Thomas Foti
Arkansas Natural Heritage Commission
Suite 1500, Tower Building
Little Rock, AR 72201
Xiaojun Li
Department of Biological Sciences
University of Arkansas
Fayetteville, AR 72701
Martin Blaney
Arkansas Game and Fish Commission
2 Natural Resources Drive
Little Rock, AR 72205
Kimberly G. Smith
Department of Biological Sciences
University of Arkansas
Fayetteville, AR 72701
We present a hierarchical classification system for existing natural vegetation of Arkansas based on the United Nations Educational, Scientific and Cultural Organization (UNESCO) system. It incorporates aspects of systems in use by the Nature Conservancy, Arkansas Natural Heritage Commission, Arkansas Game and Fish Commission, Society of American Foresters, and United States Forest Service, as well as data on potential vegetation from maps by E.E. Dale and A.W. Kuchler. A total of 18 Physiognomic cover types are recognized for natural terrestrial cover, 6 for palustrine cover, and 4 each for lacustrine cover and riverine cover. Over 200 community types are recognized, grouped into 57 cover types and 90 intermediate groupings. This system is appropriate for use with remotely sensed data and the level of detail dealt with can be rationally adjusted by working at a higher or lower level of the hierarchy. We suggest that this system form the basis for future vegetation analyses and research within Arkansas.
The Arkansas Gap Analysis Project (see Scott et al., 1993 for a complete discussion of Gap Analysis), being conducted by the University of Arkansas in cooperation with several other academic institutions, state and federal agencies, and private organizations, will produce a map of existing vegetation and potential vertebrate distribution within Arkansas. The map will be created from satellite imagery, GIS maps of geology, topography, soil and other physical features, and databases of species occurrence and habitat characteristics.
The vegetation units mapped will be those that can be distinguished on satellite imagery and GIS data layers. It is expected, based on results in other states, that approximately 50 vegetation units will be mapped. It is desirable for maximum utility of the map that these map units be related to an overall classification of Arkansas vegetation. Therefore one of the initial priorities of the Arkansas GAP was to produce such a vegetation classification for Arkansas.
Several plant community classifications exist for Arkansas (AGFC, 1948; Moore, 1959; Foti, 1974; Pell, 1981; Dale, 1986; and unpublished classifications used by several agencies cooperating in this project). A valuable published classification exists for Missouri (Nelson, 1985). Several national classifications have applicability within Arkansas (Eyre, 1980; Kuchler, 1964). However, since plant communities are not discrete entities, each classification is a reflection of the philosophy of the creator and the philosophies vary.
None of these classifications met the needs of all agencies cooperating in the Arkansas GAP. Most were too general for GAP purposes; several (Foti, 1974; Pell, 1981; Nelson, 1985) combine physical and biological diagnostic features, e.g. Dry-Mesic Oak- Hickory Forest. Including both physical and biological features simplifies a classification in that, using the above example, it is not necessary to distinguish the various combinations of Quercus alba, Q. falcata, Q. velutina, Carya tomentosa, C. texana and other species that may dominate such sites. However, such simplification of biological communities inevitably results in loss of information.
The technology being used in the Arkansas GAP, GIS and remote sensing, allows physical features of sites to be characterized using digital elevation models, geology, and soil layers while vegetation or land cover can be independently classified based on satellite or other imagery.
In addition to ecosystemic classifications which classify physical features to indirectly classify vegetation (such as those classifications discussed above) two general approaches focus strictly on vegetation: physiognomic and floristic (Whittaker, 1978).
Physiognomic classification depends on morphological characteristics (structure) of vegetation, and it is primarily determined by growth-form and life-form of the dominant or codominant plants. Physiognomic classification is extensively used to characterize vegetation over large geographical areas because it can be visually recognized and distinguished and does not require much floristic detail about the vegetation.
The floristic approach focuses on analysis and synthesis of the floristic composition of communities. The diagnostic species, which occupy ecological niches of different dimensions, are used to characterize the basic unit (association) as well as higher units of the classification hierarchy.
For the reasons stated above, it was decided that the Arkansas vegetation classification would be based strictly on biological features of the community.
Further, it was desired that the product of this effort be a hierarchical framework to allow the level of detail to vary for different users and, since GAP is a national effort, that the vegetation units recognized here should be as compatible as possible with those of other states.
Three assumptions are made for the vegetation classification system of Arkansas (based on recommendations from National GAP): 1) This system is used to describe actual or existing vegetation rather than potential or climax vegetation; 2) It does not include transition zones of vegetation (the level of detail at the lowest level is high enough that some units considered by others to be transition zones are recognized as units); and 3) It is open-ended in that categories may be added to any of the hierarchical levels as long as the additions are truly an equivalent category within the given classification level. Furthermore, it is reasonably easy to combine units recognized here and produce a classification that can be readily correlated to this one.
Based on the approach adopted for the National Gap Analysis Project (Scott et al., 1993), the vegetation classification system of Arkansas presented here follows the UNESCO (United Nations Educational, Scientific, and Cultural Organization) (1973) format with modification of the lower two levels (Jennings, 1993). This classification scheme offers a widely accepted and useful hierarchical grouping that is based primarily on 1) the physiognomic or structural expression of plant cover relative to environment at higher levels of the scheme, and 2) the floristic composition at lower levels.
There are six levels in this vegetation classification system: class, subclass, group, formation, cover type, and community type. The first four levels are physiognomic and the latter two are floristic (to the extent of defining dominant, diagnostic, or indicator species).
The following definitions of the levels of the hierarchy follow the definitions of National Gap Analysis project (Jennings, 1993) with slight modifications by us. The criteria used to define classification categories (e.g., woodland = tree canopy cover of 26-60%) are general. They are meant to be a means for grouping and discussing discrete cover types from coarse categories to finer categories. The growth-form and the life-form used in this classification system follow the growth-form categorization of Rubel (1930) and the life-form classification of Raunkier (1934), respectively.
Class: There are six primary classes. The first five of these represent vegetation cover; the sixth represents substantial bare ground. The distinctions between classes representing vegetation are based on the spacing and height of dominant vegetation growth form. 1. Forest: Forests are dominated by trees with a total canopy cover of 61% or more and tree crowns usually interlocking. 2. Woodland: Woodlands are dominated by trees with a total canopy cover of 26-60%, most tree crowns not touching each other. A herbaceous or shrub understory, or both, are usually present. They are open stands of trees, sometimes called "open forest". 3. Dwarf shrubland: These are comprised of shrubs rarely exceeding .5 meters in height at maturity. The type probably does not occur in Arkansas. 4. Shrubland: These are areas dominated by shrubs that generally range from 0.5 meter to 5 meters in height when mature, with a total canopy cover of 26% or more. A tree canopy cover of 26% or less may be present. 5. Herbaceous: These are areas dominated by grass, grass-like, or forb vegetation with a tree or shrub component not exceeding 25% cover. 6. Barren/sparsely vegetated: These are areas where vegetation cover is less than 5%. This type includes mud flats, sandy areas, and bare rock.
Note that such widely-used (but often inconsistently defined
terms as "savanna", "prairie", "glade" and "barrens" are not used
here. This is meant to reduce confusion in terminology. Some of
these traditional terms, however, are used as common names in the
classification or in descriptions of the units.
Subclass:
Subclasses are categories within each class comprised of areas in
which the main vegetation is morphologically similar. For the
classes of forests, woodlands, dwarf shrublands and shrublands
the similarities are based on these factors:
1. evergreen;
2. deciduous or mixed.
For the class of herbaceous the similarities are based on:
1. tall grasses, more than 1.0 meter in height;
2. medium-tall grasses, from 0.5 to 1.0 meter in height;
3. short grasses, less than 0.5 meter in height;
4. forbs.
Group:
Groups are categories within each subclass which may be based on
any of the following:
For forests, woodlands, and shrublands:
1. climate, e.g., tropical, temperate, subpolar;
2. morphology, e.g., broad-leaved, sclerophylous,
needle-leaved.
Formation: Formations are categories within each group comprised of areas in which the vegetation similarities are based on any of the following criteria. Tree size and crown shape: Non-giant forests are those 5-50 meters in height having 1. rounded crowns, e.g., Pinus echinata 2. conical crowns, e.g., Juniperus virginiana Life zone: 1. temperate lowland 2. montane 3. alpine Substrate: 1. alluvial 2. serpentine Kinds of associated vegetation, e.g., broad-leaved forest with or without evergreen needle-leaved trees, or with or without succulents. Amount and kind of understory.
Cover type: Cover type is a group of plant community types having the same primary dominant species and similar physiognomy; an aggregation of plant community types.
Community type: The community type is an assemblage of plant species that interact at the same time and place and have defined species composition and physiognomy, regardless of seral stage; usually named by the names of the species that dominate the canopy layer.
A modification of the national model was made to allow users to more easily find wetland community types: at the highest level we distinguished Terrestrial, Palustrine, Riverine and Lacustrine systems (following Cowardin, 1979). In the national model these have been incorporated lower in the hierarchy (at the formation level). However, there are enough wetland types in the Arkansas classification that such an approach adds to confusion by users. If a user desires, it should be straightforward to place these groups at the formation level. In either event it should be understood that these units are a classification of the emergent vegetation of wetland communities, not a classification of the physical wetland communities.
By agreement of the review committee, as well as National GAP guidance, community types that are the result of human activity, e.g., urban and agricultural areas, are not included in this scheme, but widespread successional communities that may result from previous human disturbance are included.
The first draft of the vegetation classification system was generated by correlating several of the existing classification systems that are in use within Arkansas. These included the vegetation classification of Arkansas developed by Dale (1986); the system used by the Society of American Foresters (Eyre, 1980) and that developed by Kuchler (1964), the unpublished classifications used by the U.S. Forest Service and the Arkansas Natural Heritage Commission (based on Pell, 1981), and the unpublished vegetation cover types used by the Arkansas Game and Fish Commission. We also referred to other sources, such as the unpublished southeast and midwest regional ecological community classification systems of The Nature Conservancy, and the draft ecological community classification system for Tennessee GAP Project.
These cross-correlated units were placed into the appropriate levels (either cover type or community type) of the modified UNESCO system, and revised several times based on review by the GAP vegetation classification committee.
The final vegetation classification system of Arkansas contains approximately 215 community types (Table 1). Approximately 75 of the community types are within the wetland systems, while the remainder are within the terrestrial system.
Of the higher levels in the hierarchy, 32 physiognomic types are recognized at the formation level and 57 floristic types at the cover type level. The authors were concerned about the dramatic increase in units from 57 at the cover type level to 215 at the community type level, so an intermediate level comprising 94 units was created to provide an intermediate level of detail.
In order to add to the objectivity and usefulness of the classification, the committee will develop a list of type stands or example stands that will be documented with vegetation data and will be available for further review and research. The near- term documentation will be used in classifying satellite imagery to produce the GAP vegetation map.
Review and revision of this classification by a large and diverse committee has demonstrated that it is easily compressed to provide fewer units or extended to provide more detail, and is therefore highly flexible. Furthermore, its emphasis on plant community composition and structure makes it well adapted for use with satellite and aerial imagery. The U.S. Forest Service is presently developing a site classification system which will complement this vegetation-centered approach.
Because of the flexibility, clear focus and suitability of this system for use with remote sensing data, we suggest its use in future vegetation and habitat studies in Arkansas.
Besides the authors, the GAP Vegetation Classification Committee included Belinda Ederington and Chris Ware from Arkansas Game and Fish Commission, Carl Minehart from Ozark National Forest, Bill Pell from Ouachita National Forest, Lance Peacock from Arkansas Chapter, Nature Conservancy, Philip Tappe from University of Arkansas at Monticello, Bob Bennett from Arkansas State University, and Jim Grant from the Arkansas Forestry Commission. Edward E. Dale also offered comments on an earlier draft. This research was supported by the National Biological Survey through the Arkansas GAP Project.
Cowardin, L. M., V. Carter, F. C. Golet, and E. T. LaRoe. 1979. Classification of wetlands and deepwater habitats of the United States. USDI Fish and Wildlife Service, Washington, 103pp.
Dale, E. E. 1986. The vegetation of Arkansas (text and map). Ark. Nat. 4(5):6-27.
Eyre, F. H. (ed.) 1980. Forest cover types of the United States and Canada. Society of American Foresters. Washington, 148pp.
Foti, T. L. 1974. Natural divisions of Arkansas. 11-34 in Arkansas Natural Area Plan. Arkansas Department of Planning. Little Rock, AR. 248pp.
Jennings, D. M. 1993. Natural Terrestrial Cover Classification: Assumptions and Definitions. GAP Analysis Technical Bulletin 2. U.S. Fish and Wildlife Service, Idaho Cooperative Fish and Wildlife Research Unit, Univ. of Idaho, Moscow, Idaho, 28p.
Kuchler, A. W. 1964. Potential natural vegetation of the conterminous United States. Amer. Geog. Soc. Spec. Pub. 36. 156 pp.
Moore, D. M. 1959. Trees of Arkansas. AR Forestry Commission. Little Rock, 142 pp.
Nelson, P. W. 1985. The terrestrial natural communities of Missouri. MO Dept. Nat. Res., MO Dept. Cons., MO Nat. Areas Comm., Jefferson City, 197 pp.
Pell, W. F. 1981. Classification and protection status of remnant natural plant communities in Arkansas. Proc. Ark. Acad. Sci. 35:55-59.
Raunkier, C. 1934. The Life Forms of Plants and Statistical Plant Geography. Oxford: Clarendon.
Ruble, E. 1930. Pflanzengesellschaften der Erde. Bern/Berlin: Huber.
Scott, J. M., F. Davis, B. Csuti, R. Noss, B. Butterfield, C. Groves, H. Anderson, S. Caicco, F. D'Erchia, T. C. Edwards, Jr., J. Ulliman, and R. G. Wright. 1993. Gap Analysis: A geographic approach to protection of biological diversity. Wildlife Monographs No. 123:1-41.
United Nations Educational, Scientific and Cultural Organization. 1973. International classification and mapping of vegetation. Paris. 35pp.
Whittaker, R. H. 1978. Classification of Plant Communities. Dr. W. Junk Publishers, The Hague: Netherlands. 408pp.