USGS - science for a changing world

Northern Prairie Wildlife Research Center

  Home About NPWRC Our Science Staff Employment Contacts Common Questions About the Site

Effects of Vegetation Manipulation on Breeding Waterfowl in Prairie Wetlands — A Literature Review

Management of Wetland Vegetation for Waterfowl


Komarek (1976) stated that the fire ecology of wetlands was sorely in need of scientific study. General references (Kozlowski and Ahlgren 1974; Wright and Bailey 1982) indicate that burning of marsh vegetation releases nutrients, opens the canopy and detrital layer, and allows for increased insolation and resultant earlier warming of bottom soils. Biological productivity usually increases following fire, even though plant species composition may be altered. Little change in species composition usually occurs when perennial species with meristem at or below ground level are burned during their dormant period.

Fires were common in prairie wetland vegetation in the early 19th century, as evidenced by the accounts of early traders and travelers. For example, in 1803 Henry and Thompson (1965) recorded fire rushing through "low places covered with reeds and rushes." In 1858 or 1859, Boller (1972) saw a large conflagration spread for many miles after being set by American Indians in "dry rushes in the prairie bottoms." Denig (1961), writing about his experiences during 1833-54, noted that fire would sweep over ice through wetland vegetation.

Impacts on Vegetation

Little is known about the environmental effects of fire in prairie wetlands. Much of the available information is obtained from general observations on wetlands where the fire frequency or season was unknown, or from fires set in a variety of vegetation types, usually on a nonexperimental basis. Hence, the results are often inconsistent and of minimal predictive value. Early studies by Lewis et al. (1928) indicated the changes in a few plant communities in central Alberta that could be expected in the presence or absence of a burning regime. Furniss (1938) noted that heavily lodged stands of Typha latifolia and Scirpus validus could be removed by fire in Saskatchewan wetlands. Ward (1942) found that dense beds of Phragmites australis in Manitoba wetlands could be opened up by either spring or late summer burns, but that only late summer burning killed the "roots" (rootcrowns). Grange (1949) observed that smartweeds (Polygonum spp.) disappeared because of competition from Carex spp., Typha spp., Phragmites australis, and various grasses. He stated that burning was probably the only effective method of stimulating smartweed growth in Wisconsin wetlands. Truax and Gunther (1951) used fall and winter burns to control undesirable vegetation at Horicon Marsh, Wisconsin. Annual burning was used to maintain the Carex spp. community in other Wisconsin wetlands (Thompson 1959). Tester and Marshall (1962) saw little change in species composition of wetland vegetation when Minnesota marshes containing low fuel volumes were burned. Smeins (1965) listed a few wetland plants found in North Dakota marshes with a history of burning. Schlichtemeier (1967) successfully removed dead stems of P. australis and Scirpus spp. with a winter burn, even though snow covered the bases of the plants. Vogl (1967) found burning generally favorable as a means of controlling woody plant invaders in Wisconsin wetlands. Smith (1969) stated that Typha spp. could quickly be destroyed by fire in Alberta wetlands. Beule (1979) concluded that burning was an ineffective control for Typha spp. in Wisconsin wetlands unless the peat layer was also burned. Gorenzel et al. (1981) found that fire failed to kill Typha spp. and S. americanus in a Colorado wetland. Thompson (1982) studied the seasonal effects of burning P. australis stands in a Manitoba wetland, and concluded that the changes in species composition and productivity produced by fall burns were intermediate between those produced by spring or summer burns.

In seasonal prairie wetlands, Stewart and Kantrud (1972) thought Polygonum coccineum increased after burning. However, Millar (1973) found no change in stands of Carex atherodes, Scolochloa festucacea, and Eleocharis palustris after repeated burning, which indicates these common plants of seasonal wetlands are extremely fire tolerant.

The aforementioned studies and observations do not provide managers with definitive, quantifiable information needed to formulate burn prescriptions in prairie wetlands. Research on prescribed burning for these wetlands for wildlife production was urged by Ward (1968) and Weller (1978), but to date almost all marsh burning for improvement of Waterfowl habitat has been done on migration or wintering areas (Sanderson and Bellrose 1969; Rutkowsky 1978).

Effects on Breeding Waterfowl

There is little substantive information about fire as it affects use of prairie wetlands by breeding waterfowl. Bennett (1938) and Furniss (1938) probably were the first to postulate that some benefits to breeding waterfowl could accrue from marsh burning. Bennett recommended shoreline burning to open dense stands of emergents to increase foods for blue-winged teal (Anas discors), whereas Furniss noted that crow predation on Saskatchewan duck nests may be less in marshes where heavily lodged, old-growth Typha spp. and Scirpus spp. stands were opened up or rejuvenated by fire. Cartwright (1942) suggested that burning dense, matted vegetation in Manitoba meadows would improve use by nesting ducks. Ward (1942) stated that burned openings in dense stands of Phragmites australis were heavily used by breeding ducks at the Delta Marsh, Manitoba. Grange (1949) noted that plants that produced seeds readily eaten by ducks were easily lost to competition from other plants and considered burning the only effective way to control plant succession in Wisconsin wetlands. In South Dakota, Evans and Black (1956) noted that burning often improved use of wetlands by pairs of breeding waterfowl. Drewien and Springer (1969) observed that many burned wetlands lacked roosting cover in the spring, but that overall use of the wetlands by breeding pairs was not much affected.

Only a few experimental marsh burns have been conducted to study the effects on breeding waterfowl. Ducks showed increased use of winter-burned stands of P. australis and Scirpus acutus in wetlands in the Nebraska Sandhills (Schlichtemeier 1967). Ward (1968) found that, in a Manitoba wetland, fire opened up old stands of P. australis that formerly were almost devoid of duck nests and stimulated growth of Scolochloa festucacea, which supported highest duck nest densities. However, duck nest success was low the first year after a fire on low, Poa pratensis prairie in Iowa (Messinger 1974). A more detailed study was conducted by Bjork (1976), who observed that, in a Swedish wetland "severely damaged" by overgrowth of Phragmites australis and Carex acuta, burning and mechanical methods of vegetation control resulted in much greater use of the area by breeding ducks, probably because of the presence of higher populations of chironomid insects. Prescribed burning of P. australis and Typha spp. during the dormant season is practiced on some National Wildlife Refuges (Fig. 2).

Figure 2: Photo of biologist administering a prescribed burn
Fig. 2.  Prescribed spring burn being used to open a dense stand of Phragmites australis on the J. Clark Salyer National Wildlife Refuge. (Bottineau Cty., North Dakota, 4.5 miles southeast of Westhope; photo by R. Giese.)

In the absence of water control, burning of vegetation in wetlands that naturally retain water only seasonally probably cannot be justified as a management practice for breeding waterfowl (Diiro 1982). Diiro found that increased early-season productivity of plants and invertebrates in basins burned the previous fall was offset by a general scarcity of water caused by the reduced snow-trapping ability of burned vegetation. In addition, snow accumulations tend to crush the softer vegetation found in seasonal wetlands, causing them to maintain an open or semi-open aspect during most springs. However, I believe that in pristine times vegetation in such wetlands would have burned more frequently than that found in more permanent wetlands. Long-term experiments on the effects of fire in the less permanent types of wetlands are needed.


Much more is known about the effects of grazing than of burning on wetland plant communities. Unless unusually severe, grazing results in greater plant species diversity and the development of more intricate patterns and sharper boundaries among plant communities (Bakker and Ruyter 1981). Livestock trampling may affect the height and density of wetland vegetation more than consumption (Hilliard 1974). Overgrazing may cause a decrease in primary production (Reimold et al. 1975), an increase in water turbidity (Logan 1975), and areas devoid of vegetation (Bassett 1980), as shown in Fig. 3. Adaptations of wetland plants to grazing include nodal rooting and unpalatability (Walker 1968; Walker and Coupland 1968). Marshes often show greater vegetative response to grazing than upland communities (Bassett 1980). Lists of species that increase with, or are tolerant of, grazing in wetlands in or near the prairie pothole region have been published by Evans et al. (1952), Smith (1953), Harris (1954), Smeins (1965), Dix and Smeins (1967), Walker and Coupland (1968), Stewart and Kantrud (1972), and Millar (1973).

Figure 3: Photo of sheep grazing near a prairie wetland
Fig. 3.  Long-term overgrazing can destroy nearly all emergent vegetation in those shallow prairie wetlands having firm bottom soils. (Dickey Cty., North Dakota, 9 miles northwest of Forbes; photo by H. F. Duebbert.)

Effects on Breeding Waterfowl

Most active management of waterfowl habitat through grazing by domestic livestock occurs on the wintering grounds, where the usual goal is to increase the availability of seeds of annual food plants (Griffith 1948; Neely 1967; Ermacoff 1968; Sanderson and Bellrose 1969). The effects of grazing on the quality of wetland habitat used by breeding waterfowl have received much attention during general investigations but little by experimental design. Early work by Bennett (1937) and Furniss (1938) on wetlands in Iowa and Alberta, respectively, indicated that overgrazing degraded habitat for ducks that nested along marsh borders or over water, but that nest density increased and egg predation by crows was less when densely vegetated shorelines were opened up by livestock. Sowls (1951) noted that ungrazed edges of wetlands attracted few breeding ducks and stated that ducks might increase if such areas were moderately grazed. Disturbed shorelines that would otherwise have supported dense growths of Typha spp. and Scirpus spp. probably supported higher densities of dabbling ducks in South Dakota stock ponds (Bue et al. 1952). Glover (1956) concluded that light-to-moderate grazing of shorelines after 1 July would not harm their value to waterfowl. Studies of man-made wetlands confirmed the deleterious effects of overgrazing on use of these wetlands by breeding ducks (Shearer 1960; Uhlig 1963). A study conducted in South Dakota (Sand Lake National Wildlife Refuge, unpublished annual reports, 1957-61) reported increased use of grazed shorelines by breeding ducks, especially green-winged teal (Anas crecca), northern pintail (A. acuta), and blue-winged teal. Salyer (1962) found that grazing was less harmful to breeding ducks when water areas increased in number and depth. Light grazing was recommended by Munro (1963) to help open Typha stands, thereby improving prairie wetlands for breeding waterfowl. Poston (1969b) postulated that light-to-moderate grazing would result in near optimum conditions for northern shoveler (A. clypeata) on Alberta wetlands. The moderately grazed portion of the wetland shown in Fig. 4 represents an interspersion of cover and open water that is attractive to waterfowl.

Figure 4: Photo of fence dividing a grassland into idle (left) and grazed (right) portions
Fig. 4.  Moderate cattle grazing created the semiopen and more diverse plant community shown on the right side of the fence; the portion on the left remains idle. (Stutsman Cty., North Dakota, 2.5 miles east-southeast of Woodworth; photo by K. F. Higgins.)

Drewien and Springer (1969) were probably the first to report that breeding ducks move to roost in more heavily vegetated wetlands at night. These wetlands contained patchy, moderately dense stands of Carex spp., Polygonum coccineum, Scirpus spp., Scolochloa festucacea, and Typha spp. The authors believed that lack of roosting cover did not limit densities of blue-winged teal on their South Dakota study area; sufficient roosting cover was always present because of other land-use practices, and even the overgrazed wetlands grew acceptable amounts of cover as the season progressed. During the day, the teal were found at higher densities on idle than on grazed wetlands; however, the authors inferred that this related to the proximity of upland nesting cover to the idle wetlands, rather than to differences among wetlands.

Kirsch (1969) found that pair use was lower on grazed North Dakota wetlands, but the differences between grazed versus idle wetlands were not significant. Cattle disturbance of duck nests was thought to be important during the study. Gjersing (1971) found high losses of duck nests due to livestock trampling around Montana reservoirs when the nests were within 7 yards of the shoreline. Winter grazing in Utah wetlands seemingly did not affect nesting dabbling ducks, but probably was harmful to divers (Hilliard 1974). In Denmark, Moller (1975) recommended grazing of wetlands during the nonbreeding season, and Fog (1976) believed that great portions of ungrazed marshes were lost to breeding ducks by the invasion of Phragmites australis. By using multiple regression analysis, McEnroe (1976) found that the percentage of shoreline grazed on natural wetlands was positively related to density indices (pairs per wetland) for the mallard (A. platyrhynchos), gadwall (A. strepera), and blue-winged teal; however, intensity of grazing was negatively associated with densities of blue-winged teal and redhead (Aythya americana). A similar analysis of dabbling duck use of man-made wetlands in South Dakota showed species-specific preferences associated with differences in vegetation height, density, or diversity caused by grazing (Flake et al. 1977). The northern shoveler made greatest use of pastured wetlands in England (Thomas 1980). The highest concentrations of breeding canvasback (A. valisineria) seen by Stoudt (1982) in a Manitoba study area were usually associated with pastured wetlands containing open or half-open surfaces and stands of Scirpus acutus.

Waterfowl Broods and Grazing

Relations between grazing and use of wetlands by waterfowl broods have also received some attention. Girard (1941) noted that broods would benefit if wetland shorelines in Montana were protected from overgrazing. In Manitoba, the Typha-choked wetlands containing less than 10% open water received almost no use by duck broods (Evans et al. 1952). A history of light to moderate spring and fall grazing resulted in the open Carex spp. and Scolochloa festucacea habitat which was preferred by broods; ponds with broken stands of Scirpus acutus resulting from moderate to heavy grazing throughout the growing season were also extensively used. Broods were far more abundant on South Dakota livestock ponds with grassy shorelines than on those with mud shorelines created by overgrazing (Bue et al. 1952). Short emergent growth in sparse stands caused by grazing also provided the best brood habitat on eastern Montana stock ponds (Smith 1953). Harris (1954) observed that heavily grazed areas dominated by Scirpus spp. and Juncus spp. received the most use by broods in Washington potholes. Overgrazing, especially of small wetlands, created unsuitable brood habitat in South Dakota (Evans and Black 1956). Keith (1961) noted a large increase in brood density after partial destruction of Typha spp. stands by herbicides on Alberta impoundments, and he recommended combining grazing and herbicide applications to rejuvenate marsh edges for ducks. In Colorado, broods used lightly-to-moderately grazed wetlands far more than either moderately-to-heavily grazed wetlands or those that lay idle (Hopper 1972). Evans and Kerbs (1977) identified South Dakota impoundments having gently sloping shorelines and light use by livestock as water areas that would develop the natural vegetation structure preferred by broods. Nonetheless, if such wetlands contain appreciable amounts of Typha spp., its influence on brood use may be negative (Mack and Flake 1980). Canvasback broods in Manitoba reached highest densities in pastured wetlands containing less than 33% emergent vegetation (Stoudt 1982). Hudson (1983) found that duck brood densities were positively related to the amount of vegetation in Montana livestock ponds, but no ponds totally covered with emergents were censused, and all ponds were grazed.

Invertebrate Food and Grazing

Only a few investigators have mentioned response by invertebrate animals to herbage removal by livestock. Munro (1963) stated that grazing of Typha-dominated prairie wetlands would increase the planktonic algae that are the primary foods of invertebrates. Hopper (1972) believed that light-to moderate grazing of flooded emergent vegetation would provide invertebrate foods for duck broods and also allow them easier access to shoreline feeding areas. Some very large invertebrates in salt marshes (such as crabs) may decrease under heavy grazing, but recovery is probably very rapid once grazing pressure is lessened (Reimold et al. 1975). Decreases of invertebrate animals caused by grazing of wetlands probably occur only when livestock are present in enough numbers to destroy aquatic vegetation (Logan 1975).

Invertebrates have been known to be important in the nutrition of breeding ducks since the early 1960's (Voights 1973). Indeed, invertebrate numbers and taxa may surpass all other measured physical and biological variables as indicators of wetland quality for breeding ducks (Joyner 1980), although it may be necessary to determine the behavior and distribution of the invertebrates in order to accurately predict which microhabitats will attract feeding ducks (Joyner 1982). Management of invertebrates for waterfowl was reviewed by Schroeder (1973) who recommended manipulation of cover through water control, grazing, and burning. He cautioned that such manipulations should favor a good interspersion of cover types without creating excessive siltation, undue fluctuations of water levels during the nesting season, extensive reductions in plant abundance and diversity, and contamination of water supplies by toxic chemicals. When the invertebrate resources of prairie wetlands are manipulated by mechanical methods (Murkin 1979; Kaminski and Prince 1981a, 1981b), treatments are expensive and the response of breeding waterfowl is often small or of short duration.

Previous Section -- Problems
Return to Contents
Next Section -- Research Needs

Accessibility FOIA Privacy Policies and Notices

Take Pride in America logo logo U.S. Department of the Interior | U.S. Geological Survey
Page Contact Information: Webmaster
Page Last Modified: Saturday, 02-Feb-2013 07:05:02 EST
Menlo Park, CA [caww54]