Proceedings of the Regional Framework Meeting for Use of Research Vessel Survey Data to Assess Secondary Groundfish Stocks in the Maritimes Region

Transcription

1 Canadian Science Advisry Secretariat (CSAS) Prceedings Series 2015/029 Maritimes Regin Prceedings f the Reginal Framewrk Meeting fr Use f Research Vessel Survey Data t Assess Secndary Grundfish Stcks in the Maritimes Regin December 16-17, 2014 Dartmuth, Nva Sctia Chairpersn and editr: Kristian Curran Bedfrd Institute f Oceangraphy Fisheries and Oceans Canada 1 Challenger Drive, P.O. Bx 1006 Dartmuth, N.S. B2Y 4A2 July 2015

2 Frewrd The purpse f these Prceedings is t dcument the activities and key discussins f the meeting. The Prceedings may include research recmmendatins, uncertainties, and the ratinale fr decisins made during the meeting. Prceedings may als dcument when data, analyses r interpretatins were reviewed and rejected n scientific grunds, including the reasn(s) fr rejectin. As such, interpretatins and pinins presented in this reprt individually may be factually incrrect r misleading, but are included t recrd as faithfully as pssible what was cnsidered at the meeting. N statements are t be taken as reflecting the cnclusins f the meeting unless they are clearly identified as such. Mrever, further review may result in a change f cnclusins where additinal infrmatin was identified as relevant t the tpics being cnsidered, but nt available in the timeframe f the meeting. In the rare case when there are frmal dissenting views, these are als archived as Annexes t the Prceedings. Published by: Fisheries and Oceans Canada Canadian Science Advisry Secretariat 200 Kent Street Ottawa ON K1A 0E6 csas-sccs@df-mp.gc.ca Crrect citatin fr this publicatin Her Majesty the Queen in Right f Canada, 2015 ISSN DFO Prceedings f the Reginal Framewrk Meeting fr Use f Research Vessel Survey Data t Assess Secndary Grundfish Stcks in the Maritimes Regin; December 16-17, DFO Can. Sci. Advis. Sec. Prceed. Ser. 2015/029.

3 TABLE OF CONTENTS SUMMARY... iv SOMMAIRE... v INTRODUCTION... 1 PRESENTATION AND DISCUSSION... 2 The Prxy Apprach... 2 Discussin... 2 Evaluating Uncertainty... 3 Discussin... 3 CONCLUSIONS... 4 The Prxy Apprach... 4 Evaluating Uncertainty... 4 Dcumentatin f Meeting Findings... 4 REFERENCES CITED... 4 APPENDICES... 5 Appendix 1: List f Meeting Participants... 5 Appendix 2: Meeting Terms f Reference... 6 Appendix 3: Meeting Agenda... 8 Appendix 4: Evaluating Uncertainty (Stephen Smith)... 9 iii

4 SUMMARY A Prxy Apprach fr estimating reference pints fr secndary stcks was prpsed: the gemetric mean f the stratified ttal bimass frm the RV Survey (summer r winter, dependent n the stck) as a prxy fr bimass at Maximum Sustainable Yield (B MSY ) fr a chsen time perid. Bilgical reference pints were subsequently calculated where 40% f the prxy B MSY was prpsed as the Lwer Reference Pint and 80% f the prxy B MSY was prpsed as the Upper Stck Reference. The three-year running median f the bimass was then used t examine the lng term trend in bimass against the reference pints. Meeting participants discussed applicability f this apprach fr all secndary species, as it des nt take int accunt the nuance f bimass histry (e.g. fishing pressure). It was cncluded that fr sme stcks the bimass histry, as well as additinal analytics t evaluate uncertainty, is needed t characterize prpsed reference pints. Guidance fr evaluating uncertainty includes: 1) evaluate whether survey indices reflect fishery impact n the stck; 2) evaluate the impact f chice f perid in the time series; and 3) evaluate additinal infrmatin t infrm the impact f the Prxy Apprach. Given that many secndary stcks assessed at the meeting required further analysis, it was agreed that a Science Advisry Reprt wuld nt be cmpleted and published as part f the meeting s prducts. Last, it was agreed by sme meeting participants that despite limitatins in the prpsed Prxy Apprach, it remains very imprtant t derive reference pints fr many f the secndary stcks in questin, as there are sme urgent management needs fr many f the species that were discussed (especially fr skates). Additinal research in this regard was encuraged by all meeting participants. iv

5 Cmpte rendu de la réunin réginale sur le cadre d'utilisatin des dnnées des relevés par navire de recherche pur évaluer les stcks secndaires de pissn de fnd dans la régin des Maritimes SOMMAIRE Afin d'estimer les pints de référence pur les stcks secndaires, n prpse d'adpter une «apprche axée sur des valeurs apprximatives», qui cnsiste à utiliser la myenne gémétrique de la bimasse ttale stratifiée tirée du relevé par navire de recherche (estival u hivernal seln le stck) cmme valeur apprximative de la bimasse au rendement maximal sutenu (B RMS ) pur une péride chisie. Par la suite, les pints de référence bilgiques snt calculés avec un pint de référence limite prpsé s'élevant à 40 % de l'indicateur de B RMS et un pint de référence supérieur prpsé s'élevant à 80 % de l'indicateur de B RMS. La médiane mbile de la bimasse sur tris ans est ensuite utilisée pur examiner les tendances à lng terme de la bimasse par rapprt aux pints de référence. Les participants à la réunin discutent de l'applicabilité de cette apprche à tutes les espèces secndaires, car elle ne tient pas cmpte de la nuance de l'histrique de la bimasse (p. ex. pressin exercée par la pêche). On cnclut que, pur certains stcks, il faut tenir cmpte de l'histrique de la bimasse, ainsi que des analyses supplémentaires à l'appui de l'évaluatin des incertitudes, pur caractériser les pints de référence prpsés. Les directives pur l'évaluatin des incertitudes snt les suivantes : 1) évaluer si les indices tirés des relevés témignent des répercussins de la pêche sur le stck; 2) évaluer l'incidence du chix de la péride de la série chrnlgique; 3) évaluer des renseignements supplémentaires pur clarifier l'effet de l'apprche axée sur des valeurs apprximatives. Étant dnné que de nmbreux stcks secndaires évalués au curs de la réunin exigent une analyse plus apprfndie, il est cnvenu qu'un avis scientifique ne serait pas rédigé ni publié cmme résultat de la réunin. Enfin, certains participants à la réunin cnviennent que, malgré les limites de l'apprche prpsée quant à l'utilisatin de valeurs apprximatives, il est tujurs très imprtant d'établir des pints de référence pur bn nmbre des stcks secndaires en questin, car de nmbreuses espèces ayant fait l'bjet des discussins (surtut les raies) nt certains besins urgents en matière de gestin. Tus les participants à la réunin suhaitent que d'autres recherches sient réalisées à cet égard. v

6 INTRODUCTION Fisheries and Oceans Canada s (DFO) Precautinary Apprach (PA) framewrk aims t be guided by units that relate directly t fish stck prductivity. As such, the PA framewrk suggests pririty be given t mnitring a stck and establishing data time series that allw fr the identificatin f bilgical reference pints. Fr reprting against the PA framewrk, fishery management plans strive t include harvest strategies that incrprate a Limit Reference Pint (LRP) and an Upper Stck Reference (USR) (DFO 2006). The LRP and USR define the bundaries f the critical, cautius, and healthy stck status znes fr a fish stck; the LRP is the divide between the cautius and critical znes and the USR the divide between the healthy and cautius znes. Ideally, all fish stcks incrprate reference pints int their management; hwever, secndary fish stcks generally are nt managed n this basis. T differentiate between Primary and Secndary stck categries in the DFO Maritimes Regin (DFO Maritimes), a prtcl entitled Pririty Setting Prtcl fr Fishery Assessment and Management: Primary and Secndary Stcks in the Maritimes Regin was presented at the DFO Maritimes Regin Industry Rundtable. The verall bjective f the pririty setting prtcl is t better align fishery management strategies with underlying science needs. Fish stcks that typically fall int the secndary stck categry include grundfish stcks being caught by grundfish fisheries r caught as bycatch in ther fisheries where the value and vlume f their landings are relatively small (r where the stck has nt therwise been identified as a pririty by the DFO Maritimes Regin). In supprt f fish stck mnitring, DFO Maritimes has cnducted an annual summer RV survey since 1970 in the Nrthwest Atlantic Fisheries Organizatin (NAFO) divisins 4VWX and the Canadian prtin f 5Yb, as well as an annual winter RV survey since 1987 in NAFO Area 5Z (Gerges Bank). Results frm these surveys prvide infrmatin fr many grundfish species in the regin. The DFO Maritimes summer and winter RV surveys prvide data n bimass and abundance indices fr grundfish stcks that span 45 years and 27 years, respectively. At an Advisry prcess fr the develpment f reference pints cnsistent with the precautinary apprach fr a variety f stcks in the Maritimes Regin it was prpsed that the RV survey bimass index culd be used as a bimass at Maximum Sustainable Yield (B MSY ) prxy if the survey index was thught t be reflective f a stck s ppulatin trend (DFO 2012). T explre this pssibility further fr secndary fish stcks, a framewrk assessment science advisry meeting was held December, 16-17, 2014, at the Bedfrd Institute f Oceangraphy, Dartmuth, Nva Sctia. The verall bjective f the meeting was t peer-review a prpsed methd fr estimating bilgical reference pints fr a variety f secndary grundfish stcks in DFO Maritimes using summer and winter RV survey bimass indices. The meeting Chairpersn, Mr. Kristian Curran, first intrduced himself, fllwed by an intrductin f meeting participants (Appendix 1). The Chair thanked meeting participants fr attending the DFO Science Advisry Prcess. The Chair nted that this was a science peer-review meeting in which a science advisry reprt may be cmpleted pending acceptance f the methd being prpsed. The Chair prvided a brief verview f the Canadian Science Advisry Secretariat (CSAS) science advisry prcess and invited participants t review the meeting Terms f Reference (Appendix 2) and Agenda (Appendix 3). N revisins r additins were made t the Terms f Reference r Agenda. T guide discussin, a Wrking Paper was prvided t meeting participants n Nvember 28, 2014, in advance f the meeting date. This Prceeding cnstitutes a recrd f the meeting s discussin and cnclusins. 1

7 Rapprteur: Kristian Curran THE PROXY APPROACH Presenters: Tara McIntyre and Dn Clark PRESENTATION AND DISCUSSION The lead authrs f the Wrking Paper, Tara McIntyre and Dn Clark, briefly reviewed the DFO Maritimes plicy n science in supprt f secndary stcks. They clarified that primary stcks typically receive full science assessment, while secndary stcks d nt receive similar supprt given their lwer pririty status. The authrs further nted that RV Survey trends prvide a mnitring tl fr varius fish stcks in the regin, and that DFO Maritimes aims t develp mre rbust metrics fr secndary stcks cnsistent with the PA Plicy. Based n RV Survey data, the authrs prpsed a Prxy Apprach fr estimating reference pints fr secndary stcks: the gemetric mean f the stratified ttal bimass frm the RV Survey (summer r winter, dependent n the stck) as a prxy fr B MSY fr a chsen time perid. Bilgical reference pints were subsequently calculated where 40% f the prxy B MSY was prpsed as the LRP and 80% f the prxy B MSY was prpsed as the USR. The three-year running median was then used t examine the lng term trend in bimass against the reference pints, as it dampens the effect f utliers caused by individual survey years f high bimass resultant f ne r tw large catches. Cnfidence intervals were calculated fllwing the btstrap with replacement methd prpsed by Smith (1997). Fr sme f the stcks in this prcess, a truncated perid f time was identified frm the time series f stratified ttal bimass as the reference perid when calculating B MSY. The decisin t use a truncated perid was based n an assumptin that a reductin in bimass bserved in the recent past may nt have been caused by a shift in its prductivity regime, and that there is ptential in the future fr the stck t return t the bimass bserved in the reference perid. Discussin The authrs presented results f the Prxy Apprach n a species-by-species basis. A meeting participant requested clarificatin n hw the three year running median was calculated. The authr indicated that a number f end-pint rules existed, further nting that the nly data pint that matters is the last ne the median f the past three values. There was further discussin n methds that take the slpe f survey weight t prject frward, althugh it was agreed it is difficult t prject a meaningful trend when there is much inter-annual variability in the data. In cntrast, a meeting participant suggested that the last value nly has imprtance when yu have a Harvest Cntrl Rule (HCR) r ther decisin rules in place; fr example, 4X White Hake has an HCR with a built in respnse t accunt fr inter-annual variability (i.e. cnservative management measure). Fr species such as 4X White Hake, which exhibit high levels f interannual variability in the data, applying a cnservative HCR allws fr mre predictable explitatin ver the lng-term that is nt subject t fluctuatin in the last data pint in the survey time series. It was agreed by meeting participants that when calculating reference pints and develping HCRs it is imprtant t knw the rigins, limitatins, and histry f the data. The authrs indicated that the prpsed analytical apprach is cnservative in nature, althugh the cnfidence intervals ften span bth the upper and lwer reference pints. A meeting participant nted that cnfidence acrss years is mre nuance, and that the prpsed apprach makes a lt f assumptins abut cnstancy that may nt be true fr all secndary species being evaluated (e.g. survey catch may change frm year-t year, althugh stck abundance may nt have). It was nted, hwever, that cnfidence intervals prvide imprtant infrmatin 2

8 that shuld be included in the analysis, althugh ne shuld be careful hw such intervals are interpreted. As further evaluatin f the prpsed apprach, a meeting participant inquired if it had been applied t mre well-knwn stcks t see hw rbust the apprach may be. The authr indicated that this had been lked at briefly, but nt in any significant detail. The meeting participant replied that ne may get differences when cmparing reference pints estimated frm the Prxy Apprach cmpared t thse estimated frm ther analytical appraches used fr primary species (e.g. VPA), particularly when analytically-based reference pints d nt fall within the bimass time series fr the given species. The meeting discussin then turned t suitability f applying the prpsed apprach t secndary species that have little infrmatin. Fr instance, varius appraches have been used t evaluate reference pints fr pllck that reach similar cnclusins, althugh it is knwn fr pllck that the RV Survey has many limitatins in its design and gear (the discussin briefly fcused n cnversin factrs applied t secndary stcks when DFO changed its survey vessels/gear). Similarly, fr species believed t be ver-depleted, it may be difficult t identify which prtin f the time series is mst suitable t estimate reference pints; thus, the Prxy Apprach may need t better accunt fr a stck s histry, in rder t better understand the underlying cntext f the estimated reference pints. Further, the gegraphic range f the RV Survey being limited t Canadian waters may hinder the ability t distinguish between species mvement frm actual increases/decreases in bimass (particularly fr transbundary stcks). Regardless f limitatins in the Prxy Apprach, sme meeting participants agreed that deriving reference pints fr many f the secndary species in questin is f great imprtance, as there are sme urgent management needs fr many f the species (especially fr skates). Additinal research in this regard was encuraged by all meeting participants. EVALUATING UNCERTAINTY Presenter: Stephen Smith Stephen Smith examined the use f additinal stck infrmatin t infrm the Prxy Apprach, as well as t evaluate uncertainty within the cntext f the Prxy Apprach (Appendix 4). As an example, the haddck bimass time series appeared t be made up f fur r five events, with recent prductivity being lwer than the whle series r during the fishery s perid. The presenter nted that all reference perids exhibited a trend with respect t B MSY, althugh recent annual changes exhibited mre decreases than increases the presenter questined if this culd be a fishery effect. Based n additinal analytics, the presenter cncluded that a tw-phased mdel fr survey data pst-1991 might be applicable in evaluating the haddck stck. Mr. Smith s presentatin demnstrated the imprtance f drawing upn additinal infrmatin when interpreting a fish stck, as well as the ptential need fr additinal analytics t evaluate uncertainty when applying the Prxy Apprach. Discussin It was agreed by meeting participants that the intent f bilgical reference pints is t say smething abut a stck s prductivity in this case survey bimass is used t set prxies fr verall bimass. It was further agreed that the Prxy Apprach that has been presented has been applied t ther fish stcks, althugh depending n the individual fish stck there may be additinal infrmatin (e.g. histrical cntext f fishery) that is nt being used t infrm reference pints r is nt available (e.g. bycatch species nt cnsistently mnitred) t infrm reference pints. Given uncertainty in applying the prpsed Prxy Apprach t all secndary species, it is recmmended that further research be undertaken t explre the secndary stcks that have been presented, including an evaluatin f uncertainty where it may apply. 3

9 THE PROXY APPROACH CONCLUSIONS A Prxy Apprach fr estimating reference pints fr secndary stcks was prpsed: the gemetric mean f the stratified ttal bimass frm the RV Survey (summer r winter, dependent n the stck) as a prxy fr B MSY fr a chsen time perid. Bilgical reference pints were subsequently calculated where 40% f the prxy B MSY was prpsed as the LRP and 80% f the prxy B MSY was prpsed as the USR. The three-year running median was then used t examine the lng term trend in bimass against the reference pints, as it dampens the effect f utliers caused by individual survey years f high bimass resultant f ne r tw large catches. Cnfidence intervals were calculated fllwing the btstrap with replacement methd prpsed by Smith (1997). Fr sme f the stcks that were analyzed, a truncated time perid was identified frm the time series f stratified ttal bimass as the reference perid when calculating B MSY. The decisin t use a truncated time perid fr sme stcks was based n an assumptin that a reductin in bimass bserved in recent past may nt have been caused by a shift in the stck s prductivity regime and that there is ptential in the future fr the stck t return t the bimass bserved in the reference perid. EVALUATING UNCERTAINTY Guidance fr evaluating uncertainty assciated with the Prxy Apprach is: 1. Evaluate if survey indices reflect a fishery s impact n the stck; 2. Evaluate the impact f chice f perid in the time series; and 3. Evaluate additinal infrmatin yu have t infrm the impact f the Prxy Apprach. DOCUMENTATION OF MEETING FINDINGS It was agreed by meeting participants that the final Research Dcument wuld include all species presented and discussed at the meeting, and that the dcument wuld als include the analysis presented by Stephen Smith fr further guidance n dealing with a stck s histry, as well as evaluating uncertainty, when applying the Prxy Apprach. Given that many species discussed at the meeting required further analysis, it was agreed that a Science Advisry Reprt wuld nt be cmpleted and published as part f the meeting prducts. As such, prducts f the framewrk assessment meeting are nly t include a Research Dcument and Prceedings. REFERENCES CITED DFO, A Harvest Strategy Cmpliant with the Precautinary Apprach. DFO Can. Sci. Advis. Sec. Sci. Advis. Rep. 2006/023. DFO Reference Pints Cnsistent with the Precautinary Apprach fr a Variety f Stcks in the Maritimes Regin. DFO Can. Sci. Advis. Sec. Sci. Advis. Rep. 2012/035. Smith, S.J Btstrap Cnfidence Limits fr Grundfish Trawl Survey Estimates f Mean Abundance. Can. J. Fish. Aquat. Sci. 54:

10 APPENDICES APPENDIX 1: LIST OF MEETING PARTICIPANTS Name Affiliatin Christensen, Camern* Dalhusie University / Bilgy Clark, Dn DFO Maritimes / Ppulatin Eclgy Divisin (BIO) Claytr, Rss DFO Maritimes / Ppulatin Eclgy Divisin (BIO) Cffen-Smut, Sctt* DFO Maritimes / Oceans and Castal Management Cmeau, Peter DFO Maritimes / Ppulatin Eclgy Divisin (BIO) Ck, Adam DFO Maritimes / Ppulatin Eclgy Divisin (BIO) Curran, Kristian DFO Maritimes / Centre fr Science Advice d'entremnt, Alain Sctia Harvest Seafds Inc. / O'Neil Fisheries Ltd. Francis, Cry* The Cnfederacy f Mainland Mi kmaq (CMM) Fuller, Susanna* Eclgy Actin Centre (EAC) Grant, Catharine* Eclgy Actin Centre (EAC) Hurley, Peter* DFO Maritimes / Ppulatin Eclgy Divisin (BIO) MacDnald, Carl DFO Maritimes / Resurce Management McIntyre, Tara DFO Maritimes / Ppulatin Eclgy Divisin (BIO) Perrier, Erika Mi kmaw Cnservatin Grup Schleit, Katie* Eclgy Actin Centre (EAC) Smith, Stephen DFO Maritimes / Ppulatin Eclgy Divisin (BIO) Vasctt, Kris NS Dept. Fisheries and Aquaculture / Marine Wrm, Bris* Dalhusie University / Bilgy *Participants did nt attend bth days f the meeting 5

11 APPENDIX 2: MEETING TERMS OF REFERENCE Terms f Reference Framewrk fr Use f Research Vessel Survey Data t Assess Secndary Grundfish Stcks in the Maritimes Regin Reginal Peer Review Maritimes Regin December 16-17, 2014 Dartmuth, NS Chairpersn: Kristian Curran Cntext In Fisheries and Oceans Canada s (DFO s) Maritimes Regin, a prtcl describing pririty stcks fr fishery assessment and management departmental resurce allcatin has been prduced with an verall bjective f better alignment between the science prvided and the resulting management strategy. This prtcl separates stcks int Primary and Secndary categries that define the difference in the DFO respnse t requests fr advice and allcatin f departmental human and financial resurces. Secndary grundfish stcks include grundfish stcks being caught by the grundfish fisheries, r caught as bycatch in ther fisheries, where the value and vlume f landings f that stck are relatively small, and where the stck has nt therwise been identified as a pririty by the DFO Maritimes Regin (e.g., because f an imprtant eclgical rle played by the stck r because the stck has imprtant cultural uses). Frmal stck assessments are nt cnducted by DFO Science fr secndary stcks in the Maritimes Regin. Hwever, trends in Research Vessel survey indices are reprted annually fr sme secndary grundfish stcks. Research vessel survey infrmatin has been used t set reference pints fr sme grundfish stcks in the Maritimes Regin, e.g., Eastern Pllck (DFO 2012). Objectives The bjective f this meeting is t develp a methdlgy fr using research vessel survey data t assess secndary grundfish stcks in the Maritimes Regin, and, mre specifically, t prvide advice n an apprpriate methd fr calculating reference pints fr the fllwing secndary grundfish stcks: White Hake (4X5Yb, 4VW) Haddck (4VW) Mnkfish (4X5Yb) Thrny Skate (5Z, 4X5Yb) Little Skate (5Z, 4X5Yb) Barndr Skate (5Z, 4X5Yb) Smth Skate (5Z, 4X5Yb) Lnghrn Sculpin (5Z, 4X5Yb) Expected Publicatins CSAS Science Advisry Reprt CSAS Prceedings CSAS Research Dcument(s) 6

12 Participatin DFO Science, Resurce Management, and Ecsystem Management branches Prvinces f Nva Sctia and New Brunswick Academics Abriginal cmmunities/rganizatins Fishing Industry Other invited experts References DFO Reference pints cnsistent with the precautinary apprach fr a variety f stcks in the Maritimes Regin. DFO Can. Sci. Advis. Sec. Sci. Advis. Rep. 2012/035. 7

13 APPENDIX 3: MEETING AGENDA Framewrk fr Use f Research Vessel Survey Data t Assess Secndary Grundfish Stcks in the Maritimes Regin December 16, 2014 Tuesday 1:00 1:15 Intrductin (Chair) Maritimes Reginal Peer Review Needler Bardrm Bedfrd Institute f Oceangraphy 1 Challenger Drive, Dartmuth, NS December 16-17, 2014 Chair: Kristian Curran DRAFT AGENDA 1:15 2:00 Presentatin f methd fr calculating reference pints fr secndary species 2:00 2:45 Discussin 2:45 3:00 Break 3:00 3:45 Discussin cntinued 3:45 4:00 Wrap-up Day 1 December 17, 2014 Wednesday 9:00 9:15 Review f Day 1 9:15 10:30 Cntinue discussin f methdlgies 10:30 10:45 Break 10:45 12:00 Review f SAR 12:00 1:00 Lunch 1:00 2:45 Cntinue review f SAR 2:45 3:00 Wrap-up f meeting 8

14 APPENDIX 4: EVALUATING UNCERTAINTY (STEPHEN SMITH) Slide 1 f 17 SECONDARY SPECIES RAP Stephen Smith DECEMBER 15/ Slide 2 f 17 Precautinary Apprach Limit reference pint (LRP): Avid recruitment verfishing 40% B msy Upper stck reference pint (USR): Allw time fr management decisins t be made t avid LRP. Can be a target 80% B msy Remval reference pint: Part f harvest cntrl rule Management actin thrugh changing fishing mrtality 9

15 Slide 3 f 17 Precautinary Apprach: B MSY In sme jurisdictins is a target (e.g., US, ICES) Als a reference pint fr MSC stcks Prductivity based, a functin f: Grwth Natural mrtality Recruitment DFO PA plicy allws fr estimating B MSY using mean f a stck indicatr ver prductive perid Slide 4 f 17 Precautinary Apprach: Uncertainty DFO POLICY ACKNOWLEDGES THAT UNCERTAINTY WITH RESPECT TO AVAILABLE INFORMATION AND KNOWLEDGE NEED TO BE CONSIDERED Designatin f primary/secndary species reflects wrk pririties but Als reflects amunt f infrmatin and uncertainty Table f LRPs and USRs des nt necessarily cnvey degree f uncertainty Stcks in Table 1 represent a range f knwledge, data and uncertainty Sme ther jurisdictins (e.g., US, Australia) have develped a tier system 10

16 Slide 5 f 17 Precautinary Apprach: U.S. Tier System Data rich Tier 1: Reliable pint estimates f B and B msy and reliable pdf f Fmsy Tier 2: Reliable pint estimates f B, B msy, F msy, F35%, and F40% Tier 3: Reliable pint estimates f B, B40%, F35%, and F40% Tier 4: Reliable pint estimates f B, F35%, and F40% Data pr Tier 5: Reliable pint estimates f B and natural mrtality rate M Tier 6: Reliable catch histry frm Uncertainty taken int accunt in advice Different calculatins fr ver-fishing level and allwable bilgical catch Objective is t prvide mre cnservative advice as uncertainty increases 11

17 Slide 6 f 17 Secndary Species All reference pints based n survey data even thugh there may be ther infrmatin available (e.g., 4VW haddck) N fishery data, catch series fr cntext Survey indices use all strata in management area even thugh species may nt ccupy all strata Justificatin fr prductive perid nt bvius r necessarily cnsistent All survey indices are assumed t track the impact f the fishery n the ppulatin Slide 7 f 17 Survey Indices: Optins 1. Track ppulatin dynamics: Impact f fishery: A) Tw-phase mdel (Trenkel, 2008) B t = g t-1 B t-1 + R t-1, Lg(g t ) = lg(g t-1 )+ε t, ε t ~N(-0.5σ 2, σ 2 ) B) CUSUM and SS_CUSUM mdel (Mesnil and Petitgas 2009, Pazhayamadm et al. 2013) Quality cntrl chart methdlgy Rule t pick up deviatins frm reference value in terms f standard deviatin units. 12

18 Slide 8 f 17 Prductive Perid 1. Stable perid in terms f explitatin? 2. Highest prductivity (transient?) 3. Why a gemetric mean? 4. Median mre insensitive t ccasinal large survey indices. Slide 9 f 17 Uncertainty Measures 1) Behaviur during reference perid: Runs abve and belw Bmsy estimate 2) Annual trend test: Runs test applied t annual differences 3) Directin f trend: Sign test applied t annual differences 4) Grwth assumptins fr Tw-phase mdel: Randm annual grwth vs. randm walk 13

19 Slide 10 f 17 Example f 4VWX Haddck 4VW haddck: whle time series used: 1.Fishery until 1993, bycatch and clsure after that 2.Assessment used q-crrected survey trends 3.Substantial grwth changes nted after mid-1980s (Mhn and Simn 2004) 4.Pssible habitat availability changes (Smith and Page 1994, Smith et al. 1994) 5.Recruitment available frm survey Slide 11 f 17 Example f 4VWX Haddck: B MSY Estimates Years Gemetric mean Median

20 Slide 12 f 17 Example f 4VWX Haddck Slide 13 f 17 Example f 4VWX Haddck: Uncertainty Measures Years Abve and belw B msy Annual Trend Run test Directin Sign test Randmness Trend (p<0.001) Randm (p=0.109) Decreasing (p=0.047) Randm walk (p=0.026) Trend (p=0.008) Randm (p=0.199) N trend (p=0.17) Randm (p=0.10) Trend (p=0.001) Randm (p=0.165) Decreasing (p=0.005) Randm walk (p=0.02) 15

21 Slide 14 f 17 Example f 4VWX Haddck: Uncertainty Measures Haddck time series appears t be made up f fur r five events. Recent prductivity lwer than whle series r during fishery perid All reference perids exhibit trend with respect t B MSY Recent annual changes exhibit mre decreases than increases a fishery effect? Tw-phase mdel fr survey data pst 1991 might be applicable Slide 15 f 17 Further Cmments White Hake reference perid ( ) N TAC until 1996 Landings peaked in 1987 (survey in 1985) Bycatch fishery starting in 1999 (quta caps fr fixed and mbile gear) Stck cmplex 4VWX/5Y. Bundy and Simn (2005) reprt <10% catch in 4VW after 2002 N evidence fr trends after 1991 (runs test and sign test). Randm walk test indicates memry after

22 Slide 16 f 17 White Hake CUMSUM Plt Slide 17 f 17 Main Pints: Secndary Species Lwer pririty fr assessment wrk and unlikely t be reviewed anytime sn Ignres fishery infrmatin All species here are bycatch f ther target species Management actin -> reduce fishery mrtality r use mve away prtcls fr target species Assume that survey tracks ppulatin trends withut any qualificatin/evaluatin In turn wuld expect t see impact f management actin in survey index 17