Options for Achieving Sustainability
Previous chapters have described the status of marine fisheries and ecosystems and identified some of the factors that have led to the current situation. This chapter discusses options for improving the prospects for sustaining marine fisheries. We begin with management and socioeconomic incentives because the committee believes that changes in those areas can have the largest and most immediate positive effects. We conclude with scientific considerations, many of which involve research. It is, of course, impossible to neatly categorize the following discussions as focusing on management, scientific, or socioeconomic matters. Many of the approaches discussed below include elements of all of them.
Previous chapters have suggested that management difficulties include a lack of scientific information; a lack of full appreciation and use of available scientific information; a risk-prone approach; a lack of appreciation for ecosystem and other nonfishery values; the need to balance many goals and values, some of which conflict and many of which are not clearly articulated; and space and time scales of management that do not coincide with the distribution of the target species, their ecosystems, or fishing communities. The committee concludes that to approach the goal of sustainability managers should adopt a conservative, risk-averse approach that recognizes ecosystem values. After describing the U.S. management context, the focus below is on approaches most likely to be helpful in achieving that goal.
U.S. Management Context
Marine-fishery management in the United States takes place in several interrelated ways. They include management by states for stocks found largely within the 3-nautical-mile territorial sea; by the federal government through regional councils for stocks found largely in the exclusive economic zone (EEZ) 3 to 200 nautical miles from shore; and through international bodies for certain shared or highly migratory species such as Pacific halibut, Pacific salmon, tunas, and whales. In addition, there are some highly localized systems such as town management of oysters and clams in parts of New England as well as regional systems such as the interstate marine fisheries commissions (e.g., the Atlantic States Marine Fisheries Commission). The EEZ management to some extent sets the framework for other management regimes. It was established under the authority of the Magnuson-Stevens Fisheries Conservation and Management Act (MSFCMA) of 1976, reauthorized and amended most recently in the Sustainable Fisheries Act of 1996. The primary purposes of the act are to:
- Establish a geographic zone adjacent to the United States over which the U.S. government is responsible for fishery resource management, with limited exceptions.
- Promote conservation and achieve optimum yields from the nation's fishery resources. Social and economic factors are to be given equal importance for modifying optimum yield.
- Create a legal and economic environment that stimulates harvest of fisheries resources within the area of extended jurisdiction, and subsequent processing of such catches by U.S. fishermen and companies.
- Establish an institutional structure and enforcement authority that allows the United States to carry out the objectives explicit and implicit within the Act.
- Ensure that conservation and management under the act are based on the best scientific information available (P.L. 94-265).
The organizational structure set up by the MSFCMA is based on eight regional fishery management councils, with representation from relevant state and federal agencies as well as the public. Public members have backgrounds ranging from commercial or recreational fishing to research and fishery conservation. The councils prepare fishery-management plans (FMPs) or plan amendments. They are implemented by the National Marine Fisheries Service (NMFS), if approved by the Secretary of Commerce. The NMFS also provides most of the data and stock assessments used in management.
Congress has amended the MSFCMA to revise National Standard 1 (Box 5-1) to require greater conservation; no longer can any relevant economic or social factor be used to justify fishing at levels above the maximum sustainable yield. The amendments also call for a reduction in bycatch and overcapacity and for more attention to habitat protection, specifically requiring the designation of essential fish
SEC. 106. NATIONAL STANDARDS FOR FISHERY CONSERVATION AND MANAGEMENT
habitat and consideration of actions to conserve such habitat (section 110). Congress also required NMFS to convene a panel to consider ecosystem-based approaches to U.S. fishery management (NMFS in press).
Conservative Single-Species Management
The most obvious management approach is to reduce the catch of depleted species on a single-species basis. If Georges Bank is a prime example of the effects of overfishing in the United States, the case of striped bass on the U.S. east coast is a shining example of the effects of catch controls on a single species. Striped bass populations collapsed throughout the mid-Atlantic region and elsewhere in the late 1970s (Richkus et al. 1992). Chesapeake Bay is thought to be the nursery ground for 60 to 80 percent of striped bass off the east coast of the United States. In 1984, Congress passed the Striped Bass Conservation Act, giving states authority to place moratoria on fishing for striped bass. In 1985, Goodyear published calculations showing that control of fishing for bass would lead to a rebuilding of the populations, even if the decline had causes other than overfishing (Goodyear 1985). Led by Maryland, which imposed a moratorium on striped-bass fishing in Chesapeake Bay in 1985, and Virginia in 1988, the east-coast states increasingly controlled fishing effort. In early 1995, striped bass were declared by the Atlantic States Marine Fisheries Commission to be fully recovered (NMFS 1996a).
Other species have also responded to controls applied on a single-species basis. Indeed, Myers et al. (1995) concluded, based on an examination of life-histories, that almost all overexploited fish populations would recover if fishing were stopped. For example, both king (Scomberomorus cavalla) and Spanish (S. maculatus) mackerel catches off the southeastern and Gulf coasts of the United States have been severely restricted since the mid-1980s. Spanish mackerel were removed from overfished status to fully exploited status in 1995, and their populations have shown considerable increases (NMFS 1996a). There is some optimism that king mackerel populations will increase as well. Pacific halibut have long been managed on a single-species basis and have supported a sustainable fishery since the 1920s.
Several specific methods of implementing conservative management have been described. Marine protected areas are discussed in detail below. Another approach is to adopt a fixed exploitation rate (as opposed to a fixed catch) (NRC 1996b, Walters and Parma 1996). Another is to allow fish to spawn at least once before they are fished (Myers and Mertz 1998). Myers and Mertz pointed out that this approach was recommended more than 100 years ago (Holt 1895), but that other approaches, such as maximizing yield from somatic growth, had reduced its influence on management. They also provided practical guidance, emphasizing that susceptibility of populations to overfishing is very sensitive to the age at which they are first caught. Populations that can be caught while young
but become sexually mature when much older (e.g., bluefin tuna, some cod populations) are particularly vulnerable to overfishing.
This approach is not always as successful, as described above. For example, Pacific ocean perch, severely depleted by fishing in the 1960s, supported almost no directed fisheries in the 1970s and 1980s. Their stocks were considered to be rebuilt only in the mid-1990s (NMFS 1996b, North Pacific Fishery Management Council 1997). This is not a complete surprise, as Pacific ocean perch are very long lived, but, even so, 30 years is a long time to wait for positive results. Pacific sardine populations declined drastically off the U.S. west coast in the early 1950s and were unmeasurably low by the 1970s despite essentially zero landings from about 1960. Only after the late 1980s did their populations begin to recover, and they are still low (NMFS 1996a). However, as described in Chapter 3, many small pelagic marine species like sardines and anchovies are subject to large, environmentally influenced fluctuations, so cause-effect relationships are not clear in this case. In general, a large reduction of fishing effort is a biologically effective method of conserving or rebuilding many marine fish populations, however disruptive it might be socioeconomically.
Although it is often effective, a conservative single-species approach alone is probably insufficient to sustain fisheries or ecosystems at acceptable levels of productivity. One reason is that it, like many other approaches, is difficult to implement and enforce. A considerable amount of political energy was needed to implement moratoria on striped-bass fishing; the International Pacific Halibut Commission, which manages Pacific halibut, was established by international treaty. More important, however, is that continued adverse ecosystem effects can accrue even when the target species is not depleted (see Chapter 3).
Although single-species management can be effective for maintaining population levels of individual species (e.g., Bering Sea groundfish and striped bass in the Chesapeake Bay), other organisms in the ecosystems may be affected through bycatch and trophic interactions. For example, current fisheries in the Bering Sea apparently are stable under single-species management, although earlier fisheries coupled with changes in atmospheric and oceanic circulation patterns probably contributed to declines in marine mammals and birds (NRC 1996a). Nonetheless, universal application of conservative management on a single-species basis would go a long way toward reducing overexploitation of the world's marine fisheries.
Reducing Bycatch and Discards
Reducing bycatch and discards is clearly a high priority for management and has been made a specific goal in recent national policies and international agreements. The matter has been addressed recently by the U.S. Congress in the revised Magnuson-Stevens Fishery Conservation and Management Act (see pp 78–80). The National Marine Fisheries Service has drafted a national bycatch plan (NMFS 1998). These and other efforts appear to have produced results.
Data for the years 1994 and 1995 suggest that bycatch and discard rates have declined since the mid-1980s as a result of several factors (FAO 1997d, Natural Resources Consultants 1998), including a decline in fishing effort for some important species, time and area closures, adoption of more selective fishing technologies, enforcement of prohibitions of discarding by some countries, and more progressive attitudes among fishery managers, users, and society at large with respect to problems resulting from discards. In addition, discards (but not bycatch) have been reduced by new technologies for using a variety of marine species and a greater use of many species for human consumption and for feed for aquaculture and livestock. All these efforts have reduced discards by several million metric tons since 1990 (FAO 1997d, National Resources Consultants 1998), and they have reduced bycatch as well.
Perhaps the most important overall approach is to stop treating bycatch as if it were a side effect of directed fishing. Instead, as proposed by Davis (1996), for example, the existence of bycatch should be recognized and dealth with in fishery-management plans as part of an overall exploitation of the marine community. Thus, catch quotas would be established for various gear types that reflect the mix of species those gears typically catch. Total fish removals would be accounted for if the catch quotas were based on the assessment of the species mix as a whole. Obviously, the size of the catch quota should be based as much as possible on information on interactions among the species involved (i.e., an ecosystem consideration). Under Davis's proposal, there would be no target catch or bycatch for each species; instead, there would be a total catch for groups of species. In Alaska, bycatch of halibut and groundfish is considered in setting and monitoring annual quotas, and the fisheries are closed when annual catch or bycatch quotas for individual species are reached (Pennoyer 1997).
Related to the multispecies approach to bycatch is the idea of individual bycatch quotas as opposed to fleetwide quotas or total catch quotas of bycatch (Alverson et al. 1994). The idea is that each individual fisher would be given an incentive to reduce unwanted bycatch, instead of everyone racing to catch their quota of target species before others in the fishery. This can happen even with individual quotas for the target species, because fishers want to avoid the restrictions imposed by bycatch limits on those who have not yet taken their quota of target species. One result of this is that the quota for the target species is not reached before the bycatch quota stops the fishery. However, establishment of individual quotas for bycatch as well as for the target species appears to allow better control of results (Trumble 1996): managers then have the option of keeping the target catch constant and reducing bycatch or of increasing the target catch while keeping bycatch constant. A program of individual vessel fishing quotas for halibut in the sablefish fishery in Alaska appears to have reduced discard mortality of halibut to about 136 t in 1995 as compared with 650 t in 1995 (Pennoyer 1997). In that program, the bycaught fish must be retained and landed if they are legal size. Bycatch management and enforcement often require the
use of observers on fishing vessels and it is often time-consuming and costly, even for some programs that hold individual vessels responsible for their bycatch (Pennoyer 1997).
In some cases, however, the above approach would not work. There is no rational way, for example, to develop a total catch quota for endangered species such as sea turtles caught in shrimp trawls. Even when the above multispecies approach is appropriate, there can be unwanted bycatch (i.e., bycatch of kinds and sizes of animals that cannot be used). In those cases three basic approaches to bycatch reduction have shown promise: changes in the pattern or intensity of effort, changes in the fishing gear used, and bycatch-reduction devices (BRDs) (Alverson et al. 1994, FAO 1997d, Natural Resources Consultants 1998). For example, the National Research Council (NRC) recommended a combination of two of the above approaches to protect endangered sea turtles (NRC 1990). To change the pattern and intensity of effort, the NRC recommended avoidance of certain sensitive areas at certain times and reduction of tow times. The NRC also endorsed the National Marine Fisheries Service regulations requiring the use of turtle-excluder devices, which are a form of BRD. Changing fishing gear could mean changing emphasis, for example from long-lines to trawls, or vice versa, depending on the nature and extent of the bycatch. It also includes changing details of the gear, such as mesh sizes and shapes (Bublitz 1996, Kennelly and Broadhurst 1996). Investigations of so-called active BRDs also show some promise (e.g., Loverich 1996). However, those devices, which are controlled by an operator in response to observations of catches, depend to a large degree on differences in fish behavior (as other BRDs do), and have not been easy to develop to date.
In assessing the effectiveness of approaches to reduce bycatch, it is important to consider whether the approach will result in an increase in fishing effort at other times and places, with perhaps adverse results. For example, Pereyra (1996) warned that, although midwater trawls for walleye pollock in the Bering Sea have extremely low bycatch (Alverson et al. 1994), they tend to catch smaller fish than bottom trawls, which have higher bycatch rates. Thus, requirements to use midwater trawls could adversely affect pollock populations and increase discards of small fish. This warning again can be seen as advice to consider as many aspects of the marine ecosystem as possible in developing fishery-management plans.
Discards, although related to bycatch, introduce additional difficulties because they often are unreported and sometimes are illegal. We have described the analysis of Myers et al. (1997), who implicated unreported discards in erroneous estimates of fishing mortality on northern cod (Chapter 2). Perhaps a compromise is necessary between the desire to prevent discards and the desire for accurate information on the effects of fishing. This, like the development of most fishery-management activities, is an area that requires cooperation among several groups of stakeholders (i.e., managers, scientists, users, fishers, processors, and
so on). A promising development is research by Lowry et al. (1996) on the fate of gadoid fish escaping from the cod ends of trawls that indicates fairly high mortality for small fish. Larger fish seem to have much higher survival rates. This kind of information is enormously valuable for understanding currently unobserved fishing mortality, and more such research is needed.
Another area needing knowledge is the ecosystem effects of retaining bycatch as opposed to discarding it. To the degree that discarded bycatch consists of dead animals, the question is how much the dead animals contribute to ecosystem structure and functioning. As an example, Alverson has reported that some catches counted as discards by Alverson et al. (1994) are now being retained and used on land far from the place of capture (D. L. Alverson, Natural Resources Consultants, personal communication, 1997). How does that difference in disposition affect the marine ecosystem where the animals were caught? Is it perhaps better for the marine ecosystem to discard some kinds of bycatch than to use it on land? If so, what kinds of bycatch should be discarded and in what circumstances? The answers to those questions are known poorly, if at all, but are important to an understanding and intelligent management of the ecosystem effects of fishing.
Finally, we note that some of the options discussed elsewhere in this report, in particular mariculture and marine protected areas, might achieve socioeconomic and other ecological goals in addition to reducing bycatch by reducing effort.
Marine Protected Areas
For the purposes of this report, a marine protected area (MPA) is defined as a spatially defined area in which all populations are free of exploitation. A primary purpose of such ''no-take" zones has been to protect target species from exploitation and to allow their populations to recover. Such protection has been shown to result quickly in increases in the number or size of individuals of many target species (see Table 5-1). MPAs can also protect critical habitats (like spawning grounds or nursery beds), provide some protection from pollution, protect the marine landscape from degradation caused by destructive fishing practices, provide an important opportunity to learn about marine ecosystems and species dynamics, and protect all components of a marine community (Agardy 1994, Allison et al. 1998, Bohnsack 1998). Protection against management uncertainty is another critical function of MPAs: the populations inside such areas can serve as a "bank" against fluctuations in outside populations caused by fishery-management difficulties or miscalculation. Finally, and perhaps most important, MPAs represent an opportunity to protect ecosystems.
Even small MPAs can result in rapid changes in local populations of fished species (Box 5-2, Table 5-1). Density and average size of fished populations often increase after protection. Even unexploited species can increase because of habitat protection (Russ and Alcala 1989). Larger individuals tend to have much
increased reproductive output, suggesting that overall reproduction of a particular species may increase significantly after establishment of protected areas.
Despite the overall success of MPAs that have been established and studied to date, there are important limitations to their effectiveness and huge gaps in our knowledge about how they function within broader marine ecosystems. Protected areas do not always result in higher density of target species or in higher biodiversity (Ruckelshaus and Hays 1997). This may be because larger individuals exert predation pressure that limits the number of smaller prey species or even juveniles of their own species. It might also result because the ecosystem or community being protected has been so changed by human activity that its original condition is unknown. For example, Jackson (1997) described overexploitation of large herbivores on coral reefs (manatees, the now-extinct Caribbean monk seal, and sea turtles) that devastated their populations and fundamentally changed the ecosystems in many parts of the Caribbean by about 1800. On some reefs, subsistence fishing eliminated most large fishes as well.
In many cases it is difficult to demonstrate the effectiveness of protected areas because of a lack of baseline data. For example, of the citations in Table 5-1, only about half of the studies compared target species before and after reserve establishment. The rest compared areas inside and outside reserves, which does not adequately control for differences attributable simply to habitat quality. The overwhelming differences between some protected and nonprotected areas may make such considerations minor (e.g., Alcala 1988), but well-planned studies of protected areas are required for the full range of protective effects to be understood. MPAs are also less likely to be useful for species with highly mobile life-history stages like pelagic fish or planktonic organisms. In some cases, protected areas could focus on spawning grounds (e.g., for cod or whales), nursery grounds for young of various species, or migratory corridors. However, in other cases the spatially explicit definition of MPAs may not be biologically meaningful, and other management tactics might be better (e.g., temporary full closure of the fishery on the migratory striped bass in the eastern United States).
The large gaps in our knowledge about protected areas should be addressed. Improved understanding will enable more effective MPA design, management, and evaluation. The stability of organisms within them is of primary concern. Little work has been done on this topic, probably because many MPAs are recent. Despite the gaps in our understanding, enough is known to recommend substantial increases in the number and area of reserves (Allison et al. 1998).
The following kinds of information could help make MPAs a more effective tool. Much work is needed to better inform decision makers about the details of establishing and managing marine reserves. Much of the large body of theory and practical experience pertaining to terrestrial reserves (e.g., Meffe and Carroll 1994) cannot be applied directly to marine reserves (Allison et al. 1998) because of their openness (i.e., many marine organisms, especially juveniles, are carried large distances by ocean currents and thus can enter and leave areas unpredictably)
and because marine ecosystems respond to environmental variations differently than terrestrial ecosystems do (see Chapter 3). In addition, the impact of protected areas on the full set of species in the area is generally unknown. Variation in the responses of different species with different life histories is also unknown.
A critical area of ignorance is how a protected area functions within the broader marine ecosystem of which it is a part and if protected areas export biomass or eggs and larvae into the surrounding communities. The export problem is particularly acute because the value of marine reserves as spawning banks depends on the movement of eggs, larvae, or juveniles out of the protected area. Understanding the relationship of reserve size and placement to this export function is a critical step in understanding the value of reserves (Allison et al. 1998). To date, some indications are that reserves can function as net exporters of juveniles or larvae and that population densities adjacent to reserves can be higher than in areas far from reserves (Russ and Alcala 1989), but these observations are scattered and preliminary. Other knowledge gaps include the impact of local and distant oceanic conditions on optimal reserve placement, the impact of natural patchiness of the environment on reserve function, the degree to which edge effects alter reserve-ecosystem functioning, whether populations of species with highly dispersive life histories can replenish themselves communities, and the impact of various nonexploitative recreational within reserves, the size required for reserves to support "natural" activities on reserve functioning.
Some Practical Considerations
Despite the clear advantages of marine reserves as management, conservation, and research tools, their effectiveness depends crucially on how well they are matched to managers' goals as well as to management outside their boundaries. For example, if a reserve is designed to protect an individual species, is enough known about ecosystem processes to predict the result with confidence (e.g., coral reefs as described by Jackson )? Protecting an area can affect the dynamics of the ecosystem within it, and that can affect the abundances and dynamics of individual species in surprising ways. A study of the ecosystem described in Box 3-1 would lead one to expect that protecting sea otters would result in a decrease in kelp and a resulting decrease in organisms that depend on kelp for their habitat or food. This might be an undesired result. It is also important to know if the reserve is the right size and design to achieve its goals for the species or ecosystem of concern. To help increase the effectiveness of reserves, Allison et al. (1998) offered three design guidelines or questions that would help reserve designers assess the significance and trends of the threats to organisms in the proposed reserves.
- Will organisms inside the reserve be able to persist even if exploitation outside the reserve increases?
- Will organisms in the reserve be able to persist even in the face of episodic climate events (e.g., El Niños) and directional changes in climate?
- Will organisms in the reserve be able to persist despite increased pollution, species introductions, and disease?
Allison et al. (1998) pointed out that, although those questions probably could not be answered completely, the awareness of the threats would help designers to make the reserves more effective biologically.
In addition to the need for a clear identification and statement of the goals of any proposed reserve, it is essential to design them adaptively (i.e., in such a way that their effectiveness can be assessed scientifically). This implies the need for monitoring and for controls. Judging effectiveness is not easy (Allison et al. 1998), but it is possible and must be attempted. As an example, Polovina and Haight (in press) described a case in which comparison of a marine protected area with a control clearly showed that the MPA was effective in slowing the decline of spiny lobster populations in Hawaii. However, a large-scale environmental change appears to have led to their declines even in the reserve. The existence of a control and careful monitoring allowed learning to take place, and showed how important environmental fluctuations can be.
The final—and perhaps most important—practical consideration is that MPAs are not substitutes for fishery management, but are one of several tools in the toolbox. If all other methods of fishery management were abandoned, marine protected areas would have to be enormous to protect ecosystems and fisheries, especially to protect widely ranging species. In any case, all the normal (and difficult) management methods of working to involve stakeholders, enforcement, monitoring, and adaptive management need to be used.
The Scope of the Need for Marine Protected Areas
Currently, less than one-quarter of a percent of the sea is in areas termed marine parks, marine preserves, or no-fishing zones (McAllister 1996). The degree of protection in these areas is generally far less than that proposed here. There are very few marine areas in which all species and all aspects of the habitat are protected. How much of the marine environment should be included in marine protected areas for them to fulfill their primary functions of ameliorating environmental and management uncertainty, providing a source of eggs, larvae, and recruits to adjacent areas, and protecting critical habitat? An answer to this question is critically important but is one of the most difficult aspects of the emerging discussion on marine protected areas (Box 5-2).
Goals for the scope and purposes of MPAs need to be clearly and quickly articulated. The marine environment is under continued threats, and marine
TABLE 5-1 Some Examples of Protection of Fish by Marine Protected Areas
biological resources continue to decline. Without a clear goal, it is impossible to generate the debate that expansion of MPAs requires or to begin designing and implementing protected areas before environmental damage makes that impossible. For those reasons, recent proposals to establish MPAs in 20 percent of the marine environment by the year 2020 provide a useful reference point for future consideration. The proposals also emphasize the importance of acting immediately to greatly expand the amount of area protected.
This number—20 percent—appears alarmingly and impossibly large at first but is based on a number of independent lines of argument that converge on the need for this general magnitude of commitment (Bohnsack 1994, 1996). Current understanding of marine ecosystems and populations cannot rigorously defend this number against all criticism, but it does provide a rationale for adopting a marine reserve program of this magnitude.
First, the current allocation of 0.25 percent of marine areas to reserves is too low to have an appreciable effect on marine populations across their range or
A series of studies in the Philippines illustrate several important successes and challenges related to MPAs (Russ and Alcala 1989, 1994, 1996; Vincent 1997a, 1997b; Russ 1989; Pajaro et al. 1997). The studies focused on small (less than 100 ha) reserves at Sumilon Island, Apo, and one near Handumon village, Northwestern Bohol in the central Philippines.
In all cases the reserves were effective to varying degrees in increasing the abundance and diversity of fishes, and at Sumilon fish abundance increased in adjacent waters as well. The increases in the reserves were substantial in some cases; for example, biomass increased from approximately 1.5 to 18 kg per 1,000 m2 in nine years of protection at Sumilon and from 1 to 10.5 kg per 1,000 m2 at Apo (Russ and Alcala 1996). At Sumilon, the resumption of unregulated fishing reversed the gains (Russ and Alcala 1994, 1996). Data on Handumon are not yet available but the fishers perceive an increase in the seahorse populations there (Vincent 1997a; Amanda Vincent, McGill University, personal communication, 1997).
In all cases the involvement and support of local fishers were a prerequisite for any success of the reserves; similar findings were described by Dye et al. (1994) and Odendaal et al. (1994) in South Africa and in Chile. Castilla and Fernandez (1998) also presented arguments that successful management of small-scale fisheries in general requires this involvement of fishers. In all of the above cases, enforcement was a problem that could be solved only when local fishers were sufficiently committed to the reserves and sufficiently concerned about threats to their resources that they were willing to act together to enforce the rules and prevent poaching.
These cases demonstrate that even local fisheries using little modern technology can devastate local marine ecosystems. In some of the cases, most notably that of the seahorses under study in Handumon (Vincent 1996, 1997a, 1997b; Pajaro et al. 1997), much of the fishing pressure resulted from the large international market in seahorses outside the area. A total of 32 nations are involved in trading seahorses. In particular, China and Hong Kong have large markets; in the early 1990s, annual consumption of seahorses in China, Taiwan, and Hong Kong exceeded 41 t per year, or more than 14 million animals (Vincent 1996). The animals are prized as aphrodisiacs, antiarthritic agents, and as anticholesterol therapy (Vincent 1996).
In all the cases described, success required at least some knowledge of the organisms' biology, and in all cases, additional information about that biology as well as the animals' physical and biological environments was considered likely to improve the reserve's success. The seahorse reserve at Handumon could be successful despite its small size because seahorses do not disperse widely; the reserves at Apo and Sumilon were able to increase fish abundance because the adults of those species have small home ranges. Although the larvae of many coral reef fishes disperse long distances, protection allowed arriving juveniles to survive and grow (Russ and Alcala 1996). Over the long term, it probably will be necessary to protect larger areas if all the surrounding areas are overfished. It is equally true that any attempt to establish such MPAs without adequate knowledge of local socioeconomic conditions would surely fail.
marine ecosystems, and thus this level of reserve commitment does not meet current goals for marine reserves as a management tool. In terrestrial systems, slightly more than 5 percent of the earth's land area enjoys some sort of protected status (Groombridge 1992). Clearly, even that amount of protection, while helpful, is not enough to prevent continuing loss of biodiversity. Marine systems are much more open, with more geographic exchange, than most terrestrial systems and thus need an even greater area of protection than terrestrial systems when species with dispersive life stages are involved.
A second line of evidence independent of the first is that goals of fishery management often focus on the protection of a certain fraction of the spawning stock (Clark 1996). If only 2 percent of the standing stock of a species is allowed to spawn, each individual must produce 50 offspring for the population to maintain itself. Such high reproduction per individual is very sensitive to environmental conditions and can lead to the collapse of the standing stock. As a result, fishery managers often try to protect at least 20 percent of the population. Spawners must then produce at least five offspring each, but this value is likely to be more easily achieved on a long-term basis than the 50-offspring requirement discussed above. An example of this approach was given by Bohnsack (1994), who calculated that protecting 20 percent of the habitat of red snappers (Lutjanus campechanus) would increase the total productivity of the fish population. That increased productivity would more than compensate the fishers excluded from the closed areas. These arguments are supported by information that fishing drastically reduced the numbers of spawners in several species (Mace and Sissenwine 1993, Goodyear and Phares 1990).
A third line of reasoning is an application of the inverse-square law. One of the goals of protected areas is for them to export eggs, larvae, or juveniles to other areas. The sea is a powerful dilution agent, and eggs and larvae of even coastal species will decrease in density as they spread out from a center of production. Settlement away from a protected area will decline rapidly with distance, and unless the protected area is very large, or occurs as a string of protected areas along a coast, the export of larvae and juveniles will be limited. Although research on this topic is critically needed, the large dilution capacity of the oceans suggests that a substantial fraction of habitats need to be larval exporters for reproductive individuals within reserves to have an effect outside reserves.
An Economic Argument for Marine Protected Areas
The establishment of MPAs has the potential to affect many fishers, especially to the degree that they lack mobility and the MPA excludes them from traditional fishing areas. It also might appear to impose a heavy cost on industry with no offsetting benefits. Nonetheless, there are economic considerations that can actually favor MPAs. This report has emphasized the unavoidable uncertainties involved in resource management, great enough to lead to the collapse of
resources in some cases (e.g., northern cod). MPAs could act as a hedge against such uncertainty if they are big enough and are properly designed. The expected economic return from exploiting the resource might be reduced by the MPA. If it is reduced, one can argue that there is a compensating tradeoff in the form of reduced risk: one gains protection against catastrophe.
The principle of hedging one's bets is widely followed in many economic activities where irreducible uncertainty is encountered. For example, investors in stocks and bonds who are risk averse are usually advised to diversify their portfolios of financial assets. Part of the diverse portfolio includes low-yielding, liquid, and safe assets such as U.S. Treasury bills. Although the overall yield from the portfolio is reduced in certain economic environments by this approach, risk is reduced as well. MPAs can be seen as playing a role like that of liquid assets in a financial portfolio. Reduced expectation of economic returns is offset by protection against future economic disaster.1
Much has been written about the need for management to better conform to the time and space scales of fisheries and fishing communities. Regional and global restructuring of political systems over the past few centuries has diminished, dismantled, and distorted local systems of decision making and authority, as Johannes (1977, 1978b) and others have shown for Oceania. There are often mismatches between natural and political boundaries, with either too many or too few political boundaries, contributing to risk-prone management. At one extreme, ecosystem approaches to fishery management are hindered by too many boundaries, the situation in which watersheds or other ecosystems are divided among the jurisdictions of multiple governments and agencies. Such problems require institutional change toward more effective regional management systems (NRC 1996b). The regional fishery management councils in the United States were formed to address the problem of jurisdictional mismatches and multiple boundaries, but the problem remains both within that system and in relationships between it and state and interstate management bodies.
However, at the other extreme, erasure of boundaries can diminish the capability of communities, tribes, and other local governing entities to impose controls on what is caught from and done to the local land and sea (Cronon 1983). Accordingly, one of the major institutional challenges to using ecosystem approaches is the construction, or reconstruction, of appropriate boundaries, tailored to each specific fishery and fishery-management issues. Management centralized in national or regional authorities must be balanced with involvement of
local stakeholders, communities, businesses, and property owners (Lee 1993, Pinkerton and Weinstein 1995, NRC 1996b, McCay and Jentoft 1996, Hanna 1998). A major challenge is developing institutional structures with sufficient complexity in scope and scale to be appropriate for complex and dynamic ecological systems (Ostrom in press). The boundary problem is political and must be addressed to develop effective fishery management.
International Developments in Fishery Management
Fisheries that cross jurisdictions and even national boundaries are problematic. The United States makes policy pertinent to its own fisheries; it tries to influence policies with respect to international fisheries. Those activities need to be viewed as separate efforts. Therefore, we consider the likely usefulness of international agreements, treaties, and conventions here.
International concern about the sustainability of global fisheries has resulted in the inclusion of precautionary approaches in several recent international agreements related to fisheries (e.g., Box 5-3). Three agreements form the basis for international fishery management and provide goals for national fishery management systems.
- The United Nations Convention on the Law of the Sea (UNCLOS)—This treaty entered into force on November 16, 1994, and is one of the primary instruments for the sustainable use and development of the ocean and its resources. The convention is based on a philosophy of rational use conforming with environmentally sound development. Among its many features, the convention promotes the goal of sustainability of fisheries (Articles 61.2 and 119.1). Article 61.2 requires countries to ensure (through proper conservation and management measures) that living resources in their EEZ are not endangered by overexploitation. Of the top 20 fishing countries, four have not signed the treaty: Peru, South Korea, Taiwan, and the United States, although the United States does comply with the treaty.
- Agreement for the Implementation of the Provisions of the United Nations Convention of 10 December 1982 Relating to the Conservation and Management of Straddling Fish Stocks and Highly Migratory Fish Stocks (General Assembly of the United Nations 1995)—This agreement added to the UNCLOS framework provisions for international management of highly migratory fish species (e.g., tunas, billfishes) and fish stocks that cross international borders. It was adopted on December 4, 1995, but had not been ratified as of June 1998.
- FAO Code of Conduct for Responsible Fisheries (FAO 1995c and d)—This agreement (Box 5-3) is based on Agenda 21 of the Rio Conference on Environment and Development and is notable because of its focus on the precautionary approach and because it places the burden of proof that uncertainty allows an increased catch on the fishers, not only on the managers.
These three agreements commit signatory countries to sustain their national fisheries, cooperate to sustain international fisheries, address the problems of overcapacity and bycatch, base management on sound scientific information, and conduct many other activities that should improve fisheries in the future. Because these agreements are relatively new, it is difficult to gauge their impact. Like all international agreements, their merit will depend on how many major fishing countries ratify the agreements, whether signatory countries abide by the agreements, and the effectiveness of the limited enforcement provisions. Adoption of these agreements will promote the ecosystem-based approaches recommended by this committee.
1. PRECAUTIONARY APPROACH AND BURDEN OF PROOF
Composition of Management Institutions
Fishery management is primarily a social process. To make it more successful, goals such as ecosystem resiliency (Arrow et al. 1995) should be more explicitly linked to social values (Norton 1995), and to do that, it is necessary to create institutions and organizations that help involve stakeholders in the process. The mobilization of society to reject unacceptable risks and to lobby for more conservative resource use may be a significant means of confronting uncertainty in future resource conservation and management and of developing risk-averse management (Alverson et al. 1994). A similar view was expressed by Parrish (1995), who noted that in many areas of the world one can determine when stocks are overfished and depleted, but few countries have political and fishery management systems capable of managing stocks at a fishing mortality rate that even approaches optimal levels.
Stakeholder involvement in fisheries and watershed management can be achieved through a range of mechanisms, from public hearings and advisory groups to collaborative or partnership management by governmental or non-governmental bodies to community-based management (Jentoft and McCay 1995, McCay 1995b, Pinkerton and Weinstein 1995, Hanna 1998). Butterworth et al. (1993) described a process where industry representatives—the users—were included in setting catch rates during the fishing season for the South African anchovy (Engraulis capensis). There is a great deal of uncertainty as to the stock size in that species, with a midseason survey being required to set reliable catch quotas. The industry involvement provided a more effective way of dealing with uncertainty than simply having the management agency alone set catch quotas. A somewhat similar process is used in setting catch quotas in the United States through the regional fishery management councils.
How should fishers and other individuals who gain economic benefits from natural resources be involved in management processes? A risk of participatory democracy in fishery management that must be avoided is the potential for the process to be captured by narrow interests (Jentoft and McCay 1995, McCay and Jentoft 1996). Several studies of the U.S. fishery-management system under the Magnuson-Stevens Fishery Conservation and Management Act have highlighted conflicts of interest as a major cause of regional council failure to prevent overfishing (Ludwig et al. 1993, WWF 1995). In addition, the workings of the regional councils tend to reflect and perhaps increase competition among fishing sectors and engender adversarial relationships with the National Marine Fisheries Service, as well as increased reliance on lawsuits and congressional involvement. These and other problems somewhat reduce the benefits of participatory and regionalized decision making.
On the other hand, early and meaningful involvement of members of the public can have important benefits (Hanna 1995). The approach of involving stakeholders in the regional councils is an innovative one, and in some of the regions the process works better than in others, leading to optimism that the
process can be improved. As another example, the NRC (1996b) has recommended the development of management institutions to conserve salmon in the Pacific Northwest that include a regional or watershed-scale component, that allow for shared decision making among all legitimate interests, and that ensure that local or regional interests not be permitted to override the interests of the greater region.
Sand (1992) found that failure to enunciate clear goals has constituted an institutional flaw in many environmental agreements. Miles (1994) suggested that institutional performance could be improved by
- matching fishery effort and resource availability (i.e., reducing overcapacity);
- broadening the focus of fishery management to include all sources of environmental degradation that affects fisheries;
- structuring the duty to cooperate and conserve through a set of international principles;
- implementing effective monitoring and enforcement principles; and
- creating institutions with the capacity to mandate collection and exchange of vital data.
Much of fishery management, at least from an economic standpoint, is designed to cope with the problems created by the current economic incentive system described in Chapter 4. Two broad approaches—not mutually exclusive—have been tried: regulation of effort and a rightsor privilege-based approach.
Regulation of Effort
Regulation of effort is an attempt to prevent fishers from responding to economic incentive systems in ways that society deems unacceptable. Various controls are put into effect, for example, setting total allowable catches, imposing net-mesh regulations, limiting boat size, limiting gear type (e.g., restricting ocean salmon fishing to troll gear), seasonal restrictions, and controls on the number and kinds of vessels entering a fishery. These aim to prevent overexploitation of a fishery resource, in part by reducing excess fishing capacity.
The regulatory approach has unquestionably had some successes in protecting fishery resources. Examples include Pacific halibut, walleye pollock in Alaska, and various cases where the controls became absolute (i.e., the fishery was completely halted, such as striped bass in waters of the eastern United States). However, regulations have not in general reduced excess capacity. Fishing capacity represents many inputs, and it is almost impossible
for resource managers to control all the inputs. As long as the economic incentives remain in force, fishers attempt to substitute unregulated inputs for regulated ones (Dupont 1996). As described in Chapter 4, fishers are ingenious in circumventing regulations to limit entry and fishing power. This recognition has led to the conclusion that the most effective way to reduce fishing power is to address the incentive problem. This means that one needs to identify a system that reduces or removes fishers' rewards for overcapitalization and that increases the likely future returns on their investments, so that investment by conservation or enhancement is in their interest. Two approaches to development of such systems are described below. Although at first the approaches appear to be polar opposites, they do have features in common and in some circumstances appear to have the potential to coalesce. The following discussion should be taken as providing examples of approaches to a problem that the committee believes must be addressed, rather than as an endorsement of any particular scheme over any other. Whatever system is used, it appears that including the users in the decision-making process is important to achieving successful outcomes. To ease the socioeconomic and political realities of changing from one regime to another, some process of ''grandfathering" rights or privileges should be considered, as it often is when political or legal regimes change (e.g., tax laws, changes in Social Security).
The Opportunity to Participate in Fisheries
A key issue in the management of fisheries is who has the opportunity to fish. Historically, the opportunity to fish has been open to all (referred to as open access), and this has led to the many problems described in other sections of this report. Even when access has been controlled or limited, these problems are usually not corrected. The problems are a result of fishery-management systems that make participants in a fishery competitors for a share of the available resource, which makes it rational for them to try to quickly catch a larger share. While this behavior is rational for individuals, it drives up total costs, so that the net economic benefits from fisheries dwindle or may be negative (at least for a while), and participants do not invest in conservation for the future, largely because they have no assurance that they will be beneficiaries in the future. The essence of the problem is that a fishery-management system that makes participants compete for shares of the resource does not provide incentives for efficiency and conservation.
The alternative are fishery-management systems that assign exclusive rights or access privileges to a share of the resource. In this situation, the incentive is to use the shares efficiently rather than to spend more in competition for a bigger share. And if participants' rights extend into the future, there is an incentive to make conservation investments because there is some
assurance of the opportunity to benefit in the future. This form of management is often referred to as rights-based.2
There are many forms of rights-based fishery-management systems, usually specified in terms of the definition of exclusivity, the rules about transferability of the right, and the nature of the right. This specification of the management system has a great deal of influence on the effectiveness of incentives for efficiency and conservation. Exclusivity may apply to individuals, corporations, government entities, communities, or other groups. For the specification to result in positive incentives, the exclusivity must be assigned to entities that are cohesive enough to act for the collective good of the entire entity (i.e., as quasi-individuals). Rules about transferability may range from disallowing it to allowing it without restrictions. Transferability increases economic efficiency by allowing rights or privileges to be transferred to entities that can use them most efficiently (i.e., with the lowest cost of fishing). But there are many reasons for restricting transferability, such as to prevent monopolies or for social reasons (e.g., to preserve the traditional nature of participation in fisheries). The nature of the right or privilege may be a specific amount of catch, an annual share of a total allowable catch, a specific amount of fishing effort or units of fishing capacity, or a geographically defined fishing area. Some types of specification may not completely eliminate the incentive to compete for a larger share of the resource, but they may have other positive attributes, such as lower enforcement costs or greater social acceptance. Box 5-4 describes a successful example of limiting access through permits and quotas.
The most appropriate specification of rights-based management will be a compromise between management objectives and constraints that varies from one case to another. But it is clear that some form of rights-based management that instills positive incentives for efficiency and conservation is needed.
A promising approach to rights-based management is to award rights to communities. These include place-based communities, like municipalities or the
The fishery for the herring (Clupea harengus pallasi) in San Francisco Bay is an example of a fishery that appears to be maintained in a sustainable fashion by using traditional single-species management strategies. The fishery has been characterized by three major peaks in landings in response to new demands. The first peak occurred in 1918, when most of the fish were processed into fish meal. When the reduction of whole fish into fish meal was prohibited in 1918, the fishery ended. From 1947 to 1954, whole herring were harvested and canned to make up for declining sardine stocks. The fishery again declined. Since 1973, a new international market has developed for herring roe, a delicacy in Japan (Spratt 1981).
Pacific herring live most of their adult lives in the ocean and return to San Francisco Bay only to spawn. Large schools of herring enter the bay during spawning season and may remain for up to three weeks. After spawning, they return to the ocean, where they are planktivorous, feeding primarily on zooplankton. Other marine fish and birds may forage on herring, but no higher predator depends on a diet of herring alone. Herring mature at age two and may live for 10 years, making each fish capable of several spawning migrations into the bay (Ware 1985).
San Francisco Bay provides over 90 percent of the state's herring catch (Spratt 1981). The fish must be caught within one day of spawning or while spawning is in progress, a season from November to March. A secondary fishery exists for the roe deposited on kelp fronds. Giant kelp is removed from the vicinity of the California Channel Islands and suspended from rafts in San Francisco Bay in areas where spawning is likely. After spawning, the kelp is collected with the roe attached.
Management of the fishery is based on population estimates from annual hydroacoustic surveys and spawning ground surveys. Quotas are set at about 15 percent of the amount of herring expected to return to spawning areas. These quotas are adjusted annually, and the maximum catch rate is recommended to be 20 percent (Trumble and Humphries 1985). The estimated population of Pacific herring in San Francisco Bay declined in the 1983–1984 season, probably in response to the 1983 El Niño. However, the population has been rebuilding since 1984, and spawning biomass was approximately 65,000 tons from 1987 to 1990 (Spratt 1992).
The primary tool of management is limiting entry to the fishery. Since 1983, only five new permits have been issued, and the total number of permits in San Francisco Bay is stable at about 400 for fishing and 10 for the roe-on-kelp fishery (Spratt 1992). Regulations change yearly and respond to new conflicts that arise. Several new techniques have been used: permits issued by lottery, individual vessel quotas, and allowing the selling of permits. All of these are single-species techniques, yet they have provided effective management of the fishery.
fishers of a particular area; interest-based communities (i.e., groups of fishers, fish processors and others who work in the same fishery, including cooperatives and corporations); and a broader conception of community, here called virtual communities.
There are examples of coastal communities that succeeded in managing fishery resources sustainably and avoiding overexploitation (Christy 1982, Cordell 1989, McGoodwin 1990, Hviding and Baines 1994, Pinkerton and Weinstein 1995, Leal 1996), although in some cases, the demand for food can overwhelm the ability of coastal communities to manage their resources (Simenstad et al. 1978, Jackson 1997). How have the successful communities overcome such problems as ill-defined property rights, which contribute to the overcapitalization that seems so important to the overexploitation problem? Pinkerton and Weinstein (1995) identified some prerequisites for the success of such schemes, including a minimal degree of exclusivity with respect to the resource, a high degree of community dependence on the resource, and the ability to assert management rights on an informal, if not formal, basis.
Until recently the focus has been on place-based communities, like local tribes and cooperatives or fishing villages. There are promising new directions being taken, including the use of community quotas, such as the community-development quotas allocated to certain community organizations of western Alaska (Ginter 1995). With the development of interest in "comanagement" arrangements, where government bodies and groups of resource users share the rights and responsibilities of managing resources (Jentoft 1989, Pinkerton 1994), other forms of community have been recognized as well. For example, the Canadian government is entering into "partnership" agreements with groups of licensed fishers who establish their own management plans within frameworks established by the government. There has also been discussion of the prospects of a corporate model for fishery management (Townsend 1995, Townsend and Pooley 1995). Moreover, some advisory groups for government management agencies have begun to function as communities of collaboration among industry members, scientists, enforcement officers, recreational anglers, and others for fishery management (McCay et al. 1995). Based on these and other ideas discussed at the committee's conference in Monterey in February 1996, the committee recommends expanding the notion of community in fisheries management, using the term virtual community.
A virtual community is the functional equivalent of a geographically defined community in which the members of the community may not all reside in the same area, as noted by Reingold (1993) in his discussion of a similar phenomenon among Internet users. It may be thought of as a community of interest, as distinct from a community of place. Some communities of interest might be narrowly defined groups of fishers who share a common interest in a particular management regime and are granted some exclusivity with respect to the resource, or a share of the resource, in exchange for taking on responsibilities in
managing it. Others might be made up of landowners, whose activities impinge on habitats important to fish species. Even more promising are virtual communities that constitute all parties sharing an interest in a fishery and its associated habitat. Such communities would include a very wide range—fishers and fish plant workers, landowners, biologists, community-development groups, recreational anglers, and conservationists—similar to watershed associations.
Like many small-scale, fishery-dependent coastal settlements, virtual communities have the potential to provide the social framework for managing fisheries sustainably. The essence of community is mutual communication, shared understanding for the need to solve problems, and collective action. In this broader concept of a community of interest and collaboration, information on the fishery and the larger ecosystem would be shared on a regular basis. As in virtual reality, where computers provide an opportunity to simulate what happens in the environment, a virtual community could provide computer-based modeling and data-sharing networks. In the Pacific Fishery Management Council's salmon-management process, some fishery leaders are already using fishery models to assist them in reaching consensus with fishery scientists on appropriate harvest levels.
Information would also be openly communicated about the diverse and sometimes conflicting values and needs of the human components of that ecosystem. Ideally, in such a community, trust and mutual knowledge would develop to enable those involved, however diverse their interests, to identify and work on common grounds and develop mutually acceptable goals and visions for the future of the resource and the ecosystem. Accordingly, such communities of mutual interest would come closer to realizing the objectives of ecological approaches to natural-resource management, which includes bringing user groups, local communities, and other members of the public into productive collaboration with scientists and managers (Kessler and Salwasser 1995). Experimental, game-theory, and comparative case-study research over the past decade have shown that communication, trust, and reciprocity are important requirements for groups to engage in the kind of collective action needed for community-based resource management (Ostrom 1998). The degree to which virtual communities and placebased communities have these and other conditions important to the exercise of stewardship also depends on the degree to which the activity in question—fishing in this case—is the primary reason for the community's existence.
Whether based on communities of place, interest, or collaboration, rights-based management is not necessarily dependent on exclusive rights of use or access. Use or access rights are not the only kinds of rights that make a difference in the ability of a community to manage resources sustainably. The broader conception of rights-based management that the committee advocates recognizes, first, that with rights come responsibilities, and, second, that among the critical rights are rights to information, to make policy, to plan, and to coordinate with other uses (Pinkerton 1997). It is then possible to develop more appropriate
institutional arrangements for fisheries. For example, the virtual community may not hold exclusive access rights to a fishery but may establish the conditions and constraints for those who do hold access rights in ways that combine long-term stewardship interests with shorter-term economic imperatives. Contractual arrangements among those with communal rights, for some purposes, and those with individual rights, for other purposes, also can be envisioned (Rieser 1997).
Individual Transferable Quotas
A popular, albeit controversial, management scheme for fisheries designed to alter economic incentives involves individual catch quotas. Commonly the individual quotas are transferable, and hence they are known as ITQs (individual transferable quotas). ITQ schemes have been put in place for halibut in Canada and Alaska (Box 5-5) sablefish in Alaska, wreckfish (Polyprion americanus) in South Carolina in 1992 (Gauvin et al. 1993), surf clams (Spisulas solidissima) and ocean quahogs (Arctica islandica) off the northeastern United States (McCay et al. 1995), and various species in many other parts of the world, especially Australia, New Zealand, and Iceland since about 1985. The basic idea is that the
Despite the existence of a limited-entry system in the Canadian fishery, the length of the season declined from 60 to 6 days within a decade. The industry approached the Canadian Department of Fisheries and Oceans (DFO) for help. DFO worked out an individual fishing-quota program for a two-year term in 1991 and 1992. Quota transfers were initially prohibited. The system was extended another two years with partial transferability, with no more than four shares per vessel. A modified system is still in operation, and quota owners consider the system to be a success. It is an improvement over previous systems from a conservation viewpoint because in the previous 10 years the quota had been overrun seven times and there was too much capacity in the fishery, resulting in significant waste. There are now higher exvessel prices (the price obtained by the fisher when the catch is sold to the wholesaler or seafood broker), fresh fish available throughout the year, and an increase in the number of buyers. Innovative marketing has been implemented. Users pay the monitoring cost. This is a good example of cooperative government-industry problem solving.
Similar difficulties affected the Alaska halibut fishery. The fishery's management had been successful in protecting the resource, but socially and economically it had unfortunate results. The fishery was characterized by overcapitalization and an extreme "derby" fishery (Buck 1995). As a result, an ITQ system was promulgated in 1995 (NMFS 1996a). The results have not been entirely to everyone's liking, but individual quotas have been well received by many segments of Alaska's halibut-fishing industry (Smith 1997).
authorities would set a total allowable catch (TAC), and then divide the TAC among individual fishers or groups of fishers (companies). The individuals, no longer having to compete for shares of the catch, would no longer be rewarded by overcapitalization. Described this way, ITQs—being individual transferable quotas—appear to represent a fundamentally different kind of ownership from virtual communities, in which ownership of the resource is collective. But since ITQs often develop into more than merely claims to catch shares, there is potential for ITQs and virtual communities to be complementary rather than conflicting (Scott 1993). Indeed, Alaska's community development quotas (CDQs)—catch quotas allocated to coastal communities—represent an evolution in this direction, and Castilla and Fernandez (1998) described how ITQs evolved in this fashion in an inshore fishery for invertebrates in Chile.
Combined Rights-Based Approaches
Although ITQs mitigate some adverse economic incentives, they do not necessarily reduce the discount rates experienced by fishers enough to ensure resource conservation (see, e.g., Gauvin et al. 1993, Mace 1993). That is because the productivity of the resource may be lower than the economically rational discount rate, a limitation of all forms of allocation. Furthermore, ITQs seem likely to work only if effectively monitored and enforced by authorities. However, in some cases, holders of ITQs may form virtual communities, by themselves or with others. These communities exercise collective responsibility for management of the resource (Scott 1993), as noted in New Zealand (Annala 1996), Iceland (Arnason 1996), and Nova Scotia (McCay et al. 1995), and thus can improve the enforcement of the well-defined rights conveyed by ITQs. In those countries (as elsewhere) the ITQs are expressed as percentages of the TAC rather than as fixed quantities. Thus, if the resource declines, the quotas' values decline as well, and that provides an incentive for investment in the resource (i.e., wise management). In this way, they take on some (but not all) of the attributes of shares in a corporation. They are claims to the stream of economic return from the natural capital (i.e., the resource), and their value will reflect the capitalized value of the expected returns from the resource. Expected consequences of management (and of natural fluctuations and environmental changes) will be reflected in the price of the ITQs, and thus sound management should be rewarded. Indeed, the price of wreckfish ITQ shares is stable and fish prices have increased (NMFS 1996b). To the degree that this occurs, the ITQ holders are taking on some of the characteristics of shareholders (i.e., they become de facto collective owners of the resource). An example is that of ITQ owners in a New Zealand abalone fishery who levy a charge on their sales to fund research and stock-enhancement programs of their association (Pearse and Walters 1992).
Clearly, a great deal more experience is needed with such schemes to understand their ramifications, the circumstances in which they might work well and in
which they might fail, and the appropriateness of variations on the themes. Some of these questions and practical experiences with ITQs have been discussed by McCay and her coworkers (McCay 1995, McCay et al. 1995) and OECD (1997). Concerns about ITQs include questions about equity (e.g., the assignment to individuals of exclusive rights to exploit resources that are perceived to belong to everyone and whether the shares should be given away or sold), concentration of the rights or shares in very few people's hands, questions about their effectiveness in promoting stewardship, their effect in reducing the number of participants in fisheries, the best extent and duration of the rights or shares, and other related concerns. In addition to researching those concerns, there is room to consider broadening the scope of the virtual communities, especially if one views fisheries in an ecosystem context. Should timber interests be included in salmon-fishery virtual communities, for example?
Hanna (1998) provided an analysis of ways to adapt institutions and property-rights regimes to an ecosystem approach to fishery management that reflects attributes of the ecosystem and its human users, values ecosystem services, and coordinates interest groups and managers on a broad ecosystem scale (see also NRC 1996b). Management structures need to promote the definition of multiple objectives through processes that are legitimate and flexible and that promote socially appropriate time horizons for resource use and decision making. They need to take uncertainties into account, including what Hanna calls fact uncertainty (lack of knowledge) and tenure uncertainty (resulting from unspecified property rights or uncertainty in political systems). Despite these uncertainties, the recent developments of various rights-based allocation schemes offer considerable hope for sustainable fishery management. Indeed, it not clear what other general course offers as much promise.
Managing complex biological systems is difficult because of the often large differences between the social and temporal scales of natural and socio-political systems (NRC 1996a, 1996b). Although incentive structures created by allocating transferable use rights to private entities may promote greater efficiency and new ways of valuing the resources, unless they are designed correctly, they may not by themselves adequately protect and enhance ecosystem goods and services by protecting habitat, preventing pollution, and coordinating with other fisheries (Scott 1993). Hence, other forms of organization, including comanaging and virtual communities, are needed as well. In the broader sense of rights-based fishery management advanced in this report, different kinds of rights, ranging from rights to a resource to rights of governance, would be combined with different forms of ownership, ranging from individual ownership to ownership by communities and to ownership by the public or national government. The public-trust nature of marine resources as well as public rights in the ecosystem goods and services provided by marine environments (Rieser 1991) can be combined with private and community-based ownership of rights, access, capture, and management
(Rieser 1997). Such institutional complexity is also necessary in managing the biological complexity of marine ecosystems (Ostrom in press).
The above discussion focuses on commercial fisheries, but recreational fishing is also central to the sustainability of many fisheries and is subject to economic incentives that can inhibit (or promote) conservation. An example of these incentives is provided by marine sportfishing tournaments. Although many such tournaments require release of the captured fish alive, many require at least some fish—usually dead—to be weighed to be eligible for the prize, which can exceed $100,000 in cash and equipment for the heaviest fish, with additional prizes for runners-up and for other categories. A recent (summer 1998) search of the World Wide Web turned up information on dozens of saltwater fishing tournaments with total purses of up to $1 million, and sportfishing magazines have many advertisements for such tournaments in every issue. There are also many smaller tournaments with much smaller purses, often for smaller species of fish such as bluefish and various mackerels.
For a 500-pound fish, a prize of $100,000 would amount to a price of $200 per pound, much more than the usual value of a fish caught for food. In addition, the prize money is uncertain: it depends on what others have caught as well as the willingness of the sponsor to pay. Therefore, such tournaments—which are popular throughout the coastal United States—can encourage the killing of fish for uses other than personal consumption or even sale for food. For large fishes at high trophic levels or for long-lived, slow growing species, such tournaments can contribute significantly to the overall fishing mortality. As the importance of such mortality has become clearer, many sponsors, especially of billfish tournaments, have moved toward tournaments in which prizes are given for fish that are released alive, and to the degree that such tournaments can be substituted for tournaments in which fish must be killed to earn prizes, an important economic incentive to kill fish would be eliminated.
The principles and objectives of rights-based management apply to recreational and subsistence fisheries as well as to commercial ones. For example, where there is private or association ownership of recreational fishing sites, as in many inland rivers and lakes, there may be substantial efforts to protect habitat, prevent pollution, and work toward enhancement of fish populations. A notable example is that of community-run systems for managing moose, salmon, lake trout, and other species in Quebec (Leal 1996). Moreover, if rights are allocated to commercial fishers, it is possible for groups of anglers to buy out those rights for their own purposes, including conservation; for example, the North Atlantic Salmon Fund bought rights of salmon fishers from Greenland and the Faeroe Islands beginning in 1991. Furthermore, all groups with well-defined rights are thereby in stronger positions to use the courts to protect fishery systems from pollution and habitat destruction (Brubaker 1996).
Understanding Marine Ecosystems
The ability to use ecosystem approaches to sustain marine fisheries will depend on better information. Managers need understanding and models that encompass components of ecosystems (including humans) and information about whether they are changing and if so, how; the causes of change; and how negative changes or impacts might be reduced. An understanding of the importance of the impacts of local activities to much larger spatial and temporal scales is crucial. Ocean science can contribute the requisite information and must also be tapped to develop new tools for observing and managing fish populations and marine ecosystems. Fishery managers are required to use the "best scientific information available" (MSFCMA National Standard #2). Unfortunately, the best scientific information may not be communicated to policy makers and, even if communicated, is not always adequately used. Thus, this aspect of science related to implementing ecosystem approaches must be linked closely with institutional mechanisms that specify the information needed and communicate it to managers, policy makers, and the public.
The only way to anticipate how ecosystems will respond to perturbation is to develop a better understanding of how mechanisms at lower levels of organization provide the feedbacks that govern the dynamic and nonlinear features of ecosystem responses. Marine ecosystems are assembled from loosely coevolved species into assemblages. They have the capacity for multiple stable states and complex dynamics, including chaotic fluctuations. A static view that is restricted to a description of flows is inadequate to understand the responses of a system beyond the range of conditions it has previously experienced, as well as the potential for regime shifts or other qualitative responses to climate change or fishing pressure (Steele 1998).
Ecological systems are complex interconnected nonlinear systems; as such, their dynamics may be very sensitive to past conditions, and subject to shifts in dynamics when exposed to environmental stresses or sustained fishing pressure. Keystone species such as the sea otter (see Box 3-1) are important because of their potential to mediate such domain shifts, which can have dramatic consequences for marine fisheries. The collapse of the Barents Sea herring stock because of overfishing is a case in point; with the demise of the herring, pressure on capelin was reduced, leading to an increase in those stocks; on the other hand, cod populations declined owing to inadequate food supply, and whale populations changed their range, with important effects on the dynamics of other ecosystems.
The resilience of an ecosystem can be defined as its capability to maintain essential structure in the face of perturbation and to resist significant shifts in dynamics. This is not an adaptation of the system in an evolutionary sense, but it
is an emergent property with profound importance to humans. Flexibility is the ecosystem-level equivalent of the ability of species to ''adapt" in response to environmental changes in ways that lead to persistence. For the individual species that flexibility is embedded in its genetic diversity; the same idea applies at the ecosystem level, where biotic diversity is equally important to resilience. Management for sustainability means preservation of biodiversity both for its own sake and because of its importance in maintaining ecosystem resilience.
Marine ecosystems are often defined by geographic boundaries, such as the Bering Sea. They can be large and can overlap geographic and political boundaries; their own boundaries are often not easy to delineate or define (Alexander 1990). In this discussion, we use the concept of the large marine ecosystem described by Sherman (1990) as "relatively large regions of the world, generally on the order of 200,000 km2 [or more], characterized by unique bathymetry, hydrography, and productivity…" The geographic scale of most marine ecosystems is larger than the scale of the local human communities that depend on them. Furthermore, marine ecosystems are frequently include discontinuous habitats with substantial inputs (e.g., nutrients, sediments, energy, invaders) from other kinds of habitats in the ecosystem or from other ecosystems. Because of this coupling of habitats in different areas, management schemes at the local level must include the effect of local decisions on the larger ecosystem and over long times.
Fishes and other components of marine food webs have complex life histories with different habitat requirements at each of several stages. Differences among species in spawning grounds and dispersal ability have important implications for management. Knowledge of habitat requirements and of timing of settlement of larval dispersal stages is needed to understand the effects of localized changes in environment and to predict the strength of interactions among species with complex life histories.
Understanding Policy, Institutions, and Behavior
Recent approaches to resource management place humans and their institutions squarely within ecosystems (Pickett and Ostfield 1995) and recognize the need for constructive and broad-based participation of the public in policy making and implementation (Kessler and Salwasser 1995). As experience with comanagement, virtual communities, participatory research, ITQs, and other institutional innovations increases, so does the importance of designing them appropriately. If it is important to match institutional scales of complexity with biological ones (Ostrom in press, NRC 1996a, 1996b), more work is needed to determine how this can be done in general and in particular cases. The development of comanagement and community-based management raises many issues, including ones of balancing interests of local and interest groups with those of larger publics and longer-term ecological systems and finding ways to gain the benefits
of broad-based participation without sacrificing the benefits of small-group participation and highly informed input into policy making (McCay and Jentoft 1996). The focus on rights-based management and its potential for changing incentives to foster more stewardship raises questions about whether exclusive use rights are necessary to meet that goal or whether other kinds of rights and responsibilities, including those to manage resources or habitats, can work as well or better than exclusive rights to use or access (Pinkerton 1997). The promising new direction of ITQs has a host of related issues. Most central to the question of sustainable marine fisheries is the question of how and whether ITQ management regimes can be designed to realize the benefits of a market-based approach, in terms of efficiency, while also providing incentives for greater responsibility and stewardship (Young and McCay 1995, Brubaker 1996).
Social-science investigations are discussed in various parts of this report, and they need to include research on the structure and functioning of virtual communities, on the choices people make in the face of various management and economic regimes (e.g., ITQs), on people's and communities' adaptations to changing employment opportunities in fisheries, and so on. In addition, research into institutional structures and how they function is important (e.g., NRC 1996a, 1996b).
Data and Monitoring
Valid scientific recommendations are frequently ignored when fishery regulations (e.g., regarding quotas and seasons) are implemented. Collection of reliable data over long periods in the correct locations is crucial for developing better understanding of fish population dynamics and marine ecosystem function. Monitoring is important for detecting trends and patterns of variations over time so that causal relationships among biological and physical factors can be determined. Monitoring and assessments allow evaluation of existing and new management approaches and contribute information useful to research scientists. They are also necessary to gain a better understanding of human and natural effects on fish populations and marine ecosystems, so that predictive and diagnostic models can be created and may indicate new research topics that should be pursued. Analysis of information collected at regular intervals over extended periods should lead to better understanding of patterns of variation and linkages between human population growth, utilization of resources, environmental degradation, and climate change. The importance of such information is indicated by the development of industry-based data bases, such as the Groundfish Data Bank operated in Kodiak, Alaska, for fishers in the Gulf of Alaska and the Bering Sea.
Most existing monitoring and assessment programs consist of periodic sampling by government agencies (states and the National Marine Fisheries Service in the United States; the Department of Fisheries and Oceans in Canada). There are also government-funded programs carried out to monitor fishery ecosystems for the long-term goal of understanding system dynamics. This type of monitoring
is exemplified by observations of the California Current ecosystem conducted by the California Cooperative Oceanic Fisheries Investigation (CalCOFI). This program has accumulated one of the most extensive long-term data sets of ecosystem factors related to fisheries. Other similar programs include various activities of the International Council for Exploration of the Sea and such activities as the Continuous Plankton Recorder Survey, which has been carried out monthly in the North Atlantic Ocean and North Sea since 1948.
The Global Ocean Ecosystem Dynamics program sponsored by the National Science Foundation and the National Oceanic and Atmospheric Administration is designed to study the effects of physical oceanic conditions and climate on zooplankton, fish recruitment, and adult fish populations. The international Global Ocean Observing System will monitor many relevant features of the ocean and provide data to help improve global stock assessments. These data include standard hydrographic data (e.g., sea-surface temperature, current velocities) and biological information about the distribution and abundance of larval forms. The data can be incorporated into circulation models, and indices of advective losses from the population can be obtained. The indices can be combined with standard stock assessment techniques plus estimates of larval mortality to determine the expected range of fish abundance. This could help fishery managers make better decisions by having a better idea of likely year-class abundance. These approaches are now used operationally as one tool to manage walleye pollock in the western Gulf of Alaska (Megrey et al. 1996, Herrmann et al. 1996). At present, the indices of advection are only qualitative (e.g., large, medium, small). Longer data series, coupled with transport models that have been more thoroughly tested, hold promise of allowing quantitative advection indices. In addition, pre-recruit surveys often are useful. For example, they are good predictors of recruitment of haddock into the fishery on Georges Bank, and are routinely used (S.A. Murawski, National Marine Fisheries Service, Woods Hole, Mass., personal communication, 1998).
Other types of monitoring that contribute to an understanding of marine ecosystems include satellite and in situ observations of oceanic conditions that enable estimations and predictions of physical, chemical, and biological factors that influence fisheries. The system used to monitor the El Niño/Southern Oscillation (ENSO) is a good example. Another potentially useful source of information is the long time-series stations maintained by the Joint Global Ocean Flux Study off the coasts of Bermuda and Hawaii. Monitoring and assessment of fish populations are a potential area of cooperation between fishery scientists and fishers.
New techniques for ocean observations will be available in the near future. Synoptic information on sea-surface characteristics, currents, and bottom habitat and topography are available from a variety of remote-sensing techniques. Optical and acoustical imaging techniques will be used more routinely in the future to make biological measurements and to study characteristics of water masses.
Knowledge of the genetic characteristics of fish populations will allow for better studies of population structure and diversity, gene-flow rates among populations, and the migration of species throughout their ranges. Likewise, fish tags are now available to study fish migrations and mixing of populations, the influence of environmental factors on the movement of individual fish, habitat condition in marine ecosystems, and chemical contamination. These archival tags can measure and store information about the environment (e.g., temperature, depth, irradiance) and fish behavior for up to four years. Analysis of hard parts for various isotopes in bony material laid down during growth is also used to study the migrations of Atlantic bluefin tuna and identify different stocks (Calaprice 1986). Investment in acquiring better information will not only improve assessments of fish stocks but also enhance our ability to characterize and manage marine ecosystems.
Other Scientific Tools
A variety of general tools will facilitate the scientific goals related to managing fisheries in an ecosystem context. These include laboratory and field methods ranging from molecular techniques to whole-system experiments to help understand interactions, the importance of uncertainty, and the interplay of multiple stresses on ecosystems. Improved conceptual understanding of ecosystem organization and functioning and of the relationship between biodiversity and the dynamics of component populations will also contribute to the overall success of ecosystem management.
Much of the present scientific effort related to fisheries responds to short-term tactical management needs, because of a lack of resources to proceed beyond the science needed to fulfill legal and regulatory requirements. Although stock assessments are important, emphasis on them leaves few resources for long-term activities that are necessary for constructing more realistic models of fisheries in an ecosystem context. Different kinds of information and/or more information will be needed to develop longer-term strategic fishery management plans and, in many cases, rehabilitation strategies. It is especially important to understand regime shifts and alternative stable states of marine ecosystems and their component fish species.
Obtaining reliable data from observations and experimentation is necessary to develop realistic ecosystem models. Such models, in turn, can indicate new observations that should be collected, new time and space scales for such observations, and new research approaches. At present, a number of different conceptual and mathematical models are used to guide fishery management (see NRC 1998a for a review of stock-assessment models).
Models of biota and their interactions in the complex marine environment will always be imperfect because of imperfect knowledge. Every parameter in every model will be uncertain to some degree. This type of uncertainty is straight-forward to handle; even analyses as simple as those describing propagation of errors can be used, whereas in more complicated situations more extensive sensitivity analyses may be warranted. The mathematical descriptions used in existing models may be incomplete. For example, the effects of entire trophic levels may be grossly aggregated, or neglected altogether, as is often the case for higher predators in models focused on plankton. Important details on the distribution of ages or life history stages are sometimes omitted. Environmental variability and uncertainty are rarely addressed adequately in the models of any fishery. In particular, information regarding the largest known source of interannual variability, the ENSO phenomenon, should be incorporated into models of fish populations and marine ecosystems. This could be of enormous help in managing marine ecosystems in the Pacific Ocean region; similarly, information on the North Atlantic Oscillation would help management in the North Atlantic.
Despite the prospect of more comprehensive models and better data to use in such models, it is possible that accurate fishery forecasts will never be achieved for more than a few years in advance because of the chaotic nature of marine systems (Acheson 1995). There are also modeling results that indicate that in an ecosystem composed of as few as five fish species, in which small fish of every species are eaten by larger fish, the biomass of individual species can vary unpredictably even though total biomass remains constant (Wilson et al. 1991).
Multispecies Models and Management
Recognizing that single-species management fails to embrace a realistic ecological perspective, scientists and managers have increasingly promoted the concepts of multispecies (e.g., Sissenwine and Daan 1991) or ecosystem (Sherman et al. 1993) models to supplement assessments made using single-species models. Early models of fishery ecosystems and multispecies fisheries (e.g., Andersen and Ursin 1977) demonstrated the complexity of potential interactions among species and resulting management. Subsequent model building, which adopted more limited multispecies objectives, has achieved a degree of success in understanding both species interactions and the probable consequences of management alternatives (Murawski 1991).
Institutional inertia and allocation problems, the need for more information for multispecies than for single-species models, and some cases of good performance of single-species approaches leave most fisheries under traditional single-species management. In multispecies analysis, information about abundance, coincidence of distribution, diet overlap, consumption rates, maturity, growth, catch at age, and other factors is used to analyze various forms of predatory and competitive relationships among species included in the model.
Multispecies virtual-population analysis (MSVPA), an extension of the single-species VPA that is widely used as a stock-assessment tool, recognizes the importance and complexity of species interactions by accounting for predator-prey relationships among included species (Sparre 1991, Magnusson 1995). Application of MSVPA has demonstrated that natural mortality rates of fish, especially when young, are much higher than previously estimated and that these rates vary from year to year in response to changes in the relative abundances of predators and prey. In single-species VPA, increased mesh sizes or other regulations that increase the average size and age of fish caught are generally predicted to increase overall yields. However, MSVPA models may predict decreased yields for some species if large meshes allow predator populations to expand, thus increasing their consumption of young fish (Magnusson 1995). This demonstrates the usefulness of such models for investigating ecosystem processes and for developing management strategies. However, multispecies models require more information for success than single-species models, and should be considered as supplements rather than replacements for them.
Another modeling approach deals with trophic interactions among the living elements of marine ecosystems, and drawing inferences about possible ecosystem responses to exploitation of various components from the structure and behavior of such models. Such modeling provides an opportunity to use ecosystem approaches for fishery management. The trophic modeling approach has been used for marine systems because of its ability to incorporate disparate data on the biomasses and feeding interactions of marine organisms to enable rigorous system descriptions, comparisons, and inferences related to fishery impacts on ecosystems (Sissenwine et al. 1984, Pauly and Christensen 1995, Christensen and Pauly 1998). Further, once such a model has been constructed for an ecosystem, and the balance of trophic flows established among its elements, dynamic simulations can be run. These simulations can be used to explore first-order effects of various management interventions at all levels in the ecosystem (Walters et al. 1997).
Two observations regarding the use and limitations of existing and potential models are appropriate. First, ecologists and fishery scientists have developed a great variety of models that yield a range of predictions of population and ecosystem variables. The variation of predictions from comparable models could provide an estimate of the level of uncertainty in our understanding of pertinent marine ecosystems. Second, some surprises might never be captured in models. Such surprises might include economic downturns, cultural shifts, or socio-political upheavals. For example, the dissolution of the Soviet Union has had a major impact on fisheries for Antarctic krill and other species because the combined fishing activities of the Confederation of Independent States are less than the previous effort of the Soviet Union. Perhaps economic simulations could account for small perturbations of the status quo, but predicting the most important and largest disturbances is unlikely.
Experimentation and Adaptive Management
Fisheries are large-scale perturbations that provide the opportunity for experimentation at time and space scales that would never be supported by any of the usual funding agencies. The need for experimentation and some of the difficulties inherent in experiments with marine fisheries have been described elsewhere (e.g., Larkin 1972; Pikitch 1988; McAllister and Peterman 1992; Policansky 1993a, 1993b). Indeed, the oft-repeated advice to use adaptive management (e.g., Walters 1986; NRC 1996a, 1996b) is advice to take an experimental approach, one in which the management regimes are designed to facilitate data collection and—even more important—hypothesis testing. Thus, the management plan evolves as hypotheses are either supported or rejected. Indeed, the discussion of rights-based allocation schemes above makes clear how important experimentation is, both in the natural and the social sciences.
In many cases, fishing has been going on for so long that experiments are difficult because the results have already occurred (e.g., Policansky 1993a, 1993b; Auster et al. 1996). But sometimes a long history of fishing can be an advantage when an unplanned perturbation occurs. For example, when World War II forced a cessation of fishing in the North Sea, the long data set allowed a careful analysis of its effects (Beverton and Holt 1957, Rijnsdorp 1992). Not surprisingly, there was a large increase in many fish populations.
Marine protected areas also can provide a great deal of information on the effects of fisheries, environmental fluctuations, and other factors on fishing if they are implemented adaptively (see, e.g., the study of Polovina and Haight [in press] described above). That requires that carefully designed monitoring programs accompany the implementation of protected areas. Similarly, the introduction of new fishery regulations, such as bycatch-reduction devices or turtle-excluder devices, provides opportunities for developing and testing ecological ideas as well as learning about the effects of fishing and the effectiveness of fishery management.
Deliberate experimentation with a public resource or profits is not lightly undertaken (Policansky 1993b) because there are risks involved. One is that adaptive management can take a very long time to mature into a successful program (Walters et al. 1993). Another is that populations or ecosystems could be adversely affected by experimental overfishing. But the committee presumes that the need for useful information in general outweighs the risks, although the latter must be carefully considered. For the Bering Sea, the NRC actually recommended deliberate overfishing in restricted areas as a way of gaining information about the ecosystem (NRC 1996a), and this committee agrees that useful information could be gained in that way if experimental fishing is carefully controlled.
Ecosystem-Based Approaches to Managing Fisheries
Fish populations are one portion of complex ecosystems that are affected by many natural and human-induced factors. Marine ecosystems are used by individuals with a range of perspectives and attitudes about them and about fisheries specifically. Other people rely on fish only as a food product and never visit the ocean. Some individuals depend on the ocean for their livelihood, via fisheries or some unrelated commercial activities. Still others value marine ecosystems and fish populations mostly from an aesthetic perspective. This variety of perspectives invariably produces conflicting goals regarding uses of marine ecosystems. It is one reason there have been so many calls in recent years for ecosystem-based approaches to resource management.
It is the perception of many observers that single-species fishery management has failed (Ludwig et al. 1993, Safina 1995) and that a new approach, which recognizes ecosystem values, is required to achieve sustainable fisheries. A move toward fishing and management that recognize the importance of species interactions, conserve biodiversity, and permit utilization only when the ecosystem or its productive potential is not damaged is a worthy objective. This chapter concludes with a brief discussion of just how such an approach can be applied to fishery management.
It is clear that not enough is known about most large marine ecosystems to implement a reliable whole-ecosystem approach to management. In any case, it is probably beyond our capabilities to manage some aspects of marine ecosystems, such as ENSO events and large-scale migrations. Not enough is known about trophic relationships, environmental variability, community interactions, migration patterns, and many other factors to "manage the ecosystem" as one might try to manage a terrestrial game reserve or a farm. What, then, does an ecosystem-based approach provide to fishery management that is lacking from a single-species approach or even from the multispecies approaches that have been discussed for decades?
The best answer at present is that it is uniquely useful in helping to set policy frameworks that include fish production and ecosystem goods and services; it acknowledges the critical role of ecosystem processes and a much broader focus than only the species of concern for fishing. It acknowledges that humans depend on these ecosystems for a suite of services as well as goods. An ecosystem approach includes a recognition that many segments of society have many goals and values with respect to marine ecosystems and that pursuit of any one goal is likely to affect how well other goals can be achieved. It does not provide an excuse for ignoring the biology and economics of individual species, industries, interest groups, and other segments of society. The committee concludes that an ecosystem approach, so defined, and adopted in accordance with the recommendations outlined in this report, will improve prospects for the sustainability of marine fisheries. The approach is described in more detail below.
A Suggestion for Action
An ecosystem approach to fishery management addresses human activities and environmental factors that affect an ecosystem, the response of the ecosystem, and the outcomes in terms of benefits and impacts on humans. The essence of the framework is characterized by stresses, responses, and benefits. The traditional view of a fishery narrowly fits into this framework with fishing as the only stress, the ecosystem response specified solely in terms of the effect of fishing mortality on a single species, and the outcome in terms of catch. One way of achieving an ecosystem approach is to incrementally add to the list of stresses, the scope of the ecosystem responses, and the type of benefits considered in fishery management.
Additional stressors might be forms of degradation in habitat and environmental quality. Experiments on the biological response of the resource (the exploited species) to these stresses would allow the stresses to be taken into account in population models that are traditionally used to determine the effect of fishing. In this way, several different forms of population stress can be compared quantitatively to the stress of fishing. The comparison would be helpful because much is known about how populations respond to fishing but less about population responses to other forms of stress. Even if managers decide not to experiment by deliberately stressing fisheries, they can instead take advantage of whole-system "experiments" (e.g., wars, major environmental fluctuations) and other management actions as described in the section on adaptive management above.
The type of responses to stresses can also be expanded incrementally by trophically linking species. Multispecies virtual-population analysis is an example of this approach. Ultimately, exploited species need to be linked to other components of ecosystems so that indirect responses to stress can be addressed.
Finally, there is a need for incrementally increasing the scope of benefits from fisheries. Benefits of recreational and subsistence fisheries need to be determined. Nonmonetary benefits of ecosystem services also need to be considered. Methods that express benefits in a common metric need to be applied and improved so that decisions can be made between alternative forms of management of fisheries and other human activities.
This incremental approach will take a long time to evolve to a point where it takes account of all the important stresses, responses, and benefits, but it allows immediate progress by taking advantage of the existing framework that is used in fisheries.
Elements of the Approach
Ecosystem Monitoring. An ecosystem approach to fishery management requires a long-term commitment to systematic and carefully standardized observations
of the state of ecosystem components,3 including measurements of stresses and benefits. Traditionally, only commercially exploited species and catches have been monitored. There are notable exceptions such as the California Cooperative Oceanographic Fisheries Investigation (CalCOFI) off California, Continuous Plankton Recorder surveys in the North Atlantic, and the National Oceanic and Atmospheric Administration's Marine Resources Monitoring, Assessment and Prediction program off the Northeastern United States. These programs monitor components of the plankton community and some environmental variables. An ecosystem approach will require expansion of these programs. As an example, consideration should be given to using bycatch data as a source of information about the ecosystem where practicable. Some degree of calibration could be provided by traditional sampling approaches.
Monitoring Human Systems. Much information is also needed on the behavior of people and their social, economic, and political institutions. As described above, information is needed on community-based management; matching institutional scales of complexity to biological ones; the responses of individuals and institutions to a variety of economic, environmental, and political factors; how and whether ITQ and related management regimes can be designed to realize the benefits of market-based approaches to exploitation and stewardship; and so on. Humans are part of the ecosystem and an ecosystem-based approach to management requires information on humans and their systems as well as on the other parts of marine ecosystems.
Application of Ecosystem Principles. While knowledge of how ecosystems function is still very incomplete, much is known about the structural characteristics of ecosystems, how structure relates to functioning, and how structure and functioning respond to various types of stress. For example, the role that top predators play in stabilizing ecosystems has been studied extensively. However incomplete our knowledge about ecosystem principles, what is known should be given formal consideration in fishery-management decisions.
Traditionally, fisheries have been managed by controls on the catch or the amount of fishing activity. There have also been controls on the time and place where fishing is allowed to protect certain components of the fishery resource (spawners) or to make it harder to catch fish so as to protect the resource from the fishery. These approaches have rarely addressed the broader ecological implications of fishing, and for this reason, new methods are needed. One new approach that is receiving increasing attention is marine protected areas, as discussed earlier.
Cross-Sectoral Institutional Arrangements. Marine ecosystems are affected by many human activities in addition to fishing. Traditionally, different institutions,
such as government agencies, have had responsibility for managing these activities. Institutional arrangements are needed that require decision makers to consider the effects of one sector's actions on other sectors, such as the effects of agriculture on water quality.
Large Marine Ecosystem Approach (LME). In the past decade there have been a series of meetings classifying coastal areas of the world's ocean into LMEs, and promoting an ecosystem approach to studying and managing these ecosystems. Recently, international donor agencies have shown interest in funding regional programs based on an LME approach. The approach identifies five modules that need to be addressed, including (1) productivity (i.e., the base of the food chain), (2) fishery resources, (3) ocean health (amount and quality of habitat), (4) socioeconomics, and (5) governance. Monitoring strategies for the first two modules are most developed, but the approach stresses the importance of all five modules in order to properly manage ecosystems. The LME approach incorporates several of the elements of an ecosystem approach to fishery management discussed in this report (Sherman et al. 1990, 1993).
The Precautionary Approach. Recently, this approach has gained acceptance for the management of fisheries, as indicated in the United Nations Agreement for Straddling Stocks and Highly Migratory Species. Perhaps it is even more applicable to an ecosystem approach to fishery management than it is to traditional single-species management, because of the level of uncertainty about ecosystems and the potential risks associated with their misuse. It seems unlikely that sustainability of marine fisheries will be achieved without a more pervasive and stronger commitment to the precautionary approach.