22 Turbot in Subarea 4

Transcription

1 912 ICES WGNSSK REPORT Turbot in Subarea 4 This report presents the stock assessment carried out for turbot (Scophthalmus maxima) in Subarea 4 in Following an inter-benchmark procedure for this stock in 2015 and another one in 2017, a new assessment model (SAM) was used since More details on the data used, assumptions made and the assessment model settings can be found in the Stock Annex and inter-benchmark report. The assessment output (SSB) was used as the basis for deriving advice under category 3 of the ICES DLS approach in 2017, as the assessment showed extreme recurring retrospective bias in F and SSB estimates. At WGNSSK 2017, some serious issues were found in the settings of the assessment that were causing the retrospective patterns. The assessment for this stock was therefore delayed, pending further work to be carried out during the inter-benchmark process that took place during the summer of During the inter-benchmark an evaluation was made of current input data, new catchat-age data was supplied, new catch and survey data was evaluated in terms of its performance in and suitability for the assessment, the SAM model settings were parameterised, a decision was made on the category of the assessment, and finally new reference points were estimated General Biology and ecosystem aspects Turbot is broadly distributed from Iceland in the North, along the European coastline, to the Mediterranean and Adriatic Sea in the south. In general, turbot is a rather sedentary species, but there are some indications of migratory patterns. For example in the North Sea, migrations from the nursery grounds in the south-eastern part to more northerly areas have been recorded. IBPNEW (ICES, 2012) concluded that Turbot in the North Sea (Subarea 4) can be considered as a distinct stock for management purposes. Turbot is typically found at a depth range of 10 to 70 m, on sandy, rocky or mixed bottoms and is one of the few marine fish species that inhabits brackish waters. It is a typical visual feeder and could be regarded as a top predator. Turbot feeds mainly on bottom living fishes (e.g. common gadoids, sandeels, gobies, sole, dab, dragonets, sea breams etc.) and small pelagic fish (e.g. herring, sprat, boarfish, sardine) but also, to a lesser extent, on larger crustaceans and bivalves. Despite its role as a top predator in the North Sea ecosystem, at present turbot is not included as a species in the WGSAM multispecies assessment (ICES, 2014a) Fisheries In the 1950s the UK was the biggest contributor to the landings (~50% of the landings). In recent years most of the landings stem from the Netherlands (~50-60%). In most countries turbot is caught in mixed fisheries trawls, with most of the landings in the Netherlands coming from the 80 mm beam trawl fleet (BT2) fishing for sole and plaice. In Denmark, the second largest contributor to the landings in recent times, there is a directed fishery for turbot using gillnets (~10% of the total landings). See the Stock Annex for more details.

2 ICES WGNSSK REPORT Management A combined EU TAC for turbot and brill is set for EU waters in areas 2a and 4. This TAC only applies to the EU fisheries. This management area (particularly the inclusion of Area IIa) does not correspond to either of the stock areas defined by ICES for turbot and brill. No specific management objectives or plans are known to ICES. As a primarily bycatch species, regulations relating to effort restrictions for the primary métiers catching turbot (e.g. beam trawlers) are likely to impact on the stock. Fishing effort has been restricted for demersal fleets in a number of EC regulations (e.g. EC Council Regulation Nos. 2056/2001, 51/2006, 41/2007, and 40/2008). The Dutch Producer Organisations have introduced a minimum landings size of 27 cm in In 2016 this size was increased to 30 cm first, and then to 32 cm. In the summer of 2016, the POs decided to prohibit landing the smallest market category and in October the weekly landings were capped to 375 kg wk -1. These measures were taken in order to try and keep the landings within the national quota. In 2017, these measures were continued. See the Stock Annex for more details Data used Following the inter-benchmark conducted in the summer of 2017, the assessment of North Sea turbot requires three main types of data: Catch data: estimates of removals of turbot by the fishery. Survey data and commercial LPUE (landings per unit effort): indices of trends in population abundance over time from fisheries independent and fisheries dependent sources, respectively. Biological data: estimates and/or assumptions on growth, maturation and natural mortality. Since the assessment is age-based, data for the above is required for each age. See the Stock Annex for more details on the data used in the assessment, sources and historical values Catch data InterCatch was used for the first time for the North Sea turbot stock at WGNSSK 2014, and has been used since. Dutch samples (for data from 2004 present) and Danish samples (from 2014 present) accounting for auctions, quarters and market categories, are used for estimating the age structure of the landings. Prior to that, the landings-at-age information is from an old Dutch monitoring scheme from the 1980s. Figure shows the métiers with numbers at age samples for the landings in The only usable samples were those from the Dutch TBB_DEF_70-99 (beamtrawl) métier, and these samples were used to raise all the unsampled métiers. Raising was done by quarter. The TBB_DEF_70-99 samples were evenly spread over the seasons. All beam trawl fleets were raised using the age distributions calculated from the TBB samples. There are a wide variety of métiers that land turbot, including a significant amount from the Danish > 220mm gillnet fleet.

3 914 ICES WGNSSK REPORT 2017 Figure shows the trend in total landings over time. Landing of turbot decreased during the 1990s and for the last ten years have been stable in the region of 3000t. Over this time effort by the Dutch beam trawl fleet, which contributes the most of the landings, has decreased notably. Since turbot is primarily a bycatch species, this indicates that abundance of turbot has likely increased over this period. Landings at age are presented in Table and Figure , with weights-at-age in the catch presented in Table The 2005 year class shows up clearly in the landings data. In 2016, there appear to be large landings of age 3, suggesting a strong 2013 year class. Following a decrease in minimum market size for turbot in the Netherlands in 2002, there has been a notable increase in the amount of age 1 and 2 turbot landed, accounting for half of the catch in some years but this proportion has been decreasing in recent years due to some poor year classes in 2012 and Since turbot are only fully mature at age 4, a high proportion of immature fish are the landings. However, the last 5 years have also seen an increase in the proportion of age 5+ fish in the landings compared to the five years prior to that, these are now in the same order of magnitude as the estimates in the 1980s. This could reflect a reduction in F recently leading to an increasing proportion of older fish in the landings. However, since the catch data is raised using only the Dutch 80mm TBB fleet, signals in catch at age data may not be accurate reflections of true removals from the population over time Discard data The assessment of this stock assumes that discarding of catches for this stock is negligible. However, there was a sudden increase in the landing of age two turbot following the decrease in minimum market size in the Netherlands in Given that there was no known change in the fishing behaviour of the main fleets at this time, this could indicates that previously more age 1 fish must have been caught than were actually landed. These were either discarded or, as a much sought after fish, kept by the fishermen for personal use. This would mean that the discards could be underestimated in the period up to 2002 relative to the period following this, potentially causing a bias in the assessment outputs. Alternatively, subsequent to the change in MLS, more targeting of small turbot may have occurred. Without a useable time series of discards before and after this change it is difficult to determine which of these explanations holds. However, the impact on the final year estimates is likely to be small because with the reduction in minimum market size in 2002, the assumption of negligible discards probably holds for the last 10 years. Discard data were submitted to InterCatch by various nations. However, there is very limited age sampling of the discards. Very few fish were sampled in the discards of some of the Danish métiers (<10 per métier, fewer than the number of ages in the assessment model), not enough to be used in the raising of international landings. In 2017, most countries provided estimates of discards in 2015 to InterCatch. Out of the 665 tonnes of estimated discards, 566 (85%) was reported data and 99,5 tonnes is raised in InterCatch. The proportion of landings with discards associated (same strata) is 22 percent. The discard rate (discards: / (discards + landings: ) was 16% in This is a substantial increase compared to the most recent period ( ), when discard ratios were approximately 5%. No useable age structure information was submitted for the discard estimates.

4 ICES WGNSSK REPORT Survey data and commercial LPUE Two survey abundance indices, the Sole Net Survey (SNS) and the Beam Trawl Survey (BTS ISIS), and one standardised commercial LPUE abundance index, the Dutch 80 mm beam trawl fleet (BT2), are used to tune the assessment (Table and Figure ). All abundance indices indicate an increase in the number of fish aged 4 and older in late 2000s compared to the past. An increase in the amount of older fish would indicate either strong recruitment or a decrease in mortality (e.g. fishing pressure) exerted on the stock. After a decrease in some of the older ages and no clear indications of strong year classes since 2010, year classes 2013 and 2014 (ages 2 and 3 in 2016) appear strong. There is fairly close agreement between the three indices on the general trends in abundance at age, but the data are noisy from year to year. This can be seen in the low R 2 values in the internal consistency correlations in the BTS_ISIS and SNS surveys (Figure ). The SNS survey is particularly poor at picking up cohort signals, with low R 2 values on the correlations between numbers at consecutive ages. Though all correlations between successive ages are positive, estimated numbers at age, particularly for the younger ages, fluctuate a lot from year to year. The BTS-ISIS is more internally consistent for ages 3 and up. The almost non-existent relationship between the numbers estimated at age 1 and the numbers estimated at age 2 in the following year suggest that in future removing age 1 from this index may be appropriate. The internal consistency of the NL_BT2 LPUE index is significantly better, though the removal of age 1 from this index could also be considered. However, this index is no longer used as an age-structured index. Noisy indices that are more indicative of general trends are best used in an assessment model that is able to smooth over the noise in the data. The SAM model used for this stock is able do this, but nevertheless inputting noisy data into the assessment will increase uncertainty in the outputs. By removing the age-structure from the NL BT2 LPUE index, the clearest cohort signals in the assessment of this stock are coming from the catch at age matrix Biological data All biological data used in the assessment are presented in Table Weight at age Constant annual catch and stock weights at age (long term means of all available data) were previously used in the assessment because of large gaps in the time series of weight at age data for turbot in the North Sea (Figure ). What data is available is also very noisy, due to low sample sizes for most ages. The data that are available, and trends in other flatfish species in the same areas suggest that there have been potentially significant changes in weight at age over time. At IBPturbot a method was developed to model the growth parameters over time, allowing smooth changes over the time series (see Stock Annex for full details). The results indicate an increase in weight at age from the start of the time series, peaking in the early 1990s. Since then weights at age have decreased again to slightly lower than the 1970s. Maturity At IBPNEW (ICES, 2012) turbot maturity data from the Netherlands was used to study some reproductive characteristics of turbot from the North Sea. A female maturity ogive constructed from derived from a General Linear Model fit using the maturity data from the recent time period was chosen for the stock.

5 916 ICES WGNSSK REPORT 2017 Natural mortality There are currently no accepted estimates of turbot natural mortality over time. A number of alternative methods, using different estimates of growth parameters, were used to estimate the level of natural mortality by age for turbot in the North Sea at IBPNEW (ICES, 2012). Since turbot grows relatively fast compared to other flatfish species in the same areas, results indicate that natural mortality is higher. However, due to high variability for recorded values of K (an estimated growth parameter) for turbot, it proved difficult to find agreement on natural mortality values. Hence, after performing assessment test runs, a constant value of M = 0.2 for all ages and years was chosen for this stock. This is twice the level used in the sole and plaice assessments in the North Sea Stock assessment model After the inter-benchmark protocol for this stock in the summer of 2017, a new assessment model (SAM) is used. More details on the data used, assumptions made and the assessment model settings can be found in the Stock Annex and in the inter-benchmark protocol report. In summary, three abundance indices are used (Table ): - The SNS survey, from age 1 6 without plus group over the period , - The BTS-ISIS survey, from age 1-7 with plus group over the period , and, - A standardised age-aggregated commercial LPUE of Dutch BT2 over the period As mentioned above, Danish age information for the period was supplied before the inter-benchmark protocol. This was evaluated and added to the catch-at-age matrix. The previous age data sources were revaluated and finally only the Dutch data sources were accepted. This means that the start of the catch-at-age matrix is now reduced to Catches are fitted to catch-at-age data from the Netherlands ( , 1998, ) and Denmark ( ). Finally, different SAM model configuration were tested and a final set was used. Natural mortality is fixed at 0.2 and the surveys present a maximal age plus group at age 7 (BTS-ISIS, no plus group on SNS) and catch at age 8. While the diagnostics of the new base assessment looked satisfactory, the retrospective pattern in F is still large so the inter-benchmark protocol agreed to keep considering the assessment as a category 3. The external reviewers agree that results generated by the assessment model can be used for providing fisheries advice as an ICES Category 3 stock although the stock has a quantitative assessment that could qualify as category 1. The assessment output (SSB) is used as the basis for deriving advice in 2017 under category 3 of the ICES DLS approach (2 over 3 rule) Model settings The assessment model was conducted using the settings and configuration given below. Details of the assessment model can be found in the Stock Annex and Inter-benchmark report (see also tables ).

6 ICES WGNSSK REPORT Assessment settings used in the final assessment Year 2017 (IBPTurbot proposal) Model SAM First tuning year 1981 Last data year 2016 Ages 1-8+ Plus group Stock weights at age Catch weights at age Total Landings Landings at age Yes Von Bertalanffy growth curve with time varying Linf Von Bertalanffy growth curve with time varying Linf Not used , 1998, 2000 present Discards Not used (assumed 0) Abundance indices BTS-Isis Catchability in catch at age matrix independent of age for ages >= coupling of fishing mortality STATES (Row represent Catch, olumns represent ages) Use correlated random walks for the fishing mortalities (0 = independent, 1= correlation estimated) Coupling of catchability PARAMETERS (Surveys)) Row represent fleets (SNS and BTS-only, lpue ageaggregated), Columns represent ages) Coupling of fishing mortality RW VARIANCES SNS Standardized NL-BT2 LPUE age-aggregated catchable biomass Coupling of log N RW VARIANCES Coupling of OBSERVATION VARIANCES (Row represent fleets (Catch, SNS, BTS, lpue age-aggregated), Columns represent ages) Stock-recruitment model code (0=RW, 1=Ricker, 2=BH) 0 Fbar ranges Assessment model results The stock summary is given in Table , while fishing mortality at age and abundance at age estimated by the assessment model are presented in tables and

7 918 ICES WGNSSK REPORT , respectively. Other key model outputs are given in tables and plotted in Figure Status of the stock Fishing mortality was estimated at 0.36 in 2016, an increase from 2015 (0.30). This is well below the long term geometric mean (0.45). The SSB in 2016 was estimated to be 11789t, a small increase from 2015 which was estimated at 10868t. Both years are higher than the long term geometric mean (8078 t). The estimated recruitment (age 1) for 2016 is just below the geometric mean of the time series (4950). However, this estimate is based on very little data and is unlikely to be a reliable estimate Historic stock trends SSB was at its highest in the early 1980s (possibly higher before that time for which no reliable data is available). From the mid 1980s up until the early 2000s SSB declined gradually and F increased gradually (Figure ). The lowest observed SSB was in 1999, SSB subsequently increased and has continued to increase since. Recruitment has been variable over the time-series without a clear trend. Recent recruitment (2014 and 2015) have been well above long term mean and do now contribute to the increase in SSB. Mean F peaked in 1994 at 0.80, but then declined gradually to ~0.3 in before gradually increasing again. These trends correspond well with the trends in fishing effort of the beam trawl fleet. There are no clear patterns in recruitment, though values are estimated at a slightly higher level, but with more uncertainty, during the years of missing landings at age data (1990s). Recent recruitment has been at or above average Retrospective assessments The results of five retrospective assessments, run using the same model settings but removing one year of data from the end of the time series, are plotted in figures In most years F has been severely overestimated. There is a smaller retrospective pattern in SSB or recruitment. For F, all retrospective peels are well outside the confidence bounds of the latest assessment. The disagreement between retrospective assessments is large. This is likely due to the quality of the data used by the assessment and the model itself. No clear reason for the retrospective pattern could be found in the inter-benchmark Model diagnostics Model diagnostics are provided in tables and figures , , and Please refer to the Turbot Inter-benchmark 2017 report for more detailed specifications Management considerations There are a number of EC regulations that affect the flatfish fisheries in the North Sea, e.g. as a basis for setting the TAC, limiting effort, and minimum mesh size Effort regulations The overall fleet capacity and deployed effort of the North Sea beam trawl fleet has been substantially reduced since 1995, due to a number of reasons, including the above

8 ICES WGNSSK REPORT mentioned effort limitations for the recovery of the cod stock. In 2008, 25 vessels were decommissioned Technical measures Turbot is mainly taken by beam trawlers in a mixed fishery directed at sole and plaice in the southern and central part of the North Sea. Technical measures (EC Council Regulation 1543/2000) applicable to the mixed flatfish fishery affect the catching of turbot. The minimum mesh size of 80 mm in the beam trawl fishery selects sole at the minimum landing size (24 cm). However, this mesh size is likely to catch immature turbot (age 1 and 2 fish). Mesh enlargement would reduce the catch of smaller turbot at the same time potential increasing the yield per recruit, but would also result in loss of marketable sole catches. A closed area has been in operation since 1989 (the plaice box) and since 1995 this area has been closed in all quarters. The closed area applies to vessels using towed gears, but vessels smaller than 300 HP are exempted from the regulation. An additional technical measure concerning the fishing gear is the restriction of the aggregated beam length of beam trawlers to 24 m. In the 12 nautical mile zone and in the plaice box the maximum aggregated beam-length is 9 m Combined TAC At present the EU provides a combined TAC for turbot and brill in the North Sea. This TAC seems largely ineffective in reducing F: increases in the stock at similar TACs lead to increased discarding. In addition, it is unclear how the quantitative single species advice for turbot and the qualitative single species advice for brill can/will be used to formulate a combined TAC for these two stocks. In this situation, improving the brill assessment may be necessary in order to ensure efficient management of both of these stocks. Ideally, a combined TAC is on that is not used Proxy reference points In order to get reference point proxies a SPiCT analysis was done (Table and figures ). Additionally, since the quality of the turbot assessment is of category 1 level, an EQsim analysis is run too. See the Stock Annex and the inter-benchmark protocol report for more details. In summary, the SPiCT run estimates a proxy of fishing mortality at maximum sustainable yield (Fmsy) of (95% CI ) (Table ) with some retrospective patterns (Figure ). Both methods and runs indicate that the stock is not overfished in relation to FMSY proxies. The SPiCT run is appropriate for management, and allows for using a single consistent model. Due to the retrospective issues, the SPiCT model without the prior on LPUE sdi should be used (Figures ). Both the SPiCT and Eqsim models could still be run in the future to check for consistency in status determination and management advice References ICES, 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: pp.

9 920 ICES WGNSSK REPORT 2017 ICES. 2013a. Report of the Working Group on the Assessment of Demersal Stocks in the North Sea and Skagerrak (WGNSSK), April 2013, ICES Headquarters, Copenhagen. ICES CM 2013/ACOM: pp. ICES, 2013b. (DRAFT) ICES Implementation of Advice for Data-limited Stocks in ICES CM 2012/ACOM: pp. ICES, 2014a. Report of the Working Group on Multispecies Assessment Methods (WGSAM), October 2014, London, UK. ICES CM 2014/SSGSUE: pp. ICES. 2014b. Report of the Working Group on Mixed-Fisheries Advice for the North Sea (WGMIXFISH), May ICES CM 2014/ACOM:22. ICES. 2014c. Report of the Joint ICES MYFISH Workshop to consider the basis for FMSY ranges for all stocks (WKMSYREF3), November 2014, Charlottenlund, Denmark. ICES CM 2014/ACOM: pp. ICES Report of the Inter-Benchmark Protocol for Turbot in Subarea 4 (IBP Turbot), May 2015, By correspondence. ICES CM 2015/ACOM:XXX. Xxx pp.

10 ICES WGNSSK REPORT Table Turbot in Area 4. Catch in numbers (units: thousands). age age age age

11 922 ICES WGNSSK REPORT 2017 Table Turbot in Area 4. Weights at age in the catch (units: kg). age age age age

12 ICES WGNSSK REPORT Table Turbot in Area 4. Weights at age in the stock (units: kg) age age age age

13 924 ICES WGNSSK REPORT 2017 Table Turbot in Area 4. Natural mortality. age age Table Turbot in Area 4. Proportion mature age age

14 ICES WGNSSK REPORT Table Turbot in Area 4. Fraction of harvest before spawning age age Table Turbot in Area 4. Fraction of natural mortality before spawning. age age Table Turbot in Area 4. Survey indices SNS Configuration Turbot in IV - 22/04/2015. Imported from VPA file. min max plusgroup minyear maxyear startf endf NA Index type : number SNS - Index Values Units : NA year

15 926 ICES WGNSSK REPORT 2017 age year age BTS-ISIS - Configuration Turbot in IV - 22/04/2015. Imported from VPA file. min max plusgroup minyear maxyear startf endf Index type : number BTS-ISIS - Index Values Units : NA year age year age Dutch_BT2_LPUE_ModelD - Configuration Turbot in IV - 22/04/2015. Imported from VPA file. min max plusgroup minyear maxyear startf endf Index type : biomass Dutch_BT2_LPUE_ModelD - Index Values Units : NA year age year age

16 ICES WGNSSK REPORT Table Turbot in Area 4. Stock object configuration min max plusgroup minyear maxyear minfbar maxfbar Table Turbot in Area 4. SAM configuration settings name : Turbot in IV desc : Imported from a VPA file. ( D:/Repository/Turbot/assessment runs/lowestoft files/index_raw.txt ). Wed Sep 20 16:45: range : min max plusgroup minyear maxyear minfbar maxfbar range : fleets : catch SNS BTS-ISIS NL_LPUE fleets : plus.group : TRUE states : age states : fleet states : catch states : SNS NA NA NA NA NA NA NA NA states : BTS-ISIS NA NA NA NA NA NA NA NA states : NL_LPUE NA NA NA NA NA NA NA NA logn.vars : catchabilities : age catchabilities : fleet catchabilities : catch NA NA NA NA NA NA NA NA catchabilities : SNS NA NA catchabilities : BTS-ISIS NA catchabilities : NL_LPUE 6 NA NA NA NA NA NA NA power.law.exps : age power.law.exps : fleet power.law.exps : catch NA NA NA NA NA NA NA NA power.law.exps : SNS NA NA NA NA NA NA NA NA power.law.exps : BTS-ISIS NA NA NA NA NA NA NA NA power.law.exps : NL_LPUE NA NA NA NA NA NA NA NA f.vars : age f.vars : fleet f.vars : catch f.vars : SNS NA NA NA NA NA NA NA NA f.vars : BTS-ISIS NA NA NA NA NA NA NA NA f.vars : NL_LPUE NA NA NA NA NA NA NA NA obs.vars : age obs.vars : fleet obs.vars : catch obs.vars : SNS NA NA obs.vars : BTS-ISIS NA obs.vars : NL_LPUE 8 NA NA NA NA NA NA NA obs.weight : NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA srr : 0 scalenoyears : 0 scaleyears : NA scalepars : cor.f : 2 cor.obs : NA 0 NA NA NA 1 NA NA NA 1 NA NA NA 1 NA NA NA 1 NA NA NA NA NA NA NA NA NA NA cor.obs.flag : ID AR ID ID biomasstreat : nohess : FALSE timeout : 3600 likflag : LN LN LN LN fixvartoweight : FALSE simulate : FALSE residuals : TRUE sam.binary :

17 928 ICES WGNSSK REPORT 2017 Table Turbot in Area 4. FLR, R SOFTWARE VERSIONS FLSAM.version 2.01 FLCore.version R.version R version ( ) platform x86_64-w64-mingw32 run.date :10:59 Table Turbot in Area 4. Stock Summary Year Recruitment Low High TSB Low High SSB Low High Fbar Low High Landings Landings Age 1 (Ages 2-6) SOP thousands thousands thousands f f f tonnes

18 ICES WGNSSK REPORT Table Turbot in Area 4. Estimated fishing mortality (units: na) age age age age

19 930 ICES WGNSSK REPORT 2017 Table Turbot in Area 4. Estimated population abundance (units: na) age age age age

20 ICES WGNSSK REPORT Table a. Turbot in Area 4. Predicted catch numbers at age (units: na) age age age age

21 932 ICES WGNSSK REPORT 2017 Table b. Turbot in Area 4. Catch at age residuals (units: na) age age age age Table a. Turbot in Area 4. Predicted index at age SNS (units: na). age age

22 ICES WGNSSK REPORT Table b. Turbot in Area 4. Index at age residuals SNS age age Table a. Turbot in Area 4. Predicted index at age BTS-ISIS age age age

23 934 ICES WGNSSK REPORT 2017 Table b. Turbot in Area 4. Index at age residuals BTS-ISIS age age age Table a. Turbot in Area 4. Predicted index at age NL_LPUE age all age all age all Table b. Turbot in Area 4. Index at age residuals NL_LPUE age all age all age all

24 ICES WGNSSK REPORT Table Turbot in Area 4. Fit parameters name value std.dev 1 logfpar logfpar logfpar logfpar logfpar logfpar logfpar logsdlogfsta logsdlogfsta logsdlogfsta logsdlogn logsdlogn logsdlogobs logsdlogobs logsdlogobs logsdlogobs logsdlogobs logsdlogobs logsdlogobs logsdlogobs logsdlogobs transfirardist transfirardist itrans_rho Table Turbot in Area 4. Negative Log-Likelihood Table Summary of SPiCT analyses (with priors on logn, logalpha, and logbeta). > summary(fit) Convergence: 0 MSG: both X-convergence and relative convergence (5) Objective function at optimum: Euler time step (years): 1/16 or Nobs C: 36, Nobs I1: 26, Nobs I2: 22, Nobs I3: 13 Priors inp$priors$logn <- c(log(2), 1, 0) inp$priors$logalpha <- c(log(2), 3, 0) inp$priors$logbeta <- c(log(2), 1, 0) Residual diagnostics (p-values) shapiro bias acf LBox shapiro bias acf LBox C *. I ** *** I I * * Model parameter estimates w 95% CI estimate cilow ciupp log.est alpha e e alpha e e alpha e e beta e e r e e rc e e rold e e

25 936 ICES WGNSSK REPORT 2017 m e e K e e q e e q e e q e e n e e sdb e e sdf e e sdi e e sdi e e sdi e e sdc e e Deterministic reference points (Drp) estimate cilow ciupp log.est Bmsyd e Fmsyd e MSYd e Stochastic reference points (Srp) estimate cilow ciupp log.est rel.diff.drp Bmsys e e Fmsys e e MSYs e e States w 95% CI (inp$msytype: s) estimate cilow ciupp log.est B_ e e e F_ e e e B_ /Bmsy e e e F_ /Fmsy e e e Predictions w 95% CI (inp$msytype: s) prediction cilow ciupp log.est B_ e e e F_ e e e B_ /Bmsy e e e F_ /Fmsy e e e Catch_ e e e E(B_inf) e+04 NA NA

26 ICES WGNSSK REPORT Figure Turbot in : Total landings by métier in 2016 sorted by sampled/unsampled for numbers at age in InterCatch Official landings (tonnes) Year Figure Turbot in Total landings (from the ICES database of official landings).

27 938 ICES WGNSSK REPORT 2017 Figure Turbot in Landings at age for the years with available data between Figure Turbot in Time series of the standardized indices for ages 1 to 9 from the three tuning fleets available for the assessment: BTS-ISIS (black), SNS (red) and NL beam trawl LPUE (blue; not used in the final assessment).

28 ICES WGNSSK REPORT Figure Turbot in Internal consistency of the two tuning indices available for the assessment : BTS-ISIS (top), and SNS (bottom)

29 940 ICES WGNSSK REPORT 2017 Figure Landings (left) and stock (right) weight at age from observations (points) and modelled values (lines).

30 ICES WGNSSK REPORT Figure parameter-correlation plot. It shows the correlation among all parameters that are estimated in the model. Fpar parameters refer to catchabilities, Fstates to the random walk in F, logn to the random walk in N, logobs to the observation variances, frardist to the auto-correlation in the surveys and trans_rho to the correlation in the F-random walks

31 942 ICES WGNSSK REPORT 2017 Figure plot showing the observation variance vs the CV of that estimate

32 ICES WGNSSK REPORT Figure catchabilities of the surveys for all surveys with more than 1 age-group

33 944 ICES WGNSSK REPORT 2017 Figure Estimated observation variances (scaling factor for each of the surveys), ordered from the best to the worst survey fit and has colour coding to show which bars belong to one dataset

34 ICES WGNSSK REPORT Figure estimated selectivity, grouped by a 5-year period. Values represent actual F-at-age

35 946 ICES WGNSSK REPORT 2017 Figure summary plot of SSB, F and Recruitment, including the uncertainty bounds

36 ICES WGNSSK REPORT Figure residual bubble plot of catches

37 948 ICES WGNSSK REPORT 2017 Figure residual bubble plot of SNS and BTS-ISIS survey

38 ICES WGNSSK REPORT Figure estimated correlation between age-groups, the rounder the circle, the less correlation there is, flat lines indicate a 100% correlation