Introduction to the Gap Analysis Program

Conservation of biodiversity requires that land owners, managers, planners, and decision makers know what biological resources occur in their jurisdiction and the conservation status of those resources. The Gap Analysis Program of the USGS Biological Resources Division (BRD) is the only nationwide program to produce this data and assessment.

Gap analysis is a scientific method for identifying the degree to which native animal species and natural communities are represented in our present-day mix of conservation lands. Those species and communities not adequately represented in the existing network of conservation lands constitute conservation "gaps." The purpose of the Gap Analysis Program (GAP) is to provide broad geographic information on the status of ordinary species (those not threatened with extinction or naturally rare) and their habitats in order to provide land managers, planners, scientists, and policy makers with the information they need to make better-informed decisions.

To achieve this, GAP is the first state- and national-level effort to complete the following:

GAP is one of the most demanding GIS projects ever launched with full involvement of remote sensing technology for land cover mapping, GIS modeling and spatial analysis, and database storage and analysis. Delivery of data over the internet and by CD-ROM is also pushing the electronic publishing frontier.

Development of the biological data sets and stewardship maps (ownership and management status for biodiversity conservation), provides the public with estimates of the representation of plant communities and animal species (elements) on lands managed for biodiversity maintenance. "Gaps" are those elements that are insufficiently represented and may be at risk of endangerment in the future unless changes in their management status are made. The data sets have thus far been used for numerous applications for both conservation and development planning, as well as scientific research. GAP is conducted in cooperation with over 400 institutions from all state and federal land management agencies, academia, non-profit and industry groups.

The individual GAP projects conduct the project using a variety of software packages including Esri ArcInfo, ERDAS, Intergraph, SPECTRUM, PCI, ORACLE, DBASE, and GRASS, among others. The purpose of this presentation is to provide participants with an overview of the program structure, status, operation, goals, methods, and products. We will identify methods to access and use the data, including access through the Internet.

INTRODUCTION

The purpose of this document is to provide the reader with an overview of the function and structure of the National Gap Analysis Program. While assumptions and limitations and operations of the program are presented, detailed descriptions of methods are not included since they are provided elsewhere, such as in the GAP handbook, bulletins, annual status reports, and peer-reviewed literature which are available from the GAP home page at http://www.gap.uidaho.edu.

Mission

The mission of the Gap Analysis Program (GAP) is to provide state, regional, and national assessments of the conservation status of native species and natural land cover types of the U.S. and to facilitate the application of this information to land management activities.

The Issue

The loss of biodiversity is a major concern for natural resource managers and conservation biologists. There is uncertainty and debate about the number of species being lost or at risk of extinction (Lugo 1988). However, that the number of species at risk has increased dramatically since the Endangered Species Act was passed in 1976 is unquestioned. There is a growing consensus that programs designed to prevent the extinction of endangered and threatened species do not address the larger problem of fragmentation, loss of habitat, and disruption of ecological processes, all of which contribute to a number of once common species being placed on the endangered species list (Noss et al. 1995). A "coarse-filter" approach is needed to complement the "fine-filter" approach of the Endangered Species Act (Hutto 1987, Scott et al. 1987). While considerable information has been gathered by state, federal, and private groups to implement recovery actions for species that are in immediate danger of extinction or threatened with the possibility of extinction, there is no comparable information available for the vast majority of species and ecosystems not yet endangered but potentially at risk. There is no standard landscape-level spatial data available for ordinary species, yet we are continually surprised by new endangered species crises. Until GAP began mapping land cover at levels of spatial and thematic resolution meaningful to the management of biodiversity, even basic land cover maps have not been available.

Without land cover maps at least as detailed as the dominant cover type, at a minimum of a 1:250,000 scale, there is little hope of objectively implementing a coarse-filter approach to the conservation of biodiversity. Orians (1993), for example, offers a discussion of the level of detail needed. The lack of mesoscale maps of vegetation types for states and regions has been due in part to lack of technology to create such maps. For example, past failures to assess environmental issues over large areas was partly because of the absence of remotely sensed imagery and the technology capable of creating and analyzing such large data sets. That changed with the launching of the Landsat satellites, the creation of off-the-shelf geographic information systems (Davis et al. 1990, Scott et al. 1987), higher capacity computers, and inexpensive mass data storage.

Gap analysis was developed as a proactive coarse-filter approach to protecting biodiversity (Scott et al. 1987, 1993). It has attempted to make full use of recent advances in remote sensing and geographic information systems. Gap analysis provides an overview of the distribution of terrestrial vertebrates and land cover types, relative to special management areas and land ownership, at a scale useful to land managers for making policy decisions and to ecologists for placing small study areas in an ecological context. GAP seeks to identify "gaps" (i.e., vegetation types or species not adequately represented in areas managed for long term maintenance of natural systems) that may be filled through changes in land management practices. GAP researchers use terrestrial vertebrates and vegetation alliances as indicators of, or surrogates for, biodiversity (Austin and Margules 1986, Scott 1993, National Gap Analysis Program 1994, Csuti and Kiester 1996, Noss and Cooperrider 1994, Jennings 1996). Digital maps of these elements of diversity are overlain in a GIS with maps of areas managed for biodiversity and land ownership to identify those that are under represented in the existing network of areas. It is not a substitute for intensive localized inventories, nor a replacement for traditional single-species approaches to protecting diversity. GAP permits quick, graphic display of basic data sets that can be used to identify opportunities for future management activities that will reduce future resource conflicts. We hypothesize that gap analysis, by focusing on higher levels of biological organization, will be both cheaper and more likely to succeed than conservation programs focused on single species or populations (Scott et al. 1993), while providing an ecologically-based sampling universe for more detailed studies of, for example, composition, structure, and function of species and natural community alliances.

A detailed initial description of the gap analysis concept can be found in Scott et al. (1993). Methods for conducting a state GAP project can be found in the Handbook for Gap Analysis (National Gap Analysis Program 1994). Substantial amounts of additional information about GAP can be found in the five GAP Bulletins (1991-1996), the 1994 and 1995 GAP Status Report (Jennings and others 1996), the book "Gap Analysis: A landscape approach to biodiversity planning" (Scott et al. eds. 1996) published by the American Society for Photogrammetry and Remote Sensing, the GAP home page (http://www.gap.uidaho.edu/gap), or any one of the many state GAP home pages linked through the national program page, and in over 150 peer-reviewed journal articles written directly about some aspect of GAP since 1987 (for obtaining this bibliography from the GAP home page, see "GAP Handbook, Standards and References," then select "Gap Analysis References," or use the following URL: http://www.gap.uidaho.edu/gap/litur.html).

Program Background

The development of methods for gap analysis began in 1987 in response to the need to complement species-by-species management in dealing with broad-spectrum habitat loss (Scott et al. 1987). There was a need for synoptic and geographically explicit information on the distribution of each native vertebrate species, natural community, and their management status. At the time, there were no readily available consistent data that could provide for an understanding of either a single land management decision or the occurrence of a species' habitat in the ecological contexts of landscapes or bioregions, nor the rangewide expressions of a "biodiversity element," such as a particular species or natural community alliance.

In the ensuing years since 1988, significant barriers to mapping elements of biological diversity across large areas have been overcome. A wide range of tools for mapping natural land cover and habitat types and predicting vertebrate species distributions has emerged, and procedures have been refined, tested, and further refined. Further development and testing of methods (for example, accuracy assessment), though, is still needed. A wide base of consensus and cooperation has been developed, including, for example, the classification of natural communities, a consistent set of satellite images from which to render digital base maps, and ways in which GAP information is applied to everyday resource decisions and long-range planning.

PROGRAM DESCRIPTION

Goals and Objectives

The goals and objectives of GAP are:

Goal 1: Provide focus and direction for conservation efforts with explicit documentation of under represented elements of biodiversity, realizing that only certain selected elements of biodiversity are presently known and can be mapped.

Objectives: a.Map the current land cover types of the United State to the alliance level (dominant/ co-dominant cover type; FGDC 1996), or groups of undifferentiated alliances where necessary, at a minimum mapping unit no larger than 100 hectares. b.Predict the distribution of animal species within their accepted geographic ranges. c.Determine the area occupied by alliance types and vertebrate species that are within lands managed for biodiversity. d.Determine the degree of public agency ownership of land areas having under-represented species and alliance types. e.Identify gaps in the natural area reserve system of the United States. f.Assess accuracy of various theme maps using independent contractors. g.Update data sets every five to ten years or when appropriate. h.Begin development of methods for applications of the gap analysis concept to other components of biodiversity, such as in aquatic environments and for plant species. i.Conduct research necessary to improve mapping methods, analyses, and dissemination (see page for a description of and goals for the GAP research component).

Goal 2: Maximize the utility of the GAP effort.

Objectives: a.Improve communication and coordination among natural resources agencies by: 1.Fostering joint participation in building GAP data sets 2.Using the same information for resource analysis and decision making 3.Standardizing definitions and classifications of ecosystem components. b.Facilitate development of institutional capability for ecosystem and landscape inventory, monitoring, research, analysis, planning, and information transfer at local, state, and federal levels. c.Provide the basis for a predictable and integrated strategy for managing natural biodiversity in the U.S.

Assumptions and Limitations

This section addresses the conceptual, technical, and organizational assumptions and limitations of the program.

Conceptual

The most fundamental assumption the Gap Analysis Program makes is that the best time to ensure that a species does not become endangered with extinction is while it is still relatively common. Waiting until a species is actually endangered or threatened with extinction results in reactive management activities that are expensive, exhibit a low probability of success (Tear et al. 1993), and are socially and politically divisive. This assumption underpins the GAP method of assessing the present-day conservation status of each vertebrate species as well as floristically-defined habitat types (community alliances). Additional assumptions and limitations of gap analysis are discussed below.

Lower cost by avoiding crises

Gap analysis assumes that the cost of maintaining species in their natural state, while they are relatively common and part of self-sustaining ecosystems, is less than the cost of intensive management programs needed to save species that are at the brink of extinction (Scott et al. 1987). An efficient way to avoid extinction crises is to work with the many different institutions private and public that are involved with land-use management and land-use planning to develop large-area geographic information on certain indicators, or surrogates, for overall biodiversity. Such information, developed with systematic and consistent methods, is then applied to land-use and resource management decisions via the ongoing activities of many different decision makers, whether incrementally small everyday decisions such as zoning permits or decisions of broader scope such as state land-use planning.

Use of surrogates for biodiversity

A perfect surrogate for modeling all biodiversity does not exist. It is therefore necessary to use less-than-optimal surrogates to model biodiversity. The two biodiversity surrogates used by gap analysis are vertebrate species and community alliances (primarily dominant vegetation types; see the GAP handbook [National Gap Analysis Program 1994] or Grossman et al. [1994] for more on alliances). Vertebrate species are used because of their intrinsic importance, their major role in community patterns and processes (Terborgh 1988), and because mapping their distributions at a practical and useful scale is tractable (Scott et al. 1993). Dominant vegetation types are used because patterns of natural terrestrial land cover are an integrated reflection of the physical and chemical factors that shape the environment of a given land area (Whittaker 1965). They also are determinants for overall biological diversity (Franklin 1993, Levin 1981, Noss 1990) as their structure and composition significantly affects species-level interactions. Community alliances are the finest level of biotic assemblages that can be described and mapped over large areas using remotely sensed imagery (although technical limitations to mapping every alliance type remain, see the section on technical assumptions, below). Dominant vegetation types are the constituent parts of landscapes and can be used as a currency for habitat types in conservation evaluations (Fenner 1975, Austin 1991). As such, descriptions of dominant natural land cover types and maps of their distributions offer an equivalent set of land cover mapping units which can be categorized and the categories used for research, planning, and management of natural biological resources (Jennings 1996).

The degree to which vertebrate species and community alliances are surrogates for invertebrates, fungi, or plant species remains unexplored at the level of resolution that GAP is using. Some correlation between certain invertebrates and vegetation types has been shown by Halbert et al. (1995). Prendergast et al. (1993) found only low correlation among areas of high diversity for birds, butterflies, dragonflies, liverworts, and aquatic plants. This study focused on the potential overlap among "hotspots" of diversity for these groups of organisms. However, the biogeographic relatedness of taxa-specific species richness is fundamentally a different concept than whether or not or the degree to which natural communities capture the biological diversity of non vertebrate species.

Taxonomic biases and useful spatial framework

Another assumption is that while focusing on species and dominant vegetation types will result in conclusions that are biased to these elements of biodiversity, this approach provides a synoptic spatial framework for linking information which is finer as well as coarser in both thematic description and spatial resolution. For example, maps of species distributions or habitat types produced for GAP can provide an ecological and geographic context for stand or plot data measuring population or genetic criteria while directly linking these representations to continent-level measurement of biome criteria.

Hierarchy theory for ecology

The mechanisms by which changes to the extent and distribution of plants and animals are taking place appear complex because of the hierarchical relationships between ecological systems of differing spatial-temporal extents (such as organisms, populations, species, communities, landscapes, or regions). Alterations to land and water characteristics, formerly limited in extent to populations and species, are now manifest at the scale by which natural communities and landscapes function (Noss et al. 1995).

Because the dynamics of larger systems (e.g., landscapes) constrain the behaviors and occurrences of the smaller systems that they encompass (e.g., populations or species) by means that are independent of the smaller systems (O'Neill et al. 1986), conservation efforts implemented at the population or species level are not effective when system-wide changes are being forced at the landscape level. Furthermore, the mechanisms, or "emergent properties," by which a system interacts with the forcing variables cannot be identified by a simple aggregation of its smaller components nor by a reduction of its larger components (Allen and Starr 1982). In order to slow the loss of our biological resources, the basis for solving problems and implementing decisions must be predicated on information that is extracted from the level at which the changes are being induced, in this case communities and landscapes. The most ambitious application of the hierarchical concept is the National Gap Analysis Program (O'Neill 1996).

Not a substitute for ESA or Heritage

Gap analysis is not a substitute for endangered and threatened species management, planning, and research (for example, as is currently provided under the Endangered Species Act, state Natural Heritage Programs, or similar activities). Although the program seeks to reduce the rate at which species are becoming endangered or threatened, intensive efforts to maintain those species already imperilled are vital.

Not a national survey

Finally, the program is not a thorough nationwide inventory of biological resources. Gap analysis was developed in response to rapid habitat loss and to meet the need for information critical to natural resources management agencies. It is a relatively quick and reliable assessment of the conservation status of vertebrate species and habitat types across large areas. It is needed to provide a coordinated focus and direction for the many different organizations involved in land management and land-use planning.

The process of improving knowledge in systematics and biogeography, in all their dimensions, is a complex one vital to the interests of the Unites States and all nations. The Gap Analysis Program seeks to support and expedite that process by stratifying our land area into sampling units according to attributes of dominant vegetation types and species distributions.

Technical

Maps are a display of a database

All maps are abstractions and generalizations of reality and can be thought of as testable hypotheses. The gap analysis maps of land cover are not produced by classical cartography, rather they are two-dimensional representations of large digital data sets that are manipulated by computerized database functions. As such, these data sets can be continually tested for areas of weakness and improved upon with better information.

Contextual information

In conducting land cover mapping, GAP assumes that large-area (i.e., an entire state at a time) contextual information has greater importance for GAP purposes than higher-resolution site-specific information. To this end, the program relies on Landsat Thematic Mapper (TM) satellite images to generate a digital base map and from which initial land cover patterns are delineated.

Many other sources of land cover information are then used to interpret and refine the TM-derived spatial data. These include air photos, air video, other maps, and field observations. Thematic Mapper data are used because they have: (a) a high signal-to-noise ratio; (b) desirable radiometric bands and bandwidths; (c) good cartographic accuracy; and (d) appropriate ground resolution relative to large-area mapping.

Standard Land Cover Classification

The Gap Analysis Program assumes that a standardized land cover classification system is critical to the development of data sets that cover more than one state. Pursuant to this assumption, the program has worked with its partners to develop a National Vegetation Classification System (FGDC 1996). The following land cover classification criteria were assumed as basic requirements: (a) an ability to distinguish areas of different actual dominant vegetation; (b) a utility for modeling vertebrate species habitats; (c) a suitability for use within and among biogeographic regions; (d) an applicability to Landsat Thematic Mapper (TM) imagery for both rendering a base map and from which to extract basic patterns; (e) a framework that can interface with classification systems used by other organizations and nations to the greatest extent possible; and (f) a capability to fit, both categorically and spatially with non-natural areas such as agricultural and built environments.

Minimum Mapping Unit

The GAP data sets are produced at a nominal scale of 1:100,000. They do not show habitats or features smaller than the minimum mapping unit (MMU), which in most cases is 30 m2 but may vary depending on two factors. First, mapping techniques used in some states in the earliest generation of maps used a 100 hectare MMU. Some of these explored the issue of having multiple MMUs within a map, for example, a 100 ha MMU for uplands but a 40 ha MMU for riparian areas. Second, this has changed as more and more states having fine-grained landscapes began GAP projects. Cooperators in most newer states determined to develop and maintain land cover data using an MMU equal to the processed Landsat TM ground resolution, 30 m2. Many have argued convincingly that this approach results in expanded utility of the map products by allowing for spatial aggregation of map units to thresholds defined by user needs. GAP has supported research in software development for aggregation tools, and the program expects to produce and distribute a new and more efficient software tool for this purpose. Not only will this approach optimize the utility of GAP product applications, it will contribute significantly to basic research on MMU sensitivity analyses, leading to better interpretation and understanding of map content relative to generalization.

Each homogeneous area equal to or larger than the MMU is categorized according to the land cover classification system by its dominant vegetation type or, in the absence of vegetation, by the dominant land cover feature. The GAP data sets in some states serve as a spatial framework for finer-level habitat characteristics, which can be mapped as needed with a greater level of effort. Subdominant features of known occurrence can be listed in the database as an attribute of a given mapped polygon.

Characteristics of mapped land cover units

Generally, GAP data sets do not directly portray the age, specific height, or subcanopy composition of a mapped vegetation type. The structure of a mapped vegetation type is inferred by its physiognomic classification. Each community alliance has been described and classified according to features such as life form (tree, shrub, forb), spacing (closed canopy, open canopy, sparsely vegetated), phenology (deciduous, evergreen) morphology (sclerophyllious, broad-leaved, needle-leaved) and other structural characteristics according to the National Vegetation Classification System (FGDC 1996).

Map unit boundaries

The mapped boundaries delineating land cover types are approximations. The degree of the gradient between land cover types is not depicted in GAP data sets, and these gradients may range from gradual to sharp.

Vertebrate species distributions

Maps of species distributions are predictions illustrated as a generalized abstraction, as are all maps. These maps are based on known ranges of species, derived from known locations of actual specimen collections, in combination with the best available information about each species' affinity for habitat types as represented by the mapped land cover types and other known land features. Field observations should be obtained for site-specific applications.

Maps do not imply habitat quality

Maps showing the predicted distributions of vertebrate species do not show, nor do they imply, information about habitat quality or actual population density. Gap analysis data predict the presence or absence of a species, not whether it is abundant or uncommon at any given location or at any given period of time. Field-based inventories are required in order to gain information about habitat quality and population densities.

Organizational

Business Model

The single most significant organizational assumption that the program makes is that the work is best carried out state-by-state (although in a few cases GAP projects cover more than one state, state-level implementation is still the primary level of GAP organization) and that each state project is supported by the mutual cooperation of natural resources institutions (state, federal, private) from within each state.

Information as a catalyst

State projects are viewed as an event in progress, having to do with the development of powerful new information about biological resources. The information not only affects tools and capabilities, but the process of developing the information itself catalyzes integration among the cooperating institutions and resulting in important new institutional relationships and structures (i.e, the Missouri Resources Assessment Partnership; see http://www.msc.nbs.gov/morap/). As centralized environmental management and regulation is de-emphasized, scientifically sound biogeographic information shared among institutions for managing resources becomes more important for effective and meaningful decision making.

Multi state standardization, consistency, flexibility

Related to this is the basic assumption that benefits will be derived from a more unified approach to the management of biological resources not only among institutions within a state, but among such institutions across state boundaries. This approach necessarily incorporates hierarchy theory for ecological systems. The GAP strategy corresponding to this assumption is to foster the development and use of consistent definitions and classification of species assemblages and other basic sets of information as well as the use of consistent technical methods wherever possible. At the same time, GAP recognizes that singular approaches in many areas, for example methods of land cover pattern delineation, have not been proven. Therefore, flexibility in methods is also a key assumption as well as a limitation for GAP (and the state-of-the-science in general). GAP recognizes that standardization of procedures is desirable where it is possible, and it is required where the case can objectively be made (as in, for example, the development of the National Vegetation Classification System). Standardization of all methods for the sake of standardization alone, however, is self-limiting, especially where innovation and discovery are central to achieving the objectives, as they are with the Gap Analysis Program.

The program is presently in the process of assessing strengths and weaknesses of all land cover mapping methods thus far applied by each state project. It is expected that this will lead to better standardization of methods where practical.

Products

The National Gap Analysis Program results in six major categories of products:

Applications

General

Vickerman and Smith (1995) identified three basic types of application of GAP data to conservation planning: (1) situation-specific decision making, (2) integration with existing plans or planning processes, and (3) systematic land-use planning. A survey of applications of GAP data provided an early (1993-1995) profile of some uses of GAP data for situation-specific planning. This survey (Jennings 1995a) documented 43 randomly chosen cases of data applications from 12 state projects. It resulted in identification of 8 categories of situation-specific applications, shown below in Table 1. The most frequent (34%) situation-specific application of GAP data was for direct land management decision making where quick access to ecological information was required to meet an immediate need at low cost, such as a rapid assessment of the distribution of a habitat type in response to proposed changes in management or use.

Table 1.

An early profile of some categories and application frequencies of GAP data
CATEGORY PERCENT FREQUENCY OF USE
Direct Land Management 33.5
Delineation of Critical Habitats by Agency 13.0
Environmental Assessments 12.5
County Planning 10.5
Wildlife Management 10.5
Basic Research 9.0
Private Business Applications 5.0
Development of Option for Large-Area Conservation Designations 6.0




State Applications and Activities


State GAP data sets have become fundamental sources of biogeographic information in states where they have been completed. Many, if not most, completed GAP states are applying GAP data sets to some form of biodiversity management. Some examples of significant state-level planning activities that are using GAP data may be seen in Arkansas, California, Colorado, Florida, Indiana, Missouri, Oregon, and Tennessee.

Another state-level application of GAP data of note is the NatureMapping program (Dvornich et al. 1995), a K-12 curriculum, discussed in the section on information dissemination below (see page ).

Regional Applications and Activities


Numerous regional applications are emerging from GAP activities. For example, the Upper Midwest Gap Project is resulting in a single 3-state data set covering Michigan, Minnesota, and Wisconsin. The Mid American Mapping Consortium, consisting of the GAP projects in Iowa, Nebraska, Kansas, and Missouri has a goal of a set of enhanced GAP map products for that 4-state area. The Mid-Atlantic GAP regional activity is resulting in uniform products for the region covering Maryland, Delaware and New Jersey. GAP land cover maps for the Western states are being regionalized into a single map product allowing entities to place state-level management decisions in a large regional context. A GAP-supported regional conservation assessment for the Colorado Plateau ecoregion is well underway and promises to provide greater common ground for to the many different state, federal, tribal, and private interests in that region.

To illustrate the GAP commitment to regional products and activities as well as meeting the needs of the program's Department of Interior clients, is the following example of the National Wildlife Refuges in Context GAP project. With the Western states GAP data sets, the program is embarking on the development of a customized set of biogeography conservation data for each of the 93 National Wildlife Refuges (NWRs) in the West. For each of these refuges, a CD-ROM will be produced in a user-friendly "plug-and-play" PC format oriented toward NWR personnel. The sets of information products will include:

Each CD-ROM product will contain hypertext links to the original GIS data, allowing more sophisticated users to access large basic GAP data sets for a wide variety of applications. Using this product will allow NWR personnel to quickly access current information about the ecology, management, and conservation contexts of their refuge.

In other regional applications, Thomas and Davis (1966) identified seven very threatened and 16 threatened habitat types in their application of gap analysis data in the Mojave Desert of California. Even more importantly, they provided experience showing "that in order for gap analysis to serve as a mechanism for collaborative planning, it should be presented as a dynamic tool rather than an end product." They suggest that for GAP to reach its potential as a tool for conservation planning there should be: (a) commitment to distribution, (b) commitment to support users and to correct and update data sets, and (c) development of an interactive gap analysis decision support system.

Crowe (1966) provides a case study of the use of gap analysis in regional planning in southern California, in which the Southern California Association of Governments (SCAG), in partnership with the California Gap Analysis Project:

Stoms and Davis (in preparation), using recently completed state GAP project land cover data, completed a gap analysis of vegetation alliances across the Intermountain Semi-Desert Ecoregion, showing that four percent of this area (parts of California, Colorado, Idaho, Montana, Nevada, Oregon, Utah, Washington, Wyoming) is managed on a permanent basis to maintain biological diversity, and of 48 vegetation alliances, 22 are particularly vulnerable to elimination or degradation.

These are just some of the many regional activities spawning from the National Gap Analysis Program. Integral to these developments is the program's state-based business model and the cooperative nature of data development and analysis.

Partnerships


Interagency collaboration has been a hallmark of GAP work. One important result of the Gap Analysis Program is the very strong cooperation, interaction, and support received from all sectors of the natural resources conservation community. Collaboration at the national and state levels is now serving as a significant catalyst in eliminating institutional barriers among and between zoologists, land managers, botanists, policy analysts, ecologists, planners, remote sensing experts, agency managers, geographers, information experts, computer scientists, and others involved with conservation biology (Jennings 1995b). Program partnerships are numerous and details of each partnerships are provided in the 1994 and 1995 GAP Status Report (Jennings and others 1996).

Of primary importance is that public agencies (local, state, federal) and private organizations are coming together around GAP information, in no small part because they are all mutually involved in its development. This coalescing is being facilitated by new science and better technology as well as the setting of standard methods, definitions, and database formats, including careful peer-review of products and procedures. The GAP partnership model contains the key elements identified by the National Research Council's Mapping Science Committee as necessary to promote the National Spatial Data Infrastructure (NSDI): shared responsibilities, shared cost, shared benefits, and shared control (National Research Council 1994). The utility of the GAP effort is being further extended via the application of consistent spatial data through its partnership infrastructure, meeting the call for nationally coordinated spatial data (National Research Council 1993).

At the beginning of the 1997 fiscal year, there were about 445 organizations cooperating on a state-project basis.

At the national level, one of the most notable partnership activities is the Multi Resolution Land Characteristics Consortium (MRLC). The primary purpose of the MRLC is the joint acquisition of Landsat Thematic Mapper (TM) satellite imagery covering the U.S. The 1993 MRLC TM acquisition was the first comprehensive TM data set ever assembled for the 48 contiguous states. A private sector benefit from this activity is the marketing of the data set by the EOSAT Corp. as the "Best of the U.S." collection. The consortium is currently pursuing a second nationwide acquisition of TM. Other goals of the MRLC include development of a flexible land characteristics database through the coordinated ongoing activities of member programs. The current partners in the MRLC are:

The MRLC is constituted by a Memorandum of Understanding (MOU), and membership in the MRLC is open to other federal programs. Through their activities, the MRLC partners are continuing to develop a flexible database of information characterizing land cover at multiple resolutions. There is ongoing development of consortium protocols for accuracy assessment, land cover legends, spectral classification of imagery, and product definition. In addition, the consortium members continue to better define roles for a more efficient division of labor as well as improve upon a strategy for database development.

National GAP programs are presently emerging in both Mexico and Canada. The U.S. GAP has been assisting counterparts in both of these countries and expects to continue to foster the development of standardized biogeographic data as well as standardized TM imagery for North America.

Funding


Funding is provided on an annual basis from the national program to state projects using: (a) the BRD Research Work Order process for Coop Units; (b) Cooperative Agreements for non federal institutions such as universities or state agencies; or (c) Inter-Agency Agreements for other federal agencies. Regardless of the funding mechanism, each state project operates under an approved work plan.

The GAP program has received substantial funding support in the past from the Environmental Protection Agency and the Department of Defense (see Jennings and others 1996). In fiscal year 1997, GAP anticipates receiving extra-agency funding from EPA to support the MRLC Liaison position and to assist with the mapping of land cover in South Carolina.

Reporting Schedules


The national program produces an annual status report in the second quarter of each fiscal year. Individual state projects produce quarterly or biannual status reports and a final report at the end of the project term.

Updating


The intended frequency of updating the GAP state products is every 5-10 years, depending on: (a) perceived needs by state cooperators, (b) funding availability, and (c) national needs.

Information Management


All GAP data are produced with metadata according to federal standards (Cogan and Edwards 1994). Currently, GAP data is being served by a network of state GAP data nodes which are linked through the GAP home page (http://www.gap.uidaho.edu/gap). At the national level, preparations are under way to archive GAP data at the USGS EROS Data Center. Data will be served over the internet by the Multi-Resolution Land Characteristics Consortium (MRLC; http://www.epa.gov/grd/mrlc) system, for which a prototype has been developed and is presently being tested (see http://edcwww2.cr.usgs.gov/mrlc/mrlc_server.html, then select Federal Region 3, then select Delaware to view and interact in a graphical format with land cover information). All state project data are examined and tested for functionality and content by the University of Idaho's Landscape Dynamics Laboratory prior to archiving. State projects are responsible for assessing the accuracy of their data.

Information Dissemination


Each state project final report is being produced with hypertext links to web sites having the actual data sets. These reports are being published on web sites, CDs, and on paper. Regional products will be disseminated in a similar manner. The dissemination of results will take somewhat different forms and have different emphases in different states, depending upon the state-level support, needs, and cooperator organization. The program is supporting biological extension agents in two states on an experimental basis. These extension projects are focused on both meeting cooperator information needs as well as exploring the application and delivery of GAP information to county planning offices.

One form of information dissemination of special interest is the NatureMapping Program (Dvornich et al. 1995; see http://salmo.cqs.washington.edu:80/~wagap/nm/). NatureMapping uses GAP data as the basis for a K-12 curriculum and a community involvement program. The program's vision is to create a national network that links natural resource agencies, academia, and land planners with local communities through their schools and civic organizations. The goal is to keep common animals common and to maintain our quality of life by training individuals to become aware of their natural resources, and to provide the tools to inventory and monitor their resources.

Research

The national program has an ongoing need for research and development in areas such as methods for assessing the accuracy of spatial data sets, analytical methods for identifying conservation gaps at different spatial and thematic scales, and tools for aggregation of map units. Of special interest is the development of methods for extending the gap concept to components of biodiversity other than terrestrial land cover and terrestrial vertebrate species. For example, two pilot research project, one in New York state and the other in Missouri, are developing applications of the gap analysis concept to aquatic environments. Another research effort being planned is to explore the development of gap methodology for individual plant species.

In meeting research needs, the program will develop annual research agendas. All research proposals and work plans are reviewed by a panel of peers. The following goals and objectives are a point of departure for future research on vertebrate species conservation.

Program Evaluation

The Gap program has so far undergone two review processes, one by a panel of peers (Zube et al. 1995) and another by the forest products industry (Flather et al. 1994). It is anticipated that a second peer review of the program should be undertaken within the next two years.

LITERATURE CITED

Allen, T.F.H., and T.B. Starr. 1982. Hierarchy perspectives for ecological complexity. The University of Chicago Press, Chicago. 310p.
Austin, M.P. 1991. Vegetation: Data collection and analysis. In: Nature conservation: Cost effective biological surveys and data analysis, C.R. Margules and M.P. Austin, eds., Australia CSIRO, East Melbourne.
Austin, M.P., and C.R. Margules. 1986. Assessing representativeness. In: Wildlife conservation evaluation, M.B. Usher, ed. Chapman and Hall, Ltd., London. 394p.
Burley, F.W. 1988. Monitoring biological diversity for setting priorities in conservation. In: Biodiversity, E.O. Wilson, ed., National Academy Press, Washington, D.C. 521p.
Cogan, C., and T.C. Edwards, Jr. 1994. Metadata standards for Gap Analysis (version 1). In: A handbook for Gap Analysis, National Gap Analysis Program, Moscow, Idaho, USA.
Crowe, R.E. 1966. Use of gap analysis data in the Mojave Desert of California. In: Gap analysis: A landscape approach to biodiversity planning, J.M. Scott et al., eds. American Society for Photogrammetry and Remote Sensing, Bethesda, Maryland. 320p.
Csuti, B. and A.R. Kiester. 1996. Hierarchical gap analysis for identifying priority areas for biodiversity. In: Gap analysis: A landscape approach to biodiversity planning, J.M. Scott et al., eds. American Society for Photogrammetry and Remote Sensing, Bethesda, Maryland. 320p.
Davis, F.W., D.M. Stoms, J.E. Estes, and J. Scepan. 1990. An information systems approach to the preservation of biological diversity. Int. J. Geographical Information Systems, 4(1):55-78.
Dvornich, K.M., M. Tudor, and C.E. Grue. 1995. NatureMapping: Assisting management of natural resources through public education and public participation. Wildlife Society Bulletin, 23(4):604-619.
Fenner, F., ed. 1975. A national system of ecological reserves in Australia. Australian Academy of Sciences Report No. 19. Canberra, Australia. FGDC (Federal Geographic Data Committee). 1996. FGDC vegetation classification and information standards. U.S. Geological Survey, Reston, Virginia.
Flather, C.H., K.R. Wilson, D.J. Dean, and W.C. McComb. 1994. The National Gap Analysis Program: A review and inspection of ecological assumptions. A report to the National Council of the Paper Industry for Air and Stream Improvement, Inc. USDA Forest Service, Rocky Mountain Forest and Range Experiment Station, Fort Collins, Colorado.
Franklin, J.F. 1993. Preserving biodiversity: Species, ecosystems, or landscapes? Ecological Applications, 3(2):202-205.
Grossman, D., K.L Goodin, X. Li, C. Wisnewski, D. Faber-Langendoen, M. Anderson, L. Sneddon, D. Allard, M. Gallyoun, and A. Weakley. 1994. Standardized national vegetation classification system. Report by The Nature Conservancy and Environmental Systems Research Institute for the NBS/NPS Vegetation Mapping Program. USGS Biological Resources Division, Denver, Colorado.
Halbert, S., M.D. Jennings, C. Cogan, S. Quisenberry, and J.B. Johnson. 1995. Potential use of suction trap collections of aphids as indicators of plant biodiversity. In: Insects In A Changing Environment, N. Stark and R. Harrington, eds., Academic Press, London. 535p.
Hutto, R.L., S. Reel, and P.B. Landres. 1987. A critical evaluation of the species approach to biological conservation. Endangered Species Update, 4(12):1-4.
Jennings, M.D. 1995a. A discussion of the adoption and diffusion of gap analysis as a technical innovation. Gap Analysis Bulletin No. 4. National Gap Analysis Program, Moscow, Idaho, USA.
Jennings, M.D. 1995b. A confluence of biology, ecology and geography for the management of biological resources. The Wildlife Society Bulletin, 23(4):658-662.
Jennings, M.D. 1996. Nomenclature and mapping units for gap analysis land cover data. In: Gap Analysis: A landscape approach to biodiversity planning. J.M. Scott et al., eds. American Society of Photogrammetry and Remote Sensing, Bethesda, Maryland. 320p.
Jennings, M.D., E. Brackney, P. Crist, and R. Sorbel. 1996. 1994 and 1995 Gap Analysis status report. National Gap Analysis Program, Moscow, Idaho, USA.
Levin, S.A. 1981. The problem of pattern and scale in ecology. Ecology, 73:1942-1968.
National Gap Analysis Program. 1994. A handbook for gap analysis. Moscow, Idaho, USA.
National Research Council. 1993. Toward a coordinated spatial data infrastructure for the nation. National Academy Press, Washington, D.C. 171p.
National Research Council. 1994. Promoting the National Spatial Data Infrastructure through partnerships. National Academy Press, Washington, D.C. 113p.
Noss, R.F. 1990. Indicators for monitoring biodiversity: A hierarchical approach. Conservation Biology, 4:355-364.
Noss, R.F, and A.Y. Cooperrider. 1994. Saving nature's legacy. Island Press, Washington, D.C. 420p.
Noss, R.F., E.T. LaRoe III, and J. Michael Scott. 1995. Endangered ecosystems of the United States: A preliminary assessment of loss and degradation. Biological Report 28. U.S. Department of the Interior, National Biological Service, Washington, D.C.
O'Neill, R.V., D.L. DeAngelis, J.B. Waide, and T.F.H. Allen. 1986. A hierarchical concept of ecosystems. Monographs in Population Biology 3. 53p.
O'Neill, R.V. 1996. Recent developments in ecological theory: Hierarchy and scale. In: Gap Analysis: A landscape approach to biodiversity planning. J.M. Scott et al., eds. American Society of Photogrammetry and Remote Sensing, Bethesda, Maryland. 320p.
Orians, G.H. 1993. Endangered at what level? Ecological Applications 3(2):206-208.
Prendergast, J.R., R.M. Quinn, J.H. Lawton, B.C. Eversham, and D.W. Gibbons. 1993. Rare species, the coincidence of diversity hotspots and conservation strategies. Nature 365:335-337.
Scott, J.M., B. Csuti, J.D. Jacobi, and J.E. Estes. 1987. Species richness. BioScience 37(11): 782-787.
Scott, J.M., F. Davis, B. Csuti, R. Noss, B. Butterfield, C. Groves, H. Anderson, S. Caicco, F. D'Erchia, T. Edwards, J. Ulliman, and G. Wright. 1993. Gap analysis: A geographic approach to protection of biological diversity. Journal of Wildlife Management 57(1) supplement, Wildlife Monographs No. 123.
Stoms, D.M., and F.W. Davis. In preparation. Gap analysis of vegetation alliances in the Intermountain Semi-Desert ecoregion (U.S.A.). Institute for Computational Earth System Science, University of California, Santa Barbara.
Tear, T.H., J.M. Scott, P. Hayward, and B. Griffith. 1993. Status and prospects for success of the Endangered Species Act: A look at recovery plans. Science, 262:976-977.
Terbourg, J. 1988. The big things that run the world. Conservation Biology, 2:402-403.
Thomas, K.A., and F.W. Davis. 1996. Applications of gap analysis data in the Mojave Desert of California. In: Gap Analysis: A landscape approach to biodiversity planning. J.M. Scott et al., eds. American Society of Photogrammetry and Remote Sensing, Bethesda, Maryland. 320p.
Vickerman S., and K.A. Smith. 1995. Draft recommendations for the implementation of gap analysis: A report to the National Biological Service. Defenders of Wildlife, Portland, Oregon and McCollum Associates, Sacramento, California.
Whittaker, R.H. 1965. Dominance and diversity in land plant communities. Science, 147: 250-259.
Zube, E.H., editor. 1994. Peer review panel report of the national Gap Analysis Program for the National Biological Survey. NBS/Idaho Cooperative Fish and Wildlife Research Unit, University of Idaho, Moscow.

Authors

Patrick J. Crist, Program Coordinator and Michael Jennings, National Director
GAP
530 S. Asbury Street, Suite 1
Moscow, ID 83843
Telephone: (208)-885 3555
Fax: (208)885-3618
e-mail: gap@uidaho.edu