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
jim.ianelli@noaa.gov.
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.
Discussion
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
>>>continued
|
|
Quarterly sidebar
AFSC Quarterly Research Reports July-Sept 2004
Contents
Feature
ABL Reports
NMML Reports
RACE Reports
REFM Reports
Quarterly Index
Quarterly Home
|
