Northern Prairie Wildlife Research Center
We conducted this study using water-level data collected from 36 wetlands located throughout the PPR of North and South Dakota. Overall characteristics of wetlands in the PPR have been described in detail by van der Valk (1989). The selected wetlands were equally distributed among cropland and grassland dominated landscapes. In addition, we stratified our sample by wetland class (temporary, seasonal, and semipermanent)(Stewart and Kantrud 1971, Cowardin et al. 1988) because each class responds differently to surface and ground-water hydrology (Winter and Carr 1980). We drew sample wetlands at random from populations within a sample of 10.4 km2 plots (Fellows and Buhl 1995) that are used to provide data for a mallard simulation model (Cowardin et al. 1988). In total, we sampled 12 temporary, 12 seasonal, and 12 semipermanent wetlands from 10, 10.4 km2 plots (Figure 1). Wetlands in each 10.4 km2 plot were mapped by the National Wetlands Inventory according to the classification of Cowardin et al. (1979), and the upland areas were classified as grassland or cropland as described by Cowardin et al. (1988). We adopted the same ratio of cropland to grassland that was used to index wetland condition in the U.S. Environmental Protection Agency's Environmental Monitoring and Assessment Program (EMAP-Wetlands) for the PPR (Cowardin and Peterson 1995). Due to the availability of detailed water level data, we tested the accuracy of prototype water-level recording devices in 2 semipermanent wetlands (wetlands P7 and P8) located at the Cottonwood Lake Study Area (CLSA), a long-term study area near Jamestown, North Dakota (Swanson 1987).
|Figure 1. The Prairie Pothole Region of North and South Dakota showing the location of 10.4 km2 plots used in this study.|