USGS Home
Contact USGS
Search USGS

Table 8. Total numbers of perennial, introduced, and annual (includes biennial) or native plant taxa in communities in wet-meadow, shallow-marsh, and deep-marsh zones of sample wetland basins in good- and-poor quality watersheds, 1992-1993

The initial direct disturbance to a wetland by repeated cultivation for seedbed preparation is severe. Cultivation probably affects all stages in the plant regeneration cycle, an important mechanism in the maintenance of plant species diversity (Grubb, 1977). Plowing severs rhizomes of the native perennial hydrophytes and overturns and dries the sod; repeated discing and harrowing may follow for a year or more before planting. These operations totally eliminate most of these plants. After this stage, farmers regularly cultivate entire basins of lesser water permanence for crop production or to control weeds whenever water levels permit. Farmers also regularly cultivate the wet-meadow and shallow-marsh zones of basins of greater water permanence. Indeed, during this study, over 77 percent of the area of all wet-meadow zones in poor-quality watersheds had been cultivated recently. These cultivation stresses are thus in a sense predictable and do not allow the original native plant community to re-establish itself. However, some plants are adapted to repeated disturbances of bottom substrates; such disturbances may eliminate competitive dominants, thereby allowing subordinates to occupy the disturbed sites (Wilson & Keddy, 1986). In frequently cultivated wetlands in the PPR, these subordinates consist of a few rapidly maturing annuals (e.g., Gratiola neglecta) and relatively short, deep-rooted perennials (e.g., Polygonum amphibium).

Besides direct tillage, additional stressors common to plant communities in frequently cultivated wetlands are inputs of silt, pesticides, and fertilizers. Silt comes from adjacent uplands, but wetlands often receive direct inputs of pesticides and fertilizers. It is generally unknown whether these stressors have antagonistic, synergistic, or additive effects (sensu Turner, 1985) on the structure and function of these communities. However, fertilizers normally increase productivity and decrease species richness in most wetland plant communities studied (Vermeer & Berendse, 1983). Herbicides commonly used on row crops in the region seem to greatly decrease both production and species richness of hydrophytes.

Farmers often leave wetlands of greater water permanence idle because potential crop yields are poor due to high salinities or sandy bottom substrates, or because of the potential for miring or damaging equipment on large boulders. Nevertheless, these basins also accumulate agricultural nutrients, and many have large amounts of standing dead vegetation. Deposits of silt from the adjacent cropped uplands sometimes accumulate in the peripheral zones to form a barrier or build mud deltas in interior zones. We noted silt accumulations that virtually eliminated the wet-meadow zone of several study wetlands.

Idle conditions cause many changes to the wetland biota. Woody plants, particularly Salix spp. and Populus spp., invade idle prairie wetlands, especially where earlier cultivation disturbs wet-meadow zones. Idle coastal marshes show decreased plant species richness and numbers of vegetation types present; vegetation mosaics there tend to be coarse-grained (Bakker, 1985, Andresen et al., 1990). Idleness also allows formation of monotypic stands of robust emergents that shade out shorter plants and reduce avian diversity and abundance (Jones & Lehman, 1987; Hellings & Gallagher, 1992). For some plants, such shading can be more important than the effects of herbivory and competition on seedling establishment (Bergelson, 1990). Buildup of litter and organic material from emergent species in prairie wetlands can reduce water depth or eliminate shallow-water areas (Ward, 1942, 1968; Walker, 1959; Hammond, 1961).

Native plants in the region are adapted to hydrologic changes, fire, and mammalian herbivory. In pre-agricultural times, these natural forces probably created some sort of normal, but unknown, homeostatic behavior of the grassland ecosystem. Livestock grazing is currently the dominant land-use practice in grassland-dominated watersheds in the region, although haying is also common. The most common sources of livestock water in regional pastures are natural basin wetlands. Nevertheless, ranchers construct many livestock-watering facilities, some in natural basin wetlands. Prairie wetlands are basically wet grasslands. Lack of grazing in grasslands is now abnormal; under such situations obligate grazophiles can disappear (McNaughton, 1979, 1986). Ratios of standing crop to litter can fall as plant communities age without grazing (Bazely & Jeffries, 1986). Conversely, livestock grazing, particularly in spring and fall, maintains species richness in meadow grasslands (Smith & Rushton, 1994). Grazing in long-idled salt marshes slowly enhances species diversity and creates fine-grained vegetation mosaics (Bakker, 1985). Grazing by cattle of monodominant stands of Typha Xglauca Godr. decreases live stems, dead stems, and litter (Schultz, 1987). Grazing thus may remove much organic matter and create open water areas where submersed plants flourish. Although we did not analyze the effects of grazing, most of the basins we studied in good-quality watersheds were grazed by cattle. Such basins generally had higher plant species diversity. There is a threshold to tolerance for grazing, however, even in prairie wetlands, because long-term overgrazing and trampling can reduce the shallower zones to nearly bare soil.

Although an average of less than four percent of the area of wet-meadow zones we studied was currently mowed, mowing and removing emergent vegetation (haying) for livestock feed or bedding is a common practice in the PPR. Most mowing occurs in watersheds that farmers seed to perennial forage crops such as alfalfa or mixtures of alfalfa and grasses. However, farmers also often use larger wetlands in annually tilled watersheds for hay production. Livestock producers consider some native species, such as Scolochloa festucacea, excellent forage. They sometimes seed introduced species, such as Phalaris arundinacea, in wetland basins. Forage production depends on wet land zone and water levels before and during harvest. Producers mow many wet-meadow zones nearly every year, whereas they can cut hay (sometimes only used for bedding) from deep-marsh zones only after a series of dry years. Most observers agree that long-term use of basins for hay tends to increase the abundance of certain emergents (Smeins, 1967; Walker & Coupland, 1968,1970; Stewart & Kantrud, 1972).

Cultivation of various emergent wet-meadow and shallow-marsh communities during dry years seems to create coarse-grained vegetation mosaics with fewer communities. Farmers regularly convert old annular stands of weedy annuals, drawdown species, or early successional vegetation to crop or fallow monotypes. Although not statistically significant, communities we studied in good-quality watersheds tended to be larger. This probably reflects the greater use for pastures, rather than cropland, of tracts containing larger wetlands. Landowners also used some larger wetlands in poor-quality watersheds for hay or pasture. If farmers do not cultivate wet-meadow zones and if siltation is not severe, these wetlands appear similar to those in good-quality watersheds. However, farmers cultivate nearly all basins in poor-quality watersheds for crop production or weed control whenever bottom substrates are dry.

Greater percentages of standing dead vegetation in the deeper, more permanent zones likely reflect reduced accessibility of these zones to livestock and farm equipment. We expected much greater amounts of dead vegetation and increased litter depth in deep-marsh zones in poor-quality watersheds because of siltation and lack of grazing, cultivation, or other mechanisms that reduce plant biomass. However, sample sizes for deep-marsh zones were too small (n = 3 and n = 1 for 1992 and 1993, respectively) to detect differences. Agricultural and pastoral operations tended to reduce standing dead vegetation in the less-permanent zones of both watershed types. Livestock grazing pressure in these zones in basins in good-quality watersheds probably was insufficient to greatly reduce standing dead vegetation.

As with standing dead vegetation, litter core lengths were naturally greater in zones of greater water permanence, likely because of reduced accessibility to machinery and livestock. Another factor could be greater biomass production because the more-permanent zones usually support taller, more robust plant species. Nevertheless, the presence of surface water limits access by machinery more than cattle. Thus, litter depth among zones varied significantly only in basins in poor-quality watersheds because these basins were usually ungrazed. There were no significant effects due to watershed quality alone. The irregular destruction of litter by machinery in basins in poor-quality watersheds likely was not much greater than that caused by the often season-long livestock hoof action and herbivory that compress or reduce the litter layer in grazed basins. A single pass by tillage equipment often tears narrow openings in vegetation and can leave much of the litter layer intact. Also, basins in poor-quality watersheds often receive inputs of fertilizer from their adjacent cropped uplands that could increase plant biomass in areas where root systems are not directly destroyed by tillage.

The greater percentages of unvegetated bottom found in wet-meadow zones of wetlands in poor-quality watersheds undoubtedly reflect the effects of cultivation. Communities in these watersheds tended to have large amounts of unvegetated bottom regardless of water levels. Herbicides can further reduce plant populations in cultivated wetlands. Farmers use herbicides directly on cropped wetlands but also sometimes treat non-cropped wetlands to prevent introduced perennial grasses with hydrophytic tendencies, such as Agropyron repens, from spreading to the uplands. Livestock grazing, except when extremely intense such as in heavily trampled barnyards or feedlots, seldom creates unvegetated bottoms.

Percentage open water naturally increased in all zones as water levels rose after the drought of 1992 and preceding years. The expectation that basins in good-quality watersheds would have greater amounts of open water held for all zones during the relatively dry year of 1992. But in the following relatively wet year, the differences were less obvious, particularly in zones of lesser water permanence. Wet-meadow zones in basins in poor-quality watersheds had more open water than basins in good-quality watersheds in 1993. We attribute this to the flooding of bare tilled soils created by cultivation.

In summary, in poor-quality watersheds, the hydrology of prairie wetlands combines with a variety of agricultural practices to create unnatural, coarse-grained patterns in wetland vegetation or basins devoid of vegetation. Open water or unvegetated areas can lie adjacent to areas with greater amounts of litter than are found where grazing is the predominant land use.

Plant species richness was lower in wet-meadow zones in basins in poor-quality watersheds. We generally found more annuals, both native and introduced, during the drier year. This likely reflects the increased occurrence of drawdown species that pioneer on bare mud flats and upland species that invade wetlands during drought. An obvious pattern in tilled wetlands is the replacement of native perennials with cultivated crop plants. However, tillage also creates favorable conditions for introduced perennials as well as native and introduced annuals, of which many are classified as noxious weeds. These are usually the target of additional tillage operations in and around the basins. Herbicides applied to crops growing in the basins may further reduce species richness. Special herbicides may kill nearly all vegetation ("chemical fallow") in uplands and wetlands scheduled for summerfallow.

We conclude that amounts of unvegetated bottom and plant species richness successfully discriminated between wetlands in good- and poor-quality watersheds. Intensively tilled prairie wetlands with large amounts of unvegetated bottom show poor use by aquatic or marsh birds. For example, in the 1960s, these wetlands comprised about one-fourth of the total area of basin wetlands in the North Dakota portion of the PPR. During this period, only 4.4 percent of the ducks (Kantrud & Stewart, 1977) and less than 0.5 percent of the other birds (Kantrud & Stewart, 1984) were observed on these basins. Nevertheless, there is a need for additional indicators of the general quality of prairie wetlands. Most valuable would be indicators that could be photographed or otherwise remotely sensed. A ideal set of indicators could detect the absence of stressors, as well as the presence of structures or functions, of known value to major groups of organisms. We recommend that more data from a larger, random sample of wetlands, rather than from wetlands in the extremely good- and poor-quality watersheds we studied, be collected during the operational phase of EMAP. Such efforts, if successful, should allow the construction of empirical models that will use explanatory variables measured on the uplands, such as ratios of major land use practices within the wetland watersheds, to predict indicators of wetland quality in those wetlands.