is a minor contributor to the total international landings of sea bass, so the discards will have a relatively small impact on total catch-at-age. Discard rates of gillnetters were much lower: 1% by weight and 4% by number. Beam trawlers discarded 5% by weight and 27% by number but overall catches of sea bass by this fleet are very small. Discarding was almost entirely related to the MLS of 36 cm.

The equivalent discard rate estimates for French demersal otter trawlers in 2009–2013 varied widely between years and averaged less than 10%. A smaller proportion of French otter trawl activity is close inshore compared with UK vessels, and a smaller discard rate is expected.

As discarding is consistently driven by the MLS of (mostly) 36 cm, the absence of dis-cards data is likely to mainly affect the time-series of recruitment estimates.

The addition of a constant recreational F vector represented a major change, allowing, for the first time, an ability to explore the relative fishing mortality caused by commer-cial and recreational fishing. The results are very dependent on the estimates of recre-ational fishery harvests from surveys in the last few years, each of which have relatively low precision, and on the assumption of constant F over time, supported only very weakly by some survey estimates of numbers of sea anglers in England and Wales since 1970. To avoid a combination of M and historical recreational F that may be out of line with observed maximum age of sea bass, IBPBass reduced the value of M from


0.20 to 0.15. In combination, M and recreational F increase from 0.15 (M only) at the youngest ages to 0.23 (M&F) at the oldest ages. These values are in combination not far from the constant M=0.20 used by WGCSE in 2013. Choice of M, and inclusion of a constant recreational F vector, had little impact on relative trends. Adding a constant multiplier of 3.0 to UK inshore nets and lines landings, to explore sensitivity to under-estimation of landings, also had no effect on relative trends, but caused a further in-crease in biomass without altering the estimates of commercial F.

The combination of all these changes to the assessment has had very little impact on recruitment compared to the WGCSE 2013 assessment, but leads to a general increase in biomass throughout the series, with a smaller depletion relative to the biomass in the early part of the series (Figure 4.3.1). The estimates of commercial fishery F are lower throughout the series, and do not show the sharp increase during the 2000s shown by the WGCSE assessment. A more gradual increase in F is apparent. The trend of declining biomass in recent years is apparent in both assessments, following from the apparent reduction in recruitment.

The overall picture from both assessments is of a stock that has undergone an expan-sion since the mid-1990s due to strong recruitment in 1989 and some subsequent above-average recruitment, and is now declining due to much poorer recent recruitment. The fishing mortality in this period of decline is larger than in the early 1980s when there was also a declining trend in biomass following a period of poor recruitment.

IBPBass considers the final Run 22 to be a more robust representation of stock devel-opment than the model proposed by WGCSE 2013, but cautions that there are a num-ber of uncertainties regarding recreational catches, M, non-inclusion of discards estimates, selectivity and absence of French fishery composition data prior to 2000 that impact the absolute values of biomass and fishing mortality. However, IBPBass con-cluded that the new assessment contains valuable information on fishing mortality rates that should allow the assessment to be presented as a full analytical assessment, rather than “trends only” where mean-standardised trends are given rather than abso-lute values.


Figure 4.3.1. Comparison of stock trends between Run 22 (FINAL), and the WGCSE 2013 final age–length assessment.

4.4 Final assessment model and diagnostics

The data incorporated in the assessment are shown graphically in Figure 4.4.1 (left plot). The fleet landings are also shown. A range of model outputs are shown in Figures 4.4.2 to 4.4.12. Standard summary tables, and tables of output stock numbers and com-mercial fishery F are given in Tables 4.4.1–4.4.3. Note that all selectivity parameters for the UK trawl/nets/line fleet are estimated in the final run, whereas parameter 6 (selec-tivity at oldest ages) was fixed in sensitivity runs around Run 22. This does not alter the relative trends. Also, the final run includes the 2013 Solent Autumn survey which means there is no commercial catch for 2013. This has negligible impact on biomass and fishing mortality. The likelihoods and their component values for Run 22 are given earlier in Table 4.2.7. The fishery composition data carry 82% of the total of the negative log likelihood, while surveys contribute only 12%. The inclusion of 2013 Solent survey data without any fishery data for 2013 is recognised as not being good practice, but this has no discernible impact on stock and fishing mortality estimates for 2012 or earlier (see Figure 4.2.15). This issue will not arise for the update assessment in 2014, for which 2013 catch data will be available.


Figure 4.4.1. Left: Datasets used in the final sea bass run 22. Right: landings series for the four fleets.

Figure 4.4.2. Final sea bass run 22: Fitted length-based and age-based selectivity curves.


Figure 4.4.3. Final sea bass run 22: fit to French fishery length composition data.

Figure 4.4.4. Final sea bass run 22: Fit to Channel ground fish survey length compositions.


Figure 4.4.5. Final sea bass run 22: Fit to French fishery and Channel ground fish survey length compositions, aggregated across time.


Figure 4.4.6. Final sea bass run 22: Fit to age composition data for the combined UK otter trawl, nets and lines fleets.


Figure 4.4.7. Final sea bass run 22: Fit to age composition data for the UK midwater trawl fleet.

Figure 4.4.8. Final sea bass run 22: Fit to UK fleets age compositions, aggregated across time.


Figure 4.4.9. Final sea bass run 22: Fit to Channel ground fish survey total abundance index (natural and logarithmic scales), accounting for length-based selectivity.


Figure 4.4.10. Final sea bass run 22: Fit to UK Solent autumn bass survey indices for ages 2, 3 and 4, on log scale (treated as independent indices in the model to avoid problems in fitting selectivity).


Figure 4.4.11. Final sea bass run 22: Top left: time-series of log recruit deviations. Top right: stock–recruit scatter (model is fitted assuming Beverton–Holt stock–recruit model and steepness = 0.999.) Bottom left: recruitment time-series; bottom right: SSB series (female only, based on 50:50 sex ratio at-age) with 95% asymptotic intervals.


Figure 4.4.12. Illustration of potential recreational fishery removals compared with landings of commercial fishery métiers from UK and France, based on a fixed recreational F vector (F(5–11)= 0.07) included in the Final Run 22. Note that these figures are intended to illustrate the potential magnitude of recreational catches based on the recreational fishing mortality needed to generate 1500 t of recreational fishery removals in 2012, and are completely conditional on an assumed con-stant recreational F which is unlikely to be true.


Table 4.4.1. Numbers-at-age in the population, at the start of the year. Greyed cells indicate num-bers influenced by Solent 2013 survey without fishery data in 2013 and are therefore overestimates (will not be an issue for WGCSE 2014 assessment); bold numbers for 2012 and 2013 year classes are SS3 values of average recruitment.

Table 4.4.2. Commercial fishing mortality-at-age.

Year 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16+1985 137 765 4875 4626 1925 543 784 471 490 1679 359 305 242 155 145 100 4181986 477 118 658 4195 3956 1596 423 579 341 352 1201 256 217 173 110 103 3691987 3692 410 102 566 3581 3251 1217 304 407 237 243 828 176 149 119 76 3241988 5993 3178 353 87 481 2885 2376 822 198 261 151 154 522 111 94 75 2521989 29690 5158 2735 304 75 395 2186 1694 573 137 179 103 105 357 76 64 2231990 5715 25554 4440 2353 259 61 297 1549 1174 394 94 122 70 72 242 52 1951991 4380 4919 21995 3820 2007 212 46 210 1067 802 268 63 83 47 48 164 1661992 6538 3770 4234 18921 3247 1609 152 30 136 686 512 170 40 52 30 31 2091993 3872 5627 3245 3642 16093 2619 1174 103 20 88 442 328 109 26 33 19 1531994 9569 3333 4843 2792 3106 13139 1961 826 71 14 60 299 222 74 17 23 1161995 14379 8236 2868 4168 2386 2560 10012 1420 591 51 10 43 213 158 52 12 981996 1916 12376 7089 2468 3560 1960 1935 7170 1004 417 36 7 30 149 110 37 781997 18138 1649 10652 6098 2098 2861 1417 1293 4636 641 264 22 4 19 94 69 721998 9003 15611 1420 9164 5188 1690 2074 953 845 3006 413 169 14 3 12 60 901999 18018 7749 13437 1221 7800 4193 1236 1409 629 553 1951 267 109 9 2 8 962000 8678 15508 6670 11559 1039 6286 3042 826 907 400 349 1227 167 69 6 1 652001 12349 7469 13348 5738 9841 842 4637 2087 550 597 261 227 796 109 44 4 432002 16849 10629 6428 11483 4887 7983 622 3185 1390 362 390 170 148 517 70 29 302003 17411 14502 9149 5531 9783 3966 5891 428 2130 922 238 256 112 97 338 46 392004 12968 14986 12482 7870 4702 7869 2876 3938 276 1355 581 150 160 70 60 211 532005 10571 11162 12898 10737 6688 3775 5684 1912 2524 174 848 362 93 99 43 37 1642006 11912 9098 9607 11094 9103 5314 2671 3670 1183 1532 105 506 215 55 59 26 1192007 8002 10253 7831 8263 9401 7208 3727 1709 2255 715 915 62 299 127 33 35 852008 2737 6887 8825 6736 7014 7498 5127 2435 1078 1403 440 561 38 183 78 20 732009 2671 2356 5928 7591 5720 5603 5348 3372 1552 679 875 273 347 23 113 48 582010 365 2299 2028 5099 6451 4590 4039 3568 2183 993 431 553 172 219 15 71 662011 2173 314 1978 1744 4323 5113 3221 2591 2203 1328 598 258 330 103 130 9 822012 6363 1870 271 1702 1482 3459 3659 2123 1651 1388 829 371 160 204 63 80 562013 6363 5477 1610 233 1442 1170 2397 2317 1299 1000 831 494 221 95 121 38 81

Year 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 161985 0.000 0.000 0.000 0.005 0.027 0.059 0.084 0.098 0.105 0.110 0.112 0.113 0.114 0.114 0.114 0.114 0.1151986 0.000 0.000 0.000 0.007 0.036 0.080 0.112 0.129 0.137 0.143 0.146 0.147 0.148 0.149 0.149 0.149 0.1491987 0.000 0.000 0.000 0.011 0.056 0.122 0.173 0.202 0.218 0.228 0.233 0.235 0.236 0.237 0.237 0.238 0.2381988 0.000 0.000 0.000 0.007 0.039 0.086 0.119 0.136 0.144 0.150 0.153 0.154 0.155 0.156 0.156 0.156 0.1561989 0.000 0.000 0.000 0.007 0.041 0.092 0.126 0.142 0.148 0.154 0.157 0.159 0.159 0.160 0.160 0.160 0.1601990 0.000 0.000 0.000 0.008 0.042 0.094 0.130 0.148 0.155 0.161 0.164 0.166 0.167 0.167 0.167 0.167 0.1681991 0.000 0.000 0.000 0.011 0.061 0.141 0.190 0.210 0.216 0.224 0.228 0.230 0.231 0.232 0.232 0.232 0.2321992 0.000 0.000 0.000 0.010 0.055 0.124 0.171 0.195 0.205 0.214 0.218 0.220 0.221 0.222 0.222 0.222 0.2231993 0.000 0.000 0.000 0.008 0.043 0.098 0.133 0.148 0.154 0.160 0.163 0.165 0.165 0.166 0.166 0.166 0.1661994 0.000 0.000 0.000 0.006 0.034 0.080 0.104 0.110 0.109 0.113 0.114 0.115 0.116 0.116 0.116 0.116 0.1161995 0.000 0.000 0.000 0.006 0.037 0.088 0.115 0.122 0.122 0.127 0.129 0.130 0.130 0.131 0.131 0.131 0.1311996 0.000 0.000 0.000 0.011 0.059 0.133 0.185 0.211 0.222 0.231 0.235 0.237 0.239 0.239 0.239 0.240 0.2401997 0.000 0.000 0.000 0.010 0.057 0.130 0.178 0.200 0.207 0.215 0.219 0.221 0.222 0.223 0.223 0.223 0.2231998 0.000 0.000 0.000 0.010 0.053 0.122 0.168 0.190 0.199 0.206 0.210 0.212 0.213 0.214 0.214 0.214 0.2141999 0.000 0.000 0.000 0.010 0.056 0.129 0.184 0.215 0.226 0.234 0.238 0.240 0.241 0.242 0.242 0.242 0.2422000 0.000 0.000 0.000 0.010 0.050 0.113 0.158 0.182 0.193 0.201 0.205 0.207 0.208 0.208 0.208 0.209 0.2092001 0.000 0.000 0.000 0.009 0.049 0.111 0.157 0.181 0.191 0.199 0.203 0.205 0.206 0.206 0.206 0.206 0.2072002 0.000 0.000 0.000 0.009 0.049 0.113 0.156 0.177 0.185 0.192 0.195 0.197 0.198 0.198 0.199 0.199 0.1992003 0.000 0.000 0.000 0.011 0.058 0.130 0.184 0.214 0.227 0.236 0.240 0.242 0.244 0.244 0.245 0.245 0.2452004 0.000 0.000 0.000 0.011 0.060 0.134 0.189 0.220 0.233 0.243 0.247 0.250 0.251 0.252 0.252 0.252 0.2522005 0.000 0.000 0.000 0.014 0.070 0.154 0.219 0.255 0.273 0.285 0.290 0.294 0.295 0.296 0.296 0.296 0.2972006 0.000 0.000 0.001 0.014 0.074 0.163 0.228 0.262 0.278 0.290 0.296 0.299 0.300 0.301 0.302 0.302 0.3022007 0.000 0.000 0.000 0.013 0.066 0.149 0.207 0.236 0.249 0.259 0.264 0.267 0.268 0.269 0.269 0.269 0.2692008 0.000 0.000 0.000 0.012 0.065 0.147 0.200 0.226 0.236 0.246 0.251 0.253 0.254 0.255 0.255 0.255 0.2562009 0.000 0.000 0.000 0.011 0.060 0.136 0.186 0.210 0.220 0.229 0.233 0.236 0.237 0.237 0.238 0.238 0.2382010 0.000 0.000 0.000 0.014 0.073 0.163 0.225 0.257 0.271 0.282 0.288 0.291 0.292 0.293 0.294 0.294 0.2942011 0.000 0.000 0.000 0.012 0.063 0.143 0.198 0.225 0.236 0.245 0.250 0.253 0.254 0.254 0.255 0.255 0.2552012 0.000 0.000 0.000 0.014 0.077 0.175 0.238 0.266 0.276 0.287 0.292 0.295 0.296 0.297 0.297 0.298 0.298


Table 4.4.3. Assessment summary. Recruitment, SSB, total-stock biomass TSB, and F.

YEAR RECRUITS (AGE 0) SSB TSB F(5–11) LANDINGS

1985 137 9381 12330 0.097 1076

1986 477 8591 11807 0.128 1315

1987 3692 8065 11021 0.202 1979

1988 5993 7288 9546 0.135 1238

1989 29690 6911 8741 0.140 1161

1990 5715 6187 8516 0.145 1033

1991 4380 5379 9582 0.206 1225

1992 6538 4628 11035 0.192 1184

1993 3872 5399 12989 0.146 1251

1994 9569 7501 14943 0.106 1370

1995 14379 9838 16278 0.119 1777

1996 1916 11210 16930 0.208 3023

1997 18138 10797 16431 0.196 2620

1998 9003 10231 16251 0.187 2388

1999 18018 10102 16773 0.209 2665

2000 8678 9968 17355 0.180 2397

2001 12349 10418 18610 0.178 2482

2002 16849 11056 19932 0.174 2628

2003 17411 12123 21324 0.210 3445

2004 12968 12791 22258 0.217 3730

2005 10571 13195 23099 0.253 4392

2006 11912 13124 23395 0.259 4522

2007 8002 13297 23522 0.233 4213

2008 2737 13951 23667 0.223 4244

2009 2671 14427 23189 0.207 4013

2010 365 14653 22121 0.254 4758

2011 2173 13836 19615 0.222 3870

2012 N/A1 12969 17121 0.261 4060

2013 N/A1 11114 13946

1 Recruit deviations not estimated for these year classes.

4.5 IBPBass revised stock assessment inputs and model structure/ param-eters

The structure and input data/ parameters of the IBPBass revised SS3 model are sum-marized below:

Model structure

• Temporal unit: annual based data (landings, survey indices, age–frequency and length–frequency);



Fleet definition

Four fleets defined: 1. UK bottom trawls, nets and lines; 2. UK midwater pair trawls; 3. French fleets (combined); 6. Other (other countries and other UK fleets combined).

Landed catches

Annual landings in tonnes from 1985 to final year for the four fleets from ICES Subdi-visions IVb and c, VIIa, d–h. French data were as provided by Ifremer.

Abundance indices

Channel Groundfish Survey in VIId in autumn (France). 1988 to present: total swept-area abundance index and associated length composition data. Number of stations with sea bass used as input effective sample size. Input CV for survey = 0.30 all years. First three years of composition data are excluded.

Cefas Solent survey in autumn (VIId). Years 1986 to 2011; 2013. Three independent abundance index series were defined, each being a single age group (2, 3, 4 years old). They are treated as three independent surveys (following a recommendation from R. Methot) to circumvent difficulties in estimating selectivity parameters for a survey se-ries comprising only three young age groups, although this approach loses covariance information due to year-effects in the survey.

Fishery landings age composition data: UK fleets

Age bins: 0 to 15 with a plus group for ages 16 and over. Age compositions for UK fleets expressed as fleet-raised numbers-at-age, although they are treated as relative compositions in SS3. Year range for UK trawls/nets/lines: 1985 to present; UK midwater pair trawl: 1996 to present.

Length composition data: French fleets

The length bin was set from 4 to 100 cm by 2 cm intervals. Length compositions for the following fishing fleets were used: French all fleets combined: 2000 to present.


Table 4.5.2 summarises key model assumptions and parameters. Other parameter val-ues and input data characteristics are defined in the SS3 control file BassIVVII.ctl, the forecast file Forecast.SS and the data file BassIVVII.dat as used by IBPBass 2014.


Table 4.5.1. Key model assumptions and parameters from the IBPBass final run.


Starting year 1985

Ending year 2013

Equilibrium catch for starting year 0.82* landings in 1985 by fleet.

Number of areas 1

Number of seasons 1


Number of surveys two surveys: CGFS; Solent autumn survey (Solent spring and Thames survey removed).


Number of active parameters 68


Maximum age 30

Genders 1


Ages for summary total biomass 0–30




Minimum age for growth model 2


Maturity Logistic 2-parameter – females; L50 = 40.65 cm




Maximum F 2.9

Fleet 1: UK Trawl/nets/lines selectivity Double normal, age-based

Fleet 2: UK Midwater trawl selectivity Asymptotic, age-based

Fleet 3: Combined French fleet selectivity Asymptotic, length-based

Fleet 4: Other fleets/gears selectivity Asymptotic: mirrors French fleet

Year-invariant recreational fishing mortality vector (F(5–11) = 0.07): selectivity

Asymptotic, age-based (fixed, not estimated)



CGFS survey timing (yr) 0.75


Survey selectivities: Solent autumn: [all survey data entered as single ages; sel = 1]

Survey selectivities: CGFS Double normal, length-based



















1965

1 as recommended by R. Methot after scrutinizing earlier SS3 runs during IBPNEW 2012, and used by IBPNEW and WGCSE. The WGCSE 2012 tabulated the original value of 5.78 cm at age 0 in error.


5 Forecast

Due to the additional complexity of adding a fixed recreational fishing mortality vector for removals (harvest), and the time required to configure Stock Synthesis to mirror the ICES procedures for short-term forecasts, IBPBass decided not to try to develop a fore-cast procedure within Stock Synthesis for use in the forthcoming WGCSE update as-sessment. This unfortunately loses the ability to provide MCMC confidence intervals around the assessment and forecasted variables, and the forecasts are entirely deter-ministic. Management options involving biological reference points (BRPs) adopt BRPs conditional on the assumptions in the assessment regarding M, selectivity, maturity, weights-at-age, etc.

5.1 Estimating year-class abundance

Recruitment estimates for sea bass are well below average from 2008 onwards (Table 5.1.1).

Table 5.1.1. Recruitment estimates from 2000 onwards from the final Stock Synthesis run, and their derivation.

YEAR CLASS NUMBER (‘000) SE DERIVATION IN SS3

Recr_2000 9660 1116 Recruit deviation estimated

Recr_2001 13 762 1522 Recruit deviation estimated











Recr_2012 7491 672 Long-term average

Recr_2013 7490 672 Long-term average

The long-term geometric mean for 1985–2011 is 6041 thousand fish.

Recruitment deviations estimated from 2008 to 2011 are well below the long-term GM, and the short-term GM for 2008–2011 is 1769 thousand. The recent estimated devia-tions are estimated almost entirely from the Solent autumn survey including the 2013 survey.

WGCSE (2013) reviewed some information on environmental influences on sea bass recruitment which supports a recent reduction in recruitment. Survival of 0-gp and 1-gp sea bass in nursery areas in estuaries and saltmarshes is thought to be enhanced by warmer conditions promoting survival through the first two winters, and increasing the growth rates (Pawson, 1992). Inshore sea surface temperatures in January–March were relatively low along the south coast of England prior to 1989, coinciding with a period of below-average recruitment estimates from SS3 (Figure 5.1.1, reproduced


from WGCSE 2013). From 1989 to 2008, temperatures were higher on average, coincid-ing with the very strong 1989 year class and a period of generally elevated recruitment. From 2009 onwards, spring temperatures reduced again. For a year class, the relevant temperature regimes are in their first two years in the nursery areas, i.e. for the 2008 year class from spawning in spring 2008, it is the temperatures from the time of year when the post-larvae arrive in the nursery area later in 2008 through the 2008/2009 and 2009/2010 winters. These data will be updated at WGCSE 2014. The winter of 2012/2013 was exceptionally cold in the UK, and it would appear inappropriate to use a long-term geometric mean for the 2012 year class. Hence recruitment for the 2012 year class was represented by the 2008–2011 short-term GM of 1552 thousand fish. The winter of 2013/2014 was much warmer, although characterized by widespread flooding around estuaries that include some bass nursery areas. The effect of the flooding on bass re-cruitment is not known. IBPBass concluded there was no information on which to base a recruitment other than the long-term GM of 6041 thousand fish for year classes from 2013 onwards.

Recruitment values for the short-term forecast are summarised in Table 5.1.2.

Figure 5.1.1. (a) Mean January–March sea surface temperature (SST) estimates for five coastal loca-tions along the south coast of England and the Thames estuary; (b) bass recruitment trends from the WGCSE update SS3 model including during the 1975–1985 burn-in period. (Temperature data from http://www.cefas.defra.gov.uk/our-science/observing-and-modelling/monitoringpro-grammes/sea-temperature-and-salinity-trends.aspx.).


Table 5.1.2. Recruitment estimates included in the short-term forecast for sea bass.

YEAR CLASS SS3 (AGE 0) GM 2008–2011 GM 2008–2011

2011 2477 thousand

2012 1769 thousand

2013 6041 thousand

2014 6041 thousand

The input for the short-term catch predictions is given in Table 5.1.3. The derivation of the inputs is described in Table 5.1.4.

Table 5.1.3. Inputs for sea bass short-term forecast.

ageNo. at age in

2013weight in

stock

Proportion mature

(female)H.Cons mean F

(2010-2012)H.Cons mean

weights Recreational F

Recreational removals mean

weight M

0 6041 0.002 0.000 0.000 0.007 0.000 0.007 0.151 1522 0.022 0.000 0.000 0.078 0.000 0.052 0.152 1835 0.101 0.000 0.000 0.386 0.000 0.154 0.153 266 0.219 0.000 0.012 0.581 0.002 0.295 0.154 1638 0.383 0.187 0.062 0.714 0.013 0.480 0.155 1347 0.588 0.421 0.140 0.877 0.054 0.702 0.156 2780 0.827 0.640 0.192 1.094 0.091 0.954 0.157 2693 1.091 0.793 0.217 1.334 0.099 1.228 0.158 1510 1.374 0.885 0.228 1.592 0.100 1.517 0.159 1162 1.669 0.937 0.236 1.865 0.100 1.815 0.15

10 968 1.970 0.965 0.241 2.149 0.100 2.116 0.1511 576 2.272 0.981 0.243 2.439 0.100 2.416 0.1512 258 2.570 0.989 0.244 2.728 0.100 2.711 0.1513 111 2.862 0.993 0.245 3.011 0.100 2.998 0.1514 142 3.145 1.000 0.245 3.287 0.100 3.275 0.1515 44 3.418 1.000 0.246 3.552 0.100 3.540 0.15

16+ 95 3.964 1.000 0.246 4.269 0.100 4.070 0.15


Table 5.1.4. Derivation of short-term forecast inputs.

INPUT DATA DERIVATION

Numbers-at-age 0–16+ in 2013 SS3 output. (N age zero overwritten by long-term GM 6041. N age 1 overwritten by short-term 2008–2011 GM (1769) reduced by M=0.15 (1522)

Recruitment 2014 onwards Long-term GM 6041

Mean wt-at-age in stock SS3 output

Proportion mature (female) SS3 output

Commercial fishery (H-cons) mean F at-age Average 2010–2012: SS3 output Z’s minus M=0.15 and recreational F at-age

Commercial fishery (H-cons) mean weight-at-age

SS3 output figures on mean weight in UK, French and other fleets, weighted by SS3 model estimates of landings numbers-at-age for the fleets

Recreational removals F at-age Input values to SS3 (year-invariant); based on commercial lines selectivity from SS3 Run 15b)

Recreational removals weights-at-age Output values for UK commercial fleets from final SS3 Run

M 0.15 at all ages

A detailed forecast is given in Table 5.1.5 for the status-quo F option, which is the most likely forecast given the absence of any restrictive management controls on effort or landings of sea bass.

Two sets of management options are given in Table 5.1.6a,b, showing the expected commercial landings and total recreational removals in 2013 and 2014, and SSB and total biomass in 2015, for a range of F-multipliers applied simultaneously to the com-mercial fishery and recreational fishery removals (Table 5.1.6a) or to the commercial fishery only with recreational F held constant except at commercial Fmult = 0 when the recreational F is set to zero (Table 5.1.6b). Continuation of status quo F is expected to result in SSB declining further to 8430 t (lowest observed SSB was 5200 t in 1992), and total biomass to 10 510 t (lowest observed = 9580 t in 1990).

Options for F0.1 or F35%spr (or similar) BRPs will be added to the option table for future update assessments.


Table 5.1.5. Detailed table for status quo F forecast.

Year: 2013H.cons F mult: 1 F(5-11): 0.214Recreational F mult 1 F(5-11): 0.092

AgeF(5-11): H-cons

F(5-11): Recreational

Catch Nos: H-cons

Yield: H-cons

Catch Nos: Recreational

Yield: Recreational Stock Nos Biomass

SSB nos. Jan 1

SSB tonnes Jan 1

0 0.000 0.000 0.0 0.0 0.0 0.0 6041 9 0 01 0.000 0.000 0.0 0.0 0.0 0.0 1522 34 0 02 0.000 0.000 0.5 0.2 0.4 0.1 1835 185 0 03 0.012 0.002 3.0 1.7 0.4 0.1 266 58 0 04 0.062 0.013 90.5 64.6 18.9 9.1 1638 628 307 1175 0.140 0.054 159.2 139.6 62.1 43.6 1347 792 567 3336 0.192 0.091 433.5 474.2 204.5 195.2 2780 2298 1779 14707 0.217 0.099 468.1 624.3 212.5 261.0 2693 2939 2136 23308 0.228 0.100 273.3 435.1 119.9 181.9 1510 2075 1337 18379 0.236 0.100 217.6 405.7 92.0 167.0 1162 1940 1089 1818

10 0.241 0.100 184.3 396.1 76.5 161.8 968 1907 934 184111 0.243 0.100 110.6 269.6 45.4 109.8 576 1308 565 128212 0.244 0.100 49.7 135.7 20.3 55.1 258 663 255 65513 0.245 0.100 21.5 64.7 8.8 26.3 111 318 110 31614 0.245 0.100 27.5 90.4 11.2 36.7 142 447 142 44715 0.246 0.100 8.6 30.4 3.5 12.4 44 151 44 151

16+ 0.246 0.100 18.4 78.7 7.5 30.5 95 377 95 377Total 2066 3211 884 1291 22990 16128 9360 12976


AgeF(5-11): H-cons


Catch Nos: H-cons

Yield: H-cons



SSB nos. Jan 1

SSB tonnes Jan 1

0 0.000 0.000 0.0 0.0 0.0 0.0 6041 9 0 01 0.000 0.000 0.0 0.0 0.1 0.0 5200 116 0 02 0.000 0.000 0.4 0.2 0.3 0.0 1310 132 0 03 0.012 0.002 17.7 10.3 2.6 0.8 1579 345 0 04 0.062 0.013 12.4 8.9 2.6 1.2 225 86 42 165 0.140 0.054 154.6 135.6 60.3 42.3 1309 770 551 3246 0.192 0.091 148.9 162.9 70.3 67.0 955 790 611 5057 0.217 0.099 313.5 418.2 142.4 174.8 1804 1968 1431 15618 0.228 0.100 305.8 486.8 134.1 203.5 1690 2322 1496 20569 0.236 0.100 175.4 327.0 74.2 134.6 937 1564 878 1465

10 0.241 0.100 136.1 292.4 56.5 119.5 715 1408 690 135911 0.243 0.100 113.8 277.4 46.8 113.0 592 1346 581 132012 0.244 0.100 67.8 184.9 27.7 75.2 352 903 348 89313 0.245 0.100 30.4 91.5 12.4 37.2 157 450 156 44714 0.245 0.100 13.1 43.1 5.3 17.5 68 213 68 21315 0.246 0.100 16.8 59.5 6.8 24.2 87 296 87 296

16+ 0.246 0.100 16.5 70.2 6.7 27.3 85 337 85 337Total 1523 2569 649 1038 23105 13054 7022 10792


AgeF(5-11): H-cons


Catch Nos: H-cons

Yield: H-cons



SSB nos. Jan 1

SSB tonnes Jan 1

0 0.000 0.000 0.0 0.0 0.0 0.0 6041 9 0 01 0.000 0.000 0.0 0.0 0.1 0.0 5200 116 0 02 0.000 0.000 1.3 0.5 0.9 0.1 4475 451 0 03 0.012 0.002 12.6 7.3 1.9 0.6 1127 247 0 04 0.062 0.013 74.0 52.8 15.5 7.4 1340 513 251 965 0.140 0.054 21.3 18.7 8.3 5.8 180 106 76 456 0.192 0.091 144.6 158.2 68.3 65.1 928 767 594 4917 0.217 0.099 107.7 143.7 48.9 60.1 620 676 491 5368 0.228 0.100 204.8 326.0 89.8 136.3 1132 1555 1002 13779 0.236 0.100 196.2 365.9 83.0 150.6 1048 1750 982 1640

10 0.241 0.100 109.7 235.7 45.5 96.3 576 1135 556 109511 0.243 0.100 84.0 204.8 34.5 83.4 437 993 429 97412 0.244 0.100 69.8 190.3 28.5 77.4 362 930 358 91913 0.245 0.100 41.4 124.8 16.9 50.7 214 614 213 61014 0.245 0.100 18.6 61.0 7.6 24.8 96 302 96 30215 0.246 0.100 8.0 28.4 3.3 11.5 41 141 41 141

16+ 0.246 0.100 20.2 86.4 8.2 33.5 104 414 104 414Total 1114 2005 461 804 23922 10718 5193 8639


Table 5.1.6. Sea bass in IVbc, VIIa,d–h. Management option table.

a. F multipliers applied simultaneously to commercial fleet and recreational fishery removals.

b. F multipliers applied to commercial fleet only.

2013 Commercial fishery Recreational fisheryBiomass SSB Fmult Fbar Landings Fmult Fbar Landings

16128 12976 1 0.214 3211 1 0.092 1291

2014 Commercial fishery Recreational fishery 2015Biomass SSB Fmult Fbar Landings Fmult Fbar Landings Biomass SSB Total Fbar

13054 10792 0 0.040 0 0 0.000 0 14218 11886 0.0400.1 0.021 295 0.1 0.009 119 13815 11512 0.0310.2 0.043 580 0.2 0.018 235 13425 11150 0.0610.3 0.064 857 0.3 0.028 346 13047 10799 0.0920.4 0.086 1125 0.4 0.037 455 12681 10459 0.1220.5 0.107 1385 0.5 0.046 560 12327 10131 0.1530.6 0.128 1636 0.6 0.055 662 11984 9813 0.1840.7 0.150 1881 0.7 0.064 760 11652 9505 0.2140.8 0.171 2117 0.8 0.074 856 11331 9207 0.2450.9 0.193 2347 0.9 0.083 948 11019 8919 0.275

1 0.214 2569 1 0.092 1038 10718 8639 0.306

2013 Commercial fishery Recreational fisheryBiomass SSB Fmult Fbar Landings Fmult Fbar Landings

16128 12976 1 0.214 3211 1 0.092 1291

2014 Commercial fishery Recreational fishery 2015Biomass SSB Fmult Fbar Landings Fmult Fbar Landings Biomass SSB Total Fbar

13054 10792 0 0.000 0 1 0.092 1155 13051 10802 0.0920.1 0.021 283 1 0.092 1143 12794 10563 0.1130.2 0.043 559 1 0.092 1130 12541 10329 0.1350.3 0.064 830 1 0.092 1118 12295 10101 0.1560.4 0.086 1094 1 0.092 1106 12054 9877 0.1780.5 0.107 1354 1 0.092 1094 11819 9659 0.1990.6 0.128 1607 1 0.092 1083 11588 9446 0.2200.7 0.150 1855 1 0.092 1071 11363 9237 0.2420.8 0.171 2098 1 0.092 1060 11143 9033 0.2630.9 0.193 2336 1 0.092 1049 10928 8834 0.284

1 0.214 2569 1 0.092 1038 10718 8639 0.306


6 Biological reference points

6.1 Background

WGCSE 2013 computed yield-per-recruit based biological reference points F0.1 and F35%spr based on the inputs and outputs of the stock synthesis update assessment. The F(5–11) value for F35%spr was 0.17 and the F0.1 was 0.18. The subsequent Advice Drafting Group (ADG) commented (in the meeting minutes) that this stock was data-limited stock category 3.2.0 (DLS 3.2.0; Analytical assessment. PA buffer applied due to strong increase in F). There was extensive discussions on which method to apply for provid-ing advice for 2014 and defining a biological reference point:

• First version: Advice for MSY-transition (uncertainty cap not considered rel-evant). Reduce F by 33%. The method does not quite fit into any of the earlier defined DLS methods. The case is considered rather straightforward: there is a reliable F time-series and a defined FMSY (which is considerably below recent three year average F). The time-series of biomass estimates could also be used, but is consider less reliable (recreational catch not included) and no biomass reference points are established.

• Revisited version: Due to concerns about the absolute values (as stated in the benchmark report), the ADGCS decided to apply DLS 3.2.0, a method based on trends only. Both the uncertainty cap and pa-buffer are applied, resulting in a 36% reduction.

There was an extensive discussion at the ADG on whether or not to explicitly advise for effort management in addition/as opposed to a catch constraint.

The revised SS3 model proposed by IBPBass now includes an explicit recognition of the possible recreational fishing mortality, albeit as a fixed vector of F at-age following the same selectivity as the commercial line fishery. This leads to a more complex prob-lem in defining BRPs based on yield-per-recruit because the fishing mortality now has separate commercial and recreational F components which can be manipulated sepa-rately or simultaneously in a yield-per-recruit analysis. The SS3 model is fitted with stock–recruit steepness fixed at 0.999 therefore cannot provide MSY reference points conditioned on a fitted stock–recruit curve. There are insufficient observations at low SSB to suggest the possible steepness of the relationship (Figure 6.1.1).

The recreational F vector is indicative only, in the sense that it is conditioned on a total annual recreational removals estimate of ~1500 t for recent years, which can be consid-ered as a “plausible scenario” rather than an explicit estimate of recreational F with recreational fishery survey data and their precision included in the model fitting. Dif-ferent scenarios for total recreational removals affect how total fishing mortality is split between commercial and recreational F, but the combined F estimates are minimally affected as they are driven by the fishery composition data (See Section 4, Figure 4.2.16).


Figure 6.1.1. Estimates of SSB and recruitment (age 0) from the final SS3 run conditioned on a steep-ness of 0.999 (see Section 4, Figure 4.2.12, for plots based on assessment runs 20a-l with steepness values fixed at 0.5, 0.8 or 0.999).

6.2 Potential MSY BRPs

BRPs for this assessment based on advice from ICES WKMSYREF2 (ICES, 2014) in-clude:

1 ) Setting an FMSY proxy as F0.1 or Fxx%spr based either on the commercial fishery only, or the combined commercial and recreational F.

2 ) Setting a BMSY trigger around a low percentile of the expected range of SSB when fishing at FMSY.

These options are explored below, based on the final assessment run.

Unfortunately it has not been possible in the time available to carry out a full MCMC bootstrap of the sea bass assessment and to propagate this into a forecast period to evaluate the percentiles of expected SSB while fishing at FMSY. A concern with sea bass is that recruitment has shown longer term changes in mean recruitment (and hence stock productivity) that appear related to changes in sea temperature at decadal scales.

6.2.1 Yield-per-recruit reference points

The inputs for a yield-per-recruit analysis are given in Table 6.2.1. They are identical with the short-term forecast inputs except that the values are extended to 30 years of age. SS3 provides population estimates out to this age. It is assumed that no fish sur-vive after the 30th year, as no fish older than 28 years have been observed historically.


Table 6.2.1. Inputs to yield-per-recruit analysis. F(5–11) for commercial fleets is the 2010–2012 aver-age.

The yield-per-recruit analysis can be configured to represent a range of scenarios for linked or independent management of fishing mortality for the commercial fishery and the recreational fishery. Only two scenarios are explored here:

1 ) A range of F-multipliers from 0 to 2 is applied in 0.02 steps to both the com-mercial and recreational Fs calculated for the 2010–2012 period.

2 ) A range of F-multipliers from 0 to 2 is applied only to the commercial fishery except at Fmult = 0, which is applied to both fisheries in order to obtain the SSB per recruit in absence of fishing (which is the same for both scenarios). For all commercial F multipliers >0.0, an F-multiplier of 1.0 is applied to the recreational fishery.

Age M Pmatstock

wt (kg)

F(5-11): commercia l

fleets

F(5-11): recreational

fi shery

Catch wt (kg): commercia l

fleets

Catch wt (kg): recreational

fleets

0 0.15 0.000 0.002 0.000 0.000 0.007 0.0071 0.15 0.000 0.022 0.000 0.000 0.078 0.0522 0.15 0.000 0.101 0.000 0.000 0.386 0.1543 0.15 0.000 0.219 0.012 0.002 0.581 0.2954 0.15 0.187 0.383 0.062 0.013 0.714 0.4805 0.15 0.421 0.588 0.140 0.054 0.877 0.7026 0.15 0.640 0.827 0.192 0.091 1.094 0.9547 0.15 0.793 1.091 0.217 0.099 1.334 1.2288 0.15 0.885 1.374 0.228 0.100 1.592 1.5179 0.15 0.937 1.669 0.236 0.100 1.865 1.815

10 0.15 0.965 1.970 0.241 0.100 2.149 2.11611 0.15 0.981 2.272 0.243 0.100 2.439 2.41612 0.15 0.989 2.570 0.244 0.100 2.728 2.71113 0.15 0.993 2.862 0.245 0.100 3.011 2.99814 0.15 1.000 3.145 0.245 0.100 3.287 3.27515 0.15 1.000 3.418 0.246 0.100 3.552 3.54016 0.15 1.000 3.678 0.246 0.100 3.804 3.79317 0.15 1.000 3.925 0.246 0.100 4.044 4.03218 0.15 1.000 4.158 0.246 0.100 4.270 4.25819 0.15 1.000 4.377 0.246 0.100 4.482 4.46920 0.15 1.000 4.582 0.246 0.100 4.680 4.66721 0.15 1.000 4.774 0.246 0.100 4.865 4.85222 0.15 1.000 4.952 0.246 0.100 5.036 5.02323 0.15 1.000 5.117 0.246 0.100 5.195 5.18124 0.15 1.000 5.270 0.246 0.100 5.343 5.32825 0.15 1.000 5.412 0.246 0.100 5.478 5.46426 0.15 1.000 5.542 0.246 0.100 5.603 5.58827 0.15 1.000 5.662 0.246 0.100 5.718 5.70328 0.15 1.000 5.772 0.246 0.100 5.824 5.80829 0.15 1.000 5.873 0.246 0.100 5.921 5.90530 0.15 1.000 6.085 0.246 0.100 6.124 6.107


In scenario 2, a constant recreational F(5–11) of 0.09 is applied in all circumstances other than a full moratorium. This means that at a commercial F multiplier of 1.0, the total F is split between commercial and recreational fishing in the same way as in 2010–2012, but for other multipliers the ratio of commercial to recreational F varies.

The detailed YPR output for scenarios (1) and (2) are given in Table 6.2.2a,b. In scenario (1), both the commercial and recreational YPR rise to a plateau as the combined F is increased (Figure 6.2.1a). The FMAX is not definable as there is no clear peak in the YPR curve. The F35%spr value (combined commercial and recreational) is around 0.13, which is also the value for F0.1. This is close to the value of M (0.15).

In scenario (2), the combined F does not fall below 0.09 except in a complete morato-rium. As the commercial F multiplier is increased in this scenario, the commercial YPR increases to a plateau while the recreational YPR declines, because the combined F be-comes increasingly dominated by the commercial fishery (Figure 6.2.1b, left hand plot). The F0.1 and F35%spr values, calculated as the combined commercial and recreational F, remain the same as in scenario (1) at F=0.13. However, the split between commercial and recreational F at F0.1 and F35%spr is opposite in the two scenarios. In scenario (1), the split is commercial F = 0.09 and recreational F = 0.04. In scenario (2), the recreational F is 0.09 at all commercial F multipliers >0, so the split at F0.1 and F35%spr is the opposite to scenario (1) with commercial F = 0.04 and recreational F = 0.09.

The commercial YPR in scenario (2) can also be plotted as a function of the commercial F component (Figure 6.2.1b, right hand plot), showing again that the commercial F at F0.1 and F35%spr is only around 0.04 for this scenario (recreational F = 0.09).

The SSB per recruit at F35%spr is 2.29 kg for both scenarios, which implies that SSB would vary around a very large value of ~22 600 t in the long term, if recruitment varies around the long-term arithmetic mean from the SS3 assessment (9900 thousand fish). This is well above the 2013 point estimate of SSB (13 050 t) and above any historical SSB (Figure 6.3.2.1). This conclusion is uncertain, as there are no observations of recruit-ment at such high SSB and it is unknown what density-dependent processes may op-erate. Historical recruitment in sea bass has also shown extended periods of above-average or below-average recruitment, with occasional very large or very weak year classes.

For scenario (1), the commercial yield-per-recruit at F35%spr is around 0.26 kg which im-plies long-term average commercial yield of 2590 t. For scenario (2), where the com-mercial F at the reference point is lower at F=0.04, the commercial yield-per-recruit is 0.12 kg implying long-term average commercial yield of 1170 t. The long-term recrea-tional yield for scenario (1) is 1010 t and for scenario (2) is 2350 t, this difference again reflecting the different split of the commercial and recreational F at the F reference point. Once again it must be emphasized that the relative contribution of commercial and recreational fisheries to total F is only known approximately, based on some recent recreational fishery surveys and given the uncertainties in the commercial landings figures for some fleets. However the examples given show the added complexity of defining management goals for fisheries where the resource is shared between com-mercial and recreational fisheries.

The F35%spr value estimated by WGCSE in 2013 was around 0.18, reflecting the larger M value of 0.2 used in the assessment. The IBPBass estimate of F35%spr is reduced to 0.13, due to the reduction in input M from 0.20 to 0.15. This implies a less productive stock, and hence a lower fishing mortality associated with a given SSB depletion.


The use of YPR reference points as proxy for FMSY reflects only the aspects of produc-tivity related to growth, natural mortality and age at maturity. The slow growth in sea bass, together with delayed maturity and low natural mortality (as indicated by lon-gevity to nearly 30 years) implies that sustainable fishing mortality is likely to be rela-tively low. The other component of productivity, the shape and steepness of the stock–recruit relationship, is not accounted for, although this is also a critical component to consider in relation to sustainability in general and MSY in particular. The shape is currently not definable given the very wide variations in recruitment and the occur-rence of different recruitment regimes in the dataseries that appear related to trends in environmental conditions. This leads to additional uncertainty in MSY reference points.

6.2.2 BMSY trigger point

As the shape of the stock–recruit curve is not definable, no BMSY trigger could be esti-mated. This is considered further in relation to precautionary biomass reference points.


Table 6.2.2. Detailed yield-per-recruit output for two scenarios for applying F-multipliers to the commercial and recreational fishery: (a) same multipliers applied to both; (b) constant F applied to recreational fishery for all commercial F-multipliers greater than zero.

a. Scenario 1

b. Scenario 2

Commercial fishery Fmult

Recreational fishery Fmult

F(5-11) total

F(5-11) commercial

F(5-11) recreational SSB/R

% SPR/SPR(F0)

YPR-commercial

YPR recreational

YPR-total international

0 0 0.000 0.000 0.000 6.530681 100% 0.000 0.000 0.0000.1 0.1 0.031 0.021 0.009 4.835607 74% 0.125 0.050 0.1750.2 0.2 0.061 0.043 0.018 3.720487 57% 0.194 0.077 0.2710.3 0.3 0.092 0.064 0.028 2.957745 45% 0.233 0.092 0.3250.4 0.4 0.122 0.086 0.037 2.41693 37% 0.257 0.101 0.357

0.43 0.43 0.132 0.092 0.040 2.285496 35% 0.262 0.102 0.364 F35%spr & F0.10.5 0.5 0.153 0.107 0.046 2.020865 31% 0.271 0.105 0.3760.6 0.6 0.184 0.128 0.055 1.722395 26% 0.280 0.108 0.3880.7 0.7 0.214 0.150 0.064 1.491782 23% 0.285 0.109 0.3950.8 0.8 0.245 0.171 0.074 1.309688 20% 0.289 0.110 0.3980.9 0.9 0.275 0.193 0.083 1.163169 18% 0.291 0.109 0.4001 1 0.306 0.214 0.092 1.043323 16% 0.292 0.109 0.40127

1.1 1.1 0.336 0.235 0.101 0.943879 14% 0.293 0.108 0.401311.2 1.2 0.367 0.257 0.110 0.860318 13% 0.293 0.107 0.400831.3 1.3 0.398 0.278 0.120 0.789315 12% 0.294 0.106 0.4001.4 1.4 0.428 0.299 0.129 0.728384 11% 0.293 0.105 0.3991.5 1.5 0.459 0.321 0.138 0.675631 10% 0.293 0.104 0.3981.6 1.6 0.489 0.342 0.147 0.629594 10% 0.293 0.103 0.3971.7 1.7 0.520 0.364 0.156 0.589127 9% 0.293 0.102 0.3951.8 1.8 0.551 0.385 0.166 0.553324 8% 0.293 0.101 0.3941.9 1.9 0.581 0.406 0.175 0.521459 8% 0.292 0.100 0.3932 2 0.612 0.428 0.184 0.492945 8% 0.292 0.099 0.391

Commercial fishery Fmult

Recreational fishery Fmult

F(5-11) total

F(5-11) commercial

F(5-11) recreational SSB/R

% SPR/SPR(F0)

YPR-commercial

YPR recreational

YPR-total international

0 0 0.000 0.000 0.000 6.530681 100% 0.000 0.000 0.0000.1 1 0.113 0.021 0.092 2.594142 40% 0.069 0.269 0.3380.2 1 0.135 0.043 0.092 2.268055 35% 0.121 0.236 0.357 F35%spr & F0.10.3 1 0.156 0.064 0.092 2.004465 31% 0.161 0.209 0.3700.4 1 0.178 0.086 0.092 1.788372 27% 0.193 0.187 0.3800.5 1 0.199 0.107 0.092 1.608937 25% 0.219 0.168 0.3870.6 1 0.220 0.128 0.092 1.458208 22% 0.240 0.152 0.3920.7 1 0.242 0.150 0.092 1.33026 20% 0.257 0.139 0.3960.8 1 0.263 0.171 0.092 1.22062 19% 0.271 0.128 0.3980.9 1 0.284 0.193 0.092 1.12586 17% 0.282 0.118 0.4001 1 0.306 0.214 0.092 1.043323 16% 0.292 0.109 0.40127

1.1 1 0.327 0.235 0.092 0.970924 15% 0.301 0.101 0.402091.2 1 0.349 0.257 0.092 0.907007 14% 0.308 0.094 0.402571.3 1 0.370 0.278 0.092 0.850247 13% 0.314 0.088 0.402801.4 1 0.391 0.299 0.092 0.799571 12% 0.320 0.083 0.402831.5 1 0.413 0.321 0.092 0.754101 12% 0.325 0.078 0.402711.6 1 0.434 0.342 0.092 0.713116 11% 0.329 0.074 0.402481.7 1 0.456 0.364 0.092 0.676018 10% 0.332 0.070 0.402161.8 1 0.477 0.385 0.092 0.642306 10% 0.336 0.066 0.401781.9 1 0.498 0.406 0.092 0.611559 9% 0.339 0.063 0.401342 1 0.520 0.428 0.092 0.583422 9% 0.341 0.060 0.40087


a. Scenario 1

b. Scenario 2

Figure 6.2.1. Yield and spawning biomass per recruit curves for two scenarios for applying F-mul-tipliers to the commercial and recreational fishery: (a) same multipliers applied to both; (b) constant F applied to recreational fishery for all commercial F-multipliers >0 (left hand plot: YPR for com-mercial and recreational fisheries, plotted against total combined F; right hand plot: commercial and recreational YPR plotted against commercial F only).

6.2.3 Limit and precautionary reference points

Blim could be defined as a previous historical low value from which the stock has re-covered. Given the delayed maturity in sea bass (50% maturity in females at approxi-mately six years), total-stock biomass may be a more sensitive metric to give advance warning of the need for management measures to conserve SSB. The previous historic low value for total biomass (since 1985) was 9900 t in 1990. This was followed in 1992 by the historic low SSB of 5200 t (Figure 6.2.3.1). It is however more conventional to use SSB for defining limit and precautionary reference points for biomass. The Stock Syn-thesis inverse-hessian estimate of standard deviation of spawning–stock biomass (for fixed recreational F vector) in 2013 is approximately 20% relative to the mean, equiva-lent to 2700 t. This uncertainty in SSB estimates could be used through MCMC projec-tions to quantify the probability of the stock falling below Blim, and provide a basis for defining a suitable precautionary SSB biomass. As this was not possible during IBPBass, no precautionary reference points could be established, and only a Blim for SSB was considered. This is the lowest observed biomass (5200 t).


6.2.4 Candidate biological reference points

Candidate values for reference points are given in Table 6.2.3.1 and plotted on Figure 6.2.3.1 in comparison with historical SSB and F trends. The time-series of historical total F (estimated commercial and fixed recreational F(5–11) combined) has increased grad-ually over time, but the all the values lie above the F35%spr reference point, which is also close to or the same as the F0.1. Almost all historic F values lie along the plateau of the YPR curve (see Figure 6.2.1), implying that the progressive increase in F over time has not provided any benefits in terms of increased yield-per-recruit, but have caused sub-stantial depletion in SSB per recruit to well below 35% of the unexploited value. High values of F also mean a greater truncation of the age composition, which is an issue for recreational sea angling where the availability of large fish to catch is an important objective.

There is currently no TAC for sea bass, and control of fishing mortality would have to be through other approaches to managing effort on sea bass, and including technical measures to alter selectivity and/or restrict fishing seasonally or spatially. Note that the inclusion of discards in the assessment would alter the reference points and historical series. In the absence of discards, it is difficult to infer benefits to YPR or SSB/R in im-proving the selectivity patterns of the fleets. There is currently no time-series of recre-ational fishing catches to monitor the impacts of any management measures, and the frequency and extent of future surveys remains uncertain.

Table 6.2.3.1. Potential biological reference points for sea bass.

TYPE VALUE TECHNICAL BASIS

Precautionary Approach

Blim 5200 t SSB: lowest observed (1992)

Bpa Not defined

Flim Not defined

Fpa Not defined

MSY approach FMSY 0.13 F35%spr (equivalent to F0.1)

BMSY trigger Not defined


Figure 6.2.3.1. Long-term stock and F trends for sea bass plotted with potential reference points.


7 References

Alagaraga, K. 1984. Simple methods for estimation of parameters for assessing exploited fish stocks. Indian Journal of Fisheries, 31: 177–208.

Alverson, D. L., and M. J. Carney. 1975. A graphic review of the growth and decay of population cohorts. J. Cons. Int. Explor. Mer 36: 133–143.

Armstrong, M. and Walmsley, S. 2012. Life-history estimates of natural mortality of sea bass around the UK. Working Document: ICES IBPNEW 2012.

Armstrong, M.J. and Walmsley, S. 2012. An evaluation of the bass fleet census and logbook sys-tem for estimating annual landings by gear for fishing vessels in England and Wales. Work-ing Document to ICES IBP-NEW 2012. September 2012.

Armstrong, M. and Readdy, L. 2014. Effect on sea bass Stock Synthesis model of expanding the UK fishery age compositions to a larger plus group. Working Document, ICES IBPBASS, 2014.

Beverton, R. J. H. 1992. Patterns of reproductive strategy parameters in some marine teleost fishes. J. Fish Biol. 41(Supplement B): 137–160.

Coppin, F., Le Roy, D., Schlaich, Y. 2002. Manuel des protocoles de campagne halieutique: Cam-pagnes CGFS, Système d'information halieutiques - Campagnes à la mer. Ifremer, 09/2001 – DRV/RH/DT/AN-NUMERO, 29pp. (in French).

Diodati, P. and R.A. Richards. 1996. Mortality of striped bass hooked and released in salt water. Transactions of the American Fisheries Society. 125: 300–307.

Hewitt, D.A. and J.M. Hoenig. 2004. Comparison of two approaches for estimating natural mor-tality based on longevity.

Hoenig, J. M. 1983. Empirical use of longevity data to estimate mortality rates. Fish.Bull. 82: 898–903.

ICES. 2012a. Report of the Inter-Benchmark Protocol on New Species (Turbot and Sea bass; IBPNew 2012), 1–5 October 2012, Copenhagen, Denmark. ICES CM 2012/ACOM:45. 239 pp.

ICES. 2012b. Report of the Working Group on Recreational Fisheries Surveys (WGRFS). ICES CM 2012/ACOM:23. 55 pp.

ICES. 2013a. Report of the Working Group for Celtic Seas Ecoregion (WGCSE), 8–17May 2013, Copenhagen, Denmark. ICES CM 2013/ACOM:12. 1986 pp.

ICES. 2013. Report of the Working Group on Recreational Fisheries. ICES CM 2013/ACOM:23.

ICES. 2014. Report of the Workshop to consider reference points for all stocks (WKMSYREF2). ICES CM 2014/ACOM:47.

Kupschus, S., Smith, M. T., Walmsley, S. A. 2008. Annex 2: Working Document. An update of the UK bass assessments 2007. Report of the Working Group on the Assessment of New MoU Species (WGNEW). By Correspondence, ICES CM 2008/ACOM:25. 77 pp.

Mangel, M., MacCall, A.D., Brodziak, J., Dick, E.J., Forrest, R.E., Pourzand, R. and Ralston, S. 2013. A perspective on steepness, reference points, and stock assessment. Can. J. Fish. Aquat. Sci. 70: 930–940 (2013) dx.doi.org/10.1139/cjfas-2012-0372.

Methot, R.D. 2000. Technical Description of the Stock Synthesis Assessment Program. National Marine Fisheries Service, Seattle, WA. NOAA Tech Memo. NMFS-NWFSC-43: 46 pp.

Methot, R.D. 2011. User Manual for Stock Synthesis, Model Version 3.23b. NOAA Fisheries Ser-vice, Seattle. 167 pp.

Pauly, D. 1980. On the interrelationships between natural mortality, growth parameters, and mean environmental temperature in 175 fish stocks. J. Cons. Int. Explor. Mer 39: 175–192.


Pawson, M.G. 1992. Climatic influences on the spawning success, growth and recruitment of bass (Dicentrarchus labrax L.) in British waters. ICES Marine Science Symposia 195: 388–392.

Pawson, M. G., Kupschus, S. and Pickett, G. D. 2007a. The status of sea bass (Dicentrarchus labrax) stocks around England and Wales, derived using a separable catch-at-age model, and im-plications for fisheries management. ICES Journal of Marine Science 64, 346–356.

Pickett, G. D., Eaton, D.R., Cunningham, S., Dunn, M.R., Potten, S.D. and Whitmarsh, D. 1995. An appraisal of the UK bass fishery and its management. Lab Leafl. MAFF Direct. Fish. Res., Lowestoft, (75): 47pp.

Ralston, S. 1987. 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. 375–404.

Rocklin D, Levrel H, Drogou M, Herfaut J, Veron G. 2014. Combining Telephone Surveys and Fishing Catches Self-Report: The French Sea Bass Recreational Fishery Assessment. PLoS ONE 9(1): e87271. doi:10.1371/journal.pone.0087271.

Ifremer. 2011. Synthèse des flottilles de pêche 2009 - Flotte Mer du Nord - Manche - Atlantique - Méditerranée. Edited by Leblond Emilie, Daures Fabienne, Berthou Patrick, Merrien Claude, Pitel-Roudaut Mathilde, Macher Claire, Le Grand Christelle, Demaneche Sebastien, Jezequel Michele, Bodere Erwan, Le Blond Samuel. 2011.http://archimer.ifre-mer.fr/doc/00042/15352/.

Tulp, I., Bolle, L. J. and Rijnsdorp, A. D. 2008. Signals from the shallows: In search of common patterns in long-term trends in Dutch estuarine and coastal fish. Journal of Sea Research 60 (1–2), pp. 54–73.

Van Damme, C., Bakker, K., Bolle, L., de Boois, I., Couperus, B., van Hal, R., Heessen, H., Hoek, R., Fässler, S. 2014. Handboek en protocollen voor bestandsopnamen en routinematige bemonsteringen op het water. Versie 8, december 2013.

van der Hammen, T and de Graaf, M. 2012. Recreational fishery in the Netherlands: de-mographics and catch estimates in marine and freshwater. IMARES Wageningen UR, Report Number C147/13.

van der Hammen, T., Poos, J. J., van Overzee H. M. J., Heessen, H. J. L. and Rijnsdorp A. D. 2013. Data evaluation of data limited stocks: Horse mackerel, Sea bass, Greater Silver Smelt, Tur-bot and Brill. IMARES report number C166/13.

van Beek, F. A., Rijnsdorp, A. D., de Clerck, R. 1989. Monitoring juvenile stocks of flatfish in the Wadden Sea and the coastal areas of the southeastern North Sea. Helgoländer Meeresunter-suchungen, 43 (3–4), pp. 461-477. doi: 10.1007/BF02365904.

Walmsley, S. and Armstrong, M. 2012. The UK commercial bass fishery in 2010. Working Docu-ment to ICES WGNEW 2012. August 2011.


Annex 1: Stock annex

Stock-specific documentation of standard assessment procedures used by ICES.

Stock European sea bass (Dicentrarchus labrax) in Subarea IVb,c and VIIa, d–h (BSS-47)

Working Group WGCSE

Date May 2014

Revised by Mike Armstrong, May 2014

Revisions v1. WGNEW (bass in all areas)

v2. IBPNEW 2012 (retaining only information for BSS-47)

v3. WGCSE 2013 with revisions

v4. WGCSE 2014 with revisions from IBP-bass

A. General

A.1. Stock definition

Bass Dicentrarchus labrax is a widely distributed species in Northeast Atlantic shelf wa-ters with a range from southern Norway, through the North Sea, the Irish Sea, the Bay of Biscay, the Mediterranean and the Black Sea to Northwest 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, dis-tribution of commercial catches and similarities in stock trends between areas, drawing also on extensive information contained in previous WGNEW and ICES SGBASS re-ports.

IBP-NEW and WGCSE considers that stock structure remains uncertain, and recom-mends further studies on sea bass stock identity, using conventional and electronic tagging, genetics and other individual and population markers (e.g. otolith micro-chemistry 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 (e.g. temperature and food). Such a modelling work is being carried out in France in the framework of a PhD study (R. Lopez), and in the UK.

The pragmatic view of IBP-NEW 2012 was 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). Rel-atively data-rich area with data on fishery landings and length/age compo-sition; discards estimates and lengths; growth and maturity parameters; juvenile surveys, fishery lpue trends. [Bass-47].


• 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. [Bass-wosi].

Fishery landings of sea bass are extremely small in Irish coastal waters of VIIa and VIIg and the stock assessment for assessment area 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 prespawning 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 sea-men) 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 €31 937 in 2008. 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 it’s caught, giving added value to certain métiers: in 2013 according to the database SACROIS (Ifremer-DPMA), mean price of sea bass sold in France by liners was €17 per Kg compared with €7 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 Neth-erlands and Belgium. Relatively little historical data are available on recreational fish-eries 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 IBPBass (ICES 2014).

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 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 70–89 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 fish-

ing for bass in Irish waters; plus MLS of 40cm; • 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 (Paw-son, 1992). Reynolds et al. (2003) observed a positive relationship between annual sea-water 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 ad-dition, 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 tempera-ture (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).

1 Region 1: All waters which lie to the north and west of a line running from a point at latitude 48°N, 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 lati-tude 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.


B. Data

B.1. Commercial catch

B1.1 Landings data

Data available

Landings series for use in the assessment are available from three sources:

i ) Official statistics recorded in the ICES FishStat database since around the mid-1970s.

ii ) French landings for 1999–2013 from a separate analysis by Ifremer of log-book, auction and VMS data, with earlier official data adjusted as de-scribed below.

iii ) 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 2000 onwards Ifremer has provided revised French landings from a separate analysis of logbook and auction data and VMS which allocates landings correctly by fishing ground (SACROIS methodology used at present as Official Landings). Quality of data from 2000 can be considered as high. From 1985 to 2000, quality of data is poorer. To generate a consistent series of French landings from 1985 onwards for the Subarea IV and VII assessment, IBPNew 2012 (ICES, 2012a) adjusted pre 1999 official FishStat landings by the average of the Ifremer correction factors by area from 2000–2010. Factors for adjustments calculated for each area and applied to the French land-ings data from FishStat are: 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.

To split French landings by métier for IBPBass 2014, the ratio of landings among méti-ers was used to convert total French landings prior to 2000 estimated by ICES IBPNEW 2012 to French landings by métier. This was the best way to get consistent time-series of French landings. Moreover some uncertainties remain from 1985 to 2000 because the French landings of sea bass are probably underestimated particularly for small-scale inshore fisheries. 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 official reported landings of sea bass in the UK are known to underestimate the true total landings, particularly for small-scale inshore fisheries where there has been no requirement to submit EC logbooks. Prior to the introduction of Buyers and Sellers regulations in 2006 requiring sales documentation, local fishery inspectors estimated landings of the under-10 m fleet using whatever information they had available from auctions, and frequently entered aggregated estimates for multiple vessels into the fishery landings database. Unfortunately the Buyers and Sellers regulations do not cover all landings. The EU Control Regulation allows landings of less than 30 kg to be disposed of for personal consumption without providing sales slips or other docu-ments. This is ostensibly to reduce the administrative burden for a skipper disposing of small quantities for personal use. However, for small-scale fisheries where there are very large numbers of small vessels often catching small quantities, the cumulative catch of unrecorded small landings can be relatively high. This is likely to be an issue over the full time-series.

Due to the known inaccuracies in reported landings of such vessels, Cefas (UK) imple-mented an independent logbook scheme and port survey in England and Wales in 1985 to estimate mean cpue (annual landings per vessel, based on a logbook kept by selected skippers) and total number of vessels catching sea bass (from an annual port survey covering different stretches of coastline in successive years). Total bass landings were estimated as the product of cpue and vessel numbers. The scheme was stratified by area, gear and vessel characteristics. Selected vessels from the strata kept logbooks for periods ranging from one to 25 years, and comprised what could be described as a “reference” fleet as opposed to a randomised selection of vessels each year. The scheme is described by Armstrong and Walmsley (2012), who identified some issues with the survey related to overstratification and extensive imputation needed due to small and declining numbers of logbooks, and the non-random nature of vessel selection for log-books. Sufficient logbooks were available only for inshore vessels using fixed/driftnets or lines. The scheme was terminated in 2007 and 2008, and reinstated for a further two years (2009 and 2010) before being terminated again. The scheme has now been sus-pended permanently.

Despite the potential biases, the survey results for commercial vessels confirm that the historical official reported landings of sea bass are likely to be underestimates (Figure B1.1). For fixed/driftnets, the landings including the Cefas logbook estimates for under 10 m vessels results in a landings series that is on average around three times higher than the official statistics. For lines, the ratio fluctuates around 3.0 for a large part of the series but was larger from 2000–2005.


Figure B1.1. Top: estimates of sea bass landings for under-10 m UK netters and liners based on the Cefas logbook and port survey scheme. Bottom: ratio of landings of these gears including the Cefas logbook estimates for <10 m commercial vessels, and the total official reported landings of all UK vessels using these gears.

The historical differences between official landings statistics and Cefas logbook land-ings estimates for sea bass are hard to interpret because there are no diagnostics pro-vided on how representative the Cefas logbooks are of the overall fleets in each survey stratum, and because extensive imputations have been needed due to over-stratifica-tion. Precision is likely to be low due to the small number of logbooks, although this cannot be quantified. Although the logbook estimates are subject to bias and low pre-cision, a bias factor of ~3.0 (Figure B1.1) could be applied to the input landings in Stock Synthesis to evaluate its effect on the assessment trends and advice.

Dutch landings

Landings and effort data from the commercial fleet are available from the EU logbooks; market category composition of landings is available from the auction data (sale slips); and size and age data are available through market sampling. These data have been summarized in van der Hammen et al. (2013).

EU logbook data

Official EU logbook data of the entire Dutch fleet are maintained by the NVWA (for-merly known as the General Inspection Service, AID). IMARES has access to these log-books and stores the data in a database (VISSTAT). EU logbook data contain information on:

• landings (kg): by vessel, trip, ICES statistical rectangle and species; • effort (days absent from port): by vessel, trip and ICES statistical rectangle; • vessel information: length, engine power and gear used.


Logbook data are available from the entire Dutch fishing fleet and from foreign vessels landing their catches in the Netherlands.

Auction data: landings by market category

Auction data cover both the total Dutch fishing fleet and foreign vessels landing their catches on Dutch auctions. These data are also stored in VISSTAT and contain infor-mation on:

• landings by market category (kg): by vessel, trip (landing date) and species.

Further information on availability and quality of landings data by country is provided by IBPBass (ICES, 2014).

B1.2 Discards estimates

UK data

Survey design and analysis

Estimation of bass discards is problematic because the observer scheme covers all ves-sel trips in a stratum without reference to target species, and overall it samples less than 1% of all fishing trips. As bass are absent or at very small numbers in a large fraction of fishing trips throughout the year, particularly in winter, the amount of sam-ple data on bass is very low and the estimates are likely to have poor precision and variable biases related to inclusion of under 10 m vessels. Vessels under 10 m, that are responsible for a large fraction of bass catches, were excluded until recent years on health and safety and other logistical grounds, and although under 10 m vessels are now included, the fleet of vessels under 7 m remains excluded.

Two approaches to raising the UK discards data were considered in previous assess-ments: 1) a “design-based” estimator calculating sampling probability as the ratio of numbers of trips in the fleet to numbers of trips sampled within a stratum; 2) a ratio estimator using the ratio of reported landings in a stratum to the computed retained weight of sea bass in sampled trips. Neither of these approaches is unbiased, as the variable treatment of under 10 m vessels precludes reliable estimation of sampling probabilities for method (1) for many of the years, and also affects the assumption of method (2) that the retention curve for sampled vessels is representative of all vessels in the stratum. The non-reporting of potentially large quantities of sea bass landings in the under 10 m fleet, due to the dispensation from reporting trip landings under 25 kg also means that applying method (1) may lead to raised estimates of retained catch that are much higher on average than the reported landings for the fleet. Therefore the in-clusion of absolute estimates of discards from method (1) may lead to incorrect estima-tion of retention parameters and fishing mortality in Stock Synthesis. To correct for this would mean re-scaling the discards estimates according to the ratio of estimated to reported retained catch, which is equivalent to method (2). Method (2) – raising using ratio of reported to observed landings – was therefore used for sea bass to ensure in-ternal consistency in the handling of the data within Stock Synthesis, but many caveats remain concerning the accuracy of the estimates.

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.

Numerically, the largest numbers of bass discarded from UK fisheries are from the bottom-trawl fleet, with much smaller numbers from nets and even less from beam trawls. Only eleven midwater pair trawl trips have been observed, and discarding of bass was negligible as the fishery targets mainly adult bass. No bass discards were observed in the eight longline trips observed. The raised length frequencies of dis-carded and retained bass, aggregated over all years, are available along with the reten-tion ogives. It is clear that discarding is driven primarily by the minimum landing size of 36 cm (see IBPBass 2014 report).

French data

Survey design and analysis

The French sampling schemes also utilize vessel-list sampling frames and random se-lection of vessels within strata defined by area and fleet sector. From the activity cal-endars 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 on board 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 sam-pling 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: http://sih.ifremer.fr/content/download/5587/40495/file/Manuel_OB-SMER_V2_2_2012.pdf)


Discards data are only available for French fleets from 2009 onwards. For 2012, results are described in the annual French report on:" http://ar-chimer.ifremer.fr/doc/00167/27787/25978.pdf" for 2012. Discards estimates for France are from vessel selections that for some areas and gears include relatively limited num-bers of observed trips where sea bass is caught and discarded. Precision is therefore very low at current sampling rates. In France the low sampling rate observed can be explain by the low discarding rate. For information in 2013, only 26 tons would have been discarding in the studied area. Length frequencies were not available at IBPBass for 2013.

Spain

No bass discards were observed for any métier in the 2003-2013 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 an-nual mean numbers discarded per hour in 2004–2010. No commercial fisheries for sea bass exist in Ireland.

http://sih.ifremer.fr/content/download/5587/40495/file/Manuel_OBSMER_V2_2_2012.pdf





B1.3 Recreational catches

Recreational marine fishery surveys in Europe are still at an early stage in development (ICES WGRFS 2012b, 2013b). Survey design and methods of recreational catch estima-tion are described in IBPBass 2014.

Data from France

A survey of recreational fishers, focusing mainly on bass, was conducted between 2009 and 2011. Estimates of sea bass recreational catches were obtained from a panel of 121 recreational fishermen recruited during a random digit dialling screening survey of 15 000 households in the targeted districts. 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 estimates for Subarea IV and VII were 940 t kept and 332 t released.

The precision of the combined Biscay and Channel estimate was relatively low (CV=26%; note that the figure of 51% given in IBPNEW 2012 (ICES, 2012a) was incor-rect). This gives an average and 95% confidence intervals of 3170 t [1554 t; 4786 t] for the whole Subareas IV, VII and VIII. Increasing the panel from 121 to 210 fishermen would be expected to improve precision to 20% and increasing this panel to 500 would improve precision to 13%.

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 rec-reational catch was taken by sea angling (rod and line or handline).

Taking into account a potential hooking mortality of 20%, the estimate of annual French recreational fishery removals from Areas IV and VII in 2009–2011 is increased to just over 1000 t.

A new survey was conducted from July 2011 to December 2012, based on a similar methodology to the previous study (not only on sea bass this time, but also on other marine species including crustaceans and cephalopods). A random digit dialling screening survey of 16 130 households led to the recruitment of a panel of 183 fisher-men to keep logbooks. In parallel, 151 fishermen were recruited on site by the Promo-peche association, and 30 more via the sea bass fishermen panel set up in 2009. This resulted in 364 panel members keeping logbooks describing their catches (species, weight, size, etc.) The focus of the survey on sea bass shows that in Atlantic (Bay of Biscay and Channel), the estimated recreational catch of bass in 2012 was 3922 t of which 3146 t was kept and 776 t released. At this time results have to be considered as provisional, (results split between Bay and Biscay and Channel are not available yet with relative standard error).

UK (E&W)

A new survey programme Sea Angling 2012, based on a statistically sound survey de-sign started 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. Results are available at http://www.marinemanage-ment.org.uk/seaangling/.

The surveys adopted, where possible, statistically sound, probability-based survey de-signs, building on knowledge gained through participation in the ICES Working Group on Recreational Fishery Surveys (WGRFS). Two survey approaches were adopted: first a stratified random survey of charter boats from a list frame covering ports in England, and secondly an on-site stratified random survey of shore anglers




and private boat anglers to estimate mean catch per day, combined with annual effort estimates derived from questions added to a monthly Office of National Statistics household survey covering Great Britain.

A list of almost 400 charter boats was compiled for the charter boat survey, and 166 skippers agreed to participate. Each month over a twelve month period in 2012 and 2013, 34 randomly selected skippers completed a diary documenting their activities, catches and sizes of fish. A diary was completed whether or not any fishing took place. Data from 5300 anglers were collected. Total annual catches were estimated by raising the monthly catches per vessel from the diaries to all vessel-month combinations in the frame, and raising this to all vessels including refusals. The estimated total annual catch of sea bass for the entire coast of England was 44 t (RSE 31%) of which 31 t was kept. The release rate by number was 37%. The charter boat survey has potential bias due to the large non-response rate, if non-respondents have different catch rates to respond-ents.

The Office of National Statistics (ONS) household survey covered 12 000 households during 2012, and from this it was estimated that 2.2% of adults over 16 years old went sea angling at least once in the previous year. The surveys estimated there are 884 000 sea anglers in England. Estimation of fishing effort by shore and private boat anglers proved very difficult due to the overall small number of households with sea anglers in the survey. A range of methods was explored to estimate annual and seasonal effort using the ONS data alone, and combining it with observations from on-site and online surveys. It has not been possible yet to agree on a best estimate of effort, and for that reason the estimates of total catch (cpue × effort) for shore and private boat angling are given as a range of plausible values.

The survey of anglers fishing from the shore and private boats to estimate cpue was carried out throughout 2012 using on-site interviews. A stratified random design was adopted to select shore sites and boat landing sites on a weekly basis from site lists stratified into low-activity and high-activity sites. The shore survey used roving-creel methods (collecting data from partial angling trips), and the private boat survey a rov-ing access-point survey (data from completed trips). Visits were made to 1475 shore sites and 425 private boat sites, and 2440 anglers were interviewed. The mean daily catch rate of kept and released fish of each species was estimated based on the survey design, and sizes of caught fish were recorded. Cpue for shore angling was estimated using catches for the observed trip duration and estimates of expected total trip dura-tion for that day. A length-of-stay bias correction was applied based on expected total trip duration. The catch-per-day estimates were combined with estimates of total an-nual fishing effort (days fished) obtained from the ONS survey to estimate total annual catches. Release rates, by number, were 82% for shore angling and 57% for private boats. Non-response rates were very low (<10%) in this survey. The range of point es-timates for shore-caught bass was 98–143 t (total) and 38–56 t (kept), and for private and rented boats was 194–546 t (total) and 142–367 t (kept). The relative standard errors for the individual shore and private boat estimates were relatively high at 40–50%.

Combining the catch estimates for charter boats, private boats and shore angling, the point estimates of annual kept weights of sea bass ranged from 230 t–440 t (Table 2.4.1c), compared with total UK commercial landings of almost 900 t in 2012. The com-bined estimates of bass catches had precision (relative standard error) estimates of 26%–38%for the different effort estimation methods. The relatively large standard er-rors combined with the range of plausible methods of estimating effort for shore and private boats.


Netherlands

Sea bass are taken by recreational sea anglers in the Netherlands. A recent survey in-vestigated the amount of sea bass caught by recreational fishers (van der Hammen and de Graaf, 2012; ICES, 2012b) from March 2010 to February 2011. Estimates of sea bass catches recreational were obtained from a panel of 1043 recreational fishermen re-cruited during a telephone survey of 109 293 people. Revised estimates were provided to WGCSE 2013 (ICES, 2013a). The catch weights are estimated with a limited amount of length–frequency data, and are therefore less reliable than the estimates in numbers (and may also be adjusted if more data are available). For the same reason, there are no ‘returned’ estimates by weight (yet).

The estimated total recreational catch of sea bass was 366 000 fish (RSE 30%), of which 234 000 were retained, equivalent to 138 t. These results are mainly applicable to Sub-area IV.

Other countries

Sea bass are a popular angling species in Ireland and are also caught in Belgium. A survey in Belgium in 2013, using questionnaires, estimated 60 tons of bass caught and retained. 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 WGCSE 2014 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 sam-pling 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 days, cur-rently 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 es-tablished for pelagic trawlers, beam trawlers and demersal trawl, nets and lines. Sam-pling 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 sam-pling out.

Length compositions are first vessel-raised using ratios of landed live weight to pre-dicted live weight of the length frequency calculated from a length–weight relation-ship:

W (kg) = 0.00001296 (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 en-sures sums-of-products ratios of 1.0, but will lead to some bias in numbers-at-length due to discrepancies between true fish weights and calculated fish weights from the length–weight relationship.


Age compositions for UK commercial fishery landings of sea bass are derived from biennial (January–June and July–December) age–length keys (ALK) constructed for four areas: IVb,c; VIId, VIIe,h and VIIa,f,g. These are applied to fleet-raised landings length frequencies for each of four gear groups (bottom trawls; midwater trawls; fixed/driftnets and lines) in each area. Further details are given in the ICES IBP-NEW (ICES, 2012a) and WGCSE (ICES, 2013a) reports and in the stock annex along with ta-bles giving numbers of trips sampled for length and age and numbers of fish measured and aged.

A recommendation of WGCSE 2013 was to expand the UK age frequencies to the full recorded age range and to re-evaluate the plus-group definition (previously at 12+). Sea bass have been recorded to almost 30 years of age, and it was thought that having more true ages in the Stock Synthesis input data could allow better estimates of early recruit deviations. The necessary extractions were done for IBPBass, and the data were examined in detail by Armstrong and Readdy (IBPBass Working Document_01). Bub-ble plots and catch curves showed that coherent information on year classes was pre-sent well beyond the last true age (11) previously adopted. The exploratory SS3 runs show that the different choices of plus-group (12+; 16+; 18+; 20+) have relatively little impact on the results, other than (as hoped) a slightly better estimation of early recruit deviations. Expanding the age compositions may help fit domed selection curves for fleets where this is appropriate, but risks an increasing number of zero catch entries for older ages as recent weak year classes feed into future catches and become depleted. A plus-gp of 16+ was recommended for further model development and agreed by IBPBass. Sampling of midwater trawls prior to 1996, and in 1997, was considered too poor to develop age compositions. All datasets show a very strong 1989 year class and very weak 1985–1987 year classes.

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 compo-sition and their relative importance per fishing harbours and month. Both are also de-signed 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, spe-cific 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, re-fusal rates and their reasons, time taken to sample, all this information is also available from the website, together with sampling protocol (in French:

http://sih.ifremer.fr/content/download/5587/40495/file/Manuel_OB-SMER_V2_2_2012.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 sam-pled 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 2010. For other areas the information is not available. All length samples are populated in a cen-tral database (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.


Length compositions of French sea bass commercial landings are constructed for four areas from 2009 to 2012 (IVb,c; VIId, VIIe,h and VIIa,f,g), for four métiers : GTR_DEF; LLS_DEF; OTB_DEF and OTM_DEF. It was not possible to disaggregate total length compositions, from 2000 to 2008. Only numbers-at-length for all gears are available from 2000 to 2009. The input data for French fleets in Stock Synthesis are the fleet-aggregated length compositions in 2-cm classes (20–21.9, 22–23.9, etc.) for each year from 2000 onwards.

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 num-bers of fish measured per trip.

Other countries

Fishery landings length or age compositions from other countries catching bass were not available to IBPBass 2014. The Netherlands did collect age samples of sea bass every year from 2005 to 2008. From 2010 onwards, age samples are collected only once every three year. In the IMARES market sampling data on length, age, sex and weight are collected for several commercially important species. For sea bass this is done on an irregular basis and data are only available for some years. Market sampling is done since 2005. The age sampling frequency is now set triennially (2010, 2013, etc.) Every three years four samples of 15 fish (60 fish in total) are aged, and every year the lengths of 24 samples of 15 fish (360 fish in total) are taken. 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, ESS for the combined UK trawls, nets and lines fleet are set at 100 for each year (20 for 1985–1990), and for the French combined-fleet LFDs, the ESS is set at 200 for each year.


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 labora-tories 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 17–74 cm was sampled on board 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 opera-tors. However it was noted by the operators that photographic images were more dif-ficult 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 2012.

Age validation

WGCSE and IBPBass were 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 fol-lowed 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 er-rors. 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 con-ducted 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 90 000 sea bass have been aged since 1985. 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-at-age.

All ages for fitting growth curves are referred to a nominal January 1 birthdate, accord-ing to month of capture. Parameters of the von Bertalanffy growth curve were fitted in Excel Solver using nonlinear minimisation of ∑(obs-exp)^2 for lengths-at-age of indi-vidual fish, by area and for all data combined.

Von Bertalanffy model parameters were as follows:

AREA IVBC VIID VIIE VIIAFG ALL AREAS

Linf (cm) 82.98 87.22 92.27 81.87 84.55

K 0.1104 0.09298 0.07697 0.09246 0.09699

t0 (years) -0.608 -0.592 -1.693 -1.066 -0.730

Standard deviation of length-at-age distributions increases linearly with age according to:

SD (age) = 0.1166*age + 3.5609

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 8.5–11°C, and appeared to spread east through the Channel as the surface water temperature ex-ceeded 9°C. Seasonal patterns of occurrence of advanced maturity stages in UK sam-ples 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 south-ern North Sea (information from commercial fisheries and angler reports in Nether-lands 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 31–35 cm, aged 4–7 years, and females at 40–45 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 82%


of females were mature at-age 6 and 7 respectively, with a very small percentage ma-ture 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 col-lected by the UK (Cefas) during 1982–2003 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 sam-pled 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 staging of sea bass by the UK has used the ma-turity 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 ad-vanced maturing fish underrepresented in inshore catches sampled during the period of spawning migrations. An additional spent stage (VIII) has been occasionally rec-orded.

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 ma-jority 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 represent-ing 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 intercept 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 esti-mated length at 50% maturity (L50%).

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 sam-pling which is spread over very many more trips and months than in 2009 and there-fore has better coverage. Maturity ogives were therefore fitted including and excluding 2009 data. The inclusion of 2009 data, which were 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 allow the length classes <35 cm to carry more weight, and the ogive appears to fit the data for 30–40 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 in a recent year to have excessive leverage in the model fit, and the model excluding 2009 was consid-ered 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 Recovering spent

Ovaries one-third length of ventral cavity, opaque, pink with thickened walls and may have atretic eggs

Testis one-third length of ventral cavity, often bloodshot with parts dark grey

III Developing (early)

Ovaries up to one-half length of ventral cavity, orange-red, slight granular appearance, thin, translucent walls

Testes thickness 10–20% of length, dirty white, tinged grey or pink

IV Developing (late)

Ovaries greater than one-third length of ventral cavity, orange-red; eggs clearly visible, but none hyaline

Testes flat-oval in cross section and thickness >20% of length, half to two-thirds of ventral cavity. White colour and milt expressed from vent if pressure applied to abdomen

V Gravid (ripe)

Swollen ovaries two-thirds length of ventral cavity, pale yellow-orange; opaque eggs clearly visible with some hyaline

Testes bright white and more rounded-oval in cross section. Only light pressure required to cause milt to flow from vent

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

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


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: in-cluding 2009 data. Points are proportion mature in the raw data. Dotted line is the number of ob-servations per length class.

The parameters of the model Pmat(L) = 1/(1+e-b(L+c)) are given below:

FEMALES MALES

Intercept (a) -13.556 -16.851

Slope (b) 0.3335 0.4861

c = a/b -40.6488 -34.6652

L25% 37.35 32.41

L50% 40.65 34.67

L75% 43.95 36.93

The logistic model for females and males is:

Pmat(L) = 1/(1+e-0.3335(L-40.6488)) (females) Pmat(L) = 1/(1+e-0.4861(L-34.6652)) (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 1982–2003 UK samples from all areas.

FEMALES MALES

age 2 0.00 0.00

age 3 0.00 0.27

age 4 0.17 0.54

age 5 0.21 0.61

age 6 0.55 0.91

age 7 0.95 0.98

age 8 1.00 1.00

age 9 0.95 0.98

age 10+ 1.00 1.00

Data on sea bass maturity have also been collected in the Netherlands since 2005. Meth-ods and data are described by Quirijns and Bierman (2012). For male fish, too few spec-imens were measured to estimate maturity. Maturity-at-age and length are 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 1982–2003 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 (~4 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 ma-turity-at-length for Netherlands samples is not based on enough measurements to show a reliable maturity ogive.

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 11–15 mm in length (at 30–50 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 overwintering 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 south-ern Ireland.

During 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 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’ dis-tribution 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).

Sea temperature has a strong influence on sea bass dynamics, affecting spatial distri-butions, and also the growth and survival of young bass in nursery areas during the first years of life (Pawson, 1992; WGCSE, 2014).

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 Ber-talanffy growth parameters k and Linf (asymptotic length), length or age at 50% ma-turity 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.


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 liter-ature 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 0.15–0.22 (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, 1996. 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 Di-vision of Marine Fisheries included a total of 40 different experiments by 16 different authors where striped bass hooking mortality was estimated over two or 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 prerecruit surveys

The UK has conducted prerecruit 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, alt-hough in the Thames it is deployed as a twin rig and in the Solent as a single rig.

Age at capture of 1,145 bass by anglers in Ireland

0

2

4

6

8

10

12

14

16

18

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 27

Age, years

Per

cen

tag

e fr

equ

ency


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 first 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 (arithme-tic mean numbers per 10 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.

The spring and autumn Solent survey index series are updated to include the autumn 2013 survey and to amend an error in the autumn survey indices in 2000. The surveys do not show major year-effects, but as noted in previous assessments the autumn (Sep-tember) survey shows a general increase in recruitment during the 1990s up to the mid-2000s, with very low indices for the 2008 onwards year classes, while the spring survey shows poor recruitment from around 2002 onwards. Previous Stock Synthesis runs show that the autumn survey is much better fitted than the spring survey. The spring survey is likely to be more strongly affected by weather and by temperature effects on distribution.

The Thames survey series indicates an increase in recruitment from the mid-1990s to early 2000s followed by some poor year classes, possibly a strong 2007 year class, then weak year classes in 2008 and 2009. A problem with the use of the Thames survey is that it may reflect recruitment from spawning that became established in the southern North Sea as the stock expanded in the 1990s under warmer sea conditions, and may


therefore not reflect recruitment trends that influence the larger stock components in the Channel and Celtic Sea.

A justification of using the Solent survey as an index of recruitment over the full range of the stock was the results of a statistical, UK-only fleet-based separable model devel-oped by Pawson et al. (2007) and updated by ICES WGNEW (Kupschus et al., 2008). The Pawson et al. model was fitted only using UK age compositions for trawls, mid-water trawls, nets and lines, separately for ICES Divisions IVbc, VIId, VIIeh and VIIafg, and was intended mainly to estimate fleet selection patterns. Although it excluded any tuning data, the recruitment-series for each sea area closely resembled the Solent sur-vey indices and to an extent the shorter Thames series, and was able to provide coher-ent selection patterns by fleet.

The full Solent survey series was subject to a change in gear design in 1993. Some com-parative trawling was carried out to develop age-varying calibration factors, but these are poorly documented and the original raw data and calibration results are currently being sought at Cefas. Pending an evaluation of this, the benchmark Stock Synthesis runs included a sensitivity run with the series split into two periods around the gear change. Some additional issues with calibration factors applied to the spring survey were detected during the benchmark, and this is considered later in the sections on model development.

A precision estimate was calculated for the Solent and Thames surveys based on the between-tow variations in catch rate of all the age classes used in the index. For the Solent spring, Solent autumn and Thames surveys, the relative standard errors were 0.42, 0.25 and 0.43 respectively.

IBPBass (ICES, 2014) evaluated these surveys and removed the Solent spring and the Thames survey from the assessment for reasons given in the IBPBass report, leaving the better-performing Solent autumn survey.

B3.2 Other 0-gp & 1-gp surveys

The UK has undertaken a seine net survey in the Tamar Estuary, since 1985. Additional data are available from the Camel estuary and power stations in the Thames and Sev-ern 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. Seine net 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

1972 3

1973 4

1974 1

1975 15 78

1976 127 100

1977 - 6

1978 - 5

1979 - 5

1980 9 37

1981 2 216 21

1982 123 83 56

1983 30 226 83

1984 134 8 62

1985 0.663 0.385 22 11 76

1986 0.005 0.014 1 3 14

1987 0.032 0.062 31 96 116

1988 1.484 1.284 48 98 54

1989 2.348 2.389 112 446 610

1990 1.038 1.516 89 25 433

1991 0.076 0.058 50 300 64

1992 2.216 2.431 25 280 104

1993 1.013 0.913 22 202 131

1994 1.126 0.346 134 - 26

1995 2.356 1.294 - - 27

1996 0.102 0.047 119 242 †

1997 1.119 1.299 102 †

1998 2.082 3.170 264

1999 1.215 0.937 56

2000 0.340 1.185 133

2001 0.351 0.129 †

2002 2.098 3.179

2003 0.965 1.067

2004 1.453 0.261

2005 0.522 0.169

2006 0.186 0.203

2007 0.475 1.308

2008 1.275 1.229

2009 0.460


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 1987. Abundance indices are calculated as stratified means.

Figure B3.2. Station positions for French Evhoe bottom-trawl survey (not used in assessment).

B3.4 Channel Ground Fish Survey (CGFS): France

Raw data on sea bass from the French scientific trawl survey "Channel Ground Fish Survey - CGFS" were not available for the previous benchmark in 2012 (IBPNew, ICES, 2012a). Details of the survey are given in Coppin et al. (2002), which includes a full description of the GOV trawl used in October each year at the 82 stations in ICES Di-vision VIId shown in Figure B3.3. The majority of sea bass are caught in the coastal waters of England and France (Figure B3.3). The abundance indices from all the sta-tions give similar trends as from a subset of stations in the main coastal areas, and trial runs with SS3 gave similar trends. Therefore, for further SS3 development, the indices calculated from all the area are used.

The abundance indices are calculated applying a stratified random swept-area based estimator. Strata correspond to ICES statistical rectangles. Swept-area is calculated us-ing wingspread. As this is a stratified swept-area based indicator, uncertainty is based on between haul variance within a strata and summation of variances across strata. Full methodology is presented in the WD_01, available in Annex 2 of the IBPBass re-port 2014. The trends in both the index and in the proportion of stations with sea bass show some similarities to the trend in total biomass estimates from the ICES WGCSE 2013 update assessment using Stock Synthesis, which lent a priori support to the use of the index in the assessment. The swept-area indices of abundance, the percentage of stations with sea bass, and the variance of the estimates are included in the WGCSE 2014 report and will be updated annually. The length composition of the survey index is calculated and is also input to Stock Synthesis.


The precision of the swept-area indices appears unrealistically high in some years (e.g. 0.025 in 1991), which may indicate that the index trends are driven largely by the inci-dence of positive catches. Modelling of the data using delta lognormal models may provide more realistic precision. During trial Stock Synthesis runs, the use of the CVs resulted in an inability to fit the selection curve for the survey due to individual years being given far too much weight. Relaxing the CVs to 0.30 for all years except the first three years (set to 0.6 given the very low incidence of positive stations) allowed the model to fit the length compositions more closely over the series. The annual indices are therefore input to Stock Synthesis with a CV of 0.30 for all years. The effective sam-ple sizes for the annual survey length composition data are set at the number of stations with sea bass length data. The length compositions for the first three years (1988–1990) are excluded from the assessment due to very small sample sizes although the aggre-gate indices are retained.

Figure B3.3. Left: stations fished during the Channel Groundfish Survey carried out annually by France. Right: distribution of total catches of sea bass over the survey series.

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 log-books provide time-series of varying duration on catch-rates of individual vessels us-ing specific gears. The logbooks with sufficient data cover eight gear types within trawls, nets and lines, covering mainly 10 m and under vessels, excluding recreational vessels. The total numbers of logbooks have declined from 50–60 in earlier years to below 20 in recent years. No logbooks were issued in 2008: Year

Region 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

1 7 9 11 19 9 8 15 16 15 22 16 14 18 16 16

2 0 10 10 15 17 14 13 23 10 25 24 20 24 19 17

3 2 4 6 5 7 7 4 6 7 6 9 3 8 5 3

4 5 5 7 9 7 8 7 11 11 4 6 4 4 4 4

5 7 6 10 13 9 9 10 18 8 10 9 7 11 12 11


Year

Region 2000 2001 2002 2003 2004 2005 2006 2007 2009 2010

1 16 19 14 12 13 8 6 0 3 3

2 15 15 13 14 7 10 5 0 3 2

3 2 5 3 5 5 5 7 0 3 3

4 4 5 6 7 1 3 4 0 3 1

5 9 10 9 4 2 5 6 2 1 1

(Region 1: North Sea IVbc, 2: eastern Channel VIId; 3: western Channel VIIeh; 4: Celtic Sea (VIIfg); Irish Sea (VIIa). The trend in number of records per year shows roughly the same pattern across gears:

An exploratory GAM method was developed (Armstrong and Maxwell, 2012) to ex-tract a common temporal trend in lpue from the individual series for ICES Areas IVbc&VIId, VIIeh and VIIafg (referred in the models as areas 1&2, 3 and 4&5). This is analogous to combining series of tree ring counts from timbers of various ages to give a single series describing climate changes. The general method involves estimating log-book factors and year factors (and interactions) to minimise residual model error. Fol-lowing initial model development and evaluation, a negative binomial error distribution with log link was selected. This can accommodate zero values and allows for the variance to increase with the mean. Working with a log link implies that the estimated trend with year is multiplicative not additive. The R command showing the exact options used for areas 1&2 combined (North Sea and VIId) is:

bass.gam3.12 = gam(lpue ~ factor(BookGear) + s(Year, k=10, bs="ts"), family=neg-bin(c(1,10)), optimizer="perf", data=bass.dat, subset=ARegion=="1and2")

Fitted trends and confidence intervals suggest an increasing lpue trend in regions 1&2 (North Sea & VIId), and 3 (VIIeh) (Figure B4.1). A relatively flat trend and possible recent decline is indicated in regions 4&5 (VIIafg) although the recent trend is highly imprecise. Residual checks indicate the model assumptions are reasonable. Model di-agnostics and sensitivity to smoothing and other parameters are given in Armstrong and Maxwell (2012).


Figure B4.1. Cefas bass logbook lpue: Selected model for combined regions, plots showing year effects from a fitted model with separate mean value for each book number-gear combination and negative binomial error distribution, dashed lines are a 95% confidence interval.

B4.2 UK fleet lpue based on official catch dataseries

Armstrong and Maxwell (2012) review trends in UK commercial fishery lpue for sea bass in the North Sea (IV), eastern Channel (VIId), western Channel (VIIe) and Irish/Celtic Seas (VIIafg) from 1985–2011, and evaluate the possibility of using the time-series as relative abundance estimates for tuning stock assessment models.

Gears which catch bass are targeted at a variety of species, and the fishing effort is distributed across many areas where sea bass have zero or very low probability of cap-ture. A number of approaches are possible to subset fishing trips to include only those that have a probability of catching the species for which lpue is to be estimated. One approach (Stephens and MacCall, 2004) is to cluster fishing trips according to species that occur in association, and use only the cluster with the species on interest for esti-mating lpue. This method has not yet been applied to UK data. An alternative method to subset trips was applied. This involved (a) selecting gear types that account for ~95% of the total bass landings in each area since 2005; (b) for the selected gears and areas, identify ICES rectangles accounting for ~95% of the total bass landings since 1985. An-nual lpue was then estimated for each area and gear, separately for vessels of 10 m (LOA) and under and >10 m vessels. The LOA split is important because reporting of landings and effort of 10 m and under vessels has been very uncertain historically, particularly prior to the introduction of Buyers and Sellers regulations in 2005. Lpue of 10 m and under vessels may be very inaccurate prior to 1995.

It was not possible to evaluate the effect of any increase in targeting of bass by individ-ual vessels using the selected gear types in the selected rectangles, or effects of technol-ogy creep. Increased targeting is likely to have happened for vessels with increasingly limited quotas for other species such as cod and which have switched to non-TAC spe-cies such as sea bass. For some gears, such as beam trawls, sea bass are not targeted and are purely a bycatch.

1985 1990 1995 2000 2005 2010-0.

50.0

0.51.0

Regions 1 and 2

Year

s(Yea

r,6.58

)

1985 1990 1995 2000 2005 2010

-0.6

-0.2

0.20.6

Region 3

Year

s(Yea

r,0.83

)

1985 1990 1995 2000 2005 2010

-1.0

-0.5

0.00.5

Regions 4 and 5

Year

s(Yea

r,7.77

)


Too many lpue series have been examined to reproduce in the stock annex, but can be viewed in Armstrong and Maxwell, 2012.

B4.2 French lpue sets

Lpue of French trawlers in IVb,c, VIId and VIIeh is available from 2000 when Ifremer has estimated landings by ICES Divisions. A recent study has developed indices as kg/per day based on data from auction’s sales. This study was carried out on French bottom trawlers (less than 18 m), having a fishing strategy with the least distant ran-dom sampling; this fleet usually doesn’t target sea bass. Large bias can be caused where: 1. an auction sale corresponds to several days of fishing, 2. technological ad-vances are not taken into account, and 3. changes in fishermen’s strategies are not taken into account. Never the less, for information, those from the Channel and North Sea have been compared to the UK Otter trawls lpue, and similarities shown on Figure B4.2 are observed.

Figure B4.2. UK fleet lpue based on official catch dataseries, compared to the French lpue sets based on auction hall sales.

A French study has started from 2014 on the subject of exploring the possibility of get-ting an abundance indices signal from the commercial fisheries database, and will im-prove the 2013 study. Conclusion were not available for WGCSE 2014 (the end of the study is planned for the end of the year 2014).

B.5. Other relevant data

None.

C. Assessment: data and method

C.1. Model configuration

Model used: Stock Synthesis 3 (SS3) (Methot, 2010)

Software used: Stock synthesis v3.23b (Methot, 2011)

WGCSE 2013 conducted an update assessment using Stock Synthesis 3 (SS3) (Methot, 2010). The software used was Stock synthesis v3.23b (Methot, 2011), according to the Stock Annex developed by ICES IBPNEW 2012 with inclusion of fishery data for 2012. The assessment requires a modelling framework capable of handling a mixture of age


and length data for fisheries and surveys (fleet-based landings; landings age or length compositions, age-based survey indices for young bass) and biological information on growth rates and maturity. Landings-at-age were available for four UK fleets from 1985 onwards, whereas French fleets had length composition data that were available only since the 2000s. The Stock Synthesis assessment model was chosen, primarily for its highly flexible statistical model framework allowing the building of simple to complex models using a mix of data compositions available. The model is written in ADMB (www.admb-project.org), is forward simulating and available at the NOAA toolbox: http://nft.nefsc.noaa.gov/SS3.html.

A mixed age–length model was fitted by WGCSE 2013 as the base case, with a length-only model for comparison. Some adjustments were made by WGCSE 2013 to the model: i) UK fishery compositions for 2012 were input to the age–length model as length compositions because age compositions were not available; ii) the UK midwater trawl series was reduced to 1996 onwards and was input as length compositions be-cause unusual length-based selection curve parameters were obtained when inputting the data as age compositions; iii) recruit deviations were estimated back to 1965.

IBPBass (ICES 2014) addressed the following recommendations of WGCSE 2013 for developing the assessment during the inter-benchmark meeting. Work completed is indicated in parentheses:

• Source and review information on historical catches and develop plausible scenarios including over the 20+ year burn-in period for the assessment [some investigations were pursued in France but yielded no clear information on pre-1985 landings];

• Review the derivation and quality of historical fishery length/age composi-tion data [not done beyond the information on sampling intensity and coverage already available];

• Expand UK fishery age compositions to all true ages [done, see below]; • Rationalise the fleet definitions, and reduce to the minimum sufficient to

provide robust SS3 stock trends [done, see below]; • Source and evaluate candidate lpue or effort series for tuning abundance or

fishing mortality on older ages [fishery-dependent abundance indices were not considered other than some information presented in Section 2 on lpue of fleets in the Netherlands];

• Collate and evaluate other survey data on bass abundance that could be in-corporated in the model [French Channel Groundfish Survey was evaluated and incorporated in the assessment];

• Determine the most robust approach to incorporating mean length-at-age and length-at-age distributions in SS3 [Not done];

• Investigate potential biases in using combined-sex growth parameters [Not done];

• Further explore the sensitivity of the assessment to decisions on model structure and inputs [See model development and sensitivity analyses carried out below];

• Consider if simpler assessment approaches are warranted [IBPBass focused exclusively on Stock Synthesis to try to make best use of all available data].

http://nft.nefsc.noaa.gov/SS3.html


The structure and input data/ parameters of the IBPBass (2014) model, used for the update assessment by WGCSE 2014, are summarized below:

The structure and input data/ parameters of the IBPBass revised SS3 model are sum-marized below:

Model structure

• Temporal unit: annual based data (landings, survey indices, age frequency and length frequency);


Fleet definition

Four fleets defined: 1. UK bottom trawls, nets and lines; 2. UK midwater pair trawls; 3. French fleets (combined); 6. Other (other countries and other UK fleets combined). [WGCSE 2013 assessment modelled selectivity separately for UK trawls, midwater trawl, nets and lines].

Landed catches

Annual landings in tonnes from 1985 to final year for the four fleets from ICES Subdi-visions IVb and c, VIIa, d–h. French data were as provided by Ifremer.

Abundance indices

Channel Groundfish Survey in VIId in autumn (France), 1988 to present: total swept-area abundance index and associated length composition data. Number of stations with sea bass is used as input effective sample size. Input CV for survey = 0.30 all years. First three years of composition data are excluded. [Survey not included in WGCSE 2013 assessment].

Cefas Solent survey in autumn (VIId). Years 1986 to 2011; 2013. Three independent abundance index series were defined, each being a single age group (2, 3, 4 years old). They are treated as three independent surveys (following a recommendation from R. Methot) to circumvent difficulties in estimating selectivity parameters for a survey se-ries comprising only three young age groups, although this approach loses covariance information due to year-effects in the survey. [Solent spring and Thames survey in-cluded in WGCSE 2013 assessment].

Fishery landings age composition data: UK fleets

Age bins: 0 to 15 with a plus group for ages 16 and over. Age compositions for UK fleets are expressed as fleet-raised numbers-at-age, although they are treated as relative compositions in SS3. Year range for UK trawls/nets/lines: 1985 to present; UK midwater pair trawl: 1996 to present.

Length composition data: French fleets

The length bin was set from 4 to 100 cm by 2 cm intervals. Length compositions for the following fishing fleets were used: French all fleets combined: 2000 to present.



The following table summarises key model assumptions and parameters. Other pa-rameter values and input data characteristics are defined in the SS3 control file BassIV-VII.ctl, the forecast file Forecast.SS and the data file BassIVVII.dat as used by IBPBass 2014 and by WGCSE 2013. Model inputs and year ranges (for WGCSE 2014) are in Fig-ure C1.

Key model assumptions and parameters from the WGCSE 2014 update assessment.


Starting year 1985

Ending year 2013

Equilibrium catch for starting year 0.82* landings in 1985 by fleet.

Number of areas 1

Number of seasons 1


Number of surveys two surveys: CGFS; Solent autumn survey (Solent spring and Thames survey removed).


Number of active parameters 68


Maximum age 30

Genders 1


Ages for summary total biomass 0–30




Minimum age for growth model 2


Maturity Logistic 2-parameter – females; L50 = 40.65cm




Maximum F 2.9

Fleet 1: UK Trawl/nets/lines selectivity Double normal, age-based

Fleet 2: UK Midwater trawl selectivity Asymptotic, age-based

Fleet 3: Combined French fleet selectivity Asymptotic, length-based

Fleet 4: Other fleets/gears selectivity Asymptotic: mirrors French fleet

Year-invariant recreational fishing mortality vector (F(5–11) = 0.09)

Asymptotic, age-based (fixed, not estimated). Added to M vector



CGFS survey timing (yr) 0.75


Survey selectivities: Solent autumn: [all survey data entered as single ages; sel = 1]



Survey selectivities: CGFS Double normal, length based















Age error matrix CV 12% at-age



1965

Last year for recruit deviations 2011 (last year class with survey indices)

1 as recommended by R. Methot after scrutinizing earlier SS3 runs during IBPNEW 2012, and used by IBPNEW and WGCSE. The WGCSE 2013 tabulated the original value of 5.78 cm at-age 0 in error.

Figure C1. Summary of inputs and year ranges for Stock Synthesis assessment (as at 2014).


C.2 Assessment procedure

The model is run with the executable file SS3.exe in the same folder as the following files:

BassIVVII.ctl SS3 configuration file

BassIVVII.dat SS3 data inputs

Starter.SS SS3 startup file

Forecast.SS SS3 forecast file

Results are output in the same folder (key results file is “results.sso”). Plots can be gen-erated using r4ss after calling library(r4ss), using the following code (adjusted with correct path name):

age <- SS_output(dir= 'C:/Users/ma02/Documents/ICES/WGCSE2014/Bass47/SS3 update assessment') SS_plots(replist=age,pdf=F,png=T,dir='C:/Users/ma02/Docu-ments/ICES/WGCSE2014/Bass47/SS3update assessment')#,uncertainty=F )

Retrospective analysis is done with the output files from the base run in the same folder as the file retro.bat. For five retrospectives, six Starter files are included. The base file Starter.SS includes the following code nine lines from the bottom:

-5 # retrospective year relative to end year (e.g. -4)

The five retrospective Starter files use the name convention Starter-5; Starter-4; Starter-3; Starter-2; Starter-1, amending the command -5 # retrospective year relative to end year (e.g. -4) to reflect the year peel stated in the file name. A piece of code “Retro-Plots_R4SS” is available to plot the retrospectives although an Excel file is currently used to read the results from each of the Report.sso files imported into worksheets.

For the WGCSE 2014 assessment, a year-invariant recreational fishing mortality vector (F(5–11) = 0.09) was added to the annual M of 0.15, specified as age-specific values in the .ctl file in the line:

3 #_natM_type:_0=1Parm; 1=N_breakpoints;_2=Lorenzen;_3=agespecific;_4=ag-espec_withseasinterpolate

0.150 0.150 0.150 0.152 0.163 0.204 0.241 0.249 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250

The recreational component was arrived at iteratively to generate a total recreational fishery landing close to 1500 t for 2012.

Future runs may need to revise this to generate the same recreational catch.

When the end year for the Stock Synthesis run is specified as the last year with fishery data, the Report.sso file contains estimates of biomass and numbers only to the start of the final year with data, and Zs only to the year before the final one. A work-around to get biomass and numbers for survivors at the end of the last year with data, and Zs for the final year with data, the end year can be specified as the year after the last with


data. F values, as used by ICES, are not generated automatically by Stock Synthesis but can be computed from the Zs after subtracting M.

D. Forecast

Due to the additional complexity of adding a fixed recreational fishing mortality vector for removals (harvest), and the time required to configure Stock Synthesis to mirror the ICES procedures for short-term forecasts, IBPBass decided not to try to develop a fore-cast procedure within Stock Synthesis for use by WGCSE. This unfortunately loses the ability to provide MCMC confidence intervals around the assessment and forecasted variables, and the forecasts are entirely deterministic. Management options involving biological reference points (BRPs) adopt BRPs conditional on the assumptions in the assessment regarding M, selectivity, maturity, weights-at-age, etc. The procedures for deriving inputs for the short-term forecast are described below.

D.1 Estimating year-class abundance

Stock Synthesis should be set up to estimate recruit deviations only until the last year that has survey data to make the estimates. E.g. including the Solent survey indices for ages 2–4 in 2013 means that the last year class tuned by a survey index is 2011 (two year-olds in 2013). SS3 will put a value from the fitted stock–recruit curve for later year classes. WGCSE overwrites these later year classes using the long-term (1985 onwards) geometric mean, or a short-term GM if there is a persistent reduction in recent recruit-ment. The numbers-at-age for the starting year of the forecast are also over-written for these year classes by reducing the GM recruitment by the appropriate number of years of M (as there is no catch for the first few years of age).

WGCSE (2013) reviewed some information on environmental influences on sea bass recruitment which supported a recent reduction in recruitment. Survival of 0-gp and 1-gp sea bass in nursery areas in estuaries and saltmarshes is thought to be enhanced by warmer conditions promoting survival through the first two winters, and increasing the growth rates (Pawson, 1992). WGCSE 2014 presented an argument for choosing a particular recruitment value for the 2012 year class for inclusion in forecasts, based on a consideration of past recruitment in relation to temperature. The data and arguments in WGCSE 2014 should be consulted for an explanation of the logic used.

The format for reporting the recruitment values for the short-term forecast are summa-rised using an example in Table D1.1., and an example of a short-term forecast input file is given in Table D1.2.

Table D.1.1. Example of recruitment estimates included in a short-term forecast for sea bass, from IBPBass 2013.

YEAR CLASS SS3 (AGE 0) GM 2008–2011 GM 2008–2011

2011 2648 thousand

2012 1815 thousand

2013 6057 thousand

2014 6057 thousand


Example input for the short-term catch predictions is in Table D1.2. The derivation of the inputs is described in Table D1.3.

Table D1.2. Example inputs for sea bass short-term forecast, from WGCSE 2014. Inputs for short-term forecast. F(5–11) is mean for years 2011–2013. Numbers-at-ages 0–2 in 2014 are adjusted by replacing Stock Synthesis average values in 2012–2014 (years with no recruit deviations estimated) with the short-term (2008–2011) GM in 2012 and the long-term GM in 2013 and 2014, with adjust-ments for natural mortality. Rules are below table.

ageNo. at age in

2014weight in

stock

Proportion mature

(female)H.Cons mean F

(2011-2013)H.Cons mean

weights Recreational F

Recreational removals mean

weight M

0 6057 0.002 0.000 0.000 0.007 0.000 0.007 0.151 5213 0.022 0.000 0.000 0.076 0.000 0.052 0.152 1345 0.101 0.000 0.000 0.388 0.000 0.154 0.153 1687 0.218 0.000 0.014 0.584 0.002 0.295 0.154 225 0.382 0.186 0.070 0.715 0.013 0.480 0.155 1433 0.587 0.419 0.161 0.877 0.054 0.702 0.156 830 0.825 0.638 0.220 1.093 0.091 0.954 0.157 1749 1.088 0.792 0.247 1.331 0.099 1.228 0.158 1491 1.370 0.885 0.256 1.588 0.100 1.517 0.159 808 1.664 0.937 0.266 1.860 0.100 1.815 0.15

10 625 1.964 0.965 0.271 2.143 0.100 2.116 0.1511 519 2.264 0.980 0.274 2.431 0.100 2.416 0.1512 312 2.562 0.989 0.275 2.719 0.100 2.711 0.1513 140 2.853 0.993 0.276 3.002 0.100 2.998 0.1514 60 3.135 0.996 0.276 3.276 0.100 3.275 0.1515 77 3.406 0.998 0.276 3.540 0.100 3.540 0.15

16+ 76 4.017 0.998 0.276 4.069 0.100 4.134 0.15

Age 0,1,2 over-written as follows

2014 yc 2014 age 0 replaced by 1985-2011 LTGM (6057); 2013 yc 2014 age 1 replaced by SS3 survivor estimate at age 1, 2014 * LTGM / SS3 estimate of age 0 in 20132012 yc 2014 age 2 replaced by SS3 survivor estimate at age 2, 2014 * STGM (1815) / SS3 estimate of age 0 in 2012 (weak y 2011 yc estimated from Solent survey in 2013, age 2


Table D1.3. Derivation of short-term forecast inputs (based on example from IBPBass 2014).

INPUT DATA DERIVATION

Starting numbers-at-age 0–16+ in first year (intermediate year)

SS3 output. (N age zero overwritten where necessary by long-term GM, short-term GM or other predicition, reduced by M=0.15 the the require number of years (if no commercial catches) or multiply N at-age in starting year from the assessment by ratio of the replaced recruit value with the SS3 estimate.

Recruitment 2014 onwards Long-term GM, short-termGM or other predictor.

Mean wt-at-age in stock SS3 output

Proportion mature (female) SS3 output

Commercial fishery (H-cons) mean F at-age Average last three years: SS3 output Zs minus M=0.15 and recreational F at-age

Commercial fishery (H-cons) mean weight-at-age

SS3 output figures on mean weight in UK, French and other fleets, weighted by SS3 model estimates of landings numbers-at-age for the fleets

Recreational removals F at-age Input values to SS3 (year-invariant), based on commercial lines selectivity)

Recreational removals weights-at-age Output values for UK commercial fleets from final SS3 Run

M 0.15 at all ages

An example detailed forecast is given in Table D1.4 for the status quo F option, which is the most likely forecast given the absence of any restrictive management controls on effort or landings of sea bass. See WGCSE 2014 for examples of management options tables.

Future forecast routines may be configured in Stock Syntheses, allowing MCMC esti-mation of confidence limits.


Table D1.4. Example of detailed short-term forecast (WGCSE 2014).

E. Biological reference points

E.1 Background

The Stock synthesis model currently used fixes stock–recruit steepness at 0.999. There are insufficient observations at low SSB to suggest the possible steepness of the rela-tionship. This means that MSY reference points cannot be obtained from a plausible


AgeF(5-11):

CommercialF(5-11):

RecreationalCatch Nos:

CommercialYield:

Commercial Catch Nos:

Recreational Yield:

Recreational Stock Nos BiomassSSB nos.

Jan 1SSB tonnes

Jan 10 0.000 0.000 0.0 0.0 0.0 0.0 6057 9 0 01 0.000 0.000 0.0 0.0 0.1 0.0 5213 117 0 02 0.000 0.000 0.0 0.0 0.3 0.0 1345 135 0 03 0.014 0.002 21.1 12.3 2.8 0.8 1687 369 0 04 0.070 0.013 14.1 10.1 2.6 1.2 225 86 42 165 0.161 0.054 193.1 169.4 65.3 45.9 1433 841 601 3526 0.220 0.091 146.3 159.9 60.3 57.5 830 684 530 4377 0.247 0.099 340.7 453.4 136.2 167.2 1749 1903 1385 15078 0.256 0.100 300.2 476.7 116.8 177.3 1491 2044 1319 18089 0.266 0.100 168.0 312.5 63.1 114.5 808 1345 757 1260

10 0.271 0.100 132.1 283.1 48.7 103.0 625 1227 603 118411 0.274 0.100 110.7 269.1 40.4 97.6 519 1176 509 115312 0.275 0.100 66.7 181.3 24.2 65.7 312 798 308 78913 0.276 0.100 30.1 90.2 10.9 32.7 140 400 139 39714 0.276 0.100 12.9 42.3 4.7 15.3 60 188 60 18815 0.276 0.100 16.5 58.4 6.0 21.1 77 261 77 261

16+ 0.276 0.100 16.3 66.2 5.9 24.3 76 304 76 303Total 1569 2585 588 924 22647 11887 6404 9654


AgeF(5-11):

CommercialF(5-11):


CommercialYield:


Recreational Yield:


Jan 1SSB tonnes

Jan 10 0.000 0.000 0.0 0.0 0.0 0.0 6057 9 0 01 0.000 0.000 0.0 0.0 0.1 0.0 5213 117 0 02 0.000 0.000 0.0 0.0 0.9 0.1 4487 452 0 03 0.014 0.002 14.5 8.4 1.9 0.6 1157 253 0 04 0.070 0.013 89.6 64.1 16.4 7.9 1430 547 266 1025 0.161 0.054 24.0 21.1 8.1 5.7 178 105 75 446 0.220 0.091 175.2 191.4 72.2 68.9 994 820 634 5237 0.247 0.099 102.0 135.8 40.8 50.1 524 570 415 4518 0.256 0.100 214.4 340.5 83.5 126.6 1065 1460 942 12919 0.266 0.100 186.9 347.6 70.2 127.4 899 1496 842 1401

10 0.271 0.100 102.0 218.5 37.6 79.5 482 947 465 91411 0.274 0.100 79.1 192.3 28.9 69.8 371 840 364 82412 0.275 0.100 65.8 178.9 23.9 64.8 307 788 304 77913 0.276 0.100 39.5 118.7 14.3 43.0 184 526 183 52214 0.276 0.100 17.8 58.3 6.4 21.1 83 260 82 25915 0.276 0.100 7.6 27.0 2.8 9.8 36 121 35 121

16+ 0.276 0.100 19.4 78.7 7.0 28.9 90 362 90 361Total 1138 1981 415 704 23559 9670 4698 7591


AgeF(5-11):

CommercialF(5-11):


CommercialYield:


Recreational Yield:


Jan 1SSB tonnes

Jan 10 0.000 0.000 0.0 0.0 0.0 0.0 6057 9 0 01 0.000 0.000 0.0 0.0 0.1 0.0 5213 117 0 02 0.000 0.000 0.0 0.0 0.9 0.1 4487 452 0 03 0.014 0.002 48.3 28.2 6.4 1.9 3861 844 0 04 0.070 0.013 61.5 44.0 11.3 5.4 981 375 183 705 0.161 0.054 152.7 133.9 51.7 36.3 1133 665 475 2796 0.220 0.091 21.8 23.8 9.0 8.6 124 102 79 657 0.247 0.099 122.2 162.6 48.8 60.0 627 683 497 5408 0.256 0.100 64.2 102.0 25.0 37.9 319 437 282 3879 0.266 0.100 133.5 248.3 50.1 91.0 642 1068 601 1001

10 0.271 0.100 113.4 243.0 41.8 88.4 536 1053 518 101711 0.274 0.100 61.0 148.4 22.3 53.8 286 648 281 63612 0.275 0.100 47.0 127.9 17.1 46.3 220 563 217 55713 0.276 0.100 39.0 117.1 14.1 42.4 182 519 181 51514 0.276 0.100 23.4 76.6 8.5 27.7 109 341 108 34015 0.276 0.100 10.5 37.2 3.8 13.5 49 167 49 166

16+ 0.276 0.100 15.9 64.9 5.8 23.8 74 298 74 297Total 914 1558 317 537 24901 8341 3544 5869


stock–recruit relationship and have to be derived from yield-per-recruit. WGCSE 2013 and 2014, and IBPBass 2014 computed yield-per-recruit based biological reference points F0.1 and F35%spr based on the inputs and outputs of the stock synthesis update assessment. In 2013, with an input M of 0.20, the F(5–11) value for F35%spr was 0.17 and the F0.1 was 0.18. In 2014, WGCSE reduced the M to 0.15, and this caused a reduction in the F35%spr to 0.13 for combined commercial and recreational landings, and the F0.1 was 0.12. The F reference points therefore scale directly with M, as expected, with some effect of changes in estimated fishery selection patterns and weights-at-age.

The revised SS3 model proposed by IBPBass and applied at WGCSE 2014 now includes an explicit recognition of the possible recreational fishing mortality, albeit as a fixed vector of F at-age following the same selectivity as the commercial line fishery. This leads to a more complex problem in defining BRPs based on yield-per-recruit because the fishing mortality now has separate commercial and recreational F components which can be manipulated separately or simultaneously in a yield-per-recruit analysis. The recreational F vector is indicative only, in the sense that it is conditioned on a total annual recreational removals estimate of ~1500 t for recent years, which can be consid-ered as a “plausible scenario” rather than an explicit estimate of recreational F with recreational fishery survey data and their precision included in the model fitting. Dif-ferent scenarios for total recreational removals affect how total fishing mortality is split between commercial and recreational F, but the combined F estimates are minimally affected as they are driven by the fishery composition data.

E.2 MSY BRPs or proxies for sea bass

BRPs for this assessment based on advice from ICES WKMSYREF2 (ICES, 2014), and recognising that FMSY cannot be obtained from a fitted stock–recruit curve include:

1 ) Setting an FMSY proxy as F0.1 or Fxx%spr based either on the commercial fishery only, or the combined commercial and recreational F.

2 ) Setting a BMSY trigger around a low percentile of the expected range of SSB when fishing at FMSY.

Unfortunately it has not been possible yet to carry out a full MCMC bootstrap of the sea bass assessment and to propagate this into a forecast period to evaluate the percen-tiles of expected SSB while fishing at FMSY. A concern with sea bass is that recruitment has shown longer term changes in mean recruitment (and hence stock productivity) that appear related to changes in sea temperature at decadal scales. For a biomass ref-erence point, WGCSE 2014 selected the lowest observed SSB (Bloss) of 5250 t as a value for Blim, the limit reference point for SSB. There is therefore no MSY or precautionary reference point for biomass (Table E2.1).

WGCSE chose F35%spr as a proxy for FMSY (value: F(5–11) = 0.13 for combined commercial and recreational fishery). As for many stocks, this value is close to F0.1 (F=0.12). (Table E2.1).

The yield-per-recruit curve is flat-topped and FMSY in sea bass is poorly defined (Figure E2.1). Fishing at FMAX implies a yield-per-recruit only marginally higher than at F35%SPR but requires much higher F, with much larger fishing costs where these are propor-tional to F.

E.3 Inclusion of BRP in short-term forecast

For the estimation of Yield-per-recruit and Spawner biomass per recruit, WGCSE 2014 varied the multipliers on both fisheries by the same amount. The resulting yield-per-


recruit curve therefore reflects total F and total yield, but does not assume a particular allocation between any fisheries.

For short-term forecasts, WGCSE agreed that the multiplier on the recreational F vector should be maintained at 1.0 for all management options except zero F, and only the commercial fishery vector is altered in management options. The relative contribution of commercial and recreational F to the total F will not be the same as the contribution used in calculating the FMSY based on the mean F for 2011 to 2013. However, the selec-tivity and weights-at-age for the two fisheries are sufficiently close that the agreed FMSY

can be applied for different combinations of commercial and recreational F with rela-tively minor differences. This allows managers to select an F relative to FMSY for a fore-cast year but decide how the resultant catch forecast can be allocated between the different fishery sectors. This is no different from management of different commercial fishery fleets.

Table E2.1. BRPs proposed by WGCSE 2014 for sea bass in Areas IV and VII.

TYPE VALUE TECHNICAL BASIS

Precautionary approach Blim 5250 t Lowest observed SSB

Bpa Undefined

Flim Undefined

Fpa Undefined

MSY approach FMSY 0.13 Based on F giving SSB per recruit 35% of value at zero F.

MSYbtrigger Undefined

FMAX is not definable.

Figure E2.1 Yield and biomass per recruit analysis from WGCSE 2014: conditional on mean pattern of F-at-age for 2011–2013 for the commercial fishery and the vector of F-at-age for recreational fish-ing, with the same F-multiplier applied to both. The partial YPR by fleet is shown conditional on the relative contribution of the F-at-age in the two fisheries to the combined F being the same as estimated for recent years in the assessment. The 35%spr is indicated (red line) to show where this occurs on the SPR curve.


E.4 Yield-per-recruit reference point calculations

Example inputs for a yield-per-recruit analysis are given in Table E4.1.1. They are iden-tical with the short-term forecast inputs except that the values are extended to 30 years of age. SS3 provides population estimates out to this age. It is assumed that no fish survive after the 30th year, as no fish older than 28 years have been observed histori-cally. The input data are entered into the ICES Standard Plots system on SharePoint, which produces the YPR and SPR plots together with estimates of FMAX, F0.1 and FMED. F35%SPR is not output but can easily be computed using a spreadsheet, checking that the F0.1 and SPR values are the same as the ones generated by the ICES Standard Plots soft-ware. (This software cannot at present display the partial YPR values for individual fleets).

Table E4.1.1. Example inputs to yield-per-recruit analysis. F(5–11) and weights-at-age are the recent three year average. The values for a 16+ gp are indicated although using this prevents the mean stock weight to increase in the plus group as F is reduced, leading to a small bias in SSB per recruit at low or zero F. Bass in the wild have been recorded up to 28 years old.

There is currently no TAC for sea bass, and control of fishing mortality would have to be through other approaches to managing effort on sea bass, and including technical measures to alter selectivity and/or restrict fishing seasonally or spatially. Note that the inclusion of discards in the assessment would alter the reference points and historical

Age M Pmatstock

wt (kg)

F(5-11): commercia l

fleets

F(5-11): recreational

fi shery

Catch wt (kg): commercia l

fleets

Catch wt (kg): recreational

fleets

0 0.15 0.000 0.002 0.000 0.000 0.007 0.0071 0.15 0.000 0.022 0.000 0.000 0.076 0.0522 0.15 0.000 0.101 0.000 0.000 0.388 0.1543 0.15 0.000 0.218 0.014 0.002 0.584 0.2954 0.15 0.186 0.382 0.070 0.013 0.715 0.4805 0.15 0.419 0.587 0.161 0.054 0.877 0.7026 0.15 0.638 0.825 0.220 0.091 1.093 0.9547 0.15 0.792 1.088 0.247 0.099 1.331 1.2288 0.15 0.885 1.370 0.256 0.100 1.588 1.5179 0.15 0.937 1.664 0.266 0.100 1.860 1.815

10 0.15 0.965 1.964 0.271 0.100 2.143 2.11611 0.15 0.980 2.264 0.274 0.100 2.431 2.41612 0.15 0.989 2.562 0.275 0.100 2.719 2.71113 0.15 0.993 2.853 0.276 0.100 3.002 2.99814 0.15 0.996 3.135 0.276 0.100 3.276 3.27515 0.15 0.998 3.406 0.276 0.100 3.540 3.54016 0.15 0.998 3.665 0.276 0.100 3.791 3.79317 0.15 0.999 3.911 0.277 0.100 4.030 4.03218 0.15 0.999 4.143 0.277 0.100 4.255 4.25819 0.15 1.000 4.362 0.277 0.100 4.466 4.46920 0.15 1.000 4.566 0.277 0.100 4.664 4.66721 0.15 1.000 4.757 0.277 0.100 4.848 4.85222 0.15 1.000 4.935 0.277 0.100 5.019 5.02323 0.15 1.000 5.099 0.277 0.100 5.177 5.18124 0.15 1.000 5.252 0.277 0.100 5.324 5.32825 0.15 1.000 5.393 0.277 0.100 5.459 5.46426 0.15 1.000 5.523 0.277 0.100 5.584 5.58827 0.15 1.000 5.642 0.277 0.100 5.698 5.70328 0.15 1.000 5.752 0.277 0.100 5.803 5.80829 0.15 1.000 5.853 0.277 0.100 5.900 5.90530 0.15 1.000 6.064 0.277 0.100 6.102 6.107

16+ 0.15 1.000 4.017 0.277 0.100 4.069 4.134


series. In the absence of discards, it is difficult to infer benefits to YPR or SSB/R in im-proving the selectivity patterns of the fleets. There is currently no time-series of recre-ational fishing catches to monitor the impacts of any management measures, and the frequency and extent of future surveys remains uncertain.

F. Other Issues

F.1. Historical overview of previous assessment methods

Previous assessments of sea bass in Area IV and VII are summarised below.

2007: Pawson et al. 2007. ADMB separable model on UK data; updated 2008 at WGNEW (Kup-schus et al. 2008).

2012: IBPNew (ICES, 2012). Development of age and length-based Stock Synthesis assessment.

2013: WGCSE. Update assessment using IBPNew SS model. Recommended Inter-benchmark to improve model.

2014: IBPBass. Added new CGFS surveys series; removed poorly performing surveys; improved fleet structure and selectivity model; incorporated recreational fishery information; devel-oped forecast and BRPs.

2014: WGCSE. Update assessment using IBPBass model.

E. References Armstrong, M.J. 2012. Life-history estimates of natural mortality of sea bass around the UK.

Working Document: ICES IBP-NEW 2012; October 2012. 3pp.

Armstrong, M.J. and Walmsley, S. 2012a. An evaluation of the bass fleet census and logbook system for estimating annual landings by gear for fishing vessels in England and Wales. Working Document: ICES IBP-NEW 2012; October 2012. 11pp.

Armstrong and Walmsley. 2012b. Age and growth of sea bass sampled around the UK. Working Document: ICES IBP-NEW 2012; October 2012. 15pp.

Armstrong and Walmsley. 2012c. Maturity of sea bass sampled around the UK. Working Docu-ment: ICES IBP-NEW 2012; October 2012. 14pp.

Armstrong, M.J. and Maxwell, D. 2012. Commercial fleet lpue trends for sea bass around the UK. Working Document: ICES IBP-NEW 2012; October 2012. 29 pp.

Armstrong, M. and Readdy, L. 2014. Effect on sea bass Stock Synthesis model of expanding the UK fishery age compositions to a larger plus group. Working Document, ICES IBPBASS, 2014.

Coppin, F., Le Roy, D., Schlaich, Y. 2002. Manuel des protocoles de campagne halieutique: Cam-pagnes CGFS, Système d'information halieutiques - Campagnes à la mer. Ifremer, 09/2001 – DRV/RH/DT/AN-NUMERO, 29pp. (in French).

Diodati, P. and R.A. Richards. 1996. Mortality of striped bass hooked and released in salt water. Transactions of the American Fisheries Society. 125: 300–307.

Drogou M et al. 2011. Synthèse des informations disponibles sur le Bar : flottilles, captures, mar-ché. Reflexions autour de mesures de gestion.

Dunn, M.R. and Potten, S. 1994. National Survey of Bass Angling: Report to the Ministry of Ag-riculture, Fisheries and Food. University of Portsmouth, Centre for the Economics and Man-agement of Aquatic Resources. 45pp + appendices.

Dunn, M., Potten, S., Radford, A. and Whitmarsh, D. 1989. An Economic Appraisal of the Fishery for Bass in England and Wales. Report to the Ministry of Agriculture, Fisheries and Food. University of Portsmouth. 217 pp.


Herfaut J., Levrel H., Drogou M. et Véron G. 2010. Monitoring of recreational fishing of sea bass (Dicentrarchus labrax) in France: output from a dual methodology (telephone survey and di-ary) ICES CM 2010/R: 05.

ICES. 2001. Report on the ICES Study Group on Bass. CM 2001/ACFM:25, 18 pp.

ICES. 2002. Report on the ICES Study Group on Bass. CM 2002/ACFM:11 ref.G, 59 pp.

ICES. 2004a. Report of the Study Group on Bass, Lowestoft, England, August 2003. ICES Docu-ment, CM 2004/ACFM: 04. 73 pp.

ICES. 2004b. Report of the Study Group on Bass, By Correspondence. ICES Document, CM 2004/ACFM: 31 Ref G. 56pp.

ICES. 2008. Report of the Working Group on the Assessment of New MoU Species (WGNEW). By Correspondence, ICES CM 2008/ACOM:25. 77 pp.

ICES. 2009. Report of the ICES Workshop on Sampling Methods for Recreational Fisheries (WKSMRF). ICES CM 2009/ACOM: 41.

ICES. 2010 Report of the Planning Group on Recreational Fisheries (PGRFS), 7–11 June 2010, Bergen, Norway. ICES CM 2010/ACOM:34. 168 pp.

ICES. 2011. Report of the Planning Group on Recreational Fisheries (PGRFS). ICES CM 2011/ACOM:23.

ICES. 2012a. Report of the Working Group on Assessment of New MoU Species (WGNEW), 5–9 March 2012, ICES CM 2012/ACOM:20. 258 pp.

ICES. 2012b. Report of the Working Group on Recreational Fisheries Surveys (WGRFS). ICES CM 2012/ACOM:23. 55 pp.

ICES. 2013a. Report of the Working Group for Celtic Seas Ecoregion (WGCSE), 8–17May 2013, Copenhagen, Denmark. ICES CM 2013/ACOM:12. 1986 pp.

ICES. 2013. Report of the Working Group on Recreational Fisheries. ICES CM 2013/ACOM:23.

Jennings, S., and Pawson, M. G. 1992. The origin and recruitment of bass, Dicentrarchus labrax, larvae to nursery areas. Journal of the Marine Biological Association of the United Kingdom, 72: 199–212.

Kennedy, M. and Fitzmaurice, P. 1968. Occurrence of eggs of bass, Dicentrarchus labrax, on the southern coasts of Ireland. Journal of the Marine Biological Association of the UK, 48: 585–592.

Kennedy, M. and Fitzmaurice, P. 1972. The biology of the bass, Dicentrarchus labrax in Irish wa-ters. Journal of the Marine Biological Association of the United Kingdom 52: 557–597.

Kupschus, S., Smith, M. T., Walmsley, S. A. 2008. Annex 2: Working Document. An update of the UK bass assessments 2007. Report of the Working Group on the Assessment of New MoU Species (WGNEW). By Correspondence, ICES CM 2008/ACOM:25. 77 pp.

Lam Hoai Thong. 1970. Contribution à l’étude des Bars de la région des Sables d’Olonne. Trav. Fac. Sci. Rennes, Ser. Océanogr. Biol., 3: 39–68.

Lancaster, J. E. 1991. The feeding ecology of juvenile bass Dicentrarchus labrax (L.). PhD thesis, University College of Swansea, 281 pp.

Laurec et al. 2012. Analysis of length distribution in sea bass for a given read age. WD to IBP-NEW 2012.

Mahé, K., Holmes, A., Huet, J., Sévin, K., Elleboode, R. 2012. Report of the Seabass (Dicentrachus labrax) Otolith and Scale Exchange Scheme 2011, 16 pp.

Masski, H. 1998. Identification de Frayères et Etude des Structures de Population de Turbot (Psetta maxima L.) et du Bar (Dicentrarchus labrax L.) en Manche Ouest et dans les Zones


Avoisinantes. Thèse présentée a la Faculte des Sciences de Brest. Universite de Bretagne Occdentale. 136pp + annexes.

Mayer, I., Shackley, S.E. and Witthames, P.R. 1990. Aspects of the reproductive biology of the bass, Dicentrarchus labrax L. II. Fecundity and pattern of oocyte development. J. Fish Biol. 36:141–148.

Methot, R.D. 2000. Technical Description of the Stock Synthesis Assessment Program. National Marine Fisheries Service, Seattle, WA. NOAA Tech Memo. NMFS-NWFSC-43: 46 pp.

Methot, R.D. 2011. User Manual for Stock Synthesis, Model Version 3.23b. NOAA Fisheries Ser-vice, Seattle. 167 pp.

Pawson, M. G. 1992. Climatic influences on the spawning success, growth and recruitment of bass (Dicentrarchus labrax L.) in British Waters. ICES mar. Science Symp. 195: 388–392.

Pawson, M. G., Kupschus, S. and Pickett, G. D. 2007a. The status of sea bass (Dicentrarchus labrax) stocks around England and Wales, derived using a separable catch-at-age model, and im-plications for fisheries management. ICES Journal of Marine Science 64, 346–356.

Pawson, M. G., Kelley, D. F. and Pickett, G. D. 1987. The distribution and migrations of bass Dicentrarchus labrax L. in waters around England and Wales as shown by tagging. J. Mar. Biol. Ass. UK, 67: 183–217.

Pawson, M. G., and Pickett, G. D. 1996. The annual pattern of condition and maturity in bass (Dicentrarchus labrax L) in waters around the UK. Journal of the Marine Biological Associa-tion of the United Kingdom, 76: 107–126.

Pickett, G.D. 1990. Assessment of the UK bass fishery using a log-book-based catch recording system. Fish. Res. Tech. Rep., MAFF Direct. Fish Res., Lowestoft 90: 30pp.

Pickett, G. D., and Pawson, M. G. 1994. Bass. Biology, Exploitation and Management. Chapman & Hall, London, Fish and Fisheries Series, 12. 358 pp.

Quirijns, F. and Bierman, S. 2012. Growth and maturity of sea bass sampled around the Nether-lands. Working Document: ICES IBP-NEW 2012; October 2012. 9pp.

Reynolds, W. J., Lancaster, J. E. and Pawson, M. G. 2003. Patterns of spawning and recruitment of bass to Bristol Channel nurseries in relation to the 1996 "Sea Empress" oil spill. J. Mar. Biol. Assoc. UK, 83: 1163–1170.

Rocklin et al. 2012. Assessment of the sea bass recreational catches using a large-scale network of volunteers. In prep.

Rocklin D, Levrel H, Drogou M, Herfaut J, Veron G. 2014. Combining Telephone Surveys and Fishing Catches Self-Report: The French Sea Bass Recreational Fishery Assessment. PLoS ONE 9(1): e87271. doi:10.1371/journal.pone.0087271.

Stequert, B., 1972. Contribution à l’étude du bar Dicentrarchus labrax (L.) des reservoirs à poissons de la région d’Arcachon. Th. 3ème year: Faculté des Sciences.

Thompson, B. M., Harrop, R. T. 1987. The distribution and abundance of bass (Dicentrarchus labrax) eggs and larvae in the English Channel and Southern North Sea. Journal of the Ma-rine Biological Association of the United Kingdom, 67, 263–274.

Tulp, I., Bolle, L. J. and Rijnsdorp, A.D. 2008. Signals from the shallows: In search of common patterns in long-term trends in Dutch estuarine and coastal fish. Journal of Sea Research 60 (1–2), pp. 54–73.

Van der Hammen, T and de Graaf, M. 2012. Recreational fishery in the Netherlands: catch esti-mates of cod (Gadus morhua) and eel (Anguilla anguilla) in 2010. IMARES Wageningen UR, Report Number C014/12, 61 pp.

van der Hammen, T., Poos, J. J., van Overzee H. M.J., Heessen, H. J.L. and Rijnsdorp A. D. 2013. Data evaluation of data limited stocks: Horse mackerel, Sea bass, Greater Silver Smelt, Tur-bot and Brill. IMARES report number C166/13.


van Beek, F. A., Rijnsdorp, A.D., de Clerck, R. 1989. Monitoring juvenile stocks of flatfish in the Wadden Sea and the coastal areas of the southeastern North Sea. HelgoländerMeeresunter-suchungen, 43 (3–4), pp. 461–477. doi: 10.1007/BF02365904.

Walmsley, S. and Armstrong, M. 2012. The UK commercial bass fishery in 2010. Working Docu-ment to ICES WGNEW 2012. August 2011.


Appendix 1

Content of Stock Synthesis Control File (BassIVVII.ctl) used at WGCSE 2014. The file was originally built from one used for other species, and some comments remain that are for those assessments. Rows preceded by # are skipped, and are greyed out.

#C growth parameters are estimated

#_SS-V3.04-safe;_09/09/09;_Stock_Synthesis_by_Richard_Methot_(NOAA);_using_ADMB_7.0.1

1 #_N_Growth_Patterns

1 #_N_Morphs_Within_GrowthPattern(GP)

#_Cond 1 #_Morph_between/within_stdev_ratio (no read if N_morphs=1)

#_Cond 1 #vector_Morphdist_(-1_in_first_val_gives_normal_approx)

#

##1 # N recruitment designs goes here if N_GP*nseas*area>1 #here 1 gp, 4 seasons, 1 area

##0 # placeholder for recruitment interaction request

#GP seas area for each recruitment assignment

##1 1 1 # example recruitment design element for GP=1, season=1, area=1

#

#_Cond 0 # N_movement_definitions goes here if N_areas > 1

#_Cond 1.0 # first age that moves (real age at begin of season, not integer) also cond on do_mi-gration>0

#_Cond 1 1 1 2 4 10 # example move definition for seas=1, morph=1, source=1 dest=2, age1=4, age2=10

#

0 #_Nblock_Patterns

#_blocks_per_pattern

# begin and end years of blocks in first pattern

#

0.5#_fracfemale #? Note sex ratio in bass increases with length.

3 #_natM_type:_0=1Parm; 1=N_breakpoints;_2=Lorenzen;_3=agespecific;_4=agespec_withsea-sinterpolate

0.150 0.150 0.150 0.152 0.163 0.204 0.241 0.249 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250

#_no additional input for selected M option; read 1P per morph

1 # GrowthModel: 1=vonBert with L1&L2; 2=Richards with L1&L2; 3=not implemented; 4=not implemented #note - maguire et al 2008 pg 1270, Downloaded from icesjms.oxfordjour-nals.org at ICES on October 17, 2011

2 #_Growth_Age_for_L1

999 #_Growth_Age_for_L2 (999 to use as Linf)


0 #_SD_add_to_LAA (set to 0.1 for SS2 V1.x compatibility)

0 #_CV_Growth_Pattern: 0 CV=f(LAA); 1 CV=F(A); 2 SD=F(LAA); 3 SD=F(A)

1 #_maturity_option: 1=length logistic; 2=age logistic; 3=read age-maturity matrix by growth_pattern; 4=read age-fecundity; 5=read fec and wt from wtatage.ss

#_placeholder for empirical age-maturity by growth pattern

4 #_First_Mature_Age

1 #_fecundity option:(1)eggs=Wt*(a+b*Wt);(2)eggs=a*L^b;(3)eggs=a*Wt^b

0 #_hermaphroditism option: 0=none; 1=age-specific fxn

1 #_parameter_offset_approach (1=none, 2= M, G, CV_G as offset from female-GP1, 3=like SS2 V1.x)

1 #_env/block/dev_adjust_method (1=standard; 2=logistic transform keeps in base parm bounds; 3=standard w/ no bound check)

#

#_growth_parms

#_LO HI INIT PRIOR PR_type SD PHASE env-var use_dev dev_minyr dev_maxyr dev_stddev Block Block_Fxn

#0.01 0.5 0.2 0.2 -1 0.1 -3 0 0 0 0 0 0 0 # NatM_p_1_GP_1

-1 30 19.67 19.67 -1 0.5 -5 0 0 0 0 0 0 0 # L_at_Amin_GP_1

60 100 80.26 80.26 -1 15 4 0 0 0 0 0 0 0 # L_at_Amax_GP_1

0.01 0.2 0.09699 0.09699 -1 0.05 -3 0 0 0 0 0 0 0 # VonBert_K_GP_1

0.005 0.5 0.2 0.2 -1 0.8 -6 0 0 0 0 0 0 0 # CV_young_GP_1

0.005 0.5 0.1 0.1 -1 0.8 -6 0 0 0 0 0 0 0 # CV_old_GP_1

-1 1 0.00001296 0.00001296 -1 0.05 -3 0 0 0 0 0 0 0 # Wtlen_1

2 4 2.969 2.969 -1 0.05 -3 0 0 0 0 0 0 0 # Wtlen_2

30 50 40.649 40.649 -1 5 -3 0 0 0 0 0 0 0 # Mat50%

-5 1 -0.33349 -0.33349 -1 0.03764 -3 0 0 0 0 0 0 0 # Mat_slope

-3 3 1 1 -1 0.8 -3 0 0 0 0 0 0 0 # Eg/gm_inter

-3 3 0 0 -1 0.8 -3 0 0 0 0 0 0 0 # Eg/gm_slope_wt

0 0 0 0 -1 0 -3 0 0 0 0 0 0 0 # RecrDist_GP_1

0 0 0 0 -1 0 -3 0 0 0 0 0 0 0 # RecrDist_Area_1

0 0 0 0 -1 0 -4 0 0 0 0 0 0 0 # RecrDist_Seas_1

0 0 0 0 -1 0 -4 0 0 0 0 0 0 0 # CohortGrowDev

#

#_Cond 0 #custom_MG-env_setup (0/1)

#_Cond -2 2 0 0 -1 99 -2 #_placeholder when no MG-environ parameters

#

#_Cond 0 #custom_MG-block_setup (0/1)

#_Cond -2 2 0 0 -1 99 -2 #_placeholder when no MG-block parameters

#_Cond No MG parm trends


#

#_seasonal_effects_on_biology_parms

0 0 0 0 0 0 0 0 0 0 #_femwtlen1,femwtlen2,mat1,mat2,fec1,fec2,L1,K

#_Cond -2 2 0 0 -1 99 -2 #_placeholder when no seasonal MG parameters

#

#-6 #_MGparm_Dev_Phase

#

#_Spawner-Recruitment

3 #_SR_function

#_LO HI INIT PRIOR PR_type SD PHASE

1 16 10 5 -1 1 1 # SR_R0

0.2 0.999 0.999 0.999 -1 0.2 -1 # SR_steep

0.1 2 0.9 0.9 -1 0.2 -5 # SR_sigmaR

-5 5 0 0 -1 1 -3 # SR_envlink

-5 5 0 -0.7 -1 2 -2 # SR_R1_offset

0 0 0 0 -1 0 -99 # SR_autocorr

0 #_SR_env_link

0 #_SR_env_target_0=none;1=devs;_2=R0;_3=steepness

1 #do_recdev: 0=none; 1=devvector; 2=simple deviations

1965 # first year of main recr_devs; early devs can preceed this era

2011 # last year of main recr_devs; forecast devs start in following year 2013. Youngest survey age 2gp 2013

2 #_recdev phase

0 # (0/1) to read 13 advanced options

#0 #_recdev_early_start (0=none; neg value makes relative to recdev_start)

#-4 #_recdev_early_phase

#0 #_forecast_recruitment phase (incl. late recr) (0 value resets to maxphase+1)

#1 #_lambda for prior_fore_recr occurring before endyr+1

#985 #_last_early_yr_nobias_adj_in_MPD

#1955 #_first_yr_fullbias_adj_in_MPD

#2010 #_last_yr_fullbias_adj_in_MPD 2012

#2011 #_first_re-cent_yr_nobias_adj_in_MPD 2013

#1 #_max_bias_adj_in_MPD (1.0 to mimic pre-2009 models)


#0 #_period of cycles in recruitment (N parms read below)

#-5 #min rec_dev

#5 #max rec_dev

#0 # 3 #_read_recdevs

# 2009 -0.04979586

# 2010 -0.04979586

# 2011 -0.04979586

#_end of advanced SR options

#

#_placeholder for full parameter lines for recruitment cycles

# read specified recr devs

#_Yr Input_value

#

#

#Fishing Mortality info

0.2 # F ballpark for tuning early phases

-2001 # F ballpark year (neg value to disable)

3 # F_Method: 1=Pope; 2=instan. F; 3=hybrid (hybrid is recommended)

2.9 # max F or harvest rate, depends on F_Method

# no additional F input needed for Fmethod 1

# if Fmethod=2; read overall start F value; overall phase; N detailed inputs to read

#0.3 3 0 # if Fmethod=3; read N iterations for tuning for Fmethod 3

5 # N iterations for tuning F in hybrid method (recommend 3 to 7)

#

#_initial_F_parms

#_LO HI INIT PRIOR PR_type SD PHASE

0.00001 2 0.03 0.3 -1 0.5 1 # InitF_OTB_Nets

0.00001 2 0.03 0.3 -1 0.5 1 # InitF_Midwater

0.0001 2 0.03 0.03 -1 0.5 1 # InitF_French

0.00001 2 0.03 0.03 -1 0.5 1 # InitF_Other

#

#_Q_setup

# Q_type options: <0=mirror, 0/1=float, 2=parameter, 3=parm_w_random_dev, 4=parm_w_rand-walk)

#_Den-dep env-var extra_se Q_type

0 0 0 0 # FISHERY1

0 0 0 0 # FISHERY2

0 0 0 0 # FISHERY3


0 0 0 0 # FISHERY4

0 0 0 0 # SURVEY AutBass 2

0 0 0 0 # survey AutBass 3

0 0 0 0 # Survey AutBass 4

0 0 0 0 # Survey CGFS1

#_Cond 0 #_If q has random component, then 0=read one parm for each fleet with random q; 1=read a parm for each year of index

#_Q_parms(if_any)

# LO HI INIT PRIOR PR_type SD PHASE

#-15 5 0 0 -1 1 -1 # Q_base_2_SURVEY EVHOE

#

#_size_selex_types

0 0 0 0 # UKTrawl_Nets #_RDM now all fleets have size selectivity

0 0 0 0 # UKMidwater

1 0 0 0 # French

15 0 0 3 # Other

0 0 0 0 # AutBass2

0 0 0 0 # AutBass3

0 0 0 0 # AutBass4

24 0 0 0 # CGFS1

#

#_age_selex_types

#_Pattern ___ Male Special

20 0 0 0 # 1 UKTrawl

12 0 0 0 # 2 UKMidwater

10 0 0 0 # 3 French

15 0 0 3 # 4 Other

11 0 0 0 # 5 AutBass2

11 0 0 0 # 6 AutBass3

11 0 0 0 # 7 AutBass4

10 0 0 0 # 8 CGFS1

#_LO HI INIT PRIOR PR_type SD PHASE env-var use_dev dev_minyr dev_maxyr dev_stddev Block Block_Fxn

20 91 38.8836 30 -1 0.1 2 0 0 0 0 0 0 0 # SizeSel_5P_1_French

0.01 30 5.18514 5 -1 0.01 3 0 0 0 0 0 0 0 # SizeSel_5P_2_French


20 94.1 32 32 -1 0.05 2 0 0 0 0 0 0 0 # SizeSel_2P_1_CGFS1

-6.0 4.0 -6.0 -6.0 -1 0.05 3 0 0 0 0 0 0 0 # SizeSel_2P_2_CGFS1

1.0 5.0 3.3 3.3 -1 0.05 3 0 0 0 0 0 0 0 # SizeSel_2P_3_CGFS1

3.0 6.0 4.4 4.4 -1 0.05 3 0 0 0 0 0 0 0 # SizeSel_2P_4_CGFS1

-8.0 9.0 -8 -8 -1 0.05 2 0 0 0 0 0 0 0 # SizeSel_2P_5_CGFS1

-5.0 9.0 -1 -1 -1 0.05 2 0 0 0 0 0 0 0 # SizeSel_2P_6_CGFS1

0 15.8 7.4 7.4 -1 0.05 2 0 0 0 0 0 0 0 # PEAK value SizeSel_1P_1_OTB

-6.0 4.0 -6.0 -6.0 -1 0.05 3 0 0 0 0 0 0 0 # TOP logistic SizeSel_1P_1_OTB

-3 9.0 0.3 0.3 -1 0.05 3 0 0 0 0 0 0 0 # WIDTH exp SizeSel_1P_1_OTB

-1.0 9.0 -0.8 -0.8 -1 0.05 3 0 0 0 0 0 0 0 # WIDTH exp SizeSel_1P_1_OTB

-10.0 9.0 -5 -5 -1 0.05 2 0 0 0 0 0 0 0 # INIT logistic SizeSel_1P_1_OTB

-5.0 9.0 1.8 1.8 -1 0.05 2 0 0 0 0 0 0 0 # FINAL logistic SizeSel_1P_1_OTB

0 16 7.0 7.0 -1 0.05 2 0 0 0 0 0 0 0 # SizeSel_2P_1_MWT

0.01 30 7.02399 5 -1 0.05 3 0 0 0 0 0 0 0 # SizeSel_2P_2_MWT

2 2 2 2 -1 99 -3 0 0 0 0 0 0 0 # AgeSel_10P_1_Autumn 2 min age

2 2 2 2 -1 99 -3 0 0 0 0 0 0 0 # AgeSel_10P_2_Autumn 2 max age





#_Cond 0 #_custom_sel-env_setup (0/1)

#_Cond -2 2 0 0 -1 99 -2 #_placeholder when no enviro fxns

#_custom_sel-blk_setup (0/1)

#_Cond No selex parm trends

#_Cond -4 # placeholder for selparm_Dev_Phase

#_env/block/dev_adjust_method (1=standard; 2=logistic trans to keep in base parm bounds; 3=standard w/ no bound check)

#

# Tag loss and Tag reporting parameters go next

0 # TG_custom: 0=no read; 1=read if tags exist

#_Cond -6 6 1 1 2 0.01 -4 0 0 0 0 0 0 0 #_placeholder if no parameters


#

1 #_Variance_adjustments_to_input_values

#_fleet/svy: 1 2 3 4 5 6 7 8

0 0 0 0 0 0 0 0 #_add_to_survey_CV

0 0 0 0 0 0 0 0 #_add_to_discard_stddev

0 0 0 0 0 0 0 0 #_add_to_bodywt_CV

1 1 1 1 1 1 1 1 #_mult_by_lencomp_N

1 1 1 1 1 1 1 1 #_mult_by_agecomp_N

1 1 1 1 1 1 1 1 #_mult_by_size-at-age_N

#

3 #_maxlambdaphase

1 #_sd_offset

#

0 #8 # number of changes to make to default Lambdas (default value is 1.0)

# Like_comp codes: 1=surv; 2=disc; 3=mnwt; 4=length; 5=age; 6=SizeFreq; 7=sizeage; 8=catch;

# 9=init_equ_catch; 10=recrdev; 11=parm_prior; 12=parm_dev; 13=CrashPen; 14=Morphcomp; 15=Tag-comp; 16=Tag-negbin

#like_comp fleet/survey phase value sizefreq_method

# 5 1 1 0.1 1 #_RDM reduce emphasis on age comp and wt-at-age by 10x

# 5 2 1 0.1 1

# 5 3 1 0.1 1

# 5 4 1 0.1 1

# 7 1 1 0.1 1

# 7 2 1 0.1 1

# 7 3 1 0.1 1

# 7 4 1 0.1 1

#

# lambdas (for info only; columns are phases)

# 0 0 0 0 #_CPUE/survey:_1

# 1 1 1 1 #_CPUE/survey:_2

# 1 1 1 1 #_CPUE/survey:_3

# 1 1 1 1 #_lencomp:_1

# 1 1 1 1 #_lencomp:_2

# 0 0 0 0 #_lencomp:_3

# 1 1 1 1 #_agecomp:_1

# 1 1 1 1 #_agecomp:_2

# 0 0 0 0 #_agecomp:_3

# 1 1 1 1 #_size-age:_1


# 1 1 1 1 #_size-age:_2

# 0 0 0 0 #_size-age:_3

# 1 1 1 1 #_init_equ_catch

# 1 1 1 1 #_recruitments

# 1 1 1 1 #_parameter-priors

# 1 1 1 1 #_parameter-dev-vectors

# 1 1 1 1 #_crashPenLambda

0 # (0/1) read specs for more stddev reporting

# 1 1 -1 5 1 5 1 -1 5 # selex type, len/age, year, N selex bins, Growth pattern, N growth ages, NatAge_area(-1 for all), NatAge_yr, N Natages

# 5 15 25 35 43 # vector with selex std bin picks (-1 in first bin to self-generate)

# 1 2 14 26 40 # vector with growth std bin picks (-1 in first bin to self-generate)

# 1 2 14 26 40 # vector with NatAge std bin picks (-1 in first bin to self-generate) 999


Appendix 2

Content of Stock Synthesis Data file (BassIVVII.dat) used at WGCSE 2014. The file was originally built from one used for other species, and some comments remain that are for those assessments. Rows preceded by # are skipped, and are greyed out.

#C Sea bass IV VII input data file

1985 # _styr lr

2014 # _endyr lr

1 # _nseasons lr number of quarters in a year

12 # _nmonths per season lr months in each quarter

1 # _spawn_seas ? which season is the spawning season

4 # _Nfleet #lr

4 # _Nsurveys #lr

1 # _N_areas # lr

UKOTB_Nets_Lines%UKMWT%French%Other%AutBass2%AutBass3%AutBass4%CGFS1

-1 -1 -1 -1 0.83 0.83 0.83 0.75 # _surveytiming_in_season

1 1 1 1 1 1 1 1 # _area_assignments_for_each_fishery_and_survey

1 1 1 1 # _units of catch: 1=bio; 2=num

0.1 0.1 0.1 0.1 # _se of log(catch) only used for init_eq_catch and for Fmethod 2 and 3

1 #_Ngenders # lr (?if split by male female then 2? lr)

30 #_Nages # lr

48.7 0.51 709.68 118.76 #CF _init_equil_catch_for_each_fishery (82% of 1985 value)

29 # _N_lines_of_catch_to_read #lr updated WGCSE 2014

# _catch_biomass(tons):_columns_are_ fisheries units ,year,season Updated WGCSE 2014

59.72 0.62 870.38 145.61 1985 1

115.66 2.16 1180.34 17.27 1986 1

118.7 0.02 1839.69 20.9 1987 1

164.12 7.72 1027.77 38.61 1988 1

181.52 9.13 916.96 53.14 1989 1

136.47 22.79 849.03 25.2 1990 1

222.1 14.49 970.72 17.32 1991 1

138.88 7.93 1000.57 37.07 1992 1

222.63 1.01 979.35 48.26 1993 1

523.9 0.4 786.31 59.74 1994 1

609.6 1.32 1056.62 109.63 1995 1

458.36 87.29 2395.39 82.31 1996 1

473.06 71.4 1984.07 91.37 1997 1

386.89 84.73 1773.01 143.47 1998 1


434.08 219.55 1842.63 168.35 1999 1

311.96 52.13 1805.61 227.41 2000 1

340.31 95.47 1883.4 162.36 2001 1

495.71 109 1824.17 198.72 2002 1

440.53 126.81 2470.88 406.81 2003 1

480.31 130.76 2604.02 514.61 2004 1

394.77 78.29 3161.45 757.11 2005 1

506.03 32.8 3258.97 724.37 2006 1

607.42 63.93 2770.24 771.77 2007 1

714.21 19.71 2749.77 760.36 2008 1

643.13 11.09 2649.48 709.33 2009 1

634.15 41.93 3236.34 845.17 2010 1

611.2 97.5 2525.7 635.33 2011 1

755.1 49 2606.1 650 2012 1

726 39 2770.3 596 2013 1

################################

98 #_N_cpue_and_surveyabundance_observations #lr removed 2 lines was 92 instead of 90. Up-dated WGCSE 2014

#_Units: 0=numbers; 1=biomass; 2=F

#_Errtype: -1=normal; 0=lognormal; >0=T

#_Fleet Units Errtype

1 1 0 # UK OTB_Nets

2 1 0 # UK MWT

3 1 0 # French fleets

4 1 0 # Other

5 1 0 # AutBass2

6 0 0 # AutBass3

7 0 0 # AutBass4

8 0 0 # CGFS1

################################

# yr qtr indexNumber(7-16) indexResult indexCV

## AutBass (numbers 2):

1986 1 5 0.27 0.433295234

1987 1 5 0.05 0.433295234

1989 1 5 6.68 0.433295234

1990 1 5 2.81 0.433295234

1991 1 5 3.08 0.433295234


1992 1 5 0.95 0.433295234

1993 1 5 6.65 0.433295234

1994 1 5 3.33 0.433295234

1995 1 5 4.83 0.433295234

1996 1 5 5.52 0.433295234

1997 1 5 33.62 0.433295234

1998 1 5 1.22 0.433295234

1999 1 5 19.37 0.433295234

2000 1 5 6.07 0.433295234

2001 1 5 34.42 0.433295234

2002 1 5 7.42 0.433295234

2003 1 5 8.37 0.433295234

2005 1 5 13.12 0.433295234

2006 1 5 9.51 0.433295234

2007 1 5 3.42 0.433295234

2008 1 5 18.52 0.433295234

2009 1 5 13.25 0.433295234

2011 1 5 2.25 0.43

2013 1 5 1.34 0.43


1986 1 6 4.26 0.433295234

1987 1 6 0.28 0.433295234

1989 1 6 0.37 0.433295234

1990 1 6 1.15 0.433295234

1991 1 6 0.21 0.433295234

1992 1 6 18.59 0.433295234

1993 1 6 3.59 0.433295234

1994 1 6 1.84 0.433295234

1995 1 6 4.69 0.433295234

1996 1 6 0.43 0.433295234

1997 1 6 4.52 0.433295234

1998 1 6 5.5 0.433295234

1999 1 6 0.67 0.433295234

2000 1 6 11.35 0.433295234

2001 1 6 3.92 0.433295234

2002 1 6 3.87 0.433295234

2003 1 6 4.6 0.433295234

2005 1 6 7.98 0.433295234


2006 1 6 9.21 0.433295234

2007 1 6 1.78 0.433295234

2008 1 6 6.66 0.433295234

2009 1 6 6.25 0.433295234

2011 1 6 1.39 0.43

2013 1 6 0.08 0.43


1986 1 7 1.31 0.433295234

1987 1 7 2.27 0.433295234

1989 1 7 0.0037 0.433295234

1990 1 7 0.02 0.433295234

1991 1 7 0.03 0.433295234

1992 1 7 0.16 0.433295234

1993 1 7 4.39 0.433295234

1994 1 7 0.29 0.433295234

1995 1 7 0.72 0.433295234

1996 1 7 0.11 0.433295234

1997 1 7 0.06 0.433295234

1998 1 7 0.61 0.433295234

1999 1 7 0.87 0.433295234

2000 1 7 0.03 0.433295234

2001 1 7 1.57 0.433295234

2002 1 7 0.4 0.433295234

2003 1 7 0.59 0.433295234

2005 1 7 0.84 0.433295234

2006 1 7 1.02 0.433295234

2007 1 7 0.3 0.433295234

2008 1 7 0.34 0.433295234

2009 1 7 0.33 0.433295234

2011 1 7 0.42 0.43

2013 1 7 0.1 0.43

## CGFS1:

1988 1 8 245776 0.6

1989 1 8 77716 0.6

1990 1 8 1129914 0.6

1991 1 8 4250635 0.3

1992 1 8 2617984 0.3

1993 1 8 2299918 0.3


1994 1 8 1097829 0.3

1995 1 8 1021740 0.3

1996 1 8 1224238 0.3

1997 1 8 1817599 0.3

1998 1 8 2531044 0.3

1999 1 8 1642270 0.3

2000 1 8 2570996 0.3

2001 1 8 3150674 0.3

2002 1 8 3872427 0.3

2003 1 8 8739057 0.3

2004 1 8 3598440 0.3

2005 1 8 3005317 0.3

2006 1 8 5517999 0.3

2007 1 8 3661314 0.3

2008 1 8 6468841 0.3

2009 1 8 2564696 0.3

2010 1 8 1804537 0.3

2011 1 8 1513745 0.3

2012 1 8 2034554 0.3

2013 1 8 995987 0.3

0 # _N_fleets_with_discard

0 # N discard obs

0 # _N_meanbodywt_obs

30 #_DF_for_meanbodywt_T-distribution_like

3 # length bin method: 1=use databins; 2=generate from binwidth,min,max below; 3=read vector

49 # number of population length bins to be read

4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 96 98 100

-0.001 #_comp_tail_compression

1e-007 #_add_to_comp

0 #_combine males into females at or below this bin number

41 #_N_LengthBins

14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 92 94

40 #_N_Length_obs #lr Updated WGCSE 2014


#Yr Season Flt/Svy Gender Part Nsamp datavector(female-male)

##### LANDINGS OF COMMERCIAL FLEETS (1 TO 5), ORDERED BY YEAR AND QUARTER:

############################################################################

2000 1 3 0 2 200 0 0 0 0 0 0 0 0 0 0 9931 34932 85866 126730 102836 80478 93344 80934 55399 52948 42094 26460 27357 23581 14295 18044 10773 9903 5709 5721 2345 2595 2102 888 1021 548 123 0 0 0 0

2001 1 3 0 2 200 0 0 0 0 0 0 0 0 0 0 17962 19809 68920 76594 98008 109595 106857 77694 57055 51658 36737 35839 22762 25834 18773 13532 11068 9120 11771 5733 5345 2782 1691 583 296 204 0 61 0 0 0

2002 1 3 0 2 200 0 0 0 0 0 0 0 0 1015 0 12469 38249 46427 62503 82461 91064 86723 62163 55905 46180 35998 26001 19019 14210 11129 16771 11011 5447 4795 4559 1825 1260 357 155 109 0 0 0 0 0 0

2003 1 3 0 2 200 0 0 0 0 0 0 0 3455 13054 58717 105655 125326 180475 119495 145456 104545 130023 115806 91915 93878 48742 60839 31614 33688 30691 18823 13230 7960 5374 5617 3275 1356 297 783 112 148 0 0 0 0 0

2004 1 3 0 2 200 0 0 0 0 0 0 0 0 14 13057 78811 127801 124051 227214 282390 243107 188494 126685 72581 82331 50633 60284 31334 19126 23996 14799 10650 8569 4880 2974 2675 2567 548 425 149 295 0 149 0 0 0

2005 1 3 0 2 200 0 0 0 0 0 0 0 0 0 9903 29872 97890 128022 231750 266905 344681 270532 239265 169478 115269 62106 67741 61132 43591 35774 25788 12456 13360 8908 8053 9811 5020 2378 1365 107 0 0 0 0 0 0

2006 1 3 0 2 200 0 0 0 0 0 0 0 0 15689 32459 179130 285704 217657 178250 196868 289998 285451 263272 200874 119836 99509 99674 54522 45908 23763 20607 14969 13976 9653 4521 3424 2883 731 201 261 30 0 0 0 0 0

2007 1 3 0 2 200 0 0 0 0 0 0 0 0 0 181 4715 39335 102714 146272 145122 164011 130859 100043 99210 75929 74405 55147 46087 28056 23057 18091 8715 8793 4835 2707 1962 1010 399 158 37 59 0 0 0 0 0

2008 1 3 0 2 200 0 0 0 0 0 0 0 0 0 8250 28986 229758 263071 266408 237160 270810 228996 142650 112385 74336 66260 48853 39689 29840 28335 14420 12694 9039 6821 4714 1623 1257 534 261 8 0 0 0 0 0 0


2009 1 3 0 2 200 0 0 0 0 0 0 0 292 473 2239 10714 124925 211881 225545 193030 222613 238849 155222 159658 114530 84649 96257 51578 36547 57472 24016 21415 27466 20198 12083 7551 979 1765 264 1004 0 0 0 0 0 0

2010 1 3 0 2 200 0 0 0 717 0 0 0 0 0 9811 28290 169311 177571 182105 283064 251956 230227 188149 186310 109212 120550 71590 62211 31544 19076 62005 26388 9340 8541 29128 1884 2114 182 5525 6097 863 0 1207 0 0 0

2011 1 3 0 2 200 0 0 0 0 0 0 0 0 0 1976 13885 57121 87842 128838 187586 201447 199487 194697 145447 124239 92526 72471 46869 31690 19998 17624 14720 7906 6114 2082 1163 1096 476 148 104 0 0 0 0 0 0

2012 1 3 0 2 200 0 0 0 0 0 0 0 1219 0 1583 6518 85760 172510 140273 147895 162333 180752 158490 130759 107214 90638 78934 54869 35387 33085 17714 15170 9374 8114 4147 2313 1540 1134 282 451 29 27 0 0 0 0

2013 1 3 0 2 200 0 0 0 0 0 0 0 0 239 0 1617 26153 130687 272401 311983 234498 270538 187922 166217 140323 115326 118018 78259 132959 61538 79165 46759 31584 32919 21221 11918 3418 2153 236 2743 0 193 0 0 0 0

-1988 1 8 0 0 6 0 0 0 0 0 0 0 0 0 0 0 58885 0 76121 20143 0 0 0 20143 0 39616 0 15434 15434 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

-1989 1 8 0 0 3 0 0 0 0 0 0 0 0 0 22235 0 22235 0 11011 0 0 0 0 22235 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

-1990 1 8 0 0 8 0 0 32733 98199 120021 96615 222797 345962 107244 0 0 18463 0 6106 0 6106 23906 0 17254 17254 17254 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

1991 1 8 0 0 19 0 0 0 0 0 304624 1719409 1119677 531390 110365 162253 126980 45607 16723 16703 10625 0 16885 0 5352 30272 16885 0 0 0 0 0 16885 0 0 0 0 0 0 0 0 0 0 0 0 0

1992 1 8 0 0 23 0 0 0 0 0 0 0 63079 287513 510612 638459 530944 309090 78328 79289 26490 42201 0 0 15138 5247 0 10494 0 0 0 15853 0 0 5247 0 0 0 0 0 0 0 0 0 0 0

1993 1 8 0 0 21 0 0 0 0 0 5551 38106 121556 170213 347228 237317 321637 247396 266632


204261 133186 158039 17036 0 0 14798 0 0 16962 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

1994 1 8 0 0 19 0 0 0 18458 0 5108 71094 30647 64428 59303 79319 77545 139680 69887 157500 113756 52799 63325 37992 18996 18996 0 0 0 0 18996 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

1995 1 8 0 0 17 0 0 0 0 0 65628 182886 191483 118285 68606 94174 58735 62595 72147 8650 15487 10698 0 20034 15487 0 0 17299 0 0 0 19546 0 0 0 0 0 0 0 0 0 0 0 0 0 0

1996 1 8 0 0 26 0 0 0 0 10096 71187 198585 162449 209883 102756 137006 50619 60715 0 46710 0 25805 38406 67089 0 0 0 21710 0 0 0 0 0 0 0 21222 0 0 0 0 0 0 0 0 0 0

1997 1 8 0 0 31 0 0 0 0 0 52420 172277 157267 364320 311146 265276 102658 136668 71024 62096 42836 3198 27070 0 36775 0 0 0 0 0 0 0 0 12568 0 0 0 0 0 0 0 0 0 0 0 0

1998 1 8 0 0 38 0 0 0 0 0 0 10254 42156 199875 248028 445903 469290 406992 271169 77307 56211 109682 49306 13620 13620 0 25037 27240 13620 13620 0 0 0 11417 0 26697 0 0 0 0 0 0 0 0 0 0

1999 1 8 0 0 37 0 0 0 0 18433 15128 83225 167646 146928 90595 177558 153662 195992 214407 105858 79810 47256 78075 7863 17639 0 9059 0 9059 0 15018 0 0 0 0 0 0 0 0 9059 0 0 0 0 0 0

2000 1 8 0 0 36 0 0 0 18644 65257 62040 125296 417455 440978 309620 338993 204879 181687 95559 108703 70962 23940 21506 14766 0 24695 0 14306 0 9307 13096 0 9307 0 0 0 0 0 0 0 0 0 0 0 0 0

2001 1 8 0 0 39 0 0 0 0 0 59740 413187 767358 518235 300493 248669 158653 247952 66992 125042 43191 37193 10236 24597 0 14837 24648 13083 52462 0 0 12053 0 0 12053 0 0 0 0 0 0 0 0 0 0 0

2002 1 8 0 0 44 0 0 0 0 0 10739 67655 197986 381355 832988 623508 439729 380196 270697 189153 102893 112164 50660 44892 16171 27189 21612 15932 12167 0 27484 12167 25511 0 0 9579 0 0 0 0 0 0 0 0 0 0

2003 1 8 0 0 41 0 0 0 0 22663 13218 161174 795498 720874 802826 1375758 1218315 1251975 623889 464021 274147 207776 203176 72670 136756 185043 89875


52508 26254 27514 0 13127 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2004 1 8 0 0 44 0 0 0 0 0 2692 0 122879 403673 714884 730822 416792 287359 324966 129278 137035 115126 31042 39375 29156 18125 9347 0 61811 2692 2692 9347 0 0 0 0 0 0 9347 0 0 0 0 0 0 0

2005 1 8 0 0 40 0 0 0 0 15565 0 210994 418270 315562 267295 354515 215506 225666 181642 106404 138927 146271 109108 98370 34142 9487 52772 51553 21543 10751 10751 0 0 10223 0 0 0 0 0 0 0 0 0 0 0 0

2006 1 8 0 0 36 0 0 0 0 0 9814 197122 866753 893641 788190 1120577 531053 311464 137430 190156 96298 71441 84223 86273 31873 22996 25308 19628 9814 3032 0 9814 0 0 11099 0 0 0 0 0 0 0 0 0 0 0

2007 1 8 0 0 33 0 0 0 0 0 0 47391 407308 386452 394375 609358 479157 314000 219897 169994 170481 111148 80056 57131 123245 25737 18311 0 11929 0 10662 5331 19351 0 0 0 0 0 0 0 0 0 0 0 0 0

2008 1 8 0 0 40 0 0 0 0 10729 52876 128066 266412 229414 731914 1661470 931172 880632 758996 224765 208707 112521 79753 13259 72807 14967 28587 19628 0 0 0 42166 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2009 1 8 0 0 26 0 0 0 0 0 0 21413 77225 253598 386804 279160 234313 217371 238780 238652 168707 110777 82775 99010 56555 38480 0 21413 12158 0 0 0 12249 0 0 15256 0 0 0 0 0 0 0 0 0 0

2010 1 8 0 0 30 0 0 0 0 0 0 20661 79020 106240 232673 118288 117886 316567 207944 97593 110810 191255 62479 14580 62090 0 22711 0 14580 0 14580 0 0 0 0 14580 0 0 0 0 0 0 0 0 0 0

2011 1 8 0 0 27 0 0 0 0 0 0 13881 0 45451 95055 204438 227355 208472 118040 120428 68027 88143 90627 60054 47411 9869 0 33403 36907 46184 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2012 1 8 0 0 25 0 0 0 0 0 0 29922 15009 161369 400458 247182 200887 183315 84361 197280 103031 86442 77382 31957 43195 19956 39781 11851 40610 19956 20305 20305 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2013 1 8 0 0 19 0 0 0 0 0 0 42825 113195 62476 113528 127750 96177 135950 70592 33717 44279 61410 11237 17382 17382 48087 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0


############################################################################

17 #_N_age_bins

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

1 #_N_ageerror_definitions

0.5 1.5 2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5 10.5 11.5 12.5 13.5 14.5 15.5 16.5 17.5 18.5 19.5 20.5 21.5 22.5 23.5 24.5 25.5 26.5 27.5 28.5 29.5 30.5

0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1

45 #_N_Agecomp_obs Updated WGCSE 2014

1 #_Lbin_method: 1=poplenbins; 2=datalenbins; 3=lengths

0 #_combine males into females at or below this bin number

#Yr Seas Flt/Svy Gender Part Ageerr Lbin_lo Lbin_hi Nsamp datavector(female-male)

1985 1 1 0 0 1 -1 -1 20 0 0 39 11855 24729 4971 7840 1115 2464 8117 1957 1998 1274 388 940 290 881

1986 1 1 0 0 1 -1 -1 20 0 0 0 4884 27887 15561 3502 9222 1468 5205 16306 2422 2496 3194 962 1339 3200

1987 1 1 0 0 1 -1 -1 20 0 0 0 538 36640 58747 16002 2398 3049 921 1349 7020 2644 1661 680 854 3144

1988 1 1 0 0 1 -1 -1 20 0 0 0 2002 22412 63428 41135 13637 2189 3383 1028 973 8536 3208 1898 1323 3962

1989 1 1 0 0 1 -1 -1 20 0 0 36098 5449 1940 13675 52948 30807 11658 4352 4018 2310 3753 10735 1659 1452 3893

1990 1 1 0 0 1 -1 -1 20 0 0 0 2211 2456 2443 10075 32009 22045 5886 2407 1726 1059 1388 5816 451 3327

1991 1 1 0 0 1 -1 -1 100 0 0 3297 14161 46666 2441 2480 9079 27634 22309 8985 912 3381 3336 1399 9320 4874

1992 1 1 0 0 1 -1 -1 100 0 0 0 17906 60212 38948 2021 985 2539 6440 6535 1759 546 529 258 581 4676

1993 1 1 0 0 1 -1 -1 100 0 0 20 1116 102649 83638 30222 1624 860 2851 8286 6615 2316 960 1453 1231 5544


1994 1 1 0 0 1 -1 -1 100 0 0 2 1296 33559 441976 57214 26998 2078 432 3844 10956 8893 2830 1341 983 6488

1995 1 1 0 0 1 -1 -1 100 0 0 0 5739 38805 91413 431628 28024 13059 1411 1040 2136 8541 7158 3551 4314 7272

1996 1 1 0 0 1 -1 -1 100 0 0 145 9340 41876 32357 89255 251801 12051 6450 352 794 970 4055 2652 1238 3457

1997 1 1 0 0 1 -1 -1 100 0 0 330 4027 10393 73598 50958 51074 186353 11996 5953 927 955 589 4462 3208 4262

1998 1 1 0 0 1 -1 -1 100 0 0 0 13955 58518 55500 105347 35312 26571 71835 5360 1616 202 329 1167 1931 2542

1999 1 1 0 0 1 -1 -1 100 0 0 346 140 97612 160695 56563 50986 15824 13035 43407 3277 2153 260 807 548 3431

2000 1 1 0 0 1 -1 -1 100 0 0 0 9389 2804 168794 84314 17784 18662 7809 8269 15820 1410 379 26 90 1105

2001 1 1 0 0 1 -1 -1 100 0 0 488 9410 80244 32923 123218 30484 10730 12972 8458 12042 13359 1464 1157 112 1196

2002 1 1 0 0 1 -1 -1 100 0 0 522 15068 55187 284390 21080 92341 24876 10798 12961 4029 5697 14855 823 348 876

2003 1 1 0 0 1 -1 -1 100 0 0 0 6054 71838 68737 181978 11049 53668 19350 6287 4653 3870 2998 6685 1071 911

2004 1 1 0 0 1 -1 -1 100 0 0 0 4868 44567 223462 84721 117748 4530 16798 8033 3022 2016 2195 728 2747 410

2005 1 1 0 0 1 -1 -1 100 0 0 0 13942 107126 69643 117337 40220 59321 6590 15664 2820 1599 2379 703 676 933

2006 1 1 0 0 1 -1 -1 100 0 0 0 18742 133803 158129 61442 88113 25791 41969 3889 6116 3389 1138 591 930 3977

2007 1 1 0 0 1 -1 -1 100 0 0 0 947 41363 208202 127765 59388 51729 20972 23751 15466 7284 5952 87 535 849

2008 1 1 0 0 1 -1 -1 100 0 0 0 10123 119526 326706 183313 69770 31832 29029 13352 10641 2043 1730 4019 478 1703

2009 1 1 0 0 1 -1 -1 100 0 0 0 2816 78121 183363 197627 86301 29037 22777 12978 12790 7765 2962 2268 0 225


2010 1 1 0 0 1 -1 -1 100 0 0 0 1067 77792 188027 162311 111066 42029 15853 8777 8349 3153 2497 2515 577 1761

2011 1 1 0 0 1 -1 -1 100 0 0 0 841 32410 131931 111148 84747 67313 41381 15244 16038 8388 4829 4086 1212 2092

2012 1 1 0 0 1 -1 -1 100 0 0 0 1690 15902 187155 248719 82602 59754 47075 24028 16085 6655 6120 4371 1056 1634

2013 1 1 0 0 1 -1 -1 100 0 0 0 0 46413 67860 219457 150956 51911 45186 33248 28066 17913 6654 5558 4380 4008

1996 1 2 0 0 1 -1 -1 50 0 0 0 0 289 796 3892 71665 5583 1648 21 334 154 622 485 199 560

1998 1 2 0 0 1 -1 -1 50 0 0 0 0 264 6405 12691 9161 8714 26925 2696 370 100 57 128 957 614

1999 1 2 0 0 1 -1 -1 50 0 0 0 0 2988 18438 15167 27342 13892 18263 43646 4481 1695 324 387 308 2762

2000 1 2 0 0 1 -1 -1 50 0 0 0 15 60 2475 7585 3269 4496 1459 2829 7075 633 174 39 96 421

2001 1 2 0 0 1 -1 -1 50 0 0 0 0 176 884 19449 19953 6925 5181 3072 2797 9505 843 625 121 258

2002 1 2 0 0 1 -1 -1 50 0 0 0 3 37 2349 1558 23776 9568 6215 5901 444 5591 9096 0 0 524

2003 1 2 0 0 1 -1 -1 50 0 0 0 0 2503 9885 36543 7420 37748 9582 2943 3029 576 157 3764 72 262

2004 1 2 0 0 1 -1 -1 50 0 0 0 7 1310 13054 15006 50232 3340 21608 8363 369 1876 1192 95 1011 20

2005 1 2 0 0 1 -1 -1 50 0 0 0 0 130 2404 17514 16457 19899 2180 5924 0 2093 113 0 45 686

2006 1 2 0 0 1 -1 -1 50 0 0 0 0 105 263 282 1892 660 1432 118 82 52 0 0 0 26

2007 1 2 0 0 1 -1 -1 50 0 0 0 0 659 4305 12037 9213 11685 4780 3249 1079 1380 21 64 32 207

2008 1 2 0 0 1 -1 -1 50 0 0 0 52 513 1775 3779 2060 1627 1794 870 1106 35 211 565 0 47


2009 1 2 0 0 1 -1 -1 50 0 0 0 0 101 712 2439 2911 945 880 189 334 194 12 190 0 1

2010 1 2 0 0 1 -1 -1 50 0 0 0 9 36 1741 5545 8261 6677 4755 403 3786 152 294 313 551 51

2011 1 2 0 0 1 -1 -1 50 0 0 0 0 255 4397 10231 13639 15908 13641 4424 4232 2773 1688 1003 264 424

2012 1 2 0 0 1 -1 -1 50 0 0 0 0 391 4456 10762 10003 8746 5782 2738 1133 289 433 143 127 226

0 #_N_MeanSize-at-Age_obs

#Yr Seas Flt/Svy Gender Part Ageerr Ignore datavector(female-male)

# samplesize(female-male)

0 #_N_environ_variables

0 #_N_environ_obs

0 # N sizefreq methods to read

0 # no tag data

0 # no morphcomp data

999 # end of data file marker


Appendix 3

Content of Stock Synthesis Starter file (Starter.SS) used at WGCSE 2014.

#V3.23b

#C Bass initial assessment

BassIVVII.dat

BassIVVII.ctl

0 # 0=use init values in control file; 1=use ss3.par

1 # run display detail (0,1,2)

1 # detailed age-structured reports in REPORT.SSO (0,1)

0 # write detailed checkup.sso file (0,1)

4 # write parm values to ParmTrace.sso (0=no,1=good,active; 2=good,all; 3=every_iter,all_parms; 4=every,active)

1 # write to cumreport.sso (0=no,1=like&time-series; 2=add survey fits)

1 # Include prior_like for non-estimated parameters (0,1)

1 # Use Soft Boundaries to aid convergence (0,1) (recommended)

2 # Number of bootstrap datafiles to produce

8 # Turn off estimation for parameters entering after this phase

10 # MCMC burn interval

2 # MCMC thin interval

0 # jitter initial parm value by this fraction

-1 # min yr for sdreport outputs (-1 for styr)

-1 # max yr for sdreport outputs (-1 for endyr; -2 for endyr+Nforecastyrs

0 # N individual STD years

#vector of year values

0.0001 #0.0001 # final convergence criteria (e.g. 1.0e-04)

0 # retrospective year relative to end year (e.g. -4)

0 # min age for calc of summary biomass

2 # Depletion basis: denom is: 0=skip; 1=rel X*B0; 2=rel X*Bmsy; 3=rel X*B_styr

0.4 # Fraction (X) for Depletion denominator (e.g. 0.4)

2 # SPR_report_basis: 0=skip; 1=(1-SPR)/(1-SPR_tgt); 2=(1-SPR)/(1-SPR_MSY); 3=(1-SPR)/(1-SPR_Btarget); 4=rawSPR

2 # F_report_units: 0=skip; 1=exploitation(Bio); 2=exploitation(Num); 3=sum(Frates)

#COND 3 10 #_min and max age over which average F will be calculated with F_reporting=4

0 # F_report_basis: 0=raw; 1=F/Fspr; 2=F/FMSY ; 3=F/Fbtgt 999 # check value for end of file


Appendix 4

Content of Stock Synthesis Forecast file (Forecast.SS) used at WGCSE 2014. This is not used for creating forecasts yet, but has to be available.

#V3.23b

# for all year entries except rebuilder; enter either: actual year, -999 for styr, 0 for endyr, neg

number for rel. endyr

1 # Benchmarks: 0=skip; 1=calc F_spr,F_btgt,F_msy

1 # MSY: 1= set to F(SPR); 2=calc F(MSY); 3=set to F(Btgt); 4=set to F(endyr)

0.4 # SPR target (e.g. 0.40)

0.4 # Biomass target (e.g. 0.40)

#_Bmark_years: beg_bio, end_bio, beg_selex, end_selex, beg_relF, end_relF (enter actual year, or

values of 0 or -integer to be rel. endyr)

0 0 0 0 0 0

# 2010 2010 2010 2010 2010 2010 # after processing

1 #Bmark_relF_Basis: 1 = use year range; 2 = set relF same as forecast below

#

0 # Forecast: 0=none; 1=F(SPR); 2=F(MSY) 3=F(Btgt); 4=Ave F (uses first-last relF yrs); 5=input

annual F scalar

3 # N forecast years

1 # F scalar (only used for Do_Forecast==5)

#_Fcast_years: beg_selex, end_selex, beg_relF, end_relF (enter actual year, or values of 0 or -

integer to be rel. endyr)

0 0 0 0

# 1180696575 1667592815 7631713 0 # after processing

1 # Control rule method (1=catch=f(SSB) west coast; 2=F=f(SSB) )

0.4 # Control rule Biomass level for constant F (as frac of Bzero, e.g. 0.40); (Must be > the no F

level below)

0.1 # Control rule Biomass level for no F (as frac of Bzero, e.g. 0.10)

0.75 # Control rule target as fraction of Flimit (e.g. 0.75)

3 #_N forecast loops (1=OFL only; 2=ABC; 3=get F from forecast ABC catch with allocations ap-

plied)

3 #_First forecast loop with stochastic recruitment

0 #_Forecast loop control #3 (reserved for future bells&whistles)



2011 #FirstYear for caps and allocations (should be after years with fixed inputs)

0 # stddev of log(realized catch/target catch) in forecast (set value>0.0 to cause active impl_error)


0 # Do West Coast gfish rebuilder output (0/1)

-1 # Rebuilder: first year catch could have been set to zero (Ydecl)(-1 to set to 1999)

-1 # Rebuilder: year for current age structure (Yinit) (-1 to set to endyear+1)

1 # fleet relative F: 1=use first-last alloc year; 2=read seas(row) x fleet(col) below

# Note that fleet allocation is used directly as average F if Do_Forecast=4

0 # basis for fcast catch tuning and for fcast catch caps and allocation (2=deadbio; 3=retainbio;

5=deadnum; 6=retainnum)

# Conditional input if relative F choice = 2

# Fleet relative F: rows are seasons, columns are fleets

#_Fleet:

# 0 0 0 0

# 0 0 0 0

# 0 0 0 0

# 0 0 0 0

# max totalcatch by fleet (-1 to have no max) must enter value for each fleet

-1

# max totalcatch by area (-1 to have no max); must enter value for each fleet

-1

# fleet assignment to allocation group (enter group ID# for each fleet, 0 for not included in an

alloc group)

0

#_Conditional on >1 allocation group

# allocation fraction for each of: 0 allocation groups

# no allocation groups

0 # Number of forecast catch levels to input (else calc catch from forecast F)

2 # basis for input Fcast catch: 2=dead catch; 3=retained catch; 99=input Hrate(F) (units are from

fleetunits; note new codes in SSV3.20)

# Input fixed catch values

#Year Seas Fleet Catch(or_F)

#

999 # verify end of input


Annex 2: IBPBass Working documents

Working document 01: IBPBass 2014

Effect on sea bass Stock Synthesis model of expanding the UK fishery age com-positions to a larger plus group

Mike Armstrong and Lisa Readdy, Cefas

Introduction

The previous benchmark assessment of sea bass in ICES Areas IV and VII (IBPNEW: ICES 2012), and the subsequent WGCSE update in 2013 (ICES 2013), incorporated UK fishery age compositions with a plus-group of age 12. This was because historical UK data processing generated output data files with a hard-wired plus group preventing any subsequent extension without repeating all the extractions and associated fleet-raising. IBPNEW noted that strong year classes could be tracked in the raw unpro-cessed age data well beyond age 11, and WGCSE 2013 recommended expanding the age compositions to all true ages (to allow other plus-group configurations to be eval-uated). A particular interest in expanding the age compositions was to see if early re-cruit deviations prior to the “data rich” period could be better estimated.

During 2014, the UK (England and Wales) age compositions for landings by bottom trawls, midwater pair trawls, fixed and driftnets, and lines, were re-computed to the oldest true age recorded. This Working Document examines the year-class signals pre-sent beyond age 11, and re-runs the WGCSE 2013 base-case Stock Synthesis model keeping all inputs and settings constant except for expansion of the UK age composi-tions to plus-groups of 16+, 18+ and 20+. Key outputs (likelihoods, SSB and recruitment estimates and their standard errors) and model fits to surveys and composition data are examined.

Results

Empirical information in the catch-at-age matrix

A visual expression of the year-class signals is given by bubble plots (Figure 1). Strong year classes are discernible well beyond the oldest true age (11) in the previous assess-ments. The 1976 year class, recorded first at-age 9 in 1985, can be tracked out to over 20 years of age. Earlier year classes, however, are not clearly evident. This is partly due to the smaller sample sizes for some fleets in the earlier years.

Catch curves for year classes 1970 to 1999 for the combined UK fleet data indicate a generally poor information content at-ages 20 and over, except for some year classes (Figure 2). For zero catches which cannot be logged, a value of 1 fish was added (Ln(1) = 0). This was only done for embedded zeroes; zero catches beyond the oldest non-zero catch in each cohort were ignored (this will lead to some bias in slope estimates). Track-ing is reasonable out to age 19 in the combined data. The tracking will be less evident in disaggregated fleet data where there are smaller sample sizes. The slopes of the catch curves indicate a period of increase during the late 1980s (Figure 3: top). Given that the average age in the landings is around 6–7 years, the increasing catch-curve slopes is consistent with the initial period of increasing landings (Figure 3: bottom). However, the trends in catch curve slope will also be affected by changes in contribution of dif-ferent gear types, due to their different selectivity. These changes are accounted for in Stock Synthesis where fleets are disaggregated.


Performance in Stock Synthesis

The WGCSE 2013 base case Stock Synthesis model was re-run replacing the UK age compositions (12+ gp) with the same data expanded out to 16+, 18+ and 20+ years. The WGCSE removed the age composition data for UK midwater trawls due to unusual selectivity parameter estimates, and fitted this fleet using the SS3 length model. Also, WGCSE 2013 did not have UK age compositions for 2012 and included length data for that year instead. To allow a direct comparison with the 2013 base case, the UK mid-water trawl LFDs (1996–2012) and the LFDs for bottom trawls, nets and lines for 2012 were retained, along with the French combined-fleet LFDs as used in the previous as-sessment. Hence the SS3 runs show what the WGCSE 2013 results would have been for different plus group settings.

The different plus-group settings had very little impact on time-series of SSB and re-cruitment (Figure 4), or the estimate of virgin and initial SSB. Estimates of SSB for a ten year period in the middle of the series were very slightly higher for the expanded age compositions compared with the WGCSE run using a 12+ group. Adding more true age groups resulted in an increase in the age composition likelihood and hence total likelihood (Table 1) due to the additional data components. However the CVs of the SSB and recruitment estimates over the full time-series were hardly changed (Figure 5).

Table 1. Likelihoods (total and components) for re-runs of the WGCSE 2013 base case Stock Syn-thesis run (base 12+gp) using expanded UK fleet age compositions (20+, 18+ and 16+gps) with all other inputs the same as in the WGCSE final run.

Inspection of Stock Synthesis diagnostics revealed very little difference for the four plus-group settings. The most obvious difference was in the 1970–1980 recruitment de-viations which are better estimated using an expanded age composition (a comparison between 12+gp and 18+gp is given in Figure 6.


Conclusions

The exploratory SS3 runs show that the different choices of plus-group have relatively little impact on the results, other than (as hoped) a slightly better estimation of early recruit deviations. Expanding the age compositions may help fit domed selection curves for fleets where this is appropriate, but risks an increasing number of zero catch entries for older ages as recent weak year classes feed into future catches and become depleted. A plus-gp of 16+ is recommended for further model development.

Figure 1. Bubble plots of UK bass age compositions (all fleets combined).


Figure 2. Catch curves for five year old and older sea bass; all UK fleets combined. Embedded zero catches replaced with value of 1 (zero catches beyond oldest true age recorded in a year class are omitted). Linear regressions are fitted. Vertical lines indicate lower limits of the plus-groups 12+, 16+, 18+ and 20+ explored in Stock Synthesis.


Figure 3. Top plot: Slopes of the catch curves in Figure 2. (Note these will have some bias due to censoring of zero catches beyond oldest true age with non-zero catch). A three-year running mean line is fitted. Bottom plot: international landings of sea bass.


Figure 4. Estimates of SSB and recruitment from re-runs of the WGCSE 2013 SS3 runs using UK age compositions for trawls, nets and lines expanded to include plus groups of 16+, 18+ and 20+. Plot (a) shows the estimated virgin SSB and initial SSB (for first year with recruit deviations).


Figure 5. Comparison of the standard deviations of the SSB and recruitment estimates (divided by the point estimates to give a CV) for the virgin and initial SSB and recruitment estimates and the annual values, for different plus group settings.

Figure 6. Recruitment deviations from Stock Synthesis using a 12+gp (WGCSE 2013 base case) and 16+gp and 18+gp.

18+ plus group

12+ plus group

16+ plus group


Working document 02: IBPBass 2014

Comparison of UK and French length compositions by year, area and gear

Mike Armstrong & Mickael Drogou

This document provides a direct comparison between annual length compositions for landings by four gear groups (midwater pair trawl PTB; otter trawl OTB; nets; lines) to evaluate the possibility of adopting only the UK composition data (1985–2012) in Stock Synthesis while combining UK and French landings in each gear grouping. Fleet-raised length compositions are given for gears and areas since 2009 where sampling data were available.

a ) Midwater pair trawls

The UK and French pair trawl fleets target mainly adult bass on or near offshore spawning sites in late winter through spring. Mesh size is at least 100 mm. Compara-tive LFDs are available only for VIIeh, and are similar in composition (Figure 1). The mean lengths do not deviate systematically (Figure 5).

b ) Bottom otter trawls (OTB)

The UK OTB fleet includes many under-10 m vessels, and much of the bass fishing is close inshore where there is a high contribution of young bass. Discard rates of bass are highest in this métier, which typically uses 80 mm mesh in the Channel and Celtic Sea, and this is not reflected in the LFDs which are landings only. The UK OTB length frequencies are often narrower in range than in the French OTB (Figure 2), and the mean length is consistently lower (Figure 5) suggesting these two OTB fleet compo-nents have differing selectivity. The French fleet presents a large range of activities, and the main component is not coastal. This would probably explain the difference observed in the UK and French data. Moreover a large number of species is caught and sea bass is often not a target. Mesh size is also 80 mm in general.

The length range is narrower, and the mean length lower, in the eastern Channel (VIId) which is also an area with the highest discarding and on the UK coast is close to im-portant nursery areas in the Solent. The broader length range in the western Channel (VIIeh) may partly reflect closer proximity to important spawning grounds. Hence se-lectivity, at the scale of the full stock range, will be area dependent and may therefore vary to some extent from year to year according to any changes in distribution of fish-ing.

c ) Nets

The data on UK and French nets may not be directly comparable due to different mixes of types of nets including trammelnets (GTR). The typical mesh size for bass is 100 mm. The UK net fishery for bass occurs close inshore and is mainly carried out by under-10 m vessels. Only two comparisons were available (Figure 3) and the French LFDs were dramatically different in the two years, causing a large difference in mean length (Fig-ure 5). French nets don't catch a lot of sea bass in the studied area, and different kind of nets can be used (not only GTR). This is completely different for the Bay of Biscay where a large number of vessels target sea bass during winter with GTR (mesh size 100 mm).

d ) Lines


The data on UK and French lines may not be directly comparable because the French data are for longlines (code LLS) whereas the UK data include all forms of lines includ-ing longline, handline and commercial rod and line. As with nets, the UK line fishery is mainly by under-10 m vessels operating inshore. In four comparisons, the UK line data consistently show a lower mean length (Figures 4 and 5).

Conclusions

The tendency for the French OTB fishery to catch larger bass, on average, than in the UK OTB fishery, may reflect a fishery that is less artisanal and operating over a broader depth range in the Channel and Celtic Sea. Selectivity, taken at the stock level and not locally, also appears to vary according to the location of fishing around the coast. It is probably therefore an incorrect assumption to treat the French OTB fleet as having the same selectivity parameters as the UK fleet, particularly where a domed selection pat-tern is evident in the UK fishery.

Nets are also expected to have domed selectivity due to the nature of the gear. Unfor-tunately there are insufficient comparisons to draw any conclusions on the relative se-lectivity of French and UK nets.

A direct comparison for lines is difficult because the French data appear to only repre-sent longlines. However, the mean length of bass in the UK samples is consistently lower than in the French samples, again suggesting some difference in selectivity.

The UK and French midwater pair trawl fishery are part of the same offshore fishery on mature bass, although not necessarily overlapping completely in spatial coverage. The available comparisons do not suggest a consistent difference in size composition between the UK and French fleets.


Figure 1. LFDs for midwater pair trawls (PTB), France and UK, by area and year since 2010.


a ) Otter trawls OTB

Figure 2. LFDs for otter trawls (OTB), France and UK, by area and year since 2009.


Figure 3. LFDs for nets (trammelnets GTR for France; all fixed and driftnets for the UK); by year since 2009. Comparative data only for VIIeh, in two years.


b ) Lines (longline and handline for UK; longline only for France)

Figure 4. LFDs for lines (longlines LLS for France; handlines, longlines and rod and line for the UK); by year since 2009. Comparative data only for VIIeh.


Figure 5. Mean lengths of bass by country, year, gear and area.


Annex 3.1: Summary of likelihood values from SS3 runs presented

Table 1. Likelihoods for SS3 development runs.

Run Code: 1 2 3 4 5 6 7 11b 11c 15b

Lambda2013 WGCSE final age/length model WGCSE with 16+ gp WGCSE with 18+ gp

WGCSE with 20+ gp

Run 1 with 2012 UK AFDs added

Run 4 with UK MWT AFDs added

Run 5 with all UK ESS adjusted to 50; France ESS as WGCSE

Run 6 with UK ESS 1985-90 reduced to 10; France ESS all 200

Run 7 with CGFS data entered as length compositions with length-based double-normal selectivity

As run 11b, but with UK fleets selectivity fitted as age-based, not length based (all fleets asymptotic)

Three UK and French combined metiers (trawls& nets, midwater, lines) with age based selectivity for the 3 combined fleets. "Other" selectivity = lines . Trawls_nets double normal.

TOTAL 774.567 811.25 818.368 823.165 808.066 674.572 1027.72 796.953 931.059 887.132 450.25Component logL*LambdaCatch 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.06E-12Equil_catch 1 0.0476351 0.126527 0.213076 0.309567 0.146883 0.203039 0.879775 0.317822 0.315408 1.25E-07 0.0463133Survey 206.102 207.912 208.275 208.328 202.9 203.078 214.992 210.854 216.963 0.170375 56.9506Length_comp 287.015 285.96 285.964 285.961 273.97 132.55 134.04 134.835 264.654 220.462 123.974Age_comp 258.496 293.838 300.404 305.055 306.387 313.039 649.564 423.687 421.325 263.499 239.662Recruitment 1 22.8933 23.4008 23.499 23.4978 24.6461 25.6865 28.2312 27.2427 27.5518 376.15 29.1022Forecast_Recruitment 1 0 0 0 0 0 0 0 0 0 26.6112 0Parm_priors 1 0.00583555 0.00578508 0.00578027 0.00577641 0.0058279 0.00638661 0.00646085 0.00600792 0.234847 0 0.50145Parm_softbounds NA 0.00770892 0.00738236 0.00737756 0.00737376 0.00906508 0.0100461 0.00877121 0.00992588 0.0145824 0.228055 0.0137316Parm_devs 1 0 0 0 0 0 0 0 0 0 0.0113211 0Crash_Pen 1 0 0 0 0 0 0 0 0 0 0 0

0Convergence level 6.99E-05 1.919E-05 4.662E-06 2.255E-04 2.376E-05 7.222E-04 9.244E-05 5.156E-04 6.77E-05 3.27E-05 3.166E-05Number of parameters 63 63 63 63 63 63 63 63 69 69 69


Table 2. Likelihoods for SS3 development runs.

Run code: 16a 16b 16c 17a 17b 18 18b 18c 19a 19b 19c

Lambda

Four fleets (Combined UK OTB,nets&lines; UK MWT; combined French fleet with 2000-2012 LFDs; "other"; age-based double-normal selectivity on UK OTB/nets/lines fleet; "other" selectivity = France (asymptotic)

16a with asymptotic age-based selectivity on UK fleet.

16b with domed selectivity of French combined fleet

Run 16a allowing SS3 to estimate VBGF parameter K

Run 16a allowing SS3 to estimate VBGF parameter K and Length at Amin

Run 16a with Effective Sample Sizes on UK OTB/nets/lines increased to 100 (20 for 1985-1990)

Run 18 with selectivity priors replaced with soft bounds

Run 18 with all soft bounds replaced by priors

Run 18 with all CGFS selectivity parameters fixed at run 18 estimates.

Run 19a with UK OTB/nets/lines selectivity parameter fixed for right-hand limb at run 18 estimate.

Run 18 with UK nets and lines landings and age compositions increased by factor of 3.

TOTAL 471.452 474.89 469.576 460.246 456.089 536.723 535.87 536.838 536.72 536.821 537.069Component logL*LambdaCatch 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.25E-07 5.32E-10Equil_catch 1 0.0461579 0.0739659 0.0561979 0.0555754 0.0587913 0.063079 0.0633271 0.0630827 0.0630484 0.0668912 0.0559847Survey 57.5951 56.9704 56.5334 49.4981 45.7962 64.1681 64.1645 64.168 64.2036 64.101 65.4597Length_comp 259.107 258.334 257.732 253.699 253.477 261.142 261.088 261.141 261.107 260.994 261.105Age_comp 126.056 130.752 126.481 128.454 128.011 182.105 181.969 182.106 182.107 182.384 180.719Recruitment 1 28.0338 28.526 28.1312 28.017 28.2656 28.5479 28.5522 28.5478 28.5458 28.5928 29.0172Forecast_Recruitment 1 0 0 0 0 0 0 0 0 0 0 0Parm_priors 1 0.598811 0.224429 0.628096 0.509699 0.468845 0.681463 0 0.796294 0.681804 0.67088 0.69586Parm_softbounds NA 0.0144647 0.00809223 0.0143249 0.0127851 0.0119447 0.0161337 0.0322361 0.0161334 0.0117905 0.0115452 0.0164318Parm_devs 1 0 0 0 0 0 0 0 0 0 0 0Crash_Pen 1 0 0 0 0 0 0 0 0 0 0 0

Convergence level 7.255E-06 5.902E-05 1.005E-04 5.152E-05 2.079E-05 5.678E-05 1.01E-06 1.69E-04 1.648E-04 1.464E-04 1.083E-04Number of parameters 69 65 73 70 71 69 69 69 63 62 69


Table 3. Likelihoods for runs examining M and stock–recruit steepness.

Run code: 20a 20b 20c 20d 20e 20f

LambdaRun 18 with M=0.15 and Steepness = 0.8

Run 18 with M=0.15 and Steepness = 0.9





TOTAL 548.254 545.521 545.515 538.46 537.448 536.723Component logL*LambdaCatch 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.25E-07Equil_catch 1 0.535754 0.551564 0.542918 0.0673456 0.0657106 0.063079Survey 65.1417 65.0196 64.5582 64.2145 64.1869 64.1681Length_comp 260.943 260.876 260.828 261.177 261.156 261.142Age_comp 183.796 183.429 185.794 182.212 182.148 182.105Recruitment 1 37.1692 34.97 33.3185 30.0945 29.1945 28.5478Forecast_Recruitment 1 0 0 0 0 0 0Parm_priors 1 0.653976 0.658788 0.460701 0.679362 0.680631 0.681463Parm_softbounds NA 0.0155856 0.0156792 0.0117109 0.0160938 0.0161179 0.0161337Parm_devs 1 0 0 0 0 0 0Crash_Pen 1 0 0 0 0 0 0

Convergence level 2.168E-04 9.196E-05 1.068E-04 7.099E-05 4.492E-06 9.339E-05Number of parameters 69 69 69 69 69 69

Run code: 20g 20h 20i 20j 20k 20l

LambdaRun 18 with M=0.25 and Steepness = 0.8






TOTAL 534.986 534.582 534.3 541.085 534.019 533.904Component logL*LambdaCatch 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.25E-07Equil_catch 1 0.00581474 0.00514635 0.0045178 0.000142453 1.71E-05 4.87E-05Survey 63.62 63.6161 63.6141 62.3351 63.0549 63.0576Length_comp 261.61 261.6 261.591 261.62 262.174 262.167Age_comp 181.887 181.876 181.869 190.714 182.142 182.14Recruitment 1 27.1519 26.7724 26.5081 26.0001 25.9222 25.8135Forecast_Recruitment 1 0 0 0 0 0 0Parm_priors 1 0.695438 0.695713 0.69587 0.405555 0.709273 0.709212Parm_softbounds NA 0.0164182 0.016423 0.0164256 0.0105771 0.016699 0.0166975Parm_devs 1 0 0 0 0 0 0Crash_Pen 1 0 0 0 0 0 0

Convergence level 1.672E-05 1.814E-04 3.230E-06 6.804E-05 3.717E-05 1.960E-05Number of parameters 69 69 69 69 69 69


Table 4. Likelihoods for final Run 22 and some sensitivity analyses.

Run code: 22 23a 23b 23c 23d 23e 23f

Lambda

Final run with recreational F vector; M=0.15 (based on Run 18b)

Run 22 with F multiplier 0.04 on recreational F (rec F = 0.037)





Run 22 with F multiplier 0 on recreational F (rec F = 0)

TOTAL 534.766 537.108 537.766 535.04 534.265 534.016 542.668Component logL*LambdaCatch 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.25E-07 1.25E-07Equil_catch 1 0.0192335 0.102338 0.0385686 0.0260397 0.00896997 0.00479664 0.554813Survey 64.8834 64.7879 64.4916 64.8618 64.9227 64.9652 64.9322Length_comp 261.071 260.866 261.08 261.032 261.149 261.203 260.765Age_comp 181.455 182.027 184.128 181.53 181.337 181.273 183.056Recruitment 1 27.3061 29.2941 28.0073 27.5591 26.8159 26.5387 33.3291Forecast_Recruitment 1 0 0 0 0 0 0 0Parm_priors 1 0 0 0 0 0 0 0Parm_softbounds NA 0.0312498 0.031123 0.0205526 0.0312083 0.0312474 0.0312481 0.0311721Parm_devs 1 0 0 0 0 0 0 0Crash_Pen 1 0 0 0 0 0 0 0Convergence level 8.36E-05 0.00032627 0.000171466 8.41E-05 0.000201841 0.000125392 1.36E-05Number of parameters 69 68 68 68 68 68 68

Run code: 24a 24b 24c 24d 25

Lambda

Run 22 with M=0.20 and F multiplier 0.04 on recreational F (rec F = 0.037)



Run 22 with M=0.20 and F multiplier 0 on recreational F (rec F = 0)

Final run 22 with recreational F vector; M=0.15; UK nets and lines landings increased by factor of 3

TOTAL 533.921 533.51 533.354 535.876 536.117Component logL*LambdaCatch 1.25E-07 1.25E-07 1.25E-07 1.25E-07 5.33E-10Equil_catch 1 0.00869536 0.00220128 0.000215906 0.0642675 0.0327335Survey 64.2507 64.3149 64.3769 64.14 66.1709Length_comp 261.296 261.423 261.549 261.06 261.296Age_comp 181.536 181.434 181.402 182.016 180.28Recruitment 1 26.7975 26.3039 25.9931 28.5636 28.3044Forecast_Recruitment 1 0 0 0 0 0Parm_priors 1 0 0 0 0 0Parm_softbounds NA 0.0322547 0.0322878 0.0323051 0.0321384 0.0327771Parm_devs 1 0 0 0 0 0Crash_Pen 1 0 0 0 0 0Convergence level 5.09E-05 2.24E-05 4.35E-05 3.55E-05 9.48E-05Number of parameters 68 68 68 68 68


Annex 3.2: Summary of SSB plots from SS3 runs presented

Figure 1. Runs 1–5. Result from WGCSE 2013 assessment shown top left. Horizontal lines have been manually added at approximately 4000 t, and y-axes stretched or contracted where needed, to help visual comparison of trends.


Figure 2. Runs 6–15b. Horizontal lines have been manually added at approximately 4000 t, and y-axes stretched or contracted where needed, to help visual comparison of trends.


Figure 3. Runs 16a–18a. Horizontal lines have been manually added at approximately 4000 t, and y-axes stretched or contracted where needed, to help visual comparison of trends.


Figure 4. Runs 18b–19c. Horizontal lines have been manually added at approximately 4000 t, and y-axes stretched or contracted where needed, to help visual comparison of trends.


Figure 5. Runs 20 a–l. Horizontal lines have been manually added at approximately 4000 t, and y-axes stretched or contracted where needed, to help visual comparison of trends.


Figure 6. Runs 22–23e (Final run and sensitivities). Horizontal lines have been manually added at approximately 4000 t, and y-axes stretched or contracted where needed, to help visual comparison of trends.


Figure 7. Runs 24a–25 (final run sensitivities). Horizontal lines have been manually added at ap-proximately 4000 t, and y-axes stretched or contracted where needed, to help visual comparison of trends.


Annex 4: Participants list

NAME ADDRESS PHONE/FAX E-MAIL Mike Armstrong

Centre for Environment, Fisheries and Aquaculture Science (Cefas) Lowestoft Laboratory Pakefield Road NR33 0HT Lowestoft Suffolk UK

Phone +44 1502 524362 Fax +44 1502 524511

[email protected]

Mickael Drogou

Ifremer Centre de Brest PO Box 70 29280 Plouzané France

Phone +33 2 98 22 43 74 Fax +33

[email protected]

Chris Legault Invited Expert

National Marine Fisheries Services Northeast Fisheries Science Center Woods Hole Laboratory 166 Water Street Woods Hole MA 02543 United States

Phone +1 508 4952025 Fax +1 508 4952393

[email protected]

Cristina Morgado ICES Secretariat

International Council for the Exploration of the Sea H. C. Andersens Blvd. 44–46 1553 Copenhagen V Denmark

Phone +45 33 38 67 21 Fax +45 33 63 42 15

[email protected]

Jan Jaap Poos Chair

Wageningen IMARES PO Box 68 1970 AB Ĳmuiden Netherlands

Phone +31 317 487 189 mob private + 31 6 22 79 44 89 Fax IMARES general +31 317 480 900

[email protected]

Lisa Readdy Centre for Environment, Fisheries and Aquaculture Science (Cefas) Lowestoft Laboratory

Pakefield Road

NR33 0HT Lowestoft

Suffolk

UK

Phone +44 1502 52 4319

Fax +44 1502 52 4511

[email protected]