Link to NMFS Homepage Link to NOAA Homepage Keyboard Navigation Alaska Fisheries Science Center Home banner


Status of Stocks & Multispecies Assessment Program

Management Strategy Evaluation Working Group

Evaluation of fishery management strategies has been an ongoing research activity of the AFSC for years. Most recently, the Programmatic Supplemental Environmental Impact Statement (PSEIS) for the Bering Sea – Aleutian Islands (BSAI)and Gulf of Alaska (GOA) Groundfish Fishery Management Plans (FMPs) devoted thousands of pages to evaluation of both current and alternative fishery management strategies. Nevertheless, further work remains to be done. Therefore, a working group (WG) has been established and tasked with continuing and expanding the AFSC’s research in the area of management strategy evaluation (MSE). While it is understood that MSE research will never truly be finished, it is expected that the WG will be able to make significant advancements in this area over the next few years. The WG met for the first time on 17 August 2004 at the AFSC. Present from the AFSC were Kerim Aydin, Jeff Breiwick, Martin Dorn, Sarah Gaichas, Jim Ianelli, Pat Livingston, Paul Spencer, and Grant Thompson. Present from the University of Washington were Doug Kinzey, Arni Magnuson, and James Murphy.

Harvest Strategy Review

In October 2001, the North Pacific Fishery Management Council (NPFMC) commissioned an independent scientific review of the council’s current groundfish harvest strategy. The review panel, chaired by Daniel Goodman, presented its report (the “Goodman report”) to the Council in November 2002. The AFSC presented a written response to the Goodman report in October 2003. Briefly, that response noted that many of the MSE-related suggestions made in the Goodman report had already been addressed in the PSEIS or other documents, including use of a wide set of performance measures involving both utilization and conservation objectives, use of a public process to develop alternatives, use of species-specific harvest strategies for certain species, use of group-specific harvest strategies for certain groups of species, use of alternatives to the F40% reference point, use of a utility function approach to choose reference points, and examination of multi-annual catch limits. Furthermore, in some instances the MSE contained in the PSEIS went beyond the suggestions made in the Goodman report by using, for example, a state-of-the-art technical interactions model that facilitates simulation of the effects of the Optimum Yield (OY) caps in the BSAI and GOA groundfish fisheries.

However, the Goodman report also contained several suggestions pertaining to MSE that have yet to be implemented on a major scale, including 1) use of parallel “operating” and “assessment” models to facilitate simulation of the feedback nature of the management process, 2) use of ecosystem models, 3) use of multi-attribute control rules in the lower tiers, 4) use of constraints on the extent to which Total Allowable Catch (TAC) can change from year to year, and 5) reexamination of the extent to which the current tier system correlates information quality with management precaution. The Goodman report also suggested that alternative management strategies be tested with respect to the effects of regime shifts, spatial structure, depensation, and interspecific differences in life history and with respect to imprecision in estimates of selectivity and survey catchability. WG members suggested that the scope of MSE research should not be limited to issues raised in the Goodman report.

Marine Stewardship Council

Martin Dorn reported on the progress of the Marine Stewardship Council (MSC) in responding to an industry request for certification of the BSAI and GOA pollock fisheries. It is possible that certification will be granted conditionally, with one of the conditions being completion of a more thorough MSE according to a specified timetable.

Literature Review

A preliminary list of references pertaining to MSE was distributed to the WG and is available on request from

National Standard Guidelines

Grant Thompson reported on a draft proposed rule describing a possible revision of the guidelines for National Standard 1 of the Magnuson-Stevens Fishery Conservation and Management Act (MSFCMA). Some of the new features in the proposed revision focus on the distinction between “core stocks” and “assemblages.” For example, an OY control rule would be required for each core stock and for each assemblage or each indicator stock within an assemblage. Targets, such as OY, would have to be achieved on average. The current minimum stock size threshold (MSST) would be replaced by a “biomass limit” (Blim), which has a default value of 1/2 biomass maximum sustainable yield (BMSY). The default Blim would not have to be used for a particular stock if the default does not make sense in light of the stock’s natural variability. Also, Blim would not have to be specified if the available data are inadequate to do so or if the stock is managed under a sufficiently conservative OY control rule. The term “overfished” would be replaced by “depleted” throughout the guidelines.


The WG spent the majority of the meeting hearing presentations on recent MSE research and discussing possible approaches to future MSE research. Presentations were given on recent research pertaining to the performance of the current management strategy in the presence of regime shifts and on use of an operational management procedure in which this year’s catch limit is a linear function of last year’s catch limit and this year’s estimated biomass. Recent research was discussed pertaining to the impact of biological interactions on stocks managed using reference points derived on a single-species basis was also presented. Points raised during the discussion included the following:

1) Use of ecosystem models in defining reference points:


A) There is a difference between estimating a true ecosystem MSY and incorporating ecosystem considerations into the estimation of single-species MSYs.


B) The proceedings volume from the NMFS National Stock Assessment Workshop 6 (NSAW6)included a helpful discussion (p. 55-57) on the meaning of “ecosystem MSY” and possible multi-species management objectives.


C) It might be easier to use ecosystem models to establish a reasonable buffer between ABC and OFL than to determine a truly optimal harvest strategy.


D) Perhaps the use of ecosystem models could be confined initially to examination of aggregate OY caps rather than trying to determine how ecosystem considerations ought to influence single-species ABCs.

2) Other uses of ecosystem models in MSE:


A) At least initially, parallel MSEs could be conducted, one track using ecosystem models but ignoring some of the more complicated single-species considerations and the other track using fairly sophisticated single-species models but ignoring ecosystem considerations.


B) Perhaps the operating model could include biological interactions but the assessment model would still be based on single species. (This idea seemed to have widespread support.)


C) Some concern was expressed over the extent to which future MSEs should depart from status quo tools and management (e.g., should we base our planned research on use of highly parameterized ecosystem models before we even know whether they can be developed?).


D) Ecosystem models might be useful in estimating the natural ranges of fluctuations referenced in the proposed revision of the guidelines for National Standard 1.


E) Perhaps an ecosystem model could be used as a sort of “sidecar” which, for each year in a simulation, would return an adjustment to each projected single-species ABC.

3) Technical considerations in model development:


A) If we use an ecosystem model, will the parameters be estimated statistically or by some other method?


B) Given that it will be impossible to simulate the entire stock assessment process (e.g., the evolution of assessment methodology over time), it may be useful to test how well alternative simplifications of the stock assessment perform (e.g., simply drawing an “estimated” stock size at random from a distribution may prove to be an adequate simulation of the assessment process).

4) Risk analysis:


A) One method of risk analysis consists of producing tables showing the probabilities associated with the various possible outcomes (e.g., biomasses and catches of target and nontarget stocks). Pro: lots of fishery risk analyses are done this way, provides lots of information for decision-makers to study, does not require a priori specification of objectives. Con: creates information overload, invites decision-makers to “reverse engineer” objectives, unlikely to produce optimal decisions.


B) Another method of risk analysis measures the costs and benefits of the various possible outcomes, weights them by their respective probabilities, then determines the harvest strategy that maximizes the expected value of an objective function that has been specified in advance. Pro: minimizes the amount of information decision-makers need to synthesize, produces optimal decisions given the objective function. Con: few fishery risk analyses are done this way, requires advance specification of objective function.


C) Of course, it is possible to present both methods in the same document.

Next Steps

Major advances in MSE research will take a number of years to complete. The best course of action will probably involve incremental advances, starting from the simplest problems and working toward the most complex. The WG plans to meet again in the coming year to provide its members an opportunity to report on progress towards evaluating Alaska groundfish management strategies.

By Grant Thompson and Jim Ianelli



Quarterly sidebar Contents


ABL Reports

NMML Reports

RACE Reports

REFM Reports

Quarterly Index

Quarterly Home