Using Native Vegetation in Ecosystem Restoration Projects

Ecosystem and streambank restoration projects use a wide variety and large quantities of native vegetation to economically restore wetlands and streambanks, while providing adequate armoring to withstand the velocities and shear stresses from river flows.  Vegetative materials include: wetlands or uplands seed mixes; bare rooted seedlings or saplings; balled and burlap trees; dormant live cuttings (brush, stakes, poles) and potted plants.  This photo shows a dormant live willow stake that has sprouted in Spring after being driven into a bare bank during the dormant.  This stake is about ½” diameter, and projects about 12” from the bank.  The stake is buried 3’ to 4’ into the bank.  Native cuttings can also be used to provide brush layers, fascines, dormant live stakes and poles and vegetation soil reinforced systems. 

The use of vegetation can reduce the extent of riprap armoring, while still providing the necessary resistance to shear stresses and velocities on the upper banks of a stream.   The upper range of permissible shear stresses of soil bioengineering measures has been found to range from 1 to 5 lbs/square foot, with a corresponding range of permissible velocities from 3 to 10 feet/second.  Velocities and shear stresses need to be determined through hydrologic and hydraulic modeling of the river reach to determine worst case velocities and shear stresses.  See link to publication below: http://el.erdc.usace.army.mil/elpubs/pdf/sr29.pdf

On many of the Cuyahoga Valley National Park projects, we utilized native plant materials found within the park, rather than having the contractor provide cuttings from off site.  These included: sandbar willow; black willow; red ozier dogwood and some others.  The Park’s landscape architect would identify areas within the park from which the contractor could harvest plant materials.  This is both economical and environmentally sustainable, since the harvested plant material will eventually grow back, and the cost to cut and deliver the plant material from within the park is significantly lower than purchased material.  One should be aware not to specify plant material that is under distress from an invasive species, as is the case with Green Ash.  At one time cuttings from Green Ash were specified for use in the upper zone of restored riverbanks until it became known that ash trees were being harmed by the Emerald Ash Borer.  Plant material cannot be relied upon to provide armoring below mean water level.  In those areas, a riprap revetment should be used, and should extend up the bank to a level corresponding with the channel forming discharge or ordinary high water.

Plant material, when properly engineered, is an environmentally sustainable means to effectively stabilize riverbanks.

The Importance of Engineering in Ecosystem Restoration

Here’s a streambank restoration project on a Cuyahoga River tributary that, from the photo, looks to have been missing some important engineering.  The designer placed a woven drainage filter fabric on a steep, excavated bank, and installed rounded cobbles against the fabric to restore the eroded streambank. The streambank restoration failed during a subsequent flooding event, destroying the repair and exposing bare banks that will contribute detrimental sediment to the aquatic ecosystem.  This failure, like many, can be attributed to at least three factors: the rock; the filter drainage fabric; and the toe protection.  

Rock used to protect a streambank must be engineered because the stability of the rock must conform to the laws of physics.  Those laws, when applied to streambanks, account for: the median size of the stone (D ₅₀), specific gravity of the stone, specific gravity of water, a metric to represent the uniformity of the gradation of the stone, the relationship between the thickness of the layer and the largest (D₁₀₀ ) particle size, channel bend radius, channel bottom width, water depth, bank slope, stream velocity, and a stone stability factor.  The stability factor varies depending on whether the stone is angular, typical of quarried rock, or rounded, typical of glacial material found in a gravel pit.  Studies have shown that rounded stone is 20% less stable than angular stone of the same size and specific gravity.  The angularity naturally interlocks, tying the entire rock layer together.      

Drainage fabric is available in two types: woven and non-woven.  The woven fabric is thin, smooth, and slippery.  The non-woven is thick, soft, and spongy.  In this application, the woven fabric could have contributed to the failure due to its slippery properties.  Better yet, the fabric could have been eliminated entirely and replaced with either an engineered bedding layer of smaller size stone, or by using a thicker layer of armoring stone that is self-filtering. 

The photo does not indicate whether some type of stone toe protection was provided.  All streambank restoration projects should have either a launchable stone toe or a stone toe that’s excavated into the channel bed at the toe of the slope to prevent failure of the bank when scour occurs in the channel, and it will scour.   Although scour at the toe of a slope is most likely to occur at the outside of a bend, in this photo the stream is flowing into the picture, and the failed streambank protection is located in a straight section of the channel just upstream of a bend.  Even straight sections of channel will experience toe scour and should have a stone toe.

On riverbank restoration projects, therefore, it is essential to do the engineering studies to properly design the restoration to provide long term stability to the streambank so that the aquatic, cultural and other resources can be preserved.