Ecosystem Stability/dynamics

Contributors: Dave Clay, Sharon Clay, Carter Johnson, Dave Meyer, Don Olson, Maribeth Price, Craig Spencer, Larry Stetler, Dan Swets, Kerri Vierling, Kate Webb

Prairie potholes

The prairie pothole region of the North American continent includes the glaciated regions of North and South Dakota, Minnesota, Montana, and south-central Canada. The agroecosystems of the prairie pothole region are extremely important for their agricultural aspect in world food production and their natural resource aspect in migratory waterfowl production. The prairie pothole region provides habitat and cover for a wide array of both nonmigratory species and migrating birds including ducks and geese. In fact, this region and the wetlands of the Nebraska Sandhills produce up to 90% of the North American continent’s duck population annually (Batt et al., 1990).

In addition, the wetlands of this region are an important natural ecological resource for the area. These wetlands provide water catchmen areas to reduce flooding of states bordering the Mississippi and Missouri rivers and provide a natural buffer to reduce the impact of natural and manmade contaminants.

A problem in studying agroecosystems containing wetlands is that the boundaries between agricultural areas and wetlands change due to the wetlands dynamic nature during a single season and cyclically over years and decades. The classification scheme of wetlands, permanent, semi-permanent, and temporary, reflects the dynamic nature of thee habitats. A dramatic example of such changes is the history of Lake Albert in Hamlin County, SD. During the dust bowl era, the lakebed was dry and used for cropland. Today, the 100+ acre lake is a productive fishery for walleyes and northern pike and provides habitat for ducks, geese, and other animals. The changes in the prairie pothole landscape are primarily influenced by climate (which is being influenced by human activities) and secondarily by human land use in and around these areas.

Global climate change and its possible effects on natural and agricultural ecosystems are of national concern. Climate change has been documented with four of the warmest years on record occurring in the 1990’s and 9 out of the past 11 years being warmer than any previous years on record (Kerr, 1998). Model simulations based on projected increases in greenhouse gases and aerosols indicate that the rate of warming in the next century will be greater than any increases seen during the past 10,000 years (Graham, 1995; Houghton et al., 1996). Winter temperatures in the high northern latitudes are expected to increase along with warmer land than sea temperatures (Houghton et al., 1996). Increased precipitation is expected globally; however, regional forecasts vary with some areas wetter and some drier than their current norm (Cushman and Spring, 1989; Grotch and MacCracken, 1991).

Climate change models on inland prairie wetland ecosystems have not been studied sufficiently to predict the possible changes that may occur in these areas. The northern Great Plains is classified as subhumid or semiarid, and the wetland resources in this area are affected by climate variability. Wetland water levels and vegetation already fluctuate widely in these areas (Borchert, 1950; Kantrud et al., 1989; Larson, 1995). Global climate change is predicted to influence the distribution and functions of wetlands by altering their hydrologi regimes (changing water availability and the depth, duration, frequency, and season of flooding) outside their normal range (Poiani and Johnson, 1993).

Many forces threaten to alter the productivity and biodiversity of the prairie wetland landscapes. Studies of their current dynamics and their interactions with agricultural production while protecting and maintaining the many functions of the wetlands located in the prairie pothole region.

Ecosystem-scale experiments are difficult to conduct in the field. However, historical remote sensed information, modeling, and weather data should allow accurate estimation of how the prairie pothole region has developed and changed in the past. Ground-based sampling, modeling, and current imagery may give an indication of how the region may change in the future. In addition, remote sensing will provide a tool to link scales of spatial information.

The vulnerability of these agroecosystems depends both on the direction and magnitude of the predicted climate change and the agricultural use of upland habitats. For example, model simulations that used slightly increased temperatures and slightly lower rainfall amounts than today’s norm indicate that semi-permanent wetlands (areas wet in most years, with balanced emergent cover:water ratio) could shift to completely closed based basins with no open water areas and could dry up by midsummer in most years (Poiani et al., 1996). This would lower waterfowl production and wetland values. Land management exacerbate or ameliorate the impacts of global climate change.

Several specific issues, requiring careful and complete study, need to be addressed when beginning to understand the basic dynamics of the agroecosystems in the prairie pothole region. These issues include: the impact of global climate change on water cycling, C sequestering , and nutrient cycling at the individual pothole scale and regionally; the interaction among water, carbon, plant species, and nutrient on biomass production and biodiversity; and the importance of the potholes in the reducing downstream flooding and chemical contamination to subsurface aquifers. This information is needed to be able to predict and maintain the environmental health of the systems in the region. The following discussion will expand on each of these issues as it relates to the wetland portion of the agroecosystem. The impacts of the agronomic side may be addressed in the paper by Dalsted et al.

Impacts of land disturbance on erosion and water quality via paleolimnological and remote sensing analysis.

Ongoing work including analysis of lake sediment cores from three northwest Montana lakes reveals substantial correlation between mass sedimentation rate (MSR) of fine sediments and past human land disturbance activities. Accelerated logging was followed by increases in MSR of up to fourteen times the pre-European settlement rate (Spencer 1999, Schelske 1994). While natural land disturbance activities, including floods and wildfires, appeared to have some impact on MSR, human activities, lead by logging activity and associated road construction, appeared to be the greatest contributor to increased MSR. We intend to establish relationships between the NASA Ames Research Center and the research being done at Augustana College, specifically utilizing remote sensing to characterize land cover change resulting from timber harvest and other human disturbances on a watershed basis. Most of the timber harvest activity involves large clearcuts that are distinctly visible on Landsat (MSS) images. The NASA Ames Research Center, Earth Science, Ecosystem Science research activities are "directed to advanced understanding of the physical and chemical processes of biogeochemical cycling and ecosystem dynamics of terrestrial and aquatic ecosystems through the utilization of aerospace technology." Linkages that may be established could involve, for example, the use of high resolution sensors on aircraft to enhance the preliminary remote sensing work done with the Advanced Very High Resolution Radiometer (AVHRR) satellite imagery.

Lake sediment cores will be analyzed to document historical changes in sediment accumulation rate, primary productivity, and other limnological parameters. Attempts will be made to correlate changes in these parameters with historical changes in various land disturbance activities, both anthropocentric and natural. Our primary study area will be in NW Montana where timber harvest, fires, and floods represent major land disturbance activities. Preliminary results from core sample analyses in this region have produced promising results (Spencer and Schelske 1998, Schelske et al. 1994).

This project is a multidisciplinary effort involving biogeochemical analysis of sediment cores to document changes in lake conditions linked with time series analysis of remote sensing images to help document land-cover change in the lake basins. Logging in NW Montana has been characterized by large clearcuts (> 1km2), which are visible on Landsat images which date back to the early 1970's. In some cases however, alpine meadows, avalanche chutes, and burns may be confused with clearcuts. We will use time series analyses together with geometric (e.g., Hough Transform [Ballard and Brown 1982]) or feature space transformations (e.g., SHOSLIF [Swets and Weng, 1999]) to help distinguish between these features. Stratigraphic analysis of core samples with include a wide variety of parameters including nutrients, metals, pollen, and charcoal, as well stable isotopes (for estimation of primary productivity).

This work will involve scientists in the Biology, Computer Science, and Chemistry Departments at Augustana College. We are open to collaborative work with other SD EPSCoR labs for some of the analytical work. Remote sensing work will likely involve collaboration with the EROS Data Center and NASA Ames.


  1. The paleolimnological research may eventually be expanded to South Dakota; however, such efforts will face serious additional challenges. Although it would be interesting to conduct similar research on logging and mining activity in the Black Hills, the lack of any natural lakes seriously compromises the potential for paleolimnological research in western South Dakota. In eastern South Dakota, some collaboration might be possible with the proposed prairie pothole/agricultural impacts work; however, other obstacles exist in this area. The smaller size of the lakes and their catchments may preclude the use of older low-resolution Landsat images to quantify land cover change. Periodic dredging, drying, and water diversions/rerouting projects, common in the pothole region, may produce artifacts in the stratigraphy of the lake sediment cores. Finally, agricultural development in many parts of eastern South Dakota began in the mid- to late 1800's, which complicates sediment core analyses. The best sediment core dating techniques (via radioactive 210Pb) can produce reliable dates back to 1870 at best. Thus it may be difficult to quantify baseline conditions (such as erosion rates) prior to European settlement.
  2. The sustained productivity of the prairie pothole wetlands on occasional drying and reflooding. The length of time a wetland spends in each stage and the flood or drought return time have not been closely examined. Study of climate variability in space and time is required to understand this driving force which maintains the incredible biodiversity found in the northern Great Plains. Remote sensed information linked to climate and weather data can be used to investigate these relationships.
  3. The vulnerability of wetland resources in the northern Great Plains depends on the agricultural uses of the upland habitats. Croplands and pastures dominate the land use around most of the prairie wetlands; these influence runoff, sedimentation rates, infiltration by pesticides, and nesting habitat for waterfowl. Future farming and conservation practices on agricultural lands will have a large impact on the functional and taxonomic characteristics of the wetland ecosystems. Remote sensed date seasonally and interannually with ground truthing is needed.
  4. The vulnerability of wetland resources also depends on the direction and magnitude of future climate changes. Global warming and its possible effects on wetlands are a major concern to policy makers, resource managers, and the general public. Hydrology is the single most important set of variables determining the development and persistence of wetlands in the landscape. Climate drives the entire hydrologic regime. Even small changes will cause significant changes in the characteristics of the prairie wetlands.
  5. Invasive species have devastated many of the world’s ecosystems. Several species threaten to expand rapidly into northern prairie wetlands, having become established in wetlands and lakes in the eastern and western United States. These include purple loosestrife and Eurasian water milfoil. Study of the ways to minimize the negative impacts of these introduced species are needed immediately to insure continued high biodiversity among the prairie wetlands. Remote sensing would be very helpful in defining the rate of invasion and changes that occur due to these species.
  6. Wetlands impact on carbon sequestering.
  7. Wetlands impact on reducing downstream flooding.
  8. Wetlands impact on reducing chemical contamination.

Preliminary List of Techniques:

Simulation modeling using climatic data from today and predicted.

Determine energy flux into and out of agroecosystems in the Prairie Pothole region.

Lake core sampling – to map history of land use, sedimentation rates, infiltration rates, contaminant degradation rates

Productivity measurements

Remote sensing

Mapping and transect studies to determine baseline biodiversity and changes

Natural abundance isotope work to determine C sequestering, N cycling


Return to "Info on Oct. 7, 1999 SD NASA EPSCoR Meeting" website