Fisheries Theme Team

Population Dynamics

Background

The management of both exploited and protected living aquatic resources, whether in marine or freshwater environments, is dependent on estimates of their abundance and dynamics. Such estimates are, by their very nature, imprecise owing to the inability to directly observe and enumerate populations of fish, shellfish, marine mammals, and turtles. Standardized methods for directly or indirectly sampling these populations and combining such information with harvest estimates from commercial and recreational fisheries have made it possible to estimate relevant population characteristics, although often with considerable uncertainty. The products of these methods, termed stock assessments, provide the scientific input to the management processes employed at state, regional, national, and international levels.

An assessment includes not only an estimate of a stock’s past and present abundance, size/age structure, and productivity, but can also include a quantitative forecast of future stock sizes and harvest levels associated with various management alternatives such as different exploitation rates, mesh or minimum harvest sizes, and closed areas or seasons. Accurate stock assessments and the incorporation of their results into regulatory measures are vital for avoiding potential drastic reductions in resource abundance and associated economic, social, and ecological problems. Therefore, ongoing research and development is needed to continually produce new and improved stock assessment methods incorporating state-of-the-art technological innovations and statistical theory, and to ensure that the very best methods are employed in real-world assessments.

The report from a recent study by the National Research Council (NRC) Committee on Fish Stock Assessment Methods entitled "Improving Fish Stock Assessments" (NRC 1998) reviewed the state of existing knowledge about stock assessments and presented a series of recommendations for improving the process and for evaluating existing assessments. These recommendations included: 1) adhering to a five-step process for conducting assessments, 2) ensuring the availability of at least one reliable abundance index for a stock, 3) using a variety of assessment models and independent estimates of natural mortality (M), 4) including realistic measures of uncertainty in the output variables, 5) including management tools specific to the species managed in precautionary management measures, 6) simultaneously evaluating assessment methods and harvesting strategies to determine their ability to achieve management goals, 7) developing new data and models that either can reduce uncertainty or are robust to incomplete, variable data or environmental fluctuations, 8) conducting periodic independent peer reviews of assessments, 9) establishing a national standardized and formalized data collection system, and 10) through partnerships of government laboratories, universities, and the fishing industry, ensuring the exchange of personnel and ideas and providing funding for education to keep the stock assessment process fresh and invigorating.

The National Marine Fisheries Service (NMFS) has the responsibility to perform stock assessments on all living resources residing in the U.S. Exclusive Economic Zone (EEZ) (3-200 nautical miles from shore) and to participate in international assessments on some highly migratory species (e.g., tunas, billfish, sharks). In addition, some assessments are performed by scientists working within organizations such as the International Pacific Halibut Commission (IPHC), Inter-American Tropical Tuna Commission (IATTC), International Commission for the Conservation of Atlantic Tunas (ICCAT), and the International Council for the Exploration of the Sea (ICES). The three regional interstate Marine Fisheries Commissions (i.e., ASMFC, GSMFC, and PSMFC) provide frameworks under which state and Federal scientists cooperatively conduct assessments of stocks primarily in multi-state waters; some states such as Alaska, Oregon, and Florida assume the responsibility for assessing the stocks in their own waters. There is also U.S. participation in the assessment of living resources residing outside the U.S. EEZ handled by international organizations or commissions such as the International Whaling Commission (IWC), Northwest Atlantic Fisheries Organization, ICES, and the Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR).

In the freshwater arena, individual states are responsible for assessing stocks in their respective lakes, reservoirs, rivers, and streams. In the Great Lakes, stocks assessments are performed by state, Federal, and university scientists under the auspices of the Great Lakes Fishery Commission.

Problems

A number of problems have been identified and a series of recommendations for improving stock assessments has been proposed (e.g., NRC 1998), but two specific problem areas or deficiencies were identified in an October 1998 NMFS/OAR Science Retreat relative to this topic: 1) stock assessment methodology development and 2) graduate student training.

Stock assessments constitute one of the more visible and highly scrutinized activities performed by NMFS at its five Fishery Science Centers and their supporting laboratories. These assessments provide the scientific basis for regulatory measures recommended by the regional Fishery Management Councils and enacted by NMFS. Approximately half of the NMFS budget is invested in such activities. Input data for the various quantitative methods used in NMFS stock assessments (as well as those of other agencies and organizations) are derived from both fishery-independent and fishery-dependent sources.

Fishery-independent information is obtained from research vessel surveys employing trawls, dredges, traps, and hooks, or from remote sampling methods such as hydroacoustic surveys conducted by vessels and transect sighting surveys conducted from ships and aircraft. Such surveys offer the best opportunity for controlling sampling conditions over time and are the best choice for achieving a reliable index of abundance (NRC 1998).

The lack of time series of reliable commercial and recreational catch statistics, estimates of discards and bycatch, and comprehensive biological sampling of catches (both landings and discards) nationwide has hampered the ability to derive estimates of stock size and fishing mortality rates (F) using traditional methods such as production modeling, virtual population analysis (VPA), swept area, and length cohort analysis. Such methods also underestimate the true level of uncertainty in stock status. NMFS has proposed several research priorities to improve analytical methods for stock assessment and prediction:

A national effort is now being coordinated by the NMFS Office of Science and Technology to develop an assessment "toolbox" containing software for various methods incorporated under a standard set of input and output protocols. Although academic researchers at universities in the Sea Grant network have been active in the development of stock assessment methods, university and NMFS collaboration in this effort to date has been largely informal. There is considerable potential for expansion. Successful collaboration will require the participation of researchers from both sectors in developing a research program of mutual interest and benefit, and will also require investment of supporting funds by both NMFS and Sea Grant.

Stock assessments should provide the quantitative support needed for making management decisions in the face of uncertainty. Estimates of precision are needed both for the various stock parameters calculated in the assessment (e.g., stock sizes, fishing mortality rates) and for the forecasted stock sizes and harvest levels associated with a range of possible exploitation rates or other regulatory actions. Most stock assessment methods currently in practice assume that all output error stems from variation in survey abundance indices and that the catch and age compositions are without error. Newer techniques, such as Bayesian statistical analysis which is receiving increasing use, permit probability distributions around all input data, with the assumed distributions compared to the analytical outcome to form confidence intervals around quantities of interest. Further development and implementation of such approaches, as well as the education and training of assessment scientists in their applied use, will provide opportunities for collaboration between academic researchers in the Sea Grant community and Federal and state scientists.

The goal of assessment models is to map or scale relative indices of stock abundance from fishery-independent or -dependent sources to absolute estimates of abundance. Critical to this process is the catchability coefficient (q) of the survey or fishing gear used to derive the relative index which is generally not known with any degree of certainty. Attempts have been made recently by NMFS scientists, in collaboration with university researchers and the fishing industry, to estimate survey catchabilities directly from in situ research by means of depletion experiments, direct observations, and video surveillance. Further work is needed in this area and would provide an excellent opportunity for increased collaboration between NMFS and university (both Sea Grant and NURP programs) researchers.

Graduate Student Training

NMFS, which currently employs about 100 full-time assessment staff or about two-thirds the number of such experts available in the U.S., is finding it increasingly difficult to attract and hire qualified people to fill vacant positions. The fundamental problem is that insufficient numbers of quantitative scientists, who also meet the citizenship requirement for employment by the Federal government, are being trained at the MS and PhD levels. This dilemma stems from several factors including 1) a shortage of university faculty qualified to teach graduate level population dynamics and quantitative ecology or to mentor students in research projects involving the development, application, or interpretation of stock assessment data, methods, or results; and 2) inadequate financial support at the universities to attract the best students.

Sea Grant universities, which have trained nearly all the stock assessment scientists now employed by NMFS, state agencies, and other organizations, are in a unique position to play a key role in alleviating the shortage of highly qualified people. The new Joint Graduate Fellowship Program in Population Dynamics announced in November 1999, sponsored by the National Sea Grant College Program Office (NSGO) and NMFS, will fund two highly qualified PhD-level students for up to three years (beginning in 2000) who will each be required to work summers at a NMFS facility in close collaboration with a NMFS mentor who will provide data for their thesis and serve on their committee. The goal of the program is to encourage qualified students to pursue careers in population dynamics and stock assessment methodology, increase the available expertise in this field, foster closer relationships between NMFS and Sea Grant university scientists, and provide real-world experience to graduate students and accelerate their career development. Available funds will permit two additional students to be supported in each subsequent year up to a maximum of only six students in a given year. Further financial support would allow this unique opportunity for training and research to be expanded and would encourage the recruitment of more quantitatively qualified faculty.

Literature Cited

National Research Council. 1998. Improving fish stock assessments. National Academy Press, Washington, DC, 177 pp