3

General Issues in the Collection, Management, and Use of Fisheries Data

WHAT ARE FISHERIES DATA?

The phrase “fisheries data” is a general way of referring to data that may be of use in the management of a fishery as well as for commercial, recreational, cultural, and scientific purposes. Such data usually include biological information about the exploited fish and associated species, economic information about the fishermen and the markets for the catch, and information about the environmental conditions that affect the productivity of the species. This information is collected from many sources.

A primary source of information is the commercial and recreational fishermen themselves, so-called fishery-dependent data. Logbooks (also called trip tickets) are designed to collect data on the time and place of fishing, the effort expended, catch by species, and other information. In many jurisdictions, completion of logbooks is a condition of participation in the fishery. Often, information from logbooks is the most timely information on current fishery conditions; mechanisms for self-reporting are rare in recreational fisheries.

Catch sampling programs are another important source of information. Fish can be measured and weighed either at sea (by observers) or at landing sites (by port agents). Observers are placed on commercial fishing vessels to provide information on fishing activities that are not always reported in logbooks, such as effects of fishing activities on protected species and the extent and fate of by catch and discarding. Samples can be obtained to determine the species composition, sex ratio, and age composition of the catch.

In some fisheries, scientific surveys are a vital component of the stock assessment process. Research vessels of the National Oceanic and Atmospheric Administration (NOAA) and commercial fishing vessels operating under charter agreements with NOAA are used to conduct surveys of fish abundance. These surveys are the primary source of fishery-independent data, including estimates of the age structure of fish populations and relative abundance of stocks. The National Research Council (NRC, 1998a) demonstrated the importance of accurate indices of abundance, which in many fisheries can be obtained only from fishery-independent surveys.

In fiscal year 1999, the National Marine Fisheries Service (NMFS) spent $28.8 million on ship time for surveys (not counting personnel and analyses), $3.9 million for recreational monitoring, $9.2 million on observer programs (with another $10 million provided by industry), and $2.8



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA 3 General Issues in the Collection, Management, and Use of Fisheries Data WHAT ARE FISHERIES DATA? The phrase “fisheries data” is a general way of referring to data that may be of use in the management of a fishery as well as for commercial, recreational, cultural, and scientific purposes. Such data usually include biological information about the exploited fish and associated species, economic information about the fishermen and the markets for the catch, and information about the environmental conditions that affect the productivity of the species. This information is collected from many sources. A primary source of information is the commercial and recreational fishermen themselves, so-called fishery-dependent data. Logbooks (also called trip tickets) are designed to collect data on the time and place of fishing, the effort expended, catch by species, and other information. In many jurisdictions, completion of logbooks is a condition of participation in the fishery. Often, information from logbooks is the most timely information on current fishery conditions; mechanisms for self-reporting are rare in recreational fisheries. Catch sampling programs are another important source of information. Fish can be measured and weighed either at sea (by observers) or at landing sites (by port agents). Observers are placed on commercial fishing vessels to provide information on fishing activities that are not always reported in logbooks, such as effects of fishing activities on protected species and the extent and fate of by catch and discarding. Samples can be obtained to determine the species composition, sex ratio, and age composition of the catch. In some fisheries, scientific surveys are a vital component of the stock assessment process. Research vessels of the National Oceanic and Atmospheric Administration (NOAA) and commercial fishing vessels operating under charter agreements with NOAA are used to conduct surveys of fish abundance. These surveys are the primary source of fishery-independent data, including estimates of the age structure of fish populations and relative abundance of stocks. The National Research Council (NRC, 1998a) demonstrated the importance of accurate indices of abundance, which in many fisheries can be obtained only from fishery-independent surveys. In fiscal year 1999, the National Marine Fisheries Service (NMFS) spent $28.8 million on ship time for surveys (not counting personnel and analyses), $3.9 million for recreational monitoring, $9.2 million on observer programs (with another $10 million provided by industry), and $2.8

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA million on vessel monitoring system (VMS) programs. The expenditure by NMFS for these data collection activities is thus on the order of $45 million. Additional expenditures were made by states and industry. The total fishery harvest in the United States (commercial and recreational) is valued at approximately $45.7 billion when the total economic effects are included (NMFS, 1995). WHO USES FISHERIES DATA? Fisheries data have many uses and many users—including stock assessment by scientists, strategic planning by industry, and fishery monitoring and allocation decisions by managers. Adequacy of data can be evaluated only in the context of the purposes for which they are used. Each use implies a set of users and a suite of requirements that the data must satisfy, including timeliness, level of detail, accuracy, accessibility to users, coverage or completeness, and credibility of the data collection process and the management process that uses the data. Fisheries data are vital to strategic planning activities in coastal communities that rely on fisheries. Fishery management authorities are responsible to use fisheries data for creating policies for the orderly and sustainable development and management of fisheries. Civil authorities use fisheries data to site marinas, underwater pipes and cables, and other maritime facilities, and to develop infrastructure for the fishing industry. Bankers use fisheries data to plan economic development and loan packages to fishermen, fish processors, and ship suppliers. Fishermen themselves use fisheries data to plan future fishing activities, such as shifts to new fishing grounds, changes in fishing gear, and changes in species targeted. However, fishermen often use their own data sources, including their own logbooks and observations, and what they learn from other fishermen and buyers, instead of using government data. This may occur because of some fishermen's mistrust of government data, the frequent lag time in availability of such data (often too great to use government data in business planning), and the lack of data for the geographic area and type of fishery in which a specific fisherman is engaged. Monitoring conditions in a fishery is the responsibility of regional fishery management councils and NMFS, and is the primary means of assessing compliance with and accomplishing enforcement of fishery regulations. Another major responsibility of the regional councils is allocation of harvest opportunities among different user groups. Environmental and other interest groups also have become increasingly involved in monitoring fishing activities. Monitoring often requires data with great detail in both time and space as well as frequent updates, often within a fishing season. Stock assessment is a critical use of fisheries data and is often considered its primary use. The committee devoted a significant portion of its attention to the data used in stock assessments, using the summer flounder fishery as a case study. Scientists employed by state, interstate, national, and international fishery agencies are the primary users of data relevant to stock assessments; in addition, university and private sector scientists increasingly are becoming involved in stock assessments and related research. Current stock assessment practices use data aggregated over the entire fishing ground and over a fishing season. Although assessment methods may require a greater diversity of data, the resolution in space and time is usually rather coarse and may need updating only infrequently, such as annually or semi-annually. The multiple users of fisheries data have different requirements in terms of resolution in time and space for each possible data element. Table 3-1 summarizes the requirements for data elements by various users (based on committee experience); specific details depend on the characteristics of individual fisheries. Data system designers, therefore, must consider that the demands will vary among users, and the system must be capable of accommodating users who require data at different spatial resolutions and different degrees of timeliness.

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA TABLE 3-1 Example of How Data Timeliness and Spatial Resolution Vary Among Users   USE Users   MONITORING Councils NGOs ASSESSMENT NMFS Scientists Academics Fishermen ALLOCATION Councils PLANNING Councils Local and State Governments Economists Bankers Fishermen Update Frequency (timeliness) Within season Between seasons Within and between seasons 1-5 years Spatial Resolution Detailed/mandated jurisdiction Stock-wide Stock-wide Ad hoc Required Elements         Catch by species Fishing effort Catch at length     Age composition       Sex ratio       Credibility is one of the major concerns surrounding current fisheries data collection activities. Many stakeholders believe that data collected by NMFS are neither accurate nor complete. These misgivings are exacerbated by problems of timeliness and accessibility and by perceived conflicts of interest; NMFS not only collects the data but also conducts stock assessments, makes policy recommendations to councils, enforces fishery regulations, and makes judgments about the policy recommendations and fishery management plans prepared by the regional councils. For many fishermen these multiple responsibilities of a single agency create some mistrust regarding the collection and use of fisheries data. Two recent reports stress the importance of greater collaboration among scientists and stakeholders in data collection. First, the Consortium for Oceanographic Research and Education states Finally, collaborative data collection and research efforts should be encouraged among agency scientists, independent scientists, and representatives of industry and public interest groups. Not only would this build confidence among the different groups, but it would provide access to valuable, non-traditional sources of information (CORE, 2000). Second, the General Accounting Office was asked by Congress to examine NMFS' compliance with several aspects of the MagnusonStevens Act, including use of the best available scientific information and consideration of economic effects of fisheries management on communities. The GAO recommended that NMFS: increase the involvement of the fishing industry, its expertise, and its vessels in fishery research activities in order to expand the frequency and scope of NMFS' data collection efforts, review data collection requirements placed on fishermen to limit requested informa-

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA tion to what is needed for conservation and management, regulation, and scientific purposes, and review data collection procedures for fisheries where the recreational sector constitutes a major portion of the fish caught to minimize the inconsistent treatment of commercial and recreational fishermen (GAO, 2000, p. 29). It is clear from these activities, both initiated by Congress, that improving data collection is a priority for Congress. Each region of the United States uses different methods to collect, manage, and use fisheries data. In part, such differences are based on differences in the biology and social aspects of the fisheries. Many differences may be due to tradition and familiarity with certain approaches and the accumulation of past actions, rather than rational choice. Other differences arise from state and federal legislation that requires or permits specific activities. Data Needed for Different Management Methods What biological, economic, and social data are most needed to provide assessments suited to five common management methods? current state of the fishery management goals and measures of their achievement of intended effects management actions needed to achieve management goals Five common management methods include (Table 3-2): Total allowable catch (TAC) Effort management Gear restrictions and fish size limits Closed areas (see NRC, 2000) Closed seasons Current State of the Fishery Fishery status questions address not only the current status of the stock but also the fishery as a whole, including social and economic factors. Relevant questions include What is the current spawning stock biomass level? What is the current level of fishing mortality? Is recruitment being sustained? Is growth potential maximized? What is the effect of fishing, if any, on the ecosystem? What is the essential habitat for the species and what is the status of the habitat? What social and economic benefits are realized from this resource? What is the relation between current fishing capacity and the sustainable yield of the fishery? All types of management have specific needs for answering the system status question. For TAC-based management, it is essential to know the current catch, and for effort management to know the current effort. For management based on gear restrictions or individual size limits, it is important to know about the sizes of fish currently being caught and the selectivity of the gear used, and is probably desirable to know the size of fish at maturity, if a goal of management is to allow fish a chance to spawn at least once before capture. For closed areas, it is important to know the distribution of fish relative to the extent of the closed areas and the rate at which fish move in and out of these areas. With closed seasons, it is important to know the seasonal distribution of fish and the timing of spawning. In practice, data requirements may be simplified by substituting measurements of effort for fishing mortality, catch per unit effort (CPUE) for biomass, length distributions in catch and surveys for age, and recruitment survey CPUE for recruitment. Hence, data of these types should be

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA TABLE 3-2 Requirements for Biological, Social, and Economic Data For Five Common Management Methods Management Method Data Requirements TAC-based Catch and effort data Fishing mortality rate Annual TACs and estimation of recruitment Social and economic impacts of management Likelihood that regulations will foster misreporting of fishery-dependent data, including economic and regulatory discardsa Economic contributions of recreational and commercial fisheries, including supporting industries Distribution of catch among gear types and between commercial and recreational fishermen Effort management Catch and effort data Fishing mortality rate Social and economic impacts of management Optimal harvesting and processing capacity Present participation of individuals in the fishery Dependence on the fishery How efficiency of effort has changed and how effort is allocated across different species and sizes of fish Likelihood that regulations will foster misreporting of fishery-dependent data, including economic and regulatory discards Economic contributions of recreational and commercial fisheries, including supporting industries Distribution of effort and likely impacts of capacity reduction approaches Gear restrictions and fish size limits Catch and effort data Size distribution of fish being caught Selectivity of gear Size at maturity Age at first capture Encounter rate and release mortality of undersized fish Compliance with size limits Social and economic impacts of management Economic contributions of recreational and commercial fisheries, including supporting industries Impact of regulations on fishing behavior, especially where fishery-dependent data are used for stock assessments Closed areas Distribution of fish (by size and maturity, within and outside the closed areas) When and at what rate fish move in and out of the area Catch and effort data outside the closed area Social and economic impacts of management Economic contributions of recreational and commercial fisheries, including supporting industries Distribution of catch among gear groups and between commercial and recreational fishermen

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA   Catch and effort data Likelihood of regulations to foster misreporting of fishery-dependent data and economic and regulatory discards Potential shifts in fishing areas and effort Knowledge of essential habitat Closed seasons Seasonal distribution and timing of spawning Catch and effort data Distribution of fish in closed and open seasons Social and economic impacts of management Economic contributions of recreational and commercial fisheries, including supporting industries Distribution of catch among gear groups and between commercial and recreational fishermen Likelihood of regulations to foster misreporting of fishery-dependent data and economic and regulatory discards Potential shifts in fishing times or areas a Economic discards are fish discarded because they are unmarketable (because of their quality, size, species, or sex) or because a fisherman hopes to replace them with higher-value fish. Regulatory discards are fish discarded because they are prohibited by regulation from being landed, because of their size, species (prohibited species or species for which the seasonal quota has been filled), the gear used, or area fished. desirable for almost any management system. However, for fisheries in which fishing mortality is a small proportion of total mortality, it might be argued that an intensive monitoring system is of marginal value relative to the low risk of overfishing. Whatever method of management is chosen, managers need to know approximately what portion of a stock is being exploited and how exploitation must change to achieve management goals. A minimum requirement for such assessments would be some sort of general production model that includes at least catch and effort, hence the need for catch and effort data for all management types. Management Goals and System Response System response questions involve monitoring the changes in stock status in response to changes in the management control variable (e.g., catch, effort, gear, time, or area restrictions). In the case of catch and effort quotas, stock status usually is expressed by fishing mortality level or changes in relative abundance. In the case of mesh changes or size limits, system response is usually measurable in terms of average size of fish in the fishery. Closed areas or seasons are likely to require both catch and effort data, subdivided by area and time. Management Actions Answers to management implementation questions can help managers as they select actions to achieve management goals. Such questions require that biomass be estimated for TAC management and that recruitment also be monitored. However, for heavily exploited fisheries in which the spawning stock biomass has been substantially reduced from unfished levels, recruitment is often an important component that needs to be monitored, because only a few poor recruitment years are needed for the population to crash. Effort quotas require that changes in

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA gear efficiency and targeting practices also be monitored. Mesh size or size limit management requires an understanding of gear selectivity. Closed areas and seasons require information about the distribution of fish in areas that might form part of extended closures. An important focus of both monitoring and research is how fisheries respond to regulations. This is an essential component of sustainable management, and implementation uncertainty is often one of the largest sources to total uncertainty about a stock-fishery system. Data Quality Required In an ideal world, all advice would be completely accurate and precise, but in practice data contain some level of bias and random variation, and incremental gains in precision and accuracy often require ever greater relative expenditures for sampling and analysis. Moreover, some sources of inaccuracy or imprecision may be impossible to eliminate, even with infinite sampling, because of the inherent randomness and chaos in natural systems. Pressures to maximize total allowable catch can lead to excess fishing that can harm a fishery before managers understand the dynamics of the target fish population(s). Fisheries can also be damaged when pressures to maximize total allowable catch cause managers to attempt to manage at a level of detail finer than available information will allow. Every management system should be evaluated in light of the amount of inaccuracy and imprecision in management advice that can be tolerated and still allow the system to achieve its goals. The precision of data needed depends on the management regime and objectives chosen. For example, management with closed areas would lower the precision needed for data outside the closed area.1 Taking a different management approach, with an objective of keeping catch (and employment) as high as possible, subject only to the fish being able to reproduce sustainably—the apparent goal of many U.S. fishery management plans—requires accurate and precise estimates of current stock status, minimum levels of spawning stock biomass, and fishing mortality. The higher the rate of exploitation, the more precision is needed to manage a stock adequately from a biological perspective. From economic and social views, a high level of precision may be necessary in order to avoid undue disruption in the industry. Acceptable levels of imprecision and inaccuracy also depend on the extent of annual variations in management restrictions that will be tolerated by managers and fishermen and the ability of a stock to withstand inevitable over- or underexploitation caused by inaccurate or imprecise management. Fluctuations in total allowable catch due to imprecise data would require effort to move in or out of the fishery and would probably increase costs compared to a situation in which the TAC is lower but less variable. For a given level of precision, the amount of data required (though not necessarily its cost) is, as a first approximation, independent of the size of the stock. It would be wise to compare the management method (and data collection costs) to the potential benefits of management. Managers may not be prepared or able to pay for the levels of sampling that would provide an appropriately precise fisheries assessment for some low-value stocks. If managers are not prepared to pay for greater precision, or if needed precision is not achievable at any price, managers may have to modify either their objectives or their control rules. One approach would be to select a lower level of exploitation so that stock abundances would change more slowly and fluctuations in numbers of young fish would be dampened by 1   The amount of area that needs to be closed to avoid the need for high precision data is unknown in practice, but modeling studies have indicated that as much as 30-70 percent of total fishing area may need to be protected, if this is the only form of fishery management. See NRC (2000) for a summary of the state of knowledge regarding the use of marine protected areas for fisheries management.

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA the presence of more age classes in the fish stock. Another approach would be to adopt a form of management that needs less precision in estimates of the current stock status and estimates of next year 's catch, for example, managing a fishery by limiting fishing effort rather than by limiting catch. Closed area management requires even less assessment information and is more robust to any lack of precision in assessment. If a sufficiently large part of the range of a fish stock (including spawning areas and nursery areas) were closed, it is unlikely that fishing could damage it. However, closed areas would probably not reduce the need for precise data for highly migratory stocks. Timeliness of data is a final aspect of data quality. Sampling may be designed adequately to ensure that data are accurate and precise enough for management purposes, but data analysis may not be timely. Lack of timeliness can hinder good management, particularly in the case of heavily exploited fish populations with few year classes and for fisheries that depend on inseason management. State data are sometimes only available a year after their collection and recreational data often are not available until the following season. This lag in data availability results in management that responds to the situation that existed one year ago, a situation that may no longer exist. This may explain, in part, the finding of NRC (1998a) that assessment results tend to lag behind the actual situation by one or more years in detecting stock declines and rebuilding. Timeliness of data is also affected by the frequency of surveys, discussed elsewhere in this report. METHODS OF DATA COLLECTION Data are available from a number of sources, including from ceremonial and subsistence fisheries, from fishery-independent surveys conducted by the states and NMFS, and from commercial and recreational fisheries. Data from Ceremonial and Subsistence Users Many fisheries are exploited for ceremonial or subsistence uses. For example, halibut and salmon are prominent fish species used by Native Americans in the Pacific Northwest and Alaska for both ceremonies and subsistence. Non-natives in these areas are also subsistence users. Pacific Islanders use coastal fish species and tuna for similar purposes. Data related to ceremonial and subsistence users are collected for inland waters and Pacific coastal waters and used in stock assessments. Most Pacific coast ceremonial and subsistence data relate to salmon fisheries, but groundfish catches for these uses also are included in landings data that NMFS receives from states. Subsistence use, although small in comparison to recreational and commercial use, is still significant in many U.S. fisheries. In some cases, subsistence use may be included in the recreational fishing category, accounting for individuals who regularly fish off the shore, piers, and other coastal access points to provide food for themselves and their families. Such individuals may be contacted by MRFSS intercept samplers, but they may be missed in telephone surveys because of language difficulties, mistrust of government agencies, or because they do not have telephones. Non-commercial catches may form a large percentage of the diet in some communities, but this has not been studied extensively. Another small component of subsistence use is the catch that commercial fishermen take for personal use. In any case, except for the examples given earlier, catch for ceremonial and subsistence fishing is a minor portion of the catch in most fisheries. Data from Fishery-Independent Surveys NMFS and individual states conduct a variety of surveys throughout the year in offshore and inshore waters. Some federal surveys are conducted as many as three times per year (East

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA Coast flatfish), whereas other species may be surveyed every three years (many West Coast and Gulf of Alaska fisheries) or never (Table 3-3). Appendix C illustrates and analyzes the variety of surveys conducted for summer flounder. In addition to surveys of fish abundance and population characteristics, the states, NMFS, other agencies, and academic scientists collect data related to other components of marine ecosystems and marine environmental conditions in an attempt to understand how fishing affects marine ecosystems and how marine environmental conditions affect fish populations. In general, the purpose of stock assessment activities is to monitor changes in the abundance of fish populations over time in order to evaluate the effects of past and present fishing activities on fish population trends and to predict the consequences of future fishery management decisions. Stock assessments, together with monitoring of physical and biological variables, are also needed to evaluate the effects of the environment on fish populations. Monitoring changes in abundance of fish stocks over time requires having at least one measure that reflects these changes without biases or with constant known biases. Catches from the commercial fishery may fluctuate from year to year due to causes unrelated to changes in absolute abundance. For example, changes in commercial catch can result from changes in the amount of fishing effort in any one year as a function of the price and abundance of alternate fish species, improvements in fishing technology (better nets, more precise acoustic detectors or navigational equipment), changes in management measures (closed areas, seasons, trip limits), or inaccessibility of the stock due to changes in the ranges of fish populations caused by environmental factors. Year-to-year changes in the distribution of fishing effort should be considered when using fishery CPUE data to measure fish abundance. If a fishing fleet moves from fishing grounds where fish densities are low to grounds where fish densities are high, CPUE will increase even though the overall stock abundance remains constant or declines. Analyses of CPUE data must adequately consider the spatial aspects of fish population distributions and the fishing effort applied to catching fish. Often, however, CPUE data are simply combined over broad (and inappropriate) spatial scales. In most fisheries, the best measure of relative fish abundance is obtained from fishery-independent surveys, in which the gear (and usually vessel), timing, survey design, and procedures are kept constant from year to year. As a result, annual changes in the abundance or biomass of a species are assumed to reflect actual changes in relative abundance. Surveys are intended to determine whether populations have changed relative to previous years; typically they are not designed to determine absolute abundance. In addition to tracking the relative abundance of fish stocks over time, fishery-independent surveys provide a means to gather information unattainable from landed catch (e.g., maturity indices, fishery indices for sublegal-sized fish). Operationally, the general practice of this kind of survey is to use fishing gear of a type commonly used in the fishery. However, there have been cases (e.g., the crab fisheries in the Bering Sea and the Gulf of St. Lawrence) in which bottom trawls were used for the survey while crab pots or traps were the only gear used by the fishery. Even if the gear chosen for the survey resembles that in common use when the surveys were initiated, the fishing industry can continue to upgrade and improve its gear. This usually creates the perception in the fishing industry, many years after the survey series has started, that the survey gear is old-fashioned and sub-optimal. Such perceptions also lead to charges that the outmoded survey series is not useful because of the fishing gear used in the survey. Although it is true that there have been many improvements to fishing gear and practice over the past thirty or more years, criticism of a survey series should be based more on whether the current gear is working properly and whether the selectivity of the gear is well known, rather than on whether that type of gear still is being used by the fishery (see Box 2-2 for details on gear selectivity). A further consideration is that when there is an important change in gear, both old and new gear should be used in parallel for a long-enough period of time to establish the conversion factor needed to use historical data.

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA TABLE 3-3 Research and Charter Vessel Surveys, NMFS Fiscal Year 2000 Type of Survey Area Species or Species Complex Frequency or Seasonality of Survey Number of Stations Atlantic and Gulf Surveys Autumn bottom trawl survey Cape Hatteras to Nova Scotia, 4-200 fathoms Fish and macro-invertebrates Annual 355 trawl/CTD stations Winter bottom trawl survey Cape Hatteras to Georges Bank, 15-100 fathoms Fish and macro-invertebrates Annual 155 trawl/CTD stations Spring bottom trawl survey Cape Hatteras to Nova Scotia, 4-200 fathoms Fish and macro-invertebrates Annual 335 trawl/CTD stations Northern shrimp bottom trawl survey Gulf of Maine, 50-120 fathoms Northern shrimp Annual 65 trawl stations Sea scallop survey Cape Hatteras to Georges Bank, 15-60 fathoms Sea scallop Annual 600 dredge stations and 300CTD profiles Surf clam/ocean quahog survey Cape Hatteras to Georges Bank, 4-40 fathoms Surf clam/ocean quahog Triennial 475 hydraulic dredge stationsand CTD profiles Apex predator survey Key West to Delaware Bay, 5-40 fathoms Shark Triennial 100 longline stations andprofiles Atlantic herring hydroacoustic survey Georges Bank and the Gulf of Maine, 10-200 fathoms Atlantic herring Annual 3,400 nautical miles of hydroacoustic trackline; ~70 pelagic trawl tows Small pelagics hydroacoustic survey Cape Hatteras to Nantucket Shoals, 10-100 fathoms Atlantic mackerel, butterfish, loligo and illex squid, and Atlantic herring Annual 3,400 nautical miles of hydroacoustic trackline; ~70 pelagic tows Trawl survey standardization and technology development Cape Hatteras to Nova Scotia, 5-100 fathoms Gear efficiency study Annual Variable

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA Ecosystem monitoring survey—winter Gulf of Maine, 15-200 fathoms Cape Hatteras to Georges Bank, 4-200 fathoms Multi-species eggs and ichthyoplankton Annual 30 bongo hauls 90 bongo hauls/CTD profiles Ecosystem monitoring survey—early spring Cape Hatteras to Nova Scotia, 4-200 fathoms Multi-species eggs and ichthyoplankton Annual: survey piggybacked 120 bongo hauls/CTDprofiles Ecosystem monitoring survey—late spring Cape Hatteras to Nova Scotia, 4-200 fathoms Multi-species eggs and ichthyoplankton Annual 120 bongo hauls/CTD profiles Ecosystem monitoring survey—summer Cape Hatteras to Nova Scotia, 4-200 fathoms Multi-species eggs and ichthyoplankton Annual 120 bongo hauls/CTD profiles Ecosystem monitoring survey—early autumn Cape Hatteras to Nova Scotia, 4-200 fathoms Multi-species eggs and ichthyoplankton Annual: survey piggybacked with autumn trawl survey 120 bongo hauls/CTD profiles Ecosystem monitoring survey—late autumn Cape Hatteras to Nova Scotia, 4-200 fathoms Multi-species eggs and ichthyoplankton Annual 120 bongo hauls/CTDprofiles Northern right whale survey Bay of Fundy to the Gulf of Maine, 15-200 fathoms Whales Annual Visual line transect survey with 100 plankton andCTD stations Harbor porpoise survey Georges Bank and the Gulf of Maine, 15-200 fathoms Harbor porpoise Triennial Visual line transect survey with 30 CTD profiles Marine turtle survey North Carolina to the Gulf of Maine, 4-200 fathoms Turtles Triennial Visual line transect survey with 50 CTD profiles Harbor porpoise and hydroacoustic survey Gulf of Maine, 10-200 fathoms Harbor porpoise Annual Visual line transect survey with 50 CTD profiles Pelagic delphinid survey North Carolina to the Gulf of Maine, 4 fathoms to EEZ boundary (abyssal depths) Dolphins Triennial Visual line transect survey with ~10 plankton tows and 200 CTD profiles SABRE/striped bass North Carolina to Virginia, 5-100 fathoms Striped bass and larval fish Winter 80 striped bass stations 80-100 plankton tows

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA missions, and federal agencies, including NMFS. The latter intends the FIS to complement existing data collection and management planning efforts by “providing a common thread among programs to take advantage of opportunities in technology, economies of scale, and efficiencies in re-use of survey and information management experiences, and to develop a context for assessing how to pay for these activities. ” (NMFS, undated). Legal Requirements—The Paperwork Reduction Act and other federal laws, regulations, policies, and guidance are important considerations in choosing data standards. For example, Federal Information Processing Standards (FIPS),20 Records Act provisions, the Clinger-Cohen Act, General Services Administration (GSA) regulations, and OMB Circular A-130 and other OMB guidance and memorandums may prove helpful to developers of fisheries data management systems. The regional information systems discussed earlier that are expected to participate in the FIS addressed the Privacy Act provisions and the Magnuson-Stevens Act confidentiality regulations codified at 50 CFR part 600. It is not clear how the FIS plans to provide information technology access for persons with disabilities as required by 41 CFR 201-20, 103-7. In addition, the U.S. Access Board, the federal agency set up to implement the Americans with Disabilities Act, has recently completed a set of standards for access to government Web sites. OMB Circular A-130 Section 8b(4) instructs agencies to use strategies that consist of “one or more profiles (an internally consistent set of standards), based on the current version of the NIST's (National Institute of Standards and Technology's) Application Portability Profile. These profiles should satisfy user requirements, accommodate officially recognized or de facto standards, and promote interoperability, application portability, and scalability by choosing interfaces that are broadly accepted in the marketplace to allow for as many suppliers as possible over the long term.” Software Compatibility—Daspit et al. (1997) stated that in May 1992, NMFS announced that all of its computing resources would be replaced by the UNIX operating system and the Oracle relational database management system. An example of an implementation of this standard is PacFIN's use of Silicon Graphics workstations running IRIX, a version of the UNIX operating system, along with Oracle 7 Server Release 7.3.3.5.0. The ACCSP currently under development complies with the above organizational standard by including Oracle as the backend with Businessobjects Webintelligence software providing the front-end user interface and query capabilities (M. Cahall, NMFS, personal communication, 1999). Development of organizational standards (e.g., the core set of data elements, standardized data quality assurance and quality control procedures, coding standards, and metadata) and earlier standardization on the proprietary Oracle relational database management system facilitate data interchange within NMFS. However, caution is in order when standardizing on proprietary systems rather than on open standards that facilitate interchangeability among products. For example, Oracle 7.x SQL conforms to the first or entry level of SQL-92 rather than the intermediate or full level. Entry level SQL-92 is similar to SQL-89 to which Oracle has added enhancements or extensions (Harrison, 1997). These noncompliant enhancements make migration to a competing vendor's SQL-compliant database management system more difficult. When a proprietary product is used, the organization often becomes locked into the single vendor 's products. The vendor is then under little pressure to reduce cost or improve and differentiate the product. The single vendor approach also limits systems design flexibility because the user of the proprietary product must 20   The FIPS Program was established in the 1960s to standardize federal computer usage for federal agencies and organizations.

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA wait until the vendor is willing to provide the new products that the customer needs. The latest version of Oracle (8i) includes both proprietary SQL features and somewhat greater open systems support than in previous versions. The open systems approach achieves interoperability by defining interfaces, services, protocols, and data formats favoring the use of non-proprietary specifications (OMB Circular A-130). Whether Oracle represents the best commercial practice of today, users who may someday wish to change vendors must deal with migration to a new system that will not support Oracle's proprietary SQL enhancements. Such migration issues are best handled by advance planning. Compliance with Interagency Data Standards— NMFS summarized for the committee the status of its standards of compliance by saying that “due to historical regional and programmatic autonomy, NMFS and its partners do not have a single, national, integrated system or standard for data collection.” At the same time, NMFS acknowledges in the FVR-FIS document the need for coordination in the design of data collection forms, quality assurance and quality control, coding standards, and metadata. NMFS is participating in the NOAA Biodata Working Group to ensure that all NOAA staff members document data sets, make them available, and ensure that the data are available for use in the future according to the National Archives and Records Administration's requirements for archiving data (G. Barton, NOAA, personal communication, 1999). To implement a national FIS, NMFS will provide the extensive coordination necessary to define and implement organizational standards. Since the FIS is at the conceptual stage of development, NMFS has an opportunity now to expand its current plans for developing national standards for certain data elements or coding systems to include development of an architectural framework for interoperability among fisheries data management systems. In fact, the ClingerCohen Act of 1996 instructs federal agency chief information officers to take responsibility for developing, maintaining, and facilitating the implementation of a sound and integrated information technology architecture (ITA). An ITA is an integrated framework for evolving and maintaining existing and new information technology to achieve an agency's strategic goals and information management goals. A great deal of effort must still be expended to complete the critical task of translating incompatible data formats now used by the regional systems to allow the level of interoperability necessary for the proposed FIS umbrella system. The ACCSP's use of standardized nomenclature provided by the Integrated Taxonomic Information System (ITIS) is an example of using an emerging standard to help ensure successful biological data discovery and retrieval. Documentation of how data were collected and analyzed (metadata) provides a way to understand data sets. Formal metadata standards employ a controlled or common set of terms to use when describing data. Metadata standards of potential importance to fisheries include the Federal Geophysical Data Committee (FGDC) standard implemented by the National Geospatial Data Clearinghouse and the National Biological Information Infrastructure (NBII) biological profile of the FGDC 's content standards for digital geospatial metadata. The NBII biological profile includes fields for analytical tools and methodologies. It also includes a supplemental information data element in the FGDC format for input of additional relevant information. Inclusion of information supporting the robustness of the data, such as the following, could be made available through supplying metadata along with fisheries data sets to promote understanding of the data and how it should be used: methodologies used; how much information has been directly measured as opposed to inferred, extrapolated, or produced by models; and level of uncertainty (do all included parameters hold true?), assumptions made, and uncontrolled variables. If estimates of uncertainty are incorporated directly into the assessment, this should be explained in the metadata.

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA The biological profile also includes a taxonomy data element that would be of particular use to biological data sets, such as fisheries data. More complete metadata would also facilitate peer review of the data and results and foster better quality control. The U.S. Fish and Wildlife Service has developed a process for identifying, defining, monitoring, and promoting data standards to ensure the compatibility of data management and usage throughout the agency. The standards and the process for their development are described at http://www.fws.gov/stand/. Information Management Architecture—Although the documents that describe existing and planned regional fisheries data management systems address many issues of great importance to information sharing, they need to expand on their description and justification of information management architecture. For the most part, the planning documents for fisheries data management provide comprehensive discussions of information system organization and policy and emphasize data collection and data element standards, but could benefit greatly from more detailed computational information, particularly on the topic of compliance with open-system standards. For example, the Reference Model of Open Distributed Processing (RM-ODP) ISO/IEC 10746 models an architectural hierarchy of viewpoints that is being used by the U.S. Geological Survey, NASA, NOAA, and other participants in the interagency Digital Earth Project to serve as a guide for project participants. The technology viewpoint is defined as the specific collection of technology products implemented in the system. The Digital Earth Reference Model provides a listing of interoperability standards by infrastructure category. Emphasis is placed on identifying relevant federal, national, and international standards rather than on defining proprietary products or organizational standard products that must be used by participants. As a final step, the organization identifies the specific technologies that can be used to meet the standards. For example, the FIS designers could decide at this time that they require only SQL-92 level-1 compliance, so that Oracle would be a viable technology choice to fit the fisheries information system architecture. Vendors other than Oracle, however, may meet the required standard. In fact, in its VRS-FIS document, NMFS (undated) indicates that a process will be designed to identify and evaluate candidate technologies according to specific criteria. The Raines Rules (OMB Memorandum 9702) indicate that agencies must report how well they developed information architectures and evaluated prototypes. Executive Order 12906, issued in 1994, established in the Executive Branch of the federal government a National Spatial Data Infrastructure (NSDI) and a National Geospatial Data Clearinghouse. This executive order directed the FGDC to develop standards for implementing the NSDI and directed individual agencies to use such standards or require their use by entities from which they obtain data. The executive order also directed the FGDC to submit a plan for implementing a national digital geospatial data framework. Earlier in this section, the statistical aspects of data quality were discussed. The following section describes the approaches to quality control used by planned and existing fishery data management systems. PacFIN The content of each PacFIN data file is the responsibility of the agency that provides the data; thus, the current PacFIN system does not include comprehensive validation routines, though some data validation routines are in place. An example of a current PacFIN validation routine is the standard duplicate check. If a transaction is a duplicate or includes out-of-range values, it is flagged as an error and rejected (B. Stenberg, PSMFC, personal communication, 1999). Sampson and Crone (1997) documented data collection procedures for U.S. Pacific Coast groundfish. Some data are subjected to rigorous

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA quality control before submission to PacFIN. The groundfish trawl logbooks used in Washington, Oregon, and California provide examples of the data quality procedures followed before transmission to and entry into PacFIN (Tagart, 1997). The trawl logbooks in Washington State typically are collected by the port sampler each time a fisherman completes a trip. The port sampler records the logbook data to disk with the aid of custom software that conducts cursory error checking. Coded logbook data are then sent to the state Marine Resources Division, where the data are stored and transmitted to a data specialist, who ensures that individual files are filtered through a comprehensive error-checking program, after which they are considered to be processed data. An errorscreening program checks raw logbook data for out-of-range errors in Loran or GPS coordinates, depth, fishing block, species, port, and trip type. It also screens for such errors as tow information entered in the wrong column and missing data. Records with errors are flagged in the database and a separate file is generated that describes the type of error and records the data line in the raw trawl data file in which the error occurred. The data specialist then rectifies the error by reviewing the raw data or returns the coded data to the port sampler for clarification. This procedure results in more than 95 percent of the logbooks being free of coding errors after two passes through the errorscreening program (Sampson and Crone, 1997). Processed data are then aggregated into a single file and further processed into tow-expanded logbook data to account for tows that were not keypunched.21 Tow-expanded data are next processed with fish ticket data to generate expanded trawl logbook data. Different error-checking protocols are implemented in Oregon and California, with no single standard. Oregon accepts incomplete logbooks, but codes them with a number indicating their incomplete status. According to Sampson and Crone (1997), the Oregon local port biologist “evaluates every logbook for completeness and consistency. The process includes checking the logbook for incorrect temporal sequencing of the tows or inappropriate dates or times, and filling in the following items: (1) the ticket number(s) corresponding to each trip, (2) missing depths based on the tow location and the depths indicated on the nautical charts, and (3) missing target species based on the most prevalent species hailed. The port biologist assigns each logbook a code of 1, 2, or 3, depending on its degree of completeness.” Sampson and Crone report that the port biologist will attempt to obtain missing information by interviewing the captain (the logbook has only partial information on the tow location or hail weights).22 If the captain does not provide the missing information, the port biologist will assign the logbook a code 2 if only hail weights are missing, or a code 3 if tow locations are missing. Logbooks assigned a code 3 are excluded from further processing (Sampson and Crone, 1997). ACCSP The ACCSP database planning document notes that data should be checked for accuracy and consistency before being submitted to the coastwide database. The data form review re- 21   According to Sampson and Crone (1997), Washington State did not record tow-by-tow logbook data until 1985. Originally, they keypunched every fourth tow from each logged trip, in effect subsampling tow-by-tow data. Thus, their system accommodates subsampling and provides for data expansion for subsampled trips, for example, the tows that were not keypunched. 22   The hail weight is an estimate by fishermen of either tow-by-tow landings (entered into the logbook), or of the total weight the fishermen tells the fish buyer he has on the boat as he returns from a fishing trip. When the fish are subsequently weighed, this “landed weight ” is usually used to adjust the estimated weight for each tow in the logbook. This is why it is so critical to link fish tickets (weight at the dock) with logbooks (estimated catch by depth and location).

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA quires that records should be checked for at least the following items: legibility completion of all necessary fields reasonableness of dates and times accuracy of species and gear combinations The ACCSP recommends that: incorrect or inconsistent entries follow data protocols that include contacting the data provider. (Problem data providers who consistently make errors may be given additional training or legal action may be considered [e.g., fines, license revocation].) reviewed data forms be entered into the database by adequately trained data entry clerks, with an error rate of less than 0.5 percent for the set of all data entered, through use of a double entry system (each data point is entered twice and not accepted unless both entries are identical). someone other than a data entry clerk should perform a spot check for errors on 5-10 percent of a year's entries. the following standard computer data edit checks, at a minimum, are run: species ranges, lengths, and weights dates of catch fisherman and dealer licenses fishing gear used invalid codes outliers blank fields comparisons with tracking database The edit checks flag errors and probable errors, alerting the data entry clerk and permitting changes before the data reach the database (ACCSP, undated b). The program design document also mentions that unannounced audits of dealers' and fishermen's records may be used as a data verification tool (ACCSP, undated b). Version 1.5 of the application also advocates that summary reports, similar to monthly bank statements, be sent to fishermen on a periodic basis for data verification. Benefits of such reports include (1) allowing fishermen and dealers to see the data after they have been entered, increasing confidence that their data are being used; (2) giving fishermen an opportunity to correct erroneous data, thus improving accuracy; and (3) providing fishermen with an official record of what they have caught and their revenues. All the data quality routines discussed here supplement those implemented by the data sources, since the ACCSP agreement vests responsibility for the quality and completeness of the archived records with the agencies that originally collect the information (ACCSP, 1999). ACCSP requests that the states implement standard operating procedures (SOP) and develop SOP manuals, and that members sign agreements to use a specified standard of data collection elements and reporting formats. Data will be collected by individual fishing trips, including a standardized collection of elements such as species, area fished, gear type, quantity and value of catch, and vessel identification number. Members will also use standardized units of measure, coding systems, and nomenclature whenever possible. The Georgia program that will provide data to ACCSP is illustrative. The draft SOP manual notes that the Georgia Coastal Resources Division 's statistics project is under the Commercial Fisheries Program in the Marine Fisheries Section. Historically, the project has been funded by the NMFS Cooperative Statistics Program. In 1999 the project received additional funding through the Atlantic Coast Fisheries Cooperative Management Act (ACFCMA) to implement a commercial fisheries trip ticket program that would comply with the ACCSP (Anonymous, 1999). Commercial landings data are collected with each fishery's self-coded trip tickets, which contain pre-labeled columns for the market grade and condition of the predominant species. In some fisheries, the gear quantity and area fished are

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA also labeled. Seafood dealers receive the trip ticket forms, postage-paid envelopes, and a plastic card imprinter engraved with their business name and dealer code. Seafood harvesters are provided with a personalized plastic card embossed with their name and commercial fishing license number. Port agents code the remaining fields using four-digit species codes and threedigit gear codes provided by NMFS and 6-digit area codes supplied by the Georgia Department of Natural Resources. After all data have been reviewed for completeness, they are entered by the port agents and a data clerk into a customized database implemented with the PROGRESS relational database management system. The data are edited at three different points. First, while coding and conversion take place port agents identify illegible fields and odd species and gear combinations, and track down missing data. Second, as data are entered, the data entry software checks for invalid codes, gear combinations, and price ranges. Final data editing is accomplished by running edit programs to check for outliers and by spot checking data forms against the data set. All data are entered within 10 days of receipt by the statistics program (Anonymous, 1999). According to the Georgia SOP, the PROGRESS database does not contain the fields that were added to bring Georgia into conformance with ACCSP data standards. A new database is under development in Oracle. The Oracle application is currently being tested. Sometime later, scanning of the trip ticket form will begin and images of the form itself will be created and archived on read-only CDs and stored in a safe deposit box. This will eliminate the need to store paper documents and still meet Georgia's requirements for archived records. After completed landings data are run through a PROGRESS program that converts all fields to the SEFHost format in ASCII code, data are then transferred to NMFS monthly by email (Anonymous, 1999). Fisheries Information System The VRS-FIS report to Congress (NMFS, undated) presents the following design principles for standards of measurement and quality for a future FIS: Establish standardized units of measurement and nomenclature, where possible. Establish standard coding systems, where possible, or build logical bridges or translations between separate coding systems, where necessary. Establish reasonable minimum data quality standards. Establish standard (minimum critical) data elements. Minimize number of coding systems. Develop processes to ensure the timely release of information to the public. The FIS design principles indicate that the shortcomings of the existing fisheries data management systems are well known. The need for standardized data elements to allow comparability among systems is apparent. The proposed FIS includes funding of $1.575 million to establish and implement criteria and processes for evaluation of data quality and data quality standards. These funds would be used to: research and adopt nationwide data quality standards, with help from individuals from universities, other federal agencies, and private research contractors familiar with large-scale data quality issues. establish nationwide data quality control groups to provide continuous oversight and peer review of both data collection and data quality processes. research, design, and implement validation methods for self-reported statistical systems (e.g., logbooks) to measure and document the biases and accuracy of such data. create online metadata files containing system statistical information to improve avail-

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA ability of documentation on quality of the information. Technologies for Data Management Fisheries data collection, analysis, use, and archival storage present a challenge to the organizations responsible for providing accurate, timely, and easily accessible fisheries data. It is evident not only from existing systems, but also systems in various stages of planning and development, that the existing fisheries data management systems are heterogeneous, are often incompatible and/or duplicative, are developed by different organizational entities, have a regional focus (so far), do not necessarily share data elements, and may or may not comply with federal data management standards. NMFS acknowledges that “despite some regional successes, it is clear that the current overall approach to collecting and managing fisheries data needs to be re-thought, revised and reworked. The quality and completeness of fishery data are often inadequate. Data are often not accessible in an appropriate form or a timely manner. ” (NMFS, undated). Fisheries Data Integration Since existing fisheries data management systems are heterogeneous, integration is of primary importance to enable data synthesis on regional and national scales. The lack of standardized data elements necessitates implementation of “translators” to allow incompatible data elements residing in different databases to be accessed through a single interface. The ACCSP is an example of a regional fisheries data management system that initially will employ translators to reconcile incompatible data elements now collected by participating states (ACCSP, undated b), including both commercial and recreational data. The FIS will take the process one step further by integrating ongoing regional fisheries data management activities (e.g., FIN and PacFIN) under a nationwide umbrella (NMFS, undated). The proposed system recognizes the need for implementing national standards for a core set of data elements, data quality protocols, coding standards, and metadata. The importance of fully defining and implementing these standards cannot be overemphasized. NMFS' current FIS plans could be enhanced by development of a framework that provides more detailed computational information with particular attention to compliance with open-system standards. Another aspect of data integration is use of a format that allows full exploitation of the data asset, for example, verification of different data sources used in management and research. This could be as simple as matching logbook data against trip tickets. The existence of a unique identifier for each fishing vessel as planned in the FVR system—together with the date and time of each trip—will facilitate data verification by providing positive identification of vessels and trips that can be used across databases. Such a feature should enhance the confidence of managers in data from different sources. Historically, fisheries data have focused on individual species and their population dynamics. In recent years, more attention has been given to the predatory and competitive relationships among species, as well as finding commonalties across species (e.g., meta-analysis, Myers et al., 1995). Research on biological interactions and other ecosystems research would benefit if data were collected in compatible formats and were integrated in ways that facilitate study of the complex interrelationships in marine ecosystems. This could be a useful adjunct to traditional (and still necessary) stomach content analysis. Geographic Information Systems—Geographic information systems (GISs) allow spatial data from many sources, such as sampling tows, to be referenced to a single grid of spatial coordinates. The use of GIS techniques offers promise for combining data from many sources into a single spatial grid, with new possibilities for understanding how various factors affect fish populations at various spatial scales.

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA GIS applications and visualizations show promise as outreach tools to improve communication of the results of fisheries science and management activities to the general public. They could also promote better understanding among fishery scientists. Because these applications already exist, making them more widely available and easily used could benefit all stakeholders in fisheries management and promote cooperation among data providers and other stakeholders. The Open GIS Consortium states that “GIS information is available on the Web —in stovepipes.23 Users must possess considerable expertise and special GIS software to overlay or otherwise combine different map layers. ” (Open GIS Consortium, 1999). This limits the ability of scientists, managers, fishermen, and others to use such data. The Open GIS Consortium is currently conducting a Web Mapping Testbed (WMT) project. This project is an “accelerated, multi-phase effort to meet the market's demand for interoperable geo-enabled Web technology. The project is advancing the state of Web technology to support diverse applications that access distributed geospatial information sources across the Web. Applications include environmental analysis and management. ” (Open GIS Consortium, 1999). Sponsors of WMT include the FGDC, United States Geological Survey (USGS), NASA, the U.S. Department of Agriculture (USDA), a group of 24 Australian government and commercial organizations, Pennsylvania State University, 23 vendor companies (including Microsoft, ESRI, Sun, Oracle, Cubewerx [Canada] and others [Anonymous, 1999; Open GIS Consortium, 1999]). Participants in the WMT are combining their expertise to make it possible for overlays and combinations of complex and essentially different kinds of GIS information to happen automatically over the Internet, despite differences in the underlying software (Open GIS Consortium, 1999). NMFS and other fishery organizations should consider responding to the Open GIS Consortium 's call for participants in subsequent project phases. Visualizations—Most fisheries data are presented in tabular format in reports, papers, and Web pages. In some cases, plots on two-dimensional maps represent sampling results. Many opportunities exist to provide better visualizations of fisheries data to promote better understanding by scientists, fishermen, and the general public. These could include 3-D virtual underwater worlds highlighting bottom topography, currents, concentrations of fish, and other features. Examples using Virtual Reality Modeling Language (VRML) technology are available at the Web site of NOAA 's Pacific Marine Environmental Laboratory (http://www.pmel.noaa.gov/home/visualization/visual.html). The Fisheries Oceanography Coordinated Investigations Web site (http://pmel.noaa.gov/foci/visualizations/visual.html) includes a virtual reality world, as well as other visualizations. The Pacific Fisheries Environment Laboratory (PFEL) Web site at www.pfeg.noaa.gov features a live access server that allows visitors to visualize and download selected PFEL data products. Visualizations could be used as a tool in fisheries simulations for both teaching and consensus-building purposes. DATA USE Data use is discussed only briefly here because it was the focus of the NRC's Improving Fish Stock Assessments (NRC, 1998a). Uncertainties of Data in Stock Assessments A good example of the potential complexity of data sources used in any fishery is the summer flounder fishery. All the forms of data discussed earlier in this chapter are available for this fishery (Table 3-10). Stock assessments are subject to uncertainties of various types, ranging from uncertainties in observations to implementation of management. 23   Stovepipes are systems that stand alone and do not inter-operate.

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA TABLE 3-10 Summary of Data Sources and Source Attributes for Summer Flounder   Types of Data Attributes of Sources   Spatial Information Species Composition Size Composition Age, Sex Composition Price Confidential Accessiblea       Federal       Mandatory logbooks 2 2 0 0 0 2 1 Scientific observers 2 2 1 1 0 2 1 Scientific survey 2 2 2 2 0 1 2 Recreational survey 1 1 1 0 0 2 2 Port sampling 1 2 2 2 1 0 2 VMS 2 0 0 0 0 2 0       States (see Appendix C for listing of states)   Landings 1 2 0 0 2 0 1 Surveys 2 1 1 1 0 0 2 Key: 0 = none; 1 = some; 2 = complete. a Accessible refers to the extent that data are shared between those with the responsibility to collect the data and those who may have use for the data (e.g., scientists, managers). One level is the degree of uncertainty reflected in an observation, typically referred to as the observation error. This is the variation that would be seen under repeated sampling. The observation might be survey CPUE for a given year, and so would be a statistic summarizing a number of points gathered under a prespecified sampling design. A second level of uncertainty in information comes in through the natural variation that occurs in the environment, so that even if every individual in the population is measured, variation in size and abundance is still expected from one year to the next due to the natural variation in the individual growth and population dynamics. This variation is referred to as process uncertainty and does not represent our ability to measure but rather represents expected natural variation in the process. The third level of uncertainty is reflected in how the system is characterized, which is typically done through specification of a model. The choice of model interacts with the previous two levels of uncertainty and reflects to a degree a choice between bias and variance in the estimation process. Models may characterize the system simply or with more complexity, but they may also mischaracterize the system through selection of a model that does not represent actual processes well (model misspecification). The uncertainty in model specification can be estimated only by challenging the data with different models and/or additional kinds of data, as was done through the committee's analysis of summer flounder data. Added to this is the fact that only a single realization of the data is available, a single series through time, and thus the overall uncertainty cannot be assessed with repeated time series. It is sometimes possible to get around this problem by using meta-analysis across similar systems or through adaptive management that allows informative exploration of the system. Finally, implementation uncertainty is an expression of “the

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA ability of management to achieve a particular harvest rate in any one year” (Rosenberg and Brault, 1993) and institutional uncertainty arises from “the interaction of the individuals and groups (scientists, economists, fishermen, etc.) that compose the management process ” (O'Boyle, 1993). In the end, the output from our assessments poorly represents the levels of uncertainty that exist, impeding an adequate assessment of risk in decisionmaking and informed development of research priorities. Access to Data Users of fisheries data management systems include fishery managers, scientists, fishermen, industry groups, and other interested parties. Because fishery databases contain sensitive business information, most restrict access and usage to those with a need to know. The specific approach chosen for authorization and access controls varies among existing fishery databases. PacFIN, for example, uses logins and a list of authorized or allowed Internet Protocol addresses. There is currently no interactive Web access to PacFIN. PacFIN and several other fishery databases provide summary reports on the World Wide Web for public access. The separation of general public access and restricted access to the data may impact the number of visitors recorded at fishery data Web sites. Aside from the traditional sources of fisheries data, opportunities exist for overlaying data from other studies. GISs could help in accomplishing such overlays. For example, the U.S. Global Change Research Program is supporting field studies of the dynamics of fish and plankton populations and of the causes of variations of marine biological populations in the Global Ocean Ecosystems Dynamics (GLOBEC) and other activities (Our Changing Planet, fiscal year 1999). The California Cooperative Oceanic Fisheries Investigation (CalCOFI) has compiled a long time series of data related to fisheries for the California Current System. In these cases, fisheries data management could either gain from or contribute to overlaying data from other studies. Management Information Needed by Councils The committee sent a list of questions regarding fishery data issues to the executive directors of each of the eight fishery management councils. Some councils responded by stating that they wanted annual assessments for each species rather than the staggered subset of assessments available each year. The staggered schedule is a result of inadequate funding for assessment personnel and ship time, but councils with fisheries in an overfished and rebuilding mode have said that they need annual assessment updates, so they can manage the rebuilding process more effectively. Information relevant to management priorities was requested. Several councils noted that NMFS is limited by the availability of assessment personnel and that NMFS cannot go beyond its minimal fulfillment of legal mandates. The councils identified several specific information needs (beyond what they already receive), including the following: Baseline information to manage the new national standards established by the Sustainable Fisheries Act, particularly in relation to by catch and effects on fishing communities. Although fishery management plans must contain social and economic impact analyses, NMFS and states are still not doing an adequate job of collecting data and providing them to the councils. Economic data, such as would be required to estimate consumer benefits, construct bioeconomic models, analyze vessel costs and returns, and describe fishing community structure. Enforcement data from NMFS, for example, how many landings are examined for rates of compliance and non-compliance with management measures such as trip limits. Data on the stock status of artisanal fisheries. Tagging data to study stock interactions and fish movements.

OCR for page 59
IMPROVING THE COLLECTION, MANAGEMENT, AND USE OF MARINE FISHERIES DATA Fisheries Data Discovery Users must discover fisheries data in order to use it. Most fishery databases have a public component with Web access to standardized reports. These sites have home pages indexed by the common search engines that allow discovery by the public, made possible by metatags. 24 However, not all fisheries Web sites are currently using metatags. As of June 1999, among the AKFIN, PacFIN, RecFIN, and ACCSP home pages, only RecFIN included metatags in the document source. The use of metatags should be promoted because they are used by many search engines when indexing pages and thus provide an aid to document discovery. Fishery databases also have a private component, with access restricted to researchers, managers, and others with a need to know (see earlier section on confidentiality). This restriction is related to the inclusion of sensitive business information in the database and in some cases is mandated by state or federal law or both. In general, persons who desire access to restricted fisheries information must sign a non-disclosure agreement and fill out a database access request form that must be approved by a database official. The NOAA server (http://www.noaa.gov) provides unified access to fishery data sets held by the NMFS Northwest Fisheries Science Center, with discovery through a query of FGDC metadata. The server system is now being redesigned so that access to planning documents will require a password. The existing system allows a user to select a query term, but the terms are very general. There is also a spatial query interface based on either latitude-longitude or a map. This appears to be a secondary access method with only minimal metadata included about NMFS fisheries data sets. It is often necessary to contact the individual listed in the metadata to gain access to the data. Landings information is available from the NOAA Web site for commercial (www.st.nmfs.gov/commercial/index.html) and recreational (www.st.nmfs.gov/recreational/index.html) fisheries. Cooperation and Communication The committee did not investigate all possible communication links, but it did query the regional fishery management councils about how they obtain the data they need for management decisions and whether they wanted information in a different form or wanted different information. It is obvious that cooperation in data use is essential for effective management. Different regional councils accomplish cooperation differently. In some cases, councils receive data and information on a regular basis. Most councils rely greatly on members from NMFS, the Coast Guard, states, and other organizations to ensure that the necessary information is transferred to the councils and used in management. One council expressed the desire to gain access to non-aggregated data and requested a more efficient means of data access than transfer of data disks. Another council noted the need for better ways to “analyze and reduce information so that it may be readily assimilated by council members and the public during the decision process.” The councils believe that greater efforts and resources need to be devoted to improving communication of the reasons why data are collected in specific ways, for example, using outdated trawl gear and random, stratified sampling. It was also suggested that information communication specialists be enlisted to improve communication between councils and stakeholders. Most of the councils have extensive Web sites that include their fishery management plans, other reports, meeting schedules, committee rosters, and other information. These Web sites can be very efficient communication tools, if kept current. 24   Metatags are information in World Wide Web documents that have a number of functions, including providing keywords and descriptions of the document that are accessed by search engines to categorize Web documents.