marine fisheries stock assessment improvement plan

i

Marine FisheriesStock Assessment Improvement PlanReport of the National Marine Fisheries ServiceNational Task Force for Improving Fish Stock Assessments

Pamela M. Mace (Chair), Norman W. Bartoo, Anne B. Hollowed,

Pierre Kleiber, Richard D. Methot, Steven A. Murawski,

Joseph E. Powers, and Gerald P. Scott

October 2001NOAA Technical Memorandum NMFS-F/SPO-56

U.S. DEPARTMENT OF COMMERCEDonald Evans, Secretary

National Oceanic and Atmospheric AdministrationVice Admiral Conrad C. Lautenbacher Jr., USN (ret.), Under Secretary for Oceans and Atmosphere

National Marine Fisheries ServiceWilliam T. Hogarth, Assistant Administrator for Fisheries

ii

Copies of this document may be obtained by contacting:

Office of Science and Technology, F/STNational Marine Fisheries Service, NOAA1315 East West HighwaySilver Spring, MD 20910

An online version is available at http://www.st.nmfs.gov/st2/index.html

Cover photo: Menhaden school. Credit: W. F. Hettler, National Marine Fisheries Service.

This publication may be cited as:

NMFS. 2001. Marine Fisheries Stock Assessment Improvement Plan. Report of the National MarineFisheries Service National Task Force for Improving Fish Stock Assessments. U.S. Dep. Commerce, NOAATech. Memo. NMFS-F/SPO-56, 69 p., 25 appendices.

iii

Contents

List of Illustrations ............................................................................................................................................. vi

List of Tables ..................................................................................................................................................... vii

Preface ................................................................................................................................................................ ix

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

I. Introduction ................................................................................................................................................... 5Scope of the Stock Assessment Improvement Plan ...................................................................................... 5

II. Defining NMFS� Stock Assessment Mandate .............................................................................................. 7A. What is a �Stock Assessment� .............................................................................................................. 8B. The Quality of NMFS� Assessments ..................................................................................................... 9C. Changing Demands .............................................................................................................................. 10

Northeast Fisheries Science Center .................................................................................................... 10Southeast Fisheries Science Center .................................................................................................... 11Southwest Fisheries Science Center ................................................................................................... 12Northwest Fisheries Science Center ................................................................................................... 12Alaska Fisheries Science Center ......................................................................................................... 13

D. The Credibility of NMFS� Science ...................................................................................................... 13E. Implications of the Precautionary Approach ....................................................................................... 14F. Implications of the Need to Incorporate Ecosystem Considerations ................................................... 15

III. Assessment and Management Strategy Evaluation Needs.......................................................................... 17A. Input Data ............................................................................................................................................ 17

(i) Fishery-dependent data needs ................................................................................................... 17(ii) Fishery-independent data needs ................................................................................................ 18

B. Input Data: Minimal and Optimal Requirements ................................................................................ 19C. Stock Assessment Models ................................................................................................................... 23D. Inventory of the Status of Stock Assessments: Adequacy of Input Data,

Assessment Level, and Frequency of Assessments .............................................................................. 24E. Adequacy of Technology and Infrastructure ....................................................................................... 24F. Peer Review of Assessments ................................................................................................................ 25

Northeast Fisheries Science Center .................................................................................................... 25Southeast Fisheries Science Center .................................................................................................... 26Southwest Fisheries Science Center ................................................................................................... 26Northwest Fisheries Science Center ................................................................................................... 27Alaska Fisheries Science Center ......................................................................................................... 27

G. Translation of Stock Assessment Advice into Management Action .................................................... 27H. Communication of Assessment Results and Follow-up Evaluations ................................................... 29I. Staffing Issues ..................................................................................................................................... 29

Education and training ....................................................................................................................... 29Time and motion analysis ................................................................................................................... 30Beyond assessment scientists .............................................................................................................. 30Northeast case study ........................................................................................................................... 32

iv

IV. Resource Requirements ............................................................................................................................. 37A. Programmatic Needs: Responses to questionnaires ............................................................................ 37B. Three Tiers of Assessment Excellence ................................................................................................ 37

Tier 1 � Improve stock assessments using existing data ................................................................... 37Tier 2 � Elevate stock assessments to new national standards of excellence ................................... 38Tier 3 � Next generation assessments ............................................................................................... 39

C. Timeframes and Relationships Between the Tiers ............................................................................... 43D. Region-Specific Needs to Achieve the Three Tiers of Assessment Excellence ................................... 44

Northeast Fisheries Science Center ................................................................................................... 44NEFSC current situation ............................................................................................................................ 44NEFSC programs and staffing required to meet the three tiers of excellence ......................................... 45

Southeast Fisheries Science Center ................................................................................................... 46SEFSC current situation ............................................................................................................................ 46SEFSC programs and staffing required to meet the three tiers of excellence .......................................... 47

Southwest Fisheries Science Center ................................................................................................... 48SWFSC current situation � central and western Pacific .................................................................... 49SWFSC current situation � west coast ...................................................................................................... 50SWFSC programs and staffing required to meet the three tiers of excellence ....................................... 51

Northwest Fisheries Science Center .................................................................................................. 52NWFSC current situation ........................................................................................................................... 52NWFSC programs and staffing required to meet the three tiers of excellence ........................................ 52

Alaska Fisheries Science Center ........................................................................................................ 54AFSC current situation � Gulf of Alaska groundfish ............................................................................... 56AFSC current situation � Bering Sea / Aleutian Islands ................................................................... 56AFSC programs and staffing required to meet the three tiers of excellence ........................................... 57

E. Summary: National Resource Requirements ....................................................................................... 60F. The Benefits of Implementing the Stock Assessment Improvement Plan ............................................ 61

V. Recommendations...................................................................................................................................... 65

References ........................................................................................................................................................ 67

Acknowledgments ............................................................................................................................................ 68

Acronyms ......................................................................................................................................................... 69

Appendix 1. Levels of input data (catch, abundance and life history parameters),assessment methodology, and assessment frequency for the 904 species listed in theNMFS (1999a) Report to Congress on the Status of Fisheries of the United States .............. A1

Appendix 2. Summary of the NMFS Science Quality Assurance Program .............................................. A25

Appendix 3. Executive Summary from the NOAA Fisheries Data Acquisition Plan .............................. A26

Appendix 4. Summary Description of the NMFS Stock Assessment Toolbox ........................................ A27

Appendix 5. Summary of the Objectives and Scope of the Center for Independent Experts Program ... A28

Appendix 6. Summary of the Joint NMFS/Sea Grant Graduate Fellowship Program ............................ A29

v

Appendix 7. Extract from the Executive Summary of the NRC Report on Improving FishStock Assessments (NRC 1998a) ......................................................................................... A31

Appendix 8. Executive Summary of the Report to Congress on a Proposed Implementationof a Fishing Vessel Registration and Fisheries Information Management System ............... A34

Appendix 9. Executive Summary of the NMFS Bycatch Plan ................................................................. A37

Appendix 10. Summary of the National Observer Program Initiative ....................................................... A39

Appendix 11. Extract from the NMFS Social Sciences Plan and FY2001 Budget Initiative ..................... A41

Appendix 12. Terms of Reference for the NMFS Standing Working Group on Advanced Technologies .. A43

Appendix 13. Summary of the Fisheries and the Environment (FATE) FisheriesOceanography Initiative ....................................................................................................... A45

Appendix 14. Protected Species Programs, Plans, and Initiatives ............................................................. A46

Appendix 15. Habitat Programs, Plans, and Initiatives ............................................................................. A48

Appendix 16. Extract from the Executive Summary of the NRC Report on Review ofNortheast Fishery Stock Assessments (NRC 1998b) ........................................................... A50

Appendix 17. Executive Summary from �Technical Guidance on the Use of PrecautionaryApproaches to Implementing National Standard 1 of the Magnuson-StevensFishery Conservation and Management Act� (Restrepo et al. 1998) ................................... A51

Appendix 18. Conclusions and Recommendations from the Executive Summary of the NRCReport on Sustaining Marine Fisheries (NRC 1999) ........................................................... A53

Appendix 19. Extract from the Executive Summary of the Ecosystem Principles AdvisoryPanel Report (NMFS 1999b) ............................................................................................... A57

Appendix 20. Extract from the Executive Summary of the NRC Study on Improving the Useand Collection of Fisheries Data (NRC 2000) ..................................................................... A59

Appendix 21. List of Relevant National Marine Fisheries Service Partnerships ........................................ A62

Appendix 22. Cooperative Research Programs with the Fishing Industry .................................................. A67

Appendix 23. Summary of NOAA�s Ocean Exploration Program ............................................................. A68

Appendix 24. Summary of the Census of Marine Life Program................................................................. A69

Appendix 25. A Non-exhaustive List of Other Programs and Activities that Could ProvideData and Other Inputs to Help Launch Stock Assessments Towards Tier 3 ........................ A70

vi

List of Illustrations

Figure 1. Relationship between availability of information and levels of uncertainty................................... 15

Figure 2. Theoretical effects of added information on recommended biological yields ................................. 15

Figure 3. Summary descriptions of levels of factors used to classify stocks in termsof input data and assessment status ............................................................................................... 20

Figure 4. Number of stocks with various levels of input data, assessment methodology,and assessment frequency .............................................................................................................. 22

Figure 5. Time and motion analysis for NMFS assessment scientists ........................................................... 31

Figure 6. Schematic showing relative staffing requirements to support the provision ofscientific advice for fisheries management ..................................................................................... 32

Figure 7. Schematic showing the relative costs of adding new species to be assessed .................................. 32

Figure 8. Assessment-related staffing levels by type of activity for theNortheast Fisheries Science Center ................................................................................................ 33

Figure 9. Programmatic needs averaged over responses from assessment scientistswithin each Science Center ............................................................................................................ 38

Figure 10. Impediments to the quality of assessments averaged over responses fromassessment scientists within each Science Center .......................................................................... 39

Figure 11. Fishery-dependent data needs averaged over responses from assessmentscientists within each Science Center ............................................................................................. 40

Figure 12. Fishery-independent data needs averaged over responses from assessmentscientists within each Science Center ............................................................................................. 41

Figure 13. Summary of the key features of the three Tiers of Assessment Excellence ................................... 42

Figure 14. Number of stocks assessed by assessment level at theNortheast Fisheries Science Center ................................................................................................ 44

Figure 15. Summary of FTE requirements by Science Center, Tiers of AssessmentExcellence, and activity .................................................................................................................. 61

vii

List of Tables

Table 1. Numbers of fish stocks with various levels of input data,assessment methodology, and assessment frequency...................................................................... 21

Table 2. FTEs required to meet the three Tiers of Assessment Excellence by type of activity for the Northeast Fisheries Science Center ..................................................... 45

Table 3. FTEs required to meet the three Tiers of Assessment Excellenceby type of activity for the Southeast Fisheries Science Center ...................................................... 48

Table 4. FTEs required to meet the three Tiers of Assessment Excellenceby type of activity for the Southwest Fisheries Science Center ..................................................... 50

Table 5. FTEs required to meet the three Tiers of Assessment Excellenceby type of activity for the Northwest Fisheries Science Center ..................................................... 54

Table 6. FTEs required to meet the three Tiers of Assessment Excellenceby type of activity for the Alaska Fisheries Science Center ........................................................... 57

Table 7. FTEs required to meet the three Tiers of Assessment Excellenceby type of activity for all NMFS Science Centers combined ......................................................... 62

Table 8. Total FTEs required to meet the three Tiers of Assessment Excellencefor each Science Center and all Centers combined ........................................................................ 62

ix

Preface

This report argues for greatly increased resources in termsof data collection facilities and staff to collect, process,and analyze the data, and to communicate the results, inorder for NMFS to fulfill its mandate to conserve andmanage marine resources. In fact, the authors of this re-port had great difficulty defining the �ideal� situation towhich fisheries stock assessments and management shouldaspire. One of the primary objectives of fisheries man-agement is to develop sustainable harvest policies thatminimize the risks of overfishing both target species andassociated species. This can be achieved in a wide spec-trum of ways, ranging between the following two extremes.The first is to implement only simple management mea-sures with correspondingly simple assessment demands,which will usually mean setting fishing mortality targetsat relatively low levels in order to reduce the risk of un-knowingly overfishing or driving ecosystems towards un-desirable system states. The second is to expand existingdata collection and analysis programs to provide an ad-equate knowledge base that can support higher fishingmortality targets while still ensuring low risk to target

and associated species and ecosystems. However, defin-ing �adequate� is difficult, especially when scientists havenot even identified all marine species, and informationon catches, abundances, and life histories of many targetspecies, and most associated species, is sparse. Increas-ing calls from the public, stakeholders, and the scientificcommunity to implement ecosystem-based stock assess-ment and management make it even more difficult to de-fine �adequate,� especially when �ecosystem-based man-agement� is itself not well-defined. In attempting to de-scribe the data collection and assessment needs for thelatter, the authors took a pragmatic approach, rather thantrying to estimate the resources required to develop aknowledge base about the fine-scale detailed distributions,abundances, and associations of all marine species. Thus,the specified resource requirements will not meet the ex-pectations of some stakeholders. In addition, the StockAssessment Improvement Plan is designed to be comple-mentary to other related plans, and therefore does notduplicate the resource requirements detailed in those plans,except as otherwise noted.

1

Executive Summary

� The Stock Assessment Improvement Plan is the re-port of the National Marine Fisheries Service (NMFS)National Task Force for Improving Fish Stock Assess-ments, and is a component of the Science Quality Assur-ance Program. The Task Force consisted of one represen-tative from NMFS Headquarters and 1-2 representativesfrom each of the five NMFS Science Centers. The reportalso addresses recommendations made in the NationalResearch Council study on Improving Fish Stock Assess-ments (NRC 1998a).

� Improvements in stock assessments are required forseveral reasons, including: that management entities are�managing at the edge� for many species, and thereforerequire the most accurate and precise stock assessmentspossible; it is no longer permissible to overfish; and thereare currently increased demands for adopting a �precau-tionary approach� and incorporating �ecosystem consid-erations� into stock assessments and fisheries manage-ment. This reports discusses these and other factors thatdefine NMFS� stock assessment mandate.

� Although the NRC study on Improving Fish StockAssessments (NRC 1998a) focused on improving assess-ment methodology, the Task Force agreed that the great-est impediment to producing accurate, precise, and cred-ible stock assessments is the lack of adequate input data, interms of the quantity, quality, and type of data available.

� For most stocks, there is at least basic information onlanded catch and the size frequency of the catch. How-ever, for more than 40% of the 904 stocks listed in the1999 Report to Congress on the Status of Fisheries of theUnited States (NMFS 1999a), there is no fishery-inde-pendent or fishery-dependent index of abundance, whichmakes it extremely difficult to conduct a meaningful assess-ment. Other factors, such as the need to prioritize the stocksto be assessed, result in a total of about 60% of the stocks(545 stocks) lacking assessments sufficient to evaluate stockstatus relative to overfishing. On the other hand, althoughthere are relatively few stocks with comprehensive input data,a total of 119 stocks are routinely assessed using state-of-the-art age or size structured models, some of which mayalso incorporate spatial and oceanographic effects. With afew exceptions, all of the high-valued, high-volume, orhigh-profile species are routinely assessed, while most ofthe unassessed species contribute little or nothing to totallandings.

� Stock assessments conducted by NMFS are rarely, ifever, the product of a single individual, and peer reviewis an integral part of the processes related to provision ofscientific advice in support of fisheries management thatare carried out by fisheries scientists from within and out-side of NMFS. All five Science Centers have systems inplace for peer review of stock assessments.

� The most important programmatic needs vary by re-gion, and even by species groups within regions. Overall,the two most important needs are research vessel surveysdesigned to produce fishery-independent indices of abun-dance and to collect related information on spatial and tem-poral distributions, associated species, habitat, and oceano-graphic variables; and observer programs that provide in-formation on species composition, amounts of each spe-cies kept and discarded, and fishing effort.

� Assessment scientists are faced with many demands.Within a given year, an individual assessment scientist maybe expected to: (i) participate in fishery-independent sur-veys or other field work, (ii) provide input and advice onsampling designs for research surveys and other fishery-independent data collection activities, (iii) spend time oncommercial or recreational fishing vessels, (iv) provideinput and advice on the development of data collectionobjectives and protocols for observer programs and otherfishery-dependent data collection activities, (v) conductquality control or other preprocessing of data, (vi) con-duct stock assessments, (vii) conduct research into stockassessment methods, (viii) present assessment results topeer review panels and constituent groups, (ix) participateon peer review panels, (x) participate in fishery manage-ment plan development or evaluation teams, (xi) defend astock assessment in a court of law, (xii) research and writescientific papers for primary publication, (xiii) attend col-leagues� seminars and offer critical review, (xiv) conductformal, written peer reviews of articles submitted for pub-lication in scientific journals, (xv) participate on commit-tees to advance approaches to stock assessment and fish-eries management, (xvi) undertake training to stay abreastof new methodologies, (xvii) run courses or workshops totrain others, (xviii) participate in national and internationalmeetings and conferences to enhance professional devel-opment, and (xix) undertake a variable amount of admin-istrative duties depending on supervisory level. With lim-ited exceptions, there is insufficient scope for individualscientists to focus on just one or a few of these activitiesdue to an overall shortage of assessment scientists. A sur-

2

vey of assessment scientists indicated that there is insuf-ficient time to devote to important activities such as re-search to improve the basis for assessments, professionaldevelopment, and interactions and cooperative researchwith national and international peers. The same is likelyto be true for individuals involved in data collection, dataprocessing, and data management.

� In fact, staffing needs associated with the productionof stock assessments go well beyond stock assessment sci-entists per se, who represent only the �tip of the iceberg.�Far greater numbers of staff are needed for deployment incritical data collection activities, such as commercial orrecreational catch and effort data, port sampling for bio-logical data, observer programs, and fishery-independentresource surveys. Additional staff are also required to pro-cess biological samples (e.g. to determine fish ages fromhard structures, construct age-length keys, develop growthcurves, construct maturity ogives, and possibly to identifyand count eggs and larval fish from ichthyoplankton sur-veys, and to examine stomach contents), and to enter, au-dit, integrate, and preprocess data from the myriad of datacollection activities.

� The Task Force defined three Tiers of Assessment Ex-cellence, which can be summarized as:

Tier 1 � Improve stock assessments using existing data

(a) for core species, conduct assessments that are morecomprehensive, more thorough, more timely, better quality-controlled, and better communicated;

(b) for species of currently �unknown� status, mine existingdatabases of research vessel survey data and/or commercial

and recreational statistics for archival information for newanalyses to evaluate status determination criteria.

Tier 2 � Elevate stock assessments to new national standards of excellence

(a) upgrade assessments for core species to at least Level3 [the Task Force defined six levels at which assessmentsare conducted, ranging from 0 to 5; Level 3 assessmentscomprise analytical models in which ages or species areaggregated];

(b) conduct adequate baseline monitoring for all federally-managed species (including rare species).

Tier 3 � Next generation assessments

(a) assess all federally-managed species or species groupsat a minimum level of 3, and all core species at a level of4 or 5 [size, age or stage-structured models, possibly in-cluding spatial and seasonal considerations, species asso-ciations, and oceanographic effects];

(b) explicitly incorporate ecosystem considerations suchas multispecies interactions and environmental effects,fisheries oceanography, and spatial and seasonal analyses.

� A large part of the report specifies region-by-regionprogram and staffing requirements needed to meet the threeTiers of Assessment Excellence. These are summarizedin Table 8 of the report, which is reproduced here.

� Among other things, the Task Force recommends thatNMFS should aggressively pursue a course of action fo-cusing on new budget and staffing initiatives to modern-

Table 8. Total Full-Time Equivalents (FTEs) required to meet the three Tiers of Assessment Excellence for each Science Center and allCenters combined. Estimated current FTEs include in-house staff, contractors such as observers, and �other,� which includes state govern-ment biologists, and employees or contractors associated with various regional, national, and international commissions. Numbers should becumulated across tiers.

ytivitcA tnerruCrehtO/tcartnoC/esuoh-nI 1reiT 2reiT reiT

2+1 3reiT llAsreiT

CSFEN 321 94 61 81 34 16 52 68CSFES 17 03 64 41 24 65 93 59CSFWS 08 51 +62 72 06 78 66 351CSFWN 81 33 95 31 47 78 93 621

CSFA 451 221 45 13 66 79 15 841sETFdemmuS 644 942 102 301 582 883 022 806

)K051$xETF($$ K054,51$ K057,24$ K002,85$ K000,33$ K002,19$

3

ize its data collection and assessment capabilities. At theminimum, NMFS should attempt to bring stock assess-ment science to at least Tier 2, and should initiate dialogboth within house and with the public to determine howfar-reaching and comprehensive Tier 3 should be. Thiswill require hiring or contracting considerable numbers ofadditional qualified staff for data collection, data process-ing, data management, stock assessments, and evaluationsof alternative management strategies, to ensure adequatedata and analyses on which to base conservation and man-agement decisions, now and into the future.

� It is also recommended that in order to develop morecomprehensive and integrated future budget initiativesgeared towards modernizing fisheries assessments andmanagement, NMFS should prepare an umbrella plan thatintegrates all relevant existing documents on these themes;for example, the current Stock Assessment ImprovementPlan, the NOAA Fisheries Data Acquisition Plan (Appen-dix 3), the NMFS Strategic Plan for Fisheries Research(NMFS 2001b), the Proposed Implementation of a Fish-

ing Vessel Registration and Fisheries Information Man-agement System (Appendix 8), the NMFS Bycatch Plan(Appendix 9), the National Observer Program (Appen-dix 10), the Social Sciences Plan (Appendix 11), the Ad-vanced Technologies Working Group (Appendix 12), andrelevant fisheries oceanography initiatives (e.g. Appen-dix 13).

� In order to make substantial progress towards collect-ing the data needed to improve stock assessments, par-ticularly next generation assessments, it is essential thatNMFS continue to foster partnerships and cooperative re-search programs with other federal agencies, state agen-cies, private foundations, universities, commercial and rec-reational fishing organizations and individuals, environ-mental groups, and others with a vested interest in collect-ing similar types of data, although often for different pur-poses. Programs involving cooperative research with thefishing industry should continue to be developed and ex-panded as mechanisms for providing data relevant to im-proving the quality of stock assessments.

5

I. Introduction

The Stock Assessment Improvement Plan (SAIP) is thereport of the National Marine Fisheries Service (NMFS)National Task Force for Improving Fish Stock Assess-ments. It consists of a main document with 8 tables and15 figures, an Appendix table summarizing the level ofcompleteness of data collection and stock assessments foreach federally managed stock (Appendix 1), and an ad-ditional 24 appendices (Appendices 2-25) summarizingother relevant programs, plans and reviews. The StockAssessment Improvement Plan is a component of the Sci-ence Quality Assurance Program (Appendix 2), which con-sists of several other elements including the NOAA Fish-eries Data Acquisition Plan (Appendix 3), the Stock As-sessment Toolbox (Appendix 4), the Center for Indepen-dent Experts (Appendix 5), and the NMFS-Sea Grant JointGraduate Fellowship Program (Appendix 6). The Task Forceconsisted of one representative from NMFS Headquartersand 1-2 representatives from each of the five NMFS Sci-ence Centers. Additional input from the Science Centerswas obtained via questionnaires administered to stock as-sessment scientists and managers of stock assessment pro-grams. Science Centers were also given the opportunity toreview the contents of the Plan.

This report also draws on the analyses and recommenda-tions of the National Research Council (NRC) study onImproving Fish Stock Assessments (NRC 1998a). In orderto determine which avenues should be explored to improvestock assessments, NMFS requested in 1995 that the NRCundertake a review of the agency�s current stock assess-ment methods and models and make recommendations foralternative approaches. The objective of the review was toproduce an authoritative report that documented thestrengths and limitations of stock assessment methodsrelative to the diversity of available data and types of fish-eries management systems. The resulting review (Appen-dix 7) contained ten recommendations in seven categoriesfor improving stock assessments; these are numbered andpresented in summary form below for easy referencethrough the remainder of the current report.

Recommendation #1: How should assessments be con-ducted and by whom?

Recommendation #2: Development of at least one reli-able abundance index for each stock.

Recommendation #3: Collection of auxiliary biologicaldata such as natural mortality.

Recommendation #4: More realistic assessments ofuncertainty.

Recommendation #5: Analysis of alternative harveststrategies.

Recommendation #6: Development of rigorous evalua-tion systems including simulation models.

Recommendation #7: Development of new techniques forstock assessment.

Recommendation #8: More peer reviews.Recommendation #9: Standardized data collection pro-

tocols for commercial fisheries.Recommendation #10: Education and training of assess-

ment scientists.

Improvements in stock assessments are required for sev-eral reasons, including (a) that management entities are�managing at the edge� for many species, and thereforerequire the most accurate and precise stock assessmentspossible; (b) it is no longer permissible to overfish; and(c) there are currently increased demands for adopting a�precautionary approach� and incorporating �ecosystemconsiderations� into stock assessments and fisheries man-agement. These issues are addressed in detail in SectionII, along with other factors that define NMFS� stock as-sessment mandate. Section III provides background in-formation on requirements for conducting assessments andfor evaluating alternative fisheries management strate-gies. Section IV contains detailed information on quali-tative and quantitative resource requirements for each ofthe five Science Centers, relative to three Tiers of Assess-ment Excellence. As such, it represents the core part ofthe report. Recommendations based on the preceding sec-tions of the report are summarized in Section V. Assess-ment-related information is tabulated in Appendix 1, andan additional 24 Appendices contain information on otherrelevant plans, reports, and background documents. It isanticipated that the current report will provide a foun-dation for future initiatives, including budget initia-tives, to improve the quality and quantity of NMFS�stock assessments.

Scope of the Stock AssessmentImprovement Plan

The Stock Assessment Improvement Plan (SAIP) is thelatest of a series of plans for enhancing and modernizingNMFS programs for data collection, information technol-ogy, data management, stock assessments, scientific re-search, and fisheries management. Although the SAIP isspecifically geared towards stock assessments, when ac-

6

count is taken of the diverse data needs of stock assess-ment models and the expectation of more comprehensiveecosystem-based science and management in the future,there is considerable scope for overlap or duplication be-tween several plans. With this in mind, the SAIP wasdesigned to complement plans already completed at itsinception, or otherwise to acknowledge duplication, andto avoid contradiction. The key complementary plan isthe NOAA Fisheries Data Acquisition Plan (Appendix 3),which details the need for purpose-built fishery researchvessels and chartered days-at-sea to satisfy immediate fish-ery-independent data collection needs. Thus, the SAIP doesnot include the capital and operating costs of the researchvessels, nor chartering costs, nor the permanent vesselcrews, but it does include the scientific staff that wouldparticipate in research surveys. Another important comple-mentary plan is the Proposed Implementation of a FishingVessel Registration and Fisheries Information ManagementSystem (Appendix 8), which will provide much more accu-rate, complete, and timely information on commercial fish-eries statistics. The core costs of this program are not con-sidered in the SAIP, but the in-house or contract staff re-quired to collect, manage, and process the data are included.The NMFS Bycatch Plan (Appendix 9), which includesmonitoring, data collection, and research, overlaps to adegree with the SAIP but, since the Bycatch Plan did notdevelop specific staff and other resource requirements,duplication should not be a problem.

One plan with which there is considerable potential foroverlap and duplication is the National Observer Program(Appendix 10), which was initiated subsequent to the com-mencement of the SAIP. As is evident in Section IV ofthis report, the most important overall need for improvingstock assessments is for in-house and contract staff forobserver programs for collecting data of relevance to stockassessments, in both the short- and the long-term. TheNational Observer Program examines observer needs froma slightly different perspective. First, it considers needsfor monitoring both commercially-exploited fish speciesand protected and endangered species such as marine mam-mals and sea turtles, whereas the SAIP only considers that

portion of existing and potential observer programs thatcould be attributed to obtaining data of direct relevance tocommercially-exploited fisheries stock assessments. Sec-ond, it outlines a five-year plan, whereas the SAIP is muchlonger-term. Thus, there is some degree of overlap be-tween the SAIP and the National Observer Program. Theoverlap will be quantified and controlled for as necessaryin future budget initiatives.

The SAIP focuses on field biologists who collect data;laboratory technicians who process biological samples;computer scientists who audit, manage, and analyze data;and quantitative stock assessment scientists who developand run stock assessment models. Another important fish-eries profession not represented in the SAIP is that ofsocial science (economists, sociologists, and anthropolo-gists). The need for additional social scientists is detailedin the NMFS Social Sciences Plan and Budget Initiative(Appendix 11).

Capital costs for the purchase of advanced technologiesand operating costs for research and field trials is anotheractivity that is covered elsewhere (Appendix 12), and there-fore excluded from explicit consideration in the SAIP. Alsoexcluded are major infrastructure associated with increasedstaffing, particularly new workspace and buildings that maybe required.

In order to further limit the scope of the SAIP and to re-duce overlap with other plans and initiatives, it was alsodecided not to explicitly include resource requirements forfisheries oceanography (e.g. Appendix 13), stock assess-ments and related activities for marine mammals and seaturtles (Appendix 14), habitat-related research and con-servation (Appendix 15), and stock assessments and re-lated activities for Pacific salmon.

In order to develop a comprehensive ecosystem approach tofisheries stock assessments and management, and to esti-mate the actual costs of implementing ecosystem-based man-agement (EBM), all of the above-mentioned plans, initiatives,and activities should be merged into an umbrella plan.

7

II. Defining NMFS� Stock Assessment Mandate

The central importance of stock assessments to NMFS isclear. The NOAA Fisheries Strategic Plan (NMFS 1997a)describes the agency�s mission as:

�stewardship of living marine resources for thebenefit of the Nation through their science-basedconservation and management and promotion ofthe health of their environment�

That document then outlines five �foundations for stew-ardship,� the first of which is: �Science, which is of thehighest quality, and which advances our ability to makeliving marine resource management decisions.� The goalsand objectives of the Strategic Plan are reiterated and ex-panded in the NMFS Strategic Plan for Fisheries Research(1998c, 2001b). Those which are at least partially ad-dressed by the current plan are reproduced below. To fa-cilitate cross reference, the corresponding Fisheries Stra-tegic Plan (FSP) strategy or foundation number followseach fishery research objective.

GOAL 1: Provide scientifically sound information anddata to support fishery conservation and management.(Ongoing)

Objective 1.1: Periodically assess stocks to ascertainwhether changes in their status due to natural or hu-man-related causes have occurred. These stock assess-ments require adequate fishery monitoring and resourcesurveys. (FSP Strategy 1.1.1)

Objective 1.2: Use stock assessments to predict futuretrends in stock status. Forecasts will take into accountprojected biological productivity, climatic information,economic markets, and other social forces that willaffect levels of fishing effort. (FSP Strategy 1.1.2)

Objective 1.3: Determine and reduce the level of un-certainty associated with stock assessments through im-proved data collection and advanced analytical tech-niques. (FSP Strategy 1.2.1)

Objective 1.4: Use stock assessment workshops, peerreviews, and other fora to ensure that our informationand advice are developed through an open and col-laborative process. (FSP Strategy 1.2.2)

Objective 1.5: Communicate our scientific informa-tion and advice, along with the associated uncertain-

ties, to the Councils, other management authorities,and the public. (FSP Strategy 1.1.3)

Objective 1.6: Collaborate with the Councils and othermanagement authorities to explore and develop fish-ery management regimes and alternative governancesystems that will effectively control exploitation andpromote sustainability. (FSP Strategy 1.1.4)

Objective 1.7: Provide guidelines to assist the Coun-cils in assessing and specifying maximum sustainableyield (MSY) for managed fisheries. (FSP Strategy 1.1.5)

Objective 1.8: Work with the Councils to develop ob-jective and measurable criteria for each managed stockto determine if the stock is overfished or approachingan overfished condition. (FSP Strategy 2.1.1)

Objective 1.9: For each stock which is overfished orapproaching an overfished condition, we will develop,in collaboration with the Councils, measures to elimi-nate or prevent the overfishing. (FSP Strategy 2.1.2)

Objective 1.12: Support recommendations providedby the National Research Council [NRC 1999] and theReport to Congress [NMFS 1999b] by establishing cri-teria to define and delineate marine, estuarine, and river-ine ecosystems for management purposes, and identifyindicators for assessing the status and detectingchanges in the health of such ecosystems. (FSP Strat-egy 7.3.2)

Objective 1.14: Incorporate assessments or indices ofclimate variability into stock assessments.

Objective 1.15: Monitor climate change on inter-an-nual, decadal, and centennial scales and its impact oncurrently sustainable fisheries.

GOAL 5: Improve the effectiveness of external part-nerships with fishers, managers, scientists, conserva-tionists, and other interested groups. (Ongoing)

Objective 5.1: Promote a cooperative network of part-ners in the coordination of fisheries research.

Objective 5.2: Develop infrastructure for long-term, con-tinuous working relationships with partners to addressfisheries research issues.

8

As reported in the first four annual Reports to Congresson the Status of Fisheries of the United States (NMFS1997b, 1998b, 1999a, 2001a), the status relative to over-fishing of the majority of the fish stocks covered by fed-eral Fisheries Management Plans (FMPs) is unknown. Incontrast to the first three reports, the NMFS (2001a) re-port broke stocks out into �major� or �minor� categoriesbased on landings. Thus, even though the Stock Assess-ment Improvement Plan is mainly based on informationup to January 2000, the following statements are based onthe NMFS (2001a) report, which tabulated information on905 stocks (as compared to a slightly different mix of 904stocks in the 1999 report). In the 2001 report, 623 of the905 stocks were recorded as having unknown status. Al-though it is often overlooked, most of the 905 stocks tabu-lated can be classified as �minor� stocks: 618/905 or 68.3%have recent landings less than 90.74 metric tons (200,000pounds) annually. In total, �minor� stocks have accountedfor only about 0.11% of total landings in recent years.However, it should be noted that �minor� stocks are oftennot landed or identified to species, and discarded catchesmay not be recorded, particularly where observer programsare lacking. Whether or not actual removals constitute arisk to the long-term viability of these species is unknown.Of the 287 �major� stocks, 35.2% are of unknown or un-defined status relative to threshold fishing mortality lev-els that define �overfishing,� while 41.8% are of unknownor undefined status relative to threshold stock sizes thatdefine whether a stock is �overfished.� While the costs ofdetermining the status of all 623 stocks in the unknowncategory may be prohibitive, additional efforts to obtainthe information necessary to assess the major stocks withunknown status is certainly warranted. In addition, ma-jor stocks of �known� status also require special and vigi-lant attention because many of them (25.3%) are experi-encing overfishing and many (36.5%) are overfished orapproaching an overfished condition. The need to elevatethe level of knowledge of many of the unknown species,even those of �minor� importance, will escalate as fisher-ies management progresses towards ecosystem-basedmanagement (EBM). Thus, there is a need to constantlyimprove both the quality and quantity of stock assessments.

It is also important to keep in mind that NMFS� mandateis actually a dual one of both sustainability and exploita-tion, which can often create conflict. The fact that ourscience is used for regulation means that stock assessmentswill often be challenged. Thus, as stated in the NMFSStrategic Plan for Fisheries Research, �all of the agency�sinformation must be comprehensive, objective, credible,and effectively communicated.�

A. What is a �Stock Assessment?�

The term �stock assessment� is used to describe the pro-cesses of collecting, analyzing, and reporting demographicinformation for the purpose of determining the effects offishing on fish populations. The production of stock as-sessments requires quantitative information on the rela-tive or absolute magnitude of a fish population, estimatesof the total removals due to human activities (due to fish-ery landings, discarded bycatch, and cryptic mortality dueto encounters with fishing gear), life history data includ-ing rates of growth, average age of the onset of sexualmaturity, maximum longevity, and the proportion of eachage group dying each year due to natural causes, and otherfactors that affect stock productivity. These data are com-bined using simple or complex mathematical models (NRC1998a) to derive �best� estimates of vital statistics such ashistorical and recent trends in the number and biomass ofthe resource, recruitment levels (number of small fish en-tering the fishery each year), and the fishing mortality rateor the fraction of the stock alive at the beginning of theyear that are killed by fishing (commonly referred to asthe exploitation rate).

The results of stock assessment calculations provide in-formation necessary to estimate the current abundance andexploitation rates of resources in relation to predefinedgoals for these two attributes, also termed �status determi-nation criteria.� If the biomass is determined to be signifi-cantly below a minimum threshold, the stock is in an �over-fished condition.� If the current exploitation rate is signifi-cantly higher than a maximum exploitation rate threshold,overfishing is deemed to be occurring. The Magnuson-Stevens Fishery Conservation and Management Act(MSFCMA) mandates that if stocks are in an overfishedcondition or if overfishing is occurring, managers mustimplement measures to rebuild the stock and/or to elimi-nate overfishing. In addition, assessment results providethe technical basis for setting the level of biologically ac-ceptable yield for healthy stocks, and the expected rate ofrebuilding for depleted stocks.

The stock assessment evaluation is thus a key element ofthe fishery management process since it is used to deter-mine whether additional regulations are necessary, or ifgreater fishing opportunities can be allowed consistent withthe objective of sustainable fishing. In addition to enablingdetermination of overfishing, stock assessment results havethree additional important purposes: (1) for each managedstock, a history of estimates of catches, stock sizes andexploitation rates is used to establish reference levels forthe two status determination criteria; (2) assessment re-sults (biomass, recruitment levels and exploitation rates)are combined to provide short- (1-2 year), medium- (3-10

9

year) and sometimes long-term (10+ year) projections ofhow fish populations and catches will change over time;and (3) stock assessment results are typically combinedwith other research results or assumptions to evaluate vari-ous alternative sets of management measures proposed bymanagers to attain specified objectives. The latter analy-ses are termed �alternative management strategy analy-ses� and are important because there are often many dif-ferent types of measures that can be used to manage fish-eries (e.g. regulating the characteristics of the gear such asminimum mesh size, or regulating the amount of effort byfleet sector, area, or season). The selection of the �pre-ferred� set of management measures is often complex andrequires quantitative evaluation since even if one or sev-eral options allow a stock to rebuild to optimal levels, theymay have dramatically different implications for the prof-itability of fisheries and the allocation of benefits amongcompeting harvest sectors. Quantitative stock assessmentsprovide the essential information necessary for the analy-ses required in the fishery management process.

The quality of a particular stock assessment (i.e. the accu-racy and precision of stock size and exploitation rate esti-mates) is directly related to the quality and completenessof the input data used for the assessment. No stock as-sessment is perfect because the information used is de-rived from a modest number of observations that are as-sumed to be statistically representative of the populationas a whole. Elsewhere in this document (Sections IIA,IIB), information on the quality of data on catches, abun-dance measures and life history data supporting stock as-sessments for all managed or assessed fishery stocks oc-curring in EEZ waters is provided. In most cases, the so-phistication of the stock assessment model used to esti-mate stock size and the accuracy and precision of the re-sults is directly related to the quality of the assessmentdata. It cannot be overemphasized that improving the qual-ity of fish stock assessments (consistent with the focus ofthe NRC report) primarily involves improving the qualityof basic input data on catches, abundance and life history,and that these improvements will lead to a progressionthrough Tiers of Assessment Excellence.

The goal of improving fish stock assessments is to allowsociety to extract maximum benefits from fish stocks whileminimizing the risk that stocks will become overfished.As assessments are improved, the types of questions posedby managers will increasingly emphasize multispecies as-pects (technological and biological interactions amongstocks and fisheries), and will require greater temporal andspatial detail to evaluate fine-scale time/area componentsof management measures. Thus, the requirements for thenext generation of fish stock assessments will necessitatecontinued improvements to data and refinements to models.

In addition to requiring fishery data and selection of ap-propriate mathematical models, the process of producingstock assessment, as practiced by NMFS, involves explicitand intensive QC/QA through a process known as �peerreview.� Owing to the implications of stock assessmentresults for the ecosystem and the economy, the public mustbe assured that data and procedures used by NMFS and itscooperators meet accepted standards for the production ofsuch analyses. Assessments undertaken by all of the NMFSScience Centers include an element of peer review, whichinvolves review of data and calculations by experts inde-pendent of the people responsible for the work being re-viewed. In many cases peer reviews have involved aca-demic researchers, inter-Center exchanges of personnel,experts hired by fishery industry groups, and internationalscientists. An increasing and serious impediment to theimprovement of stock assessments nationwide is the diffi-culty of providing for adequate peer review under the bur-den of increasing numbers of stock assessments with higherlevels of complexity being produced more frequently, incombination with a relatively small pool of experts in thisspecialized area.

B. The Quality of NMFS� Assessments

The stock assessment activities within the National Ma-rine Fisheries Service have produced strengths and accom-plishments that are globally recognized. The strength ofthe national assessment activities lies in the developmentof periodic stock-wide status determinations for majorspecies; i.e. those species which are economically mostimportant and which comprise the majority of the fisher-ies biomass. The accomplishments of NMFS scientists inthis regard compare favorably with any national effortworldwide. These assessments have allowed these impor-tant stocks to be monitored effectively. The precision andaccuracy of these assessments has proven to be extremelyhelpful in management. These results have also formedthe basis for much of the understanding of fishery popu-lation dynamics and the historical trends of these species.

However, NMFS assessment efforts have been less effec-tive in several other areas. Specifically, there are manyspecies that are not assessed even though in many instancessome relevant data may exist. While these species are noteconomically dominant and do not comprise a high pro-portion of the biomass, they often interact ecologically withthe economically important species and they may be sig-nificant keystones in the functioning of the ecosystem. Atthe present time we have little understanding of the roleof these species either in the ecosystem or in local econo-mies. There has been a lack of resources to obtain suffi-cient data to evaluate these species.

10

Another limitation to present assessment efforts is the un-derstanding of the spatial and temporal dynamics of thespecies. Under what conditions do fish move into alter-native areas of their range and what are the stimuli? Thesequestions have become important as managers attempt toprovide for the diversity of users of the resources.

C. Changing Demands

The demand for stock assessments has shifted both quan-titatively and qualitatively throughout the Nation. Fish-eries have expanded to target heretofore lightly exploitedresources, as traditional stocks have been fully exploitedor over exploited, leading to needs for increased numbersof stocks to be assessed. More detailed and complex regu-latory mechanisms are being devised to distribute the lim-ited resources equitably between fishing sectors, commer-cial, recreational, and bycatch users. In order to evaluatethese alternative mechanisms, spatial and temporal pro-jections of management scenarios are required, taxing thelimits of the available data and the number of assessmentscientists. The MSFCMA has imposed new managementrequirements that have increased both the detail and thenumber of assessments that are to be conducted. Thesegeneral changes have manifested themselves in theNation�s fisheries in a variety of ways which are describedbelow for each of the NMFS Science Centers.

Northeast Fisheries Science Center

The Northeast Fisheries Science Center (NEFSC) typicallyproduces stock assessments, in one form or another, forabout 51 managed species/stocks (Appendix 1). Not allof these stocks are managed under Federal FMPs (e.g. someare managed by the Atlantic States Marine Fisheries Com-mission, ASMFC, or individual states, or under interna-tional agreements). These assessments run the gamut fromindex-level assessments using trawl survey information,to stage-based analytical assessments incorporating mul-tiple abundance indices and catches. During the past 15years, stock assessments have been peer-reviewed underthe jurisdiction of the Northeast Stock Assessment Work-shop (SAW) process. Typically, the SAW has reviewedabout dozen high priority stock assessments per year, sixeach in spring and autumn SAW meetings.

The demand for more timely and comprehensive stockassessments has increased greatly, primarily as a result ofthe need to respond to information needs associated withthe MSFCMA. The exploitation history of most of the eco-nomically-important stocks of the region has necessitatedstringent rebuilding plans, and managers are requestingmore frequent assessment updates on a wider array of spe-cies, and are considering more complex types of indirect

controls on fishing mortality, including manipulating thetemporal and spatial patterns of fishing.

A major new demand on the assessment process is to in-crease the precision of estimates of exploitation rates andstock sizes. For many resources, fishing rates historicallyhave greatly exceeded standard biological reference pointsand stock sizes were considerably below those consideredoptimal for sustainable fisheries. Thus, even impreciseassessments were considered useful enough to give clearadvice about the direction of fishery management andrough indications of the magnitude of fishing effort re-ductions required (e.g. in many cases fishing mortality rateswere five times the target levels, and stock biomasses were1/5th of the biomasses generating MSY). Managementprograms instituted in the mid-1990s have resulted in re-duced harvest rates for species such as haddock, yellow-tail flounder, sea scallop, and other economically-impor-tant resources. As a result of these changes, managersrequire more precise information to determine the prob-abilities of attaining fishing mortality rate targets associ-ated with the harvest control rules they have adopted. Ad-ditionally, because so many of the region�s resources weredetermined to be overfished, 10-year rebuilding programsare now in effect. Consequently, management requiresmedium-term projections to determine which exploitationstrategies will allow attainment of biomass targets with aspecific probability.

In addition to the increased demand for precision of analy-ses, the requirement for estimates of biomasses and fish-ing mortality rates or proxies associated with MSY for allmanaged species necessitates that the level of many stockassessments be increased from monitoring of indices ofabundance to greater levels of analytic complexity. Thismeans that the demands for biological sampling of fisherycatches and abundance indices will increase (e.g. moreroutine age determinations for managed stocks). Likewise,managers require more specific information on all com-ponents of fishery catches, and especially fishery discards,which have historically not been sampled adequately inthe Northeast fisheries. Thus, in addition to increased needsfor more stocks to be assessed, increased quality of assess-ments means significantly augmented data collections andbiological information to support improved assessments.

The frequency of stock assessment updates has increasedprimarily as a result of the many changes recently incor-porated in many of the region�s FMPs. Because of reduc-tions in fishing effort, the closure of large areas of produc-tive fishing grounds, trip limits and other measures, man-agers want to know the incremental effect of these mea-sures on attaining required fishing mortality and biomasstargets. Thus, for example, the New England Fishery Man-

11

agement Council has requested detailed assessment up-dates on all managed stocks well in advance of each newfishing year, so as to determine the suite of new measuresnecessary to meet the requirements of the law. This is instrong contrast to recent history when a small group ofimportant species (e.g. cod, haddock, yellowtail flounder)were used as �bellwether� indicators of the exploitation ofthe status of a complex of about 25 managed stocks. Addi-tionally, the MSFCMA requires that stocks that were his-torically �written-off� as virtually commercially extinct,be rehabilitated towards sustainable stock levels (e.g. At-lantic halibut, redfish). Thus, new stock assessments arerequired to determine the feasibility and impacts of effortsto do so. Even stocks for which no commercial uses existare subject to increased demand for assessments when theirstatus may be impacted as a result of fishery bycatch (e.g.barndoor skates).

Management advice based on analytic stock assessmentsis also increasingly required to support complex measuresaccounting for technological interactions among theregion�s stocks (which are generally significant) and theincreased demand for finer spatial and temporal scale in-formation supporting area rotation strategies, and othercomplex management approaches. Supporting these scalesof management will require improvements in basic infor-mation collected from fishers (e.g. logbook-type data andobserver data) and dealers. There is an important newdemand for integration of single species assessment infor-mation to support assemblage management (stock trade-offs). Managers are also increasingly concerned with thetrophic implications of attempting to increase all managedstocks to BMSY simultaneously, necessitating more researchon biological interactions.

Southeast Fisheries Science Center

The fisheries under the research jurisdiction of the South-east Fisheries Science Center (SEFSC) are diverse in boththe species being exploited and the fishing sectors pros-ecuting these fisheries. The fisheries include a large num-ber of snapper-grouper fisheries, mackerel fisheries, croak-ers, shrimp and other invertebrates, sharks, and Atlantic-wide tunas and billfish. Characteristically, these fisheriesinvolve large recreational sectors along with the usual com-mercial sector (in some instances the recreational sectortakes the majority of the catch). Additionally, there is alarge bycatch sector; i.e. bycatch of commercially andrecreationally important species occurs in large numbersin some fisheries. Interactions between these sectors havemanifested themselves in numerous allocation conflicts.

Another characteristic of southeastern U.S. fisheries is thatthe productivity of many of the species being exploited is

low, supporting relatively small catches (there are over 400species within Southeast FMPs or international conven-tions). However, some of the species are extremely valu-able and many are very important to local communities.Also, in aggregate the species catches are significant andthe fisheries often have the capability to exploit a varietyof species, switching target species as conditions change.These characteristics create unique dynamics which af-fect the research and management of these resources.

The Southeast Fisheries Science Center has conductedassessments for the most important stocks for the last twodecades. These assessments have been reviewed throughStock Assessment Panels of the Councils and through in-ternational scientific working groups within the Interna-tional Commission for the Conservation of Atlantic Tu-nas. Through these scientific groups the assessment in-formation is integrated into the scientific advice on Al-lowable Biological Catch (ABC).

However, there are a number of recent events that havealtered the assessment landscape in the Southeast. Firstand foremost, several important stocks have been severelydepleted leading to very restrictive limitations on catch.As the stocks begin to recover, more detailed regulationssuch as trip limits, area closures, minimum sizes and baglimits are being devised to distribute quotas more equita-bly. Evaluation of these alternatives require an order ofmagnitude greater spatial and temporal details. The evalu-ation activities require an increasingly larger proportionof the assessment scientists� time. Additionally, as fish-ers look for new alternatives to replace depleted stocks,they have begun to target stocks that traditionally havenot been exploited. In many cases, detailed assessmentshave not been conducted for these stocks but as fishingpressure has increased, there has been an increased needto evaluate their status. The ability to do this is limited byboth the available data and the human analytical resources.

Finally, the MSFCMA has shifted the focus of manage-ment from limiting fishing mortality rates under the 602Guidelines to limiting both the maximum allowable fish-ing mortality rate and the minimum allowable stock abun-dance. Scientifically, it is easier to estimate the referencefishing rates than to determine appropriate reference abun-dance criteria, especially when data are limited and thespecies are numerous. The MSFCMA requirements haveshifted the focus to determining abundance criteria. Thisrequires examination of each stock individually. It is un-clear whether the Southeast Fisheries Science Center hasthe data or human resources available to address theseissues for all 400+ species in the southeast and Caribbean.

12

Southwest Fisheries Science Center

On the Pacific coast and in Hawaii demands on the South-west Fisheries Science Center (SWFSC) stock assessmentscientists have been increasing significantly due to newFMPs, increasing lawsuits and other pressure from indus-try and environmental groups, and new demands to man-age international fisheries for which little data are avail-able. At the same time, SWFSC resources for this researchhave diminished and little funding beyond basic salariesis available.

New FMPs are in place or being developed for coastalpelagic species and highly migratory species, and increasedattention is being demanded for stock assessments of squid,marlins, swordfish, albacore, groundfish, sharks, marlinsand tropical tunas. Environmental groups are pressingfor increased research on sharks, monk seals, turtles androckfish. Marine mammal stock assessments are partiallycompleted for the 38 west-coast species but have not beenstarted for Hawaiian or other U.S. Pacific waters. A newfishery has developed for near shore rockfish; coral fish-eries are resuming; and California salmon issues are highlycontroversial.

A new international agreement on western Pacific HighlyMigratory Species is being developed that will demand ahuge and complex stock assessment effort by the U.S. incollaboration with other fishing nations. Also, new legis-lation requires that research be strengthened on the majorinternational fisheries in which U.S. fishers participate,harvesting swordfish, tropical tunas, albacore, plus inciden-tals such as marlins, mahimahi, opah, wahoo, and others.

Northwest Fisheries Science Center

The Northwest Fisheries Science Center (NWFSC) en-gages in assessments of west coast groundfish and salmon.The demands for accurate assessments for both groups ofspecies are high and increasing. For groundfish, only 26of the 82 species have been quantitatively assessed. Ofthese 26 species, several have experienced severe declinesbecause of overly optimistic historical harvest rates dur-ing a 20-year regime of poor ocean productivity with in-adequate stock assessment information to adequatelymonitor and forecast the declines. As of 2001, rebuildingplans are being developed for seven groundfish speciesthat have declines below the overfished threshold, andthere are concerns that others of the 60+ species with un-known status may also be in danger. For salmon, the statusof Pacific salmon species on the west coast has been re-viewed under provisions of the Endangered Species Actand 26 of the populations (Evolutionarily Significant Units)have been listed as threatened or endangered. A tremen-

dous effort is being mounted by the NWFSC to developsalmon recovery plans that incorporate all aspects of hu-man and natural risks to salmon.

Groundfish and salmon are managed according to FisheryManagement Plans developed by the Pacific Fishery Man-agement Council. Although the NWFSC has the lead rolein coordinating assessment information for both FMPs,there are major contributions by other NMFS Science Cen-ters and by the state fishery agencies of California, Or-egon and Washington. For salmon, nearly all of the es-capement monitoring and run forecasting is based oninriver information and is done by the state agencies. Thisinformation is used by the Salmon Technical Team of thePFMC to develop harvest options for consideration by thePFMC. The Scientific and Statistical Committee of thePFMC provides reviews of methodology for this work.For groundfish, shoreside catch monitoring is done by thestate agencies with coordination through the Pacific StatesMarine Fisheries Commission (PSMFC) which maintainsa centralized database of fisheries data (PacFIN). In 2001,a coastwide observer program was implemented by NMFSin collaboration with PSMFC and the states. Most resourcesurveys are conducted by NMFS, with the triennial bot-tom trawl and hydroacoustic surveys providing a majorsource of data for most assessments. Approximately sixgroundfish stock assessments are conducted each year byNMFS, state agencies and others. The NWFSC coordi-nates a stock assessment review process in conjunctionwith the PFMC�s SSC, that involves external peer-review-ers and public input. These groundfish assessments havebeen controversial. The west coast groundfish industryseeks an increasing role in gathering of relevant stock as-sessment information, and in participating in the stock as-sessment process.

Passage of the Sustainable Fisheries Act strengthens themandate to improve the west coast stock assessment ca-pability. Assessments need to be conducted for more ofthe groundfish species. The level of uncertainty in ground-fish assessments and the current information indicating lowproductivity for these species needs to be combined in anadequately precautionary approach to managing these spe-cies. Rebuilding plans, which are expected to have timehorizons longer than 10 years, need to be developed andsubsequently monitored for several long-lived species. Allof these tasks will be extraordinarily difficult given the lackof a dedicated research vessel for surveying these resourcesand the low level of current resource survey efforts. Fur-ther, increased stock assessment effort will primarily resultin increased knowledge about what changes the fish popu-lations have undergone, but knowing what is only the firststep. In order to develop a better understanding of whychanges are occurring, programs need to be developed to

13

investigate the role of decadal scale changes in ocean climate,and the role of ecosystem shifts such as the major increase inpinniped abundance that has occurred off the west coast.

Alaska Fisheries Science Center

The Alaska Fisheries Science Center (AFSC) stock assess-ment staff have experienced increased demands stemmingprimarily from requests for information regarding the im-pacts of fishing on marine ecosystems. These requestsrequire added activities including (1) development of newassessments for minor species, (2) development of modelsthat address predator-prey interactions, and (3) develop-ment of models that incorporate environmental forcing onstock production.

The AFSC produces stock assessments for fish and shell-fish in the Eastern Bering Sea, Aleutians and Gulf of Alaska(Appendix 1). These stocks are managed under FederalFisheries Management Plans (FMP) or under internationalagreements. The North Pacific Fisheries ManagementCouncil (NPFMC) FMP covers 100 species/stocks in theGOA and 127 species/stocks in the Bering Sea AleutianIslands. In some circumstances (e.g. Pacific salmon andcrab), fish stocks are assessed jointly between federal andstate fishery scientists. The NPFMC has a long traditionof conservative management of Alaskan fishery resourcesthat includes bycatch limits to protect weak stocks or pro-hibited species.

The infrastructure supporting this type of managementconsists of complex fisheries-dependent and fishery-inde-pendent data collection programs as well as detailed as-sessments of core species. The NPFMC reviews stock as-sessments for groundfish and crab stocks or stock com-plexes on an annual basis. Stock assessments are peer-reviewed by the North Pacific Fisheries ManagementCouncil�s (NPFMC) Plan Team and Scientific and Statis-tical Committee. In addition, AFSC solicits peer reviewsof selected species by outside stock assessment experts.These review panels evaluate all aspects of the assess-ment from data collection to model formulations.

Passage of the MSFCMA in 1996 increased the need forcomprehensive assessment of marine resources in Alas-kan waters. The MSFCMA mandated adoption of over-fishing definitions for exploited marine resources managedunder Federal FMPs, and an evaluation of impacts of hu-mans on essential fish habitat. These two mandates re-quire increased data collection, data analysis, and impactreview. Efforts are currently underway to increase the fre-quency and regional coverage of bottom trawl and EchoIntegration Trawl surveys. For some species, new surveymethods and fishery-dependent data collection programs

are necessary to develop indices of stock abundance andcatch. Demands for data processing, data analysis andstock assessment modeling occur as these new sources offishery-dependent or fishery-independent data becomeavailable.

There is a growing need for the development and imple-mentation of complex assessment models. Recently, sta-tistical assessment models (SAMs) have been modified toconsider uncertainty stemming from process and measure-ment errors. SAMs have also been developed to explorethe impact of temporal trends in predator abundance.These models are used to explore assumptions regardingpredator satiation, natural mortality, and predator selec-tivity on uncertainty in estimating biomass.

In recent years, NMFS has experienced a number of chal-lenges to regulatory decisions based on AFSC assessments.These challenges require assessment scientists to evaluatea number of alternative harvest strategies. For example,assessment scientists have been asked to develop ecosys-tem based harvest strategies that encompass impacts offishing on the structure and function of marine commu-nities. Recent declines in sea bird and marine mammalpopulations that share exploited resources with commer-cial fishers have triggered interest in designing harveststrategies for minimizing impacts of fishing on protectedresources. These demands require analysis of marinemammal-fisheries interactions on finer spatial and tem-poral scales, and they require review of more complexmanagement approaches.

D. The Credibility of NMFS� Science

Assessment activities form the apex of the scientific sup-port of management (see Section III, I). The assessmentprocess integrates a wide array of scientific informationand the results are directly communicated to managers andconstituents by the assessment scientists. These activi-ties are required to fulfill the Agency�s dual role of main-taining conservation stewardship responsibilities and thepromotion of optimum usage of resources. This dual roleresults in a natural tension between scientists and con-stituents. The scientific results are viewed as a constrainton allowable catch and allocation decisions. Indeed, in theshort term they are. This has prompted several externalreviews over the years; for example, the NRC Review ofNortheast Fishery Stock Assessments (NRC 1998b; Ap-pendix 16).

Criticism of the science occurs for a variety of reasons. Attimes, valid concerns are raised. However, much criticismstems from other causes, such as lack of understanding ofscience�s role in the decision process by constituents, lack

14

of understanding of scientific methods and issues by con-stituents, poor communication of these methods and is-sues by the scientists, poor communication of the issuesand methods by the managers to constituents, limited dataresources to support the science, limited resources to sup-port communication efforts, deliberate use of criticism ofthe science as a strategy or negotiation tactic to alter orhinder implementation of unfavorable management deci-sions, and the perception that scientists have a limitedand unrealistic view of the status of fisheries resources.

Credibility with fishing constituents often is related to thedata supporting the assessments: constituents argue thatthe data are of poor quality, that they are unrepresentativeof their direct observations, and that the assessment scien-tists are not cognizant of these features. Hence, the con-stituents view the scientific opinions as unrealistic. Thisis a source of criticism against which scientists of NMFSmust continually be on guard. New scientific perspectiveson old problems must continually be integrated into theassessments. Mechanisms to do this are through regular,periodic peer-review and scientific program review pro-cesses which are discussed in Section III F. These re-views should make clear to managers and constituents,alike, appropriate interpretations of existing data for theexisting management questions being asked, limitationsof those data, and relative benefits to the decisions withimprovement of the data.

Communication effects on credibility are a difficult prob-lem. NMFS scientists are presently being asked to com-municate and prepare documents in three arenas: to pre-pare documents with scientific details that will communi-cate to scientists and withstand scientific challenges; toprepare documents that will communicate to managers/politicians/lawyers to withstand legal challenges; and toprepare documents to communicate the nature of the bio-logical and fisheries issues to lay constituents. Of thesethree arenas, NMFS scientists are required to put a dis-proportionate amount of effort into the second category.NMFS scientists, in general, probably get the most careerfulfillment from the first and it is in that arena where NMFSscientist�s activities overlap with academic scientists. Thethird category (preparing documents for lay constituen-cies) is admittedly lacking. Communication with constitu-encies is mostly spoken and often within the final deci-sion-making process. That atmosphere does not promotedialog. But beyond that, the three activities require re-sources: people (or proportions of people�s time) and travel.In addition, those individuals who actually conduct thescience are not likely to have equally good skills in com-munication in all three arenas. Each arena poses differentchallenges, and also offers different degrees of career ful-

fillment. Reallocation of a scientist�s time toward one activ-ity will often be at the expense of other required activitiesof the individual.

Another related credibility issue is that within NOAA.Unfortunately, credibility of NMFS science within NOAAis often filtered through constituent interactions rather thanthrough NMFS scientific interactions. Also, since muchof the assessment work results in regulations and limita-tions on the public users of fishery resources, it is hard todevelop �good news� communications.

Perhaps the best solution to the perceived credibility andcommunication problems is a recognition that under ex-isting management frameworks criticism will continue toexist; that communication to the lay public (includingwithin NOAA) takes specialized skills and resources; thatNMFS should develop new resources to achieve commu-nication goals; and that these new resources should not beachieved at the detriment of the data and assessment basesof the scientific advice.

E. Implications of the PrecautionaryApproach

A major goal of fish stock assessments is to provide scien-tific advice on sustainable harvest strategies. In reality,this is a dual goal involving determination of the harveststrategy that will, on one hand, approach the maximumlong-term average yield, yet, on the other hand, have alow probability of overfishing and causing depletion ofthe resource or other harm to the ecosystem. Maximizinglong-term average yield while at the same time minimiz-ing the risk of overfishing is impossible without a highlevel of knowledge about the abundance and productivityof the resource, especially when one considers the paucityof knowledge regarding the impact of climate, interactionswith other species, and habitat changes (Figure 1). Theprecautionary approach is concerned with maintaining abalance between high yields and low risks of depletion inthe face of uncertainty that is often substantial. In thepast, it has been common to treat uncertainty as a reasonto forestall implementation of restrictive management mea-sures. Although the precautionary approach has many fac-ets, one of the more important and universal features isthat uncertainty is perceived as a reason to exercise cau-tion by, for example, scaling back the recommended har-vest rate in relation to the level of uncertainty in estimatesof stock abundance and productivity (Figure 2) and todevelop other methods, such as marine reserves, to con-tribute to safeguarding these resources. Greater uncertaintyshould result in greater caution in fishing activities.

15

Attempts to deal with uncertainty in the context of theprecautionary approach have enlivened the developmentof limit reference points, target reference points, harvestcontrol rules, and management procedures simulation mod-els. Harvest control rules specify the management action(e.g. a specific fishing mortality) to be implemented de-pending on the status of the stock (e.g. the estimated bio-mass), and generally include target reference points (to beachieved on average) and limit reference points (to beavoided with high probability). The greater the degree ofuncertainty in the assessment of stock status or in the abil-ity to effectively implement management actions, the greaterthe difference between targets and limits should be. Mod-els that include the entire system of observation-assess-ment-management with the attendant suite of errors (com-monly called management procedures simulation models)deal with uncertainty in a more comprehensive way andenable evaluation of the robustness of alternative man-agement strategies. A previous NMFS Working Group de-veloped technical guidance on these and related topics asthey relate to the development of definitions of overfish-ing (Restrepo et al. 1998; Appendix 17).

The need to develop precautionary approaches, target andlimit reference points, harvest control rules, managementprocedures simulation models, and related methods hasadded considerably to the duties of stock assessment sci-entists and, in many cases, has strained the limits of avail-able data. In order to implement a precautionary approach,fishery scientists must deliver to fishery managers a de-scription of this uncertainty and an assessment of the riskscreated by overfishing and other impacts on the stock. It isnot adequate to simply report the best estimate and de-scribe its uncertainty. The analysis must be broadened toinclude evaluation of the possible consequences of alter-native harvest strategies given the amount of uncertaintyabout current and projected stock status.

F. Implications of the Need to IncorporateEcosystem Considerations

Trends of increasing intensity and specialization of fisher-ies, and needs to more fully integrate fisheries and pro-tected species management, argue for greater attention toecosystem effects not addressed by traditional overfish-ing concepts and stock assessment models supportingthem. Recent legal challenges to NMFS resource manage-ment decisions relative to the National Environmental Pro-tection Act (NEPA), the Endangered Species Act (ESA)and the MSFCMA illustrate the growing need for NMFS todevelop a comprehensive understanding of the effects offishing on marine food webs and the effects of fishing on

Figure 1.

Initial levels of investment in information (e.g. catch monitoring)establish a baseline; intermediate levels of information (e.g. an-nual resource surveys) produce substaintial gains; final levels ofinvestment (e.g. ecosystem research) may have diminishing re-turns with repect to short-term recommendations, but cansubstaintially affect long-term recommendations. Our goal is toobtain at least an intermediate level of information for the primarycommercially or recreationally-exploited species.

Figure 2.

With a low level of information, there is a broad potential rangeof estimated biological yields. A precautionary approach in re-sponse to this uncertainty is to set recommendations safelybelow the �best estimate� to avoid accidentally exceeding thetrue value and overfishing. The progression towards reduceduncertainty is not as linear as indicated in this diagram. Someinformation, especially fishery-independent annual surveys, cangreatly reduce uncertainty as long as it is combined with ad-equate information of other types (Figure 1).

Information

Cer

tain

ty

DiminishingReturns

CriticalMinimumInformation

Goal

Overfishing

SAFE Recommendation

True ABC

Unc

erta

inty

in R

ecom

men

ded

Bio

logi

cal Y

ield

Typi

cal

Dat

a Le

vel

Land

edC

atch

Ann

ual

Sur

vey

Rec

ruitm

ent

For

cast

s

Eco

syst

emU

nede

rtan

ding

Added Information

16

marine habitats. In addition, there is currently consider-able interest from the public, stakeholders and the scien-tific community to move towards more comprehensive eco-system-based fisheries stock assessments and manage-ment. Several thorough studies on this topic have recentlybeen completed, including an NRC report entitled �Sustain-ing Marine Fisheries� that focused on sustainability in anecosystem context (NRC 1999; Appendix 18).

These issues necessitate the development of a new era inresource monitoring that requires collection of informa-tion on seasonal movements of fish, the response of fishto oceanographic factors, and trophic interactions. Withadded information regarding the functional relationshipsgoverning the spatial and temporal distribution of fish,ecosystem considerations could be incorporated into stockassessments by: (1) modifying existing single-species over-fishing paradigms and stock assessment approaches toaccount for ecosystem attributes, (2) coupling fully mecha-nistic �bottom-up� models that incorporate the influenceof trophic interactions and oceanographic factors on re-cruitment success to stage-based assessment models, and(3) developing aggregate system models to extract princi-pal properties of marine ecosystems that can be utilized todevelop single species harvest objectives. While NMFS isactively pursuing research in support of all three types ofmodels, in the short term modification of existing singlespecies models will be the most useful tool for providingmanagement advice to our constituents. These modelsallow assessment scientists to modify concepts of growthand recruitment overfishing, maximum sustainable andeconomic yields and protected species management toaccount for ecosystem attributes such as technological andbiological interactions among assemblages, restrictedpredator-prey communities, or shifts in carrying capacitydue to decadal scale variability in ocean conditions.

As our understanding of the mechanisms controlling theproductivity of marine communities improves, NMFS muststrive to develop fishing strategies that ensure sustainedcommunity production. Ecosystem overfishing referencepoints, based on metrics of biodiversity, sustainability, andtrophic considerations, have been proposed but not widelyapplied and typically fail to address the role of naturaldisturbance on shifts in community structure. While exist-ing single species overfishing concepts and modificationshave the advantage of a strong theoretical basis for evalu-ating choices between alternative management strategies(including risk assessment) and much practical use, theydo not yet provide guidance on issues such as serial deple-tion of economically-valuable stocks, changes in bio-di-

versity, habitat-modifying effects of fishing methods, orsome trophic impacts of fishing practices such as �fishingdown the food chain.� On the other hand, measures thatprevent overfishing of single species partially serve theobjectives of multispecies management by ensuring thatno stock is intentionally overfished. With adequate ob-server coverage, the bycatch can be closely monitored andtarget fisheries can be closed to protect bycatch species ifnecessary. The Ecosystem Principles Advisory Panel re-port on ecosystem-based fisheries management (Appendix19) calls for the development of Fisheries Ecosystem Plans.A long-term goal would be the development of a theoreti-cal basis for defining ecosystem overfishing. A short termactivity in support of these goals would be to improve ourdocumentation and monitoring of metrics of ecosystemssuch as diversity indices, slopes of size or diversity spec-tra, or average trophic level. The performance of these in-dices as predictors or heuristic reference points for man-agement remains untested, but could be evaluated retro-spectively for candidate situations of obvious ecosystemoverfishing.

Regardless of approach, there appears to be a need to ac-count for ecosystem implications in a more formal waywhen looking at the entirety of fishery management mea-sures applied to systems. There is, then, a need to developand implement more widely quantitative models to assistmanagers in accounting for ecosystem considerations ex-plicitly when choosing between alternative managementstrategies. Nevertheless, resource management will, in alllikelihood, always be driven by the real or perceived im-portance to society or the ecosystem of a subset of ex-ploited and protected species. Rather than substituting forexisting overfishing and assessment modeling concepts,ecosystem considerations will increasingly be used toevaluate and modify primary management guidance ap-plied to the important species. In order for ecosystem over-fishing definitions to assume a greater role in resourcemanagement, quantifiable, predictive, and unambiguousassessments of ecosystem states and fluxes must be de-veloped and evaluated. This implies research focused onprocesses and interrelationships, complementing�ratherthan replacing�traditional stock assessment and moni-toring activities supporting existing overfishing definitions.In all likelihood, advice resulting from the explicit incorpo-ration of ecosystem effects will even further emphasize theneed for conservative management of the fishing capacityof single- and multipurpose fleets, supported by refine-ments in the use of technical measures such as marineprotected areas and gear restrictions.

17

This section provides background information on require-ments for conducting assessments and for evaluating al-ternative fisheries management strategies. Topics coveredinclude input data, stock assessment models, assessmentfrequency, adequacy of technology and infrastructure, peerreview processes, translation of stock assessment adviceinto management action, communication of assessmentresults and analyses of alternative management strategies,and staffing issues.

A. Input Data

Calibration of stock assessment models requires three es-sential categories of data: catch, abundance, and life his-tory characteristics. These data come from fishery-depen-dent and fishery-independent sources. The role of catchdata in stock assessment models is to indicate the magni-tude of fishery removals during the time period in whichthe surveys have measured a change in abundance. Totalcatch is determined from monitoring by port samplers andobservers, and mandatory or voluntary reporting systems.The most reliable indicators of changes in population abun-dance are fishery-independent resource surveys (NRC1998a). In some cases, it is possible to conduct taggingstudies, depletion experiments, or absolutely calibratedsurveys that result in an absolute estimate of stock abun-dance rather than a relative index which must be trackedover time. Fishery-dependent data (e.g. logbook data) canalso be used to develop indices of changes in abundance;however, validation that these fishery-dependent indicesare truly proportional to changes in stock size usually re-quires comparing the fishery-dependent index to a fish-ery-independent survey index. Life history data (stockstructure, growth, reproduction, and natural mortality rates)indicate the geographic limits of the stock and its inherentproductivity. Inclusion of life history data in stock assess-ment models helps assure biologically realistic resultswhich properly separate fishing mortality from naturalchanges. With incomplete data on catch, abundance, or lifehistory characteristics, the results of assessment modelswill be less precise because of uncertainty in the assump-tions used in place of the missing data.

The need for improving the collection, management anduse of fisheries data was recognized in a recent report en-titled, �Improving the Collection, Management, and Use ofMarine Fisheries Data� (NRC 2000; Appendix 20).

III. Assessment and Management Strategy Evaluation Needs

(i) Fishery-dependent data needs

Fishery-dependent data include the landed catch, at-sea dis-cards, biological characteristics (age and size composition,sex ratio, maturity stage) of the catch, fishing effort, andspatial distribution of catch and effort. Accurate stock as-sessments require that the total removals (landed plus dis-carded catch) be known for all significant commercial andrecreational fishery segments. The primary methods to ob-tain these total catch data vary regionally and are stronglyinfluenced by the scale of typical fishing operations andby the degree of historical development of federal and statereporting systems. Methods to track large volume land-ings by trawl vessels at a few locations may be ill-suited toestimating total landings by large numbers of commercialor recreational hook and line fishers individually landingsmall amounts of fish at many locations. For example,mandatory reporting of landed commercial catch by thewest coast states provides a census of total commerciallandings. Off Alaska, mandatory observer programs deter-mine total catch for major species. For recreational fisher-ies, statistical sampling procedures are used to estimatetotal recreational catch and effort from samples of anglersnationwide. However, throughout the nation there are gapsin coverage for particular fishery segments, concerns aboutunder-reporting of total catch or misreporting of speciesand the areas in which they were caught, low levels of sam-pling coverage, and insufficient statistical and databasecapabilities to ensure timely access to well-audited data.

Information on the size and age composition of the catchis needed to accurately estimate the fishing mortality causedby that catch. These data are typically obtained by sam-plers in the fishing ports and by observers on board fishingvessels. When comparable data are available for each fish-ery segment, evaluation of the biological impacts of dif-ferent allocations among the segments is facilitated. Fur-thermore, size and age data from the fishery contribute in-formation on variability in recruitment.

Collection of commercial and recreational fishery datafaces significant logistic hurdles due to the need to imple-ment sound statistical sampling procedures. The potentialfor bias and inefficiency exists in current procedures, andthe NRC review of stock assessment methods recommendsthat a standardized and formalized data collection proto-col be established:

18

NRC Recommendation #9: �The Committeerecommends that a standardized and formalizeddata collection protocol be established forcommercial fisheries data nationwide. TheCommittee further recommends that a completereview of methods for collection of data fromcommercial fisheries be conducted by anindependent panel of experts.�

One step that has been taken towards addressing this rec-ommendation is the recent NMFS Report to Congress on a�Proposed Implementation of a Fishing Vessel Registra-tion and Fisheries Information Management System� (Ap-pendix 8), as required under the 1996 reauthorization ofthe MSFCMA. The report lays out a plan for implement-ing a Fisheries Information System (FIS) by integratingand expanding on the current regional fisheries coopera-tive statistics activities in three major areas: data collec-tion, information management, and institutional arrange-ments. However, the plan is to integrate existing activities,rather than to overhaul the system completely and developprotocols to be used nationwide. The plan has been sub-mitted to Congress as required, but to date, it has not beenfunded.

(ii) Fishery-independent data needs

Fishery-independent data include information on the dis-tribution, abundance, and biology of the species being as-sessed. A suitable fishery-independent survey method musteither be calibrated to measure absolute fish abundance,or it must be directly proportional to fish abundance sothat relative trends can be tracked. When the time series ofa survey is short, there is greater value in calibrating thesurvey for absolute abundance; however, such estimatesare critically dependent on obtaining good estimates ofcatchability. As the time series gets longer, the trend infor-mation becomes more useful.

A common survey approach is to use carefully standard-ized sampling gear (e.g. trawls, hooks, or pots) to collecthundreds of samples distributed over the expected rangeof the stock. Such a resource assessment survey providesinformation on distribution and abundance, and providesspecimens for age, growth, genetic stock structure, foodhabits, maturity, and other biological studies. However, suchmethods can be difficult to standardize completely becausefish behavior and gear performance may vary with habitatand environmental conditions. Other methods are valuablefor directly calibrating such surveys, providing informa-tion from habitats not accessible to the primary samplingtool, and providing alternative measures of fish abundance.Acoustic methods have been developed to provide cali-brated information on distribution and abundance, but must

be coupled with other sampling tools to collect biologicalspecimens. Egg and larval methods have been developedto provide measurements of abundance (spawning biom-ass) that are not susceptible to the same types of samplingproblems that may affect trawl surveys. Imaging systems(visual, laser) are an appropriate tool in high reliefnearshore habitats and have been useful in understandingthe interaction between fish and other sampling tools. Mark-recapture methods, like egg and larval methods, can pro-vide a direct estimate of absolute abundance but must relyon other tools to measure distribution and to collect bio-logical specimens. More generally, a single survey methodmay not be suitable for the entire age range; for example, aseparate survey may be necessary to provide an index ofrecruitment. In many instances, it is likely that at least twosurvey methods may need to be deployed in order to pro-vide appropriate input for stock assessments and projections.

The NRC (1998a) evaluation of stock assessment meth-ods recommended that each stock assessment contain atleast one reliable index of relative stock abundance, pref-erably from fishery-independent surveys because incom-pletely calibrated fishery-dependent indices can lead tobiased stock assessment results:

NRC Recommendation #2: �At the minimum, atleast one reliable abundance index should beavailable for each stock. Fishery-independentsurveys offer the best choice for achieving areliable index if designed well with respect tolocation, timing, sampling gear, and otherstatistical survey design considerations.�

Attempts to satisfy this type of recommendation haveplayed a key role in NMFS� research planning for severaldecades. The most recent document directed specificallyat this type of recommendation is the NOAA Fisheries DataAcquisition Plan (NMFS 1998a; Appendix 3), which callsfor a combination of purpose-built fishery research ves-sels and chartered days-at-sea to satisfy immediate fish-ery-independent data collection needs.

Reliable fishery-independent indices are already availablefor several key stocks, primarily in the northeastern UnitedStates and Alaska which have long time series of researchsurvey data. Such indices will become even more widelyavailable as NMFS and partnering agencies and institu-tions acquire additional research platforms, including dedi-cated research vessels. But, even with additional resourcesfor research, some important variables will always be dif-ficult to estimate; for example, natural mortality, which isa key assessment variable singled out for attention by NRC(1998a):

19

NRC Recommendation #3 (in part): � ...Greaterattention should also be devoted to includingindependent estimates of natural mortality inassessment models.�

To obtain reliable independent estimates of natural mor-tality, the types of fishery-independent research requiredare likely to involve extensive mark-recapture studies and/or collection and analysis of food habitats data from largenumbers of potential predators covering extensive spatialand temporal scales.

B. Input Data: Minimal and OptimalRequirements

The great diversity of data available for the world�s fisher-ies has fostered the development of a wide range of stockassessment modeling methods that can take advantage ofthese data. As the scope of the data and their quality andquantity improves, several improvements in stock assess-ment results will accrue. As data become more precise andas the time series of data become longer, the precision ofstock assessment results should improve, and there shouldbe greater stability in resulting recommendations on thestatus and potential yield from the stock. As more types ofdata become available, it will be possible to test and vali-date model assumptions and reduce the possibility thatmodel results are biased because of inappropriate assump-tions about the data. Appropriate data are also needed toreliably forecast likely future conditions of a stock, in ad-dition to obtaining a retrospective view of a stock�s his-tory. It has sometimes been argued in the scientific litera-ture that well-calibrated fishery catch per unit effort (CPUE)data is an adequate measure of relative stock abundance,and that useful stock assessments can be based solely onsimple models tuned to such data. While this may be truefor some fisheries, there are many case studies demonstrat-ing that the assumption that commercial CPUE is directlyproportional to resource abundance may lead to large bi-ases in results, and that such bias is often detected too lateand only when additional sources of data are obtained andincluded in the assessment.

In each of three major categories of information requiredas input to stock assessments; viz, catch, abundance, andlife history, the Task Force defined 5-6 progressively morecomplete levels of data availability (Figure 3). Such a pro-gression will fit no fishery perfectly, but gives a generalguide to the progression of information improvement thatshould be the goal of comprehensive stock monitoring pro-grams. A balanced development in these three categoriesof input data is also beneficial; generally, a stock assess-

ment model will not be able to fully utilize detailed catchdata if there is an inadequate survey index and lack of keybiological data.

Levels of catch data

0 � No catch data.1 � Landed catch provides a minimum estimate of fish-

ery removals and is typically obtained from man-datory landing receipts. In some cases, particu-larly recreational fisheries, a statistical samplingprogram is used to expand estimates of sampledcatch up to the total angling population.

2 � Catch size composition provides a measure of thesizes of fish being impacted by the fishery, andwhen tracked over time can provide an index ofrecruitment to the fishery and total mortality rates.

3 � Spatial data on catch from logbooks can provideinformation on range extensions and contractions,and other changes in stock or fleet distribution.

4 � Catch age composition requires the developmentof age determination techniques and an invest-ment in the collection and processing of appro-priate samples. The result is much greater stockassessment accuracy than can be obtained withsize composition data alone.

5 � Accurate and complete data on total removals(including landed catch, discards, bycatch in otherfisheries, and cryptic mortality induced by fish-ing gear contact) will contribute to accurate stockassessment results. An at-sea observer programcan monitor total removals, cross-check logbookdata, and collect site-specific biological samples.In many fisheries, the relative merits of observerprograms for collecting data on total removals and/or age composition data may warrant consider-ation before or instead of investing in a fisherylogbook program.

Levels of abundance data

0 � No abundance data.1 � Relative abundance index from fishery catch per

unit effort or an imprecise, infrequent survey.Another Level 1 situation would be a singlesurvey from which an estimate of absoluteabundance has been made. At this low level ofinformation there will only be a limited ability totrack changes in stock abundance because ofuncertainties in the calibration of the index, or ahigh level of noise in the data relative to themagnitude of the expected changes in stockabundance.

20

Catch Abundance Life History0 = none1 = landed catch2 = catch size composition3 = spatial patterns (logbooks)4 = catch age composition5 = total catch by sector (observers)

0 = none1 = fishery CPUE or imprecise survey with size composition2 = precise, frequent survey with age composition3 = survey with estimates of q4 = habitat-specific survey

0 = none1 = size2 = basic demographic parameters3 = seasonal or spatial information (mixing, migration)4 = food habits data

0 = none1 = index only (commercial or research CPUE)2 = simple life history equilibrium models3 = aggregated production models4 = size/age/stage-structured models5 = add ecosystem (multispecies, environment), spatial & seasonal analyses

0 = never1 = infrequent2 = frequent or recent (2-3 years)3 = annual or more

Assessment Frequency

Figure 3. Factors used to classify stocks in terms of input data and assessment status.

Data

2 � Precise, frequent surveys with age compositionwill provide more accurate tracking of changes instock abundance and the associated agecomposition data will enable better estimation ofhistorical and current levels of recruitment.

3 � Research surveys with known or estimatedcatchability, acoustic surveys with known orestimated target strengths, and statistically-designed tagging studies can provide estimatesof absolute abundance. This is especially valuablewhen the time series of the survey is so short thatno trend is detectable.

4 � Habitat-specific surveys refine the concept ofstratified random surveys so that survey resultsare more closely associated with particularhabitats. The result is improved knowledge of therelationship between fish assemblages and habitatfeatures. In addition, these surveys use alternativemethodologies to extend survey coverage into allrelevant habitats.

Levels of life history data

0 � No life history data.1 � The size composition of harvested fish provides

a simple index of a stock�s growth potential andvulnerability to overharvesting.

2 � Basic demographic parameters such as age,growth, and maturity rates provide informationon productivity and natural mortality.

3 � Seasonal and spatial patterns of mixing, migration,and variability in life history characteristics,especially growth and maturity, provides improvedunderstanding of how a population responds toits environment.

4 � Food habits information defines the predator-preyand competitive relationships within the fishcommunity, thus providing a first step towardsdirect estimation of natural mortality rates andecologically-based harvest recommendations.

The availability of data at these various levels is tabulatedin Appendix 1 for each of the 904 stocks included in theNMFS (1999a) Report to Congress on the Status of Fish-eries of the United States. The data are also summarized byindividual and combined Science Centers in Table 1 andFigure 4 and discussed in Section III D.

21

Table 1. Numbers of stocks with differentlevels of input data (catch, abundance andlife history parameters), assessment meth-odology and assessment frequency for the904 stocks listed in the NMFS (1999a) Re-port to Congress on the Status of Fisheriesof the United States. Zero indicates noinformation; otherwise, the higher the level,the better the information. See Figure 3and the text for a description of the levels,Figure 4 for graphical comparisons, and Ap-pendix 1 for the stock-by-stock information.

SLEVEL hctaC ecnadnubA yrotsiHefiL tnemssessAleveL

tnemssessAycneuqerF

DENIBMOCSNOIGERLLA

0 59 473 69 545 3445.0 04 41

1 555 553 915 351 7012 27 521 702 06 9213 54 23 96 72 5224 17 4 31 1115 62 8muS 409 409 409 409 409

CFMSA&CMFAM,CMFEN0 4 1 0 2 75.0 0 0

1 7 03 9 51 222 21 22 42 51 813 8 3 31 3 94 02 0 01 915 5 2muS 65 65 65 65 65

SMHcitnaltA&CMFC,CMFMG,CMFAS0 0 582 94 872 2135.0 0 0

1 483 711 292 58 822 21 91 86 81 183 22 1 31 31 14 1 0 0 825 3 0muS 224 224 224 224 224

CMFP0 0 14 1 26 265.0 04 41

1 62 43 95 2 412 7 0 72 9 413 9 81 91 4 914 62 2 3 235 1 0muS 901 901 901 901 901

CMFPW0 31 31 51 82 825.0 0 0

1 73 14 0 22 212 5 3 73 0 63 3 6 21 7 814 6 1 0 45 0 3muS 46 46 46 46 46

CMFPN0 87 43 13 571 435.0 0 0

1 101 331 951 92 132 63 18 15 81 013 3 4 21 0 8714 81 1 0 825 71 3muS 352 352 352 352 352

22

0

100

200

300

400

500

3210

0

100

200

300

400

500

600

543210-0.50

100

200

300

400

500

600

543210

Figure 4.

Number of stocks with different levelsof input data (catch, abundance, andlife history parameters), assessmentmethodology, and assessment fre-quency for the 904 species listed inthe NMFS (1999a) Report to Congresson the Status of Fisheries of the UnitedStates. See Figure 3 and the text fora description of the levels, Table 1 fortabulated summaries, and Appendix1 for the stock-by-stock information.

Catch

Abundance

Life history

Assessment level

Assessment frequency

NPFMC

WPFMC

PFMC

SEFMC & HMS

NEFMC & MAFMC

0

50

100

150

200

250

300

350

400

43210-0.5

0

100

200

300

400

500

600

43210

23

C. Stock Assessment Models

The complexity of assessment methods used for a givenstock generally reflects the availability of data and the valueor importance of the fishery. To indicate the current levelsof analysis of the status of various stocks, two columnswere added to Appendix 1, one giving a numerical codethat roughly indicates the level of modeling effort/ com-plexity/ sophistication applied to each species in Appen-dix 1 and the other giving the frequency with which stockassessments are conducted. To be classified as an assess-ment, an analysis must produce some measure of stock orfishery status relative to a benchmark such as a fishing tar-get or an overfishing limit. The assessment level codes havethe following meanings:

0 � Although some data may have been collected onthis species, these data have not been examinedbeyond simple time series plots or tabulations ofcatch.

1 � Either:a) a time series of a (potentially-imprecise) abun-dance index calculated as raw or standardizedCPUE in commercial, recreational, or survey ves-sel data, orb) a onetime estimation of absolute abundancemade on the basis of tagging results, a depletionstudy, or some form of calibrated survey.

2 � Simple equilibrium models applied to life historyinformation; for example, yield per recruit orspawner per recruit functions based on mortality,growth, and maturity schedules; catch curveanalysis; survival analysis; or length-based co-hort analysis.

3 � Equilibrium and non-equilibrium production mod-els aggregated both spatially and over age andsize; for example, the Schaefer model and the Pella-Tomlinson model.

4 � Size, stage, or age structured models such as co-hort analysis and untuned and tuned VPA analy-ses, age-structured production models, CAGEAN,stock synthesis, size or age-structured Bayesianmodels, modified DeLury methods, and size orage-based mark-recapture models.

5 � Assessment models incorporating ecosystemconsiderations and spatial and seasonal analy-ses in addition to Levels 3 or 4. Ecosystem con-siderations include one or more of the following:a) one or more time-varying parameters, eitherestimated as constrained series, or driven by en-vironmental variables,b) multiple target species as state variables in themodel, or

c) living components of the ecosystem other thanthe target species included as state variables in themodel.

According to the above scheme, an �assessment� is a dataanalysis at Level 1 or greater, provided that analysis al-lows statements about relative stock or fishery status to bemade. The frequency column in Appendix 1 give codesdefined as follows:

0 � Never: an assessment has never been conducted.1� Infrequent: the most recent assessment was

conducted more than three years ago.2 � Frequent or recent: the most recent assessment

was conducted within the last three years but isnot conducted annually.

3 � Annual or more: assessments are conducted atleast annually.

The assessment levels listed above were designed to rep-resent a series of increasing analytical effort and sophisti-cation. Lower levels are amenable to use of standardizedsoftware, but upper levels, particularly Level 5, probablyrequire that models be tailored to deal with the particulari-ties of each stock assessment or group of related assess-ments. Such newly crafted models will most likely requireadditional types of input data concerning oceanographicconditions, and/or biological features of various compo-nents of the ecosystem, and/or operational details of thefishing gear. Thus the progression of assessment levelsimplies a progression of increasing data needs. In addi-tion, NRC (1998a) recommends that, where possible, morethan one assessment model should be applied for a givendata-set or fish stock:

NRC Recommendation #3 (in part): �Becausethere are often problems with the data used inassessments, a variety of different assessmentmodels should be applied to the same data; newmethods may have to be developed to evaluatethe results of such procedures. The differentviews provided by different models shouldimprove the quality of assessment results...�

Another NRC (1998a) recommendation is to develop newtechniques for stock assessment:

NRC Recommendation #7: �NMFS and otherbodies responsible for fisheries managementshould support the development of newtechniques for stock assessment that are robustto incomplete, ambiguous, and variable data andto the effects of environmental fluctuations infisheries.�

24

Development of new computational and analytical tech-niques is most likely to take place at the upper end of thelist of assessment levels where it will likely be necessaryto craft a wide variety of new assessment models. How-ever, because these new kinds of models may require newkinds of data, there may also be a need to develop newsampling or survey techniques, or to perform experiments.Models that can incorporate the effects of environmentalfluctuations in fisheries are already under active develop-ment, but they suffer from a lack of understanding of theeffects of major environmental regime shifts on individualspecies and species interactions. Thus, their predictabilityis currently limited. On the other hand, pursuit of this lineof research is likely to be more fruitful than attempts todevelop models that are robust to incomplete or ambigu-ous data � the extent to which models can compensate fordata deficiencies is extremely limited.

NRC (1998a) also recommended the development of morerealistic assessments of uncertainty:

NRC Recommendation #4: �The Committeerecommends that fish stock assessments includerealistic measures of the uncertainty in the outputvariables whenever feasible. Although a simplemodel can be a useful management tool, morecomplex models are needed to better quantify allthe unknown aspects of the system and to addressthe long-term consequences of specific decisionrules adequately. The implementation of thisrecommendation could follow the methodsdiscussed in Chapter 3 [of NRC 1998a].�

While expression of uncertainty is not included in the abovedefinitions of assessment levels, one would expect that in-creasing sophistication of assessment models would gohand in hand with increasing sophistication in calculatinguncertainty, and one would hope that more sophisticateduncertainty assessments would also be more realistic. Tothe extent that more realistic uncertainty assessments in-corporate more components of variation, there is the pos-sibility that they would show wider confidence bounds.Thus the higher level assessments are not at all guaranteedto yield tighter uncertainty distributions and consequenthigh levels of allowable catch as promised elsewhere.Therefore, it is important that lower level assessments beaccompanied by uncertainty calculations that are sophisti-cated and comprehensive enough to make them as realisticas they are for higher level assessments.

D. Inventory of the Status of StockAssessments: Adequacy of Input Data,Assessment Level, and Frequency ofAssessments

The status of input data and assessment analyses for the904 stocks listed in the 1999 Report to Congress on theStatus of Fisheries of the United States (NMFS 1999a) istabulated in Appendix 1 and summarized in Table 1 andFigure 4. The first point to note is that for most stocks,there is at least basic information on landed catch and thesize frequency of the catch. However, for more than 40%of the stocks, there is no fishery-independent or fishery-dependent index of abundance, which makes it extremelydifficult to conduct a meaningful assessment. Other fac-tors, such as the need to prioritize the stocks to be assessed,results in a total of about 60% of the stocks (545 stocks)lacking assessments sufficient to evaluate stock status rela-tive to overfishing. Although there are relatively few stocksat the highest levels of each of the input data categories, atotal of 119 stocks are routinely assessed at Level 4 orhigher. Detailed examination of the information containedin Appendix 1 shows that most of NMFS� data collectionand assessment resources have been directed towards thosespecies that dominate in the catch or have previously beendeemed to be overfished. With a few exceptions, all of thehigh-valued, high-volume, or high-profile species are rou-tinely assessed, while most of the stocks with few inputdata and analysis are bycatch species that contribute littleor nothing to total landings. In other words, they are usu-ally relatively unimportant from an economic perspective.Their importance from an ecological perspective and theirbiological status with respect to risk of reproductive fail-ure is generally unknown.

E. Adequacy of Technology andInfrastructure

In some respects, the job expected of stock assessmentscientists is impossible: to estimate the numbers and bio-mass of each harvested species in the ocean even thoughthey cannot be seen; to determine demographic parameterssuch as growth and mortality even though such are affectedby unobservable and complex interactions between spe-cies and with the environment; and to forecast catches andpopulation responses ahead 1-10 or more years even thoughincoming recruitment is known to be highly variable andaffected by environmental events that may not yet haveoccurred. The difficulty of these tasks necessitates high-technology solutions. Improved technologies are neededto sample, survey, or experiment with species of interest insitu, in order to decrease sampling error, increase samplingintensity, or increase the area or number of species cov-

25

ered. Such technologies, many of which are actively beingdeveloped at present (Appendix 12), include developmentof specialized sampling nets and other methods of directsampling, multifrequency acoustics, multi-beam acoustics,LIDAR, laser line scan systems, remotely-operated vehiclesequipped with underwater cameras, and electronic acous-tic or satellite tags.

Improvements in fishery-dependent data sampling are alsorequired to reduce the reporting burden on fishers, reducereporting errors and mistranslation of information, and in-crease the timeliness of availability of such information.Vessel monitoring systems are already in use by severalfishing fleets, but these are mainly used to record and moni-tor vessel location. Several prototype electronic logbooksystems have been developed and tested and, if these canbecome part of the standard operating procedures of allmajor fisheries, they will have tremendous benefit to fish-ers, scientists and managers alike. Tools for remote moni-toring of fishing behavior and catch quantity and composi-tion are under development but complex problems remainto be solved.

The availability of hardware and software for processingthe complex and voluminous data collected by some sam-pling tools is often a limiting factor in the implementationof innovative assessment methodologies. Lack of adequatecomputing power may also be an obstacle in the stock as-sessment and stock projection processes, particularly whenrealistic representations of uncertainty are attempted. Forexample, if uncertainties in assessment inputs are modeledsuch that probability distributions of current status are pro-duced rather than point estimates, and then future stock orfishery status is projected from these distributions incor-porating uncertainty in future events as well, the numberof iterations required can quickly mount up and bog downexisting computer systems. Thus, proposals for improvingstock assessments need to be linked to advanced technol-ogy initiatives and information technology (IT) planning.In fact, it is now a requirement that IT staff be included inprogrammatic planning activities and the budget formula-tion process.

The final infrastructure-related concern voiced by the TaskForce was the availability of space to house the additionalstaff required to improve stock assessments. Office spaceis already at a premium in most NMFS facilities. Any planto increase on-site staff will also need to address this issue.

F. Peer Review of Assessments

Stock assessments conducted by NMFS are rarely, if ever,the product of a single individual. Peer review is an integral

part of the process conducted by fisheries scientists fromwithin and outside of NMFS. The NRC (1998a) recom-mended that:

NRC Recommendation #8: �NMFS conduct (atreasonable intervals) in-depth, independent peerreview of its fishery management methods toinclude (1) the survey sampling methods used inthe collection of fishery and fishery-independentdata, (2) stock assessment procedures, and (3)management and risk assessment strategies.�

With regards to the three classes of peer reviews listed bythe NRC, NMFS routinely conducts peer reviews of stockassessments and stock assessment procedures, and occa-sionally conducts reviews of survey sampling methods, butrarely conducts reviews of management strategies. One ofthe problems that arises is in the interpretation of the word�independent.� To some, it means non-government, or atleast non-NMFS. A more liberal interpretation is simply areview conducted by experts who have not been directlyinvolved in the work being evaluated. NMFS Science Cen-ters frequently recruit scientists from other Science Cen-ters, regional offices, or headquarters to participate in peerreviews. It is also common to invite state fisheries scien-tists, academics and non U.S. nationals to serve as review-ers, particularly now that the Center for Independent Ex-perts (CIE) has been formed. The CIE (Appendix 5) pro-vides a mechanism for accessing a worldwide pool ofhighly-qualified fisheries scientists, statisticians, and otherexperts.

All five Science Centers have systems in place for the peerreview of stock assessments and sampling methods. Theseare described briefly below.


The current peer review forum for stock assessments con-ducted in the Northeast dates back to 1985. At that time aregion-wide process was initiated to subject selected stockassessments to a two-level peer review. The goals of thisprocess are to assure that scientists reviewing the assess-ments are not those responsible for the conduct of the work,and that experts independent of the process are included.Although the details of the structure have changed, theStock Assessment Workshop (SAW) has been, and contin-ues to be, the main vehicle for critical evaluation of stockassessment results and the crafting of management advicein the region.

The SAW consists of two parts: two week-long meetingsof the Stock Assessment Review Committee (SARC), usu-

26

ally conducted in June and November each year, and for-mal SAW meetings with the New England and Mid-Atlan-tic Fishery Management Councils where results of theSARC are presented and feedback is solicited. Overseeingthe process is the SAW Steering Committee, consisting ofthe Northeast Regional Administrator, the Science andResearch Director of the Northeast Fisheries Science Cen-ter, and the executive directors of the New England andMid-Atlantic FMCs and the Atlantic States Marine Fisher-ies Commission (ASMFC). The steering committee setsthe schedule of which stock assessments will be reviewedat upcoming SARC meetings, and determines the specific�terms of reference� establishing assessment informationrequirements of managers. The SARC committee usuallyconsists of about 10 members selected from the staffs ofthe Center, ASMFC (member states), the Councils, anddesignated outside experts from academia, state agencies,other NMFS Centers and foreign (usually Canadian) re-search institutes.

Stock assessments reviewed at the SARC are conductedby standing working groups (WGs) responsible for thevarious species: Northern Demersal WG, Southern Demer-sal WG, Coastal/Pelagic WG, Invertebrate WG. Stock as-sessment methods are addressed by the standing Meth-ods WG. Currently the chairs of the WGs are NEFSC scien-tists, but membership consists of state and academic sci-entists as well. Some assessments are contributed to theSARC directly by ASMFC assessment committees. Thechair of the SARC meetings has, in recent years, been se-lected by the Center for Independent Experts (CIE; Appen-dix 5), and has included individuals from Canada and Scot-land in the past two years.

In addition to the SARC/SAW process, which primarilyaddresses assessments of state and national importance,stock assessments of transboundary (international) impor-tance are peer reviewed in additional fora, including theTRAC (Transboundary Resource Assessment Committee),a joint U.S.-Canada committee responsible for cod, had-dock and yellowtail flounder on Georges Bank, NAFO (theNorthwest Atlantic Fisheries Organization), which is re-sponsible for Illex squid and various other stocks, andNASCO (the North Atlantic Salmon Conservation Orga-nization), which receives scientific advice from ICES (theInternational Council for the Exploration of the Sea) NorthAtlantic Salmon WG.


As the SEFSC provides quantitative stock assessment ad-vice to three Fishery Management Councils plus the NMFSAtlantic Highly Migratory Species Management Division,

there are a number of peer-review processes that are un-dertaken. Most typically, SEFSC assessments are conductedby a team of Center scientists. Assessment documents areprepared and distributed to Council review panel mem-bers in advance of review meetings. The Councils� stockassessment review panels, which are typically comprisedof regional experts who have not been involved in the workbeing evaluated, comment on the adequacy of the assess-ment and provide management advice to each Council. Inthe case of many Atlantic HMS fisheries, assessments areconducted in an international assessment working groupsetting (through ICCAT), with subsequent additional peerreview conducted by ICCAT�s Standing Committee onResearch and Statistics. Within ICCAT, it is not unusualfor the U.S. scientific delegation to be comprised of scien-tific representatives of a wide array of interest groups. It isalso common for the U.S. scientific delegation to be com-prised of non-U.S. nationals. For other HMS species (e.g.coastal sharks), assessments are carried out in a workshopformat in which state fisheries scientists, academics andnon-U.S. nationals participate in the assessment. Furtherreview of any of these assessments is also undertakenthrough the Center for Independent Experts (CIE; Appendix5), if the assessment results appear to raise controversy.


The Southwest Fisheries Science Center provides peer re-view for all its stock assessments and uses a variety ofmechanisms to do so. The choice of mechanism is oftenbased on the customary approach for the forum receivingthe assessment. For assessments produced by the SWFSCfor the Pacific Fishery Management Council (PFMC) aStock Assessment Review Panel (STAR Panel) is formedwith members chosen from the Council�s SSC and othernominated non-NMFS individuals to review and verify theassessment. For assessments produced by the SWFSC forthe Western Pacific Fishery Management Council(WPFMC), peer review is accomplished using the Centerfor Independent Experts (CIE; Appendix 5), currently co-ordinated through the University of Miami, or other desig-nated panels. Protected resource stock assessments are peerreviewed by panels of external reviewers constituted byexternal organizations such as the Inter-American tropicalTuna Commission (IATTC), or the Marine Mammal Com-mission, or the SWFSC. For SWFSC assessments presentedto international scientific bodies such as the Interim Sci-entific Committee for Tunas and Tuna-like Species in theNorth Pacific Ocean (ISC) or the Standing Committee onTuna and Billfish (SCTB) of the Secretariat of the PacificCommunity (SPC), either as finished assessments or asNMFS input for collaborative assessments, the receivingforum and its scientists provide the peer review.

27


The stock assessment review (STAR) process for ground-fish assessments off the U.S. west coast has been devel-oped as a shared responsibility of the National Marine Fish-eries Service and the Pacific Fishery Management Coun-cil. The STAR process helps make groundfish stock as-sessments the �best available� scientific information andfacilitates use of the information by the Council. The pro-cess operates under the direction of a NMFS Stock As-sessment Coordinator and reports primarily through theCouncil�s Scientific and Statistical Committee (SSC). Theprocess has a detailed calendar, explicit responsibilities forall participants, and specified outcomes and reports. STARpanels meet in a public setting in which all interested par-ties are legitimate meeting participants. This increases un-derstanding and acceptance of groundfish stock assessmentand review work by all members of the Council family.

The STAR Panel�s terms of reference concern technicalaspects of stock assessment work. The Panel is expectedto identify scenarios that are unlikely or have a flawed tech-nical basis, while reporting information, discussions, anddisagreements which reflect uncertainty in the assessment.The Panel operates by consensus and strives for a risk neu-tral approach in its reports and deliberations.

STAR Panels normally meet for one week to review twoassessments. Typically 2-3 Panels meet each year, and Pan-els reviewing transboundary assessments are informallycoordinated with the Canadian stock assessment reviewprocess. Each Panel normally includes a chair, at least one�external� member (i.e. outside the Council family and notinvolved in management or assessment of West Coastgroundfish), and one SSC member. In addition to Panelmembers, STAR meetings will include representatives fromCouncil technical and advisory committees with responsi-bilities laid out in their terms of reference. The STAR�sSSC representative attends Council meetings where stockassessments are discussed to explain the reviews and pro-vide other technical information and advice.


The AFSC provides stock assessment advice to the NPFMCon an annual basis. Stock assessments are reviewed inter-nally for consistency and accuracy. External technical re-views are conducted by the NPFMC BSAI and GOA PlanTeams and the Scientific and Statistical Committee (SSC).The Plan Teams and SSC are composed of scientists whorepresent federal, state and academic institutions. The PlanTeams provides a detailed technical review of the assess-ment methods and analytical approaches. The SSC pro-

vides a similar level of technical advice and is responsiblefor establishing the Allowable Biological Catch (ABC) andOverfishing Level (OFL) for FMP species. Preliminaryassessments are prepared for the September Council meet-ing and final assessment documents are completed in No-vember for Council action in December. Preliminary as-sessment documents are required when assessment scien-tists introduce a new analytical method, or utilize a newdata source in their model. The preliminary SAFE pro-vides an opportunity for the analyst to incorporate com-ments and suggestions from the Plan Teams and Scientificand Statistical Committee in their final SAFE chapter. InNovember, the Plan Team meets to review the final SAFEdocuments. The Plan Teams prepare reports documentingtheir recommendations for ABCs and OFLs and they com-pile the SAFE document for distribution at the DecemberCouncil meeting.

AFSC schedules detailed reviews of selected assessmentson an annual or semiannual basis. Stock assessment ex-perts are invited to conduct a thorough review of the meth-odology used. This review process provides time for theassessment expert(s) to work one on one with the assess-ment scientists. Reports derived from this process are pre-sented to the NPFMC advisory bodies. This assessmentreview is similar in scope to the reviews now provided bythe Center for Independent Experts (Appendix 5).

G.Translation of Stock Assessment Advice into Management Action

The translation of stock assessment advice to managementaction is where science and management interface and isan important but often controversial activity. Conflicts fre-quently arise over the �proper� roles of scientists and man-agers. At one extreme, it is argued that there should begreater separation of the science and the management, inorder to ensure that the science is unbiased. Scientists wouldthen provide information on stock status in a form such asgraphs giving the probability that current or projected fish-ing mortalities will be above or below some benchmark(specified previously by the managers), and managerswould decide what action to take on the basis of this infor-mation. At the other extreme, it is argued that there shouldbe greater co-mingling of science and management withmost if not all science being specifically focused on man-agement-oriented questions, and the priorities for sciencebeing driven by management priorities. In reality, scien-tists provide information on stock status but, because theyhave the data, the quantitative skills, and the infrastruc-ture, are subsequently asked to evaluate the likely outcomesof alternative management actions with respect to theireffect on future stock status. Stock assessment scientists

28

are frequently members of Plan Development Teams andrelated groups that evaluate the effectiveness of alterna-tive management tactics and strategies in meeting manage-ment goals. Assessment scientists often also evaluate andprovide advice on management benchmarks.

Many of the problems addressed by Plan DevelopmentTeams and related groups are tactical; i.e. short-term mea-sures to solve the immediate problems of reducing fishingmortality and/or rebuilding stock biomass. Tactics that arefrequently evaluated include size limits, gear restrictions,closed areas, closed seasons, trip limits, total allowablecatches, limited entry, and restricted days at sea. When fish-ing mortality and fishing capacity are under control, andthe stock biomass is near long-term sustainable levels, it isappropriate to conduct strategic (long-term) analyses of�optimum� management strategies. Such strategies mightinclude constant fishing mortality strategies at various lev-els of fishing mortality, constant escapement strategies,constant catch strategies, alternative strategies that havevariable effects on the bycatch of protected species or non-target species or nontarget sizes, pros and cons of perma-nent closed areas, and the social and economic implica-tions of alternative fleet configurations. These types ofanalyses tend to be conducted only sporadically�typicallyat the beginning of development of a new management plan,during major overhaul of a plan, or as a research topicundertaken by one or more internal or external scientistson their own initiative.

The process of translating assessment advice to manage-ment action is also where conflicts arise over the �proper�amount of influence by, or interaction with, stakeholderssuch as the commercial, recreational and environmentalsectors. The process set up by the MSFCMA theoreticallyinvolves public participation at every step. In general, how-ever, there is relatively little public involvement in the as-sessment process itself, possibly because relatively fewpeople have the training or interest in the technical aspectsof the quantitative analyses conducted. There tends to beconsiderably greater involvement at the stage of formulat-ing management actions to improve stock or fishery sta-tus. This mainly takes the form of attendance at FisheryManagement Council meetings and public hearings and,increasingly, by challenging particular management actionsor the stock assessment itself in courts of law. The prob-lems addressed in these forums also tend to be mainly short-term and tactical.

It is likely that conflicts could be lessened considerably ifmore resources were to be devoted to improving this inter-face between science and management. First, more atten-tion should be paid to analyzing the long-term implica-tions of alternative management strategies, and a greater

array of alternatives should be examined. The NRC (1998a)study recommended evaluation of a wide array of alterna-tive management strategies in terms of their robustness toassessment and other errors.

NRC Recommendation #5: �Precautionarymanagement procedures should includemanagement tools specific to the speciesmanaged, such as threshold biomass levels, sizelimits, gear restrictions, and area closures (forsedentary species)�

Second, simulation models should be constructed to allowmanagers and other stakeholders to evaluate the implica-tions of alternative actions and strategies themselves. Suchmodels have been in existence for at least 25 years, but ittakes considerable time to program them and to constructa user-friendly interface, particularly if a wide array ofmanagement alternatives is incorporated. In addition, theymay need to be reprogrammed each time a new stock as-sessment is conducted. Third, models for analyzing assess-ment methods and harvesting strategies simultaneously,called management procedures simulation models, shouldbe constructed for each major stock or fishery. The struc-ture of management procedures simulation models variesbut they generally include an operating model that pro-vides a simulation of a �true� population, a procedure forsampling the true population, an assessment model that usesthe sampled data to produce a �perceived� population, amanagement model that implements specific harvest rules,and performance statistics and feedback associated witheach of these components. This is essentially the approachrecommended by NRC (1998a).

NRC Recommendation #6: �Assessmentmethods and harvesting strategies have to beevaluated simultaneously to determine theirability to achieve management goals. Ideally, thisinvolves implementing them both in simulationsof future stock trajectories. For complexassessment methods, this may prove to be verycomputationally intensive, and an alternative isto simulate only the decision rules while makingrealistic assumptions about the uncertainty offuture assessments. Simulation models should berealistic and should encompass a wide range ofpossible stock responses to management andnatural fluctuations consistent with historicalexperience. The performance of alternativemethods and decision rules should be evaluatedusing several criteria, including the distributionof yield and the probabilities of exceedingmanagement thresholds�

29

This framework is particularly useful for investigating therobustness of various types of biological reference pointsand management actions, but it is extremely labor andcomputationally intensive.

H. Communication of Assessment Results and Follow-up Evaluations

Communication, or the perception of a lack of communi-cation, may be one of the greatest threats to the credibilityof stock assessment science. Translating complex techni-cal information into formats that a wide array of constitu-ents can identify with can be extremely time consumingand not all assessment scientists are equally adept at it. Inaddition, it is often not pleasant or rewarding to presentassessment results and evaluations of alternative manage-ment strategies to a sometimes hostile audience with vary-ing agendas and views about the future of the fishery. Sci-entists are often accused by one or more sectors (e.g. small-scale commercial fishers, large-scale commercial fisher-ies, for-hire recreational fishers, private recreational fish-ers, and environmentalists) of being biased in favor of oneor more other sectors.

Another communication problem affecting the credibilityof stock assessments is the disassociation between the gen-eration and analysis of fishery-dependent data. A fisher fill-ing out a logbook detailing the catch at a certain time andplace may believe that the size of a particular catch wasmore a function of weather or oceanographic conditionsor the way the gear was deployed, rather than actual abun-dance, but this information will probably not be conveyedto those analyzing the data. In addition, fishers may some-times have an incentive to under- or over-report catches.And some fishers may provide incomplete data because,like most people, they simply dislike filling out forms. Fish-ers may distrust stock assessment results because (i) theyare aware of the problems inherent in the generation offishery-dependent data, and (ii) since fishers are adept atfinding fish, they may have a more optimistic view of thestate of the stock than is implied by the assessment. It isoften stated that it is impossible for a scientist to produce avalid assessment unless s/he spends time out on the waterobserving fish and fleet behavior. However, since individualvessels tend to focus on �hot spots,� a few days at sea wouldonly give a very localized view of a fishery or stock. Sci-entists also need to spend time on statistically-designedfishery-independent surveys to develop a more holistic viewof fish distribution and abundance.

Lack of time to communicate with other groups of con-stituents is not just a problem for scientists. All groups ofconstituents would probably benefit from participating

more in each others activities, but this would take timeaway from their own specialty. Assessment scientistsshould, however, devote more time and attention to com-munications about data deficiencies, to cooperative re-search with constituents, to communication of assessmentresults, and to interactive analysis of the implications ofalternative management tactics and strategies.

I. Staffing Issues

Education and training

NMFS employs the largest collection of world-renownedfisheries scientists of any agency, university, or other orga-nization worldwide. In general, these scientists have strongbackgrounds in both biology and either mathematics orstatistics. However, biologists with solid quantitative skills,or quantitative experts with some biological background,are relatively rare and the pool of qualified applicantsgraduating from appropriate university courses is actuallyshrinking. This situation was recognized by NRC (1998a)who recommended that:

NRC Recommendation #10: �NMFS and otherbodies that conduct stock assessments shouldensure a steady supply of well-trained stockassessment scientists to conduct actualassessments and to carry out associated research.NMFS should encourage partnerships amonguniversities, government laboratories, andindustry for their mutual benefit. This can beaccomplished by exchanging personnel and ideasand by providing funding for continuingeducation at the graduate, postdoctoral, andprofessional levels, including elements such ascooperative research projects and specializedcourses, workshops, and symposia.�

In fact, NMFS has numerous cooperative programs withacademic institutions (see Data Acquisition Report, NMFS1998; Appendix 3), provides funding for continuing edu-cation of employees, and frequently organizes topical work-shops and specialized courses. However, the paucity ofqualified applicants for advertised stock assessment sci-entist positions is evidence that insufficient people are be-ing encouraged to enter this field and receive appropriatetraining. A relatively new program designed to alleviatethis problem has been established jointly by NMFS andNOAA Sea Grant. Each year (beginning in 2000), this pro-gram will provide up to three years of funding, mentoringand summer employment for two Ph.D. candidates in quan-titative assessment-related areas of research, up to a maxi-mum of six students at any one time (Appendix 6).

30

In addition, NMFS does not have a comprehensive con-tinuing education program for technical staff, although therehave been several attempts to initiate such programs. Asshown below, assessment scientists do not feel that theyhave sufficient opportunity to participate in professionaldevelopment activities, including training.

Time and motion analysis

As detailed in Sections III G and III H and elsewhere inthis report, it is evident that there are many other demandsplaced on assessment scientists aside from the basic re-quirement of a background in biology and mathematics.Within a given year, an individual assessment scientist maybe expected to: (i) participate in fishery-independent sur-veys or other field work, (ii) provide input and advice onsampling designs for research surveys and other fishery-independent data collection activities, (iii) spend time oncommercial or recreational fishing vessels, (iv) provideinput and advice on the development of data collectionobjectives and protocols for observer programs and otherfishery-dependent data collection activities, (v) conductquality control or other preprocessing of data, (vi) con-duct stock assessments, (vii) conduct research into stockassessment methods, (viii) present assessment results topeer review panels and constituent groups, (ix) participateon peer review panels, (x) participate in fishery manage-ment plan development or evaluation teams, (xi) defend astock assessment in a court of law, (xii) research and writescientific papers for primary publication, (xiii) attend col-leagues� seminars and offer critical review, (xiv) conductformal, written peer reviews of articles submitted for pub-lication in scientific journals, (xv) participate on commit-tees to advance approaches to stock assessment and fish-eries management, (xvi) undertake training to stay abreastof new methodologies, (xvii) run courses or workshops totrain others, (xviii) participate in national and internationalmeetings and conferences to enhance professional devel-opment, and (xix) undertake a variable amount of admin-istrative duties depending on supervisory level. With lim-ited exceptions, there is insufficient scope for individualscientists to focus on just one or a few of these activitiesdue to an overall shortage of assessment scientists.

To better understand the allocation of NMFS� stock as-sessment scientists� time, and to determine whether thereis a difference between the actual and optimal allocation,the Task Force prepared a questionnaire and administeredit to working stock assessment scientists. Activities com-monly undertaken by assessment scientists were dividedinto ten categories: (i) the mechanics of stock assessments,(ii) modeling research to improve stock assessment meth-odology, (iii) other (field or related) research to improve

stock assessments, (iv) participation in data collection ordata management activities, (v) provision of scientific ad-vice to Fishery Management Councils and others, (vi) par-ticipation in FMP development, evaluation of the conse-quences of alternative management strategies, and otherCouncil-related activities, (vii) other interactions with con-stituents, (viii) professional development including re-searching and writing scientific papers, reading scientificjournals, attending conferences, and training, (ix) admin-istrative duties, and (x) other. Assessment scientists wereasked to estimate the percentage of their time roughly av-eraged over the previous two years spent in each of theseactivities, and also to estimate the ideal percentage alloca-tion of time averaged across a group of stock assessmentscientists, recognizing that there may be some degree ofspecialization between individuals. Results are summarizedin Figure 5 for all respondents combined and separatelyfor each Science Center except the Northwest (due to avery small sample size). Overall, about 22% of an averageassessment scientist�s time is spent on the mechanics ofconducting stock assessments, and this seems to be closeto ideal, although there are notable differences betweenCenters. The other features of the summarized results thatstand out are a desire to spend less time on data collectionand data management activities, providing scientific ad-vice, FMP development and, in particular, administrativeduties; and more time on modeling and other research andprofessional development. This is an important result thatsupports the belief of the Task Force that assessment sci-entists are �stretched too thin.� Production of an assess-ment and provision of advice are activities that usually havea critical time horizon associated with them, whereas re-search to improve the basis for assessments does not. Yet,such research is crucial to advance the discipline. Also, inthe hectic world of stock assessments, professional devel-opment is almost perceived as a luxury when, in fact, it isessential for maintaining a workforce of high caliber, in-ternationally renowned and respected assessment scientists.Interaction with peers both nationally and internationallyis also essential given the small size of the profession andthe need to have a critical mass to discuss and debate ideas.

Beyond assessment scientists

A wide diversity of staff is required to produce a stockassessment. In fact, stock assessment scientists just repre-sent the �tip of the iceberg� (Figure 6). Far greater num-bers of staff are deployed in critical data collection activi-ties, such as commercial or recreational catch and effortdata, port sampling for biological data, observer programs,and fishery-independent resource surveys. Additional staffare required to process biological samples (e.g. to deter-mine fish ages from hard structures, construct age-length

31

Figure 5.

Time and motion analysis for NMFS assessment scientists, av-eraged over all responding individuals, and individuals within eachof four of the Science Centers (the sample size for the NorthwestCenter was too small).

Percent of time

Southeast Center

Alaska Center

Time & Motion Analysis

Percent of time

Percent of time

Southwest Center

Percent of time

Northeast Center

Percent of time

Averaged over individuals

Now Ideal

0 5 10 15 20 25 30

Mechanics

Modeling research

Other research

Data collection

Give advice

FMP development

Other constituents

Professional development

Administrative duties

Other

0 5 10 15 20 25 30

Mechanics

Modeling research

Other research

Data collection

Give advice

FMP development

Other constituents

Professional develoment


Other

0 5 10 15 20 25 30

Mechanics

Modeling research

Other research

Data collection

Give advice

FMP development

Other constituents



Other

0 5 10 15 20 25 30

Mechanics

Modeling research

Other research

Data collection

Give advice

FMP development

Other constituents



Other

0 5 10 15 20 25 30

Mechanics

Modeling research

Other research

Data collection

Give advice

FMP development

Other constituents

Professional develoment


Other

32

Figure 6.

Schematic showing relative staffing requirements in support ofproviding scientific advice for fisheries management. Staff re-quirements for conducting stock assessments, developing newstock assessment methods, and communication of results andmanagement strategy evaluations represent just the tip of theiceberg.

Figure 7.

Schematic showing the relative costs of adding new species tobe assessed. Often the existing infrastructure can be used tocollect the basic data. However, the higher up the pyramid, theless the ability of the existing infrastructure to absorb new re-sponsibilities.

keys, develop growth curves, construct maturity ogives,and possibly to identify and count eggs and larval fish fromichthyoplankton surveys and examine stomach contents),and to enter, audit, integrate, and preprocess data from themyriad of data collection activities. Support staff such assecretaries, administrators, and human resource personnelare required to support the data collection and stock as-sessment staff and their activities. Assessment scientiststhemselves are involved in three primary assessment-re-lated activities: conducting assessments (using assessmentmodels), methods research (developing assessment mod-els), and analyzing management alternatives and provid-ing advice to managers and constituents based on assess-ment results (management strategy evaluations). Relativestaffing requirements for assessment-related responsibili-ties can be roughly represented by a pyramid, with datacollection activities forming the base of the pyramid, andthe assessments themselves at the apex using informationfrom all lower levels (Figure 6).

Thus, when a new species needs to be assessed, the entirepyramid of activities needs to be considered. If the exist-ing infrastructure can be used to collect the basic data forthe new species (or basic data are already being collectedbut have never been processed), it may only be necessary

to expand slightly on data collection and data managementactivities. However, the higher up the pyramid, the less theability of the existing infrastructure to absorb new respon-sibilities (Figure 7). If an entirely new program or infra-structure is needed to provide the basic data for the newspecies, one or more levels of the pyramid may requiresubstantial expansion.

Current assessment-related staffing requirements by typeof activity are detailed below using the northeast region asa case study.

Northeast case study

A careful inventory was conducted for staffing levels ex-pressed as Full Time Equivalents (FTEs) for all major datacollection, research, and modeling activities of direct rel-evance to northeast stock assessments. It should be notedthat these analyses apply only to staffing levels and otherresources contributed on the federal side; however, formany of the region�s resources, data collection and analy-ses are undertaken by staff in state marine fisheries or-ganizations. Totals FTEs by category are summarizedin Figure 8.

Fishery-dependent data collection(catch, effor t , landings, biological sampling)

Observer programs for bycatch, discards, etc.

Fishery-independent surveys

Processing of biological samples(age, growth, maturity, etc.)

Data entry, audit ing/database management

Preprocessing of data

Other support staff(secretarial, administrative)

Conductassessments

Methodsresearch

Communicationof results and

follow-upevaluations







Other support staff(secretarial , administrative)

Conductassessments

Methodsresearch







Other support staff(secretarial, administrative)

Conductassessments

Methodsresearch

Communicationof results and

follow-upevaluations

Add NewSpecies

CurrentSpecies

33

Figure 8.

Assessment-related staffing levels by type of activity for the North-east Fisheries Science Center. DMS = Data Management Ser-vices; R/V = Research Vessel.

Northeast Center / RegionStock Assessment Activities

Numbers of FTEs

Commercial Catch and Biological Sampling:(49 FTEs; 30 in-house, 19 contract)

Commercial landings data are primarily collected througha network of NMFS �port agents� stationed in major fish-ing ports throughout the region, and mandatory dealer andfisher-supplied data. Dealer records are required for mostmajor regulated species. Port agents assure that dealer dataare entered into computer files and audited. Individual fish-ers are required to submit vessel trip reports (VTRs or log-books), which are entered into computer files through acentral processing facility located at the Northeast RegionalOffice in Gloucester, Massachusetts. In addition to basicdata on fishery landings (lbs.), VTRs are also used to allo-cate landings to water area fished, which is an importantelement when more than one stock of a particular speciesis assessed and managed, and for analyses of managementstrategies involving measures such as closed fishing areas.Discard data are requested in VTRs, but the data providedin these self-reported documents are generally consideredunreliable for stock assessment purposes (although the datahave been used in some limited circumstances). These datacollection programs generate information for activitiesother than stock assessment (e.g. compliance monitoring,economics, and management), but without such data, moni-toring of the effects of fishing on the stocks would not bepossible.

Biological sampling of landings (length sampling, collec-tion of structures for subsequent ageing) is also carried outby port agents and additional sampling staff allocated tothe ports. Sampling priorities are developed by stock as-sessment scientists, and port agents attempt to fulfill mini-mum sample sizes for length and age sampling.

Recreational Catch and Biological Sampling:(50 FTEs; 3 in-house, 47 contract)

Recreational fisheries in the Northeast are an importantsource of fishing mortality on regulated species such asstriped bass, bluefish, Atlantic cod, winter flounder, scupand black sea bass. Data on the magnitude of the recre-ational catch (numbers of fish caught) are derived fromthe Marine Recreational Fishery Statistical Survey(MRFSS) conducted by NMFS. This nationwide surveyemploys a contractor who deploys field personnel in a two-phase sampling scheme. The phone survey element of theprogram identifies the number of households participatingin recreational fisheries in the region. The intercept por-tion of the survey estimates catch numbers and speciescomposition of fishing trips by anglers. The data are com-bined to generate estimates of recreational landings anddiscards, by species.

Biological sampling of recreational catches is currently lim-ited to length composition and individual weight data. Inseveral states, the basic sampling scheme is augmented (in-creased sample sizes) in order to provide more precise catchestimates for important species.

In the Northeast region, the contractor utilizes 47 FTEsfor the phone survey and intercept portions of the survey.An additional three FTEs are required to administer theprogram and provide statistical oversight and management.

Research Vessel Surveys:(15 FTEs; 8 in-house, 7 contract or volunteers)

Standardized research vessel surveys provide the backbonefor stock assessment activities in the region, and have done

Commercial Catch& Biological Sampling

N=49

Recreational Catch& Biological Sampling

N=50

R/V SurveysN=15

Sea SamplingN=14

Age & GrowthN=11

DMSN=5

ConductingAssessments

N=16

N=4

Follow-upAnalysis

N=8

FTEs

Methods

34

so for nearly 40 years. The primary survey activities in-clude spring and autumn bottom trawl survey series (broad-based multispecies trawling surveys), a winter bottom trawlsurvey in the Middle Atlantic and Southern New Englandregion (primarily providing data for stocks such as sum-mer flounder, scup, squids, dogfish, and skates), a sea scal-lop dredge survey, an hydraulic dredge survey for surfclamand ocean quahog, and a trawling survey for northernshrimp. These surveys require scientific personnel equiva-lent to about 10 person-years to collect data in the field.Currently this requirement is fulfilled by assigned person-nel, volunteers from various agencies and universities, andstock assessment staff. In addition to field data collectionpersonnel, about 5 FTEs are required for data entry andauditing.

Additional research vessel survey data are provided bysome states (and by Canada for transboundary resources).More recently, cooperative NMFS-fishing industry surveyshave been undertaken to increase the spatial resolution ofsurveys for sedentary resources (scallop and surfclam), andto develop approaches to real-time management (squid).These activities have significantly increased the require-ment for at-sea personnel and for analysts to design thesurveys and analyze the results.

Sea Sampling:(14 FTEs; 3 in-house, 11 contract)

Most sea sampling (fishery observer) activities in the North-east Region are directed to assessing the impacts of fisher-ies on marine mammal populations of the region, includ-ing harbor porpoise in relation to sink gill net fishing. Thetotal sea sampling program includes about 56 FTEs; how-ever, the majority of the program is focused on monitoringfisheries interactions with protected species, includingmarine mammals and sea turtles. About 25% of the seasampling program is devoted to sampling for fishery-re-lated problems (e.g. stock assessment and compliancemonitoring for fished resources). The magnitude of theprogram is not sufficient to provide reliable estimates offishery catches and discards for all the region�s fisheries.Consequently, the limited resources of the fisheries-por-tion of the sea sampling program have been focused onseveral high priority problems: (1) discards of summerflounder in the Middle Atlantic and Southern New Englandtrawl fishery, (2) estimates of fishery catches and size com-position and bycatches of the sea scallop dredge fisheries,(3) estimates of finfish bycatches in the northern shrimptrawl fishery, and (4) monitoring of finfish bycatches inthe sea scallop dredge fishery in an area recently reopenedto fishing on Georges Bank.

Age and Growth:(11 FTEs; 8 in-house, 3 contract)

Analyses of year class strength and projections of stockabundance require measurements of the age-length and age-weight relationships of fishery resources. Because of highinterannual variation in recruitment survival, the abundanceand growth rates of adjacent year classes may differ greatly.Accordingly, where age-based stock assessments are per-formed, age information must be collected each year fromthe fisheries and from research vessel abundance surveys.

The NEFSC currently ages about 50,000 individual fishand invertebrates per year. These ageing studies supportpriority age-based stock assessments, depending on whichstock assessments need to be updated. In addition to age-ing work, biological studies supporting stock assessmentsinclude validation of ageing structures (e.g. fish otoliths orscales, clam shells) and factors controlling the rate ofgrowth and onset of sexual maturity.

Data Management Services:(5 FTEs; 4 in-house, 1 contract)

Data management activities (data entry, data auditing, da-tabase maintenance, custom programming for high prior-ity tasks, and support of data processing activities such asgeographical information systems) requires about fiveFTEs.

Stock Assessment Scientists:(28 FTEs; 23 in-house, 5 contract)

Stock assessment staff include individuals involved in dataassembly and quality control (technical functions), as wellas stock assessment model execution, development of newanalytical approaches to stock assessment methodology,computer programming of models, the provision of man-agement advice, and peer review of assessment science.These tasks can be divided into three broad categories de-scribing the general functions of stock assessment research:(1) conducting stock assessments, (2) developing andimplementing stock assessment methods, and (3) assess-ment follow-up activities including analysis of the impli-cations of alternative management strategies and other sci-entific input into the management process. Within theNortheast Region, approximately 16 FTE are involved inthe conduct of stock assessments, four in methods-relatedresearch, and eight in communicating assessment resultsand evaluating alternative management strategies. In allcases, no single individual exclusively performs one ofthese tasks; rather, individuals may function in all threeareas during part of an assessment cycle.

35

Apart from scientists at the NEFSC, other stock assess-ment professionals from several states, ASMFC, the tworegional Fishery Management Councils, Canada, and vari-ous academic institutions all contribute to the stock assess-ment and peer review processes in the Northeast Region.

Total (172 FTEs)

Based on the above, there is a minimum of 172 FTEs in-volved in various data collection, data management, dataanalysis, and communication functions related to the pro-

vision of scientific advice for 59 species or stocks of fish-ery resources in the Northeast. On average, this repre-sents about three staff per assessed species or stock, sothat the minimum additional staffing needed to assess anew species using existing infrastructure is at least three.However, if entirely new major programs are required (e.g.a new logbook reporting system, a new port sampling pro-gram, a new observer program, a new type of resource sur-vey), the Task Force estimated that as many as 20 newstaff may need to be added. These estimates also do notinclude administrative support staff.

37

IV. Resource Requirements

A. Programmatic Needs: Responses to questionnaires

In preparation for addressing the question of resource re-quirements for improving stock assessments, the TaskForce prepared a questionnaire and administered it toworking stock assessment scientists and to managers ofstock assessment programs. Not surprisingly, program-matic needs varied by program with, for example, someassessment groups having reasonable fishery-independentdata but poor fishery-dependent data, and others the re-verse. Thus, on average, all types of data commonly re-quired as input to assessment models are lacking (Figure9).

Overall, the need for fishery-independent indices of rela-tive abundance is the greatest of all, although less so inthe Northeast Center. Information on target catch appearsto be relatively the least problematic except that the North-east Center identifies it as its most important program-matic need (Figure 9). Similarly, the lack of a reliable fish-ery-independent index is the greatest impediment to pro-ducing high-quality stock assessments, particularly in thesoutheast, although less so in the northeast (Figure 10).

On average, lack of adequate data seemed to be onlyslightly more of an impediment than staffing levels tothe quality of assessments but again this varies consider-ably by program. Data and research needs for recreationalfisheries were low in Alaska where such fisheries are rela-tively much less important (Figure 10). Overall, observerprograms and analyses of biological samples were iden-tified as the two most important fishery-dependent dataneeds, with improved information on recreational catchmonitoring and commercial fishing effort being relativelythe least important, although still in need of substantialimprovement (Figure 11). Overall, tagging programs andstaff to process biological samples were identified as thetwo most important fishery-independent data needs, withtraining in species identification and improved under-standing of benthic habitat associations being relativelythe least important (Figure 12).

The general conclusion from these questionnaire sum-maries is that, overall, no single activity stands out asbeing disproportionately deficient; however, it is equallytrue that none of the inputs to stock assessments approachthe ideal situation of �no real need for improvement.�

Figures 9-12 give a qualitative indication of the variationin data and staffing needs between Science Centers, butthe raw data (not included with this report, but available onrequest) indicate that there is greater variability in data andstaffing needs between programs than there is betweenCenters.

B. Three Tiers of Assessment Excellence

The Task Force developed three scenarios to consider inthe analysis of the resources required to improve stockassessments. These are detailed below and summarizedin Figure 13.

Tier 1 � Improve stock assessments usingexisting data

(a) for core species, conduct assessments that aremore comprehensive, more thorough, more timely,better quality controlled, and better communicated;

(b) for species of currently �unknown� status,mine existing databases of research vessel surveydata and/or commercial and recreational statisticsfor archival information for new analyses toevaluate status determination criteria.

Tier 1 essentially addresses the question of what improve-ments in stock assessments can be made without initiatingnew data collection programs. Although the Task Forceagreed that new and/or expanded data collection programsare of paramount importance to the improvement of stockassessments, it was concluded that a certain limited amountcould be accomplished even in the absence of new pro-grams. Although most data collected by NMFS are ana-lyzed in a timely manner, there are many databases thathave not been examined exhaustively. In particular, theremay be considerable unanalyzed data for �minor� or non-target species. In some cases, there may even be histori-cal data that has never been computerized, thus necessi-tating �data-rescue� operations. One reason that some datahave been left unedited or unanalyzed is simply a lack oftechnical and quantitative staff to do the work. Inadequatestaffing levels have also compromised the timeliness, qual-ity and thoroughness of assessments conducted to date.Thus, the main requirements for Tier 1 are increased staff-ing levels, particularly database managers, statisticians,technicians, and assessment scientists.

38

Tier 2 � Elevate stock assessments to newnational standards of excellence

(a) upgrade assessments for core species to atleast Level 3;

(b) conduct adequate baseline monitoring for allfederally-managed species (including rarespecies)

The focus for Tier 2 is new or expanded data collectionand research initiatives. The task of upgrading assess-ments for core species to at least Level 3 would likely berelatively simple if there were adequate baseline moni-toring for all federally-managed species. A key questionis, �what is �adequate�?� The definition of �adequate�will differ by species or stocks and will depend on theirgeographic range, extent of migration, and magnitude ofinter-annual variations in stock size and recruitment. The

Task Force agreed, however, that in most cases adequatecoverage would require sampling throughout the rangeof a species or stock at least every 1-3 years, and prefer-ably at least every 1-2 years. For most species, fishery-independent research surveys are the method of choice;for some species, tagging experiments may be more prac-tical; and where neither of these are possible, fishery-de-pendent surveys may suffice. There are currently very fewstocks that can be characterized as having adequatebaseline monitoring (Appendix 1 and Table 1). In addi-tion, a minimal requirement for conducting ecosystem-based management and for fully satisfying the standardsset forth in the Sustainable Fisheries Act (e.g. standardsassociated with bycatch issues) is that there be adequatebaseline monitoring of all commercial and recreationalspecies and also all associated species, not just federally-managed species.

Average Response

Figure 9.

Programmatic needs averaged over responses from assessment scientists within each Science Center. An average response of 1 fora given Science Center indicates that the lack of a given program is a major impediment to producing credible assessments and has highpriority for improvement; 2 indicates a major impediment, but not amongst the highest priorities; 3 indicates adequate for accuracy butnot for sample size; 4 indicates that relatively fine tuning is needed; and 5 indicates that the current program is adequate with no realneed for improvement. Thus, for the five stacked histograms combined, a total of five would represent the greatest need and 25 wouldindicate the least need. The difference between 25 and the summed histograms is an overall indication of the need for improving thespecified programs. FI=Fishery-independent; FD=Fishery-dependent.

Programmatic Needs

0 5 10 15 20 25

AK

NW

SW

SE

NE

Target catch

Bycatch & discards

FD biology

FD effort

FI index

FI biology

Stock structure

Seasonal/spatial data

39

Tier 3 � Next generation assessments

(a) assess all federally-managed species or spe-cies groups at a minimum Level of 3, and all corespecies at a Level of 4 or 5;

(b) explicitly incorporate ecosystem considerationssuch as multispecies interactions and environmen-tal effects, fisheries oceanography, and spatial andseasonal analyses

The Task Force struggled to define reasonable limits toTier 3. The most recent Report to Congress on the Statusof Fisheries of the United States (NMFS 2001a) lists 905federally-managed stocks, most of which are not routinelymonitored, and many of which may not even be identifiedto species in commercial or recreational landings. The num-ber of data collection activities and staff resources requiredto enable 900+ assessments of stock status to be under-taken on a regular (e.g. annual) basis is enormous. Addi-

Figure 10.

Impediments to the quality of assessments averaged over responses from assessment scientists within each Science Center. Anaverage response of 1 for a given Science Center indicates that the quantity or quality of data and staff resources is a majorimpediment to producing credible assessments and has high priority for improvement; 2 indicates a major impediment, but not amongstthe highest priorities; 3 indicates adequate for accuracy but not for sample size; 4 indicates that relatively fine tuning is needed; and5 indicates that the current program is adequate with no real need for improvement. Thus, for the five stacked histograms combined,a total of five would represent the greatest need and 25 would indicate the least need. The difference between 25 and the summedhistograms is an overall indication of the need for improving the specified data collection programs or staffing levels. FI=Fishery-independent; FD=Fishery-dependent.

tionally, if associated species and other ecosystem con-siderations were to be taken into account, the task is mindboggling. It then becomes necessary to ask the question,what would be the utility of having 900+ annual assess-ments; is this a reasonable long-term objective? Wouldthis substantially enhance fisheries management, or arethere simpler ways of achieving a similar result? Certainly,it is hard to imagine that 900+ catch quotas would thereforebe set, monitored and enforced simultaneously.

From a management perspective, a more realistic aimwould be to manage only the primary (core) species bycatch quotas, effort controls, or similar high maintenancemanagement methods, and to manage other species usingclosed areas (e.g. marine protected areas, MPAs), closedseasons, gear restrictions and other indirect managementmeasures. From a stock assessment perspective, a morerealistic aim would be to assess groups of species fromwithin the same fishery or geographic area in an aggregateLevel 3 assessment, but to also have separate fishery-in-

Average Response

Impediments to Quality of Assessments

0 5 10 15 20 25

0 5 10 15 20 25

Catch

FD CPUE

Recreational data

Biology

FI index

Environment

Multispecies

Recreational research

Data collection staff

Database managers

Assessment scientists

Methods scientists

Training

Communication

AK

NW

SW

SE

NE

40

dependent indices of relative abundance that could be moni-tored over time to make sure that no individual species wasbecoming severely depleted. Nevertheless, it is obviousthat any reasonable attempt to even partially satisfy theobjective of assessing all federally-managed species at alevel of 3-5 will require substantial new or expanded datacollection and research initiatives, and staff to collect, man-age, process, and analyze the data, and to communicatethe results.

In essence, Level 4 assessments can be considered �state-of-the-art,� while Level 5 assessments are �next genera-tion assessments.� Level 4 stock assessments are the stan-dard to which NMFS Science Centers currently strive forthe stocks of primary importance. Level 4 assessmentscomprise analytical age, size, or stage-based calculationsthat provide relatively precise time series of stock abun-dance estimates, estimates of exploitation rates and thedistribution of the exploitation across size or age groups.From such analyses, short- and medium-term stock and

fishery projections and detailed analyses of alternativemanagement scenarios can ensue.

One goal of the Stock Assessment Improvement Plan is toincrease the proportion of stocks that can be evaluatedwith Level 4 stock assessments. This step alone will re-quire a major commitment of resources to enhance datacollection activities and analysis functions. However, it isimportant to consider enhancements beyond high qualitysingle species stock assessments, recognizing longer-termneeds of fishery management and emerging issues relatedto management of species assemblages, communities andecosystems. Clearly, there is increasing demand for in-formation to allow finer scales of management in space(geographic distribution) and time (seasonally, monthly,and even weekly) than are typically provided in Level 4assessments. In many cases, these needs are immediate,as managers attempt to manipulate the spatial and tem-poral pattern of fishing effort to change exploitation ratesand patterns on individual stocks, to harmonize the man-

Average Response

Fishery-dependent Data Needs

Figure 11.

Fishery-dependent data needs averaged over responses from assessment scientists within each Science Center. An averageresponse of 1 for a given Science Center indicates that a new or greatly expanded data collection program of the specified type wouldgreatly enhance the ability to produce accurate, precise, and timely assessments; 2 indicates that the program would help moderately;3 indicates that the program would only help marginally; and 4 indicates that the program would not help or is irrelevant. Thus, for thefive stacked histograms combined, a total of four would represent the greatest need and 20 would indicate the least need. Thedifference between 20 and the summed histograms is an overall indication of the need for improving the specified data collectionprograms or staffing levels.

0 5 10 15 20

Observer programs

Commericial catch monitoring

Recreational catch monitoring

Logbook program

Data management

Biological analysis

Sampling designs

Fishing effort

Effort & economic data

AK

NW

SW

SE

NE

41

Figure 12.

Fishery-independent data needs averaged over responses from assessment scientists within each Science Center. An averageresponse of 1 for a given Science Center indicates that a new or greatly expanded data collection program of the specified type wouldgreatly enhance the ability to produce accurate, precise, and timely assessments; 2 indicates that the program would help moderately;3 indicates that the program would only help marginally; and 4 indicates that the program would not help or is irrelevant. Thus, for thefive stacked histograms combined, a total of four would represent the greatest need and 20 would indicate the least need. Thedifference between 20 and the summed histograms is an overall indication of the need for improving the specified data collectionprograms or staffing levels. To group main headings (upper case labels) and subheadings (lower case), it is necessary to read frombottom to top on the y-axis.

Fishery-independent Data Needs

Average Response

AK

NW

SW

SE

NE

0 5 10 15 20

SAMPLING FREQUENCY

More routine surveys

Seasonal surveys

Spawning surveys

Increase sampling density

TRAWL SURVEY

Gear properties

Benthic habitat associations

Trawl mortality

Benthic impacts

LONGLINE

Gear properties


ACOUSTIC SURVEYS

Gear properties

Contaminant species

TAGGING

Archival tags

Tag retention studies

Tag recovery studies

SUBMERSIBLES

Line transect

Avoidance

Benthic impacts


RECRUITMENT STUDIES

Gear properties

Benthic habitat studies

PROCESS STUDIES

Oceanographic data

Intensify sampling

DATA MANAGERS

SAMPLE PROCESSORS

STATISTICAL ANALYSIS

SPECIES ID TRAINING

42

agement of co-occurring stocks, and to deal with alloca-tion issues. There is also growing interest and need forquantitative information on predator-prey and competitiveinteractions among managed stocks and associated spe-cies. Assessments incorporating biological interactions willbecome increasingly important because of the requirementsof the Sustainable Fisheries Act to maintain all managedstocks near or above biomass levels that can support MSY.The feasibility of achieving this simultaneously for allstocks warrants further investigation, as do the trade-offsbetween fishery yields and stock sizes that will accrue dueto manipulation of the abundances of interacting species.Such models are now available in limited situations, prima-

rily as research tools, but the next generation of assess-ment models will be required to allow more detailed man-agement scenario analyses of such trade-offs for a widerdiversity of situations.

Next-generation assessments are also envisaged as pro-viding the foundation for ecosystem-based management.While considerable work on incorporating ecosystem con-siderations into assessment models and management ad-vice is currently underway, both within and outside ofNMFS, ecosystem science is still in its infancy. Ecosys-tem research is also prohibitively expensive and labor-intensive. The U.S. Global Ocean Ecosystem Dynamics(GLOBEC) program on Georges Bank serves as an example.Planning for this program was initiated in 1991 with prelimi-nary studies in 1993, and a full program including broad-scale monitoring of physical and biological variables andfine-scale process-oriented studies was conducted overthe period 1994-99. During this period, research vesselswere deployed on Georges Bank for about 250 vessel-daysper year, with data collection and analysis involving about70 scientists, plus support staff, at an overall cost of about$5 million per year. It is projected to take at least anotherfour years to process all of the ichthyoplankton samples,analyze the data, and synthesize the results. Although thisprogram went well beyond simple monitoring of speciesand collected considerable physical oceanographic dataas well as investing in new technologies, for practical andlogistic reasons, the program focused only on a few targetspecies: cod and haddock (primarily only at the egg andlarval stages), and two species of calanoid copepods. Evenaccounting for the transferability of knowledge gained fromthe process-oriented studies and technological spin-offbenefits, a data collection program of this intensity wouldhardly be practical if the targeted species included all spe-cies inhabiting Georges Bank.

Thus, the goal of performing frequent individual assess-ments for all 900+ FMP species, incorporating ecosystemconsiderations for as many stocks as possible, and con-sidering the effects on associated non-FMP species, isprobably not realistic. It is highly likely that the cost ofconducting this amount of research would far exceed thelanded value of the fisheries. This would not necessarilymean that the overall benefit of such research was nega-tive because research on marine ecosystems has utilitybeyond simply providing advice on optimal harvest levels.However, there are many competing priorities for govern-ment spending. Even if this stock assessment improve-ment plan and related initiatives are aggressively pursuedand actively supported by stakeholders, it is unlikely thatNMFS will ever have sufficient data collection and analy-sis capabilities to conduct more than double the number ofassessments currently undertaken per annum, meaning that

Figure 13. Summary of the key features of the three Tiers ofAssessment Excellence.

TIER 3Next generation assessments

♦ Assess all managed species or species groupsat a minimum Level of 3

♦ Assess core species at a Level of 4 or 5♦ Explicitly incorporate ecosystem

considerations, including environmentaleffects, oceanography, and spatial analysis

TIER 2Elevate all assessments to new

national standards of excellence♦ Upgrade to at least Level 3 for core species

♦ Adequate baseline monitoringfor all managed species

TIER 1Improve assessmentsusing existing data

♦ More comprehensive for core species♦ Mine existing databases for

species of unknown status

43

some species will probably always be assessed either in-frequently or as part of a larger group. However, with suf-ficient resources, it will be possible to also conduct betterassessments for the core species. In particular, it may befeasible to anticipate conducting Level 5 assessments (in-corporating some but not necessarily all elements listedunder Level 5) for as many as 4-8 core species per region.Ecosystem-based research is also likely to yield usefulancillary information about associated species, as well asimproving our understanding of the dynamics of marineecosystems.

Another important future consideration for next-genera-tion stock assessment models is that people and groupsinfluenced by the results of such models (commercial fish-ers, recreational fishers, environmental groups, and man-agers) will increasingly request greater access to the dataand models themselves, and greater participation in datacollection and analysis functions. In the next generation,user-friendly models to analyze the implications of alter-native management strategies (e.g. stock projection mod-els simulating the biological and economic consequencesof various patterns of future catches or exploitation rates)should be developed and made available to the public sothat affected parties can conduct their own analyses of al-ternative management scenarios. While all of this is pos-sible with current technology and agency expertise, theresources required to develop the necessary interfaces withthe public at large are not inconsequential. Greater flex-ibility in analysis options should be one of the hallmarksof next-generation assessment models, as should accessto data and models over distributed computer networks.An important element of improving NMFS� stock assess-ments is planning for and moving forward with the next-generation of stock assessments immediately, consistentwith these considerations.

In conclusion, models addressing more species, and moredetailed spatial, temporal, environmental and species in-teractions questions will require significantly more pre-cise, timely and comprehensive fishery-dependent and fish-ery-independent data. Next generation models will beextremely data-intensive, requiring much-augmented, com-prehensive monitoring data. Gathering and analyzing suchdata will require even greater cooperation from harvest-ers, fish dealers, and others, more agency staff and fund-ing, and more partnerships and cooperative research pro-grams with other federal and state government agencies,academic institutions, private foundations, fishers, andenvironmental groups with a vested interest in similar orrelated data. Many such partnerships already exist (Ap-pendix 21), but many more are needed. Recent initiativesto develop cooperative research programs with the fishingindustry (Appendix 22) are showing considerable promise

as a mechanism for augmenting existing programs to col-lect data of relevance to stock assessments, and a NationalCooperative Research Program is now being developed. Itmay also be fruitful to pursue participation in broad-scaleprograms such as NOAA�s Ocean Exploration Program(Appendix 23), the Census of Marine Life (Appendix 24),and other initiatives involving science policy, data collec-tion and scientific research (Appendix 25).

C. Timeframes and Relationships Betweenthe Tiers

Attainment of the three Tiers of Assessment Excellenceinvolves both short and long-term horizons which, in turn,are dependent upon other complementary programs andinitiatives being put in place, and the ability to recruit quali-fied personnel for the various tasks at hand.

Tier 1: With adequate additional trained staff, most usefulwork based on existing data will probably have been com-pleted within 3-4 years, by which time new data from Tiers2 and 3 would hopefully be beginning to become avail-able for additional species. Tier 1 benefits will be almostimmediately obvious as data on species of currently �un-known� stock status are analyzed; however, moving cer-tain species from �unknown� to �known� status may notbe the highest priority. For example, improved analysisof major target stocks, currently overfished stocks, or newor expanded data collection programs for such species maytake precedence.

Tier 2: Contingent on initiation of needed new data collec-tion programs and appropriate additional staff, benefitswould become obvious within 5-10 years as time seriesdevelop to sufficient length to be of use in stock assess-ment models. There are also likely to be some immediatebenefits; e.g. immediate improvements in the knowledgeof the fine-scale distribution of some species and assem-blages which could improve management decisions.

Tier 3: Next generation assessments represent a long-term(10+ years) objective and investment because consider-able research and development is required and becausenew time series of consistent data collection must be initi-ated. In addition, Tier 3 is dependent upon an adequate,purpose-built fleet of dedicated research vessels, contin-ued development of advanced technology that will facili-tate sampling of marine organisms, and development ofpartnerships and cooperative research programs with otherfederal agencies, state agencies, private foundations, uni-versities, commercial and recreational fishing organizationsand individuals, environmental groups, and others with avested interest in collecting similar types of data, althoughoften for different purposes.

44

D. Region-Specific Needs to Achieve theThree Tiers of Assessment Excellence

The number of species covered by FMPs differs substan-tially between regions. This is less a reflection of regionaldifferences in species diversity or fishing intensity than itis of regional differences in the philosophy of which spe-cies to include in FMPs. For example, the Gulf of Alaskagroundfish plan includes 100 species, but most of the land-ings are comprised of only a dozen or so species; in con-trast, in the New England and Mid-Atlantic regions, lessthan two dozen of the nearly 200 shelf species are explicitlyincluded in FMPs. For the purposes of reporting on thestatus of U.S. fisheries, tracking progress in conserving orrestoring resources, and comparing region-specific needsand achievements, it would be useful to have greater con-sistency. However, for the purposes of this plan, the au-thors all approached the question of region-specific needsin a similar way, regardless of differences in regional phi-losophies about the degree of inclusiveness. Core speciesare those with the highest value, highest volume, or high-est profile. Minor species are those that contribute little ornothing to landings, but need to be considered in someway in an ecosystem context, regardless of whether or notthey are explicitly included in FMPs.

Unless otherwise specified, the current and required re-sources detailed in this section apply as of January 2000.In some cases, there have been several staff hired or con-tracted to perform assessment-related activities subse-quently. These are highlighted in the appropriate sections.


The following two sections contain an analysis of the cur-rent staffing and status of assessments in the northeast re-gion relative to defined assessment levels, and an analysisof the staffing resources necessary to meet the three Tiersof Assessment Excellence based on data and resourcescurrently used in the region. It should be noted that theseanalyses apply only to staffing levels and other resourcescontributed on the federal side; however, for many of theregion�s resources, data and analyses are undertaken bystaff in state marine fisheries organizations. Thus, exist-ing and required staffing resources should be consideredminimum.

NEFSC current situation

A total of approximately 172 staff involved in stock assess-ment related activities within the Northeast region (SectionIII, part I and Figure 8) currently provide advice on 59 man-aged or otherwise important species/stocks (Figure 14).

The distribution of these stocks is bimodal with respect to�assessment level,� with 24 stocks assessed at Level 3 andabove, and 35 stocks at Level 2 and below. This mix ofassessment quality and completeness is primarily a func-tion of historical interest in various species (e.g. ground-fish, summer flounder, and surfclams), an increasing needfor higher-level assessments to support management pro-grams, and new legal requirements for population biologydata. The situation is not static, with managers increas-ingly requesting more frequent assessment updates, withmore extensive �terms of reference.�

In particular, stock assessments are now often required toincorporate discussion and evaluation of �control rules�used by management to meet the requirements of theMSFCMA. These control rules provide managers with aformulaic approach to scientific advice, pre-specifying therelationship of target fishing mortality rates to biomassconditions in the stocks. The construction and testing ofcontrol rules makes use of absolute biomass and fishingmortality rate estimates, or proxies for these quantities, ifadequate approaches can be developed. As part of thiseffort, medium term simulations of the performance ofcontrol rules in recovering and maintaining stocks are re-quired so as to evaluate the efficacy of a proposed controlrule in meeting the 10-year or one generation time con-straints imposed by the MSFCMA. Center stock assess-ment scientists have been involved in developing proposedcontrol rule strategies for various species, and in scien-tific research for providing realistic simulations of the per-formance of stocks in relation to control rule management(e.g. simulating population status in the medium term us-ing various approaches for determining recruitment re-

Figure 14.

Number of stocks (N=59) assessed by assessment level at theNortheast Fisheries Science Center.

Assessment Level

Num

ber o

f Sto

cks

0

5

10

15

20

25

543210

2

20

13

3

19

2

45

sponses). As the need for more complex stock assessmentshas been increasing, so has the need to upgrade index-level assessments to assessments incorporating age/sizestructure to support sophisticated simulations of controlrule performance.

NEFSC programs and staffing required to meet thethree tiers of excellence

Based on the current distribution of stock assessment lev-els, data and technical limitations, and staffing in data col-lection and analytical tasks, the following represents ananalysis of augmented staffing levels required to meet thethree Tiers of Assessment Excellence for Northeast stockassessments (staffing increases by activity are summarizedin Table 2).

Tier 1: Improve stock assessments using existing data

The intent of upgrading assessments of Tier 1 is to morefully utilize existing information, to upgrade the synthesisof available data and to provide information to users on amore timely basis. In order to meet Tier 1 requirementsfor Northeast stocks, additional FTEs are required prima-rily in data management and synthesis activities (Table 2).

Data on species age and growth have been collected andarchived, but not analyzed, for many stocks currently as-sessed in the �index level� category. A total of five addi-tional FTEs in the Age and Growth activity will allow morecomplete biological data for assessments of some of thesestocks. Additionally, this would enable more timely pro-duction of age data.

Improving Tier 1 assessments will also require additionalstaff to better archive and extract sea sampling informa-tion, and biological sampling data collected from commer-cial fisheries. Data management support is needed to main-tain databases and improve access to a wider array of us-ers. A modest increment in stock assessment and supportstaff is required to produce more frequent and improvedassessments under this tier.

Tier 2: Elevate stock assessments to new nationalstandards

The major increment in FTEs required under the stock as-sessment improvement plan occurs when moving to Tier 2requirements to upgrade assessments for core species toat least Level 3 and for providing adequate baseline as-sessments for all managed species. In order to meet these

Table 2.

FTEs required to meet the three Tiers of Assessment Excellence by type of activity for the Northeast Fisheries Science Center. Numbersof FTEs in each category do not necessarily reflect the actual number of individuals involved in these activities, in that some individualsmay divide their time between several activities. Estimated current FTEs include in-house staff, contractors such as observers, and�other,� which includes state government biologists, and employees or contractors associated with various regional, national and interna-tional Commissions. Follow-up evaluations include the production of additional assessment outputs, evaluations of alternative man-agement strategies, and participation in plan development teams. Numbers should be cumulated across tiers.


2+1 3reiT

&hctaClaicremmoCgnilpmaSlacigoloiB 03 91 01 01&hctaClanoitaerceRgnilpmaSlacigoloiB 74 3 5 5

smargorPrevresbO 3 11 31 31

syevruStnednepedni-yrehsiF 8 7 01 5 5 2selpmaSlacigoloiBssecorP).cte,ytirutam,htworg,ega( 8 3 3 5 5 01 5

&tnemeganaMataDataDfognissecorperP 4 1 7 2 9 5

stnemssessAtcudnoC 31 4 3 2 1 3 5

hcraeseRsdohteMtnemssessA 3 2 1 3 5&stluseRfonoitacinummoC

snoitaulavEpu-wolloF 7 1 2 1 3 3

)stsitneicstnemssessA(latotbuS 32 5 3 6 3 9 31)sreht0(latotbuS 001 44 31 21 04 25 21

latoT 321 94 61 81 34 16 52

46

requirements, there needs to be major improvements in thequality and timeliness of commercial and recreational fish-eries data and required biological sampling. Currently, bio-logical sampling of catches is concentrated on a few corestocks, and sampling levels are barely adequate in manyother cases. Improvements in the collection, management,and synthesis of fishery dependent data are needed. Age-ing structures are not currently collected from recreationalcatches and this needs to be rectified. Likewise there is nouniversal sea sampling program providing routine estimatesof discards and bycatch from the region�s fisheries, andthis needs to be improved. Increased biological samplingunder these programs will require the processing of greaternumbers of ageing structures, and data entry and manipu-lation. For some stocks not currently indexed by researchvessel survey programs (e.g. tilefish, deep water fisheriesand components too deep for surveying in current pro-grams) additional types of fishery-independent data arerequired and will need to be developed. Additional stockassessment scientists are required to produce higher qual-ity and more frequent assessments called for under thisrequirement.

Tier 3: Next generation assessments

Next generation assessments are intended to provide moreexplicit accounting for biological and technological inter-actions, longer-range predictions and integration of bio-logical and environmental data. A major component ofthese assessments will be the incorporation of feeding datainto routine stock assessments and modeling and spatialdynamic models and data to examine the fine-scale impli-cations of alternative management strategies. In order tosupport such requirements, additional data collections forbiological analyses (e.g. stomach sampling) are required,as well as fishery oceanographic data bases, geographicalinformation systems, and data management necessary tosupport these activities.


The fisheries under the research jurisdiction of the South-east Fisheries Science Center are diverse in both the spe-cies being exploited and the fishing sectors prosecutingthese fisheries, including large recreational sectors andbycatch sectors, as well as commercial fisheries. In somefisheries, the productivity of many of the species beingexploited is low, supporting relatively small catches (thereare over 400 species within Southeast FMPs or interna-tional conventions). However, some of the species are ex-tremely valuable and many are very important to local com-munities. Also, in aggregate the species catches are sig-nificant and the fisheries often have the capability to ex-

ploit a variety of species, switching target species as con-ditions change. These characteristics result in unique re-search and management requirements.

SEFSC current situation

The Southeast Fisheries Science Center has unique re-source requirements to achieve each of the three Tiers ofAssessment Excellence. This results largely from the di-versity of fishery resources occurring within the Region.Current stock assessment efforts have focused on a smallnumber of core species (those of greatest public interestand often of greatest political importance due to conflictbetween constituents). Thus, detailed assessments are con-ducted on 10-15 stocks, annually. However, there are alarge number of stocks upon which little assessment workis done other than to monitor catches. The catches of anyindividual one of these unassessed stocks is often smalland of small socioeconomic significance; however, in ag-gregate they are an important part of the fisheries economicsector and fishing communities.

Fisheries of the southeast are managed by the South At-lantic Fishery Management Council, the Gulf of MexicoFishery Management Council and the Caribbean FisheryManagement Council through fishery management planson shrimp, reeffish, snapper-grouper, spiny lobsters, coastalpelagics, red drum, stone crabs, corals and others. Thenumber of FMPs requiring stock assessment data is in-creasing. Data collection in support of assessment of thesespecies comes through the SEFSC efforts and through jointagreements with the individual states (plus Puerto Ricoand the U.S. Virgin Islands) and with the Atlantic andGulf States Marine Fisheries Commissions. Additionally,since the recreational sector is large in many fisheries (insome cases larger than the commercial sector), several jointagreements have been made to obtain recreational catchdata from various survey mechanisms.

A major issue that impacts stock assessments in the south-east United States is bycatch, particularly discarded by-catch resulting from Gulf of Mexico shrimp trawlers. Themortality resulting from this activity impacts stocks of fishfor which there are directed fisheries, therefore limiting theproduction from those fisheries. There are also major con-cerns with the impacts of gill-net fisheries on marine mam-mals, and hook and release mortality in the substantialrecreational fisheries that exist in the region. Incorporatingbycatch estimates into stock assessments requires a newlevel of commitment to data collection through observerprograms. Initial ad hoc projects have been conducted toobtain estimates of bycatch, but the precision is lacking.

47

The Southeast Fisheries Science Center has the responsi-bility for providing the United States scientific support forassessing stocks of Atlantic tunas, swordfish and billfishin conjunction with the International Commission for theConservation of Atlantic Tunas (ICCAT) of which the UnitedStates is a signatory nation. Assessments of the tuna,swordfish and billfish stocks are conducted jointly withscientists from various nations. The Southeast FisheriesScience Center has the responsibility for monitoring catchand scientific data from throughout the U.S. Atlantic coastand report these to ICCAT. The U.S. scientists, also, take alead role in the joint assessment working groups withinICCAT�s scientific committee.

SEFSC programs and staffing required to meet thethree tiers of excellence


The first Tier of Assessment Excellence is a goal of improv-ing assessments with existing data for both core stocksand those stocks whose status is largely unknown. Thecore stocks for which detailed assessments are currentlybeing conducted can be improved, even with existing data.Improvements can be achieved largely through more com-prehensive characterization of the uncertainty associatedwith various management parameters arising from the as-sessment. Characterizing the uncertainty requires stochas-tic modeling activities which are time consuming both intheir development and in the actual running of the mod-els. Uncertainty characterization also requires extendedinteraction with managers and constituents in order toappropriately formulate the statistical questions.

The first Tier can be achieved for the �non-core� stocks bydeveloping and organizing the data bases necessary forfirst pass assessments for these species. This will requirestatistical determinations of catch by size and other rel-evant strata, the collation of biological data and the analy-sis of appropriate survey and catch-effort trend data. Inmany cases some data exist within Federal, State and aca-demic institute data bases on each of these aspects. But itremains to integrate the information and make �first-pass�assessments. These initial assessments are impor-tant for management, as they will allow initial overfish-ing/overfished determinations to be made; additionally, theresults will be extremely useful in guiding further scien-tific prioritization of data collection activities for thesestocks. Due to the large number of these stocks within thepurview of the SEFSC, this will require increased moni-toring by assessment scientists.

Tier 2: Elevate stock assessments to new national standards

The second Tier of Assessment Excellence expresses thegoal of upgrading assessments of core species to a level inwhich dynamic changes in stock abundance are estimatedand monitored over time; and that there should be a baselinemonitoring of all managed species. To achieve this Tier,expanded data collection activities and extensive moni-toring activities by assessment scientists will be required.Of particular importance is the need for fisheries-indepen-dent data. As noted above, catches for many stocks arerelatively small; therefore, assessments with adequate lev-els of precision will require monitoring of appropriateabundance indices. SEFSC scientists indicate the highimportance of developing fishery-independent indiceswithin their responses to the questionnaire (Section IV(A)). The scientists� responses also placed emphasis onobserver programs to address important issues of bycatch,discards, collection of biological data, and collection ofbetter effort data. Thus, Tier 2 efforts should focus ondeveloping and improving data collection mechanisms.Fishery independent efforts require extended ship timewhich is addressed in other initiatives. However, improve-ments will require more than simply conducting more trawlsurveys. Extensive research is needed to explore avenuesfor monitoring stocks and life stages of stocks that are notconducive to trawl surveys; for example, mackerels andother coastal pelagics; billfishes and tunas; and reef dwell-ing species. This will require creative interaction betweenassessment scientists, survey statisticians, ecologists andgear-specialists in order to design appropriate survey strat-egies. Additionally, second Tier goals will also require im-proved characterization of bycatch, discards and other fish-eries and biological data. Observer programs are essentialfor these activities. Management of statistically usefulobserver programs will require the close cooperation ofbiologists, assessment scientists, data managers and pro-gram management with the constituents.


The third Tier of Assessment Excellence expresses a goalof having minimal assessment levels (dynamic monitoringof abundance - production modeling) for ALL stocks withall core stocks being addressed by size, age, sex-struc-tured assessments with possible inclusion of ecosystemfactors. As noted above, the diversity of fishery stocksunder SEFSC purview indicates the importance of eco-system considerations. What effect are major ecosystemperturbations such as bycatch or environmental changeslikely to have on species distributions? Can species shiftsbe predicted even in a probabilistic sense? Can manage-ment strategies be devised to avoid chances of deleterious

48

socioeconomic consequences of species shifts? Can man-agement strategies be devised to achieve short term localobjectives of the fishers? These questions pose importantresearch goals. Steps to achieve these goals require exten-sive research, monitoring and data collection activities. Inparticular, spatial and temporal scales of data collectionwill need to be improved. This will require finer scale infor-mation on catches, survey abundances and oceanographicvariability. Additionally, the monitoring of a large numberof stocks (the components of the ecosystem) is needed todiscern patterns of variability.

Specific resource requirements are outlined in Table 3.Note that resource requirements are additive; i.e., require-ments for Tier 2 are additive to those in Tier 1.


The NOAA Fisheries Southwest Fisheries Science Center�sarea of responsibility encompasses a vast expanse of open

ocean and international waters, including the coastal wa-ters of California, parts of the Antarctic, and the HawaiianIslands, and the U.S. Territories of Guam and AmericanSamoa. The SWFSC is responsible for the research andmanagement of some of the nation�s most intriguing spe-cies, and is a major force in the nation�s effort to buildinternational cooperation for the stewardship of these spe-cies. Research extends over an area of more than 64.2 mil-lion square miles of open ocean � greater than 18 times thesize of the U.S. land mass � including 1.8 million squaremiles of EEZ. The Southwest Region is home to over 72protected marine species, and 153 fishery stocks, of whichthe status of 101 remains unknown. In addition to the com-plexity of the marine ecosystems in this region, the interna-tional and domestic mix of culturally diverse fishing com-munities present complex challenges for managing spe-cies, conducting research, and collecting data necessaryto ensure sustainable fishing practices. The collective valueof tuna, swordfish, sharks, and billfish from U.S. Pacificwater fisheries exceeds $1.5 billion annually. Near shore

Note that the above table specifies the labor resources (FTEs) needed to address the three Tiers of Assessment Excellence for stockassessment responsibilities. The labor resources have an associated cost which is not addressed in the table. However, in addition tothe labor costs there will be additional resources needed to conduct experiments to achieve Tiers 2 and 3. Vessel time on researchvessels will, of course, be important, but this is discussed elsewhere. Activities which would likely be prominent in achieving Tiers 2 and3 are archival tag experiments, stock identification sampling (mtDNA and microconstituents) and other activities.

Table 3.

FTEs required to meet the three Tiers of Assessment Excellence by type of activity for the Southeast Fisheries Science Center. Numbersof FTEs in each category do not necessarily reflect the actual number of individuals involved in these activities, in that some individualsmay divide their time between several activities. Estimated current FTEs include in-house staff, contractors such as observers, and�other,� which includes state government biologists, and employees or contractors associated with various regional, national and interna-tional Commissions. Follow-up evaluations include the production of additional assessment outputs, evaluations of alternative manage-ment strategies, and participation in plan development teams. Numbers should be cumulated across tiers.


2+1 3reiT

&hctaClaicremmoCgnilpmaSlacigoloiB 01 2 51 2 2 3&hctaClanoitaerceRgnilpmaSlacigoloiB 01 3 52 2 2 3

smargorPrevresbO 4 01 4 22 22 11

syevruStnednepedni-yrehsiF 8 7 2 2 6selpmaSlacigoloiBssecorP).cte,ytirutam,htworg,ega( 51 7 1 3 4 5

&tnemeganaMataDataDfognissecorperP 9 1 7 4 11 1

stnemssessAtcudnoC 8 2 4 3 7 3


snoitaulavEpu-wolloF 6 1 2 3 1


latoT 17 03 64 41 24 65 93

49

landings of sardines, mackerel, tuna, rockfish, and flatfishin California and Hawaii totaled 370 million pounds in 1998and were valued at $173 million. During the same year,aquaculture in the Southwest Region produced 30 millionpounds of fish and shellfish worth $55.6 million to growers.From the recreational perspective, a quarter of a millionsaltwater anglers, 40 annual fishing tournaments, andprized game fish like marlin, tuna, wahoo, and mahimahiplace Hawaii among the top 10 states adding significantlyto the economy from sport fishing. California, with over amillion recreational anglers, ranked second in the U.S.catching about 23 million pounds of fish.

The SWFSC supports two fishery management councils.The Western Pacific Fishery Management Council(WPFMC), located in Honolulu, Hawaii, manages the in-sular resources in the central and western Pacific. The Pa-cific Fishery Management Council (PFMC), located in Port-land, Oregon, manages fishery resources along the U.S. westcoast. Both councils manage highly migratory species.

SWFSC current situation � central and western Pacific

The SWFSC has assessment responsibility for 56 specieslisted in the FMPs under the jurisdiction of the WPFMC.In terms of domestic and international market value, �core�species include the large pelagic fishes (tunas and bill-fish), and in terms of political interest blue sharks are addedto the list. These species readily migrate far beyond theU.S. EEZ and into waters where they may be caught byfleets of other countries. Assessments of these stocks can-not be based solely on catch data within U.S. territorialwaters from U.S. domestic fleets. Stock assessment workmust be conducted in an international context, taking ac-count of all catches that affect the population being as-sessed. The two Atlantic coast science centers face a pros-pect somewhat similar to the SWFSC in this respect, al-though assessment-related procedures in the Atlantic havelong been defined by participation in well-established in-ternational agencies, such as ICCAT, that orchestrate col-lection of data and workshops involving scientists frommember countries who conduct comprehensive stock as-sessments. To date there are no comparable fishery man-agement agencies in the temperate and tropical Pacific be-yond that covered by the Inter-American Tropical TunaCommission (IATTC) in the eastern tropical Pacific.

In the absence of such management agencies in the centraland western Pacific, it has been necessary for SWFSC sci-entists to forge their own cooperative arrangements withfishery scientists in other parts of that region. This takesconsiderable time and effort. Various international coop-erative arrangements have operated on a relatively infor-mal basis for the purpose of assembling fishery data from

disparate sources and conducting stock assessments. Fore-most of late is the Standing Committee on Tunas and Bill-fish (SCTB) of the Secretariat of the Pacific Community(SPC) which for core tuna species has been fosteringfishery data collection and high level stock assessments,with the participation of an international cadre of fisheryscientists.

In the past few years, two prongs of diplomatic effort havebeen undertaken in the Pacific to formalize comprehensive,international management agencies for the region or a por-tion thereof. One, the Interim Scientific Committee (ISC),has met several times and established several subcommit-tees to promote assessment of various species and to es-tablish a comprehensive fishery data base. The other ef-fort, dubbed the Multilateral High Level Conference(MHLC), has met five times, is considerably more com-prehensive in its membership than the ISC, and has theambitious goal of establishing a fishery monitoring andmanagement agency by June 2000.

At present, for most of the major fleets harvesting tunaspecies in the central and western Pacific (including U.S.fleets), catch and effort data are available to NMFS (orother) scientists for the purpose of conducting stock as-sessments thanks to the work of the SCTB and similar in-formal cooperative arrangements. The resolution in timeand space is not always as fine as desired and there aresome holes in the data, particularly catch at size data.Fortunately, good tag return data exist for the four majortuna species (skipjack, yellowfin, bigeye, and albacore)which in conjunction with the fishery data have enabledthe high level stock assessments mentioned above. How-ever, as sophisticated as the assessments have been, theyhave not had the benefit of regular survey data or othertypes of fishery independent abundance indices becauseno regular scientific surveys are conducted for pelagic fish-eries in the region. Though scientific observer coveragehas been very poor, it is improving with the observer re-quirements of the South Pacific Forum Fisheries Agency(FFA). Observer coverage on U.S. purse seine vessels inthe region is 23%, and coverage of the Hawaii longlinefleet is less than 5%.

For domestic insular fisheries for demersal fishes and crus-taceans, the data coverage is mixed. Regular abundanceindex surveys for lobsters have been conducted in con-junction with release of tagged lobsters. In addition, therehas been occasional observer coverage of the lobster fleet.Commercial catch/effort data for demersal fish are collectedby the state of Hawaii, but data for significant amounts of�recreational� catch (much of it sold at roadside stands)are not collected.

50

The Honolulu Laboratory has seven stock assessment sci-entists, including those involved in conducting assess-ments, methods research, and follow-up activities such asinput to plan development, for the 56 species in theWPFMC FMPs. These scientists are additionally chargedwith investigating the magnitude and gravity of interac-tions between domestic fisheries and protected species ofturtles and sea birds. Assessment duties for tunas, bluemarlin, swordfish, and blue shark are shared with scien-tists from other countries and agencies around the region,but many of the other pelagic species are neglected throughnecessity.

SWFSC current situation � west coast

The SWFSC has assessment responsibilities for four FMPsunder the jurisdiction of the Pacific Fishery ManagementCouncil (PFMC). Under the Groundfish FMP, the SWFSCshares assessment responsibilities with the NWFSC. TheSWFSC produces the assessments and the NWFSC pro-vides data collection and overall coordination for 82groundfish species, including over 40 species of rockfishdistributed from Southern California to Canada. Only 26 of

the 82 groundfish species have been assessed, and almostnone from central California southward. Five species havebeen quantitatively assessed as overfished. This hascaused a crisis due to severely reduced catch allocations.Many unassessed species are thought to be overfished aswell, and there is at least one case where an unassessedstock may be threatened or endangered.

Under the Salmon FMP, the SWFSC has sole responsibil-ity for assessing and developing recovery plans for 10endangered salmon and steelhead runs from California af-fecting three species.

Under the Coastal Pelagics FMP, the SWFSC has sole re-sponsibility for assessing sardine, Pacific and jack mackerel,northern anchovy, and market squid. This is done in coop-eration with the State of California on an annual basis.

Under the Highly Migratory Species FMP currently beingdeveloped, the SWFSC will have sole responsibility forassessing six tuna and billfish species caught by fisheriesoriginating from the U.S. west coast. This will be doneusing the same agreements and mechanisms described

Table 4.

FTEs required to meet the three Tiers of Assessment Excellence by type of activity for the Southwest Fisheries Science Center. Numbersof FTEs in each category do not necessarily reflect the actual number of individuals involved in these activities, in that some individualsmay divide their time between several activities. Estimated current FTEs include in-house staff, contractors such as observers, and�other,� which includes state government biologists, and employees or contractors associated with various regional, national and interna-tional Commissions. Follow-up evaluations include the production of additional assessment outputs, evaluations of alternative manage-ment strategies, and participation in plan development teams. Numbers should be cumulated across tiers.

* A loosely determined number of collaborating assessment scientists at SPC, CSIRO, NRIFSF and elsewhere.


2+1 3reiT

&hctaClaicremmoCgnilpmaSlacigoloiB 1 6 5 4 4 2&hctaClanoitaerceRgnilpmaSlacigoloiB 2 2 2 1

smargorPrevresbO 7 7 7 7

syevruStnednepedni-yrehsiF 8 1 2 11 11 9selpmaSlacigoloiBssecorP).cte,ytirutam,htworg,ega( 12 2 2 6 11 71 41

&tnemeganaMataDataDfognissecorperP 91 6 3 8 01 81 51

stnemssessAtcudnoC 21 +01 * 6 8 41 01

hcraeseRsdohteMtnemssessA 3 2 2 3 5 5&stluseRfonoitacinummoC

snoitaulavEpu-wolloF 9 5 4 9 3

)stsitneicstnemssessA(latotbuS 42 0 +21 31 51 82 81)sreht0(latotbuS 65 51 41 41 54 95 84

latoT 08 51 +62 72 06 78 66

51

above in the section on central and western Pacific and willinclude Mexico. Additionally, this FMP will include fourpelagic shark species. None of these have been assessed.The complexities of coordinating international assessmentsare similar to those discussed for the central and westernPacific.

The SWFSC is currently assessing the white abalone popu-lation which has been petitioned to be listed as endan-gered under the Endangered Species Act. This speciesneeds to be assessed throughout its range, which requirescooperation with Mexico.

The California recreational fishery bridges FMPs and com-plicates management, research and assessment efforts. Thisfishery is composed of both commercial passenger fishingvessels and private fishers and generates effort in millionsof days annually. It targets many of the same species ascommercial fleets, and is highly significant economically.

SWFSC programs and staffing required to meet thethree tiers of excellence

The SWFSC currently has 24 assessment scientists with atotal of 80 staff in the assessment processes (Table 4).The SWFSC is aided in commercial catch, recreationalcatch and biological sampling by various state agencieswhich independently or by contract conduct sampling. Thejump to Tier 1 requires a large increase in stock assessmentscientists, necessitated by the current low staffing of as-sessment personnel at the SWFSC. Movement to Tiers 2and 3 is highlighted by the need for increases in methodsresearch along with additional assessment capacity.


Moving to the first Tier of Assessment Excellence, improv-ing assessments with existing data, could be readilyachieved by additional SWFSC scientists for many spe-cies. Current data collections, including indices of abun-dance and key biological data exist with various sources �state, federal, and international � and could be preparedfor assessment use in relatively short order. In the South-west Region, considerable numbers of species have notyet been assessed or have been inadequately assessed.These include some high profile species currently fishedsuch as striped and blue marlins, all of the pelagic sharkscaught in the Pacific coast HMS fisheries, and several tunaspecies such as skipjack and bigeye tunas. The rockfishesin central and southern California have large numbers ofunassessed species including some mainstays of the rec-reational fishery such as Pacific bonito, vermillion rockfishand black abalone which are expected to be declared over-fished once Tier 1 assessment are completed, or, as in the

case of the white abalone move directly from unassessedto endangered. The jump to Tier 1 requires appreciableincreases in database managers and analysts to retrieve,audit and preprocess the data; biological technicians toprocess archived samples; and stock assessment sci-entists to conduct assessments, explore new method-ologies appropriate to data-poor situations, and com-municate the results (Table 4).


Moving or upgrading assessments to Tier 2 where dy-namic changes in stock abundance can be assessedand monitored for core species and all managed speciesare monitored will require expanded data collection aswell as extensive monitoring by assessment scientists(Table 4). SWFSC scientists are engaged in developingadvanced technology survey methodologies including,for example, ROV strip census, advanced acoustics, LI-DAR strip census and integrated acoustic and net sur-veys for krill-sized organisms. These methods, as wellas more established methods, will form the core for fish-ery-independent data monitoring. Considerable effortwill be focused on providing the basic biological pa-rameters needed to move assessments to age and sizebased methods from current Tier 1 efforts. Genetics willplay an important role in developing early life stage in-dices from fishery-independent survey methods suchas continuous underway egg and larval sampling forbiomass, which was pioneered at the SWFSC. Coordi-nation of creative interactions between the various stockassessment specialities will require careful management.


Tier 3 moves to the goal of providing basic assess-ments for all stocks with core stocks using age/size/sexstructured methods and considering ecosystem effects.To reach this goal for core stocks, extended research toestimate key biological parameters will be needed. Thiswill require substantially increased scientific effort (Table4). The SWFSC at its Pacific Fisheries EnvironmentalLaboratory is engaged in developing environmental datasets related to decadal climate shifts and shifts in oceanproductivity, and researching methodologies for incor-porating these effects into assessment models. Cur-rently, SWFSC scientists have provided a managementmodel to the Pacific Council which uses temperature asa forcing factor for determining harvest guidelines. Ex-tension of these emerging methods for incorporatingecosystem effects will require interdisciplinary coop-eration and facilitation between assessment scientistsand other disciplines.

52


NWFSC current situation

The Northwest Fisheries Science Center has lead respon-sibility for assessment of west coast groundfish and evalu-ation of recovery options for Pacific salmon. The demandsfor accurate scientific investigations for both groups ofspecies are high and increasing. Groundfish and salmonare managed according to Fishery Management Plans de-veloped by the Pacific Fishery Management Council(PFMC). Although the NWFSC has the lead role in coordi-nating assessment information for both FMPs, there aremajor contributions by other NMFS Science Centers andby the state fishery agencies of California, Oregon andWashington.

The status of Pacific salmon species on the west coast hasbeen reviewed under provisions of the Endangered Spe-cies Act and 26 populations (Evolutionarily SignificantUnits) have been listed as threatened or endangered. Atremendous effort is being mounted by the NWFSC andthe SWFSC to develop salmon recovery plans that incor-porate all aspects of human and natural risks to salmon.Nearly all of the salmon escapement monitoring and runforecasting is based on in-river information and is done bythe state and tribal agencies. These results are used by theSalmon Technical Team (STT) of the PFMC to developharvest options for consideration by the PFMC. Becausethe assessments of salmon are primarily conducted by otheragencies, and because the primary west coast salmon ac-tivity occurs under Protected Species, the salmon researchand monitoring needs are not considered further in thisdocument.

For groundfish, only 26 of the 82 species have ever beenquantitatively assessed. Of these 26 species, several haveexperienced severe declines. Harvest rates, climate, andassessment precision all contributed to this decline. Thedefault harvest rate during most of the 1990s (35% spawn-ers per recruit), while conservative by global standards,was overly optimistic during what has become a 20-yearregime of poor ocean productivity. The precision and fre-quency of stock assessments did not allow forecasting themagnitude or duration of the decline in recruitment untilseveral stocks had already crossed into an overfished state.As of 2001, rebuilding plans are being developed for sevengroundfish species. Even among the 26 assessed species,there are some for which there has not been sufficient in-formation to adequately determine their status with respectto overfishing thresholds. There are concerns that othersof the 60+ species with unknown status may be in dangerof overfishing. Further, some populations of groundfish in

Puget Sound have declined to such low levels that theirstatus was reviewed in 2000 for potential listing under theESA.

The majority of shoreside groundfish catch monitoring isdone by the state agencies with coordination through thePacific States Marine Fisheries Commission which main-tains a centralized database of fisheries data (PacFIN). Mostresource surveys are conducted by NMFS, with the trien-nial bottom trawl and hydroacoustic surveys providing amajor source of data for most assessments. Approximatelysix groundfish stock assessments are conducted each yearby NMFS, state agencies, and others. The NWFSC coordi-nates a stock assessment review process in conjunctionwith the PFMC�s SSC, that involves external peer-review-ers and public input.

Passage of the MSFCMA strengthened the mandate toimprove the west coast stock assessment capability. As-sessments need to be conducted for more of the ground-fish species. The level of uncertainty in groundfish as-sessments and the current information on low productiv-ity for these species needs to be combined in a soundprecautionary approach to managing these species. Re-building plans, which are expected to take more than 10years, need to be developed and subsequently monitoredfor several of these long-lived species. All of these will beextraordinarily difficult given the lack of a dedicated re-search vessel for these resources and the low level of cur-rent resource survey efforts. Further, increased stock as-sessment effort will primarily tell us what is occurring tothese species. Knowing what is only the first step. In orderto develop a better understanding of why these changesare occurring, programs need to be developed to investi-gate the role of decadal scale changes in ocean climate,and the role of ecosystem shifts such as the major increasein pinniped abundance that has occurred off the west coast.

NWFSC programs and staffing required to meet thethree tiers of excellence

The great diversity of habitat, life history, and knowledgefor west coast groundfish defies simple description of thedata needs for improvement. The 82 species have a collec-tive distribution which spans 1300 miles of coastline andfrom estuaries out to at least 1500 m bottom depth. Somespecies are schooling midwater, others are on the benthiccontinental slope, and others are associated with high-relief nearshore habitat. Species with the greatest accumu-lation of relevant stock assessment data tend to be thosethat have historically been targeted by the trawl fisheryand are amenable to either trawl or hydroacoustic surveys.Species that have the greatest data needs tend to be those

53

that are associated with high relief habitat and are subjectto growing commercial and recreational hook and line fish-eries. Today, only 26 of the 82 groundfish species haveever been assessed, and many of these assessments havehad insufficient data to allow adequate determination ofthe status of the species.


Bringing all west coast groundfish species to a Tier 1 levelwill require additional stock assessment, data processingand ecological staff to make the best use of the limitedexisting data. Some groundfish assemblages have no fish-ery dependent or fishery independent index of abundanceand limited biological sampling from the fisheries.

The Tier 1 focus of stock assessment modelers needs to beon developing a first-cut assessment for all species so thatany overfishing can be identified and corrected. There arethree general areas of improvements. One area will be inthe development and application of relevant assessmentmethods for more of the species that do not have sufficientdata to support current data-hungry quantitative assess-ment methods. This will require innovative use of stockassessment, biological and ecological data so that infor-mation from better known species can be used to developproxies for poorly studied species. A second area of im-provement is the development of assessment modelingprotocols that better quantify and communicate the uncer-tainty in current assessments. Such improved models willstructure implementation of a more formal precautionaryapproach to harvest management. A third area of improve-ment is in the spatial integration of fishery and surveydatabases, particularly through advances in linkage of fish-ery logbooks, landings data, and fishery biological samples.


Medium-term improvements in major data sources can leadto substantial improvements in assessment precision withinabout 10 years. These include major programs such as pe-riodic resource assessment surveys, more comprehensivefishery logbook programs and at-sea monitoring of totalcatch, collection of genetic stock structure data for morespecies, and evaluation of fish association with particularhabitats. Beyond routine monitoring, survey effort alsoneeds to be devoted to studies that will improve under-standing of how environmental and other factors affectefforts to standardize surveys. Studies are needed to in-vestigate bycatch mortality and gear impact studies. Manyof these medium-term efforts are large scale and expensive,but have the greatest likelihood of significantly improving

the accuracy of the assessments and our ability to con-duct assessments for all assemblages of groundfish.Current efforts are far from meeting Tier 2 assessmentneeds because:

1. The NWFSC has no dedicated fishery research ves-sel to do standardized resource assessment surveys orother field research;

2. Surveys to assess most of the continental shelf rock-fish and lingcod are conducted only triennially, yet sev-eral of these species are overfished and their rebuildingplan calls for a biennial assessment;

3. A small coastwide observer program to assess bycatchand total mortality of target species was not implementeduntil 2001, yet estimates of discard for some target spe-cies range up to 30%;

4. Fishery monitoring has historically focused on thetrawl fleet. There are no fishery logbooks and insuffi-cient fishery-dependent data for the hook&line fisherywhich accounts for the majority of many nearshore rock-fish species catch. Further, there are few if any fishery-independent data from which to assess the status ofthese species.


Further improvements in assessments can be made byincreasing the frequency and precision of fishery-inde-pendent surveys, and by increasing the number of spe-cies for which there is age composition data from thefishery and surveys. However, major improvements inour ability to forecast future stock conditions and toprovide assurance of ecologically safe harvest strate-gies will require qualitatively different kinds of informa-tion. Among these longer term efforts are recruitmentsurveys that will directly forecast changes in fish abun-dance, climate studies to provide longer-term predic-tions of average recruitment levels, and ecosystem stud-ies that will provide better understanding of the interac-tions among species and with their habitat. For westcoast groundfish, recruitment surveys are particularlyrelevant for species such as whiting which have tre-mendous variation in recruitment and recruit to the fish-ery at a young age. With a recruitment survey, we canbetter adjust harvest levels to track these short-termnatural fluctuations in abundance. Recruitment surveysare also relevant for the very long-lived species thathave delayed recruitment to conventional surveys andthe fishery. Here the recruitment surveys will provideadvance notice of longer-term shifts in abundance

54

caused by shifts in average recruitment levels. Climatemonitoring and fishery-oceanography investigations willhelp interpret these shifts in recruitment and further ad-vance predictive capability. The result of these ecosystemstudies will be a better assessment of the ecological impactof fishing, better understanding of the impact of factorssuch as the increased abundance of piscivorous pinni-peds, and potential adjustment of fishing strategies to ob-tain the best multispecies yields from the system.

Specific resource requirements for west coast groundfishare outlined in Table 5. The information labeled �cur-rent� in Table 5 describes the situation in January 2000.A partial step towards meeting Tier 2 needs occurred in2001 when the NWFSC received funding to establish asmall west coast groundfish observer program and con-duct coastwide trawl and hydroacoustic surveys.


The Alaska Fisheries Science Center is responsible for Gulfof Alaska, Bering Sea and Aleutian Island groundfish as-sessments. The Center conducts activities in support ofthese assessments that include fishery independent andfishery dependent data collection programs, and fisheriesoceanographic studies. The Center is also responsible forconducting fishery independent surveys and research insupport of Pacific salmon and Alaskan crab assessments.

Alaskan groundfish and crab are managed according toFishery Management Plans developed by the North Pa-cific Fishery Management Council (NPFMC). For the BeringSea / Aleutian Islands region, assessment scientists cur-rently contribute to thirteen annual groundfish assess-ments: walleye pollock, Pacific cod, Atka mackerel, yellow-

Table 5.

FTEs required to meet the three Tiers of Assessment Excellence by type of activity for the Northwest Fisheries Science Center. Numbersof FTEs in each category do not necessarily reflect the actual number of individuals involved in these activities, in that some individualsmay divide their time between several activities. Estimated current FTEs include in-house staff, contractors such as observers, and�other,� which includes state government biologists, and employees or contractors associated with various regional, national and interna-tional Commissions. Follow-up evaluations include the production of additional assessment outputs, evaluations of alternative manage-ment strategies, and participation in plan development teams. Numbers should be cumulated across tiers.

1. This table is limited to resources devoted to stock assessment of groundfish and other marine fish. Significant additional NMFS andstate resources are devoted to work on salmonids.2. �Partner� column contains minimum PSMFC and WA, OR, CA personnel working on groundfish. Many of these are supported throughfederal grants, including PacFIN.3. The in house staff column represents the total number of positions as of January 2000. New funding in FY2001 is allowing develop-ment of an observer program and expansion in survey and assessment programs. Approximately 15-20 Tier 1+2 positions will be filled.# includes 25 observers and infrastructure hired in FY 2001* excludes 25 observers hired in FY 2001; includes a proposed additional 30 contract observers (20 in Tier 2 and 10 in Tier 3)


2+1 3reiT

&hctaClaicremmoCgnilpmaSlacigoloiB 02 6 6&hctaClanoitaerceRgnilpmaSlacigoloiB 7 6 6

smargorPrevresbO 52 # 13 * 13 * 21 *

syevruStnednepedni-yrehsiF 7 2 11 11 01selpmaSlacigoloiBssecorP).cte,ytirutam,htworg,ega( 3 5 1 5 6 8

&tnemeganaMataDataDfognissecorperP 2 1 91 5 4 9 2



snoitaulavEpu-wolloF 3 1 3 2 5 7 1


latoT 81 33 95 31 47 78 93

55

fin sole, Greenland turbot, arrowtooth flounder, rock sole,flathead sole, other flatfish, Pacific ocean perch, other rock-fish, sablefish and squid and other species. For the Gulf ofAlaska region, 11 assessments are produced annually:walleye pollock, Pacific cod, Atka mackerel, thornyhead,slope rockfish, pelagic shelf rockfish, demersal shelf rock-fish, arrowtooth flounder, other flatfish, sablefish and otherspecies. AFSC staff assist in the development and reviewof stock assessments for Bering Sea crab stocks. Staff rep-resent the Center on numerous technical and decision mak-ing boards including the NPFMC Groundfish Plan Teamsand Scientific and Statistical Committee, the North PacificAnadromous Commission, the North Pacific Halibut Com-mission, and international technical committees dealingwith trans-boundary stocks such as Pacific hake and BeringSea pollock.

Bycatch limits for several species have been imposed toensure that individual species quotas are not exceeded.Species that are designated as prohibited species includeAlaskan crab (e.g. Tanner crab, blue and red king crab andsnow crab), Pacific halibut, and some stocks of Pacificsalmon (chinook, pink, sockeye, chum, coho and steelhead).Retention for sale of prohibited species is prohibited todissuade any targeting by groundfish fishers. Gulf of Alaskaand Bering Sea Aleutian Islands groundfish fisheries alsohave some bycatch and discard of unmarketable speciesand small sized fish that is typical of any multispecies fish-ery. There is some discard of marketable fish caused by theNPFMC management system for the groundfish fishery.When bycatch limits are exceeded for a species, the spe-cies can no longer be retained to discourage further catchof this species.

In-season catch composition is monitored by a major fish-ery-dependent data collection program. Catch is moni-tored by an observer program and shoreside data collec-tion. Roughly 30,000 observer days (equivalent to 114FTEs) are expended annually to collect data from the NorthPacific groundfish fishery. All vessels capable of hostingan observer may be required to do so at the vessel�s ex-pense. As currently implemented, vessels over 125 feetlength overall (LOA) are required to have an observer onboard at all times when ground-fishing, vessels of 60 to124 feet LOA are required to have observers on-board 30%of the time, and vessels under 60 feet LOA are generallyexempt from the requirements for observer coverage. Mostof the fishing vessels operating in the Bering Sea and Aleu-tian Islands exceed the 125 foot limit, while most of thefishing vessels in the Gulf of Alaska are smaller than 125feet. The recreational harvest of groundfish in Alaskan wa-ters is a minor component of the total catch. Observers col-lect biological data such as otoliths, length frequencies, stom-ach samples and maturity stage for a variety of species.

Conducting fishery independent surveys in Alaskan wa-ters requires a major investment of shiptime and person-nel. The continental shelves off Alaska make up about 74%of the total area (2,900,785 km2) of the United States con-tinental shelf. The region is marked by adverse seasonalconditions that necessitate sophisticated equipment to en-sure the safety of the crew and the accomplishment of thesurvey mission.

AFSC stock assessment scientists conduct research to im-prove the precision of their assessments, and provide tech-nical support for the evaluation of potential impacts of pro-posed fishery management measures. Research activitiesare designed to improve the quality of stock assessmentsand to expand the scope of assessments to quantify theecological impact of fishing on the Gulf of Alaska and BeringSea ecosystems. Stock assessment scientists often servea dual role acting as the scientific interface between theAlaska Fisheries Science Center and the North Pacific Fish-ery Management Council (NPFMC).

Numerous laws govern the implementation of fisheries infederal waters. AFSC staff often conduct research to evalu-ate the impact of fishing to comply with these legal re-quirements. The MSFCMA directs NMFS stock assess-ment scientists to provide annual status evaluations for allspecies managed under the NPFMC FMPs. AFSC stockassessment scientists also provide analytic assistance onmany current fisheries management issues such as researchactivities leading to implementation of precautionary re-source management, consultations and Biological Opin-ions regarding protected resources, and NEPA impactanalyses regarding the effects of fishing on the marineenvironment.

The National Standard Guidelines for overfishing state that�If environmental changes affect the long-term produc-tive capacity of the stock or stock complex, one or morecomponents of the status determination criteria must bere-specified.� This requirement necessitates new researchon the mechanisms underlying shifts in production. Stud-ies have demonstrated that several groundfish, crab andsalmon stocks exhibit shifts in production that show markedsimilarity to the time scales of distant atmospheric forcingphenomena such as the El Nino Southern Oscillation andthe Pacific Decadal Oscillation. To determine whether shiftsare due to human actions or environmentally induced shiftsin the productive capacity of a stock or stock complex re-quires new research to investigate the mechanisms under-lying the apparent response of key species to decadal scalechanges in ocean climate.

Several species protected under the provisions of the En-dangered Species Act are present in the region. Among

56

these, the western stock of Steller sea lion has been listedas endangered. A tremendous effort is being mounted byAFSC and the North Pacific Regional Office to develop aSteller sea lion recovery plan that incorporates all aspectsof human and natural risks to this marine mammal popu-lation. Principal prey items in the sea lion diet include Atkamackerel, Pacific cod, and walleye pollock. These speciesare also targets of large commercial fisheries. Efforts areunderway to explore methods to reduce the potential forcompetition between commercial fisheries and Steller sealions at crucial times of the year.

Court challenges underscore the demands on AFSC staffto conduct new research surveys, process oriented research,and assessment activities to improve our understandingof the mechanisms underlying recent declines in the Stellersea lion population and the potential role of commercialfishing in limiting its recovery. These decisions also un-derscore the need for additional staff to evaluate manage-ment alternatives to provide reasonable and prudent alter-natives to current fishing practices.

AFSC current situation � Gulf of Alaska groundfish

Among the 100 groundfish species covered by the GOAFMP of the NPFMC, 67 are assessed at a Level 1 or better(Appendix 1). These species have been the targets of fish-ery monitoring and resource survey programs that providethe basic information for quantitative stock assessments.Not all these assessments have the same level of informa-tion and precision. Of the 67 assessed species, only 8 areassessed using staged base models (Level 4 or above). Inthe case of 91 of the 100 species covered by the FMP,there is insufficient information to determine whether ornot the stocks are overfished or approaching an over-fished condition. Some of these species are targets ofdeveloping fisheries.

The 100 groundfish species can be roughly broken intofour assemblages based upon their adult habitat and co-occurrence in the fishery. This breakout will facilitate dis-cussion of fishery monitoring and resource survey programs:

1. Midwater schooling- Walleye pollock, eulachon andsquid are amenable to acoustic survey methods. Walleyepollock supports a midwater trawl fishery with annual catchnear 100,000 mt.

2. Deep slope (mostly trawlable habitat on shelf break andcontinental slope extending out to about 1500m bottomdepth) includes primarily sablefish, dover sole, shortrakerand rougheye rockfish, shortspine thornyheads, longspinethornyheads, Pacific grenadier. This assemblage supports

a valuable trawl fishery, plus sablefish is a target of pot andhook&line fishers.

3. Demersal Shelf (mostly trawl caught species on conti-nental shelf and upper slope, but many species occur overrocky habitat and some species have significant off-bot-tom tendencies). This assemblage includes rockfish spe-cies, flatfish, Atka mackerel and Pacific cod. The fisheryis trawl for most species; however, Pacific cod is taken byhook&line and pot gear.

4. Pelagic shelf rockfish (mostly in high relief habitat) in-cludes several of rockfish species.

AFSC current situation � Bering Sea / Aleutian Islands(BSAI)

Among the 145 groundfish species covered by the BSAIFMP of the NPFMC, 133 are assessed at a Level 1 or better(Appendix 1). These species exhibit great diversity in lifehistory traits. Many have been the targets of fishery moni-toring and resource survey programs that provide the ba-sic information for quantitative stock assessments. Not allthese assessments have the same level of information andprecision. Of the 133 assessed species, only 15 are as-sessed using staged base models (Level 4 or above). In thecase of 128 of the 145 species covered by the FMP, there isinsufficient information to determine whether or not thestocks are overfished or approaching an overfished condi-tion. Some of these species are the target of developingfisheries.

The 145 groundfish species can be roughly broken intofour assemblages based upon their adult habitat and co-occurrence in the fishery. This breakout will facilitate dis-cussion of fishery monitoring and resource survey programs:

1. Midwater schooling- Walleye pollock, eulachon andsquid are amenable to acoustic survey methods. Walleyepollock supports a midwater trawl fishery with annual catchnear 1,000,000 mt.

2. Deep slope (mostly trawlable habitat on shelf break andcontinental slope extending out to about 1500m bottomdepth) includes primarily sablefish, Greenland turbot,shortraker and rougheye rockfish, and shortspinethornyheads. This assemblage supports a valuable trawl fish-ery, plus sablefish is a target of pot and hook&line fishers.

3. Demersal Shelf (mostly trawl caught species on conti-nental shelf and upper slope, but many species occur overrocky habitat and some species have significant off-bot-tom tendencies). This assemblage includes rockfish, flat-

57

Table 6.

FTEs required to meet the three Tiers of Assessment Excellence by type of activity for the Alaska Fisheries Science Center. Numbers ofFTEs in each category do not necessarily reflect the actual number of individuals involved in these activities, in that some individuals maydivide their time between several activities. Estimated current FTEs include in-house staff, contractors such as observers, and �other,�which includes state government biologists, and employees or contractors associated with various regional, national and internationalCommissions. Follow-up evaluations include the production of additional assessment outputs, evaluations of alternative managementstrategies, and participation in plan development teams. Numbers should be cumulated across tiers.

fish, Atka mackerel, crab and Pacific cod. The fishery forthis assemblage is primarily trawl for most species how-ever, Pacific cod are taken by hook&line and pot gear, andcrab are taken with pot gear.

4. Pelagic shelf rockfish (mostly in high relief habitat) in-cludes several of rockfish species.

Three crab stocks are currently listed as overfished: BeringSea Tanner crab, Bering Sea snow crab and Saint MathewsIsland Blue King Crab. Rebuilding plans need to be de-veloped and subsequently monitored for these crab stocks.Building such plans will be difficult given the lack of lifehistory and stage based information for these resources.Increased stock assessment effort will primarily elucidatethe underlying factors contributing to recent declines inproduction.

AFSC programs and staffing required to meet the threetiers of excellence


Bringing all Alaskan groundfish and crab species to a Tier1 level will require additional stock assessment, data pro-cessing and ecological staff to make the best use of exist-ing data. New personnel identified under Tier 1 (Table 6)would utilize existing fishery dependent and fishery inde-pendent data to facilitate the development of assessmentsfor several new species and to improve existing assess-ments to the extent possible. To accomplish the first partof this objective, additional staff are needed to constructassessments for species where historical data is spotty oruncertain. AFSC scientists are exploring modeling ap-proaches that draw on life history information from simi-lar species to parameterize first generation assessmentsfor poorly studied species. Additional assessment scien-

* Observer program includes shore-side samplers.+ Recreational data not applicable.


2+1 3reiT

&hctaClaicremmoCgnilpmaSlacigoloiB * 6

&hctaClanoitaerceRgnilpmaSlacigoloiB +

smargorPrevresbO 02 411 12 12 01

syevruStnednepedni-yrehsiF 5.36 1 72 91 91 01selpmaSlacigoloiBssecorP).cte,ytirutam,htworg,ega( 12 3 7 7 8 51 12

&tnemeganaMataDataDfognissecorperP 5.42 3 3 6 6 1


hcraeseRsdohteMtnemssessA 8 1 3 3 3 6 4&stluseRfonoitacinummoC

snoitaulavEpu-wolloF 7 5 5 5 01 3


latoT 451 221 45 13 66 79 15

58

tists could assist in developing overfishing criteria wheninformation regarding the status of the stock is missing orintermittent. Assessment scientists and statisticians areneeded to assist in reviewing the sampling design of pro-posed or existing fishery independent and fishery depen-dent sampling programs. This research effort would re-quire a retrospective analysis of existing data to evaluatethe efficiency of the current data collection program and tomake recommendations for improvements in sampling de-sign. Additional analytical staff could conduct research tobetter quantify and communicate the uncertainty in cur-rent assessments. Improved models will structure imple-mentation of a more formal precautionary approach to har-vest management. A final area of improvement is in thedevelopment of assessment models that fully utilize exist-ing information on top down (predator/prey) influences ontime trends in natural mortality, and bottom up on marinesurvival at early life stages in spatially explicit modes. AFSCis well positioned to advance this type of state-of-the-artstock assessment. The combination of a long history ofdata collection on the food habits of groundfish in theEastern Bering Sea and Gulf of Alaska make the develop-ment of models that model top down forcing a realisticgoal. Likewise, the long history of fisheries oceanographicprocess oriented research supported by the FisheriesOceanography Coordinated Investigations provides thenecessary knowledge of lower trophic level forcing requiredto implement a fully coupled model.

Tier 2: Elevate stock assessments to new national standards

Medium-term improvements in major data sources canlead to substantial improvements in assessment preci-sion within about 10 years. These include major programssuch as periodic resource assessment surveys, expand-ing and improving at-sea monitoring of total catch, col-lection of genetic stock structure data for more species,and evaluation of fish association with particular habi-tats. Beyond routine monitoring, research should be de-voted to studies of factors that may influence surveystandardization, and development of cost-effective sur-vey technologies that are not susceptible to environmen-tal influences on standardization.

Additional staff would be required to achieve a Tier 2 level ofanalysis for BSAI and GOA groundfish. Tier 2 envisions thatassessments of core species would be upgraded at least Level3 and would provide adequate baseline assessments for allmanaged species. Fishery dependent and fishery indepen-dent data collection are needed to achieve Tier 2. These staffmembers would be responsible for compiling and analyzingdata for species currently managed as species groups (e.g.other flatfish, other rockfish and other species).

GOA assessment needs to achieve Tier 2

In the current implementation of the observer program,observers monitor catch and collect biological informa-tion on 70 of the 100 groundfish species in the Gulf ofAlaska. Several minor species are classified into generalcategories. Skates are almost always recorded as �skateunidentified,� with very few exceptions between 1990-1998. In the Gulf of Alaska, at least 80% of the recordedsculpin catch by year is recorded as �sculpin unidentified,�with the remainder of the catch identified to the genuslevel. Only small amounts (<2%) of the sculpin catch eachyear were identified to species. Likewise, octopus andsquid are generally not identified to species in the ob-server database. Octopus can only be recorded as �octo-pus unidentified,� or �pelagic octopus unidentified.�Eulachon and capelin are recorded to species more oftenthan sculpins but in 1998, approximately 80% of theircatch was recorded as �smelt unidentified.� Monitoringthe catch of these minor species would require additionalstaff to train and implement an expanded observer program.

Groundfish populations are routinely monitored by fish-ery independent surveys. A longline survey is conductedannually for sablefish. A gulf-wide trawl survey of theshelf areas of the Gulf of Alaska has been conducted on atriennial basis since 1984. Current operating plans callfor future surveys to occur on a biennial basis. An acous-tic survey of a major spawning concentration of walleyepollock in Shelikof Strait is conducted on an annual ba-sis. These surveys provide a calibrated abundance mea-sure (Level 2 or above) for only 4 species (Appendix 1).These surveys provide an index of abundance for 83 spe-cies (Appendix 1). To achieve Tier 2 level analysis addi-tional effort should be devoted to obtaining and analyz-ing the life history of characteristics of species capturedin the longline or trawl surveys (e.g. regional differencesin growth, maturity, and habitat association).

For species such as small soft-bottom roundfishes(sculpins, poachers, eelpouts, and skates) the existing timeseries of trawl survey data is inconsistent because of dif-fering levels of species identification. Starting in 1999 thisproblem was nearly eliminated because all survey vesselshad new species identification guides that included pho-tos of all known species. Species identification has there-fore been greatly increased with very little cost. An exist-ing problem for these species in the GOA, however, is thatthey are likely to have very low catchability by the surveytrawls and it is uncertain how well research vessel CPUEtracks stock size. Assessment of these species could beimproved using auxiliary trawl experiments to measure es-capement under the footrope.

59

Many species of rockfish are not well sampled becausethey occur in areas that are too rough to be sampled withour usual survey nets and, additionally, some species (e.g.Pacific Ocean perch, northern rockfish, and dusky rock-fish) are extremely patchy and not likely to be well sampledin the present bottom trawl survey. Considerable workhas been done in attempt to develop a rockfish specificsurvey, but the best approach has yet to be developed. Toimprove the survey assessment of rockfish we need moreresearch on gear design and sampling techniques. Oncethe appropriate technique is developed, it will undoubt-edly require a distinctly different survey design than is nowused and could not be incorporated into the current nor-mal survey operations. Additional work will be requiredto develop appropriate techniques for the semi-pelagicspecies.

Fishery independent collections of age, length frequencyand size at maturity are obtained for the core species (about20 species, mostly rockfish and slope species, split be-tween GOA, AI, and EBS). Expanding the age collec-tions to include the remaining species would require col-lecting otoliths for additional species on surveys and couldbe accomplished without a large increase in money ormanpower. However, additional staff would be needed toconduct the age determinations. Obtaining size at matu-rity information would require a considerable increase inresearch cruises to collect species at a time that is closeenough to spawning so that mature or recently spent fishare easily recognized. Most survey or research cruises atthe AFSC are currently conducted in the summer, aftermost species have completed spawning. In addition, asampling strategy must be worked out so that a sufficientnumber of small and immature fish are collected.

Acoustic-trawl surveys in Alaska conducted by the Re-source Assessment and Conservation Ecology Division(RACE) focus on walleye pollock as a target species. Allaspects of survey design (e.g. area, timing, sampling inten-sity, etc.) are devised to assess the distribution and abun-dance of pollock. Pollock is ideally suited for acoustic as-sessment due to its semi-demersal nature, widespread dis-tribution, and tendency to form monospecific aggregations.During RACE acoustic-trawl surveys, other pelagic fishspecies are encountered in very low numbers. Existingacoustic data could provide some information on eulachonoccurrence observed during the 1980-1998 winter-springShelikof Strait surveys. Expanding the current acousticprogram to routinely monitor eulachon would require a sig-nificant effort, including both staff and vessel time. Addi-tional trawling would be needed and extended tracklinesmay be necessary.

Application of an acoustic-trawl survey approach to otherFMP species (e.g. rockfish, capelin, squid, etc.) has beensuccessful under certain circumstances, but would requirea substantial amount of work (e.g. literature reviews andfeasibility studies) merely to make a good guess of theresources required. A significant amount of preliminaryresearch would be necessary to simply estimate the staffand funding necessary to fund each project.

Bering Sea / Aleutian Islands assessment needs toachieve Tier 2

Bering Sea and Aleutian Island groundfish populations areroutinely monitored by fishery independent surveys. Alongline survey for sablefish is conducted in alternate yearsin either the Bering Sea or the Aleutian Islands. Ground-fish trawl surveys of the Eastern Bering Sea shelf havebeen conducted on an annual basis since 1979. Ground-fish trawl surveys are conducted on a triennial basis in theAleutian Islands region. Current operating plans call forfuture surveys of the Aleutian Islands region on a biennialbasis. Acoustic surveys of major spawning concentrationsof walleye pollock near Bogoslof Island are conducted onan annual basis. An acoustic survey of walleye pollock onthe Eastern Bering Sea shelf has been conducted on a tri-ennial basis since 1979. These surveys provide a calibratedabundance measure (Level 2 or above) for 76 species (Ap-pendix 1). These surveys provide an index of abundance(Level 1) for an additional 47 species (Appendix 1).

As in the case of the Gulf of Alaska, the existing time se-ries of trawl survey data for species such as small soft-bottom roundfishes (sculpins, poachers, eelpouts, andskates) may provide inconsistent results because of differ-ing levels of species identification. This problem has beenaddressed through the addition of new species identifica-tion guides which included photos of all known species.However, as in the GOA, it is likely that these species havevery low catchability by the survey trawls in the AleutianIslands region. Assessment of these species could be im-proved using auxiliary trawl experiments to measure es-capement under the footrope.

Many species of rockfish are not well sampled by the Aleu-tian Island trawl survey because they occur in areas thatare too rough to be sampled with our usual survey netsand, additionally, some species (e.g. Pacific Ocean perchand northern rockfish) are extremely patchy and not likelyto be well sampled in the present bottom trawl surveys.Nevertheless, the current Aleutian Island trawl survey doesprovide an index of abundance for several rockfish spe-cies, and rockfish age data are collected during the sur-veys. Additional research is needed to design a calibratedsurvey for rockfish.

60

Staffing needs to expand fishery independent collec-tions of age, length frequency and size at maturity werediscussed in the section on GOA fishery independentsurveys.

During Midwater Assessment and Conservation Engineer-ing (MACE) acoustic-trawl surveys, other pelagic fish spe-cies are encountered in very low numbers. Existing acous-tic data could provide some information on eulachon oc-currence observed during the 1980-1998 winter-springBogoslof Island surveys. Expanding the current acousticprogram to routinely monitor eulachon would require asignificant effort - including both staff and vessel time.Additional trawling would be needed and extendedtracklines may be necessary.


A substantial increase in stock assessment staff would berequired to achieve a Tier 3 level of analysis for BSAI andGOA groundfish. Tier 3 assessments would account forboth biological and technological interactions and inte-gration of biological and environmental data that may leadto more reliable long-range predictions. To accomplish thisgoal necessitates the implementation of fisheries oceano-graphic research programs for a broad spectrum of spe-cies. At the current time AFSC primarily supports fisheriesoceanographic research on walleye pollock. Likewise, ad-ditional staff would be required to provide information onpotential trophic interactions between species. Assess-ment scientists would be required to develop a broaderspectrum of assessment modeling tools to address the com-plex interactions envisioned under Tier 3. In addition tothe complex modeling activities envisioned for core spe-cies, additional stock assessment scientists would be re-quired to conduct basic assessment functions for all spe-cies covered by the FMP.

E. Summary: National ResourceRequirements

Current FTEs and FTEs required to achieve the objectivesof the three Tiers of assessment Excellence are summarizedby Science Center, Tiers of Assessment Excellence, andactivity in Figure 15. Similar but more detailed summariesare provided in Tables 7 and 8. Table 7 sums the FTE re-quirements for Tiers 1, 2 and 3 by major activity for all fiveNMFS Science Centers combined. Almost three times asmany additional staff are needed to collect, manage andprocess data, as compared to additional staff needed toconduct and communicate stock assessments, to evaluatealternative management strategies, and to conduct research

into assessment methods. By far the greatest overall needis for observers for Tier 2, particularly in the Southeast,Northwest and Alaska Science Centers. The second great-est overall need is for staff to participate in fishery-inde-pendent surveys (note, however, that this is contingent onthe acquisition of adequate Fisheries Research Vessels, asoutlined in the NOAA Fisheries Data Acquisition Plan,NMFS 1998c; Appendix 3).

Table 8 summarizes the total FTEs requirements for Tiers 1,2 and 3 for each Science Center and all Centers combined.In terms of current in-house staff, contract employees, andothers who provide assessment data (e.g. state govern-ment biologists, and employees or contractors associatedwith various regional , national and international Commis-sions), the Alaska Center is the largest with 330 FTEs, theNortheast Center is second with 188, the Southeast Centeris third with 147, the Southwest Center is fourth with 121,and the Northwest Center has 110. (These numbers applyto the baseline of January 2000, except where otherwisenoted in Tables 2-6; in particular, the Northwest Centertotal includes 25 observers hired in FY2001). Consideringthe sum of Tier 1 and 2 requirements, the Alaska, North-east, and Southeast Centers require additions of about 30-40% to existing staff, whereas the Southwest Center re-quires an addition of about 70%, and the Northwest Centeran addition of about 80%. To calculate the approximatecosts of new FTEs to satisfy Tiers 1, 2 and 3, a multiplyingfactor of $150,000 per annum was used. This number takesinto account salary and benefits, travel, training, equip-ment and individual Information Technology needs (al-though not the core systems needed for data managementand communications); i.e. the multiplier covers everythingexcept major infrastructure, particularly new workspace andbuildings that may be required.

The numbers of additional staff indicated in Tables 2-8may seem staggering, but these numbers have been care-fully thought through by the Task Force members. Theysimply represent the increasing demands being placed onNMFS to assess more stocks more frequently, and withgreater accuracy, precision and timeliness; to incorporateassociated non-target species and other ecosystem consid-erations into the analyses; and to evaluate a wider array ofmanagement options on increasingly finer temporal andspatial scales. In addition, as outlined in the Introduction,the FTE requirements detailed here are meant to comple-ment other related NMFS plans such as the Data Acquisi-tion Plan (NMFS 1998a; Appendix 3), which is primarilyconcerned with the costs of operating dedicated fisheryresearch vessels and purchasing charter boat days at sea;the Stock Assessment Toolbox Plan (Appendix 4); the Cen-ter for Independent Experts Program (Appendix 5); the Pro-

61

posed Implementation of a Fishing Vessel Registration andFisheries Information Management System (Appendix 8);the NMFS Bycatch Plan (Appendix 9); the National Ob-server Program (Appendix 10), the Social Sciences Plan(Appendix 11), the Advanced Technologies Working Group(Appendix 12), and relevant fisheries oceanography initia-tives (e.g. Appendix 13). In order to develop a comprehen-sive ecosystem approach to fisheries stock assessmentsand management, and to estimate the actual costs of imple-menting ecosystem-based management (EBM), these andrelated plans, initiatives and activities should be mergedinto an umbrella plan.

F. The Benefits of Implementing the StockAssessment Improvement Plan

The benefits of implementing the Stock Assessment Im-provement Plan are numerous and diverse. With adequateadditional trained staff, existing databases can be minedfor material to improve analyses for major target stocksand for currently overfished stocks, and to develop newanalyses for stocks of currently unknown status. The ben-efits arising from Tier 1 alone will, however, be limitedbecause the most important need is for new and expandeddata collection programs. Ultimately, these will lead togreater numbers of stocks being assessed with higher fre-quency, and greater accuracy, precision and timeliness.

FTEs (including contractors and partners)FTEs (including contractors and partners)

Figure 15.

Summary of FTE requirements by Science Center, Tiers of Assessment Excellence, and activity. FI=Fishery-independent.

Current Situation Tier 1 Requirements

Tier 3 RequirementsTier 2 Requirements

0 50 100 150 200

Commercial catch

Recreational catch

Observers

FI surveys

Process samples

Data management

Assessments

Methods research

Follow-up analysis

0 20 40 60 80 100

Commercial catch

Recreational catch

Observers

FI surveys

Process samples

Data management

Assessments

Methods research

Follow-up analysis

0 20 40 60 80 100

Commercial catch

Recreational catch

Observers

FI surveys

Process samples

Data management

Assessments

Methods research

Follow-up analysis

AKNWSWSENE

0 20 40 60 80 100

Commercial catch

Recreational catch

Observers

FI surveys

Process samples

Data management

Assessments

Methods research

Follow-up analysis

62

Table 8.

Total Full-Time Equivalents (FTEs) required to meet the three Tiers of Assessment Excellence for each Science Center and all Centerscombined. Estimated current FTEs include in-house staff, contractors such as observers, and �other,� which includes state governmentbiologists, and employees or contractors associated with various regional, national and international commissions. Numbers should becumulated across tiers.

Table 7.

FTEs required to meet the three Tiers of Assessment Excellence by type of activity for all NMFS Science Centers combined. Numbersof FTEs in each category do not necessarily reflect the actual number of individuals involved in these activities, in that some individualsmay divide their time between several activities. Estimated current FTEs include in-house staff, contractors such as observers, and�other,� which includes state government biologists, and employees or contractors associated with various regional, national andinternational Commissions. Follow-up evaluations include the production of additional assessment outputs, evaluations of alternativemanagement strategies, and participation in plan development teams. Numbers should be cumulated across tiers.


2+1 3reiT llAsreiT

CSFEN 321 94 61 81 34 16 52 68CSFES 17 03 64 41 24 65 93 59CSFWS 08 51 +62 72 06 78 66 351CSFWN 81 33 95 31 47 78 93 621

CSFA 451 221 45 13 66 79 15 841sETFdemmuS 644 942 102 301 582 883 022 806

)K051$xETF($$ K054,51$ K057,24$ K002,85$ K000,33$ K002,19$

ytivitcA tnerruCrehtO/tcartnoC/esuoh-nI 1reiT 2reiT 2+1reiT 3reiT sreiTllA

&hctaClaicremmoCgnilpmaSlacigoloiB 14 72 64 22 22 5 72&hctaClanoitaerceRgnilpmaSlacigoloiB 75 6 43 51 51 4 91

smargorPrevresbO 43 061 4 49 49 04 431

syevruStnednepedni-yrehsiF 5.49 81 93 84 84 73 58selpmaSlacigoloiBssecorP).cte,ytirutam,htworg,ega( 56 81 71 02 23 25 35 501

&tnemeganaMataDataDfognissecorperP 5.85 21 52 33 02 35 42 77

stnemssessAtcudnoC 94 4 +32 62 62 25 42 67

hcraeseRsdohteMtnemssessA 51 2 5 9 11 02 22 24&stluseRfonoitacinummoC

snoitaulavEpu-wolloF 23 2 8 51 71 23 11 34

)stsitneicstnemssessA(latotbuS 69 8 +63 05 45 401 75 161)sreht0(latotbuS 053 142 561 35 132 482 361 744

latoT 644 942 102 301 582 883 022 806

63

Incorporation of ecosystem considerations into the analy-ses will facilitate analysis of trade-offs between harvestingtarget species and protecting non-target species such asmarine mammals. The enhanced data collection and analy-sis activities proposed herein will also result in more accu-rate projections of future stock status under various alter-native management strategies, and will enable evaluationof an increasingly wider array of management options onfiner temporal and spatial scales, both of which will im-prove the basis for management decisions.

An improved knowledge base, improved ongoing data col-lection programs, and more comprehensive models shouldreduce the frequency of risk-prone management decisions,which have been common in many regions of the UnitedStates to date. This in turn will enable higher catches onaverage, at less risk to fisheries resources. The risk of non-target marine species becoming rare or extinct should alsobe considerably diminished, particularly in comparison tothe current situation in which species could potentially bedisappearing without us even being aware of it.

Overall, implementation of the Stock Assessment Improve-ment Plan will result in a greatly improved knowledge basefor marine species, and a better basis for risk-averse man-

agement decisions which will result in fewer depleted oroverfished stocks and greater stability and profitability inthe fish harvesting sector. However, it should be notedthat improved knowledge and enhanced stock assessmentcapability will not by themselves result in fewer overfishedstocks and a more stable fishing industry; there must be aconcomitant commitment to responsible fisheries manage-ment and fisheries policy development.

Another benefit of implementing the SAIP will be to im-prove relations between NMFS and other line offices withinNOAA, other federal agencies, state agencies, academia,the commercial and recreational fishing industries, andenvironmental groups by promoting cooperative researchand other types of partnerships. NMFS� own programsand those developed through such partnerships shouldalso result in spin-offs in terms of monitoring informationand research that can provide input into other programs;for example, risk and damage assessments. The resultingdatabase of spatial and temporal distributions of marinespecies, associations between species, oceanographic vari-ables, and habitat relationships will also be an invaluablesource of raw material with which to develop and test hy-potheses about population dynamics and ecosystem struc-ture and function.

65

V. Recommendations

1. NMFS should aggressively pursue a course of ac-tion focusing on new budget and staffing initia-tives to modernize its data collection and assess-ment capabilities. As a minimum, NMFS shouldattempt to bring stock assessment science to atleast Tier 2 (Section IVB), and should initiate dia-log both within house and with the public to deter-mine how far-reaching and comprehensive Tier 3should be. This will require hiring or contractingconsiderable numbers of additional qualified stafffor data collection, data processing, data manage-ment, stock assessments, and evaluation of alter-native management strategies, to ensure adequatedata and analyses on which to base conservationand management decisions, now and into the future.

2. In order to improve the credibility of its stock as-sessment science, in addition to acquiring the re-sources needed to produce the best possible sci-ence, NMFS must improve its public image, bothwith constituents and within NOAA itself. Thereappears to be little awareness that NMFS employsthe largest collection of world-renowned fisheriesscientists of any agency, university, or other orga-nization worldwide, and that fisheries science is afield where new and useful methodologies havemostly originated within government agencies (in-cluding those of foreign governments), rather thanwithin academia.

3. NMFS also needs to make fishers, politicians, andthe public aware of the benefits of truly precau-tionary management which will reduce the risks ofoverexploiting fisheries resources and associatedspecies, and will ultimately lead to greater stabilityin the fishing industry.

4. Another avenue of public awareness which NMFSshould pursue is to educate and discuss with in-terested parties (especially constituents and con-gressional aides) the implications of calls to incor-porate ecosystem considerations into fisheries as-sessment and management. In particular, NMFSshould request input on what different groups ofpeople actually mean by �ecosystem consider-ations,� and then jointly evaluate the costs andbenefits of adopting such approaches. NMFSneeds to work harder to align public expectationwith reality.

5. NMFS needs to be more proactive in communicat-ing the fact that the methodologies employed toconduct stock assessments are far less problem-atic than is the quality, quantity, and type of dataavailable for analysis. NMFS needs to seek out anddevelop cooperative arrangements with stakehold-ers to improve the quality, quantity, and type ofdata provided.

6. In order to make substantial progress towards col-lecting the data needed to improve stock assess-ments, particularly next generation assessments, itis essential that NMFS develop further partnershipsand cooperative research programs with other fed-eral agencies, state agencies, private foundations,universities, commercial and recreational fishing or-ganizations and individuals, environmental groups,and others with a vested interest in collecting simi-lar types of data, although often for other purposes.Many such partnerships already exist (Appendix21), but many more are needed. Programs involv-ing cooperative research with the fishing industry(Appendix 22) should continue to be developedand expanded as mechanisms for providing datarelevant to improving the quality of stock assess-ments.

7. In order to enhance progress in the development ofnew models and methodologies for conductingstock assessments, performing risk analyses andstock projections, and constructing multispeciesand ecosystem models, NMFS must free up moretime for existing quantitative staff to pursue suchresearch and engage more fully in professional de-velopment activities which, in turn, implies the needto also hire or contract additional qualified quanti-tative staff.

8. In order to ensure a future supply of quantitativescientists to perform stock assessments and relatedactivities, NMFS must augment existing programsthat fund graduate study in appropriate fields.

9. In order to maintain the high caliber of current ana-lytical staff, NMFS must develop a comprehensivetraining program to enhance the quantitative skillsof in-house staff.

66

10. In order to develop more comprehensive and inte-grated future budget initiatives geared towardsmodernizing fisheries assessments and manage-ment, NMFS should prepare an umbrella plan thatintegrates all relevant existing documents on thesethemes; for example, the current Stock AssessmentImprovement Plan, the NOAA Fisheries Data Ac-quisition Plan (Appendix 3), the NMFS StrategicPlan for Fisheries Research (NMFS 2001b), the Pro-

posed Implementation of a Fishing Vessel Regis-tration and Fisheries Information Management Sys-tem (Appendix 8), the NMFS Bycatch Plan (Appen-dix 9), the National Observer Program (Appendix10), the Social Sciences Plan (Appendix 11), theAdvanced Technologies Working Group (Appen-dix 12), and relevant fisheries oceanography initia-tives (e.g. Appendix 13).

67

NMFS 1997a. NOAA Fisheries Strategic Plan. NationalMarine Fisheries Service, NOAA, Department of Com-merce. Silver Spring, Maryland.

NMFS 1997b. Report to Congress on the Status of Fisher-ies of the United States. National Marine Fisheries Ser-vice, NOAA, Department of Commerce. Silver Spring,Maryland.

NMFS 1998a. NOAA Fisheries Data Acquisition Plan.National Marine Fisheries Service, NOAA, Departmentof Commerce. Silver Spring, Maryland.

NMFS 1998b. Report to Congress on the Status of Fisher-ies of the United States. National Marine Fisheries Ser-vice, NOAA, Department of Commerce. Silver Spring,Maryland.

NMFS 1998c. NMFS Strategic Plan for Fisheries Research.National Marine Fisheries Service, NOAA, Department ofCommerce. Silver Spring, Maryland.

NMFS 1998d. Managing the Nation�s Bycatch: Programs,Activities, and Recommendations for the National MarneFisheries Service. National Marine Fisheries Service, U.S.Department of Commerce, Silver Spring, Maryland. 174 p.

NMFS 1999a. Report to Congress on the Status of Fisher-ies of the United States. National Marine Fisheries Ser-vice, NOAA, Department of Commerce. Silver Spring,Maryland.

NMFS 1999b. Ecosystem-Based Fishery Management: aReport to Congress by the Ecosystem Principles Advisory

Panel. National Marine Fisheries Service, U.S. Departmentof Commerce, Silver Spring, Maryland. 54 p.

NMFS 2001a. Report to Congress on the Status of Fisher-ies of the United States. National Marine Fisheries Ser-vice, NOAA, Department of Commerce. Silver Spring,Maryland.

NMFS 2001b. NMFS Strategic Plan for Fisheries Research.National Marine Fisheries Service, NOAA, Department ofCommerce. Silver Spring, Maryland. 88 p.

NRC 1998a. Improving Fish Stock Assessments. NationalResearch Council. National Academy Press. Washington,D.C. 177 p.

NRC 1998b. Review of Northeast Fishery Stock Assess-ments. National Research Council. National AcademyPress. Washington, D.C. 128 p.

NRC 1999. Sustaining Marine Fisheries. National ResearchCouncil. National Academy Press. Washington, D.C. 164 p.

NRC 2000. Improving the Collection, Management, andUse of Marine Fisheries Data. National Academy Press.Washington, D.C. 236 p.

Restrepo, V.R., Thompson, G.G., Mace, P.M., Gabriel, W.L.,Low, L.L., MacCall, A.D., Methot, R.D., Powers, J.E., Tay-lor, B.L., Wade, P.R. and Witzig, J.F. (1998) Technical Guid-ance on the Use of Precautionary Approaches to Imple-menting National Standard 1 of the Magnuson-StevensFishery Conservation and Management Act. NOAA Tech-nical Memorandum NMFS-F/SPO-31. 54 p.

References

68

The National Task Force for Improving Fish Stock Assess-ments would like to thank the NMFS Science Center as-sessment scientists who participated in the questionnairesurvey. We are also grateful for the many insightful com-ments received from various scientists at the Science Cen-

Acknowledgments

ters and at NMFS Headquarters during the review period.Lastly, we thank Allen Shimada (NMFS Office of Scienceand Technology), Betty Holmes (NEFSC), and ShelleyArenas and Dave Stanton (NMFS Scientific PublicationsOffice) for their help in producing the final report.

69

Acronyms

ABC � Allowable Biological CatchAFSC � Alaska Fisheries Science CenterAI � Aleutian IslandsASMFC � Atlantic States Marine Fisheries Com-missionBSAI � Bering Sea and Aleutian IslandsCIE � Center for Independent ExpertsCPUE � Catch Per Unit EffortCSIRO � Commonwealth Scientific and IndustrialResearch Organization (Australia)DAS � Days At SeaEBM � Ecosystem-Based ManagementEBS � Eastern Bering SeaEEZ � Exclusive Economic ZoneESA � Endangered Species ActESU � Evolutionarily Significant UnitsFFA � South Pacific Forum Fisheries AgencyFIS � Fisheries Information SystemFMC � Fishery Management CouncilFMP � Fishery Management PlanFRV � Fishery Research VesselFSP � Fisheries Strategic PlanFTE � Full-Time Equivalent (applied to numbers ofin-house staff or contractors)FY � Fiscal YearGLOBEC � GLOBal ocean ECosystem dynamicsGOA � Gulf of AlaskaGOM � Gulf of MexicoHMS � Highly Migratory SpeciesIATTC � Inter-American Tropical Tuna CommissionICCAT � International Commission for the Conser-vation of Atlantic TunasICES � International Commission for the Explora-tion of the SeaISC � Interim Scientific Committee for Tunas andTuna-like Species in the North Pacific OceanLOA � Length OverAllMFCMA � Magnuson Fishery Conservation andManagement ActMHLC � Multi-lateral High Level Conference

MPA � Marine Protected AreaMRFSS � Marine Recreational Fisheries StatisticsSurveyMSFCMA � Magnuson-Stevens Fishery Conserva-tion and Management ActMSY � Maximum Sustainable YieldNASCO � North Atlantic Salmon Conservation Or-ganizationNEFSC � Northeast Fisheries Science CenterNEPA � National Environmental Protection ActNMFS � National Marine Fisheries ServiceNOAA � National Oceanic and Atmospheric Admin-istrationNRC � National Research CouncilNRIFSF � National Research Institute for Far SeasFisheriesNWFSC � Northwest Fisheries Science CenterOFL � Overfishing LevelPDT � Plan Development TeamPSMFC � Pacific States Marine Fisheries Commis-sionROV � Remotely Operated VehicleSAFE � Stock Assessment and Fishery EvaluationSAM � Statistical Assessment ModelSARC � Stock Assessment Review Committee(NEFSC)SAW � Stock Assessment Workshop (NEFSC)SCTB � Standing Committee on Tuna and BillfishSEFSC � Southeast Fisheries Science CenterSFA � Sustainable Fisheries Act (1996)SPC � Secretariat of the Pacific CommunitySQAP � Science Quality Assurance ProgramSSC � Scientific and Statistical CommitteeSTAR � Stock Assessment Review PanelSWFSC � Southwest Fisheries Science CenterTRAC � Transboundary Resource Assessment Com-mitteeVTR � Vessel Trip ReportWG � Working Group

marine fisheries stock assessment improvement plan

Documents