Swan Creek Bioengineering Project
Sumpter Township, MI

Grantee: Sumpter Township
Basin Program Funds: $3,970
Non-federal Funds: $1,130
Project Duration: 09/1994 - 08/1997
Status: complete

Problem Statement
Removal of much or all of the streambank vegetation along creeks, drains, and ditches has considerable ecological consequences, including increased biological oxygen demand, loss of habitat, and increased stream temperatures and nutrient loads. By providing local governments responsible for drain maintenance with proven, cost-effective techniques of replanting and stabilizing streambanks with local native vegetation, the negative effects of drain maintenance will be minimized.

Background
The goal of the multi-partner Swan Creek Bioengineering Project is to research and report on the effectiveness of three bioengineering techniques to stabilize the banks of newly cleaned or maintained stream channels in the Swan Creek watershed, a Lake Erie tributary. In an effort to prevent flooding, and to promote the drainage of agricultural fields, many communities in the Great Lakes basin clean out, deepen, or widen streams and drains. In Sumpter Township, more than 35 miles of drains are scheduled for cleaning. This maintenance procedure often involves the removal of much or all of the riparian and streambank vegetation on one side of the channel to allow access for machinery. Also, streambanks are often left with slopes that are too steep to reestablish successful vegetation. These factors increase both sloughing and the amount of sediment entering the stream from channel erosion and overland runoff. The conditions that initially necessitated the maintenance are thus exacerbated. The Great Lakes Basin Program is providing $3,970 over a two-year period to support the implementation and evaluation of three bioengineering techniques.

Activities
1) Live stakes were installed in early April 1995 along 35 yards of the Number 6 Drain to stabilize an 840-square-foot area (0.02 acres) of streambank following recent channel excavation. Approximately 75% of the 150 live stakes planted were willow species, and about 25% were dogwood species. Nearly 60% of the stakes sprouted by the second week of May. In proportion to the species of stakes planted, the majority of stakes that sprouted were willows. Willows exhibited the greatest amount of growth, with shoots reaching lengths of 3 ½ feet. Among dogwoods that sprouted, shoots grew less than 1 foot.

2) In mid April 1996, live fascine bundles were installed along 25 yards of Swan Creek to stabilize a 750-square-foot area (0.02 acres) following channel excavation. This site showed evidence of slope instability due to the removal of a large tree and root wad from the channel bank. Approximately 400 branches for fascine bundles were cut in lengths of 6-10 feet and 50 three-foot stakes were prepared. Plant materials consisted of approximately 80% willow and 20% dogwood species. Two rows of fascine bundles were placed approximately 2 ½ feet apart in shallow benches excavated on the western slope of the creek. Vegetative growth of the fascines amounted to approximately 50% of the fascine bundles one month after installation; 10% of the fascines showed vegetative growth one year after installation -- the majority of the vegetative growth was from the dogwood species. No live stakes used in the live fascine installation were viable one year after installation.

3) In early April, a vegetative live crib wall was installed along 35 yards of Swan Creek to stabilize a 1,365-square-foot site (0.03 acres). The site was chosen because of severe erosion at a sharp bend on the east side of the drain that was washing away the outside corner and creating a steep slope. Over 50 dead logs and 75 live stakes were used to construct the crib wall. Approximately 500 branches were used for the vegetative layers. The vegetative layers within the crib wall were backfilled with over 40 cubic yards of sand. After construction, the project site was seeded with

Approximately 75% of the vegetative layers sprouted the first month after installation; three months after installation approximately 75% of the vegetative layers were still viable. Approximately 60% of the live stakes used to secure the crib wall sprouted and remained viable three months after installation. Vegetative coverage on the project site, including willow, dogwood, and grass, was approximately 85% by July 1997 with the crib wall intact.

Results
Sediment and attached nutrient reduction values were calculated for all three installations using the Channel Erosion Equation, nutrient concentration values, and soil texture correction factors. Total sediment reduction is 28.95 tons per year, total phosphorus reduction is 24.60 lbs per year, and total nitrogen reduction is 49.22 lbs per year.

Contact: Stephen Kunselman, (313) 461-6201

print-ready factsheet