Hydrology and chemistry of selected prairie wetlands in the Cottonwood Lake area, Stutsman County, North Dakota, 1979-82

Abstract

The relation of hydrologic setting and temporal variability in hydrology to nutrient content and geochemical characteristics of a group of prairie wetlands and adjacent ground water was studied during the period 1979-82. Although data were collected from many wetlands and wells at the study site, emphasis in this report primarily is on four wetlands two seasonal and two semipermanent and four wells contiguous to them along a hydrologic section. The seasonal wetlands, T8 and T3, contained water only for a few weeks to months after filling in spring and early summer; both were completely dry by August. The semipermanent wetlands, PI and P8, contained water throughout each year and were ice covered in winter. One wetland, T8, recharges ground water. Wetlands PI and P8 are in areas of ground-water discharge. None of the wetlands received water by channelized surface-water inlets. Only wetland P8 had a channelized surface-water outlet. Ground-water-level data showed that high points of the water table did not always occur beneath land-surface highs. Reversals of ground-water flow occurred occasionally between two of the wetlands, T3 and PI. Significant differences existed in the chemical composition of the wetlands based on their hydrologic setting. In general, the dominant cation and anion in the wetlands were potassium and bicarbonate in wetland T8, calcium and sulfate in wetland T3, magnesium and sulfate in wetland PI, and magnesium and bicarbonate in wetland P8. Significant seasonal differences existed in the water chemistry of the wetlands in ground-water discharge areas. Water in three of the wetlands, T3, Pi, and P8, was most dilute while they filled in spring after icemelt. Concentration increased during the open-water period, and two of the wetlands, PI and P8, became most concentrated under ice cover. Concentrations of total phosphorus and total nitrogen were greatest in wetlands in areas of ground-water recharge and least in wetlands in areas of ground-water discharge. Differences in the chemistry of water from wells in the adjacent ground water resulted primarily from the positions of the wells in the ground-water flow system. The chemical type of water from well 12, which was located in a ground-water recharge area, was calcium sodium bicarbonate. Water from well 4, located downgradient from wetland T8, and from well 16, located downgradient from wetland PI, typically was a calcium sulfate type. Water from well 13, located between wetlands T3 and PI in an area of changing ground-water flow directions, was a magnesium sulfate type. Data from this study show that an understanding of hydrologic conditions is important in the interpretation of the water chemistry of wetlands in the study area. 1U.S. Fish and Wildlife Service. INTRODUCTION Numerous studies published in the proceedings of symposia on the subject of wetlands have indicated that the relation between hydrologic processes and the function and structure of wetland ecosystems is poorly understood (Good and others, 1978; Greeson and others, 1979). Probably the principal reason for this lack of understanding is that comprehensive, multidiscipline studies of wetlands rarely have been done; hydrologic studies seldom include comprehensive analyses of chemical and ecological processes, and, conversely, few ecological studies of wetland ecosystems include comprehensive investigations of hydrologic processes. Some ecological studies of wetlands have presented evidence that the structure of biological communities is affected by hydrologic processes. Gooselink and Turner (1978) indicated that plant communities change because the frequency of inundation of wetland soils affects the availability of oxygen in the root zone. Studies of wetland seed banks were used by van der Valk (1981) to show how changes in water level alter plantcommunity structure. The ability of seeds or propagules to become established in areas of either standing water or no standing water was attributed to waterlevel changes. Hydrologic setting also can affect structure of plant communities in wetlands, through its relation to wetland water chemistry; for example, Stewart and Kantrud (1972) determined that the composition of plant communities in prairie wetlands is correlated with specific conductance of water in those wetlands. Wetland water chemistry is related to hydrologic processes because those processes are a major factor in controlling the movement of chemical constituents to and from wetlands. Hemond (1980) noted that information on the movement of water to and from wetlands is essential for biogeochemical studies. The wetland-biogeochemical studies most dependent on investigation or measurement of hydrologic procHYDROLOGY AND CHEMISTRY, SELECTED PRAIRIE WETLANDS, NORTH DAKOTA esses are those of chemical mass balance. Quantification of water inputs and outputs is an integral part of determining chemical inputs and outputs in massbalance studies; for example, Valiela and Teal (1978) indicated that the quantification of water and nutrient inputs and outputs is important in understanding nu trient dynamics of wetlands. Yet, in only a few studies have nearly all relevant hydrologic processes been measured as a basis for understanding biogeochemical processes in wetlands (Crisp, 1966; Valiela and others, 1978; Mitsch and others, 1979; Hemond, 1980; Verry and Timmons, 1982). However, even in these studies, at least one hydrologic process that was a component of the water budget was measured some distance away from the study area or was calculated as the difference between measured inputs and outputs. Ground water commonly is calculated as the difference between measured inputs and outputs in chemical mass-balance studies of lakes (Winter, 1981); this approach also commonly is used in the study of wetlands. A major problem with using the approach of calculating ground water as a residual in wetland studies is that hydrologic instrumentation usually is not optimal in either accuracy or placement relative to the wetland; the difference between measured inputs and measured outputs can have little hydrologic meaning because of errors in hydrologic measurement (Winter, 1981). Conceptual models of ground-water flow systems near wetlands have been developed primarily by Canadian hydrologists; for example, Meyboom (1966, 1967) studied the interaction of ground water with several prairie wetlands and lakes in Saskatchewan. Many of Meyboom's study sites were of wetlands that received ground-water inflow because water-table highs underlaid land-surface highs. In some cases, the water table beneath hills would decline to the point where the wetland would have seepage from it to ground water for part of the year. Meyboom also conducted studies of open-water evaporation (1967) and of transpiration from phreatophytes near wetlands (1966,1967). Lissey (1971) also studied ground-water flow near prairie wetlands in Canada; he proposed that most ground-water recharge and discharge in that environment takes place in land-surface depressions that commonly are occupied by wetlands. A few studies of wetlands have attempted to relate wetland water chemistry to the hydrologic concepts of Meyboom (1966,1967) or Lissey (1971). These few studies have been confined to prairie potholes in western Canada (Rozkowska and Rozkowski, 1969; Rozkowski, 1967, 1969; Sunde and Barica, 1975) and have indicated that a relation exists between the water chemistry of lakes and wetlands and nearby ground water. Hydrologic data were not available in those studies. Consequently, water chemistry was interpreted by using a conceptual model of hydrologic processes instead of in conjunction with directly measured hydrologic processes. In most cases, wetlands have been studied individually rather than as a group within a hydrologic unit, such as a ground-water flow system. Many ecologists are unaware of ground-water flow systems and associated theory on the way ground water may effect the wetlands under investigation. This is manifested in studies of wetland biogeochemistry, where chemical mass-balance calculations include estimates of ground water by difference. Before chemical mass balances, including direct measurement of ground water, can be determined, wetlands need to be shown to be an integral part of ground-water flow systems. Such a system then can be used to begin to examine the relation between the water chemistry of wetlands and adjacent ground water. Wetlands in the prairies of North America are ideal systems for such an investigation because many have no channelized surface-water inlet or outlet. Therefore, the hydrologic processes most likely to effect such ecosystems are those related to atmospheric exchange (rainfall-evaporation-evapotranspiration) and ground-