Ecosystem Data to Guide Hard Choices

Although the native fish of Lake Victoria in Africa have long supported a productive local fishery, several other species of fish were introduced during the 1900s in an attempt to increase production. Those introductions succeeded beyond expectations: The harvest grew dramatically and one of the nonnative species, Nile perch, now accounts for 80 percent of the catch. But was this really a success? Taking into account the side effects of these introductions on other features of the ecosystem, the unforeseen costs may outweigh the benefits. To begin with, the introduced species contributed to the devastation of the native fauna. More than half of the lake’s 350 species of cichlid fish (80 percent of which are found nowhere else in the world) are now either extinct or have been reduced to populations that are only a fraction of their original size. In addition, pressure on the limited forest resources around the lake grew because fuelwood was needed to dry the oily Nile perch for transport and sale. That forest loss, combined with other land use changes, increased water pollution and lake siltation. This in turn led to an increased frequency and extent of eutrophic and anoxic conditions, which placed still more pressure on native species and may ultimately threaten the long-term productivity of the fishery.

Examples abound of vast and uncontrolled ecosystem “experiments” such as this, where people have altered ecosystems to meet one need, only to encounter an array of unforeseen side effects. The expansion of agricultural land into natural habitats increased food production but changed the quantity and quality of freshwater runoff as well. Fertilization increased crop yield but also caused eutrophication of nearby rivers and estuaries and is responsible for anoxic “dead zones” found in coastal areas near major agricultural river basins. Timber harvest and the transformation of forest land to agriculture helped to meet needs for food and fiber but also released carbon into the atmosphere and changed Earth’s albedo (surface reflectivity), contributing to the risk of global climate change.

The case of Lake Victoria thus provides a dramatic example of a more widespread problem: Historically, when we have modified ecosystems through an act as simple as adding a new species or through more sweeping changes in land cover or resource use, we have not had the ability to understand or forecast the complex ways in which our actions might affect other ecosystem goods and services. As a consequence, our management of ecosystems has not produced the net benefits that could have been achieved and in all too many cases has needlessly degraded or destroyed valuable resources. Our growing scientific capabilities and the growing human needs for these goods and services suggest that it is time to confront this problem.

Our historical approach to managing ecosystems can be characterized as sectoral and reactive. Sectoral approaches–focused on single objectives such as food production or timber supply–made sense when tradeoffs among goods and services were modest or unimportant. They are insufficient today, when ecosystem management must meet conflicting goals and take into account the linkages among environmental problems. Reactive management was inevitable when ecological knowledge was insufficient to allow more reliable predictions. Today, given the pace of global change, the escalating demand for resources associated with growth in population and consumption, and the significant social and economic costs associated with unwise resource management decisions, human welfare is utterly dependent on forward-looking and integrated management decisions.

The challenge of effectively managing Earth’s ecosystems and the consequences of failure will increase significantly during the 21st century. A “step change” is needed in the amount of information on ecosystem goods and services that is available to meet decisionmakers’ needs and, in the technical capacity, for effective ecosystem management. To help meet these needs, a partnership of scientists, United Nations (UN) agencies, international conventions, governments, the private sector, and nongovernmental organizations is attempting to launch an unprecedented worldwide initiative to mobilize scientific knowledge pertaining to ecosystems. This initiative, known as the Millennium Ecosystem Assessment, is similar in some respects to the Intergovernmental Panel on Climate Change (IPCC), but with a focus on biological systems rather than climate systems. It would capitalize on a unique convergence of data availability, scientific advances, and policymaker demand and, if successful, could dramatically accelerate the pace at which integrated and forward-looking ecosystem management approaches are adopted around the world.

Ecosystem tradeoffs

Humans have profoundly changed the world’s ecosystems. Some 40 to 50 percent of land has been transformed (through change in land cover) or degraded by human actions; more than 60 percent of the world’s major fisheries are in urgent need of actions to restore overfished stocks or to protect stocks from overfishing; natural forests continue to disappear at a rate of about 14 million hectares each year; and other ecosystems such as wetlands, mangroves, and coral reefs have been substantially reduced or degraded.

These changes in ecosystems have had significant effects on the goods and services they provide. Some of the impacts of these changes have been intended, such as the tremendous growth in crop production around the world, and many have been inadvertent, such as the degradation of water sources and the loss of biological diversity. Human development relies on ecosystem goods such as food, timber, genetic resources, and medicines, and on services such as water purification, flood control, carbon sequestration, biodiversity conservation, disease regulation, and the provision of aesthetic and cultural benefits. These goods and services are in turn dependent on various essential ecosystem processes such as pollination, seed dispersal, and soil formation. The loss and degradation of ecosystem goods and services hinder national development and take the most serious toll on the poor, who often depend directly on forests, fisheries, and agriculture for their livelihoods and who tend to be most vulnerable to problems resulting from ecosystem degradation such as floods and crop failures.

The sheer magnitude of the human impact on Earth ecosystems, combined with growing human population and consumption, mean that the challenge of meeting human demands for these goods and services will grow. Models based on the UN’s intermediate population projection suggest that an additional one-third of global land cover will be transformed over the next 100 years, with the greatest changes occurring over the next three decades. By 2020, world demand for rice, wheat, and maize is projected to increase by some 40 percent and livestock production by more than 60 percent. Humans currently appropriate 54 percent of accessible freshwater runoff, and by 2025 demand is projected to increase to an equivalent of more than 70 percent of runoff. Demand for wood is projected to double over the next half century.

These growing demands for ecosystem goods and services can no longer be met by tapping unexploited resources. The magnitude of human demands on ecosystems is now so great that tradeoffs among goods and services have become the rule. A nation can increase food supply by converting a forest to agriculture, but in so doing decreases the supply of goods that may be of equal or greater importance, such as clean water, timber, biodiversity, or flood control. It can increase timber harvest, but only at the cost of decreased revenues from downstream hydro facilities and an increased risk of landslides.

In order to make sound decisions about the management of the world’s ecosystems and to adequately weigh the tradeoffs among various goods and services that are inherent in those decisions, a dramatic increase is needed in the information brought to bear on resource management decisions. More specifically, effective management of the goods and services produced by ecosystems requires an integrated multisectoral approach, and it requires significantly greater use of ecological forecasting techniques.

The technical capacity to support integrated and forward-looking management decisions, particularly on a regional scale, is vastly greater today than even a decade ago. Three advances in particular have made this possible. First, the coverage and resolution of the new generation of remote sensing instruments, combined with long-term data sets pertaining to ecosystem conditions obtained through various national and international monitoring systems, provide scientists with the basic global and regional data sets needed to monitor ecosystem changes. Second, significant advances have been made in techniques and models that can be used for ecological forecasting. For example, watershed models enable relatively accurate predictions of the consequences of various changes in land use and land cover patterns on downstream water quantity and quality. Nutrient flow models enable predictions of the likelihood of eutrophication in watersheds subjected to increasing nitrogen or phosphorous inputs. And new combined climate/ecosystem models enable improved forecasting of the likely ecosystem effects of climate change.

Finally, considerable advances have been made in the field of resource economics. Using the tools and approaches now available for the valuation of nonmarketed ecosystem services, decisionmakers are better able to weigh economic tradeoffs among management choices. For example, within the United States more than 60 million people in 3,400 communities rely on National Forest lands for their drinking water, a service estimated to be worth $3.7 billion per year–an amount greater than the annual value of timber production from these lands. Only with this type of information at hand can a manager (or citizens) hope to make sound decisions balancing the tradeoffs and benefits obtained from various ecosystem goods and services.

A number of examples of the application of integrated ecosystem management now exist, such as New York City’s 1996 decision to invest in watershed protection to meet its clean water needs. In order to meet federal water quality standards, the city faced the choice of either filtering its water supply from the Catskill Mountains at a cost of $4 billion to $6 billion or protecting its water quality through watershed management. After examining the economics of the alternatives and after several years of negotiation with the local, state, and federal governments, the city opted for the watershed management approach at a cost less than half that of the filtration option.

The planned restoration of the Skjern watershed in Denmark provides another example of the application of integrated approaches to ecosystem management (as well as a cautionary tale about the problems that sectoral approaches can create). The course of the Skjern River has been modified several times since the 18th century, with the greatest change taking place in the 1960s when the lower 20 kilometers of the river were straightened and confined within embankments. River channelization had its intended effects of increasing the area of farmland and reducing the frequency of floods, but a series of unforeseen or unappreciated impacts more than offset these benefits: The frequency of rare but catastrophic floods increased, salmon populations plummeted, the new farmland subsided, salt intrusion and waterlogging reduced the agricultural productivity, and the increased agricultural use led to severe eutrophication of the river. Today, after careful modeling and cost-benefit analysis, Denmark is attempting to return the Skjern to something like its original state by eliminating embankments and recreating wetlands.

These examples illustrate the benefits of an integrated and forward-looking approach to ecosystem management that examines the consequences of various management alternatives for the full range of goods and services provided by the ecosystem involved. But significant obstacles prevent the more widespread adoption of this approach. Perhaps the greatest obstacle is not a technical one, but rather the current mindset that guides environmental management and is embodied in resource management institutions. It is the exception rather than the rule for a land manager to be asked to balance multiple objectives in his or her land use decisions. Farmers grow crops and foresters grow trees; ministries of agriculture support crop production, ministries of forestry support tree production. It is a nontrivial task for these institutions and managers to begin viewing their responsibility to be one of managing a bundle of goods and services and for policymakers to provide the incentives necessary to achieve this end.

And even where recognition exists of the desirability of considering the impact of management decisions on the full array of ecosystem goods and services, decisionmakers are often constrained by the lack of the basic technical capacity, modeling tools, and necessary data. For example, the watershed models mentioned above can be used to forecast the affect of various land use changes in a particular watershed on the timing and quantity of river discharge and on the sediment levels in the river, but these models require extensive site-specific information (about slope, soil type, land cover, timing of rainfall, etc.) that many regions do not have available.

Finally, an important barrier to the application of such approaches at a local scale stems from the dependence of local outcomes on regional or even global ecological processes. For example, for managers in a coastal region to effectively forecast trends in nearshore fisheries, they would need better information on trends in agricultural development in the watersheds draining into the coastal zone, since this will have a major impact on the likelihood of eutrophication in coastal zones. They would also need to know the potential impact of global phenomena such as climate change on temperature and precipitation in the region.

Millennium ecosystem assessment

There have been several instances in recent decades where a well-timed scientific or public policy initiative has capitalized on a foundation laid in basic scientific research to help bring emerging science into the mainstream of commerce or decisionmaking. In the late 1960s, institutions such as the World Bank, the UN Food and Agriculture Organization (FAO), the Rockefeller Foundation, and the Ford Foundation saw an opportunity to build on advances in crop breeding to spread the technologies and benefits of international agricultural research worldwide. In 1971, these institutions and other governments and foundations created the Consultative Group on International Agricultural Research, which went on to support agricultural advances that dramatically transformed the path of agricultural development in less developed countries. The 1990 establishment of the Human Genome Project, with a budget of more than $300 million per year, was a similar attempt to galvanize the findings of basic science and accelerate the application and use of those findings. The IPCC, established in 1988, also served to mobilize the findings from a growing body of work on climate science and effectively bring it to bear on the needs of public policy decisions related to climate change.

An analogous situation exists today in the case of ecosystem management. The data and tools needed for better management exist, but the obstacles noted above (awareness, capacity, and scale) now prevent the widespread application of these approaches. With these issues in mind, in 1998 the World Resources Institute, UN Environment Programme (UNEP), UN Development Programme (UNDP), and World Bank established a steering committee to explore whether a process could be developed to bring better scientific information on ecosystem goods and services to bear on public policy and management decisions. The Millennium Ecosystem Assessment Steering Committee is composed of leading ecological and social scientists from around the world and of representatives of many of the international bodies that might be either sources of information or users of information. The committee has proposed the following design for the initiative.

The proposed Millennium Ecosystem Assessment (MA) would be a four-year initiative to (1) use the findings of leading-edge natural and social science research on ecosystem goods and services to help make regional and global policy and management decisions, and (2) build capacity at all levels to undertake similar assessments and act on their findings. The MA would focus on the capacity of ecosystems to provide goods and services that are important to human development, including consideration of the underlying ecosystem processes on which those goods and services depend. Like the IPCC, the MA would be repeated at 5- or 10-year intervals to meet the changing needs of decisionmakers and to periodically update the state of the science regarding key policy choices.

The MA would address the following:

(1) Current ecosystem extents, trends, pressures, conditions, and value. The MA would provide baseline information for the year 2000 on the geographic extent of different ecosystems (including terrestrial, freshwater, and marine environments) and the land or resource use patterns associated with them. Building on the findings of other national, sectoral, and global assessments as well as on newer remote sensing data, it would present information on trends in ecosystem goods and services, their condition and value, their contribution to human development, and the pressures affecting them.

(2) Ecosystem scenarios and tradeoffs. The MA would present a range of plausible scenarios showing how the quantity and quality of ecosystem goods and services may change in coming decades in different regions of the world. For example, just as scientists within the IPCC tackled the question of how rising CO2 concentrations would affect climate, the MA would examine such questions as: How will the expected 50 percent increase in flows of fixed nitrogen over the next 30 years affect water quality and fisheries productivity in different regions of the world? Given projected trends in land use change over the next 30 years, what will be the likely effect on the availability and timing of freshwater supplies in different regions? Given the expected continued growth in species introductions worldwide, what will be the likely impact on biodiversity and on various ecosystem goods and services?

(3) Response options. The MA would identify policy, institutional, or technological changes that could improve the management of ecosystems, thereby increasing their contributions to development and maintaining their long-term sustainability.

The proposed assessment would examine conditions, scenarios, and response options at a global and national scale and would also include a small number of assessments at smaller scales to help to catalyze more widespread use of integrated assessments and to develop the methodologies and modeling tools needed by those assessments.

There is now considerable experience that can aid in the design of such a process. Experts that have examined successful and unsuccessful initiatives to better link scientific information to public policy actions point to three prerequisites for success: saliency, credibility, and legitimacy. Scientific information is salient if it is perceived to be relevant or of value to particular groups who might use it to change management approaches, behavior, or policy decisions. It is credible if peers within the scientific community perceive the scientific and technical information and conclusions to be authoritative and believable. It is legitimate if the process of assembling the information is perceived to be fair and open to input from key political constituencies, such as the private sector, governments, and civil society.

The IPCC successfully meets these criteria. Considerable effort has been made to ensure that the MA will also meet these criteria. In order to ensure that the findings are salient and legitimate, the key users need to participate in designing the focus and content of the process. More specifically, the intended users must be more than a hypothetical audience; they must be actively requesting such an assessment, or else there is a risk that the findings will not be used. However, unlike the IPCC, which has a single “audience” in the form of the Framework Convention on Climate Change, it was apparent from the outset that the set of potential users of the findings of the MA is quite diverse. At the international level, a number of different ecosystem-related conventions, such as the Convention on Biological Diversity, the Convention to Combat Desertification (CCD), and the Ramsar Wetlands Convention, need the type of information that the MA would produce. In addition, national governments, regional institutions, national ministries, and the private sector are also important users, because they are most directly engaged in the specific management actions that can benefit from improved understanding of the potential impacts of various management decisions on ecosystem goods and services.

For this reason, the steering committee has proposed that a board of users drawn from this array of institutions be established to govern the MA, with particular representation from the ecosystem-related conventions. This board, in consultation with the scientists that will undertake the assessment, will identify the questions that the MA will seek to answer, thereby ensuring that the scientific findings address the issues of relevance to key users. The board also will ensure the legitimacy of the process by involving key stakeholders and setting the policies for such issues as peer review. In this way, the MA will become a joint assessment undertaken in partnership by several ecosystem-related conventions and other key users to meet the specific needs of the decisionmakers represented by those institutions.

Equal effort has gone into designing the process so that it will be highly credible within the scientific community and with various users. Again, the IPCC provides useful lessons. Part of IPCC’s success can be traced to the fact that IPCC Assessments are not a particular institution’s interpretation of the findings of climate science but rather the direct conclusions of the experts themselves. This arrangement will be emulated in the MA. The MA will be conducted through a set of working groups, each chaired by leading scientists in the fields in question. The chairs of each working group will make up the ecosystem assessment panel, which in turn will be co-chaired by leading natural and social scientists. The assessment panel will interact closely with the MA board in identifying the questions that should be answered by the MA, but the scientific assessment will then follow an independent peer-reviewed process.

A number of different institutions will facilitate the assessment. A small core secretariat will be established at the institution, housing one of the chairs of the process. Other institutions will house coordinators for the various working groups or perform administrative, logistical, or outreach functions. For example, the working group focused on scenario development is likely to be organized through the Scientific Committee on Problems of the Environment (SCOPE), a program of the International Council on Science. Within the UN system, the MA will be conducted through a partnership arrangement among UNEP, UNDP, FAO, and the UN Educational, Scientific, and Cultural Organization. Finally, the assessment will be closely linked to a number of processes such as the International Geosphere Biosphere Program, the IPCC, the Global International Waters Assessment, and the UNEP Global Environmental Outlook process.

From rhetoric to reality

The MA has already built substantial momentum, but its creation is by no means ensured. The successful launch and completion of the MA will require political buy-in, financial support, and scientific engagement. From the political standpoint, governments and other users, in particular the international environmental conventions, must “own” the assessment. It will not succeed if the various users view it as an external process being conducted by scientists that may or may not generate useful information. Instead, it must be a joint initiative of the various users that is designed to meet their needs.

Considerable progress has already been made in establishing this level of ownership among the various users. In May 1999, the Ramsar Convention noted “the scope of the proposed Millennium Assessment of the World’s Ecosystems, currently under development, to deliver valuable related information of relevance to the application of the Convention.” In September, ministers of environment or their representatives from Australia, Canada, Cote D’Ivoire, the Czech Republic, Denmark, Finland, Germany, Ghana, Japan, Kenya, Mozambique, the Netherlands, Nigeria, Norway, South Africa, Sweden, Togo, the United Kingdom, the United States, and Zimbabwe stated that: “The concept of a global ecosystem assessment . . . should be supported as a means of helping decision makers in assessing the impact of their various actions on their national as well as on the global ecosystem.” At the November 1999 Conference of Parties to the CCD, Senegal introduced a statement recommending support of the MA, and this was supported by Brazil, Norway, China, Kenya, and the United States. And, at the January 2000 meeting of the scientific body of the Convention on Biological Diversity (CBD), parties requested that the convention’s executive secretary explore ways and means of collaborating in the MA with other relevant conventions and organizations.

But an important additional step is needed: The parties to the ecosystem-related conventions should now authorize those conventions to engage as partners in this joint ecosystem assessment. This does not mean that the MA would be the only mechanism available to the conventions to meet their science assessment needs; indeed, each convention is likely to design other more targeted assessment processes. However, in areas related to ecosystem goods and services where the scientific information needs of these and other related conventions overlap extensively, it will be important that these conventions directly communicate their information needs to the MA and, in turn, have a formal channel for receiving the findings of the assessment process.

On the financial front, the Global Environment Facility and the UN Foundation have indicated strong interest in the MA, subject to their council and board approval, and these sources of funding could cover about half of the $20 million budget. Funding is not yet secured, however, for many of the most essential components of the assessment, including the work to develop scenarios for ecosystem change and the catalytic local, national, and regional assessments. Just as important, a process such as the MA depends heavily on in-kind contributions by experts. (By way of comparison, the budget of an IPCC assessment is comparable to the MA budget, and in-kind contributions of time to the IPCC process are estimated to be equal to its budget.)

In the case of the MA, one of the most valuable in-kind contributions could be in the form of data. A major goal of the MA will be to improve baseline information related to ecosystem goods and services for the year 2000. Although remote sensing cannot provide all of the information needed for a baseline assessment, it is nonetheless one of the most important sources of new data that could be synthesized and disseminated through the MA process. The United States, for example, could make an invaluable contribution to the MA and to the capacity of other countries and institutions to effectively manage their ecosystems by contributing a global coverage of Landsat 7 data (worth some $10 million) for the year 2000 to the MA process. Although those images would require further processing by individual nations and researchers, they would nonetheless provide an extraordinarily useful common baseline data set for measuring and monitoring changes in ecosystems. These data, combined with the satellite information on land cover and ocean characteristics that will be available within the next few years from such instruments as the Moderate-Resolution Imaging Spectroradiometer aboard the recently launched Terra satellite would represent a quantum leap in the amount of information countries have on hand for making wise decisions regarding the use of their ecosystems.

On the scientific front, leading scientists from around the world have already been engaged in designing the MA, as members of either the steering committee or the advisory group, and an article published by the steering committee in Science has helped to generate awareness within the scientific community. But broader engagement of social and natural scientists is essential. This requires the commitment of a few leading scientists to play key roles in chairing the various components of the assessment process, and it will require convincing the scientific community that the time they devote to the process will be time well spent. A convincing case can be made only if it is clear that decisionmakers will use the findings.

Scientists in the United States were some of the early proponents of creating an IPCC-like ecosystem assessment process. Furthermore, the United States provides models of integrated assessment approaches, such as the Heinz Center Report on the State of the Nation’s Ecosystems, and examples demonstrating the utility of these approaches for ecosystem management. However, on the political front, the proposed MA is somewhat unusual among international scientific initiatives in that the United States has not been a key force behind the idea. Instead, a wide range of European countries and developing countries in Africa, Latin America, and Asia have thus far provided the leadership. Greater engagement of the United States with its scientific and data resources could greatly strengthen the process.

The linkages among political buy-in, financial support, and scientific engagement are direct. Buy-in requires that the users be convinced that the process will meet their needs. If convinced of the utility, the various users are likely to provide the financial support needed by the process. If the buy-in and financial support exist to show that the work will be used, then the scientific community is likely to make the commitment needed to undertake the assessment. And assurance that the scientific community will be effectively engaged and that the results will be of the highest credibility is key to obtaining the political buy-in.

The proposed MA is a novel institutional arrangement and process being built by an equally novel alliance of governmental, intergovernmental, scientific, and nongovernmental institutions to meet a very real set of issues with profound impacts on human lives. Decisions taken by local communities, national governments, and the private sector over the next several decades will determine how much biodiversity will survive for future generations and whether the supply of food, clean water, timber, and aesthetic and cultural benefits provided by ecosystems will enhance or diminish human prospects. The scientific community must mobilize its knowledge of these biological systems in a manner that can heighten awareness, provide information, build local and national capacity, and inform policy changes that will help communities, businesses, nations, and international institutions better manage Earth’s living systems. The MA could help dramatically speed the adoption and use of leading-edge ecosystem science, management approaches, and tools, but its success–and even existence–now require much broader engagement by institutions and agencies in the United States as well as other countries.