Marine conservation requires a new ecosystem-based concept for fisheries management that looks beyond sustainable yield for individual fish species
In the 1950s, Jamaica's coral reefs were thriving, with branching forms of staghorn and elkhorn corals ruling the shallows and plate-like corals dominating deeper waters. The reefs were popular habitats for large predatory fish such as sharks, snappers, groupers and jacks, and Jamaicans came to the reefs to trap these species.
The scene looked much the same in the 1970s—except that the large fish had disappeared onto dinner plates. Struggling to feed the island's swelling population, Jamaicans began using motorized canoes to improve their efficiency in trapping smaller, herbivorous fish, such as surgeonfish and parrotfish. The reefs still looked healthy, however, and sea urchins were doing well. As herbivorous fish were removed, the urchins no longer had to compete with them for the mainstay of their diet: algae.
By the 1990s, Jamaica's reefs were depleted of both carnivorous and herbivorous fish and smothered with fleshy algae, such as Sargassum. Sea urchins were uncommon, and the vast diversity of the reefs was reduced. Because the algae had taken over the real estate, there was little space for renewal by larval corals. Back in the late 1970s, surveys off the northern coastline had found corals over half the reefs' surface, with fleshy algae covering only 4 percent. By the early 1990s, Terry Hughes from James Cook University demonstrated that those figures were more than reversed: 3 percent of the space occupied by corals, 92 percent by algae. Thus reefs that had existed for thousands of years as one of the most diverse habitats on earth had changed into algal mats in a few decades.
What tipped the balance? Most scientists point to overfishing, although pollution and natural disasters certainly also took their toll. (See "Mud, Marine Snow and Coral Reefs," January–February.) When carnivorous fish became scarce, populations of one of their prey species, the sea urchin Diadema antillarum, exploded. These black, spiny creatures flourished even more with the demise of their competitors, the herbivorous fish. But in 1983, a mysterious pathogen sickened the urchins, reducing their number by 99 percent. With few herbivorous fish and now few urchins, the algae were free to grow unchecked.
In the past decade, Jamaican reefs have begun to revive—corals now cover about 10 percent of the surface. But it is unclear whether they will regain their dominance or whether the algae will continue to hold sway.
This example makes an important point: Harvesting a few types of fish affects not only those species but disturbs whole ecosystems, which in turn can reduce target species. We shall argue in this article that, although the primary goal of "sustainable fisheries" is to preserve the long-term viability of target species, even harvest levels considered sustainable can impact marine ecosystems. Protection of the world's oceans will in the future require a broader and more integrated scientific view than one that focuses on one or a few species.
A spectacular diversity of habitats lies in the world's oceans, which cover 361 million square kilometers and constitute more than 99 percent of the biosphere's volume. Some habitats, such as coral reefs, seagrass meadows and kelp forests, are relatively well known. Others, such as deep-sea plains and vast open waters, are alien to most people.
We are only beginning to appreciate the depth and breadth of marine diversity, but it surely rivals that on land. Coral reefs, for instance, occupy only 0.1 percent of the earth's surface, yet they may support as many as 9 million species. There also could be millions of species in the deep sea, once thought to be a vast desert. At higher taxonomic levels, marine systems are much more diverse than those on land. Of the 34 recognized animal phyla, 33 can be found in the ocean, and 15—including comb jellies, echinoderms and lampshells—are exclusively marine. Because certain marine groups, especially bottom-dwelling invertebrates, are poorly known, enormous diversity remains to be discovered. For instance, 20 new families, 100 new genera and 200 new species were recently found in or near hydrothermal vents.
Humans have exploited marine ecosystems for at least 10,000 years, but overharvesting was not a significant issue when human populations were small and fishing gear was mostly lines and hooks. Nonetheless, the first known anthropogenically caused collapse of a marine stock occurred about 3,000 years ago, when peoples on the Peruvian coast continued to harvest shellfish that had been depleted by a natural disturbance. With the technological advances of the Industrial Revolution, such as the internal combustion engine and refrigeration, the ability of humans to extract fish from the sea, particularly in offshore regions, increased dramatically. The resulting increased frequency of stock collapses led to the realization that the sea is not infinitely bountiful. More recent developments include factory trawlers, which harvest a disproportionate share of the world's fish.
Commercial fishing now directly kills enormous numbers of fish: More than 90 million metric tons are captured annually for consumption, and about 30 million metric tons are discarded as bycatch. Worldwide, 25 to 30 percent of exploited stocks experience overfishing, despite numerous incidents of stock collapse. Moreover, fishing trawls dragged across the seafloor have devastating effects on benthic communities, with impacts similar to clear-cutting in forests. By killing benthic invertebrates as well as key predators, fishing affects food webs and fundamentally alters community composition.
Why are many marine fisheries overexploited? Fisheries provide 19 percent of the animal protein consumed by human beings worldwide and generate jobs for more than 200 million people, mostly in small-scale fisheries. Understandably, this creates enormous social, economic and political pressure to maintain current fishing levels. As fisheries managers juggle diverse and conflicting objectives, political agendas and scientific uncertainties, they often err on the side of overexploitation, despite calls for caution from organizations such as the Food and Agriculture Organization of the United Nations. Such conflicts in fisheries management are typically characterized simply as science versus politics, but we suggest that there may be a deeper problem: The fundamental goals of modern fisheries management may be incompatible with sustaining marine ecosystems. Currently, a primary goal of "sustainable" fishing is to preserve the long-term viability of target species. But sustaining those species may depend also on protecting the vast biodiversity of the world's oceans, which by definition will require a more holistic approach to resource management.
The goal of long-term sustainable harvest has been a mainstay of fisheries science for the past half century. This concept was crystallized with the development of a model in 1954 by Milner Schaefer of the Scripps Institution of Oceanography that incorporates both fish-population dynamics and harvest. An important feature of this and other early models was the recognition that a population's size determines its growth rate. Thus the growth rate (overall numbers of new organisms produced per year) is low when a population is small. It is also low when a population nears its carrying capacity, because of density-dependent processes such as food availability (Figure 2). Intermediate-sized populations have the greatest growth capacity and ability to produce the most harvestable fish per year. The key realization of these early models was that fisheries could optimize harvest of a particular species by keeping the population at an intermediate level and harvesting the species at a rate equal to the annual growth rate. This strategy was called the maximum sustainable yield.
Although models are now able to capture more of the complexity of the dynamics of fish populations, two concepts remain integral parts of most management plans. The first, as noted above, is that average harvest rates should equal growth rates. The second is that harvests are sustainable even when fish populations fall well below unfished levels. The widely used term "surplus production" implies that populations produce biomass beyond that required to sustain them—and therefore that this surplus can be harvested without impacts. There is a growing sentiment, however, that we need to go beyond considering only target species in fisheries management. For example, the U.S. Congress, as part of the Sustainable Fisheries Act of 1996, directed the National Marine Fisheries Service to establish an Ecosystem Principles Advisory Board. An emerging question is: Do levels of exploitation consistent with sustaining marine fish populations have long-term, detrimental effects on ecosystems?
Fishing and Ecosystems
Ecosystems are characterized by their communities of plants, animals and microorganisms, and by the local physical, chemical and structural environments in which those communities reside. Fishing alters marine ecosystems both by modifying community composition and by altering local environments, as when trawls drag across the ocean floor. Because ecosystem components interact, the effects of fishing are more complex than the simple removal of a few species. Ecosystems can suddenly collapse when some of their components are damaged, just as houses can fall if their foundations rot away.
Marine community members interact in many ways: They feed on each other, compete for key resources and, as with kelps and corals, provide habitat structure. A species' importance in a community depends on its relative abundance and the strength of its interactions with other components. Some species have little impact on others, whereas other species greatly influence their fellow community members. A species' position in the food web determines its average trophic level. Primary producers are said to occupy the first trophic level because they derive energy from the sun or chemicals in seawater. Phytoplankton, seaweeds and sulfur bacteria are primary producers. Herbivores occupy the second trophic level because they eat primary producers. Carnivores occupy the third because they eat herbivores. The top species is called the apex predator.
Fishing can directly affect communities, as when it changes key life-history traits. Fisheries typically seek larger, more valuable fish, decreasing the average size of fish in target populations. Because reproductive ability relates to size in many species, this decrease has the potential to reduce the number of fecund fish, altering population dynamics. Moreover, large species often have long life cycles, and long-lived populations cannot rebound as quickly as can shorter-lived ones. Yet fishing has shifted many marine communities toward shorter-lived species.
Community structure is also altered by the practice of targeting organisms at lower and lower trophic levels as higher levels become depleted. Such a progression, called "fishing down food webs," is exemplified by the shift from harvesting large groundfish, such as Atlantic cod (Gadus morhua), in the Gulf of Maine, to the current reliance on herring, lobsters, sea urchins and shrimp. Work by Daniel Pauly from the University of British Columbia and others has demonstrated that the trophic level of fish landings worldwide has declined in recent decades, particularly in the Northern Hemisphere. (See "Fishing Down Aquatic Food Webs," January–February 2000.)
Many of fishing's effects on ecosystems are indirect. Trophic cascades, for example, are changes in biomass propagated across three or more trophic levels, as when depletion of a predator allows populations of prey to increase, which in turn suppresses the prey's own prey. If the cascade reaches primary producers, the ecosystem may become less productive. For example, the massive overfishing of coastal Atlantic cod and other predatory groundfish increased the abundance of the sea urchin Strongylocentrotus droebachiensis, which transformed forests of the kelp Laminaria, a primary producer and important habitat for young cod, to ecosystems dominated by coralline algae. Similarly, overhunting of sea otters in the Aleutian Islands allowed sea urchin populations to expand; the urchins then severely overgrazed kelp. Overharvesting played a key role in major shifts of entire communities in both these examples, as in the shift from corals to algae on Jamaica's reefs.
In most cases, we do not know how alterations in community composition affect ecosystem function, although they are generally believed to decrease stability. Because marine systems are immense, variable and affected by many abiotic and biological forces, they are inherently difficult to study. Qualities such as species diversity are easily measured but not explicitly linked to community function. More relevant qualities, such as stability (the ability of a community to maintain a state) and resilience (the ability of a community to return to a state after perturbation), are nearly impossible to assess in natural communities. Furthermore, communities are dynamic—some species are seasonal because they migrate, many populations naturally fluctuate in size, and all communities are at the mercy of natural disturbances, such as hurricanes. To compound these problems, we lack the comparisons in time or space that would allow us to quantify the effects of fishing or other disturbances. This lack of perspective severely limits our ability to predict how communities will respond to human pressures.
The Baltic Sea
We have tried to predict the effects of fishing on marine communities by deriving a model of an ecosystem and varying the levels at which different species are harvested. For this task, we used the computer software Ecopath (available at www.ecopath.org), which simulates food-web structure by modeling the production and transfer of biomass among species. Ecopath can be tailored to any food web and has been applied to more than 130 marine communities. It handles data on abundance, production and feeding rates, diets and harvests of key food-web components (primary producers, invertebrates, fish, marine fishing pressure on a particular species). In addition, we used Ecosim, a dynamic model that simulates how a community would likely respond to changes in fishing intensity and other forces. Ideally, field studies are used to validate model parameters, thus improving model performance and our understanding of community dynamics.
We applied Ecopath/Ecosim to the Baltic Sea, a northern European arm of the Atlantic Ocean. This marine system is ideal for this type of application because it is relatively simple and well-studied. Because fisheries are a mainstay of the region's commerce, there is a rich body of data on local fish species and the prey, predators and fisheries that interact with them.
The most visible components of the Baltic Sea food web (Figure 3)—fish and seals—are currently in undesirable states: Sprat (Sprattus sprattus), a fish of low commercial value, dominates the fish biomass; herring (Clupea harengus) and cod (Gadus morhua) are near all-time lows; and seals are slowly recovering from near-total collapse during the past century. These conditions have resulted largely from overharvesting. But how has overfishing affected the rest of the community?
We do not know how the Baltic Sea community was structured before fisheries developed, so we simulated an unfished state by eliminating all fishing from the model for 100 years. Compared with the year 2000, the 2100 community featured much larger biomasses of herring and cod at the expense of sprat (Figure 4a). These changes cascaded through the rest of the community. For example, herring's increased abundance reduced the level of one of its main prey, pelagic invertebrates such as small shrimp (Figure 4b), to half the 2000 level. Similarly, the recovery of cod cut down the biomass of benthic macrofauna, allowing one of their major prey, benthic meiofauna, to double in biomass by 2100. The overall growth in total fish biomass in the unfished scenario tremendously increased seal biomass (Figure 4c).
We also determined what the system might look like if fishing pressure were maintained at the status quo. After 100 years of status quo fishing, sprat continued to dominate the fish community, cod biomass was low, and herring were totally extirpated. The rest of the community was profoundly different from the unfished state: The abundance of benthic macrofauna was the same as at present, pelagic invertebrates were more abundant because they were not being eaten by herring, but seal recovery was only modest.
We introduced fishing into the unfished system by incrementally increasing fishing pressure on cod, herring and sprat. As we moved from no fishing to 20 percent of the status quo, the three species and the overall community responded most strongly. Seals also responded dramatically: This relatively small increase in fishing allowed human competition for fish to greatly impede seal recovery.
We also used the model to predict the consequences of a more precautionary fisheries management plan, which some Baltic fisheries scientists advocate. Their approach would cut fishing pressure on cod and herring by about 35 percent and 65 percent, respectively, from current levels and would slightly increase sprat fishing. It aims to bolster cod and herring populations by protecting them against overfishing. We detected a desirable response from the fish community: Biomasses of cod and herring clearly were higher in 2100 than in 2000 or status quo conditions. Unfortunately, the rest of the community retained the characteristics of a heavily fished system, falling somewhere between the 40 percent and 60 percent status quo scenarios. These results suggest that this type of fishery rehabilitation would do little to shift community structure toward the unfished state.
An additional simulation looked at the effects of fishing pressure on community trophic structure. As seen above, fishing directly and indirectly changes the relative abundances of species in a food web. While others have focused on the change in average trophic level of the harvested species, we calculated the average trophic level (weighted by the abundance of each community component) of the entire Baltic community across a range of fishing pressures. As pressure increased from zero to the status quo level, the average trophic level decreased, following a curve shaped like a decay function (Figure 5). Much of that decrease resulted from a shift of biomass from relatively large fish to a community increasingly dominated by phytoplankton and benthic macrofauna. The dominant fish species changed from a major predator, cod, to one of its prey, sprat, and the biomass of the overall community shifted closer to the primary consumer level (trophic level 2). Our model therefore supports the idea that fishing can change an ecosystem's trophic structure.
The Baltic Sea ecosystem is but one of hundreds of marine ecosystems that provide commercially useful fish and other resources. Taken together, these systems operate within the larger global system, which is closed and therefore sets important limits on exploitation rates. Because harvestable biomass depends on assimilated solar and chemical energy, as well as on individual animals capable of reproduction, the absolute limit on the sustainable rate of harvest is determined by the rate at which primary producers can assimilate the energy that supports higher trophic levels. If fisheries remove a substantial proportion of the energy acquired from primary producers, the system becomes potentially unstable.
Daniel Pauly and Villy Christensen estimated that 8 percent of the ocean's primary productivity is necessary to sustain global fisheries, which is four times greater than previous estimates. However, this figure was dominated by the open ocean, which represents most of the ocean's surface area but relatively little of its productivity. For commercially important regions, the primary production required to sustain fisheries was much greater: 25 percent for upwelling zones, 24 percent for tropical coastal shelves and an extraordinary 35 percent for temperate coastal shelves.
It is apparent that we are close to the limit on sustainable harvests and that simply improving fishing efficiency or management practices will not allow us to significantly increase global marine exploitation. In fact, since Pauly and Christensen published their study, global landings have leveled off (Figure 6), suggesting that we might already have reached some sort of threshold. Humans are now the apex predators in many marine communities, but, unlike other predators, we generally do not recycle energy back into the appropriate ecosystems. We also have little understanding of the long-term effects of removing huge amounts of energy and biomass from the sea.
Ecologically Sustainable Yield
We have seen that the traditional approach to fisheries management relies on single-species models of population dynamics that aim to sustain harvests of commercial species. Such an approach ignores a broad suite of interactions among exploited species and between exploited species and other members of their communities. Widespread recognition of this shortcoming has spawned considerable interest in ecosystem-based fisheries management, and scientists are beginning to identify suitable metrics for monitoring marine ecosystems. Ecosystem-based management requires a long-term commitment to monitor all trophic levels of marine organisms and the physical forces that influence their communities. The well-studied Baltic Sea community and a few others, such as the Bering Sea community, lend themselves to such analyses, but these studies must be extended to include all impacts of fishing, including habitat destruction and bycatch, which in some cases depletes more biomass than does harvest.
Achieving the goal of sustainable harvest of fish, however, does not mean that the impact on marine ecosystems of fisheries is eliminated. The Baltic Sea example shows us that careful assessment of fish stocks in concert with limited levels of fishing effort allow sustainable fisheries. At the same time, prosecution of the Baltic Sea fishery, even at a limited level, fundamentally alters community structure at all trophic levels. In our hypothetical scenarios, fishing above intermediate levels (approximately 40 percent of the status quo) elicited no appreciable positive response from the marine community. This inertia implies that even precautionary levels of harvest can diminish the ecological importance of target species by significantly weakening their linkages to other community members.
Thus, the goal of sustainable fisheries is simply different from, and in some cases may be incompatible with, the goal of maintaining natural marine communities. This should not come as a shock. Sustainable fishing will lower the size of fish populations well below the size of unfished populations. Thus we might expect community shifts in systems where fish strongly influence the population size of their prey or predators. We argue, however, that the cost of mismanaging a community might be far greater than the cost of mismanaging a fishery. Although overfished stocks have been known to recover, revival of communities that have changed states can be excruciatingly slow or even impossible. Such potential impacts are rarely considered in cost-benefit or risk analyses.
Fisheries agencies around the globe focus on the management of natural "resources." To the extent that ecosystems affect exploited resources, these agencies have embraced ecosystem research that allows better resource management. However, a shift in the species composition of benthic invertebrates or seaweeds has not historically been the concern of agencies unless such changes affected fish production. We suggest the time has come to consider the status of marine ecosystems, in addition to the status of fisheries.
One solution to the environmental crisis in our oceans might be to redefine sustainability. If we can shift our way of thinking about fisheries and ecosystems we may prevent the shifting of the ecosystems themselves. We propose a change from maximum sustainable yield to ecologically sustainable yield (ESY): the yield an ecosystem can sustain without shifting to an undesirable state.
To determine ESY, one will have to simultaneously consider the impacts of all harvested species on an ecosystem, quantifying important qualities such as community stability or resilience. This will be challenging because of the uncertainty and indeterminacy inherent in ecological systems and because ecosystems change in response to natural processes in ways that we do not fully understand. However, we can and should improve our understanding of the bounds of expected ecosystem behavior and define ESY within the limits of predictability. Maximum sustainable yield or other more conservative fisheries targets are critical tools in achieving the goal of sustainable fisheries. We submit that a concept such as ESY is a necessary instrument in achieving the objective of marine conservation.
Widespread implementation of ESY is obviously an issue of policy, rather than science. For many ecosystems, harvesting at ESY levels will require even stricter limits than those used to achieve maximum sustainable yields. But we believe that until we fully incorporate the notion of ecological sustainability into our philosophy of fisheries management, we will continue to degrade the communities that nurture commercially valuable fish.
Rick Methot, Church Grimes, Marc Mangel, Paul Dayton, Mike Schiewe, John Williams, Jon Drake and Doug Dey provided stimulating discussions that greatly improved this article. The views expressed in this article are those of the authors and not necessarily of the National Marine Fisheries Service.
Note: This article resides in the public domain; its use is not restricted by copyright.
Botsford, L. W., J. C. Castilla and C. H. Peterson. 1997. The management of fisheries and marine ecosystems. Science 277: 509–515.
Estes, J. A., and J. F. Palmisano. 1974. Sea otters: Their role in structuring nearshore communities. Science 185:1058–1060.
FAO (Food and Agriculture Organization of the United Nations). 2000. The State of World Fisheries and Aquaculture, 2000. Rome: FAO.
Grassle, J. F., and N. J. Maciolek. 1992. Deep-sea richness: Regional and local diversity estimates from quantitative bottom samples. The American Naturalist 139:313–3412.
Harvey, C. J., S. P. Cox, T. E. Essington, S. Hansson and J. F. Kitchell. In press. An ecosystem model of food web and fisheries interactions in the Baltic Sea. ICES Journal of Marine Science.
Hixon M. A., P. D. Boersman, M. L. Hunter, F. Micheli, E. A. Norse, H. P. Possingham and P. V. R. Snelgrove. 2001. Oceans at risk. Research priorities in marine conservation biology. In Conservation Biology: Research Priorities for the Next Decade, ed. M. E. Soule and G. H. Orians. Washington, D.C.: Islands Press, pp. 124–154.
Hughes, T. P. 1994. Catastrophes, phase shifts, and large scale degradation of a Caribbean coral reef. Science 265:1547–1551.
International Council for the Exploration of the Sea. 2001. Report of the Study Group on Multispecies Predictions in the Baltic. ICES CM 2001/H:04.
Jackson, J. B. C., et al. 2001. Historical overfishing and the recent collapse of coastal ecosystems. Science 293:629–637.
Jennings, S., M. J. Kaiser and J. D. Reynolds. 2001. Marine Fisheries Ecology. Oxford, U.K.: Blackwell Science.
Link, J. S., J. K. T. Bodziak, S. F. Edwards, W. J. Overholtz, D. Mountain, J. W. Jossi, T. D. Smith and M. J. Fogarty. 2002. Marine ecosystem assessment in a fisheries management context. Canadian Journal of Fisheries and Aquatic Sciences 59:1429–1440.
McGoodwin, J. R. 1990. Crisis in the World’s Fisheries—People, Problems, and Politics. Stanford, Calif.: Stanford University Press.
National Research Council. 1999. Sustaining Marine Fisheries. Washington, D.C.: National Academy Press.
Pauly, D., and V. Christensen. 1995. Primary production required to sustain global fisheries. Nature 374:255–257.
Pauly, D., V. Christensen, J. Dalsgaard, R. Froese and F. Torres, Jr. 1998. Fishing down food webs. Science 279:860–863.
Reaka-Kudla, M. L. 1997. The global biodiversity of coral reefs: A comparison with rain forests. In Biodiversity II: Understanding and Protecting our Biological Resources, ed. M. L. Reaka-Kudla, D. E. Wilson and E. O. Wilson. Washington, D.C.: Joseph Henry Press.
Schaefer, M. B. 1954. Some aspects of the dynamics of populations important to the management of the commercial marine fisheries. Bulletin of the Inter-American Tropical Tuna Commission 1:25–56.
Walters, C. J., V. Christensen, D. Pauly and J. F. Kitchell. 2000. Representing density dependent consequences of life history strategies in aquatic ecosystems: Ecosim II. Ecosystems 3:70–83.