Northern Prairie Wildlife Research Center
During the past 40 years several classification systems have been applied to wetlands in the glaciated prairie region of the United States and Canada. Metcalf (1931) used a system based on salinity and vegetation to differentiate several types of prairie ponds and lakes in North Dakota. Hayden (1943) followed the concepts of ecological succession and classified Iowa wetland communities according to their position in the evolutionary sequence from early hydroseres to prairie climax. Bach (1950) described a system of wetland types in North Dakota on the basis of their longevity or permanency. Nord, Evans, and Mann (1951) evolved a "chain type" wetland classification using Bach's permanency types in combination with a series of other factors including density, distribution, and species composition of marsh plants when classifying wetlands in Minnesota, North Dakota, and South Dakota. Leitch (1966), using Bach's basic permanency types and the criteria employed in the system of Nord, Evans, and Mann, added various physical, ecological, and historical factors when describing wetlands in Canada.
Mason (1957) presented a classification outline for wetlands based on a sequence of environmental factors treated in order of decreasing importance and including water movement, water permanence, relation of cover to open water, and water chemistry. Millar (1964) proposed a complex system for the prairie wetlands of Canada which combined the vegetational factors used by Stewart and Kantrud (1963) with topographic features, including basin area, capacity, drainage, and configuration.
The system of Martin et al. (1953) currently used by the Bureau of Sport Fisheries and Wildlife is based primarily on water depths during the growing season, cover interspersion, and occurrence of certain groups or species of wetland plants. Evans and Black (1956) combined the system of Bach with that of Martin et al. for adaptation to South Dakota wetlands. Extensive field experience has shown that the classification system of Martin et al. is too general for detailed investigations of wetlands. It has often been misinterpreted by placing too much emphasis on water depth and cover interspersion. Water depth of a given wetland type varies considerably; by itself it is generally a poor indicator of prairie wetland types and should not be equated with water permanence. Water depths and related stages of cover interspersion often change drastically from year to year and season to season because of unstable climatic conditions. The descriptions by Martin et al. of water relationships of the designated types found in the glaciated prairie region do not always correspond to the habitat requirements of the characteristic plant species listed. Lastly, insufficient distinction is made of the complex of vegetational zones within wetlands and of the particular relations of vegetation to various environmental factors.
Our ecological investigations of wetlands in central North Dakota from 1961 through 1966 indicate that the use of prairie ponds and lakes by waterfowl is strongly influenced by water permanence, depth, and chemistry, and by land use. Although these factors are complex and interrelated, any marked variations are usually reflected in differences in life form, cover interspersion, species composition, and species dominance. These vegetational differences are readily discernible in the field, and they have been used as the principal criteria in the classification system described here. This publication supersedes a preliminary paper on this subject (Stewart and Kantrud, 1969).
For purposes of this publication, natural ponds and lakes refer to wetlands occurring in natural undrained basins or kettles. Ponds are arbitrarily defined as natural nonfluvial wetlands less than 50 acres in area; lakes are larger than 50 acres. Other wetland types in the glaciated prairie region are not covered by this classification; they include natural fluvial habitats and manmade wetlands such as stock ponds, dugouts, reservoirs, and sewage lagoons.
The glaciated prairie region includes parts of the northern prairies in the Central Lowland and Great Plains which were covered with glacial drift deposits during the middle advances of the Wisconsin stage glaciation (Lemke et al. 1965). It is characterized by numerous undrained depressions and is well represented in southern Alberta, southern Saskatchewan, extreme southwestern Manitoba, extreme northeastern Montana, northern and east-central North Dakota, eastern South Dakota, and small portions of western Minnesota and northwestern Iowa. The glaciated prairie region referred to here does not include the contiguous transitional belt of aspen parkland, which is usually considered to be an ecotone between prairie and boreal forest.
While the system of Martin et al. does not provide a precise, dynamic classification of prairie wetlands, it has been used to categorize in general terms the average long-term condition of the most common prairie wetlands. Furthermore, it is referred to in certain legal documents (Public Law 87-732 and the Reuss Proviso of the Agricultural and Related Agencies Appropriations Acts, 1963 et seq.) and must be used for classification purposes, until otherwise modified, in matters concerning Federal drainage referrals for all wetland types in the Dakotas and Minnesota and for Types 3, 4, and 5 wetlands nationwide. The system described here, however, can be applied in either a detailed or a broad manner, as explained in the section on Application of the Classification System. A broad interpretation of this classification would correspond to the system of Martin et al., and would be no more difficult to apply. It would also have the advantage of providing a more precise and realistic means of ecologically classifying wetlands in the glaciated prairie region. While most readily applied to conditions existing at the time of a survey, it can be used during the nongrowing season to interpret conditions during the previous growing season. When applied in a year of average water conditions or over a span of years encompassing the full range of conditions, it may be used to accurately determine the average class of a wetland and thus meet the classification requirements imposed by current legislation.