The largest wetland drainage project in the history of Michigan was initiated in 1912 when ditching began on lands that are now part of the Seney National Wildlife Refuge in the Upper Peninsula. Ditches totaling about 30 km in length were dug across an undeveloped wetland approximately 20,000 ha in area in an attempt to convert the land to agricultural use. The agriculture venture failed, the drainage system was abandoned, and the land eventually was included in the refuge, much of it now designated wilderness. The main ditch in the area intercepted the flow of Walsh Creek and Marsh Creek, incised the ground-water table, and directed the flow of water to the Manistique River.

Lowering of the water table in the wetland allowed oxidation and subsidence of wetland peat, fostered an extensive peat fire, and altered vegetation in the hydrologically affected portion of the wetland. Habitat for wildlife has thus been affected. In addition, erosion of the banks and bed of the ditch during spring runoff continue to send large volumes of sand to the Manistique River, increasing its bed load of sand and adversely affecting fish habitat.

U.S. Fish and Wildlife Service managers at Seney National Wildlife Refuge proposed to restore habitat in an approximately 3,500 ha section of the Marsh Creek wetlands and reduce deposition of sand into the Manistique River by minimizing excessive flow through the ditch during spring runoff, promoting beaver activity to block the ditch, and diverting the water to the original Marsh Creek channel, the Walsh Creek channel, and other outflows, including Sweeney Ditch. These proposed actions have many potential ramifications. Improper distribution of diverted water may harm wetlands or downstream property; improper treatment of the abandoned ditch may cause failure to restore natural ground-water flows and wetland vegetation; diverted sheet flow of water may flood areas where peat subsidence and burning have lowered the land surface; wetland vegetation will likely change in some areas as a result of new hydrologic alterations; water chemistry in wetlands may change as a result of changes in supply; and beaver activity may affect both hydrology and vegetation.

Because so little is known about the ramifications of the proposed actions, Region 3 of the U.S. Fish and Wildlife Service requested that USGS Great Lakes Science Center and USGS Water Resources Division (WRD), Michigan District, perform studies in association with the restoration project to provide guidance to the refuge manager and personnel performing the actual restoration. The current ecological conditions in the unaltered portion of Marsh Creek north of C-3 Pool, the degraded portion south of C-3 Pool, and Walsh Ditch south of C-3 Pool were examined to provide baseline data for restoration activities. These studies evaluated the current ecological conditions and the role of the ditch in determining environmental conditions. These studies also evaluated potential methods for restoration and used pre-restoration data to make predictions regarding the effects of restoration actions, thus providing scientific information to guide management activities. Future studies will also examine ecological and environmental changes that occurred during and following restoration.

The research began in 1998 with an evaluation of environmental conditions in the wetland, including surface- and ground-water flow patterns and the ecological impacts caused by a 5-km earthen dike built across Marsh Creek in the early 1940s.

WRD scientists took in-stream measurements and strategically placed water-table wells, piezometers, and staff gauges to calculate the water budget for the pool behind the dike. Plant communities were studied at varying distances from the main ditch and from natural streams that flow through the wetland. Potential changes in land elevation caused by drying and burning of the peat were investigated by surveying areas near the ditch. Aerial photography, global positioning systems, and geographic information systems provided a geospatial framework for data analyses.

Ecological and hydrologic information is being used to assist the refuge manager in developing a plan for future distribution of water across the wetland that more closely approximates pre-drainage conditions. New FWS funding will support the replacement of a water-control structure to allow much more flexibility in distributing the water and returning it to natural creek channels. Following implementation of the plan, studies will continue to provide an evaluation of the ecological changes resulting from management actions and to assist in refinement of management plans. Knowledge gained at this site will hopefully provide important information for future peatland restoration projects at other locations.

Michael Sweat, USGS, Water Resources Division-Lansing, Hydrologist
Tracy Casselman, Seney National Wildlife Refuge, Refuge Manager
Mike Tansy, Seney National Wildlife Refuge, Refuge Biologist

Email Kurt Kowalski , Research Ecologist, Ecosystem Health and Restoration