Resource Ecology & Fisheries Management (REFM) Division

Resource Ecology & Ecosystem Modeling Program

Fish Stomach Collection and Lab Analysis

Laboratory analysis was performed on 4,905 groundfish stomachs from the eastern Bering Sea, 0 from the Gulf of Alaska, and 1,066 from the Aleutian Islands. During this quarter, 1,979 stomachs were returned by observers. No stomachs were collected aboard research vessels in the Bering Sea or the Gulf of Alaska, and 19,928 records were added to the groundfish food habits database.

By Troy Buckley, Geoff Lang, and Mei-Sun Yang

Seabird-Fishery Interactions Research

Shannon Fitzgerald continues his work with the Pollock Conservation Cooperative and Washington Sea Grant Program to develop seabird mitigation measures for large catcher/processor vessels in Alaska. The first part of the project is currently taking place on board the fishing vessel Northern Jaeger, where a streamer line apparatus and third wire crane are being tested as seabird bycatch reduction devices. The next phase will involve a vessel with no meal plant.

By Shannon Fitzgerald

Multispecies Modeling

The project "Modeling of multispecies groundfish interactions in the eastern Bering Sea" has been awarded 2 years of funding from the North Pacific Research Board. The main goals for this project are to update Bering Sea multispecies models (multispecies virtual population analysis (MSVPA), multispecies forecasting (MSFOR), and multispecies statistical catch-at-age (MSM)) and to include recent food habits and fisheries (catch-at-age) data. A new module that includes technological interactions will be developed and tested for the MSFOR and MSM. The new version of MSM (with the technological interactions module) will make available the current tools used in single-species stock assessment in a complete multispecies context and will provide probabilistic statements on the future state of commercially important components due to alternative management scenarios.

By Jesus Jurado-Molina

Multispecies and Ecosystem Modeling

NOAA’s 5-year plan, "Towards Understanding and Predicting Earth’s Environment," lists as a 3- to 5-year goal "developing the next generation of multispecies fisheries and food web production models." To aid in this development for the Alaska region, Kerim Aydin, Jesus Jurado-Molina, and Ivonne Ortiz helped organize a workshop on multispecies and ecosystem modeling, hosted by the North Pacific Fishery Management Council’s Scientific and Statistical Committee in February 2005. At the workshop, 12 scientists from the Alaska Fisheries Science Center (including those mentioned above), University of Washington, and University of Alaska presented recent and upcoming work in developing assessment-quality models of species (predator/prey), technical (fishing gear) and management (management strategy) interactions for input into stock assessment, and other decision-making processes within a multispecies, multisector fishery.

Of specific interest to the workshop attendees was the investigation of simulations in which single-species models and multispecies models gave contrasting results. For example, Figure 1 (below) shows the results of a simulated scenario consisting of removing all fishing pressure within the eastern Bering Sea ecosystem, then predicting the biomass of target species 20 years later under conditions of average recruitment.

Figure 1.  Simulated percent change of adult biomass for several target species in the eastern Bering Sea, after removing all fishing pressure for 20 years and assuming average recruitment over that time period, as reported by single species, multispecies (MSVPA), and ecosystem (Ecosim) models.

This same scenario was run with three models: a set of single species models similar to the current stock assessment models, a MSVPA model, and an Ecosim model. Results between models were similar for top predators such as Pacific cod and Greenland turbot. However, results for walleye pollock, a key forage species, showed different results when predator/prey interactions were included. Both the multispecies and ecosystem models predicted much more modest increases in pollock biomass than did the single-species model, as predation increased to compensate for the increase in food supply. It is hoped that the continued presentation of results such as these will become a useful addition to the management process.

By Kerim Aydin

>>>continued

AFSC Quarterly Research Reports Jan-Mar 2005