MEETING THE INFORMATION NEEDS OF NPBS CLIENTS
One of the most important uses of the scientific information gathered by the National Partnership for Biological Survey is to assist decision-makers in addressing existing issues and in anticipating future ones related to biological resources. Recommendations made in the previous chapter for setting priorities for research and data-gathering emphasize factors relating to both near-term and long-term resource management. The research activities outlined in Chapter 2 are essential for gathering scientifically reliable information on which credible and reliable decisions can be made with a high level of confidence. Decisions based on inadequate, unreliable or incorrect information might be unwise and costly.
To respond effectively to information needs at all levels (local, regional, national, and international), NPBS and National Biological Survey (NBS) data and information policies and programs must move from problems to solutions. To do so, the National Partnership must address the following questions:
Who are the current and potential users of the kinds of information that the NPBS will produce, and how can their needs be met? In particular, how can research findings be communicated to decision-makers in ways that foster understanding and promote appropriate interpretation and use?
How should NPBS information be collected and managed to serve modeling and prediction efforts in support of biological-resource management?
How can the NPBS make the best use of the rapidly evolving technology and systems in information management?
How should information collection and management be coordinated among NPBS participants?
Those questions should be addressed as soon as possible to produce a detailed information model describing the flow of information from collection and analysis to dissemination and use. The model should address general conceptual issues, such as designing stronger linkages between scientific information and decision-making, identifying key environmental variables and relevant spatial and temporal scales, and assigning clear institutional responsibilities to minimize redundancy. It should also deal with technical issues, such as data format, data-exchange protocols, quality assurance and control, and hardware and software requirements.
This chapter highlights some central issues for linking scientific information to decision-making and offers recommendations on the quality, availability, and dissemination of information.
The National Partnership will have to provide information on biological resources in an easily interpretable form and in a timely manner to a wide array of resource managers, city planners, conservationists, scientists, and others. Its databases will have to
be able to answer standard questions about plants, animals, and microorganisms, such as ''Where does this species occur?'' and questions that require more complicated data analysis, such as "What species occur in a given vegetation type, and which of them occur in other vegetation types as well?" Answers to those kinds of questions should be available simply, quickly, and conveniently, to anyone who needs them.
Mechanisms must be developed to provide reliable answers to complex user queries that require new analysis and interpretation of existing data. Resource-management decisions are themselves sources of information about biological processes, and data accumulated from them should be incorporated in databases and made available to be used in future decision-making.
Some specific needs for NPBS data are the following:
Public agencies need better biological data to guide acquisition of land and water rights for reserves; set priorities for research, management, and restoration programs; zone local land use and regulate use of public and private lands; locate and design public-works projects; and coordinate resource management both domestically and across international borders.
The private sector needs better information on the distribution of legally protected resources; on the impacts of metropolitan growth on ecological resources, renewable land use and nonrenewable natural resources and ways to reduce those impacts; on the distribution of species of potential economic value; and on the biological impacts of pollutants.
Decision-makers need readily available information on status and trends to alert them to issues that require legislative or regulatory attention and to assess the effectiveness of current programs.
Research scientists need improved access to biological data to help direct and design their research.
The public needs information on local organisms and habitats, on the ecological role of humans in the environment, and on regional and national trends.
Despite increasing management attention to ecosystem patterns and processes, much planning and development is still based on single species. In the absence of abundance or trend data for most species, inferences about status and about conservation risk must be drawn from incomplete distribution data and habitat models. Box 3.1 illustrates how spatial data can be integrated to model distribution and management status, in this case for the orange-throated whiptail lizard (Cnemidophorus hyperythrus). The example highlights several important points:
Species distributions are almost never known with certainty. Rather, they are modeled by integrating information from such widely disparate sources as biogeographic atlases, museum collections, habitat-relationship information, and habitat maps derived from environmental data and remote sensing imagery. Users of distribution information must be clearly informed as to who produced it, how the results were obtained, the information scale, potential biases, and inaccuracies. This knowledge of information quality is needed to ensure appropriate use of the information. Even for an easily observed and relatively well-known species, existing collection and sighting data are likely to be dated or biased. Computerization of existing data must be weighted in a way that recognizes the differences in precision of different records.
For conservation and management, knowing where a species occurs is less useful than knowing what processes allow the species to persist in some areas and not others. This requires observation, experimentation, and modeling to develop good life-history information and an understanding of ultimate controls on species' distributions and abundances. For both practical and scientific reasons, we are unlikely to have this knowledge in detail for most species in the near future. Until it is available, we will have to depend largely on the best available survey and monitoring data, correlative analyses, and scientific judgment.
Hierarchical integration and spatial display of existing
distribution and habitat information is a powerful tool for revealing information content and sampling biases in the data and for setting priorities among taxa and areas for research and management. Integration also helps to highlight possible spatial interactions among neighboring regions and ecosystems.
Only a small fraction of existing biological data and information is now used in planning and decision-making. These activities require information that differs in kind and formats from what is typically generated by systematists and field biologists, (see below), and the results of academic and agency research are communicated poorly to decision-makers (NRC, 1993b). Managers complain that relevant information is scarce, hard to find, and scattered among many institutions. Furthermore, available information is often out of date and not applicable to local problems. Regional planners are stymied by the uneven geographic coverage of existing data and by the difficulty of integrating information produced by different agencies at different times with different methods of sampling, classification, and mapping.
There is wide agreement on the urgent need to organize existing biological information and make it more readily available and to coordinate future data collection and exchange (e.g., NRC, 1993a, b). This is a substantial task, and neither its importance nor its difficulty should be underestimated. Data on a wide array of topics, ranging from species distributions to socioeconomic activity, and from many disciplines and sources must be integrated. Only a tiny fraction of these data is now readily accessible for integration. Existing data are unevenly documented and do not constitute a representative sample of all the nation's biological resources.
Despite the size of the task, a national biological information system is clearly attainable. Although they are far smaller than what is proposed in this report, the experiences of ERIN in Australia (see Box 2.1), INBio (the National Biodiversity Institute) in Costa Rica, and a number of U.S. state initiatives show
Box 3.1: The Utility and Limitation of Species Distribution Data
The accompanying diagram and figures show how disparate spatial data can be integrated in a GIS to represent the distribution and conservation status of a species. The process is illustrated for the orange-throated whiptail (Cnemidophorus hyperythrus), a lizard native to coastal Southern California and Baja California. The species is considered threatened in California because it depends on coastal sage scrub, a habitat that has been severely reduced and fragmented by urban development. Together with the California Gnatcatcher and the Cactus Wren, the whiptail now serves as a "target species" for ongoing multispecies conservation planning efforts in the region (Hollander et al., in press).
Distribution data for the whiptail are exceptionally good because it is easy to observe, occurs in accessible areas, and has been sampled intensively during the course of recent environmental impact studies. Available information includes life-history notes, a coarse map of range limits, and a table of habitat preferences, which are provided by the California Wildlife Habitat Relationships (WHR) System (Zeiner et al. 1990); remotely sensed Thematic Mapper imagery of Orange, Riverside, and San Diego Counties (Panel A); digital maps of land ownership (Panel B); museum collection data and sightings spanning the past century (Panel C); and maps of suitable habitats as defined by vegetation in the WHR System and mapped at two different scales with 1990 Thematic Mapper satellite imagery (map boundaries from the California Gap Analysis) and 1991 aerial photography (regions with small crosshatching in Panel D were mapped as part of San Diego Counties Multispecies Conservation Plan-MSCP). The processing steps necessary to produce Panel D—which is a composite of field sighting data, habitat maps within the range limits of the species, and the location of existing nature reserves—are shown in the accompanying flow chart.
This example illustrates both the utility and limitations of existing information for mapping species distributions and assessing their conser-
vation status. Specifically, existing museum collection data vary in date and locational precision. In this case, 349 museum specimens could be mapped with certainty only to the nearest USGS 7.5-minute quadrangle, providing a crude representation of the range of the species.
Sighting data are both very dynamic and potentially very biased. Most observations of the lizard were collected after 1985 for projects requiring environmental impact reports. These sightings are clustered at the margins of urban and agricultural areas in San Diego County, where most development is occurring.
Habitat mapping and GIS modeling provide a means of extrapolating the potential distribution of the species. In this instance, we have highlighted areas within the range limits of the species that support vegetation classified as suitable habitat by a statewide habitat relationships system. The potential distribution is somewhat different, depending on whether 1:24,000 or 1:100,000 scale vegetation maps are used. Even the very general pattern predicted by WHR and vegetation data is suspect, because the small amount of life-history information available for the species suggests that it is limited more by its major food item, a termite, than by vegetation.
Compiling the information in Panels A-D took considerable time, effort, and technology, yet it provides only a sketch of the species distribution, to say nothing of abundance or trends. Most other species will be even more challenging than the whiptail, which is readily observable and has received a great deal of attention. However, with GIS relational modeling and display, the distribution of a species can be inferred by extrapolating along several lines of evidence. Different data sets can be compared and contrasted, helping to reveal biases and uncertainties. Further study of the species can be productively focused on poorly surveyed areas, and existing habitat models can be refined on the basis of patterns of agreement and disagreement between predicted and observed distributions.
that biological data collection, exchange, and integration are both practical and cost-effective. Nevertheless, an effective National Partnership will require substantial cultural and institutional changes to build stronger bridges among the broad spectrum of producers and users of biological data.
Recommendation 3-1: Under the leadership of the NBS, the National Partnership should develop a National Biotic Resource Information System. This should be a distributed federation of databases designed to make existing information more accessible and to establish mechanisms for efficient, coordinated collection and dissemination of new data and information. The NBS should take the lead in promoting standards for sampling, measurement, data recording, and data transfer. It should support continuing state efforts to develop regional and statewide environmental databases and should work closely with and support database development in museums, universities, and other appropriate organizations. It should also participate in interagency initiatives to coordinate collection and management of biodiversity data by the federal government.
The Need for Spatial Data
Until recently, most research biologists have received little training in cartography or spatial analysis. They might have consulted maps or even prepared them as part of their investigations, but most biologists view mapping as peripheral to their research. Most ecological field research has been conducted on small geographic scales, and collections and field plots have rarely been precisely located spatially (i.e., "georeferenced"). In contrast, maps are a mainstay of planners and decision makers, who must locate and quantify resources and site projects over large planning areas.
That difference in perspectives has contributed to the separation of research and management and has reduced the usefulness of much existing biological information for management purposes. However, the separation is rapidly disappearing as ecologists and conservation biologists pay increased attention to spatial heterogeneity at landscape and regional levels and to the role of spatial pattern in ecosystem functioning, species persistence, and the maintenance of biological diversity.
New technologies have revolutionized mapping over the last 20 years. They include aircraft-and satellite-borne remote sensors, global positioning systems for satellite-based location and navigation, and spatial-data handling tools, such as geographic information systems (GISs), visualization, and spatial-decision support systems. Maps, once primarily a vehicle for communicating research results, are now themselves used as data to derive new information. Furthermore, many spatial data are now acquired and stored in what has been referred to as "value-neutral" form (e.g., as actual measurements, rather than categories or assessments). While creating new opportunities for spatial analysis, these advances place a new burden on producers of spatial data for fuller documentation of data sources, data content, accuracy, scale, appropriate use, and other characteristics. This documentation is called "metadata"—e.g., data about data.
Those scientific and technological advances have led to rapid growth and evolution of a National Spatial Data Infrastructure (NSDI) that is leading to a far more accurate and detailed representation of earth features and phenomena (NRC, 1993a). Biologists lag well behind physical scientists in establishing coordinated efforts to develop standard spatial-data sets. For example, detailed maps of vegetation do not exist for the United States, nor for many of its regions. We lack even reliable range maps for most organisms, including many vertebrates and plants. The NPBS will need to remedy these deficiencies to take full advantage of current and future capabilities in spatial mapping.
Recommendation 3-2: Under the leadership of the NBS, the National Partnership should recognize and participate actively in the evolving National Spatial Data Infrastructure. It should promote greater awareness and use of existing spatial data and technologies; increase efforts to locate field data spatially; adopt, where appropriate, existing standards for mapping and spatial-data handling; and increase the involvement of biologists in federal efforts to develop spatial-data and metadata standards.
Specific Needs Include
Increased training of biologists in modern spatial analytical theory and spatial statistics (Levin, 1993).
Fuller use of existing spatial data and technologies, such as global positioning systems, remote sensing, and GISs for biological survey and monitoring, specifically to achieve fuller coordination of ground sampling and mapping activities through carefully designed, multistage mapping and monitoring schemes. This will require a new level of interaction between traditional field biologists and systematists, landscape ecologists, and earth-system scientists and increased federal and state support for application of remote sensing to biological survey and monitoring, especially by the National Aeronautics and Space Administration (NASA) and the Department of the Interior.
Increased standardization of collection and documentation of spatial biological data, including development of a detailed model for biodiversity data (i.e., defining biodiversity variables and their possible transformations, functional relationships, computer representations, and documentation) and then establishment, adoption, and implementation of data and metadata standards for describing, classifying, and mapping biological features, as described by the Interior Geographic Data Committee (1992).
COORDINATION AND MANAGEMENT OF DATA AND INFORMATION
A Federated Database Approach
Biological information is produced and maintained by diverse federal, state, and local agencies and private organizations. Collectively, this is a vast effort, and no organization or centralized facility could effectively compile, maintain, and distribute all relevant information. A more realistic goal is to link existing and new biological databases into a distributed federation of NPBS databases. Such a linkage must occur both physically over networks and logically through the use of appropriate software and data standards.
Many pieces of a distributed biological database are already in place or under development, notably taxonomic databases (such as the Flora of North America), conservation databases (such as the Natural Heritage and Gap Analysis databases), monitoring data (such as the National Science Foundation's [NSF] Long Term Ecological Research [LTER] sites and the Breeding Bird Survey), and the databases maintained by statewide natural history surveys (such as those in Illinois and Kansas).
Consistency and documentation of data content, quality, performance, and exchange are essential for ensuring the usability of scientific data. Data that are to be useful decades after they are collected have special documentation requirements (Bowser, 1986). The integration of existing biological databases is seriously impeded by the lack of data standards. To be effective, the development of new standards must build from heterogeneous sources and be phased to allow existing data to be transferred to a more uniform database environment.
Recommendation 3-3: The NPBS should develop and adopt appropriate standards for data quality assurance and quality
control and metadata content. The standards should be established by federal agencies in close collaboration with states and the private sector and should build on current efforts, such as those of the Federal Geographic Data Committee, the Interior Geographic Data Committee, the Interagency Task Force on Water Quality Monitoring, the Freshwater Imperative, the Interagency Working Group on Data Management for Global Change, and Systematics Agenda 2000.
Environmental and socioeconomic databases are evolving rapidly through numerous state efforts, as well as such federal initiatives as the U.S. Global Change Data and Information System (GCDIS); the NSF LTER network; the NASA Earth Observing System Data and Information System (EOSDIS); the Environmental Protection Agency (EPA) Environmental Monitoring and Assessment Program (EMAP); EPA's revised water-quality information system, STORET; the Census Bureau's Topologically Integrated Geographic Encoding and Referencing system; the National Oceanic and Atmospheric Administration's climatological, meteorological, and oceanographic databases; the U.S. Geological Survey National Mapping Division and National Water Quality Information System; and the Department of Energy's Carbon Dioxide Information Analysis Center database. Those are some of the major building blocks of the National Spatial Data Infrastructure referred to above. They have developed independently of most biological databases, notably the museum collections and state biological surveys.
Recommendation 3-4: The NPBS should develop its databases in conjunction with the major federal environmental and socioeconomic databases to minimize redundancy, to avoid conflicting terminology and classification systems, and to maintain consistent data standards and formats.
Limits to Federation
Much of the information generated and managed by the NPBS will be housed in and exchanged over computer networks. Many producers and consumers of NPBS data do not yet have network access, but most will have it soon. Already, 5-10 million people in 50 countries are linked via Internet.
Free sharing of data raises issues of proprietorship and appropriate use. Some individuals and institutions distribute primary data only after they have had time to conduct their own analyses and publish the results. Even more important, integrating data can require a commitment of time and resources far beyond what most institutions can afford.
Recommendation 3-5: Full and open sharing of biological data on a timely basis should be an objective of NPBS data management, and adequate funds should be made available to meet this objective. The objective should apply foremost to data acquired with public funds. Conventions and protocols for sharing of primary data should be developed cooperatively among NPBS participants. Under NBS leadership, the NPBS should move quickly to produce a data and information policy for establishment, maintenance, and distribution of long-term national and international biological data and derived information.
Learning from Others
The human-genome project has been described as the first truly ''big-science'' project in biology. It has as its primary product, huge databases of complex genetic information. The National Center for Atmospheric Research stores and manipulates large volumes of data that can readily be visually browsed and retrieved across Internet. The four national supercomputers supported by
NSF manage large amounts of astronomical data. When fully functional, NASA's EOSDIS will process at least 30,000 digital images per day while managing or linking to a wide range of biological and geophysical databases.
Recommendation 3-6: NPBS data managers should consult with the Human Genome Project, EOSDIS information scientists, the National Center for Atmospheric Research, the NSF supercomputer centers, and other large database programs to take advantage of the lessons and products from these efforts to manage large volumes of biological and geophysical data and to expedite the development of an effective, distributed database environment.
Data quality assurance and control are best achieved through the clear assignment of custodianship for subsets of the data to appropriate experts or organizations. Curators must be appointed with the responsibility for overseeing the long-term quality, currency, and consistency of the data. That will help to transform databases from passive collections of information into something more similar to the scientific literature, which relies upon an active set of editors and reviewers to ensure its quality. To support the custodians, tools must be developed to encourage those who contribute to databases but do not have curatorial obligations to take responsibility for the quality of their submissions. Examples of custodial tools include standardized submission formats and intelligent software for screening submitted data.
Recommendation 3-7: NPBS data managers should develop mechanisms to ensure the clear assignment of custodial relationships for data sets and to develop appropriate support tools for the custodians.
The utility of data on status and trends will increase with the length of the record. A National Biotic Resources Information System will be most useful if it can archive biological data for many years or even centuries. A highly distributed system with many different custodians, as proposed here, is potentially vulnerable to loss of a dataset if the custodianship falters. Mechanisms should be developed to ensure that the data archives of NPBS custodians are preserved beyond the life span of individuals and institutions. For example, regional and national sites might have to be designated as repositories for data for which custodians cease to exist.
Recommendation 3-8: NPBS data managers should identify a small network of organizations to form the core of a national biological archive for data that merit preservation but have no active custodian. The NPBS should also develop clear guidelines for documentation, storage, and retrieval of these data.
Functional Requirements for NPBS Data Management
To be effective and useful, distributed databases must meet a number of functional requirements, among which the key ones are the following:
Network interfaces. Distributed databases work best when the data are available through standard interfaces over networks, such as Internet. Effective interfaces between databases can be developed only if all the databases have a standard means of allowing other computer programs to interact with them. Public-domain programs that are now appearing—such as GOPHER, MOSAIC, WAIS, and WWW—make it possible to provide integrated access into multiple databases easily and inexpensively.
Distributed queries. A user of NPBS information should be able to query multiple databases. Success depends on the degree of consistency among databases regarding taxonomic conventions, use of terms, geographic addressing schemes, data-management systems, and analytical software. Users should not be required to know all synonyms and geographic transformations, nor should they be expected to maintain many functionally equivalent software packages. Meeting the goal of a fully distributed database will require adoption of standard names and terms (or system-supported synonyms) and vendor-independent data representations.
Queries on different scales and levels of organization. Users of NPBS information systems will want to query the systems at different conceptual levels. For example, how many cricetine rodent species live in Ingham County? In Michigan? In the United States? What are the status and trends of fish populations of the Upper Peninsula of Michigan? Which of these populations are endangered? Supporting those different queries will require that the survey information system support hierarchical classifications so that questions can be posed and answered easily at different taxonomic or geographical levels. Because the classifications will change over time, they must be implemented in a manner that allows easy modification.
Vertical and horizontal integration of spatial data. Relevant information on biological resources must flow to local and county decision-makers and resource managers. Information must also support larger-scale analyses to place local actions in perspective and to address regional and national management issues and those which span administrative boundaries. Meeting these needs will sometimes require vertical integration—combining information about a place from many scales (e.g., species collection points, 1:24,000 vegetation maps, and 1:100,000 soil maps); information must be brought to a common scale of representation for analysis. Alternatively, information might be available on one scale but from different sources for different parts of the area being examined; this information requires horizontal integration—information pieced together with a common classification system and
definitions. In other words, information must be managed to support vertical integration (across space and time scales) and horizontal integration (across space and time on a fixed scale).
Flexible system design. Modern hardware and software systems have useful lives of at most a few years. Therefore, NPBS databases must be designed in a way that will facilitate their transfer across multiple future computer platforms. That is best achieved by using, where possible, commercially available products that adhere to extant standards. Custom software can be difficult, sometimes impossible, to transport to new systems. The National Institute of Standards and Technology has substantial expertise in this regard and should be consulted in the planning of the NPBS data system.
Recommendation 3-9: NPBS data managers should meet the key functional requirements for an effective distributed database. Among them are standard interfaces, an ability to query multiple databases easily and across taxonomic and geographic scales, and easy transport of the data to new computer platforms as they are developed.
INFORMATION MANAGEMENT IN NBS
Consolidation and Coordination of DOI Survey and Monitoring Data
DOI has proposed initiating a National Biological Status and Trends (NBST) program at the NBS. It will be a coordinated monitoring program based on existing activities in DOI, including the National Wetlands Inventory, the Waterfowl Inventory, the Park Service Inventory and Monitoring Program, the Gap Analysis Program, the Breeding Bird Survey, the Global Change Research Program, Biomonitoring of Environmental Status and Trends, and the Great Lakes Fisheries Assessment. DOI proposes
to initiate the program by consolidating databases, designing a statistically valid monitoring network capable of biennially assessing biological status and trends, developing standardized data collection and management protocols, and establishing rigorous quality-assurance and quality-control procedures.
The committee believes that such an effort is central to the NBS and strongly endorses the concept, but the program's scope should evolve as the NBS and other DOI survey and monitoring activities are extended to other taxonomic groups, ecosystem types, and geographic areas, as discussed in Chapter 2. The NBST program should include the management, archiving, and analysis of DOI survey and monitoring data of national extent or significance. At the same time, the NBST program must be coordinated with other important monitoring programs outside DOI. Continuing efforts like the joint Fish and Wildlife Service-Smithsonian initiative, Standard Methods for Measuring Biological Diversity, should be accompanied by similar initiatives concerned with management, exchange, and integration of biological survey and monitoring data, such as The Nature Conservancy's Heritage Program, and broadly based projects, such as the Flora of North America. The integrated management schemes being developed for southern Florida (see Box 2.3) and California (see Box 3.2) constitute excellent regional examples of the need for infrastructure sufficient to bring together status and trends data of organisms and conditions in disparate environments.
Recommendation 3-10: The NBS should establish a National Biological Status and Trends (NBST) program that builds from existing DOI survey and monitoring programs eventually to include a broader range of taxonomic groups, ecosystem types, and geographic areas. At the outset, staff should be appointed and mechanisms developed to ensure that the NBST program is effectively linked to related federal survey and monitoring initiatives and to other pertinent databases.
Box 3.2: New Approaches to Information Management and Conservation Planning in California
California's biodiversity is extraordinary for its richness (e.g., 5,862 flowering plant species, or one-fourth of all flowering plant species in the conterminous United States) and a high level of endemism (e.g., 24% of all flowering plant species). Unfortunately, this biodiversity has been seriously threatened by human activities, and the state leads the nation in federally listed threatened and endangered species.
As with other states, there is no single repository for data on California's biodiversity. The data are scattered across many institutions, and general standards have not been established for data collection, classification, or transfer or for database design. The state of California is now moving rapidly toward coordinated collection and management of biodiversity data. In 1991, a Biodiversity Memorandum of Understanding was signed by major federal and state agencies, with the intent of promoting interagency cooperation in conserving biodiversity across administrative boundaries. A California Council on Biological Diversity, established to promote this new approach, was charged to
To meet those goals, the council has divided the state into 10 "bioregions" that are defined mainly by physiographic and biogeographic features. A new process, Natural Community Conservation Planning (NCCP), has as its objective active conservation of native plants and animals and their habitats through an expanded system of natural reserves.
An NCCP pilot project in the threatened coastal sage-scrub natural community of southern California has recently been completed; it exemplifies the kind of effort required to implement proactive conservation strategies in areas under intense population and economic pressure. Applying GIS modeling to available socioeconomic and ecological data, a scientific review panel defined core and satellite habitat areas on the basis of the extent and quality of coastal sage-scrub vegetation and identified three target species—the California Gnatcatcher, the Cactus Wren, and the orange-throated whiptail lizard (Box 3.1)—to help efforts to plan and design reserves. The panel also attempted to define the extent and location of allowable development of remaining habitats. Finally, a research agenda was proposed to help resolve unanswered questions bearing on conservation of the community and its component biodiversity. It addressed six basic subjects: biogeographic inventory and mapping, monitoring of trends in selected taxa, dispersal characteristics and corridor-use patterns of target and other animal species, demography and population viability analysis, survey and autecology of sensitive plants and animals, and genetic studies.
The NCCP project in southern California underscores several user needs that the NBS must meet if it is to be effective: the need for better organization of and access to highly dispersed biological data; the need for information at all levels of biological organization, ranging from genes to communities; the need for distribution and monitoring data on species and habitats at both local and regional scales; the need to account for socioeconomic pressures and trends; and the critical need for information on spatial characteristics such, as species' dispersal traits and effects of habitat patch size and arrangement on population viability.
NBS COORDINATION OF A NATIONAL DISTRIBUTED DATABASE
Most biodiversity data are produced locally and are dynamic. Therefore, the NBS should not attempt to develop a large, central-
ized database facility to store and distribute biodiversity data. Instead, both the National Partnership and the NBS specifically should evolve a highly distributed network of databases connected through the Internet and by other appropriate means. Because most biodiversity data are maintained within states or by local jurisdictions and institutions, the NPBS will always be a highly decentralized, largely bottom-up effort. It is critical that this effort be well coordinated to minimize redundancy and to allow integration of data, but there will probably never be a single database of databases that will direct users to all relevant sources. The ERIN approach (as described in Chapter 2) was logical for Australia, where relatively few electronic databases existed earlier, but the United States already has a large number of such databases. Instead, data-sharing must be fostered through creative use of network tools. The NBS can contribute to this effort by taking the following steps:
Recommendation 3-11: The NPBS and the NBS should develop a highly distributed federation of databases, rather than a larger centralized database facility. To facilitate the sharing of data among the databases, the NBS should:
Establish a moderate-sized facility with personnel and computing capabilities for archiving and distributing regional and national NBS data sets and for meeting the proposed goals of a National Biological Status and Trends program.
Use existing tools and develop new ones to help NPBS users locate data and information, both digital and nondigital. For example, the NBS should explore the establishment of region-specific and program-specific NBS database directories and a national directory of these directories that are accessible through powerful national information networks, such as Internet.
Promote the adoption of data standards by NPBS partners by cooperating with existing efforts (e.g., ABC, 1993), by
convening technical workshops and committees, and by providing leadership in developing and applying new standards for biological data.
The National Partnership will supply data and information ranging from raw measurements to complex analysis and interpretation, and the results will be published in maps, graphical products, and reports. The Partnership (including the NBS) should use existing media and channels of information exchange whenever possible, including scientific journals, other publications, extension services (e.g., those of the U.S. Geological Survey and the U.S. Department of Agriculture), and information networks. Carefully reviewed technical reports containing original data or their interpretation should also be prepared as needed. Primary printed products must include map and graphical representations of spatial or time-series data for biogeographic areas, ecosystems, habitats, and species.
To ensure appropriate use, data must be subject to strong quality assurance and control. Published data sets should receive professional review before they are distributed. General guidelines for documentation and review of biological data are few, and development of such guidelines should be given high priority. For reasons of efficiency and utility, dissemination of NPBS information should be structured in the context of user interest in particular regions, habitats, or taxa. That means that users will have to define their queries precisely and consistently, including the level of detail and scale. In answering those queries, NPBS information should be presented to highlight what is and is not known by means of diverse, user-friendly formats that are readily accessible to users of varied background and expertise. That will help to ensure correct interpretation and appropriate use. For example, the implications of the reported presence or absence of a
given species depend greatly on the nature of the studies on which the report is based.
Software tools must be responsive to and readily accessible to all users, from both the public and the private sectors. Needs range from primary data to various kinds of derived products.
Recommendation 3-12: To facilitate use of the data that it generates, the National Partnership should develop software tools for data visualization and analysis, for data reformatting and conversion, for trend analysis of monitoring data, for spatial interpolation of sighting and collection data, and for GIS habitat modeling.
Assessments and Communication of Results to Decision-Makers
One of the challenges that the National Partnership must address is the effective communication of research results to resource managers, planners, and legislators. Much more is required than data summaries, technical reports, and professional articles. Those users have questions whose answers are not obvious in existing research products. Rather, policy and management actions often require reinterpretation of existing information. It is here that data and information are most vulnerable to inappropriate use and interpretation and that scientifically based assessments are extremely valuable. These assessments should be based on protocols developed to ensure that they clearly state, to the extent possible and in nontechnical language, such information as the current scientific understanding of the issues, the scientific uncertainties, the predicted consequences of possible actions, the uncertainties associated with those predictions, and what additional information is likely to reduce the uncertainties most rapidly. Deciding on what should be included in such assessments is not a trivial task, and it will take committees of scientists, natural
resource managers, and decision-makers to determine what information is possible and how such information should be presented.
Recommendation 3-13: The National Partnership should develop scientifically based protocols for the preparation of biological assessments for decision-makers.
Other possible mechanisms for providing such guidance include ensuring ready access to the scientists who produced the information, identifying other qualified scientists through directories of scientific expertise (using such institutions as State Heritage Programs and biological surveys), and forming standing scientific councils to support local and regional planning and decision-makers. In addition, clear standards for documenting biological datasets (metadata) should be implemented. Complete and accurate metadata are the only means of ensuring that researchers decades from now can use current data effectively.
Ensuring Scientific Quality
Information provided by the National Partnership must be of the highest scientific standards to provide the basis for objective and impartial management decisions and policy. The best ways to ensure such quality are to develop and implement standards as proposed in Recommendations 3-1, 3-3, and 3-11; and to use mechanisms to ensure the scientific review of information by appropriate professionals before it is disseminated to users. That will require a large increase in formal review by scientists both within the NBS and in the NPBS scientific community.
Recommendation 3-14: The NBS and other appropriate participants in the NPBS should adopt measures to ensure a uniformly high quality of information analyzed, interpreted,
or disseminated by them. Technical reports, articles submitted for publication in scientific journals, and similar products should be subject to peer review.
Achieving Widespread NPBS Product Communication
The National Partnership must be highly responsive to public and private-sector needs for primary data and derived products. Existing protocols for retrieving primary data can be cumbersome and slow. Where networks have been used to share data or products (e.g., the Gap Analysis Program and Breeding Bird Survey trend analyses), use of the information has increased substantially. Online data dictionaries must be adequate to guide appropriate use of the data. The NPBS should also maintain an online bibliography of its publications and project descriptions. Access should be facilitated by user-friendly interfaces with intelligent search and browse capabilities.
The U.S. Global Change Research Program has considered these and many other issues related to data management and exchange and has drafted a set of policy statements (Committee on Earth Sciences, 1992). These policies articulate a commitment to establishing and maintaining long-term data sets, full and open sharing of the full suite of global data, data archiving, data standards, data access and affordability, and clear definition of the duration of the period in which scientists retain exclusive use of original data. These policies extend beyond primary data to include enhanced data products that are especially useful to users outside the scientific research community. The Global Change Research Program data-management policies are consistent with those adopted by the Data and Information System for the International Geosphere-Biosphere Programme (IGBP) and could serve as the core of an NPBS data and information policy.
The NPBS should develop a plan for publication of selected,
peer-reviewed, regional and national primary biological data sets in CD-ROM form or in other digital media. The Partnership should also expand electronic publication of data summaries in a form that supports management applications. Selected regional and national databases and data summaries should be published in digital formats.
Many users of biological data do not have network access or adequate hardware and software. Even those who do often require information in other forms. The NPBS must continue to make print products available. The most important of those are peer-reviewed scientific articles and reports. Examples of end products in printed or electronic format that are of fundamental practical value to those who manage, use, or study biological resources include taxonomic monographs and revisions; regional floras; faunas; field guides and manuals; detailed range and habitat maps; atlases of the distribution and trends in regional biodiversity, ecological communities, and ecological systems; and scientific studies that interpret existing biological diversity in light of evolutionary history.
Recommendation 3-15: To meet the growing needs of all sectors of society for biodiversity data and information, the National Partnership should increase its capability for publication and product communication substantially. That should include increased online access to data, reports, and bibliographies; publication of selected data sets in CD-ROM format and in other media; and expanded publication of synthetic documents, such as atlases and summaries of surveys and trends.