USGS:Biology Biological Informatics Program Vegetation Characterization Program Standardized National Vegetation Classification System 4.0 The Nature Conservancy's Vegetation Classification System

4.0 The Nature Conservancy's Vegetation Classification System

4.1.1 Background

Over the past twenty years, The Nature Conservancy has developed a science-based approach to conserving biological diversity. The Conservancy's approach to conservation science relies on the consistent and systematic accumulation, management, and analysis of information on the "elements of biological diversity" specifically the status and location of plants, animals, and ecological communities. This information is collected and managed by the Association for Biodiversity Information (ABI), an international network of cooperating Natural Heritage programs and conservation data centers.

For more than a decade, the Conservancy and the Natural Heritage programs have employed a "coarse filter/fine filter" approach to preserving biological diversity. This approach involves the identification and protection of ecological communities (coarse filter) as well as rare species (fine filter). The protection of the best examples of all ecological communities will assure the conservation of most species, biotic interactions, and ecological processes. Those species that "fall through" the community filter are generally the rare species. Identification and protection of viable occurrences of rare species serves as the fine filter for preserving biological diversity (Jenkins 1976, Hunter 1991). Using communities as a coarse filter has ensured that the Conservancy is working to protect a more complete spectrum of biological diversity, not just those species whose priority conservation status has been documented.

Ecological communities were first used to help direct conservation priorities on a state-by-state basis. Community information was systematically collected by ecologists from the state Natural Heritage programs to develop and refine state-level community classifications and conservation ranks. These state classifications were developed for most states, but often used different classification approaches (White 1978, Nelson 1985, Reschke 1990). This strategy to identify conservation priorities was implemented at the state level to assure protection of ecological communities. However, national conservation efforts require compilation and analysis of community data from a rangewide perspective.

A major obstacle to using communities as conservation units at the national level was the lack of a consistent national vegetation classification system. To overcome this problem, the Conservancy, in conjunction with the ABI, has developed a standardized hierarchical system to facilitate the identification and classification of vegetated terrestrial communities across the United States.

4.1.2 Guiding Principles

4.1.2.1 Appropriate for Conservation Planning and Management

The Conservancy's national vegetation classification system was primarily developed for the purposes of conservation planning and biodiversity protection. The intent of the classification system is to provide a complete, standardized listing of all communities that represent the variation in biological diversity and to identify communities that require protection. The classification will be consistent throughout the United States at appropriate scales for conservation planning and the management and long-term monitoring of ecological communities and ecosystems. It is also intended to have applications as a vegetation data layer for mapping and landscape and ecosystem analyses.

4.1.2.2 Efficient Use of Existing Information

Because The Nature Conservancy's mission is to protect biological diversity, the classification system emphasizes biota as the major attribute. Vegetation is the primary attribute used to classify terrestrial communities. When designing the classification system, the existing standards for vegetation classification and characterization were recognized and used wherever possible (see Section 3.0 above). Various classification systems were researched that had national or international applications, used widely accepted standards, and were practical for conservation applications. Several widely accepted classification approaches were adapted and modified as necessary to meet conservation objectives. When identifying individual vegetation types within the classification system, vegetation types from existing classification schemes were analyzed and refined to bring them to a common and consistent scale.

To efficiently use existing community information across the United States, the relationships between communities in the Conservancy's classification and those from other classifications must be documented. As no single system will be completely compatible with all other classification systems, the intent was to build this system and then create cross-references to other classification schemes as needed (see Section 5.6.1 for an example). Numerous data fields are included in the ecological database records to identify these relationships (see Section 4.3.1.2 below). These features were designed to help The Nature Conservancy utilize the information in other systems as well as to help the users of other systems to understand how their classifications fit into the Conservancy's system.

4.1.2.3 Flexible

In addition to meeting the objectives for protecting biological diversity, another goal of the classification system is to meet the objectives of other federal and state agencies, academic institutions, and other conservation organizations involved in the science and practice of conservation and ecosystem management. Recognizing that the objectives for using a national vegetation classification vary among these groups, the classification system was designed to be as flexible as possible while maintaining certain standards. For example, the system is open ended new classes can be added as needed, provided they follow the guidelines developed for the classification system. In addition, information not explicitly used to classify vegetation can be incorporated as attributes in associated data records, maps, and reports (see Sections 5.4.3 and 6.2.2 below). The classification is also meant to be updated and refined as further inventory and classification efforts provide additional data and knowledge about the vegetation.

4.1.2.4 Emphasis on Natural and Seminatural Vegetation

For purposes of prioritizing classification research, the existing vegetation types have been categorized to reflect their level of disturbance and management. "Natural," "seminatural," and "modified" vegetation types are recognized to reflect differences in the natural and anthropogenic disturbance regimes. In addition, a "cultural" land cover class is recognized which includes anthropogenic vegetation types (e.g., lawns, crops) and structures (e.g., buildings, parking lots). All of these classes can be described within the Concervancy's classification system. These distinctions, while somewhat arbitrary, have been used to categorize the landscape and focus conservation efforts on the more natural types. However, in mapping vegetation, all vegetation types and land cover classes must be portrayed under a single classification approach (see Section 5.2.4 below).

4.2 Development of the Nature Conservancy Classification System

4.2.1 Identification of Classification Units

The classification units defined thus far have been primarily developed from existing vegetation data collected by state Natural Heritage programs from federal agencies, researchers, and from vegetation data or summary descriptions reported in the literature. Thousands of references and unpublished data sets have been reviewed and analyzed to create the classification units. However, there is considerable variation among states in the amount of community information available and the degree of development of the state classifications. The degree of development of the national classification on a state-by-state basis reflects the a mount of information available (see Section 5.7.1 below).

The classification process is implemented through a variety of qualitative and quantitative means depending on the amount and type of information available. The classification is continually refined and updated as additional field data are collected and analyzed. The development of the national vegetation classification has proceeded from the development of state classifications to the production of regional classifications, and finally to the generation of a consistent classification at the national level. Although the state classifications vary widely in the level of detail and classification approach, each region has cross-referenced its respective state communities within the national hierarchical framework. Problems of scale and nomenclature continue to be rectified at the regional level in close association with the state ecologists.

The Conservancy is comprised of four regions that support Natural Heritage programs in the United States: west, midwest, east, and southeast (Table 1). Each of the Conservancy's U.S. regions has now completed a regional vegetation classification which employs the standards developed for the national vegetation classification (Allard 1990, Bourgeron et al. 1994, Faber-Langendoen 1993, Sneddon et al. 1992).

4.2.2 International Efforts

The classification system is applicable worldwide. Conservancy ecologists are currently working with the ecologists in conservation data centers in Canada to employ the classification system in British Columbia, Manitoba, Saskatchewan, and Ontario. Previous versions of the system have also been applied in Jamaica and Belize. Although the specific classification units have not been identified for other countries, the classification system is developed with the expectation that it will become an international standard.

4.2.3 Support from Federal and Academic Partners

Development of this classification system has been supported by a number of federal and academic partners who have interest in using the system. Support has been provided to the Conservancy's national and regional offices as well as directly to state natural heritage programs. A summary of the support granted to the Conservancy's national and regional offices is provided below. In addition, federal agencies, such as the Bureau of Land Management, United States Forest Service, the Environmental Protection Agency, and the National Park Service, have provided support directly to state natural heritage programs for community classification and inventory. This funding has been critical to the development of the national vegetation classification.

4.2.3.1 National Biological Survey

The National Biological Survey's Gap Analysis program has supported the development of the "alliance level" units (see below) in the eastern and western regions of the Conservancy and is planning to support the similar work in the southeastern and midwestern regions. The Gap Analysis program uses the alliance level of this classification system as the standard for their vegetation maps at scales of 1:500,000 to 1:100,000 (depending on the region) across the United States.

4.2.3.2 United States Fish and Wildlife Service

The United States Fish and Wildlife Service has supported the development of a list and descriptions of all of the known rare communities of the conterminous United States through their Land Acquisition Priority System (LAPS) (Grossman et al. 1994). Rare communities are among the measures of biological diversity that make up the LAPS "Biodiversity Target," a system that helps determine priorities for the acquisition of new refuges.

Individual refuges have also supported development of the classification. For example, the Stillwater Wildlife Refuge in Nevada supported ecologists in the Conservancy's western region to develop a classification for the refuge.

4.2.3.3 United States Forest Service

Region 8 of the U.S. Forest Service has worked with the Conservancy for several years to develop a classification and description of and keys to existing vegetation for the national forests in the region. In addition, the Conservancy is working with Region 1 of the Forest Service to develop data management and analytical tools to support vegetation classification and ecosystem characterization.

Several USFS regions have supported the Conservancy to crosswalk the USFS ecological land classification with the Conservancy's vegetation classification and further expand the classification. For example, the Conservancy and Region 9 of the USFS are working together to complete a analysis of their Research Natural Areas using a combination of the USFS ecological land classification and the Conservancy's classification. Similar work is in progress with the USFS Northeast Forest Experiment Station in New Hampshire and in the Shawnee National Forest in Illinois.

Several individual USFS ecologists have collaborated with Conservancy ecologists to develop the Conservancy's classification in the Conservancy's western region. Ecologists from USFS Regions 3 and 4 collaborated with Conservancy ecologists to relate their habitat type classification to the Conservancy's classification. They also provided data and reviewed drafts of the classification.

4.2.3.4 National Park Service

In the Conservancy's southeastern region, the National Park Service is currently supporting the generation of vegetation maps for five small national parks using the regional portion of the Conservancy's classification. This project was initiated prior to the more comprehensive program to map the vegetation of all national parks and will be coordinated with the larger effort.

The National Park Service also funded the Conservancy to do a literature review to support the development of the classification of the vegetation in the Colorado Plateau, Utah.

4.2.3.5 Environmental Protection Agency

The Environmental Protection Agency's Environmental Monitoring and Assessment Program (EMAP) is using this classification system as the vegetation standard for their land-use maps. Region 5 of the EPA is supporting the Conservancy to apply the midwestern portion of the Conservancy's classification system to the vegetation of Wisconsin.

EPA's Region 7 is currently funding the Conservancy to synthesize vegetation data from states in the Great Plains using the midwestern, western, and southeastern portions of the Conservancy's national vegetation classification.

EPA Regions 1 and 3 are funding a coordinated effort by the Conservancy and state natural heritage programs to inventory and classify selected rare wetland communities in the eastern United States.

4.2.3.6 Inter-agency Groups

The Upper Great Lakes Biodiversity Committee is a group of federal and state agencies, academics, industry, and nonprofit environmental organizations in Michigan, Wisconsin, and Minnesota working to complete a regional biodiversity assessment. The assessment will include use of the Conservancy's classification system.

4.2.3.7 Academic Partners

The Nature Conservancy works closely with the Vegetation Section of the Ecological Society of America. ESA has now initiated a special panel on vegetation classification, where we will work in partnership to develop standards and a review process for the future development and refinement of the national vegetation classification system. Conservancy ecologists also collaborate with individual vegetation scientists to generate portions of the classification and solicit peer review.

4.3 Tools and Methods that Support the Documentation and Development of the Classification

4.3.1 Field Sampling

The Nature Conservancy utilizes standard methodologies for the collection of field data (Bourgeron et al. 1991, Sneddon 1992, Faber-Langendoen 1993). These methods apply to vegetation mapping (see Field Methodology report in this series) and the development of the vegetation classification and descriptions.

The field sampling methodology is usually based on the collection of plot/relevé samples of appropriate size and shape for the particular vegetation type being classified (e.g., square 10x10m plots are used to collect information on shrub-land communities, rectangular plots are generally used to collect information in riparian habitats). Within the plot, standard information is collected on the identity and abundance of all plant species, the structure/architecture of the vegetation, and a set of variables such as moisture regime, soil type, depth, organic content and pH, bedrock type, topographic setting, aspect, slope, geographic location, and others that characterize the immediate environment.

Given the extensive area covered by the classification, two methods, community-based and site-based, are commonly used to allocate samples. Community-based sampling is used to refine the classification for a targeted group of related communities. Site-based sampling is used to identify and classify the communities on a given site by identifying units which are representative of the biological associations across the major environmental gradients.

In community-based sampling, data collection is focused on a particular alliance or broader group of related communities of interest and a detailed set of criteria for site inclusion are determined a priori. For example, if sampling "fens" across six New England states, the sampling might be restricted to communities which (1) are dominated by graminoids or shrubs, (2) occur in areas of similar ecological setting (e.g., shallow to deep peat areas influenced by contact with basic groundwater), and (3) contain at least some members of a larger set of suspected characteristic species. In practice, restrictions are redefined as more is learned about the vegetation patterns.

In contrast to the community-based stratification, site-based stratification partitions the area of interest into units that reflect important environmental and topographic gradients (e.g., slope, aspect, elevation, moisture regime, soil type) (Gillison and Brewer 1985, Austin and Heyligers 1989). Transects that contain the strongest environmental gradients in a region are selected in order to optimize the amount of information gained in proportion to the time and effort spent during the vegetation survey (Austin and Heyligers 1989). Once the major environmental gradients are identified, they are partitioned into environmental cells that reflect unique combinations of the variables. Aerial photo analysis is used to further partition the units into areas of apparently homogenous vegetation. A subset of the cells that represent the entire range of variation of the site (wet-dry, low elevation-high elevation, disturbed-undisturbed, etc.) are then selected for sampling.

Once sample sites are located (by either community-based or site-based methods), plots are placed in areas of homogenous vegetation which are determined to be representative of the vegetation type. Regions of transitional nature are avoided. Random, restricted random, and stratified random schemes are all used to locate the plots within a site, though stratified random schemes are generally preferred. Because the objective for sampling is the characterization of vegetation types, the analysis methodologies are quantitative rather than probabilistic, and the defined units are scale-dependent. The representative placement scheme is substantially more efficient than other methods and appropriate for these objectives.

4.3.2 Community Descriptions

4.3.2.1 Characterization Variables and Vegetation Keys

The Nature Conservancy describes communities in the classification using a standard set of more than 100 characterization variables. Fields of information that can be completed for each community element include variables which portray the physiognomic and biotic traits of the vegetation, as well as variables that relate to key environmental factors, dynamic processes, landscape relations, community variability, threats, and management and protection needs. Cross-references to other major classifications, including the Federal Geographic Data Committee's classification standard, are included in the fields used to characterize the community elements.

On local and regional levels, complete community descriptions can be converted into vegetation keys so that users of the classification can identify occurrences of the community on the ground. National-level keys will not be possible until the classification is more complete.

While all of the fields can be used to describe a given community, such complete characterization is beyond the scope of many projects. As a result, the Conservancy has identified a minimum subset of the fields that provide a satisfactory description of a vegetation type (Table 2). Examples of basic community descriptions are included in Appendix 9.2.

4.3.2.2 Biological and Conservation Data System Community Records

Community characterization variables have been captured in a database system, the Biological and Conservation Data System, in which heritage information is managed. These files contain both data fields (single- and multi-valued) and summary fields (text) which carry information on individual occurrences (stands) of communities (Element Occurrence Record) as well as the general descriptions of the vegetation type across its range (Community Characterization Abstract). Information on communities carried in these files includes a basic description of the vegetation, its physiognomic structure, and biotic composition. Also included is information on the key environmental factors, dynamic processes, landscape relations, community variability, threats, and management and protection needs associated with each community. Fields that identify the relationship of the community to communities from other major classifications are included in the data structures. This supporting information allows the classification of each type to be documented and occurrences of types tracked by state heritage programs. Brief descriptions of the fields in the Element Occurrence Record and the Community Characterization Abstract are included in Appendixes 9.4 and 9.5.

4.3.3 Conservation Ranking

After a community's type is recognized, it is ranked according to its relative rarity or endangerment. Individual occurrences of each community type are also ranked according to their relative condition. The combination of classification and ranking systems provides a framework for identification of the most significant community types and community occurrences, a critical step in identifying priority sites for biodiversity conservation.

Communities are ranked on a global, national, and subnational (state or provincial) conservation scale of 1 to 5 in a manner similar to the system developed by The Nature Conservancy for ranking species (Master 1991). A rank of G1 (Global 1) indicates that a community is highly endangered due to rarity, endemism, and/or threats, and a rank of G5 (Global 5) indicates no risk of extinction. Similarly, a rank of N1 (National 1) or S1 (Subnational 1) indicates that the community is endangered at the national or subnational level, respectively. The two primary criteria in determining a community's rank include total number of occurrences and total area (acreage) of the community rangewide. Measures of geographic range, trends in status (expanding or shrinking range), trends in condition (declining condition of remaining acreage), threats, and fragility are secondary ranking factors which are considered when assigning a rank. The criteria used to assign a rank to a particular community are documented using a standardized format. See Appendix 9.6 for a description of Element Ranking Criteria.

In a fashion similar to ranking of community types, the occurrences of a particular community are ranked using a scale from "A" to "D." These community occurrence ranks are based on the occurrence's relative condition, size, quality, viability, and defensibility. "A" ranked community occurrences are generally large, pristine examples of the community type with relatively little disturbance and no threats, whereas "D" ranked occurrences are generally small, highly degraded, threatened examples of the type which may not be "protectable."