National Standard 2: From Origin to Application
LEGISLATIVE HISTORY: ORIGINS AND RATIONALE
Congress invoked the first “best available science” clause of the modern genre in the Marine Mammal Protection Act (MMPA) of 1972. It allowed the Secretary of Commerce to waive the general moratorium on mammal takes required elsewhere in MMPA “on the basis of the best scientific evidence available and in consultation with the Marine Mammal Commission”1 (National Oceanic and Atmospheric Administration, 1998). Congress turned to a “best available science” clause again in the Endangered Species Act (ESA) of 1973, requiring that endangered species listing decisions proceed “on the basis of the best scientific and commercial data” (Fishery Conservation and Management Act, sec. 2) and after consultation with appropriate parties. Congress has used various “best available science” formulations repeatedly in amendments to these two laws: 12 similar “best available science” clauses are found in MMPA and 8 more appear in ESA. These and other environmental laws enacted in the 1970s provide context for Congress’ use of the phrase “best scientific information available” in the Fishery
Conservation and Management Act (FCMA) of 1976 (later amended and renamed the Magnuson-Stevens Act).
Congress declared in its original statement of purpose in 1976 that its intent with passage of the Fishery Conservation and Management Act was to ensure “that the national fishery conservation and management program utilizes, and is based upon, the best scientific information available” (Fishery Conservation and Management Act, sec. 2). This purpose was embodied in National Standard 2, which has its origins in the draft Law of the Sea Treaty in circulation in the early 1970s. Neither the congressional statement of purpose nor National Standard 2 has changed since 1976. A report of the Senate Commerce Committee described National Standard 2 “as an important adjunct” of National Standard 1 and stated that it “must be recognized as one of the most important standards” (Senate Committee, Fishery Conservation and Management Act, 94).
There is little doubt, given the context of the times and the paucity of knowledge of fish populations, that the original intent of National Standard 2 was that management and conservation measures would proceed in a timely fashion despite recognized uncertainties in the scientific information. In its report, the Commerce Committee insisted that “there should be no uncertainty that the basic goal of management is to protect the productivity of fish stocks” (Senate Committee, Fishery Conservation and Management Act, 94).
The Commerce Committee recognized that if certainty were required before a management action could be taken in the inherently uncertain arena of natural resource ecology, policies already recognized as detrimental would be continued under the guise of doing no harm. Hence, the consequences of not taking action must be assessed with the same level of scrutiny as other management alternatives (Dayton, 1998). This idea was later included as part of the precautionary approach to fisheries management recommended by the United Nation’s Food and Agriculture Organization (1995) in the code of conduct for responsible fishing and required by the United Nation’s Agreement on Straddling and Highly Migratory Fish Stocks.
Nevertheless, the U.S. Senate Commerce Committee did not view decision making with limited information as a panacea for future fisheries management, as evidenced by these statements from its report:
As just stated, a basic management objective is to harvest a stock of fish at the level of optimum utilization. If little is known about the size of the stock or environmental effects on other stocks or similar
relationships, however, even the best management regime will fail. Therefore another primary goal must be to achieve [emphasis added] the best available scientific information about the stocks. The term “scientific information” is meant to include not only biological and ecological data but also economic and sociological information as well. (Senate Committee, Fishery Conservation and Management Act, 94)
The practical consequence of the congressional intent in using the phrase “achieve the best available scientific information” is that the National Oceanic and Atmospheric Administration (NOAA) Fisheries and the councils also have the responsibility to improve scientific information for decisions in future years. Thus, being able to make decisions with limited scientific information should not be used as an excuse for not doing research to improve scientific information. Because NOAA Fisheries has a limited capacity to support data collection and analysis to improve the scientific information used in stock assessments, “achieving” the best scientific information has not been possible for all fisheries. In fisheries that are managed with outdated or insufficient information, some stakeholders have therefore argued that scientific information should meet an independent standard before it is deemed worthy to be used as a basis for decision making.
DECISION MAKING UNDER UNCERTAINTY
Much of the dissatisfaction with the quality of science used in fishery management decisions arises from uncertainty in the scientific information available. An understanding of both the origins of this uncertainty and how it affects management decisions is necessary to completely understand the application of National Standard 2. First, it is important to realize that uncertainty is characteristic of scientific research, especially as it relates to natural resources. This uncertainty exists due to the complexity of natural systems, the length of time required to conduct experiments in the natural world, and in some cases, the limited availability of funds for research (Rice and Richards, 1996; Francis and Shotton, 1997). There is substantial uncertainty in the scientific knowledge of fish population dynamics and the effects of fishing because the measurements of populations are imprecise, and the interactions among human, biological, and physical systems are especially complex and usually unknown.
Given that uncertainty exists, it is important that the science underpinning management decisions assess the level of uncertainty by developing and testing alternative hypotheses and conducting sensitivity analyses to determine where uncertainties in parameter estimations are likely to have the greatest management consequences. Also, such analyses will identify important gaps in information where additional research is needed. Under an adaptive management framework, new information can be incorporated as it becomes available.
There are a number of ways of expressing uncertainty (National Research Council, 1998; Patterson et al., 2001). The simplest way is to determine the statistical imprecision of an estimate based on the standard error and/or the confidence interval. More complex methods are used by stock assessment scientists to determine the relative probability that a given catch level will result in population persistence or population collapse over time. This requires translating uncertainty into an expression of risk that is then available to decision makers charged with managing risk. When presented with an explicit description of risk, decision makers are better able to evaluate actions relative to the potential consequences of undesirable and irreversible outcomes.
Currently there is no standard for determining an acceptable level of risk. An example of a court decision in which the judge ruled that the probability of obtaining a desired result (e.g., rebuilding within some time frame) must be at least 50 percent is presented in Box 2.1.
Compliance with the objective to avoid undesirable outcomes should be considered along with the costs and benefits associated with other management objectives. In fisheries, the most common objective is to first consider maximum sustainable yield and use it as the basis for defining optimum yield, as stated in National Standard 1 in the Magnuson-Stevens Act. However, it is at least equally important to avoid population levels that are so low that they substantially increase the probability of collapse of the fish stock and, by extension, collapse of the fishery.
A management decision that allows a fishery to continue at a rate that ultimately forces its closure is undesirable for many biological, economic, and social reasons. Recognizing this situation, Congress added the mandate to avoid overfishing to the 1996 reauthorization of the Magnuson-Stevens Act. This objective is consistent with the precautionary approach of “acting before there is strong proof of harm, particularly if the harm may be delayed and irreversible” (Harremoës et al., 2002).
Summer flounder (Paralichthys dentatus), a commercially valuable species harvested off the Atlantic coast, has been overfished since the early 1990s. When stocks are overfished, the fishery management plan (FMP) (or plan amendment) is required to specify a time period for rebuilding the fishery that is as short as possible (not to exceed 10 years) given the status and biology of the stock, the needs of fishing communities, and the interaction of the overfished stock within the marine ecosystem. The target fishing mortality rate for a stock maximizes the yield of a single year-class of fish over its entire life span. It represents the maximum mortality rate that will avoid overfishing while providing the optimum yield.
Despite the overfished condition, NOAA Fisheries recommended a quota for summer flounder in 1999 that afforded only an 18 percent likelihood of achieving the target fishing mortality rate. The National Resource Defense Council challenged NOAA Fisheries’ quota on the grounds that it did not provide sufficient assurance that it would meet the conservation goals of the Magnuson-Stevens Act. The court decided that the 1999 quota was unreasonable because the proposed plan had at least an 82 percent chance of resulting in a mortality rate higher than the target rate. The court suggested that the management plan should have at least a 50 percent chance of achieving the target mortality rate, observing that “only in Superman Comics’ Bizarro world, where reality is turned upside down, could [NOAA Fisheries] reasonably conclude that a measure that is at least four times as likely to fail as to succeed offers a ‘fairly high level of confidence’” (Natural Resources Defense Council v. Daley, 209 F. Supp. 3d 747, 754 [D.C. Cir. 2000]).
The strategy for avoiding undesirable states has other implications for the decision-making process that raise a number of questions about the use of the “best scientific information available.” Should there be a threshold level of uncertainty allowed in the scientific information used in fishery management decisions? It is unrealistic to require a specific level of certainty in scientific information in an inherently uncertain science such as fisheries. Such a goal would endlessly delay management decisions required to protect or utilize the resource effectively. Delay is in itself a decision with consequences that must be weighed. One way of dealing with this problem is to quantify the level of risk associated with a suite of management actions and possible states of nature to explicitly examine the trade-offs between action and inaction.
Should the same level of certainty be required in a decision that moves a system toward an undesirable state as in one that moves it away from an undesirable state? For example, to guard against overfishing, it seems reasonable to demand greater scientific certainty for a decision to increase fishing effort than is demanded for one to decrease it. This commonsense concept seems consistent with congressional intent and is a part of the precautionary approach. It is also relevant to court cases regarding the “best scientific information available,” cases that predominantly pertain to economically important, heavily fished stocks whose status is in question (i.e., there is debate about their abundance).
When fisheries initially develop, landings are typically high. Although there is often little information about the underlying fish populations that support these fisheries, high catches lead to management decisions to allow higher catches and ultimately increased fishing capacity. Over time, catches may decline due to population depletion, resulting in a need to reduce fishing capacity. The immediate consequence of any decision to reduce capacity would be economic: loss of jobs and income for fishermen and processors and a difficult economic transition for the fishing community. These decisions naturally engender greater scrutiny, often leading to questions about the scientific information highlighting the uncertainty in fish population assessments and the unpredictable effects of proposals for reducing catch. Resource users may then focus on protecting their livelihoods for the short term rather than protecting the resource for the long term and, on that basis, advocate greater certainty in decisions that reduce fishing rather than in those that increase fishing. Choosing to err on the side of short-term economic security over long-term stock stability can lead to an undesirable state such as collapse of the fish stocks and consequent collapse of the fishery.
A major shift in thinking about uncertainty and decision making has surfaced in recent years. In the scientific arena, null and alternative hypotheses are proposed and tested, leading to either rejecting or accepting the null hypothesis. In fisheries management, the null hypothesis represents the situation in which there is no fisheries impact and, thus, management action is not needed. The alternative hypothesis is that the fisheries cause impacts and, thus, management action is needed.
The traditional approach in decision making has been to set a low probability for making a Type I error, that is, the error of incorrectly taking action when none is needed (e.g., harvest restrictions in excess of those needed for sustainability). The shift has been toward setting a low probability of making a Type II error, the error of incorrectly taking no
action when action was needed (e.g., failing to regulate a fishery, resulting in overharvesting and stock collapse). Avoiding one type of error increases the probability of making the other type. This shift to emphasize Type II error, known as “reversal of the burden of proof” (Dayton, 1998), is changing the shape of scientific advice and the resulting natural resource policy. Yet formal guidance on how to balance Type I and Type II errors is not currently available.
NATIONAL STANDARD 2 IN APPLICATION
There are no federal guidelines that explicitly describe what constitutes “best scientific information available” as required by National Standard 2. However, NOAA Fisheries has published regulations that provide some specifics about the type of information to include in FMPs, and the importance of determining what type of new information is necessary to improve management (Appendix D). These regulations state explicitly that the “fact that scientific information concerning a fishery is incomplete does not prevent the preparation and implementation of an FMP” (Fishery Conservation and Management Act, sec. 2).
To understand how scientific reports and FMPs are produced in different management regions, the committee requested summaries of the process from each of the fisheries science centers and their associated fishery management councils (Table 2.1; Appendix C). Responses to the questionnaires were not complete or useful in all cases and therefore could not be used to conduct an in-depth analysis of the different procedures employed to evaluate scientific information among the regional management councils and science centers. Therefore, the guidelines recommended in Chapter 4 call for uniformity in the application of the “best scientific information available” rather than uniformity in the process. The council and science center responses, public input at the National Research Council workshop (Appendix B), NOAA web sites, and published literature provided an overview of the production and application of scientific information by the science centers and management councils.
Overall, there is broad overlap among regions in the interpretation of National Standard 2; however, there are both cultural differences among the science centers and the councils and region- and species-specific differences among the various centers and councils in its application. Regional differences result in large part from regional ecological conditions and socioeconomic standing. Indeed, the multiple fishery
TABLE 2.1 Regional Fisheries Science Centers and the Fishery Management Councils They Support
Regional Science Center
Fishery Management Council
Alaska Fisheries Science Center
Northwest Fisheries Science Center
Southwest Fisheries Science Center
Northeast Fisheries Science Center
Pacific Islands Fisheries Science Center
Southeast Fisheries Science Center
Gulf of Mexico
management councils were established to accommodate these differences. The councils have evolved as a result of regionalization to accomplish their required tasks. For example, “the eight councils take different approaches to decision making and management, as anticipated and intended by the act” (Hanna, 2002). This is evident in how the councils are structured as well as in how they address scientific questions. It is clear that the councils differ strongly in the ways they apply ecosystem principles to the fishery management process, respond to uncertainty (National Marine Fisheries Service, 1999), and organize and address problems (Miller, 1987). The quality of the data (e.g., data-rich versus data-poor regions), the value of the fishery, and the types of species being assessed may also vary from region to region, making it difficult in some cases to assess the differences in the way scientific information is used from council to council.
NATIONAL STANDARD 2 AS INTERPRETED BY NOAA FISHERIES SCIENCE CENTERS
NOAA Fisheries science centers consistently interpret “best scientific information available” as data systematically collected through established procedures and analytical products based on commonly accepted statistical techniques or models developed specifically for resource management.
Data sources and collection methodologies across science centers include both fishery-dependent and fishery-independent data gathered by numerous individuals, groups, and government agencies. Fishery-dependent data are collected by fishermen and processors through log books, trip tickets, and landing bills. They are also collected by state and federal agencies (or their contractors), through dockside intercepts (for both commercial and recreational fishers), through telephone surveys that relate to recreational fishing activities (e.g., Marine Recreational Fisheries Statistical Surveys), through telephone surveys that gather socioeconomic information, and through observer programs that provide detailed commercial catch, effort, and bycatch data.
Fishery-independent data are obtained through routine surveys conducted by NOAA Fisheries research vessels and chartered fishing vessels as well as through scientific research conducted by federal, state, and university scientists. New efforts at cooperative research between scientists and fishermen are also providing important sources of data (National Research Council, 2003). These cooperative projects engage fishers directly in the collection of data for stock assessments and management-related issues.
Typically, anecdotal or experiential information is not gathered in a systematic fashion (except as a part of specific sociological or anthropological studies) but is obtained and incorporated into the record through public comment from stakeholders (discussed below).
Not all data types are available for all fisheries. There are differences in the magnitude, frequency, and timeliness of data collection that characterize the various fisheries and regions. In addition, in some parts of the country, there is more trust and cooperation between the science center and the fishing fleet. Science benefits from that trust because of the flow of information in both directions. Alaskan fisheries tend to be data rich, with both fishery-dependent and independent surveys, while Caribbean fisheries suffer from a lack of data.
Science centers consider all of the data available to them, but do not necessarily incorporate all data in analyses. Data may be excluded a
priori if they do not meet quality (precision and accuracy) or appropriate relevance standards, are dated, or appear to represent outliers due to equipment failures. In the North Pacific, all data may be included through weighting procedures incorporated into the stock assessment models. All relevant, high-quality data are included in the analyses using assessment models designed to account for different levels of uncertainty associated with specific data sources. The emphasis in the science centers is placed on ensuring that the analyses correctly communicate the risks and uncertainties involved with a variety of possible management decisions based on the information.
The regional fishery management councils, not the science centers or their advisory committees, are responsible for selecting among management options. This helps separate scientific from political issues. Science advisory committees are charged with advising the councils on the merits of natural and social scientific information presented, not with proposing policy. Doing so would have at least two undesirable effects: (1) the advisory committee would be perceived as advocating the policy options it proposes, and (2) its meetings would become politically charged with stakeholders attending with the intent of influencing the policies identified.
NATIONAL STANDARD 2 AS INTERPRETED BY FISHERY MANAGEMENT COUNCILS
The councils generally interpret “best scientific information available” as the most recent and relevant information available to them at the time of FMP development, typically as it appears in stock assessments and other reports generated through the science centers. Several councils, while lamenting the paucity of information on a number of stocks they are charged with managing, noted explicitly that in accordance with the mandate in the Magnuson-Stevens Act, limited information did not prevent them from making management decisions. The Western Pacific Fishery Management Council stated that “although comprehensive scientific information may be lacking in our fisheries, we do our best to provide [the public, council bodies, and council members] with the best information possible to aid the decision-making process” (Western Pacific Fishery Management Council, 2003).
Management councils typically do not collect scientific data. Rather, they rely on science centers to collect the bulk of the data used in the assessments and ultimately in FMPs. They also expect the science
centers to ensure that those data meet data quality standards, as determined by stock assessment scientists.
However, the councils collect and record verbal and written anecdotal and experiential information, opinions, and recommendations from stakeholders and the interested public through responses to stock assessments and other reports, Federal Register publications, council meetings, and as part of the National Environmental Policy Act (NEPA) scoping process. This information is used in the preparation of FMPs, particularly in the development of possible management alternatives. Further, the councils often rely on the experiential information from fishermen as a means of corroborating scientific information, determining changes in stock distributions, and revealing data discrepancies. If the two types of information conflict, however, councils report that they more often than not defer to the scientific information.
NATIONAL STANDARD 2 IN PRACTICE: STOCK ASSESSMENTS AND FISHERY MANAGEMENT PLANS
The primary concern with the application of National Standard 2 for science centers is the stock assessments they conduct, and for councils it is FMPs they develop. Stock assessments contain all the available information (both published and through input from university and state agency scientists and stakeholders) on how stock demographics are collected, analyzed, and interpreted to determine the effects of fishing on fished populations (National Marine Fisheries Service, 2001). These assessments form the heart of FMPs, which in turn outline how the councils will achieve and maintain the optimum yield for each fishery.
Whereas data collection by the science centers occurs on an ongoing basis, stock assessments and other types of reports are produced primarily in response to requests from the councils and NOAA Fisheries regional offices. Each science center meets with the appropriate council(s), regional office, and relevant regional state fisheries commissions on a periodic basis to develop an operations plan for stock and economic assessments.
Stock assessments are generally performed by NOAA Fisheries scientists, although on occasion they may be conducted by paid consultants. The completed assessments undergo rigorous peer review (Box 2.2) both internally (within NOAA Fisheries) and, in many cases, externally before they are submitted to the councils for FMP development. This includes reviews from the Center for Independent
Peer review in science is one of the more important processes to which a body of scientific work is exposed. It is the process through which practitioners with technical expertise in a particular field provide objective, constructive criticism on the validity of a body of work to ensure its compliance with scientific methods. The review process uncovers scientific problems of method, interpretation, approach, or failure to provide sufficient detail to reproduce analytical results.
Although peer review is not perfect, it is an essential component in determining what constitutes the “best scientific information available” for use in policy decisions and ensures that managers focus on the science, free of economic, historical, and cultural factors (Meffe et al., 1998). It provides scientific advice on the quality of a body of work and therefore differs substantially in weight and substance from the opinions of those lacking similar scientific expertise.
Peer review of scientific information is applied extensively to the fishery management system, to manuscripts intended for publication in journals, and to the gray literature (stock assessments, dissertations, agency reports, white papers, and other types of scientific documents) that form the bulk of the science supporting management decisions. Gray literature may be subject to both internal and external peer review. The intent is to ensure that any questions about the science can be identified and, if need be, corrected.
In some cases, internal peer review—review conducted by scientists within the institution producing the report—is quite adequate. It reduces costs and allows decisions to be made in a timely manner. Problems may arise, however, when the issue is complex, is controversial, or has far-reaching implications for management. In this case, the internal review process used by NOAA Fisheries may be viewed as biased or insufficiently independent of the source of the report. An “inbred” review results if the relationship between the report authors and the report reviewers is too close to allow an independent assessment of the report. It is considered exclusionary if data sets—particularly those arising outside the agency—are categorically eliminated from consideration. It is therefore important that controversial reports be subject to independent, external peer review to avoid the perception of bias and conflict of interest.
Experts, which consists of a pool of scientists contracted to provide independent peer review for science conducted by NOAA Fisheries (Rosenstiel School of Marine and Atmospheric Science, 2003). Each
regional science center has its own process for determining the “best scientific information available.” The Northeast Science Center has for about the past 20 years used a two-part system consisting of stock assessment development conducted by the Stock Assessment Workshop and external peer review conducted by the Stock Assessment Review Committee.
In 2002, the Southeast Science Center developed a formal process, which is embodied in the Southeast Data Analysis and Review and involves three separate meetings that occur for the following purposes: (1) accumulation and review of the data by agency and academic scientists as well as fishermen and environmental stakeholders; (2) the conduct of stock assessments; and (3) the external peer review. Southeast Data Analysis and Review participants include agency and academic scientists, nongovernmental organizations, and recreational and commercial fishermen, as well as council members.
The Pacific region uses a combination of a stock assessment team and a review panel process called Stock Assessment Review (STAR). The different STAR panels are comprised of knowledgeable scientists who were not involved in the stock assessment and include members from outside the region. STAR panels hold working meetings, open to the public, in which they review draft stock assessment documents and any other pertinent information and then work with a stock assessment team to make necessary revisions to the stock assessment. A STAR panel report is written and used by the council to develop management recommendations. In the North Pacific, stock assessments are done primarily by the Alaska Fisheries Science Center and the Alaska Department of Fish and Game. Plan teams review the stock assessments at two meetings, and the scientific and statistical committee (SSC) performs subsequent reviews. In practice, the council treats SSC recommendations as upper limits for its catch recommendations. Outside reviews of stock assessments are sometimes conducted by either the Alaska Fisheries Science Center or the council when needs dictate.
The councils are the entities that initiate development of an FMP or an FMP amendment using scientific information provided by the centers. The Magnuson-Stevens Act specifies the contents of FMPs as described in 14 required and 12 discretionary provisions (Magnuson-Stevens Act, sec. 303). Relative to National Standard 2, this includes requirements that FMPs provide summaries of the information used to determine “the present and probable future condition of, and the maximum sustainable yield and optimum yield from, the fishery” (Magnuson-Stevens Act, sec. 303). Use of the “best scientific information available” as required by
National Standard 2 is inextricably linked with the uncertainties in biological systems and how that information is used in decision making. The role of science in this process is to quantify the risks involved in taking management actions by explicitly accounting for these uncertainties.
Also, FMPs must specify objective and measurable criteria for determining when stocks are overfished and provide an analysis of how the criteria were developed, including the relationship of the criteria to the reproductive potential of stocks in that fishery. The councils obtain advice (on FMP development, monitoring, and revision) from the states, the fishing industry, consumer and environmental organizations, and other interested parties through council membership and a number of advisory panels (Appendix F). Indeed, each council is required to “establish and maintain, and appoint the members of, a scientific and statistical committee to assist it in the development, collection, and evaluation of such statistical, biological, economic, social, and other scientific information as is relevant to such councils’ development and amendment of any fishery management plan” (Magnuson-Stevens Act, sec. 302).
SSCs and other advisory bodies, which include members from a variety of disciplines, also help the councils establish FMP objectives and criteria for judging FMP effectiveness (Magnuson-Stevens Act, sec. 302) (Figure 2.1). They review stock assessment and fishery evaluation reports and ensure that the “best scientific information available” is being used by the science centers in developing FMPs. It is noted, however, that the councils have different structures for their advisory panels (Appendix F) and SSCs often serve different functions for different councils. Indeed, the Magnuson-Stevens Act does not specify how SSCs should operate. A number of councils underutilize the expertise of the advisory panels to help determine the most important research issues, according to some sources (e.g., Miller, 2002).
Typically, stock assessments appear in annual or semiannual stock assessment and fishery evaluation documents. An FMP requires an environmental assessment and regulatory impact review or an environmental impact statement for NEPA compliance. FMP development may involve stock assessment issues related to overfishing, but it also typically involves a variety of other issues related to access, seasons, and allocation, among others. As the Gulf of Mexico Fishery Management Council notes, “The quality control is inherent in the NOAA Fisheries entities providing the information and in the [stock assessment panels], [the socioeconomic panels], and SSCs using the
information” (Gulf of Mexico Fishery Management Council, 2003). Where discrepancies occur, solutions are provided by the councils primarily with advice from NOAA Fisheries scientists, council advisory panels, and council staff. The scientific information is presented at council meetings, at advisory panel workshops, and at public meetings. The council staff (composed of scientists with expertise in population dynamics, sociology, and economics) then summarizes the assessments, the public input, and council deliberations to develop a draft FMP that contains alternative regulatory strategies, based on advice contained in the stock assessment. After further deliberation and public input, the council chooses among the alternatives, the council staff makes a final draft FMP and transmits it to the Secretary of Commerce for review, and the Secretary publishes a notification in the Federal Register that the plan is available for review and comment for 60 days. After that period, either the Secretary accepts or rejects the plan. If accepted, the final rule is published in the Federal Register and the plan is implemented. If the plan is not approved, the Secretary may either ask the council to change the plan or develop an alternative FMP, or the Secretary may submit an FMP amendment (Secretarial Plan), following the same format for review and input as the councils.
COMMUNICATING SCIENTIFIC RESULTS
Public participation in the management review process depends on the council’s decision-making framework. In most cases, stakeholders may review assessments and help guide the decisions made by council members by submitting written or oral comments at open stock assessment workshops and public meetings publicized by the councils (except as noted in the Magnuson-Stevens Act).
The primary shortcoming of this format is that “open” and “accessible” cannot be construed as equivalent. Many fishermen do not attend the open meetings because they cannot afford the time away from work or the travel expenses associated with attending meetings. Further, even if they can attend, the information is not always accessible because of the form in which it is presented. Thus, the transfer of information is not always successful.
Scientists present their findings at public meetings to help council members, stakeholders, and the general public understand the scientific basis for the alternative management options being considered. Some scientists fail to do this effectively because their presentations are replete with complex terminology, methodologies, and theoretical concepts. Many fishermen lack this expertise, although nongovernmental organizations and fishing organizations often hire representatives who are conversant in the science and can interpret the information for their members.
Some council members may not be conversant in fishery science. Indeed, nearly all of the current 118 council members across the eight regional councils have no background in stock assessment science. The councils have expert scientists at their disposal on advisory panels and on review committees (Appendix F). Providing more training in scientific principles to council members is one means of making the translation of scientific information more effective. In addition, council members would benefit if those scientists who present information to the councils made a concerted effort to develop communication skills that effectively inform audiences with diverse, and often nontechnical, backgrounds.
FLAWS IN PRACTICE
Poor communication skills are not the only flaws in the application of National Standard 2. Data acquired by the science centers in a transparent fashion can ultimately contribute to flawed policy when the
methods of data selection and analysis are not transparent or the limitations of the data are not acknowledged.
Good methodology is the bedrock of science. Without robust methods, the scientific information will be suspect. Further, the problem that looms largest in any regulatory milieu—whether fisheries management or environmental protection—comes at the point decisions are made about what scientific information to include and what scientific information to exclude from use in reaching important policy conclusions (Greer and Steinzor, 2002). The former is a problem for the science centers and the latter is typically a problem for the councils and ultimately NOAA Fisheries, the entities responsible for managing and conserving the fish stocks in federal waters.
The science centers have an incentive to apply “best scientific information available” correctly. They are scientific institutions with staff who apply scientific principles in their work and whose professional integrity is at stake. Councils, on the other hand, have a different mandate to use scientific advice to manage and conserve resources while balancing competing fishery interests (National Academy of Public Administration, 2002; Wilson and Degnbol, 2002). As outlined in this report, the generation of scientific information is itself a very complex problem. The interpretation and application of this information to fisheries management may be hindered by difficulties in communicating scientific conclusions, potentially leading to a disconnect between the two as management actions are developed.
The councils sometimes delay regulatory action concomitant with requests for additional information (Box 2.3), while disregarding peer review of stock assessments and the advice of the their own advisory bodies (Eagle and Thompson, 2003; U.S. Commission on Ocean Policy, 2004). NOAA Fisheries has the authority to approve, disapprove, or partially approve FMPs based on consistency with the conservation and management goals mandated by the Magnuson-Stevens Act. NOAA Fisheries also has the option to develop independent FMPs or amendments in the face of council inaction (Gulf of Mexico Fishery Management Council, 2002). However, both NOAA Fisheries and the councils are subject to political pressure that at times has moved management plans away from the recommendations of the scientific advisory bodies (Hanna, 2002; National Academy of Public Administration, 2002; U.S. Commission on Ocean Policy, 2004). Stakeholders, whether fishermen or environmentalists, have turned to the courts in disputes over the use of or disregard for scientific information.
NOAA Fisheries recognized that the vermilion snapper stock in the Gulf of Mexico was headed for trouble as early as 1991 when the first vermilion snapper stock assessment reported the species to be experiencing an exponential increase in fishing mortality. This forecast was followed by reports of stocks approaching an overfished condition (1998), experiencing overfishing (2000), and finally reaching an overfished condition (2001). NOAA Fisheries compiled a considerable body of scientific information—the “best scientific information available”—that consistently revealed the classic signs of overfishing. These included an overall decrease in landings, in mean size of fish in the commercial catch, in catch per unit effort, in recruitment of age-1 fish, and in the consolidation of the fishery into the most productive fishing areas.
The Gulf of Mexico Fishery Management Council failed to take any action, however, despite the requirements of the Magnuson-Stevens Act to end overfishing immediately once recognized; despite advice from NOAA Fisheries scientists, the Council’s Reef Fish Stock Assessment and SSC advisory panels, and the NOAA General Counsel, and despite an explicit request from the Deputy Director of NOAA Fisheries. Rather, the council stated that it had little confidence in the status reported for vermilion snapper. Further, it declared that no action could be taken until additional data and follow-up analyses were available on age and growth, catch at age, fecundity rates, bycatch estimates, release mortality rates, and the effect of changes in fishing behavior on catch per unit effort—tasks that took several years to complete.
Currently, the council is working on proposed regulations to end vermilion snapper overfishing in the form of Draft Amendment 23 of the Reef Fish Fishery Management Plan, with a decision expected in 2004. However, the council is also asking for reevaluation of catch per unit effort and a new stock assessment. There are currently no regulations that would slow fishing mortality in this fishery.
INTERPRETATION OF BEST AVAILABLE SCIENCE BY THE COURTS
Courts are not immune from tactical forays in the use of “best science.” They are the ultimate arbiters of agency decision making, and their rulings are based largely on what advocates present to them. Perhaps the best-known case of this type is the Supreme Court’s Daubert decision
(Box 2.4). Disappointed stakeholders go to court to challenge agency decisions and must convince the court that some error was made in the decision-making process. This means that each link in the complex chain of administrative judgment is vulnerable to judicial review and can be tested after the fact by determined, informed, and highly interested critics.
The difficulties in defining the respective roles of science and law are revealed at the boundaries of environmental policy (Houck, 2003). “Good science” is often presented as that which supports the advocate’s case. When scientific information is presented as the primary incentive for making a difficult or unpopular policy decision, the science will be attacked. This will bring added costs, inquiry, criticism of the scientists, and disruption of ongoing research and management activities. Scientific uncertainty will be used by those who object to the management action as a means to reject the conclusions of the scientific experts.
Congressional use of the term “best scientific information available” is one of several techniques commonly used to facilitate the preparation and influence of scientific information in the regulatory process along with mandates for scientific studies and the strengthening of scientific advisory apparatus. ESA contains “best science” clauses but also mandates by law that listing decisions be driven chiefly by the biology of the species, not subject to refutation for economic or social reasons (Endangered Species Act, sec. 1533). The “override” mechanism in ESA (commonly known as the “God squad”) is so severely constrained by procedural hurdles that decisions to list are rarely appealed (Endangered Species Act, sec. 1536).
The rules for judicial review of science-based administrative choices are well known, but only in a general and frustratingly indeterminate fashion. Operative here is the so-called hard-look doctrine of judicial review that insists courts require agencies to explain, justify, and defend their decisions with a comfortable wrap of good sense, plausibility, and fair process. In several “best scientific information available” cases, the courts disapproved of the agency’s treatment of science, condemning uses of poor analogy, stale data, end-run procedures, implausible assumptions, unexplained and erratic changes of course, failures to answer forceful objections, and fanciful guesswork (Appendix H).
A recent example of a court taking the agency to task over its failure to ground its actions in the “best scientific information available” is that involving essential fish habitat (American Oceans Campaign v. Daley, 183 F. Supp. 2d 1, 5 [D. D.C. 2000]) (Box 2.5).
In Daubert v. Merrell Dow Pharmaceuticals (509 U.S. 579 [S. U.S. 1993]), the U.S. Supreme Court prescribed the rules for the trial judge to follow in deciding whether expert testimony will be admitted to give guidance to the trier of fact (either the judge or the jury). This process is often described as the “evidentiary gatekeeper” function, and it obliges the trial judge to make a preliminary decision on the reliability of proffered scientific testimony. The Supreme Court explained:
[The inquiry] entails a preliminary assessment of whether the reasoning or methodology is scientifically valid and of whether that reasoning or methodology properly can be applied to the facts in issue… The focus, of course, must be solely on principles and methodology, not on the conclusions that they generate.
Factors to be weighed include whether the theory or technique has been tested; whether it was subjected to peer review; the known or potential error rate; and whether it is generally accepted in the relevant scientific community.
In the course of its opinion, the Supreme Court adopted the so-called Popperian approach to science, asserting that “the criterion of the scientific status of a theory is its falsifiability, or refutability, or testability” (Popper, 1989). Most state supreme courts have followed Daubert in defining exclusionary rules for expert testimony. Surveys of federal judges and attorneys confirm the belief that the Daubert rule has resulted in closer scrutiny of expert testimony and therefore more frequent exclusions of testimony.
Courts also afford agencies ample room to make predictions, order their own affairs, and experiment with process. However, NOAA Fisheries has recently begun to lose more cases under National Standard 2 (Figure 2.2). The losses may reflect poor advocacy, poor records, or simple mistakes. Also, cases may be lost because events move more rapidly than the judicial process or because understanding of the “best scientific information available” has undergone revision.
Still, NOAA Fisheries should not be indifferent to the instructions and lessons of judicial review. Policies and actions that win court approval enjoy stability, credibility, and longevity. Wins are better than losses, for many obvious reasons. Courts afford a continuing scrutiny of and commentary on agency performance on matters of scientific information that are not available from other entities. These judicial
In 1996, Congress passed the Sustainable Fisheries Act amending the Magnuson-Stephens Act. One of its “main thrusts” was the long-term protection of essential fish habitat (EFH).
In August 1997, NOAA Fisheries contracted with the American Fisheries Society to undertake a comprehensive literature survey of scientific reports addressing fishing impacts on habitat. This survey, by Auster and Langton (1999), reviewed 90 studies from around the world “and concluded that 88 of them found some impacts resulting from fishing gear.” Auster and Langton (1999) also concluded “that the overall impact of fishing-related activities in North American waters is unknown despite research efforts spanning 80 years.”
On December 19, 1997, NOAA Fisheries promulgated EFH regulations, to become effective January 20, 1998 (Fishery Conservation and Management Act, sec. 2). It sent the Auster and Langton (1999) study to the regional councils and noted that it “was only a starting point, not a replacement for the EFH assessments for which the Fishery Management Councils were responsible” (American Oceans Campaign v. Daley, 183 F. Supp. 2d 1, 5 [D. D.C. 2000]). The regional councils affected by the Magnuson-Stevens Act submitted draft EFH amendments to NOAA Fisheries for review and comments. In their final EFH amendments, “all Councils identified some EFHs within each of their jurisdictions, yet none adopted measures that restrict fishing gear in order to minimize adverse effects of fishing related activities on EFH” (American Oceans Campaign v. Daley, 183 F. Supp. 2d 1, 5 [D. D.C. 2000]).
NOAA Fisheries partially approved these several amendments and wrote environmental assessments (EAs) for each of them. Each EA concluded that the council amendment would have no significant environmental impact. For the most part, consideration of alternatives was limited to continuing the status quo (which would violate the Magnuson-Stevens Act) and approving the amendment.
The court found that the administrative actions did not violate the Magnuson-Stevens Act, but the court found that each EA failed to comply with NEPA and implementing rules. The court said:
It does not appear that [NOAA Fisheries] took a “hard look” at the problem with respect to any of the EAs. There is no substantive discussion of how fishing practices and gear may damage corals, disrupt fish habitat, and destroy benthic life that helps support healthy fish populations. Instead, a great deal of the discussion revolves around describing the limited number of proposed alternatives and what the agency’s statutory obligations are under NEPA. There is only minimal or
vague discussion of the actual environmental consequences and impacts on the designated EFHs. In several of the EAs, [NOAA Fisheries] simply states that no data is available, and therefore it cannot assess the environmental impact. Several EAs merely note that further action is deferred to future amendments. (American Oceans Campaign v. Daley, 183 F. Supp. 2d 1, 5 [D. D.C. 2000])
The court enjoined enforcement of the amendments “until the Secretary performs a new, thorough, and legally adequate EA or Environmental Impact Statement for each EFH Amendment, in compliance with the requirements of NEPA” (American Oceans Campaign v. Daley, 183 F. Supp. 2d 1, 5 [D. D.C. 2000]). The court was strongly critical of the agency’s approval of five council decisions that had taken no action to address gear-related habitat damage.
That outcome subverts the very purpose of NEPA, which is to ensure that agencies are fully aware of any adverse environmental effects of their actions, and of all feasible alternatives which may have lesser adverse effects on the environment, so that final decision-making will be informed by a full understanding of relevant environmental impacts. (American Oceans Campaign v. Daley, 183 F. Supp. 2d 1, 5 [D. D.C. 2000])
performance audits are usually easily accessed, thoroughly contested, and empirically rich. Judicial decisions discussing “best science” issues should be made readily available in summary or abbreviated form to all agency personnel. Court cases examining the reach and meaning of “best scientific information available” provide NOAA Fisheries with a hard-look doctrine of the courts. Procedural consistency would provide the agency with a stronger basis for defending decisions in court. More specifically, guidelines that address issues of relevance, objectivity, transparency, timeliness, peer review, and the treatment of uncertainty are consistent with the procedural cues that have been sought in the court cases documented in Appendix H. Courts have reversed and remanded agency decisions contrary to “best scientific information available” concepts that are intuitive, ad hoc, and derived from values articulated in individual judicial decisions. However, the “common law” of judicial review of “best scientific information available” is insufficiently mature, elaborate, and credible for day-to-day use within NOAA Fisheries.