A Review of the Accomplishments and Plans of the NOAA Coastal Ocean Program (1994)

The Bering Sea Fisheries Oceanography Coordinated Investigations (FOCI) received its first funding from the Coastal Ocean Program (COP) in FY 1991. This report fulfills a requirement of the Coastal Fisheries Ecosystems (CFE) theme, that a thorough review of a program be conducted after three years of support. The review was conducted by the Coastal Fisheries Ecosystems subgroup of the panel on 13-14 January 1994 at NOAA's Pacific Marine Environmental Laboratory (PMEL) in Seattle, Washington.

The concept and scientific justification for Bering Sea FOCI had their origins in the International Scientific Symposium on Bering Sea Fisheries (the Sitka Conference, 1988 (COP, 1991), which identified the major research needs for management of walleye pollock in the Bering Sea. Bering Sea FOCI has two major research objectives, both intended to provide information useful for management of Bering Sea pollock stocks: 1) determine stock structure and 2) gain an understanding of mechanisms and processes leading to recruitment. These objectives have been approached through a coordinated program of physics, biology, and modeling. Significant progress has been achieved (COP, 1993a).

Substantial efforts to assess pollock stocks and to provide logistical support by the NMFS and PMEL laboratories have significantly supplemented Bering Sea FOCI studies. Because the pollock resource is important both ecologically and economically, the NMFS Alaska Fisheries Science Center has invested substantial base

resources to assess the stocks of pollock for management purposes. The Bering Sea FOCI program interacts with the Pollock Research Program of the NMFS Alaska Fisheries Science Center, particularly on stock structure research.

Bering Sea FOCI has forged partnerships between NOAA and academic scientists to complement NOAA expertise in areas of physics, biology, and modeling. The research being carried out in the Bering Sea FOCI program is consistent with the NOAA Strategic Plan (NOAA, 1993) element “Build Sustainable U.S. Fisheries” because results could lead toward predictive capabilities for managing marine fishery resources.

Bering Sea FOCI draws heavily upon talent and experience from the Shelikof FOCI project. The emphasis on patches of pollock eggs and larvae, the role of eddies, and the directed transport of patches that have been the major research efforts in Shelikof FOCI have been carried over, in part, to Bering Sea FOCI. Initial observations with respect to both physics and biology have indicated that the environment and processes in the Bering Sea are more complex than in the Shelikof Strait, however, and that results of Shelikof FOCI may not be directly applicable to the ecology and behavior of pollock in the Bering Sea.

Results to date from Bering Sea FOCI indicate that a major spawning aggregation of pollock utilizes the southeastern Bering Sea, and investigations by Bering Sea FOCI are now focused on that area. There is evidence of a cross-slope flow toward the shelf in the southeastern Bering Sea that could advect larvae spawned offshore to this area. There also is evidence that eddies may form in the Aleutian Basin and slope waters, which potentially could entrain pollock larvae. Preliminary studies of zooplankton and larval nutritional condition suggest that prey are less abundant in the Bering Sea than in the Shelikof Strait and that pollock larvae may be more likely to starve in the Bering Sea. A food-chain model supports the observations that food may be limiting to pollock larvae in the Bering Sea.

Initial observations, analyses, and syntheses have demonstrated significant progress in understanding basinwide circulation processes and the probable importance of a link between physical oceanography and larval pollock biology. Studies of the dynamics of larval pollock populations and their food resources, which are crucial to Bering Sea FOCI, will be linked closely to the studies of Bering Sea physical oceanography proposed for 1994. Sophisticated molecular biological approaches suggest that several different genetic stocks of pollock may exist in the Bering Sea.

The program is making significant progress in four areas: (1) stock structure determination, (2) circulation and physics, (3) recruitment mechanisms, and (4) food chain relationships. For the site review, the Bering Sea FOCI program provided the

subgroup with presentations, documents, and notes regarding each project, in addition to an overall summary document (COP, 1993b). The site review report comments on each research area, in addition to making specific recommendations and observations about present and future Bering Sea FOCI activities.

The subgroup was provided with information on stock structure, including knowledge derived from stock assessment research by NMFS scientists on Bering Sea walleye pollock and recent results of Bering Sea FOCI-funded research on (1) genetic discrimination of stocks and (2) circulation patterns in the Bering Sea. Significant emphasis is being focused on the pollock resource apart from research in the Bering Sea FOCI program, through NMFS annual and triennial surveys. Within the framework of traditional fisheries assessment (based upon the assumption of a unit stock in the eastern Bering Sea shelf), stock structure is acknowledged to be an important source of uncertainty in assessments. Estimates of pollock recruitment show significant variability that is not easily explained by stock size and environmental variability. In species with multiple stocks, recruitment may vary among stocks and may respond differently to environmental variability. Consequently, if recruitment variability in Bering Sea pollock is to be understood, stock structure must be defined.

The site review presentations gave evidence of distinct spawning aggregations of pollock in the Bering Sea and also variability in the biological characteristics of pollock in different areas. Successful identification of stocks often requires diverse approaches to the problem. In the Bering Sea, suggested research on stock structure includes biological studies of stock separation, circulation and larval drift research, a large-scale tagging study, and assessment models that include between-area migration. Past research on stock structure of Bering Sea pollock included measurements of chemical and isotopic composition of otoliths and limited electrophoretic studies; neither conclusively demonstrated separate stocks. A stock assessment of eastern Bering Sea pollock, now underway at the University of Alaska (funded by NMFS), will incorporate information presently known about stock structure and inferred movement patterns.

Genetics studies by Bering Sea FOCI investigators have identified differences between western and eastern Bering Sea populations of pollock, but the relationship of genetic differences to spatial and temporal distribution patterns remains unclear. Studies funded by Bering Sea FOCI to develop genetic tools to assess stock structure apparently are of high quality. The program is to be commended for attracting a talented researcher to address the stock structure questions. Although the preliminary results are encouraging, continued sampling should address both spatial and temporal

variability; i.e., migratory patterns of pollock dictate that time (season) as well as geographic location need to be considered in designing a sampling strategy. Information from the assessment model being developed at the University of Alaska would be useful in this effort. The genetic studies of Bering Sea FOCI seem to be carried out in isolation from other program elements. To remedy this problem, the program's molecular biologist should increase interactions with the NMFS stock assessment scientists as well as with other Bering Sea FOCI researchers to ensure that genetic stock structure research is integrated with more traditional physical and fisheries oceanography research to benefit the program fully.

The immediate need from genetic studies is to identify separate spawning stocks, should they exist, to facilitate planning for long term recruitment and stock assessment research. This is particularly important now that the scope of the research has narrowed to concentrate on the eastern Bering Sea. While it remains a worthy goal, efforts to develop a “magic bullet” to identify the source stock of an individual larva or juvenile may be premature until the source stocks or populations are defined genetically. The apparent lack of integration of genetics with other techniques may stem from the geographic separation of the investigators involved. Special effort should be made by the Bering Sea FOCI Executive Council to ensure inclusion of all program scientists in planning efforts so that research is coordinated to promote efficiency and to increase the relevance of results to overall program goals.

The subgroup was impressed with results of the initial research and syntheses of existing data on the physical environment of the Bering Sea. The decision to shift the investigation from the larger-scale environment of the Aleutian Basin to more process-oriented studies in support of biology in the southeastern Bering Sea is timely. As Bering Sea FOCI progresses, field programs must include integrated physical oceanography and biology observations. When integrated data sets are obtained, they should be made available to all participants as quickly as possible. The initial efforts suggest that timely distribution of data will be accomplished and that truly biophysical descriptions of the Bering Sea ecosystem, and the role of pollock in it, will result.

The subgroup noted one important facet of Bering Sea physical oceanography and associated biology that is neglected in Bering Sea FOCI. Sea ice has not been considered, but it probably plays a major role in controlling the spring bloom and secondary production. The impact of sea ice on the Bering Sea shelf ecosystem and its interannual variability are potentially important factors affecting pollock recruitment, as well as broader aspects of shelf ecology.

Much of Bering Sea FOCI research revolves around the question of exchange between the deep basin and the shelf. Quantification of this flux is described as a program goal. With the available funding and vessel time, it is uncertain how well this goal can be met. Initial computations of cross-shelf exchange using drifters seem promising, but a complete characterization is still lacking. Bering Sea FOCI has developed and deployed the “Peggy” mooring that includes a sophisticated instrument array, but it is not placed in a location where it can address the issue of cross-shelf exchange adequately. There is a need to link Lagrangian determinations of circulation with the population biology of pollock larvae, which presumably is to be carried out in planned 1994 cruises. If the physical oceanographers in Bering Sea FOCI do not believe that they can obtain good cross-shelf flux estimates, their efforts should be redirected toward more achievable goals in support of the recruitment variability objective.

The dynamics of eddies (which proved to be of critical importance in Shelikof FOCI) also are addressed in various contexts in Bering Sea FOCI. The role of eddies in the transport of larvae across the shelf break and in biological enhancement will require carefully planned mesoscale measurements. A plan and sampling design are needed that will allow food chain processes to be studied in relation to the eddy field and which will allow the transport processes for key organisms to be elucidated.

The subgroup views larval pollock sampling as the cornerstone of future Bering Sea FOCI efforts. Previous and preliminary surveys have set the stage for intensive, synoptic sampling in the 1994 field season. We encourage the investigators to utilize early results of the larval sampling and models of larval dynamics to guide them in developing a rigorous and extensive sampling program. In addition, simultaneous measures of zooplankton abundance and conductivity-temperature-density (CTD) data must be obtained during the larval surveys to help focus and calibrate the food-chain modeling effort.

Given the complexity of the recruitment problem and the risks associated with choosing a single, synthetic approach, we encourage the investigators to focus their effort by recognizing three distinct sets of questions and scales, which are nested in a hierarchy of factors that regulate recruitment success. Bering Sea FOCI investigators should recognize that recruitment may be defined at each of three levels: the cohort, the year class, and the population. Those distinctions help define specific goals and can provide the basis for formation of small working groups with a common interest. Three general goals should be:

1. Couple physical dynamics to larval survival.This issue focuses on larval mortality by considering small-scale patch effects and the short time scales of early larval life history. Components would include research efforts on nutritional status, egg and larval transport, larval survival, and larval feeding ecology. A component that considers predation on eggs or larvae is also desirable. Results will yield information about the factors that govern the success of larval cohorts. Short-term, food-chain dynamics modeling is an appropriate general tool.

2. Evaluate the success of juvenile cohorts. This effort should focus on the issue of mortality on intermediate time and space scales and incorporate research on cannibalism and the effect of non-pollock predators. Its goal should be to provide a basis to estimate survival of pollock to one year of age. The results can be linked to population models based on results of assessment surveys and could provide the basis for forecasting strengths of year classes. Age-structured pollock population models would serve as the analytical tool. Those models might be modified to allow monthly or shorter time steps important to growth-rate dynamics and changing rates of predation. A working group session might quickly define what is needed to proceed.

   The proposed field studies near the Pribilof Islands in 1994 will provide an excellent opportunity to develop and test mechanistic hypotheses. The design of this research should utilize results of laboratory studies of juvenile pollock behavior. Structuring field studies around the questions of interactive effects of food, predators, and temperature will help maximize knowledge gained from the limited ship time available in 1994. Results also may prove to be important for marine mammal conservation because of the role pollock plays in the ecology of some marine mammals in the Bering Sea. This element of research also might become the cornerstone of future and more extensive work on the juvenile life stage.

3. Develop understanding of pollock interactions in the context of community and ecosystem dynamics. This effort would emphasize the role of large-scale, long-term effects such as those resulting from interannual variation in climate, fishery exploitation rates, and changes in the intensity of species interactions involving pollock as prey, competitors, and/or predators. Research would emphasize feedbacks that result from variable recruitment. It would develop hypotheses concerning the nature of compensatory changes in the structures of pollock populations and food webs. Models of food web dynamics will need to be developed.

Small working groups could develop the most appropriate (i.e., parsimonious) models in the coming months. Large and complex ecosystem models seem unwarranted at this stage of planning and research.

For each of these goals, the formation of small working groups is desirable. They should meet or communicate frequently as they develop the conceptual framework, information sources, sampling protocols, and modeling tools required to evaluate key questions. By focusing on the scales most pertinent to those goals, their results will promote more rapid and complete understanding. Syntheses of those results could help to prepare Bering Sea FOCI for the next larger level of effort.

Knowledge of food chain dynamics of larval fish plays an important role in developing a working hypothesis on recruitment variability, and in designing field and laboratory research. The subgroup saw little evidence of a collective working hypothesis on this topic, and is concerned about a lack of coordination among research projects in this area. To date, the zooplankton and larval feeding studies are based on only a few samples at a few stations, and the larval fish surveys were not coupled with the basic studies on physical circulation and hydrographic studies. Little attention has been devoted to phytoplankton production and its relationship to larval dynamics.

Bering Sea FOCI investigators believe that larval patches may be associated with mesoscale eddies. They suggested that larvae are in better environments when entrained in eddies, or when transported onto the southeast Bering Sea shelf. However, the role of eddies and/or transport onto the shelf is not well defined in the context of what nutritional benefits may be derived, including possible ice-edge effects. Bering Sea FOCI investigators are encouraged to develop a working hypothesis that is mechanistic, not merely correlative. It is not sufficient to suggest that larvae survive and grow better inside eddies or on the shelf. It is important to understand why this might be the case. Also, it is important to consider how high-density patches of larvae of small spatial extent contribute to recruitment compared with potential contributions from lower densities of larvae dispersed over the broad extent of the Bering Sea. The results of such calculations might guide development of sampling plans and allocation of sampling effort.

A new project will soon be initiated in Bering Sea FOCI to gain a greater understanding of the functional aspects of planktonic food production and its relationship to larval survival. A major component of this project will be a new mooring. But, a

single chlorophyll-measuring mooring may not be sufficient to test ideas about food chain dynamics in eddies, slope waters, or shelf waters. If logistically possible, field studies should be expanded to include measures of nutrients and phytoplankton (cell counts and at a bare minimum, chlorophyll concentrations) in conjunction with the CTD grids and sample transects conducted on larval survey cruises.

Survey grids should be designed to test ideas about larval survival and larval food production in eddies and on the shelf. Because survey cruises cannot always set out to sample an eddy (eddies are discovered only after the fact), investigators need to collect comprehensive information at each station during surveys. The surveys conducted to date have lacked coordination in obtaining data on variables of interest, but improvements were evident in planning for the 1994 season. It is important that samples be collected to test hypotheses about probable benefits to larvae in eddies or on the shelf versus larvae swept off the shelf into the slope and basin areas. The possibility of detecting and tracking eddies in real time from sea surface temperature and/or ocean color measurements from satellites should be explored.

Apparently, after survey samples are analyzed and interpreted, the food-chain model will be employed to draw conclusions. This leads to a number of issues that must be addressed and which must be considered by Bering Sea FOCI investigators. (1) What questions do the program scientists want the model to answer? (2) Will the model need to be reparameterized and refocused to answer those questions? (3) Will there be a need for additional measurements, for example profiles of photosynthetically active radiation and diffuse extinction coefficients? And, (4) are the investigators who are providing input data for the model expecting too much from it? A working group should address these issues.

The subgroup developed a number of specific recommendations:

1. Lagrangian studies on cross-slope and cross-shelf transport must be coordinated with coincident biological studies on pollock larvae and plankton. It appears that the developing 1994 sampling plan will incorporate this approach.

2. Small working groups of Bering Sea FOCI investigators should be organized, and should meet frequently to develop hypotheses, plan research, evaluate progress, and assess the need to change directions in each of the research areas. Working in groups will have a focusing effect and will improve collaborations, and leadership will emerge.

3. Significant progress is being made in the food-chain modeling of pelagic ecosystem processes but care should be taken to ensure that data collection does not become an effort solely to satisfy needs of the model. The model itself should be expanded to estimate more than growth potential of pollock larvae. Survival rates of larvae are one obvious extension that should be pursued.

4. The molecular biological approaches to determine stock structure show promise. However, the research (to an extent) is being carried out in isolation. The investigators conducting genetics studies should interact with NMFS stock assessment scientists and Bering Sea FOCI investigators. All would benefit from increased interactions and faster progress will be made toward answering stock-related questions of recruitment variability in pollock.

5. It is critical to understand the flux of water from the deep basin to the shelf in the southeastern Bering Sea to determine how larvae might be transported from offshore to inshore areas where feeding conditions may be better. If the planned efforts by program scientists are not sufficient to address this issue, new plans or redirected efforts may be required.

6. There is a need for mechanistic studies of eddies and their dynamics, particularly of the mechanisms that can advect eddies in slope waters onto the shelf. These studies must be linked closely to biological studies of pollock larvae distributions.

7. Sea ice, polar water, spring blooms, and possible relationships to pollock recruitment should be considered in Bering Sea FOCI, because of their dominant effects on the Bering Sea shelf system.

8. Research on pollock larvae abundances and distribution, their relationship to zooplankton prey resources, and to larval condition, survival rate, and dispersal are central issues with respect to recruitment mechanisms. A major effort should be made in 1994 to address these issues.

9. Juvenile pollock research is encouraged, but it should be focused on the most important issues, which may be primarily effects of predation and cannibalism on dynamics in this stage. The proposed research in 1994 is a preliminary effort to develop this element of Bering Sea FOCI.

10. The larval and juvenile pollock behavior research (which is not funded by Bering Sea FOCI) has produced results that may be highly relevant to on-

going studies of the vertical distribution of pollock in the Bering Sea. It is important to consider results of the behavior research in designing Bering Sea FOCI field studies. Collaboration between the behavioral scientist and Bering Sea FOCI investigators should be encouraged.

11. The subgroup does not believe that a goal (as presently stated) of Bering Sea FOCI should be to increase the yield of Bering Sea pollock. We think that a more appropriate goal is to increase understanding of production and recruitment processes, so as to insure continuing high yields. This “risk aversive” goal will mesh nicely with goals of the NOAA strategic plan, which emphasize “sustainability.”

12. In the longer term, Bering Sea FOCI may wish to shift its emphasis toward ecosystem-level processes on the Bering Sea shelf. Pollock should remain a key component of Bering Sea FOCI, but the goals of the research could expand to address the problems of species interactions and their consequences in this coastal fishery ecosystem.

13. Successful leadership in a CFE program requires equal and communicative participation by the three partners, i.e., scientists from the NOAA Office of Oceanic and Atmospheric Research, the NOAA National Marine Fisheries Service, and academia. Without equal participation, maximum contributions and coordination may not be achieved. At present, the Bering Sea FOCI Executive Council does not represent the three partners equally. Its composition could be improved by adding additional academic representation.

14. The process of project solicitation, selection, and review could be improved in Bering Sea FOCI by making the solicitation more open and treating academic and NOAA partners equally with respect to proposal review. The Technical Advisory Group of Bering Sea FOCI could be utilized to help develop a standardized and acceptable procedure.

The subgroup recommends that COP continue to support Bering Sea FOCI for the remainder of its planned 5-year term. In addition, the subgroup summarizes below its findings regarding Bering Sea FOCI's goals and objectives, accomplishments, utility, responsiveness to past review, and future plans.

Goals and Objectives — The Bering Sea FOCI program has made significant progress toward its goals of understanding recruitment processes (oceanographic processes and stock contributions) of walleye pollock in the Bering Sea, now specifically in the southeast Bering Sea. The subgroup endorses the overall approach but recognized a need for greater coordination in planning and observations among biologists, physical oceanographers, and modelers. In the long term, as Bering Sea FOCI progresses to the end of its first 5 years and begins to plan subsequent studies, a more ecosystem-oriented approach, rather than a pollock-specific approach, may be warranted to investigate recruitment processes of Bering Sea fishes.

Accomplishments — NOAA-academic partnerships have been developed. Bering Sea FOCI has made progress in defining circulation, stock structure, and the probable importance of the southeast shelf as a spawning area and probable larval nursery. The program has gathered preliminary evidence that eddies may be important for pollock recruitment in the Bering Sea and that larval food may be limiting in this system. A food-chain model has been developed which, while helpful, needs to be expanded to account for larval pollock mortality in addition to growth. The “Peggy” mooring is useful to obtain physical data in the eastern Bering Sea, although limited because of its location in the deep off-slope water. The program is becoming more coordinated and in the 1994 field season will conduct its first integrated biological and physical measurements.

Utility — Bering Sea FOCI clearly has potential to contribute information important for management of pollock in the Bering Sea. The program promises to yield knowledge about mechanisms of recruitment variability and stock structure. Its research will contribute to the “sustainability ” and “predictability” goals of the NOAA strategic plan. In addition to increased knowledge about pollock recruitment variability, better understanding of the Bering Sea ecosystem, with its valuable fish, marine mammal, and seabird resources will be an important product of the research.

Responsiveness to Past Reviews — Bering Sea FOCI has not been reviewed in the past. The program is to be commended for the recent appointment of an excellent Technical Advisory Group, who will advise Bering Sea FOCI and provide independent review on a regular basis.

Program Management — Bering Sea FOCI has formed a six-person Executive Council. To be effective, this council must represent the NOAA and academic partners and must provide strong leadership to coordinate the interdisciplinary research of Bering Sea FOCI. Procedures to solicit, select, and review proposals apparently have differed for NOAA and academic partners. These procedures must be standardized and should be equally rigorous throughout the program. Working groups should be constituted in specific research areas to promote coordination and planning.

Future — Bering Sea FOCI should move gradually toward ecosystem science. Understanding pollock ecology should be an important, but not exclusive, goal of the research. Program plans should insure that physical and biological elements of the program are fully coordinated and synoptic. It is emphasized again that working groups should be formed, not only to plan and coordinate ongoing research, but to develop visions of how elements of Bering Sea FOCI should evolve in the future.

Coastal Ocean Program. 1991. Bering Sea FOCI. Fisheries-oceanography coordinated investigations. Project study plan 1991-1996. National Oceanic and Atmospheric Administration, Department of Commerce , Washington, D.C. 28 pp.

Coastal Ocean Program. 1993a. CFE FY 1994 Implementation Plan. National Oceanic and Atmospheric Administration, Department of Commerce , Washington, D.C.

Coastal Ocean Program. 1993b. Bering Sea FOCI. Fisheries-oceanography Coordinated Investigations. Overview and Plans. November 1993. Alaska Fisheries Science Center and Pacific Marine Environmental Laboratory, Seattle, Washington, National Oceanic and Atmospheric Administration, Department of Commerce , Washington, D.C. 25 pp. (prepared by A. Macklin).

National Oceanic and Atmospheric Administration. 1993. National Oceanic and Atmospheric Administration 1995-2005 Strategic Plan. July 1993. Department of Commerce, Washington, D.C.