Conventional Management of Marine Fisheries
~ enlarge ~
PROBLEMS AND ISSUES IN FISHERY MANAGEMENT
Current interest in marine protected areas (MPAs) reflects dissatisfaction with conventional approaches to the conservation of marine ecosystems, especially fishery management, which often have failed to meet societal goals for sustainable use of marine resources and protection of biological diversity and productivity. Overfishing represents one of the most challenging problems in marine conservation. On a global basis, 44% of the world's fish stocks are now fully exploited, and 25% are overexploited and clearly in need of urgent conservation and management measures (Garcia and Newton, 1997; FAO, 1999). Collapses or dramatic declines of marine fish stocks, for example the Atlantic cod (Gadus morhua) off Newfoundland and on Georges Bank, call attention to the potential for failure resulting from the limitations of fishery science and of the current management system (Walters and Maguire, 1996; Fogarty and Murawski, 1998). The view of fishery scientists is that it was partly a failure of science that caused the collapse of the Newfoundland cod. Scientific advice to managers was not always correct or timely, and management failed to act in time on either the erroneous advice or the corrected advice. In the United States, the overexploitation of bluefin tuna and of mixed-species groundfish stocks convincingly illustrates the consequences of overcapitalization and excess effort.
Problems associated with conventional fishery management that threaten the sustainability of marine fisheries have been highlighted in recent reviews (e.g., NRC, 1999a). Overfishing and attendant fishing mortality rates that are too high and poorly regulated lead the list. Overcapacity, the presence of too many partici-
pants or units of effort in fisheries, is a related and serious problem. Open access to fisheries tends to favor overexploitation rather than stewardship, a common impediment to effective management (NRC, 1999b), although very few commercially important fisheries in North America and Europe are completely open access in the sense that most are regulated, some have limited entry, and many have restrictions on effort.1 Another problem is the failure of management to act expeditiously and conservatively or to respond appropriately. Ecosystem-based approaches to management (e.g., NMFS, 1999) have emerged from concern about bycatch, habitat destruction, and the failure to consider important biological interactions (e.g., predator-prey). Shortfalls in the ability of scientists to produce accurate stock assessments have at times provided poor advice to managers (NRC, 1998a, b). Stock assessments and resultant management measures always contain a level of uncertainty. To be effective in the face of this uncertainty requires that the assessments be interpreted conservatively so that stock size is not overestimated and subsequently overfished. Conservative, flexible, and adaptive approaches can compensate for the uncertainty of stock assessments, but frequently these features are lacking from conventional management.
Overfishing is in large part a consequence of excessive effort and capacity in fisheries (NRC, 1999a). Too often, fishery managers have been unable to control fishing effort, resulting in unsustainable levels of catch. This has been a particular problem for open-access fisheries where management does not limit the number of participants or high individual effort (see Chapter 4). In this situation, the economic incentives favor short-term exploitation over long-term sustainable use because the economic benefits of sacrificing current catch to rebuild the stock are intangible compared to short-term needs (bills to be paid), and long-term benefits may have to be shared with newcomers when the fishery recovers (Hilborn et al., in press). As more people enter the fishery or improve their fishing capabilities, the future yield to the individual fisher decreases. This often fosters competition to maintain or even increase individual catch levels even as stocks decline. In response, managers may shorten fishing seasons; participants then increase their fishing power, and effort becomes concentrated in time, sometimes resulting in “races for fish” or “fishing derbies.” In the worst cases, derbies are absurdly brief, lasting only two days in the Pacific halibut (Hippoglossus stenolepis) fishery in the United States during the early 1990s before individual quotas were implemented (NRC, 1999b).
In addition to depletion of fishery stocks, there are unintended consequences of fishing, such as bycatch and degradation of habitat from destructive fishing practices (Dayton et al., 1995; Watling and Norse, 1998). Bycatch here refers to
1 Open access is defined as the condition in which access to a fishery is in effect unrestricted (i.e., no license limitation, quotas, or other measures that would limit the amount of fish an individual fisher can land) (NRC, 1999b).
the incidental catch by fishing gear of adult and juvenile fish that are not the target of the fishery (Alverson et al., 1994). Bycatch and habitat loss not only may have deleterious effects on fishery yields, but also may degrade the ability of marine ecosystems to support biological diversity. Therefore, effective regulation of fishing activity in the oceans is not just a fishery management issue. For example, unique features and habitat such as coral reefs need prohibitions on fishing, as well as protection from shipping, diving, recreational boating, and destructive coastal development. Ecosystem approaches, including marine reserves, will have to be added to the conventional management toolbox to conserve biodiversity, maintain biocomplexity, and ensure that ecosystem services are maintained for posterity. The public's interest in ecosystem approaches in part represents the existence values that the public places on preserving the diverse biota and habitats of the sea (see Chapter 4). To ensure the future of living resources and habitats in many stressed marine ecosystems, some areas of the ocean could be zoned in MPAs for limited access and use. This chapter describes conventional fishery management tools, noting both limitations and failures, to provide a context for evaluating MPAs and reserves as complementary or alternative tools.
CONVENTIONAL FISHERY MANAGEMENT
In general, conventional fishery management seeks to maintain high, yet sustainable, yields by regulating the number or weight of fish caught, the size of fish caught, or the time and space (area) within which fishing is allowed. The intent in each case is to control fishing mortality rates. Conventional approaches to fishery management in the United States can be succinctly characterized by three main components: (1) an underlying fishery science and management paradigm, (2) a set of conventional management tools, and (3) the fishery management system.
Fishery management relies on estimates of the population size of a target species to determine how many fish or what fraction of the population's biomass can be caught without damaging its reproductive potential. To make these determinations, management depends on a conceptual model of a fishery that makes three simplifying assumptions: (1) the fishing fleet targets and exploits a single-species stock, (2) the stock of interest is segregated temporally or spatially from other stocks, and (3) the individuals are perfectly mixed so that the effects of fishing are well spread over the whole stock. These assumptions, which are far from true in most situations, can have serious consequences for the effectiveness of fishery management.
Single-Species, One Stock
Most management measures are directed at individual stocks of a single species and do not take into account species interactions, such as predator-prey relationships. A basic assumption of most models used to determine a catch level is that the catch rate a stock can sustain can be designated based upon the average productivity of the stock. Productivity, in turn, is presumed to depend primarily on the size of the adult stock. In this scenario, controlling adult stock size is the primary means of ensuring sustainability of the fishery. Furthermore, stocks are assumed to respond in a density-dependent manner and therefore are postulated to have maximum productivity at intermediate stock sizes. Thus, maintaining the stock size that allows maximum sustainable yield (MSY; see Appendix B for definition) historically has been a major management goal, and fishing at a rate that produces MSY on average over good and bad years has been the target. In fact, fishing at the MSY level (a fixed exploitation rate policy) does not ensure constant catches in the future or a stable adult population size because of substantial variability in reproductive success and recruitment. It was recognized more than two decades ago (see Larkin's, 1977, famous epitaph on the concept of MSY) that it is therefore too risky to set a constant quota for catch at MSY. In good years, fishers may prosper with MSY-based catches, but in years when the environment is less favorable and recruitment and productivity decline, the stock will diminish and MSY may quickly lead to overfishing. The fishing rate corresponding to MSY (F MSY) still remains a criterion in determining optimal yield, the regulatory target used to manage marine fish stocks in the United States (NOAA, 1996a). However, FMSY now is viewed by many as a threshold that should not be exceeded, rather than as a target at which to aim. More conservative quotas and exploitation rates are now recommended, due in part to recognition of our limited ability to estimate and implement F MSY or, for that matter, other target fishing mortalities via catch or effort control. As Hilborn and Walters (1992) noted, obtaining an estimate of MSY (or F MSY) usually requires fishing at levels that already exceed it.
Fish Stocks That Are Temporally or Spatially Segregated
Although it is obvious that management must be tailored to individual species' life histories, the individual stock and single-species approaches to management are ineffective for multispecies or even mixed-stock associations, in which many different species or stocks with similar habitat and prey requirements overlap in their ranges. Good examples are reef fish off the southeastern United States and rockfish in the northern Pacific Ocean. Warm-temperate species such as gag (Mycteroperca microlepis), scamp (Mycteroperca phenax), and red hind grouper (Epinephelus guttatus), for instance, co-occur to such an extent
that catch restrictions placed on one species in the complex typically result in increased regulatory discards while fishing for associated species. The same is true for the 83 species of rockfish managed as a complex off the Pacific west coast. Inevitably, regulatory discards will increase the mortality of the restricted species and threaten its recovery.
Individuals Are Perfectly Mixed
A key element of the fishery management paradigm is the concept of a well-mixed stock. Migration patterns and more general spatial processes are fundamental components of fishery science. Knowledge of spatial processes serves to delineate management units, each viewed as a “dynamic pool” isolated from the rest. Most theory and management have been conceived for large-scale, commercial fisheries that target relatively mobile species (e.g., tuna, plaice, gadoids), for which dynamic pool assumptions may provide a reasonable simplification, at least at the scale of a fishing ground. In this conceptual model, because the effects of fishing are “diluted” in the pool, the use of spatially explicit approaches to manage each unit has been largely missing (Shea et al., 1998). This conceptual model has dominated marine fishery management. As a result, it has been applied indiscriminately to low-mobility species for which the paradigm is clearly inappropriate, such as rockfish (Sebastes spp.) and sedentary invertebrates. Fishing effort on relatively sedentary species preferentially targets the highest fish concentration and results in a mosaic of fishing mortalities and serial depletion of fishing grounds. Management by conventional means is complicated because the dynamics of the resource, fishing process, and monitoring are dominated by local processes that are often impossible or impractical to assess. In reef fisheries, for example, the relevant scale for assessment and management may correspond to a single reef.
Another tool to limit fishing effort is control of access to the fishing grounds. This strategy, known as “spatial management,” can be applied as either temporary (seasonal or year-to-year) or permanent closure of portions of the fishing grounds. Spatial management can be used to control access to resources and probably has been practiced for centuries (Cushing, 1988), but it has not played a central role in the management of marine fisheries. Its importance in the management of benthic shellfish is now accepted (Orensanz and Jamieson, 1998; Perry et al., 1999), as are novel management schemes involving area rotation (Bradbury 1990, 1991; Perry et al., 1999). However, many more species could benefit from spatially explicit management, particularly those with a relatively stationary adult stage. A major constraint on implementation of spatial management is the lack of spatial catch data for many species and the complexity of spatially explicit stock assessment models.
Although spatially explicit components usually are missing, some existing fishery models do recognize that variables other than adult stock size affect
productivity. However, these variables usually are treated as random variations beyond human control. Consequently, management focuses on regulating the size of the catch and the effort directed at obtaining it, while environmental factors are downplayed. Even when not controllable, persistent trends or variations in productivity driven by environmental conditions should be considered in policy evaluation. In addition, environmental variables that affect habitat quality and are directly affected by human activities are now addressed explicitly in fishery management plans, following the reauthorization of the Magnuson-Stevens Fishery Conservation and Management Act (MSFCMA, NOAA, 1996a) and its emphasis on essential fish habitat (EFH). This focus on habitat has initiated a shift toward spatial considerations and designation of areas important for the productivity of economically important species. These designations, of essential fish habitat and habitat areas of particular concern, have bolstered interest in protected areas and supported the use of MPAs as a legitimate tool for fishery management.
Long-term, large-scale closures have been instituted as single-species refuges from fishing to promote rebuilding of depleted stocks. In many respects, they resemble permanent reserves, except that such areas may revert to their former status when restoration is attained. Long-term closures are becoming more common as a population-rebuilding tool (e.g., northern cod stock off Newfoundland), in which depleted populations may require many years to restore. Long-term closures might achieve some goals of reserves, although benefits may be transient if subsequent fishing mortality cannot be controlled.
Another form of spatial management is the use of rotating fishing areas. Here only a fraction of the fishing grounds is opened in a given season, the rest being closed for specified periods (often years) to promote growth of young animals and allow them to reach more valuable sizes. The area opened is rotated from year to year. This approach combines temporal and spatial closure to regulate fishing. In addition to rebuilding fish stocks, rotating closures may allow habitat and biological communities to recover from the effects of fishing, such as damage to bottom habitats by trawling and movement of “fixed” gear. However, recovery of habitat and biological communities may require closures on the order of 5 to 10 years (Collie, 1997).
Conventional Management Tools
Management based on the fishery paradigm above centers on measures that regulate fishing activities and the level of catch, rather than on measures that directly promote management of habitat or consideration of environmental variables affecting fish productivity. Generally, the goal is to manage exploited populations such that they are maintained at productive levels (close to or above MSY) that support a high yet sustained fishing yield and to require rebuilding plans when yields fall below a minimum stock size threshold.
Conventional management approaches to control exploitation rates fundamentally rely on placing limits on the amount or efficiency of fishing effort (effort-controlled fisheries) or on the total amount caught by specifying catch quotas and allocations (quota-based fisheries). In both cases, a target fishing rate is first specified, whether constant or variable in response to stock condition, based on analysis of historical experience with the fishery, on experience with fisheries for similar species, or on modeled responses of the fishery to simulated fishing mortality.
There are many forms of effort controls, including restrictions on gear, vessels, time fished, and number of fishers. These are usually the first controls applied to a fishery to slow the rate of catch. Gear restrictions can include the type, amount, or dimensions of gear or specific features of the gear such as net mesh size, hook spacing on longlines, or configuration of fish traps. Vessel restrictions may include design, length, or engine horsepower. The number of fishers can be regulated by allocating licenses to either fishers or vessels. Time fished can be regulated through limiting the amount of time available for fishing through seasonal closures, “days-at-sea” restrictions, or specific days and hours when fishing is permitted.
To implement a target fishing rate by means of effort regulations, managers need a reliable estimate of the fishing mortality caused by each unit of fishing effort to be allowed, a parameter known as catchability. The estimate of catchability is then used to determine the amount of fishing effort (e.g., the number of total days at sea to be allowed) that is compatible with a chosen exploitation rate or fishing mortality so that
Ftarget = catchability × effort.
Catchability is generally assumed to remain constant as stock size varies. However, catchability may change—for instance, fishers may increase their efficiency when the stock declines—and this could result in overexploitation, a common reason for the failure of conventional management.
In addition to measures that attempt to regulate exploitation rate directly, other forms of effort control may be implemented to reduce fishing power, protect vulnerable life-history stages, or increase the market value of the fish. Temporal closures, for example, are frequently used to protect fishery resources at times when they are particularly vulnerable to fishing, such as when fish aggregate on spawning grounds. Also, temporal closures can be used to allocate fishing over the season in a manner that increases the value of landings. For example, fishing might be prohibited during parts of the year when the population is composed of small or poorly conditioned fish. Closure of the Dungeness
crab (Cancer magister) fishery during the molting season when meat quality is poor provides one example (Methot, 1986). Broader application of temporal closures to protect whole communities or complexes of species and habitats is less common, but perhaps of greater relevance to developing marine reserves as an ecosystem approach to fishery management.
A common approach to controlling fishing is to regulate the catch or the amount of fish landed. This is the favored method used to regulate fisheries in Alaska and along the west coast of the United States. Quota-based management relies on the ability to model relative trends in abundance over time and to estimate the absolute size of the exploitable stock. This information is used to set the total allowable catch (TAC) that meets a chosen target exploitation rate.
TACs or quotas are typically calculated as the product of a target exploitation rate µ and an estimate of current stock biomass BÌt:
TACt = µtBÌt.
In principle, TAC-based management can be a direct, efficient way to limit catches. However, the success of TAC-based systems depends on accurate estimates of stock abundance and biomass. Because the required level of accuracy is usually not available, the risk of overfishing may be high (Walters and Pearse, 1996; Walters, 1998).
In many heavily exploited fisheries, both effort and catch controls are used to manage the fishery. In the Pacific halibut fishery, for example, a catch quota is used to control the exploitation rate, there is a minimum size limit on landed fish, and all fishing methods except setline gear are prohibited for the directed fishery (http://www.iphc.edu). Seasonal closures are in place, which prevent the interception of fish from different regulatory areas when the fish migrate from feeding to spawning.
Both forms of regulation, catch and effort controls, have significant shortcomings because both depend on the quality of stock assessments. In turn, the accuracy and reliability of stock assessments depends on data that are frequently limited or unavailable. Conventional methods used to estimate stock abundance and current rates of fishing mortality require good historical catch statistics under significant levels of exploitation and indices of stock abundance that reliably show population trends. Numerous methods are used for stock assessment (reviewed in NRC, 1998a); most are based on analyzing the way abundance changes in response to known catch levels. In addition to estimates of stock abundance or fishing mortality, conventional management depends on knowledge of what exploitation rates are adequate to derive “biological reference points” that designate both threshold and target levels of biomass and fishing rates. The
sophistication of procedures used can lead to overconfidence in their ability to estimate abundances of stocks and their resilience to fishing pressure. This misplaced confidence contributed to the collapse of the Newfoundland cod fishery (Walters and Maguire, 1996). Although stock assessments usually are done competently by fishery scientists in the United States, the statistical uncertainty associated with estimates and biological reference points can lead to failed management (NRC, 1998a).
Fishery Management Systems
Management of fisheries in the United States typically is undertaken at geographic scales that range from local to national. Assignment of responsibilities and implementation of effective management is complex. Jurisdictions of responsible institutions and agencies may overlap in some fisheries. The eight regional fishery management councils (NOAA, 1996a) have primary responsibility for management in the U.S. exclusive economic zone (EEZ), but they may share responsibility with other regional management institutions for coastal migratory species, especially those that occur in the nearshore and estuarine regions of the coast. For example, the Mid-Atlantic Regional Fishery Management Council shares responsibility for managing coastal species such as bluefish (Pomatomus saltatrix), weakfish (Cynoscion regalis), and summer flounder (Paralichthys dentatus) with the Atlantic States Marine Fisheries Commission, which represents state interests in migratory species that are fished in the coastal zones and estuaries of Atlantic Coast states. Management systems are even more complex for such species, cause state agencies also are engaged in the regulatory process within their jurisdictions. Furthermore, management of these species is conducted at additional regional levels (e.g., the Chesapeake Bay, in which the States of Virginia, Maryland, and Pennsylvania; the Potomac River Fisheries Commission; and the District of Columbia exercise jurisdictional control).
Management systems typically institute a variety of output and input controls to regulate fisheries over their geographic ranges. Quota allocations, often among sectors of the fishery (e.g., commercial and recreational), are common; minimum sizes or other size regulations may apply. Restrictions on gears, seasons, seasonally closed areas, and combinations of methods, often with specific geographic regulations within the range of the targeted species, are the tools that managers commonly apply. Not only is it difficult to attain consensus to manage resources, but the success of management measures is often uncertain.
The uncertainties in the success of management systems lie in the attendant uncertainties that usually characterize the science and management of fishery resources. The science of stock assessment itself may be uncertain for many fished stocks. Political pressures on managers and institutions can dictate management policies and responses, sometimes to the disadvantage of long-term benefit to fisheries. Disputes among sectors of fisheries—for example, different
gear users, or recreational versus commercial fishers—can dominate the dialogue and sometimes result in compromises that do not constitute best management policy. Faced with uncertainty in science and social conflict, managers historically have been slow to act to conserve fishery resources. Legislation, such as the national standards of the MSFCMA (NOAA, 1996a), presents, at least to some, conflicting goals of conservation, economics, and social interests that delay or misdirect management actions. Finally, to be effective, management systems must encourage compliance, either through enforcement or by providing proper institutional incentives to comply with regulations. As discussed in Chapter 4, management systems that confer user rights and participation of stakeholders in the management process can improve compliance.
As noted above, management in most of the U.S. EEZ is regulated by regional management councils (NOAA, 1996a). Currently, there are 37 fishery management plans submitted by the regional councils and approved by the Secretary of Commerce. The eight councils have jurisdiction over broad, discrete geographic areas, although sometimes their management authority is shared for migratory species. Jurisdictional issues may be significant for migratory stocks, especially coastal stocks that cross the boundary between state and federal waters at 3 miles from shore (for most states). It is important to note that if fishery reserves become an important and integral part of management plans, state and coastal regional authorities (e.g., the Atlantic States Marine Fisheries Commission and its Gulf of Mexico and Pacific Coast counterparts) will have shared, possibly complex, jurisdictional authority for spatial management and enforcement. At the time of this report, some of the regional management councils are considering and developing strategies for fishery reserves and other spatially restricted fishery management plans.
In some fisheries, managers have adopted methods that control access by establishing individual fishing quotas (IFQs) (NRC, 1999b), which assign shares of the fishery-wide TAC to selected individuals or sectors of a fishery. The privileged access that is afforded by IFQ management has been criticized by some, but it represents a step by conventional managers to match capacity and effort with available fish. Assigning rights or privileges to access is not, of course, sufficient to manage fisheries unless additional conventional tools of fishery management are also applied, such as quotas and gear restrictions, and special attention is given to controlling bycatch and discards, which can be problematic in IFQ fisheries.
New paradigms are emerging to guide management of marine fisheries in the new millennium. Although many of these paradigms build on conventional management practice, they have significantly changed the philosophy of management agencies in the past decade. The precautionary approach and the riskaverse policies that it implies have been advocated globally (FAO, 1995) and in the United States (NOAA, 1996a, 1999; NRC, 1999a; Restrepo and Powers, 1999). The burden of proof is being shifted away from demonstrating a negative
effect of fishing before curtailing effort, to demonstrating that fishing practices will not damage the stock, habitat, or other ecosystem properties before allowing fishing to increase (Dayton et al., 1998). Although progress is slow, management is moving toward multispecies approaches, and ecosystem approaches eventually may be widely applied in managing marine fisheries (NMFS, 1999). Finally, the concept of embedding fishery management in the broader context of coastal zone management is being debated. It is here that MPAs can make an important contribution to accomplishing integrated management of our nation's coastal resources.
Long-term, single-species area closures represent a move toward MPA-style management. Although they have some features in common with reserves, single-species closures lack many key conservation benefits of permanent reserves and their objectives are generally narrowly drawn. Few temporal closures are designed to address multispecies or ecosystem concerns; rather temporal closures are a tool for single-species fishery management. An exception is the closure of areas 10-20 miles offshore of haulouts and rookeries occupied by endangered Steller sea lions (Eumetopias jubatus) to fishing for walleye pollock (Theragra chalcogramma). Other time and area restrictions have been implemented for the pollock fishery within and outside critical habitat for Steller sea lions.
UNCERTAINTY, FISHERY MANAGEMENT, AND A ROLE FOR MARINE RESERVES
Many scientists believe that a primary cause of fishery management failures is the inherent uncertainty in stock assessments. This uncertainty contributes to ineffective or untimely management actions and the reluctance of fishers to accept the economic costs of reducing effort even when stocks are in decline or their status is uncertain (Ludwig et al., 1993). To provide insurance against stock collapse, scientists have proposed establishing fishery reserves when the lack of accuracy in stock assessments and lack of resolve to fish conservatively make it difficult to achieve sustainable fishing levels under conventional management. The specific causes leading to the collapse of a fishery are controversial because it is difficult to discern the relative contributions of fishing pressure and environmental forces. Also, management generally does not account for the effect of environmental degradation on MSY (e.g., Myers et al., 1996, 1997; Orensanz and Jamieson, 1998; Caddy, 2000). Fishing fleets are ever more efficient at locating and catching remaining fish aggregations, with the result that once the fishery collapses, it may require long periods of time to recover, on the order of a decade or more, even in the absence of fishing (Hutchings, 2000). Ensuring against collapse is a primary but elusive goal of marine fishery management.
Central to the problem of uncertainty in fishery science and management is our difficulty in confronting it. Conventional fishery management relies on
science, particularly our ability to determine appropriate target catches and to estimate actual fishing mortality or stock size as a basis for recommending effort or catch controls to meet these targets. Even when science is adequate, the effectiveness of management in achieving the desired control (i.e., control the exploitation rate) may be uncertain (Walters and Parma, 1996; Walters and Pearse, 1996; Walters, 1998). Experience and simulation analyses have shown that stock assessment methods sometimes are prone to errors exceeding 50%, even when costly monitoring programs are in place (NRC, 1998a). Worse, errors tend to be correlated from year to year, compounding their effects over time. Retrospective analysis often reveal biases, with stock size initially overestimated or underestimated for several consecutive years (Sinclaire et al., 1991; Parma, 1993). When scientists and managers depend on catch data from the fishery itself (i.e., fishery-dependent data), levels of bycatch and discards at sea often are unknown, and these sources of fishing mortality may not be included properly in assessments. Fundamental parameters, such as the rate of natural mortality, can be specified only in a rather broad range, based on life-history correlates. Indices of abundance derived from research surveys are valuable, but they too can be imprecise or, in many fisheries, simply unavailable.
It has been argued (Walters and Pearse, 1996; Lauck et al., 1998; Walters, 1998) that uncertainty in stock assessments is simply too large to manage fisheries sustainably using conventional tools. Three main approaches have been proposed to address this uncertainty: (1) choose substantially lower catch rates as fishing targets than in the past (Mace, 1994; Restrepo and Powers, 1999); (2) implement management tools that are less dependent on stock assessments, such as reserves (Roberts, 1997a; Lauck et al., 1998; Walters, 1998; Murray et al., 1999) and size limits (Myers and Mertz, 1998), and (3) generate institutional incentives that encourage responsible behavior on the part the fishers, such as different forms of user rights (NRC, 1999b; Hilborn et al., in press). These three approaches are not exclusive, and all may have to be considered for fishery management to be successful. Marine reserves, as an alternative to conventional management, also have uncertainties associated with their performance. Sources of costs and benefits of some of these approaches are discussed in more detail in Chapter 4.