Status of Stocks & Multispecies Assessment Program

SSMA Workshop on Tagging and Spatially Explicit Models

Figure 1.  Prototype net for sampling walleye pollock for future tagging experiments.  Photo by David King.

The Status of Stocks and Multispecies Assessment (SSMA) Program held an inter-divisional and inter-agency workshop on tagging studies and the status of spatially explicit stock assessment models. Participants included fisheries scientists and economists from REFM, RACE, ABL, industry, and the University of Alaska Fairbanks (UAF). Participants developed ideas to address the over arching question “How does climate influence spawning location, larval dispersal, and selection of feeding locations of a core group of managed species?” Target species selected for study included pollock, Pacific cod, rock sole, and crab. This report focuses on reports from presentations on research on tagging studies and spatial modeling.

Dr. Robert Foy (UAF) reported on North Pacific Research Board-funded pollock tagging experiments. Dr. Foy deployed an experimental net equipped with a “swimming pool” cod end in spring 2006. The net was fragile and proved inadequate for sampling pollock spawning concentrations. He is working with David King (RACE Division) to develop a sturdier version of the net that will be deployed this fall on pollock schools off the coast of Kodiak Island (Fig. 1). Dr. Foy also reported on survival rates of pollock sampled in the spring. Results indicated mortality rates were high during the first 72 hours after capture, after that time pollock survival was high.

Dan Nichol (RACE) and Peter Munro (REFM) described results of recent cod tagging experiments. Dan also described a new technology (F-RAFOS) that emits low frequency sound from moorings. Continuous records of fish location can be estimated by tagging fish with receivers that record sound at different frequencies. Locations are determined by triangulation if the fish swims within listening distance from the transmitters.

The group agreed that the combination of F-RAFOS technology and traditional mark recapture methods were desirable. F-RAFOS tags provide detailed information on the location of a small number of individuals, traditional tags provide information on the potential variability in fish movement.

Analytical methods are already being developed for combining data on individual fish tracks (as would be provided by F-RAFOS data) with traditional mark-recapture data. The two independent data sources are brought together in likelihood models for estimating movement rates. F-RAFOS and traditional mark-recapture data are expected to work synergistically to provide much better estimates of movement rates than either method would allow independently.

Sara Miller (UAF), Grant Thompson (REFM), and James Ianelli (REFM) reported on sampling designs and analytical techniques that could be used to assess fish movement from tagging data. Sara reported that movement probabilities of walleye pollock could be resolved with reasonable power using mark recovery methods. Grant Thompson discussed a new analytical technique he developed that provided estimates of the distance of fish above the bottom when records of fish depth were collected from archival tags. Jim Ianelli and Vidar Wespestad (Resource Analysts International) reported on cooperative efforts to collect acoustic and oceanographic data from fishing vessels.

Michael Canino (RACE) presented results from a recent analysis of the stock structure of Pacific cod. Preliminary results of a microsatellite study show a significant genetic difference between Asian and North American cod stocks. Examination of North American cod revealed structure at local scales.

The meeting concluded with a general discussion of analytical techniques for modeling fish movements. In addition to the modeling approaches discussed earlier in the day, participants noted that the Atlantis modeling approach may hold promise. Economists also recommended examining recent modeling efforts led by Dan Holland in the New England region. Participants agreed to work together to form proposals for the development of spatially explicit stock assessments and ecosystem models.

By Anne Hollowed and Peter Munro


Pollock and Flatfish Recruitment Workshop

The SSMA program organized an interdivisional meeting in August to discuss the status of knowledge of processes underlying recruitment of walleye pollock and flatfish. Scientists from the Center’s Resource Ecology and Ecosystem Modeling program, the Recruitment Processes (EcoFOCI) program, the Ocean Carrying Capacity program, Auke Bay Laboratories’ Marine Ecology and Stock Assessment program, and the Pacific Marine Environmental Laboratory attended. The purpose of this meeting was to discuss: recent results in recruitment research, new and existing techniques for forecasting recruitment of managed species, and techniques for incorporating ecosystem indicators into stock assessments.

A major portion of the meeting was devoted to review of the EcoFOCI recruitment prediction for walleye pollock. The history of its development, its current state, and future improvements were discussed. Participants recognized that accurate recruitment prediction is the best measure of our understanding of mechanisms underlying recruitment success for pollock. In addition, it was noted that predicting future fish production contributes to NOAA’s mission to understand ecosystem processes and forecast future states of nature. The current performance of the EcoFOCI prediction is moderate, indicating that our understanding is incomplete. EcoFOCI will continue development and testing of its pollock Individual-Based Model (IBM) model incorporating juvenile life history (and predation mortality) into the model and will begin development of a new prediction scheme that follows a year class through time, much like hurricane prediction follows the storm along its path and predicts its intended target. EcoFOCI investigators presented information prepared for a white paper summarizing the history, status, and future steps for the walleye pollock recruitment prediction. SSMA and EcoFOCI staff will continue to work collaboratively to couple techniques for incorporating recruitment forecasts in stock assessments.

Participants reviewed field and laboratory information supporting hypotheses that explore the use of retrospective and forecasting tools for other species compared to pollock. In the case of flatfish, several lines of evidence indicated that cross-shelf transport of eggs and larvae might be a critical factor influencing recruitment success. In the case of pollock, the processes influencing recruitment appear to be more complex. Complex behavior may influence the distance larvae move and their response to the local environment. Future research might focus on developing models of larval behaviors that reduce movement to bring simulated transport models into alignment with observed larval distributions.

It was also noted that knowledge of population structure might be needed for successful predictions. For example, individuals from different regions may have different local adaptations or social facilitation. Participants noted that improvements in pollock recruitment forecasting may be realized by incorporating processes influencing juvenile predation, including seabird predation. Participants also recommended renewed efforts to assess larval condition. It was noted that new indices of environmental forcing could be developed from new modeling tools (e.g., indices of the frequency and intensity of eddies).

Participants reviewed the performance of a variety of retrospective modeling studies including: correlative modeling, individual based models, neural networks, multivariate analyses, and modeling processes on a stage-specific basis. It was noted that each approach had different strengths and that applying a variety of techniques should be continued. No single approach emerged as the best approach. Participants recommended that scientists active in field and modeling activities in support of recruitment forecasting should continue to meet on an annual or semi-annual basis.

Participants considered how the forecasts would be used to inform managers. It was noted that there is growing interest in issues of long-term ecosystem responses to fishing or climate change. These issues require forecasts of future fish production over 10-50 years. The accuracy of these predictions will depend on correctly linking fish responses to changes in spawning biomass and climate-induced changes to ecosystems. It was also noted that there is increased interest in reviews of the suitability of harvest control rules under different states of nature.

SSMA scientists provided examples of analytical techniques used to perform Management Strategy Evaluations. Recruitment forecasts would be used to develop scenarios regarding future states of nature. At a broader scope, the Center is striving to develop whole ecosystem models to assess the impacts of fishing on the ecosystem. Forecasts from these models also depend on accurate prediction of climate impacts on key processes such as consumption rates, predation probabilities, and prey growth rates.

Contributed by Anne Hollowed, Kerim Aydin, Jeff Napp, Bern Megrey and Allen Macklin