Robbin B. Sotir
Principal, Robbin B. Sotir & Associates
The role woody vegetation plays in restoring and protecting streambanks is widely recognized. In addition, it has a variety of environmental benefits, including providing overhanging cover, temperature modification, supplying organic debris, and providing streamside habitat for nesting areas and protected wildlife corridors. While this role has become well understood, concerns are often expressed regarding the use of woody vegetation in streambank protection relating to its resistance to erosion in channels with high velocities. This is especially true of streams in urban areas as well as those that have been relocated and/or shortened. Additionally, questions arise regarding how it is affected by ice and debris, and what guidelines are available for determining the lowest bank elevation for establishing vegetation.
During the last fifteen years, several soil bioengineering streambank protection systems employing a soil bioengineering woody vegetative approach have been con-structed in the U.S. These systems have experienced flood events, some which have exceeded the100 year recurrence interval. Data were collected for several streams with soil bioengineering streambank protection installations that have experienced significant flood events. Data on flow velocities and impacts on vegetation are summarized.
Preliminary guidelines are presented for setting minimum stream stages above which woody vegetation can be expected to survive. Plant survival is related to both frequency and length of inundation. Minimum stage for establishment on streams that receive prolonged runoff from snowmelt and on other ephemeral arid climate and permanent humid climate streams.
Streambanks that are well vegetated by woody plants with deep root systems are resistant to scour by high velocities and to ice and debris gouging. Erosion of such banks usually results from bed scour and undercutting of the bank toe. Banks with a dense growth of woody vegetation also provide maximum environmental benefits that include shade and overhanging cover, organic debris for the aquatic ecosystem, and valuable streamside habitat for a variety of animals. Removal of woody vegetation may lead not only to streambank erosion but also to loss of habitat.
Although woody vegetation is a valuable component of streambank protection systems, many professionals are reluctant to incorporate vegetation into their designs as a structural element. There are several reasons for this, ranging from an unfamiliarity with such designs to the actual execution of installing vegetative systems. This paper will focus on concerns about its resistance to scour during floods, its impact on flood conveyance, and to ice and debris gouging. These concerns are addressed in this paper using data collected from nine (9) soil bioengineering streambank protection projects that have experienced major flooding. The projects are in several locations across the United States including three (3) where streams freeze over during the winter season.
While woody vegetation increases flow resistance, the effect varies depending on overall channel size and the width-depth ratio. Additionally, maintenance practices on the bank vegetation will also affect flow resistance. The larger and wider the channel, the less the effect of vegetation on flow resistance. Shields and Nunnally (1984) have summarized some of the available information on this subject. Impacts on flood conveyance on small and medium size streams can be determined through modeling by assigning appropriate roughness values to vegetated banks and calculating a weighted roughness coefficient. Table 1 was developed to model the effects of various vegetation conditions and maintenance practices on Lincoln Creek in Milwaukee, WI.
Table 1. Suggested n Values - Lincoln Creek Under Summer Conditions
Condition |
Suggested n Value |
|---|---|
Condition 1 Dense growth of trees and brush. Natural succession without maintenance. |
0.12 |
Condition 2 Dense tree canopy. Light to moderate undergrowth due to shading or selective maintenance. |
>0.09 |
Condition 3 Open tree canopy with some undergrowth of woody plants, irregular maintenance. |
0.075 |
Condition 4 Open tree canopy with little undergrowth of woody plants. Maintained by annual or biennial removal of undergrowth. |
0.06 |
Condition 5 Scattered trees, light growth of shrubs. Non-woody plants controlled by irregular maintenance. |
0.05 |
Condition 6 Scattered trees, no shrubs or woody vegetation. Non-woody plants controlled by mowing. |
0.045 |
Condition 7 No trees and no woody plants over six feet tall. Dense cover of grass/weeds up to four feet. |
0.035 |
Condition 8 Scattered trees, mowed grass, park setting |
0.030 |
| Subchannel | |
Condition 9 No obstructions to flow. |
0.022 - 0.025 |
Condition 10 Scattered vegetated bars, some debris. |
0.028 - 0.030 |
Condition 11 Numerous obstructions to flow. |
0.035 - 0.040 |
Source: Robbin B. Sotir & Associates
In undisturbed, forested watersheds the resistance of vegetated banks to erosion by scour is readily evident along stream systems. Even high gradient streams tend to be relatively stable, with most erosion resulting from toe erosion. To relieve some of the concern regarding the susceptibility of woody vegetation to scour during flood events, data were collected from nine (9) soil bioengineering streambank protection projects that have experienced significant flooding. The selected site includes three (3) rural stream and river sites and six (6) in urban watersheds. These were installed between 1983 and 1997. Several sites were flooded during and immediately after installation. The results are summarized in Table 2, which names the streams, the year of installation, the highest flows that have occurred since the projects were completed, an estimate of the return interval of the flood events, the mean velocity (calculated by dividing discharge by cross sectional area), and the project condition immediately following the flood. Most of the projects in the table were installed above rock toe protection.
Table 2. Soil Bioengineering Data
Name of Stream |
Stream Location |
Construction Completed |
Maximum Discharge Since Completed (cfs) |
Flood Event Return Interval (1 yr) |
Average Velocity (f/s) |
Condition of Vegetation After Event |
|---|---|---|---|---|---|---|
Raccoon Creek |
Dallas, GA |
1983 |
2,000* |
2 |
6-7 |
no damage |
Buffalo Bayou |
Houston, TX |
1988 |
7,500 |
50** |
4-5 |
no damage from scour |
Crow Creek |
Henry, IL |
1989 |
5,000 |
10 |
4-5 |
minor scour at nose of veg dike |
Buffalo Bayou |
Houston, TX |
1993 |
4,550 |
10 |
4-5 |
no damage |
Green River |
Kent, WA |
1992 |
7,280 |
5 |
4-5 |
no damage |
Menomonee River |
Milwaukee, WI |
1993 |
10,000 |
150 |
9 |
toe scour, with damage to veg |
Johnson Creek |
Milwaukie, OR |
1994 |
1,730 |
5 |
5-6 |
no damage |
Kenai River |
Soldotna, AK |
1994 |
42,000 |
100** |
minimal damage from scour |
|
Little Sugar Creek |
Charlotte, NC |
1997 |
11,100 |
100+ |
4-5 |
no damage |
Source: Robbin B. Sotir & Associates
* Bankfull discharge calculated using slope area method.
** Largest flood of record.
Only two of the nine (9) soil bioengineering installations has received damage from flood flows. The flood on the Kenai River, estimated to be a 100-year return interval event and the largest flood on record, caused minor damage to the vegetation in the live cribwalls. These were constructed one foot lower than the specified design. Water completely covered the installation to a depth of several feet during the flood, which occurred in the second year following installation. Local observers described the damage from erosion as severe in public access areas and other reaches lacking a good vegetative cover, but as minimal along most vegetated banks and the soil bioengineering installation at Soldotna Creek Park. At least three (3) of the soil bioengineering sites, Johnson Creek, Crow Creek and Raccoon Creek were flooded immediately following installation and before the vegetation had leafed out. None suffered any damage, although there was widespread erosion in other reaches, especially along Crow Creek. The Buffalo Bayou sites received several high water events, both during and after construction without sustaining damage.
Little Sugar Creek's event occurred four (4) months after installation and exceeded the 100-year event. It, too, was under several feet of water but sustained no damage.
The Menomonee River in Milwaukee received several smaller events with average velocities of 9 -10 fps, sustaining some toe scour. In June of 1997, however, the project experienced a storm event recurrence interval, estimated to be150 years based on existing land use. At this time, significant toe scour and damage was realized by the vegetation in the downstream section of the project. Several things may have contributed to this: the project was changed significantly during construction; the bank was steepened and the rock toe protection reduced; and, it is located in an outside meander immediately upstream of an abrupt right-angle curve. These factors, as well as the project being a short, isolated reach, may have contributed more to the damage than the storm event and associated velocities.
The Menomonee River, Kenai River and Crow Creek freeze over during the winter, and some trees along the banks show evidence of ice scraping. Ice erosion is not a significant problem along any of these streams, however, and the soil bioengineering installations have shown no evidence of damage due to ice erosion.
The height above the streambed at which flood tolerant species can survive varies with a stream?s flow regime. On ephemeral streams, which flow only during and for a short time following storm events, woody vegetation can be planted on the banks all the way down to the bed. Flood flows are of short duration, and flood tolerance usually is not a concern. Ephemeral streams are most often associated with dry climates, but many headwater channels, gullies, and drainage ditches in humid climates are ephemeral streams.
The hydrographs of small and medium sized permanent streams in humid climates are dominated by low baseflows upon which are superimposed stormflow hydrographs. Depending on the size and character of the drainage basin and the type and duration of the storm, storm flows may last from a few hours to several days. Species with high flood tolerances are able to root and grow above the Average Low Water (ALW) elevation. ALW can be defined as the median daily flow, or the daily mean flow exceeded 50 percent of time. This can usually be readily identified in the field as the bank elevation below which permanent woody vegetation is generally absent. Bank vegetation below this line is usually dominated by herbaceous annual plants.
On large streams with flood hydrographs that last several weeks to several months and on streams that receive prolonged runoff from snowmelt, woody vegetation is restricted to elevations above Ordinary High Water (OHW). OHW is defined as the average elevation of extended high flow over a period of years. The stream gauge will be useful in determining the cfs as well as the stage in feet. However, on ungauged streams, the line of permanent woody vegetation is assumed to be OHW elevation. As an example, the vegetation in the soil bioengineering bank protection system at Soldotna Creek Park was designed one foot above OHW because of uncertainty in determining the elevation of OHW at this site. While numerical information is important experience, good judgement and a healthy respect for streams and rivers plays an important role in the design and implementation of these systems.
The information presented in this paper for the placement of woody vegetation in streambank protection is tentative. The information is based on field evidence observed on natural vegetated streambanks and on information derived from monitoring the performance of soil bioengineering streambank protection projects that have been installed for varying lengths of time. Although the soil bioengineering systems have performed well when subjected to flooding, the sample size is very small and the geographical domain is somewhat limited. Additional monitoring of vegetative streambank protection systems and onsite measurements of flow velocities during flood events are needed to develop improved guidelines. Information regarding performance is being gathered which also considers the location of the installation on a particular stream reach, as well as the length of installation and the upstream and downstream connections.
Gray, D.H. and R.B. Sotir (1996), Biotechnical and Soil Bioengineering Slope Stabilization - A Practical Guide for Erosion Control; New York: John Wiley & Sons.
Hahn, Mike (1997) personal communication, South Eastern Wisconsin Regional Planning Commission (SEWRPC) from monitoring station at N. 70th Street, Milwaukee, WI.
Nunnally, N.R. and R.B. Sotir, (1997), Criteria for Selection and Placement of Woody Vegetation in Streambank Protection, Oxford, MS, May 1997, pp. 816-821
Pace, S. and J. Aug (1997) personal communication, Mecklenburg County, Charlotte, NC.
Robbin B. Sotir & Associates, unpublished files (1983-1987).
Shields, F. D., Jr. and Nunnally, N. R. (1984) "Environmental Aspects of Clearing and Snagging Projects,"
Journal of Environmental Engineering, ASCE, Vol. 110, No. 1,152-165.
Sotir, R.B. (1996), Soil Bioengineering for Streambank Protection and Riverine Restoration; 1996 ASAE Annual International Meeting, Phoenix, AZ, July 1996, Paper No. ASAE 962043.
Sotir, R.B. (1997), Designing Soil Bioengineering Streambank Protection for Multiple Objectives. Proceedings, Conference on Management of Landscape Disturbed by Channel Incision Stabilization Rehabilitation Restoration, Oxford, MS, May 1997, pp. 325-350.
Tacoma Street Interchange Soil Bioengineering Project - Monitoring and Evaluation of Stability and Habitat - Fall 1996. Seasonal Monitoring Report Prepared for Oregon Department of Transportation. Prepared by: Beak Consultants and Robbin B. Sotir & Associates.
USDA/NRCS, Chapter 16: 1997 Streambank and Shoreline Protection - Engineering Field Handbook.