Comprehensive Riparian Restoration Along Burd Run
by Christopher J. Woltemade, Ph. D. & Alan Wood, P.E.
The meandering planform of Burd Run was reconstructed to replace an artificially straightened reach. Herbaceous plants were seeded under the biotextile and native shrubs and trees will be planted throughout the riparian zone in spring 2002.
A 2000 site assessment indicated water quality degradation due to accelerated bank erosion and nonpoint sources of nutrients. Artificial channel straightening and removal of riparian trees and shrubs had destabilized the stream, resulting in 1200 linear feet of accelerated bank failure. In addition, the straight stream channel lacked diversity important to both aquatic habitat and site aesthetics. The draining of floodplain wetlands and removal of riparian vegetation had greatly reduced their ability to mitigate nutrient loads.
Water quality studies conducted by students and faculty at Shippensburg University identified elevated concentrations of nitrogen and phosphorus in the stream as a result of watershed land use and geology. Agricultural land dominates the lower half of the watershed, which is typified by thin soils overlying highly permeable carbonate bedrock. This combination is highly conducive to delivering nutrients to the stream via both surface water and groundwater.
Restoration Project Goals
It was important to develop a project vision that would garner support from diverse partners and increase the likelihood of winning a grant to fund construction. A coalition of over 20 project partners was built, led by Shippensburg University, the Cumberland County Conservation District, and Shippensburg Township. Other partners include federal agencies (Natural Resources Conservation Service, U.S. Army Corps of Engineers), interstate organizations (Susquehanna River Basin Commission, Alliance for the Chesapeake Bay), state agencies (Game Commission, Bureau of Forestry, Fish and Boat Commission), environmental organizations (Audubon Society, Trout Unlimited, Conodoguinet Creek Watershed Association), local municipalities, and school districts. The support of this coalition was essential to our successful application to the Pennsylvania Growing Greener grant program, which awarded $129,000 for final design, construction, and monitoring.
Most project partners were interested in improving environmental quality. However, a key selling point to the township that owns and manages the property was avoiding a maintenance problem with a sewer line that lies parallel to the stream. There was an obvious need to address bank erosion that had the potential to undermine the sewer line that was in places only a few feet from the channel. Establishing a stable, meandering channel in its historic locationmuch further from the sewer lineoffered a win-win option.
The preliminary design of a 1280-foot meandering stream channel to replace the 965-foot straightened reach was developed as part of a geo-environmental studies masters thesis at Shippensburg University (Herrmann 1999). This study utilized multiple design methods, including analog and analytical approaches. Empirical approaches based on regional relationships (for example, predicting channel bankfull channel area from drainage area) were determined to be inappropriate. Existing regionalized or universal relationships do not apply to this site due to local conditions. Burd Run, like many streams in karst limestone terrain, conveys surface discharges that are diminished by substantial groundwater flow through enlarged fractures and conduits. In addition, urban development immediately upstream of the project site has resulted in a channel geomorphology in disequilibrium, a condition in which empirical approaches to channel design are inappropriate (Skidmore et al. 2001).
Stream channel design analogs were developed from existing natural meanders (reference reaches) immediately downstream of the straightened reach as well as historical channel locations estimated from a 1-foot contour interval topographic map prepared from a site survey (the carbon copy approach, see FISRWG 1998). These analogs were used to develop preliminary channel planform and cross-sectional dimensions. Given that the downstream analog site does not represent geomorphic equilibrium conditions, it was inappropriate to directly base channel design on those channel dimensions.
Therefore, the preliminary channel design was refined via an analytical approach. The HEC-RAS computer program was used to compare the existing and design channels in terms of their bankfull flow capacity and the shear stresses generated under a range of discharge conditions. Through an iterative process, channel design adjustments were made to insure that shear stresses in the design channel would be less that those in the existing straightened reach.
Excessive bank erosion had occurred in the existing channel due to high flow velocities in the straightened reach. The channel meanders were designed to increase channel length and reduce slope, such that channel stability could be achieved. We decided on a nested channel design that allows small, frequent storm discharges to inundate a small bench of wetland plants and floods to inundate floodplain wetland areas. The HEC-RAS modeling confirmed that the final plan meets two design criteria: (1) under a range of flood discharges the flow velocities and shear stresses in the design channel are less than those under existing conditions, (2) floods will more frequently exceed bankfull capacity on the lower end of the park and inundate floodplain wetlands.
Bank protection is based largely on bioengineering, with eleven root wads along meander cut banks, two log vanes placed in a critical transition reach, and over 1,000 live stakes to be planted along much of the riparian corridor. Hard engineering measures were used where channel deformability was not desired. Two vortex rock weirs were placed near the upper end of the project to maintain channel alignment below a road bridge. Rip-rap was used sparingly along meander cut banks and will be vegetated with willows in spring 2002. To provide immediate erosion control, BonTerra CF7 biotextile was used on the toe of channel banks and a straw blanket bio-mat was installed up-slope at the transition from upper channel banks to floodplain. The biotextile sites were pre-seeded with facultative wetland seed mix and the bio-mat sites were pre-seeded with a park mix.
An 80-foot wide riparian buffer (40 feet on each side of the channel) of native tree and shrub species will be planted in spring 2002 to provide bank protection as well as wildlife habitat, shade, and aquatic nutrient uptake. This width was selected based on a review of successful similar projects and site availability.
The wetland restoration was designed to provide greater retention of water discharging from an on-site spring largely to achieve reductions in the concentration of nitrate. This site had been artificially ditched and drained, limiting the natural functions of the wetland. Weekly sampling over the past 18 months indicates a mean nitrate-nitrogen concentration of 7.4 mg/L at the wetland spring, compared to a mean of 3.8 mg/L in the adjacent stream channel. Previous studies (Woltemade 2000) indicate that increasing the time that water is retained in such wetlands can facilitate nitrate-nitrogen reductions of up to 68 percent. The final design includes a low earthen berm that blocks the drainage ditch and allows water levels to be maintained up to two feet higher than before. An Agri-Drain flow control structure allows water levels to be controlled in 6-inch increments. The berm was also seeded with wetland mix and covered with biotextile.
The two-phase approach was developed to allow work to be completed during optimal hydrologic conditions and to allow vegetation to establish before flowing water was introduced. This minimizes the risk of bank failure during and immediately following construction. The vegetation included both grasses seeded on the banks and facultative plants transported from the old stream bank. We planned to excavate the new channel in the fall, when stream flows and adjacent ground water levels are typically low. We were especially fortunate in that 2001 was the driest year in 68 years of official weather records at Shippensburg. During the October construction, the stream ceased to flow at the project site due primarily to lack of precipitation, an unusual but not unprecedented occurrence.
In addition to the drought, groundwater storage was rapidly diminished as water flowed through alluvial gravels and cobbles into the newly excavated channel. During excavation, these coarse sediments, a legacy of earlier locations of the stream, were exposed throughout the site. This bed material was stockpiled and later used to create riffles, substantially reducing materials costs.
Even through the unusually dry conditions, numerous springs and seeps maintained wet soil conditions in wetland areas at the site. The excavating contractor, John F. Walter Excavating, Inc., initially planned to make the first channel cut with an 8-foot wide pan. The soil conditions proved too wet for this equipment and a track hoe with a 33 bucket was used instead. With an experienced operator, this machine proved to have an excellent combination of agility and speed to complete the channel excavation. Very little finish grading was needed.
Bank protection measures were also installed largely with the track hoe. We attempted to drive root wads directly into the bank with no excavation, but the ubiquitous layer of alluvial cobbles could not be penetrated. Instead, root wads, log vanes, and vortex rock weirs were installed using standard approaches of excavation and placement.
A site initially identified as the source of material needed to construct the wetland berm proved to contain sandy alluvial sediments too coarse for the berm. Nevertheless, approximately 10,000 square feet was excavated to lower the surface by two feet to allow wet soil conditions to expand from the present wetland area into this site. This wetland expansion mitigates a small wetland loss caused by stream channel excavation. We decided to use fine-textured soil excavated from the new stream channel to construct the wetland berm. In addition to its low permeability, this material also contained a large supply of wetland seeds that should germinate in 2002. A thin layer of this wetland soil was also spread across the original borrow site after excavation in order to better support wetland plants.
A number of local organizations (watershed association, student organizations, boy scouts, etc.) have been directly involved in project planning and construction. A Shippensburg University student group assisted with transplanting wetland plants from the existing stream to the new channel. Similarly, a wide range of organizations will be used to plant live stakes, shrubs, and trees along the stream in the spring. A local fishing club is working with the Pennsylvania Fish and Boat Commission to continually monitor and improve fish habitat in the new stream channel.
Much of the environmental education program is still under design. The interpretive nature trail is being developed in conjunction with community user groups and will be established in summer 2002. A training workshop for local teachers planning to use the site for field studies is planned for spring 2002 and will include elementary, middle school, and high school teachers. Shippensburg University already uses the site to support field studies in geography, earth science, and biology.
Stream geomorphology will be monitored by repeat surveys of twenty-two established cross sections. This will provide data on changes in channel width, depth, area, bankfull flow capacity, and planform. The results will be related to the bank stabilization measures associated with each cross section.
Site hydraulics will be monitored to evaluate the magnitude of discharges associated with stream channel changes. In addition, retention time of water in the wetland will be monitored, as it is often an important influence on the success of wetlands at reducing aquatic concentrations of nutrients.
Water quality studies will focus on the transport and fate of nutrients though the site. Several separate flow paths will be monitored: shallow groundwater flowing through the riparian zone, spring water discharged into the restored wetland, and stream flows entering and leaving the project site. The effect of the restoration strategies on nutrient fluxes through each of these pathways will be assessed.
Many floodplain restoration projects are fundamentally motivated by improving ecology. This project is expected to improve habitats for fish, amphibians, birds, and other animals. Ecological monitoring will focus on the aquatic invertebrates that serve as the base of the food chain as well as important indicators of general ecological health.
Our experience also illustrates the importance of subsurface site exploration during design. If we had more complete information on the extent of sand, gravel, and cobble deposits within the floodplain, we could have made fewer adjustments during construction. While the unexpected soil types provided some benefits (the cobbles used to construct riffles), it also caused some problems (abandoning the original borrow site for berm construction material). L&W
For more information, contact Dr. Christopher J. Woltemade, Dept. of Geography-Earth Science, Shippensburg University, 1871 Old Main Drive, Shippensburg, PA 17257 (717) 477-1143, fax (717) 477-4029, e-mail: firstname.lastname@example.org, website: www.ship.edu/~cjwolt or Alan Wood, P.E., State Project Engineer, USDA-Natural Resources Conservation Service, Suite 340, One Credit Union Place, Harrisburg, PA 17110-2993, (717)-237-2211, Fax. 717-237-2239, e-mail: email@example.com
Herrmann, J., 1999. A stream meander restoration study on a small ungauged stream in south central Pennsylvania. M.S. Thesis, Shippensburg University, Shippensburg, PA, 82pp.
Skidmore, P. B., F. D. Shields, M. W. Doyle, and D. E. Miller. 2001. A categorization of approaches to natural channel design. Proceedings of the 2001 ASCE Wetlands Engineering and River Restoration Conference, Reno, Nevada. 12pp.
Woltemade, C.J. 2000. Ability of restored wetlands to reduce nitrogen and phosphorus concentrations in agricultural drainage water. Journal of Soil and Water Conservation, 55(3): 303-309.
©2002, 2001, 2000, 1999, 1998 Land and Water, Inc.