Understanding Human Modifications of the Ocean: Opportunities for Large Scale Experiments

James F. Kitchell, Center for Limnology, University of Wisconsin, Madison, Wisconsin

Michael L. Pace, Institute of Ecosystem Studies, Millbrook, New York

Until recently, our vast oceans were viewed as wilderness - often visited by humans but little affected by them. That is no longer the case. The development and spread of the nineteenth century whaling industry serves as one marker of the beginning of massive human impacts on marine resources at a global scale. When the Challenger set sail, fishing and whaling were well established in all of the oceans. Now the impacts are multiple and include fisheries, eutrophication, introductions of alien species, coastal habitat destruction, and alterations of biogeochemical cycles. We increasingly view these human modifications of the sea not just as local concerns but in a regional, ocean basin and even global context.

The greatest long-term potential human modification of the ocean is climate change driven by fossil fuel combustion and the build-up of greenhouse gases in the atmosphere. Changes in the ocean-atmosphere interaction have the potential to massively alter climate. Human enterprise is currently so dependent on energy sources that produce greenhouse gases that many climate modelers are now moving beyond previous worse-case double CO2 scenarios to think about a triple CO2 world some time in the next century.

How do we understand these large, multiple and pervasive human impacts on the sea? Here we advocate that this is the greatest challenge to oceanographic science. This challenge calls for experiments appropriate to the scale of the problems and it calls for a change in mind-set. We provide examples of successful large scale experiments from limnology and argue for an accelerated use of this approach in oceanography. We illustrate the issues by examples of opportunities arising from human-driven changes of marine fisheries.

Human Impacts on the Ocean

We cannot present a comprehensive overview of the major human impacts on the global ocean. Instead, we provide a few examples of sufficient portent to portray the problem. First, open ocean environments typically have vanishingly low concentrations of trace metals such as iron, copper, and zinc. Yet, many of these elements are vital to life. Oceanic microbes are uniquely adapted to acquire trace elements at low,in situ, concentrations and find higher concentrations toxic, such as those now characteristic of many estuarine and nearshore environments. However, contemporary concentrations of some trace metals such as lead and mercury are almost completely the result of human activity. This loading of lead and mercury to the sea is trace pollution in an environment of unique sensitivity.

A second example is the export of nitrate by large rivers that empty to the sea. A sample of over 40 major rivers that in aggregate account for 37% of the freshwater input to the ocean reveals the major correlate of river nitrate is human population density in the watershed. This is in spite of vast differences among the rivers in geological, biogeochemical, and ecological processes. The sources of river nitrate vary according to population density but result primarily from atmospheric deposition, sewage and fertilizer.

Finally, the harvest of marine species for food represents a pervasive, large-scale impact of humans. During the past decade, annual global fisheries peaked at about 100 million mT. As a result, the majority of the world's fish stocks have been or are being exploited in excess of sustainable levels. Growth in world fish production owes mostly to the increase in aquaculture which produces 30% of the current total. Capture fisheries for small pelagic fishes (anchovies, anchovettas, herrings, sardines, etc.) yield another 33% of the global total and nearly all of that is rendered to fishmeal to feed pigs, chickens, cattle and aquacultured fishes and shrimps. These small pelagic fishes are the primary prey of many piscivores and their exploitation is now viewed as an intense, continuing depletion of forage fishes that has effects in many marine ecosystems. Of the current world catch, approximately 25% is killed and discarded as waste. Included in that total is an array of seabirds, sea turtles and marine mammals.

Commercial fishing proceeds in virtually all ice-free marine habitats. Thus, on a global basis, fishing should be viewed as something analogous to agriculture or forest harvest in terrestrial systems--a human-induced and continuing change in the structure and function of ecosystems. Like agriculture and forestry, the impacts of fishery exploitation expand in concert with human population growth and the advent of new technologies.

These examples illustrate the scope of the problem but hardly deal with the consequences. In the case of nitrate loading to the sea the consequence is clear - eutrophication which leads to massive algal blooms, deep water anoxia, and declining habitat quality for vital resource species. Such effects are well documented for the Gulf of Mexico near the outflow of the Mississippi River. Trace metal pollution - as noted for oceanic mercury and lead, is a more subtle and difficult question where impacts at present may be few but increases in these constituents should raise significant concerns. Global fisheries harvest has widespread ramifications for the alterations of marine ecosystems as further developed below.

Need for Large Scale Experiments

Oceanography is well adapted to the study of human alterations because of long emphasis on understanding oceanic regions as integrated systems. There has already been considerable effort on large-scale measurement programs and integration of data via models. This has been a primary avenue for global change research (e.g. the Joint Global Ocean Flux Study - JGOFS).

We suggest, however, this method of research is incomplete. Some human driven alterations are so rapid that we cannot hope that extensive measurement programs or models will allow suitable predictions either in the short or long term. Experiments taking advantage of either intentional or accidental manipulations represent an important, but underemphasized approach that must be increasingly used. In limnology, long and bitter debates characterized research in eutrophication and acidification until simple, powerful results of whole-lake manipulations conducted at Canada's Experimental Lakes Area made it clear that phosphorus was the key for controlling primary productivity of lakes and that acidification effects acted on fish populations through their indirect effects on important prey species. We hope that oceanographers will learn from those experiences - microcosms, mesocosms, field survey and process studies yield interesting results but compelling lessons are also derived from bold, large-scale experiments.

One of the most exciting recent advances in oceanography has come from such experiments. Here, the work was driven not by a human perturbation but by the basic question of what controls primary productivity in high nitrogen, low chlorophyll regions of the ocean. The argument for iron limitation was strongly substantiated by purposeful, experimental additions of iron to the open sea. These iron addition experiments are not the first, but certainly are among the most visible system-level experiments ever performed in oceanography.

The success of the iron experiments might be a harbinger of a new era in oceanography. This potential, however, requires broader recognition and support for large-scale experiments. Anderson (1997, Nature 388: 513) has written that a change in mind-set is warranted in dealing with the complex problems of harmful algal blooms. He notes the traditional approach of better understanding of the organisms and their ecology is insufficient. Instead, he argues for bold efforts to attempt to control actual blooms. The ensuing failures and successes will provide a form of learning that cannot be achieved in culture flasks. The analogue for this type of research is agricultural pest control where candidate methodologies must ultimately be field tested.

We see a fruitful path where large scale experiments provide a platform for both major scientific advances as well as potential understanding of how to control human effects on marine ecosystems. In calling for `large scale experiments' we are not referring to extensive measurement programs often dubbed `experiments' in oceanographic research. Instead, we mean either taking advantage of human-induced manipulations, or as in the case of the iron addition experiments, purposefully altering marine ecosystems. The unique feature of these experiments is that they are conducted over large scales of space and time. These scales are particularly relevant to management and to enhancing our understanding of ecological systems.

We next develop this perspective with examples from own experience with lake-ecosystem experiments and review analogous cases for the world's largest lakes, enclosed seas and extensive oceanic regions.

Keystone Predators and Trophic Cascades

We adopt the ecological framework offered by the keystone predator concept and the idea of a cascade of effects that stem from manipulating apex predators. The marine foundation for these principles derive from studies of the rocky intertidal and from the role of sea otters in maintaining sub-tidal kelp forests. An analogous idea developed among limnologists based on selective predation by fishes and among terrestrial ecologists who considered the role of predators in controlling herbivory. The basic expectation is that reciprocal changes occur in adjacent trophic levels. Removal of a major predator in a food web reduces mortality on prey populations at the next level and alters the competitive interactions at both trophic levels. Community composition changes radically in response (keystone predator effects) and ecosystem processes are potentially altered at all trophic levels (trophic cascade effects). Surviving predators experience lessened competition, grow more rapidly in response to increased prey resources, and that can alter their life history attributes.

The trophic cascade and keystone predator hypotheses were tested in an ecosystem context by manipulating food webs in small experimental lakes and in a larger, more complex context where apex predator populations were enhanced and exploited. An extensive literature has developed around these ideas and their application in lakes. While the freshwater literature may not be familiar to some marine scientists, space limitations allow only a brief overview. The main results follow. Adding piscivores to a planktivore-dominated lake caused extensive changes in the zooplankton community, increased herbivory rates, and reduced primary production. Reversing the manipulation (removing the piscivore) reversed the effects. Simultaneous manipulations of food web structure and nutrient loading illustrated that food webs dominated by piscivores with large-bodied zooplankton can effectively limit phytoplankton biomass even at very high nutrient loads. In contrast, lakes with plantivores and small zooplankton sustain high algal biomass and exhibit dramatic variability - in short these lakes have all the hallmarks of eutrophication. Recent results of this work also demonstrate that the interactions of nutrients and food web structure can alter the net direction of carbon flux between lakes and atmosphere.

Lake water quality managers call this approach "biomanipulation", it is among the tools of their trade and its successful application requires attention to site specific conditions. Responses are often non-linear because animals engage in predator avoidance behaviors. Removing piscivores is a goal of many fishers, so the success of biomanipulation efforts often traces to effective management of fishery exploitation.

Trophic cascade effects are demonstrated in the largest of temperate zone lakes (e.g., Lake Michigan, Lake Ontario). Keystone predator effects are demonstrated in the largest of tropical lakes (Lake Victoria). Although untested, the powerful advective effects in large marine ecosystems would probably diminish transmission of a trophic cascade to all levels of pelagic food webs in much the same way that modest manipulations are attenuated in freshwater systems. However, there is strong reason to believe that interactions stemming from apex predators would be substantial and discernible at the upper levels of pelagic food webs. For example, a recent National Research Council panel review of food web interactions in the Bering Sea concludes that trophic cascades there can be traced to the exploitation of whales. Reductions of whales allowed huge increases in prey resources such as krill, which fueled an explosive population response by walleye pollock and, through competition-predation interactions, depressed the local herring and capelin populations. The latter are key, energy-rich prey for juvenile fur seals and sea lions; their survival and numbers declined in concert with declines of their preferred prey resources.

The Black Sea offers another lesson worth repeating. Intensive fishery exploitation and habitat degradation due to eutrophication contributed to a collapse of Black Sea fish populations. A predaceous jellyfish replaced them and that, in turn, was replaced by an exotic ctenophore. Both jellyfish and ctenophores prey on zooplankton, fish eggs and fish larvae thereby constraining the recovery of fish populations. At one point the biomass of ctenophores in the Black Sea was estimated at 750 million mT which is about seven times the current global fisheries harvest. The fisheries of the Black Sea now yield a small fraction of their former state. Like the experimental lakes where food web manipulations revealed mechanisms operating at the larger scale, the Black Sea evidences the possible ecological consequences of intensive overexploitation and habitat degradation in a marine ecosystem.

A recent result from the North Pacific demonstrates that trophic cascades are also evident in the open ocean. During the past two decades, the Salmon Enhancement Program has helped double the abundance of salmon in the north Pacific. Pink salmon, a planktivore, have strongly increased and exhibit biennial population cycles. Japanese oceanographers recently reported a reciprocal interaction at three trophic levels in the central subarctic north Pacific. In odd-numbered years when pink salmon are most abundant (10X more than even-numbered years), summer zooplankton biomass is depressed and chlorophyll increased, both by several-fold. The reverse is evident in even-numbered years. Trophic cascades are not unique to the simple food webs of lakes.

Note that the marine cases above have not been treated as experiments, but they could have been. The Salmon Enhancement Program is clearly a management experiment while the ctenophore invasion of the Black Sea is a natural experiment. Both are subject to rigorous analysis of the patterns and potential causes of change if measurements are carried out before and after manipulations. Analysis of such experiments is also strengthened if it is possible to observe additional sites where either the manipulation did not occur or occurred with a different strength. Methods for analyzing large-scale (usually unreplicated) experiments are rapidly developing. A review of these methods is beyond the scope of this paper, but we note the need to integrate the philosophy and methodology of these experiments into oceanographic education.

Fisheries - Concepts and Approaches

We can extend the trophic cascade and keystone predator concepts to fisheries. Here, the apical predator is humans. The change in behavior of this predator offers vivid opportunities for experimentation. These opportunities require that we evaluate current concepts and approaches.

Much of fisheries science is based on the assumption that fish populations are most strongly regulated by single species population dynamics such as the stock-recruit relationship. While that approach continues to be the primary basis of research pursuits and management actions, its shortcomings are well-known and widespread. Those frustrations and failures (especially the population collapses) have led to many and varied pleas for alternatives.

One alternative appears in the calls for ecosystem management, i.e., an attempt to set the population(s) of concern in a larger context. A number of whole-ecosystem efforts have been launched in that pursuit. While these efforts are encouraging and informative, they are not yet practical. In short, the single species approach is tractable but insufficient and the ecosystem approach is sufficient but not tractable. The urgency of fisheries management and conservation issues has produced a compromise view--that an alternative ecological context must be adopted, that a predator-prey approach will be most useful, and that the framework of differences in life history characteristics set in a food web context can give greatest, immediate insights.

Discerning the causes of ecological change is a serious challenge. In coastal areas, local eutrophication may alter basic productivity. On the continental shelf, changes in many habitats have occurred as a consequence of repeatedly dragging trawls across the substrate. Extensive recent reviews document severe impacts (e.g., Georges Banks, the North Sea and Gulf of Mexico). Many fisheries for apex predators occur in the same waters as those that exploit the primary prey species (e.g., herring, anchovies, sardines, shrimp) and many have altered the prey community through the mortality due to bycatch and discards of non-target species. Many fisheries have operated for extended periods, have already evoked major changes in the local communities, and, because of their long history, offer only sparse documentation of the catches in the early stages of "fishing down". In other words, effects would be difficult to see because the major changes have already occurred before adequate recording was developed (e.g., Georges Banks). Some ecosystems exhibit dramatic environmental variability (e.g., upwellings), and others operate in environments (e.g., the Bering Sea) where shifts in climatic regimes have strong, masking effects.

Appropriate scales for management and experimentatal analyses are often defined by life histories of key predators in the system. For example, some reef fishes may carry out their entire life within a few hundred meter area while many tunas operate at the scale of entire oceans. Albacore in the North Pacific spawn south of Japan, then migrate to the California coast. Northern bluefin do much the same and they respond to prey availability at the largest scale; juveniles leave the western Pacific earlier when prey fishes (sardines) are less abundant near Japan. Both tunas feed and grow in transit and they make that trip several times during their adult life. Yellowfin and bigeye tunas migrate in the orthogonal direction; they spawn in equatorial waters, then move north to feed and grow. These migrations are thought to be a consequence of selection for placing small, vulnerable offspring in an environment of lesser predation risk.

Apex predators represent a wide range of life history strategies. Some, such as mahimahi (dolphin fish or dorado) grow very rapidly, mature early, have high fecundities, spawn often and have short life cycles. Others, such as swordfish also grow rapidly, but mature later, have high fecundities, and can accomplish large size during a longer lifetime. In colder waters, growth rates and maturity schedules are slowed but fishes such as cod exhibit the high fecundities indicative of species that live in environments where stochastic effects play a large role in early life history survivorship. In contrast, most sharks grow slowly, delay maturation, have very low fecundities, and typically have long lifespans. Their populations are particularly susceptible to exploitation effects.

To summarize, fish populations often operate over extensive areas. Their life histories set important time constraints for assessing responses. Human harvesting manipulates target populations with effects that ramify through predator-prey interactions. These trophic relationships occur in variable marine systems. Experimentation at this scale must, therefore, produce strong manipulations and will of necessity rely on multiple lines of indirect evidence.

Longline Fisheries in the Pacific: An Example of Opportunity

We use the development of longline fisheries in the Pacific as an example of how fisheries are an agent of ecological change and how such changes should be exploited in an experimental framework. The distant-water longline fishery operates by setting 800-1200 large, baited hooks hanging at intervals along a 40-60 km line suspended between several hundred drift floats - the equivalent of a sampling transect across 50 km of pelagic habitat. Developing after WW II, this fishery tripled in the 1960s over the levels reported in the mid-1950's. By the mid-70's, there was intense fishing in all international waters of the Pacific between 40N and 40S. The effect of this expansion is the equivalent of a huge sustained manipulation applied through time. The record of catch rates for blue marlin indicate that effect; catch rates have dropped continuously for a 35-year period and are currently about 20% of those recorded as the fishery began to develop. By contrast, the catch rates for mahimahi and skipjack tuna, both short-lived predators, showed only modest changes over this period.

Another large-scale change has occurred in this region. During the period of 1970-92, the drift-net fishery for squid expanded in the northern waters of the Central North Pacific (CNP) and the Transition Zone between the CNP and subarctic water. At its peak, this fishery was based on >50,000 km of drifting gill net and harvesting up to 300,000 mT of squids per year. The primary target species was the neon flying squid, Ommastrephes bartrami, which is among the preferred prey of large tunas, sharks and billfishes. Bycatch of marine mammals, seabirds and large fishes created growing international objections. The United Nation passed resolutions in opposition to this fishing method, and the fishery precipitously ceased operating in 1992. A jig-boat fishery continues, but catches smaller squid and at substantially lower total rates (about 25% of the previous exploitation levels). Thus, a sudden perturbation has allowed recovery of a major prey resource for apex predators that forage in the Transition Zone and to the north.

In summary, the long-term record of continuous exploitation in the longline fishery has reduced some apex predator populations to a small percentage of their former abundance. Those species with r-selected life histories have been less affected. Cessation of the Pacific-wide drift-net fishery has increased abundances of key prey--large squids--by reducing their exploitation to one-fourth of that prior to 1992. Thus, the relative densities of predators have been reduced and their prey may have suddenly increased. These changes are the focus of this vignette, because they maximize the potential for measuring compensatory responses in predator-prey interactions, predator growth rates and life history attributes.

The catch from a wide-ranging fleet of longline fisheries provides the opportunity for sampling integrated ecological information. The fishes sampled represent the equivalent of the Eulerian series--a simultaneous sample across species in the same space. Research efforts based on archival tagging of highly migratory fishes offer another important potential. These tags contain recording electronics that provide a continuous record of water temperature, latitude and longitude. The archival tag program presents the equivalent of a Lagrangian series--fishes are sampled through time and across space.

An ecologically complete view would include observations before exploitation began and would include members of a trophic guild that might experience some benefit as a consequence of the increased prey resources made available when target species came under exploitation. If this occurs, compensation might be broadly allocated among several species. The preferred approach should be based on evaluation for all members of an interacting trophic unit, both predators and prey.

The "complete" view, however, is not possible for most fisheries. Absolute, fishery-independent measures of predator and prey abundances or the dynamics of species interactions are simply impossible at the temporal and spatial scales pertinent to highly migratory apex predators. No research or resource management agency budgets are currently sufficient or will become so in the foreseeable future. Yet, the questions are vitally important. Indirect measures of density-dependent effects offer the tractable alternative. Using data collected from the fishery and interpreting that from the basis of ecological theory offers the benefit of an extensive sampling effort. One important research goal would be to validate and quantify the uses of such indirect measures in making ecological assessments such as for the case of the longline fishery. Another would be to recognize that fisheries should not be viewed as distinct and different from oceanography. They are part of the same system and that system is being intensively manipulated.

Another Example of Opportunity - Shrimp Fisheries and BRDs

During the past four decades annual yields of shrimp from the Gulf of Mexico and southeast Atlantic coasts of the U.S. have sustained at about one million mT. Bycatch in this trawl fishery includes a diverse array of fishes and invertebrates. Most of that catch is discarded at sea and much of that is dead or dying when released. Bycatch of other species has shifted from a ratio of 15 shrimp per 1 ancillary species to 4:1. In other words, intense trawling has reduced non-target demersal and benthic species to about 25% of their former abundance. There are virtually no assessments of the ecological implications of these changes. Further, there will soon be changes in bycatch pressure as states (e.g. Florida) are moving toward requiring Bycatch Reduction Devices (BRDs which are the equivalent of Turtle Excluder Devices) that will allow a large proportion of the juvenile fishes and larger invertebrates to escape the nets. There is a remarkable opportunity here. Benthic and demersal communities should begin to recover from the effects of trawling in areas where BRDs are required. Marine ecologists should be among the first to document the effects of this large scale change due to a release from a long term manipulation (decades of trawling). The sequential development of regulations by states may also offer design possibilities as the manipulations may not be imposed simultaneously. Thus, powerful methods of comparison between manipulated and reference systems can be brought to bear.

Conservation Issues - More Opportunities for Experiments

Large billfishes, tunas, and sharks are now among the charismatic megafauna receiving concerted attention from conservationists. From many quarters and in the media we hear statements such as: "...taking the top off the food web will have big effects on the ecosystem." That's a logical argument but there is simply no objective evidence to support or reject it.

Conservation concerns for apex predators in the pelagic realm are constrained because many of these fishes ignore national jurisdictional boundaries. Our Endangered Species Act and the Marine Mammal Protection Act are limited to U.S. territorial waters. The world's most aggressive fishing nation, Japan, has no legal analogues; nor do Korea, Taiwan, Spain, or Russia. Instead, regulatory processes must emerge from the persuasion of international politics and the handful of treaty organizations intended to create consensus.

Many views are raised in the international forum. For example, the National Audubon Society took a very aggressive and active role in the recent review of Atlantic bluefin tuna management. Their goal was simple, to protect the populations of these large, "noble" fishes from over-exploitation. Similarly, assertive voices are raised over the conventional practices of billfish tournaments and now encourage catch and release policies. Shark finning and shark mortality due to bycatch are also major conservation issues of the day.

The ecological context for these conservation arguments is presented in a document representing the consensus among 42 eminent scientists. Mangel et al. (1996 Ecological Applications 6: 338) assert that humanity must: "...fundamentally change the way it interacts with the ecological systems that directly and indirectly support it." Their seven principles for conservation offer guidance to that goal and call for an ecosystem approach instead of a species-level approach. Given the levels of uncertainty that must be accepted, Mangel et al. call for a system of reserves where research and adaptive management approaches can be pursued. In short, conservation biologists agree that species must be set in an ecosystem context, that uncertainty is a major concern, and that reserves might help by preserving options and providing insights. These recommendation if they are put into practice represent another set of potential experiments.

Marine reserves or refuges have been developed in many areas and are the focus of growing encouragement. They produce striking results of ecological contrast and strong evidence of the power of exploitation. As expected, protected areas generally harbor greater species diversity, allow increased population densities of many species and produce larger, older individuals. Their reproductive output enhances populations in adjacent areas and, in some cases, at substantial distances from the source. The effects of protection foster ecologically complex networks of indirect effects, a variety of feedback mechanisms and a suite of lag effects that make them both resistant and resilient to perturbation. Re-application of the exploitation manipulation can rapidly reverse those benefits. Taken in aggregate, these reserves offer powerful evidence of the ecological potency of fishing. Unfortunately, the current suite of refuges are almost exclusively confined to reefs and rocky shorelines or bays. At this writing, there are no refuges or reserves specifically designed to protect large pelagic predators. There is both a need and opportunity here.

Summary

Oceangraphy is a powerful and well-established environmental science. In the face of the growing human impact on the global environment, the discipline must lead in understanding causes and consequences of human-driven change. This will require science that begins to routinely exploit large experiments for learning. The potential for these experiments is widely evident in fisheries where human activity creates powerful manipulations. This essay, however, could have been written based entirely on examples drawn from the literature on eutrophication or species invasions. In other words, these opportunities are extensive. Oceanographers - by their very training, are well equipped to take the broad view and to overcome measurement limitations with powerful technical and intellectual innovation. A wise funding agency will facilitate this work and recognize that these efforts may well evolve from new alignments among non-governmental organizations, management agencies, and scientists. In our view, those alignments will be most advantageous to our science and to the society that sponsors it if they can adopt an approach based on bold, large-scale experimental manipulations.

General References

Botsford, L.W., J.C. Castilla, C.H. Peterson. 1997. The management of fisheries and marine ecosystems. Science 277: 509-515.

Carpenter, S.R. and J.F. Kitchell (eds.). 1993. The trophic cascade in lakes. Cambridge University Press.

Dayton, P.K., S.F. Thrush, M.T. Agrady, and R.J. Hofman. 1995. Environmental effects of marine fishing. Aquatic Conservation: Marine and Freshwater Ecosystems 5: 205-232.

Myers, R.A., J.A. Hutchings, and N.J. Barrowman. 1997. Why do fish stocks collapse? The example of cod in Atlantic Canada. Ecological Applications 7: 91-106.

National Research Council 1996. The Bering Sea ecosystem. National Academy Press.

Pace, M.L, and P.M. Groffman (eds.). 1998. Successes, limitations, and frontiers in ecosystem science. Springer-Verlag (in press).

Power, M.E., D. Tilman, J.A. Estes, B.A Menge, W.J. Bond, L.S. Mill, G. Daily, J.C. Castilla, J. Lubchenco, and R.T. Paine. 1996. Challenges in the quest for keystones. Biosciences 46: 609-620.

Roberts, C.M. 1997. Ecological advice for the global fisheries crisis. Trends in Ecology and Evolution 12: 35-38.

Rowley, R.J. 1994. Marine reserves in fisheries management. Aquatic Conservation: Marine and Freshwater Ecosystems 4: 233-254.

Walters, C.J., V. Christensen, and D. Pauley. 1997. Structuring dynamic models of exploited ecosystems from trophic mass-balance assessments. Reviews in Fish Biology and Fisheries 7: 1-34.

Return to OEUVRE Topics Page

National Science Foundation