GENERAL THEMES

The Board reviewed and considered hundreds of recommendations from reports and policy documents; from scholars in every scientific discipline and a broad range of professional societies; from local and Federal agency officials; and from nongovernmental

			A NOTE OF THANKS The Board is grateful to all of the individuals and organizations that provided comments during the process of developing this report. The thought and care that went into these responses were obvious, and this report has benefited accordingly. The Board does not endorse all of the comments received, but appreciates the intent behind them and the perspectives that were brought to the table. The findings and recommendations offered in this report reflect a careful process of developing coherent policy guidance for the Foundation that has necessitated difficult choices. The context for this consideration is evident throughout the report.

organizations, community groups, and concerned citizens (see Appendices B, C, and D). Many of the suggestions transcend NSF's mission and relate more properly to the entire Federal portfolio of environmental activities. Nonetheless, we include them as a record of those points made repeatedly and as a basis for many of the findings and recommendations presented in this report. In addition, the Board examined a variety of programs at NSF to determine the factors most likely to result in effective research, education, and scientific assessment activities.

Several themes emerged from this diverse input. Foremost among them was a strong endorsement of NSF's fundamental operating principles. In particular, the following strengths were highlighted:

• Credibility. NSF's merit review approach is considered key to the credibility of its environment portfolio.

• Program flexibility. The ability of core NSF programs to evolve over time as different fields of study emerge, change, and combine is widely supported.

• Emphasis on education. NSF gets positive marks for its support of education and the integration of education with research.

• Leadership. One of NSF's major strengths is its ability to activate the intellectual assets of the research and education communities and to mobilize resources for addressing substantive scientific and engineering challenges.

• Flexible funding. The ability of program officers to allocate funds to facilitate the early development of emerging fields is both beneficial to nascent disciplines and an excellent mechanism for attracting outstanding scientists to serve in the critical role of program officers.

These strengths place the Foundation in a unique position to expand its efforts to enable a broad spectrum of advances in the research community and to strengthen and expand its partnerships with other Federal agencies in support of environmental research, education, and scientific assessment.

Also from this input, the Board heard many ideas that framed ways in which NSF could and should develop its environmental portfolio. The repeated suggestions are summarized below.

INPUT RECEIVED DURING THE HEARING PROCESS^*

Enable Significantly More Interdisciplinary and Multidisciplinary Research to Address Environmental Issues and Problems

This recommendation has been repeated frequently over a number of years as researchers have grappled with the extraordinary complexity of environmental systems and the factors influencing those systems. For example, the Corson Report (NRC 1993) notes that "the research establishment is poorly structured to deal with complex, interdisciplinary research …" Expertise from multiple disciplines—including the physical, biological, and social sciences and engineering—is required to advance understanding and solve environmental problems. Many of the individuals who spoke to the Board in its public events or via its web site emphasized this as an area that NSF needs to strengthen, and a sizable fraction of the approximately 250 reports in Appendix Balso mentions this issue. Many also emphasized the inherent difficulties in establishing interdisciplinary and multidisciplinary projects within the context of disciplinary programs through which funding is presently available.

The best interdisciplinary science must be firmly grounded in rigorous disciplinary research. Enabling productive interdisciplinary efforts, however, requires significantly more than simply assembling outstanding disciplinary researchers. Successful interdisciplinary research requires different ways of conceptualizing problems; an openness and respect for other disciplines; and the availability of time for the development and maturation of new interactions, language, understanding, methodologies, and concepts. Fostering interdisciplinary research thus needs to occur in parallel to the conduct of disciplinary research. The report of the USGS Workshop on Enhancing Integrated Science explores this area and suggests a draft set of principles for the conduct of interdisciplinary science endeavors (USGS 1999b).

The Board heard that interdisciplinary grant competitions at NSF suffer from weak continuing institutional commitment and planning. Environmental research takes at least a 2- or 3- year startup period to become fully effective. Once a competition is announced, program officers within NSF and researchers in outside communities must assemble new alliances and learn to work together. It takes a couple of times through the process to learn it well. The Board heard that by the time the process becomes well focused, changes within the Foundation shift budgets to other priorities and personnel rotate out or are reassigned elsewhere. Many NSF interdisciplinary programs operate in only startup or lame duck mode, and this obstructs real progress toward addressing important, interdisciplinary environmental issues.

The Board also heard endorsements of the core programs at NSF and was urged to secure funding for environmental research that complements and expands existing activities. There has been, and continues to be, a tremendous amount of important knowledge that has been generated by the solid foundation that the core programs at NSF provide. Environmental research that has major political elements has the potential to substantially diminish the stability of NSF's environmental research efforts. While it would be very helpful to strengthen and expand existing NSF programs and to increase the capacity for interdisciplinary environmental research, such expansions should not be made at the cost of the long-term stability of disciplinary environmental research at the Foundation.

*Prior to the July 1999 release of the Interim Report.

Recognize The Inherent Complexity and Nonlinearity of Most Environmental Systems

Many individuals suggested that NSF's new focus on biocomplexity is timely and urgently needed, but felt that support for a far greater effort in this area is required. They pointed out the importance of recognizing the inherent differences between reductionist approaches (which focus on smaller and smaller units of a process or system) and more synthetic approaches (which emphasize interactions among components, complex behaviors, and emergent properties). Significant advances in synthetic, holistic approaches are required to understand environmental systems.

Environmental issues are often characterized by both interdisciplinarity and complexity. For example, scholars concerned with conservation of biodiversity must synthesize advances in evolutionary systematics, biogeography, and ecological genetics in order to understand genetic diversity and how it can be conserved.

In another example, the Board learned that synthesis of advanced process understanding in atmospheric science, hydrology, and geology is necessary to quantify mass flux and energy balance in certain natural systems. This is specifically important to our understanding of the complexities of flow in the "vadose zone": the region of soil and fractured rock where we are intentionally (Yucca Mountain) and inadvertently (Hanford and other sites) storing high-level radioactive waste.

The Board also heard testimony urging NSF not to make biocomplexity the lens through which all environmental research should be focused. The concern was that it risks making the term so broad as to be meaningless and could devalue disciplinary research not central to understanding biocomplexity.

Consider Questions At The Appropriate Temporal And Physical Scale By Taking Into Account Long-Term And Large-Scale Research Needs

The Board heard from a variety of sources that the need for long-term research, monitoring, and assessment of environmental trends far exceeds what is generally being delivered. A whole new level of effort is needed to complement the excellent examples of long-term, large-spatial-scale research that were identified (e.g., certain Global Change Research endeavors and the LTER program).

The vast majority of field studies are of insufficient duration or spatial scale, or both, to capture important phenomena. For example, in a survey of the duration of research projects published in the journal Ecology between 1977 and 1987, Tilman (1989) found that 40 percent of those studies had time periods of less than 1 year and that more than 92 percent of experimental field studies had durations of 5 years or fewer. Given that many organisms require more than a few years to complete their life span and that most ecological processes require a long period to exhibit their potential range, an emphasis on shorter term projects can substantially constrain the development of environmental understanding. Similarly, the spatial scale of most research projects does not approach the scale at which whole system patterns and processes begin to emerge.

The idea of environmental research and education hubs—physical and/or virtual centers, or collaboratories—was advanced as one way by which researchers could synthesize the findings from long-term and large-scale research. A parallel goal for such hubs could be the integration of research with education.

The Board also heard that long-term and large-scale research offers opportunities for partnerships with other Federal agencies as well as state, tribal, and local agencies and NGOs. The LTER program could serve as a model for how such partnerships might be established and maintained.

Include Appropriate Human Components (e.g., Economics and Social Sciences) In Environmental Research And Education

Over the last decade or so, an increasing number of environment-related reports have noted that great leaps in our understanding of environmental systems will be made as system paradigms expand to include human sciences. New areas encompass theoretical and empirical research to develop measures of sustainable consumption levels; quantitative studies on the efficient use of resources; research on the relationships between environmental regulations, private sector investment decisions, and productivity growth; and research on participatory processes, scientific and technological innovation, and resource management.

A particularly critical area of study, research on environmental valuing and decision-making, has shown that humans weigh concerns for social justice, aesthetics, history, and economic factors in assessing the merits of policy and practice. Further research is needed to identify the kinds of participatory processes and educational approaches that enhance human ability to make good use of scientific information in developing stable, sustainable environmental policies, frequently in the face of substantial scientific uncertainty (see Box 11).

The Board heard testimony that the human sciences have developed with impoverished spatial information, in part because until recently the capacity to create such large data sets was constrained by enormous costs and the capacity to analyze such data was poor. Information technology has now advanced to the point that spatially explicit problem solving in the human sciences can be integrated meaningfully with similar approaches in ecology, the geosciences, and other fields. For example, these capabilities could be applied to concerted, long-term research into the historical effects of human communities on local environments . This type of research could provide fine-grained, spatially explicit, historical data on changing ecosystems and on the dynamic relationship of human communities and ecosystems.

Create A More Effective Information Infrastructure To Facilitate Significant Advances In Informatics, Data Management, Modeling, Synthesis, And Dissemination Of Information

It is generally acknowledged that effectively addressing environmental issues requires utilizing the powerful new tools of information technology to manage, use, and communicate the scientific data and information already in existence and to be generated by future research and monitoring.

The Board learned that approximately $600 million per year is spent on environmental information generation through research, data collected by monitoring efforts, and the storage and analysis of data (PCAST 1998). But existing high-quality information is not currently being incorporated into management decisions because of lack of electronic availability of the information and inadequate capabilities to interpret, synthesize, and analyze that information.

For example, the United States possesses approximately 750 million biological specimens in its natural history museums and herbaria. The georeferenced data (geographic coordinate data attached to the biological information) from these specimens are urgently needed as a tool to study the status and trends of ecological systems, but the vast majority of this information has not been digitized.

The Committee for the National Institute for the Environment, in testimony to

Today we speak easily of collaborations between molecular biosciences and ecology. What we quickly forget is the sometimes long period of incubation before such collaborations take hold and lead environmental science in new directions. To realize the Nation's environmental research agenda, we need to understand the process of scientific collaboration better. Perhaps the vehicle here is information. Therefore, the Board could well explore how we bring information technology more fully to the environmental research agenda — W. Franklin Harris, University of Tennessee

the Board, called for an overarching electronic network for this spectrum of information activities. This network would feature the combined use of Internet-centered information technology, services, products, existing organizations and systems, and information specialists organized into an environmental information infrastructure. The recommended network would facilitate linkage of distributed information and databases, improved quality control of databases, increased support for data standardization and information management, and improved access to information for the public.

The Digital Library Interoperability project at the University of California—Santa Barbara, Stanford University, and the University of California—Berkeley may provide lessons. The Internet allows computers to exchange data, and the web gives computer users interactive access to information. But users of digital libraries and information grids need services to help them manage raw information and organize data. This NSF-NASA-Defense Advanced Research Projects Agency-supported project is building tools and services to allow people to exploit the remarkable opportunities for collaborative creation and sharing of knowledge that a digital world makes possible.

Develop And Exploit State-Of-The-Art Technology To Advance Environmental Studies And Address Environmental Problems

New computational algorithms, remote sensing, new kinds of sensors, genome sequencing, laser technologies, and other advanced approaches are moving environmental research into a new era. Previously inconceivable advances are being suggested. A variety of tools from molecular biology (e.g., oligonucleotide probes) are letting us interrogate microbial assemblages to find out what microbial types are present, what they can do, and what they are doing. One scientist testified to the Board, for example, that genomic barcoding of the pathogen Pfiesteria in the Chesapeake Bay may become a reality thanks to microchips that will identify the organism's genome as quickly as a supermarket scanner. Tools from molecular chemistry (e.g., advanced x-ray methods) allow scientists to collect unprecedented kinds of information about geochemical environments at the microbe scale.

Powerful new computers and algorithms are letting scientists construct models that include the true complexity of biogeochemical systems. We are beginning to access the information processing capability to connect the many processes of environmental and human systems coherently so that we achieve a comprehensive understanding. But the kinds of profound advances that we foresee require integrated research and application of these advanced tools across a broad front of fundamental questions and environmental issues.

Other kinds of advances should be supported in the newly emerging environmental technology area of industrial ecology, a field that takes a systems view of the use and environmental implications of materials, energy, and products in industrial societies. Specifically, it places industrial activity in its environmental context and draws on nature as a model for the processes involved in industrial activity. The rich research agenda for industrial ecology has grown from more traditional research on particular materials and economic sectors to include needs for cross-sector and multiscale approaches.

The Board also heard that fundamental research is needed to enable the shift from waste management and remediation to avoidance of environmental harm. For example, fundamental studies in chemistry and engineering have led to environmentally benign alternatives to chlorinated hydrocarbons for use in the synthesis of chemicals and pharmaceuticals and in manufacturing processes. Industries have been quick to adopt new products such as these, as well as new approaches to polymer production, drycleaning, and paint application that prevent pollution and thereby avoid environmental harm.

Support Inventory And Monitoring Programs To Characterize Animal And Plant Resources And To Determine Their Status And Trends

Plant, animal, and microbial species provide the basis for economically productive enterprises, including crop and timber agriculture, livestock husbandry, fishing, and consumptive and nonconsumptive wildlife recreation. The Board learned that protecting the basis of these endeavors calls for a more extensive understanding of the wild relatives of these species (as rich sources of new genes), of threats from invasive species including pests and pathogens, and of the ways in which the relevant ecosystems will respond to the plethora of ongoing global changes. In addition, studies of genetic diversity and the rich array of chemicals and structures found in plants, animals, and microbes contribute directly to many facets of the biotechnology industry and biomedical research. The need for evaluation of patterns and causes of change goes beyond the need for information on individual species. Assessing the status and trends of ecosystems providing essential services is increasingly recognized as vital to economic and health interests. Ecosystem services of particular interest include pollination, pest control, water purification, and flood control (PCAST 1998).

Support Research That Connects More Effectively With Decision-Making (Policy, Regulatory, Management, Institutional, And Individual)

There has been a growing interest over the last decade in improving the scientific basis of environmental decision-making. Several recommendations the Board heard and read on this topic are of relevance here: (1) research results should be communicated to potential users in a useful and understandable form; (2) research should include a focus on those environmental problems where users need better information (see Box 12); and (3) public understanding of science, in particular in the environmental area, needs to be improved.

Knowledge assessments are one route toward providing a common base of understanding. A model for such knowledge assessments might be the Issues in Ecology series produced by the

Unlocking Our Future, the Report to Congress of the House Committee on Science (1998), emphasizes that the role for science in helping society make good decisions will take on increasing importance, particularly as we face difficult decisions related to the environment.

Ecological Society of America. These peer-reviewed publications report, in lay language, the consensus of a panel of scientific experts on issues relevant to the environment.

The Corson Report, the Committee for the National Institute for the Environment, the American Institute of Biological Sciences, and the Ecological Society of America, among others, suggest specific ways to improve the use and usefulness of knowledge resulting from the research enterprise (see e.g., NRC 1993, CNIE 1994, Blockstein 1997). Suggestions include: improved coordination across the environmental research portfolio; setting priorities to produce a more comprehensive knowledge base; better mechanisms for the communication of urgent societal needs to the research community; better communication of research results to multiple audiences; improved mechanisms for organization, management, and distribution of data; and improved public understanding of science and environmental issues.

Include Educational Elements In Environmental Programs And Plans

The Board heard that education and training in the Nation's universities are strongly disciplinary, whereas solution of environmental problems also requires broadly trained people and multidisciplinary approaches. Opportunities for broadly based interdisciplinary graduate degrees are few, and faculty are often not as well rewarded for interdisciplinary activities as they are for disciplinary work. Additionally, environmental scientists are often not appropriately trained to address pressing needs and fill positions in career paths outside academe.

Complexity, and biocomplexity in particular, offers roadmaps for training the next generation of scientists. Creativity in building the educational support system for this new integrative environmental science is an especially important challenge. It will require new models of institutional cooperation and new degrees of freedom on the part of NSF program officers to assess and build creative, integrative research/educational programs.

Improve Coordination Among Programs And Agencies

The need for good communication and coordination across agencies was highlighted as an ongoing challenge (see Box 13). CENR provides a mechanism for this coordination and has overseen a variety of highly successful interagency activities. For example, USGCRP has for a decade focused multiple Federal agencies on understanding the components of the Earth system and modeling at the global scale. Within a coordinated framework, progress has been made in understanding the loss of stratospheric ozone, the important roles of terrestrial and marine ecosystems in the overall carbon cycle, and past changes in the Earth's environment that provide a context for anthropogenic changes now ongoing. USGCRP has also provided predictive information about El Ni&#ntilde;o that has been useful to natural resource management and agencies concerned with human health and safety.

Not all testimony supported coordination of Federal agency activities through a committee structure. The Board heard testimony that interagency programs may lack the necessary ownership within each agency and can lead to renaming of existing activities rather than major new initiatives.

The Board also learned of excellent examples of interagency coordination that have not involved CENR. One example is NSF's interaction with universities and other Federal agencies to develop and implement the network of LTER sites. Many of these projects involve complex partnerships with mission agencies, and the scientific yield has been extraordinary. The Board heard that NSF must continue its leadership role and its partnering efforts with other agencies.

Finally, the Board heard that scientific assessments, by establishing and communicating a base of scientific knowledge on a given topic, can provide a mechanism for improving collaborations between Federal agencies and between the Federal and private sectors.

Improve Predictive Capabilities In A Variety Of Environmental Areas

Our ability to predict the behavior of environmental systems has grown steadily with an increase in understanding of many of these complex systems. For example, interdisciplinary paleoclimatic research is improving our understanding of the Holocene climate. This is important because it is within the Holocene that the boundary conditions for modern natural climate variability can be identified and from which the relative importance of natural versus anthropogenic climate forcing can be assessed. Understanding of modern climate and prediction of future climate will require a detailed understanding of Holocene climate forcing and response.

Most environmentally related scientific inquiry focuses on components of the environment or the individual effects of one component on others. Simulation and other models provide a framework within which to place our understanding of all the components simultaneously as they occur in nature. This framework allows quantitative accounting of the interaction of the component parts with factors outside the system and the sometimes surprising responses resulting from feedback among interacting components. For example, at the Central Plains Experimental Range LTER site, scientists have studied and modeled the long-term effects of grazing on vegetation succession dynamics. Surprisingly, heavy grazing in this system resulted in little change in annual net primary production, increased plant density, and decreased abundance of exotic invading species. LTER scientists speculate that this response reflects the importance of native herbivores in these ecosystems over evolutionary time. This particular response to grazing has not been generalizable to all grassland ecosystems, however: Sensitivity to grazing varies with gradients of productivity and environmental conditions.

Comparison of model output with data from environmental experiments indicates how much confidence can be placed in the models. And models that have been tested successfully in a variety of situations permit more robust predictions about the complex behavior of the environment. Modeling experiments can be conducted to help design research in unexplored areas. Additionally, sets of environmental drivers can be used in models to represent management or impact scenarios of particular interest to scientists or society. Simulation models have thus become tools of necessity for environmental research.

Get Input On Priority Setting From People And Organizations Familiar With Research, Education, And Assessment Issues

No multifaceted program can be accomplished without setting priorities. The Board examined several examples where research or education agendas were defined in an inclusive and integrated manner. For example, the Freshwater Imperative Research Agenda (Naiman et al. 1995) was developed with NSF support over a 2-year period of study and consensus building involving a broadly interdisciplinary array of scientists, managers, and educators. This research agenda sets priorities and develops detailed research questions as well as makes recommendations for implementation. Such research agendas are the exception rather than the rule, however, and it became clear to the Board that this is an area that needs much more attention, in particular where priorities are set in interdisciplinary areas.

Throughout the public input process, it became increasingly clear that citizens, government officials, representatives of other Federal agencies and of professional scientific and engineering societies, and individual scientists look to NSF for leadership in environmental research, education, and scientific assessment. The expectation that NSF will play a key role was highlighted for the Board in a number of ways and by groups ranging from National Research Council committees to advocacy groups. The strong message running throughout the hearings was that NSF can, and is expected to, respond vigorously to the new challenges of providing and communicating the fundamental knowledge base, and educating and training the workforce to meet the environmental challenges of the new century. A parallel message underscored the requirement for significant new resources to accomplish these goals.

INPUT RECEIVED IN RESPONSE TO THE INTERIM REPORT

After the release of the report as an interim document in July 1999, the Task Force on the Environment web site received almost 7,000 "hits." The Task Force also received comments on the Interim Report from a variety of individuals and professional organizations representing several thousand environmental scientists, engineers, and educators (see Appendix D). The vast majority of these comments were quite positive, reinforcing the input received earlier and supporting the recommendations. A number of the suggestions were very helpful to the task force in strengthening and clarifying the report. Several additional points were made by multiple respondents:

The Report Should Be Implemented In A Way That Is Recognizable As A New Approach In Order To Be Successful In Receiving Funding And Providing Scientific Information That Will Make A Difference

In addition to this explicit point, several organizations commented that the report should be implemented as a cohesive program, not treated as a menu from which selections might be made. Several organizations were interested in how NSF would collaborate with interested parties to pursue implementation and in how NSF would integrate research, assessment, education, and information. The point was also made that outcome assessment tools should be developed by which the success of the report could be measured.

In recommending a unique implementing entity, multiple respondents suggested that interdisciplinary programs be established and separated from disciplinary units to most effectively nurture interdisciplinary approaches. The underlying concern expressed was that as long as interdisciplinary programs compete for resources within a single budgetary organization, they will be at a disadvantage for the simple reason that interdisciplinary proposals will be perceived as less relevant to the core goals of the disciplinary unit. This point was coupled to the observation that most interdisciplinary activities cannot be sustained over the necessary time periods without an organizational home within NSF. It was of interest to the Board that these comments were made multiple times and across virtually all disciplines.

NSF Should Consider Different Approaches To Establishing Research Priorities

Several respondents suggested that NSF include stakeholders—established environmental groups, scientists, policy-makers at all levels of government, and NGOs—in the process of

With regard to the NSB report overall, we applaud the Board's recommendation that environmental research be made one of NSF's highest priorities and agree that funding should be substantially augmented, particularly in five specific areas emphasized in the report: interdisciplinary research; environmental education; economic valuation of ecological goods and services; long-term, large-scale research; and improving environmental assessment capabilities.—President's Committee of Advisors on Science and Technology, 1999 (Appendix E).

determining research priorities. Others suggested that priorities should be developed in large part through a series of scientific assessments. Several respondents advised NSF to palce the discussion of research needs into a broader national context and fully consider leveraging opportunities.

NSF Should Work To Break Down The Barrier Between Applied And Basic Environmental Research

Many respondents commented on what has emerged as a continuum between applied and basic research, and several suggested that NSF allow policy-relevant basic research to flourish alongside more traditional approaches. It was pointed out that the illustrative boxes in the Interim Report concentrated on problem-focused science and engineering, while the text emphasized fundamental research. Several respondents wondered if this was a disconnect or an intentional effort to highlight where the most exciting advances were occurring. Others asked if NSF would now support science that meets intellectual merit criteria but is primarily directed toward environmental improvement rather than scientific advancement.