Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures (2016)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

14 2.1 NCHRP Report 544 In 2005, the results of NCHRP Project 24-19 were published as NCHRP Report 544: Envi- ronmentally Sensitive Channel- and Bank-Protection Measures (McCullah and Gray 2005). After conducting an extensive literature review and evaluation of commonly used environmentally sensitive techniques, McCullah and Gray identified 44 environmentally sensitive channel- and bank-protection techniques for study. The channel- and bank-protection techniques were grouped into four major categories: (1) River Training Techniques, (2) Bank Armor and Pro- tection, (3) Riparian Buffer and River Corridor Treatments, and (4) Slope Stabilization. Tech- nique descriptions and guidelines for their applications were developed. In many respects the work by McCullah and Gray (2005) can be viewed as a starting point and foundation for the present study. 2.2 Synthesis of Current Practice 2.2.1 Introduction Since publication of McCullah and Gray (2005), numerous reviews, handbooks, and measure- specific guidance documents for environmentally sensitive bank protection have been pub- lished by federal, state and local agencies. These documents emphasize the steps describing various measures and provide construction or installation guidance. Similarly, large numbers of case studies of biotechnical bank-protection projects have appeared, but few contain hydrau- lic data. Hydraulic design criteria are scarce and with few exceptions rely on the literature that was summarized within NCHRP Project 24-19. The data underlying the hydraulic criteria are drawn from a variety of sources and vary in quality from qualitative anecdotal rules of thumb to detailed laboratory measurements. Many are based on isolated spot measurements of velocity. Considerable progress has been made in recent years in understanding the biology of riparian plants and developing technology for improving planting success. Limited but impor- tant advances have been made in understanding and simulating the complex fluid mechanics of open-channel flows adjacent to vegetated banks. Slope stability models have been modified to include contributions of roots to soil strength, and such models are becoming more widely employed. 2.2.2 Environmentally Sensitive Channel- and Bank-Protection Measures State departments of transportation (DOTs) are seeking ways to incorporate environmental criteria into projects that impact streams in order to comply with legislation and cooperate C H A P T E R 2 Findings

Findings 15 with conservation agencies and interests. Environmentally sensitive measures for control- ling erosion of channel banks, beds, and floodplains have been described by many authors and usually feature use of living and nonliving plant materials in combination with stone, geotextiles, and soil. However, guidance for design, construction and maintenance of such measures is often qualitative, forcing agencies to rely on individual experience or accept high levels of risk and liability. Further, documented experience with many types of measures is in short supply. In response to this situation, NCHRP Project 24-19 was initiated. The effort was targeted at compiling existing information on measures for controlling channel erosion that simultane- ously provided benefits to terrestrial habitat, aquatic habitat, or aesthetics relative to tradi- tional bed and bank erosion control measures. The final report from this effort (McCullah and Gray 2005) contains detailed descriptions including typical design drawings along with as much design criteria as could be gleaned from the literature extant as of 2001–2002. Data assembled for each technique included allowable hydraulic loadings, dominant modes of failure, and research needs (see Appendix A of this report for a summary). In addition, a rule-based technique selection system was also developed for NCHRP Report 544. The selection system is presented as an interactive software program titled “Greenbank,” which can be found on the CD-ROM that accompanied NCHRP Report 544 (CRP-CD-58). The user is queried regarding environmental objectives, dominant erosion processes at the site in question, hydraulic conditions, and other site characteristics. A short list of suggested techniques or measures is then provided, along with links to additional information. Related content useful to designers was published by McCullah (2006). The following sections syn- thesize results of a literature review focused on developments subsequent to the NCHRP Project 24-19 effort. 2.2.3 Literature Prior to NCHRP Project 24-19 The review presented by McCullah and Gray (2005) includes an annotated list of key docu- ments and websites, and the CD that accompanied NCHRP Report 544 contained a compilation of .pdf versions of many of the documents. Accordingly, this section presents only a very brief overview of the pre-2002 literature dealing with environmentally sensitive bed- and bank- protection measures. Earliest work in stream bank erosion and protection did not differentiate between standard and environmentally sensitive measures, as practitioners then were largely agnostic about eco- logical impacts. However, basic principles underlying river engineering in general and stream bank erosion and its control were considered relevant. For example, the extensive national research and demonstration program in stream bank protection conducted by the USACE in the late 1970s and early 1980s (“Section 32 Program”) produced a wealth of information that is often overlooked because little of it appeared in the open literature. Results (summarized in Table 2.1) highlighted the need for designers to consider both geotechnical and hydraulic processes and to think about erosional processes across a wide range of spatial scales. Noteworthy items within the early literature are the texts by Schiechtl (1980) and Schiechtl and Stern (1994) that provide rather comprehensive overviews of European bioengineering practice, which has a very long history (Evette et al. 2009). Bache and Coppin (1989) present similar material from a U.K. perspective, while Gray and Leiser (1982) and Gray and Sotir (1996) write for a U.S. audience. Henderson and Shields (1984) and Henderson (1986) also provided an early review of environmental features for stream bank erosion control in use within the U.S. Although case studies and project reports may be found in the early literature, scientific research to support development of design criteria for environmental channel erosion control measures

16 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures is very scarce prior to 1990. Most of the content of the earlier works is composed of photos and drawings of measures under construction or sometime after completion, descriptions of con- struction techniques, and occasionally cost data. Laboratory data are available for permissible hydraulic stress on certain types of flexible channel linings, but less is provided for measures that include woody plants. Design guidelines in these works are mostly qualitative and descriptive rather than quantitative, and reflect the experience-based design approaches based on judgment, although the details vary greatly by Section 32 refers to Section 32 of Public Law 93-251, the "Streambank Erosion Control Evaluation and Demonstration Act of 1974," which authorized $50 million for a national program of demonstration and research. The program, which ended in September 1982, included: Laboratory demonstrations, Construction of demonstration projects at 125 bank-miles of eroding bank line at 68 legislated and selected sites, Observation of 50 existing projects, and Extensive literature surveys. The final report, U.S. Army Corps of Engineers, 1981. Final Report to Congress, The Streambank Erosion Control Evaluation and Demonstration Act of 1974, Section 32, Public Law 93-251. Main Report, Supplemented by Appendices A-H in separate volumes. U.S. Army Corps of Engineers, Washington, D.C., contains a summary of the program findings, including the items below. Causes The causes of stream bank erosion are complex and varied. Erosion along major rivers may involve different processes than for small streams, and dominant processes frequently vary from one site to another along the same stream. In most cases, erosion at a site is the result of several causes, and each cause may require a specific cure. Prediction of future bank erosion is beyond the current state of the art. Although the basic processes are understood, the complex interactions between many variables in natural systems make prediction of erosion indeterminate. Streams displaying very active tendencies to erode their banks often seem to reverse themselves and display periods of relative stability. Selection and design of effective bank protection often depends on a good understanding of the geotechnical characteristics of the site and the fluvial geomorphology of the reach and the watershed. Cures Where bank failure is due to bed degradation, grade control structures may be needed to restore stability. There is no universal method that offers low-cost solutions for all stream bank erosion problems. Low-cost methods are best suited for short-term protection on small streams. There are no cheap solutions, but good understanding of the important processes can yield cost-effective projects. Much erosion can be prevented with toe protection, especially with vegetation on the upper bank. In fact, the final report stated that "…the most important conclusion is to provide effective protection at the toe of the bank." Bank shaping and vegetation planting without providing toe protection is usually ineffective. Bank-protection materials (rock, stone, armor blocks, etc.) in river reaches subjected to high velocities, waves, or high levels of turbulence must be placed on appropriate granular or geotextile filters to prevent the loss of bank material to penetrating currents. In low-energy environments, however, a blanket of quarry-run riprap of sufficient size and thickness performs well. Riprap blanket is generally the most cost effective, flexible, and widely used bank-protection technique. Other structures made of stone riprap such as spurs, jetties, groins, stone toe, and windrow revetment provide adequate protection when properly designed, but some initial erosion should be anticipated before the structures become effective. The alignment of bank-protection structures is critical. During periods of high flow, the location of the major point of attack by the current will usually vary from low or moderate flow conditions. The most common mistake in designing bank protection for an eroding bank is to extend bank protection too far upstream and not far enough downstream. Table 2.1. Lessons learned from the Section 32 program.

Findings 17 technique. When hydraulic loading criteria are provided they are often based on the author’s experience rather than specific test data. Literature on all types of environmentally sensitive stream treatments has grown rapidly since about 1990. In the late 1990s, Fripp prepared the tabulation of allowable velocities and shear stresses for a variety of channel boundaries, including environmentally sensitive treatments shown in Table 2.2 (personal communication, Jon Fripp, USDA NRCS). The source documents for Table 2.2 vary from peer-reviewed journal papers describing measurements performed under controlled conditions in hydraulic laboratories to rough measurements made under field conditions, to claims made by vendors of various products, to nonspecific reports based on experience. This compilation provided the basis for a modified table published by Fischenich (2001b) (Table 2.3) as part of a series of technical notes issued by the USACE ERDC. Refinements of these tabulations were used to generate rules for specific measures in the Greenbank selection system (McCullah and Gray 2005). Since definite numerical limits were required for the selection of system logic, allowable shear and velocity values for Greenbank could not be expressed in approximate terms, or as a range of values. Where the source docu- ments provided a range of values, the midpoint of the range was adopted for use in Greenbank. Furthermore, Greenbank contained criteria for all 44 measures described by McCullah and Gray (2005), a much longer list than those presented in Tables 2.2 and 2.3. Techniques based on stone structures such as vanes, bendway weirs, longitudinal stone toe, and Newbury rock riffles were assigned maximum permissible velocities of 11.5 ft/s based on upper limits for well-designed stone structures. Measures intended to address geotechnical slope stability such as drop inlets, chimney drains, and live pole drains were not assigned maximum hydraulic loading values since allowable velocity and shear in such an application depends on the treatment applied to the bank face. In some cases, Greenbank used findings published subsequent to Fischenich (2001b) such as Lipscomb et al. (2001) for vegetated, articulated concrete blocks or the Erosion Control Technology Center (ECTC) (2001) for turf reinforcement mats. Since the completion of work on NCHRP Project 24-19 and publication of McCullah and Gray (2005), publications dealing with stream erosion control, riparian zone and floodplain res- toration, and interactions among plants, slope stability, and stream hydraulics have multiplied prolifically. For purposes of this synthesis, these documents may be categorized as either applied or fundamental literature. 2.2.4 Applied Literature Subsequent to NCHRP Project 24-19 Applied literature includes how-to guides, handbooks, case studies, and studies and docu- ments focused on the performance of a specific measure (e.g., willow stakes or rootwads). A subset of these documents consists of guidelines and research reports on techniques for han- dling plant materials (e.g., soaking or refrigeration) to improve survival and performance. Most of the applied literature has been published by governmental agencies and is not published in peer-reviewed scientific or engineering journals. Much of the content in these documents is duplicative or redundant. Handbooks and Reviews This section places all documents containing catalogs of environmentally sensitive measures in the same category, even though some (e.g., Admiraal et al. 2007, Landphair and Li 2001) are literature reviews while others provide design guidance (e.g., McCullah 2006, Allen and Leech 1997). Handbooks have been produced by the city of Denver (Denver Urban Drainage and Flood Control District 2001a,b,c), King County, Washington (Johnson and Stypula 1993),

18 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures Bank Material/Protection Shear Stress (lb/ft2) Velocity (ft/s) Type of Criteria Source Bermuda grass, erosion resistant soils, 0%–5% slope 8 design USDA 1947 (rev. 1954) Bermuda grass, erosion resistant soils, 5%–10% slope 7 design Bermuda grass, erosion resistant soils, over 10% slope 6 design Bermuda grass, easily eroded soils, 0%–5% slope 6 design Bermuda grass, easily eroded soils, 5%–10% slope 5 design Bermuda grass, easily eroded soils, over 10% slope 4 design Grass mixture, erosion resistant soils, 0%–5% slopes 5 design Grass mixture, erosion resistant soils, 5%–10% slopes 4 design Grass mixture, easily eroded soils, 0-5% slopes 4 design Grass mixture, easily eroded soils, 5%–10% slopes 3 design Grasses: Lespedeza sericea, Weeping lovegrass, Yellow bluestem, Kudzu, Alfalfa, Crabgrass, Common lespedeza; erosion resistant soil, 0%–5% slope unless on side slopes 3.5 design Grasses: Lespedeza sericea, Weeping lovegrass, Yellow bluestem, Kudzu, Alfalfa, Crabgrass, Common lespedeza; easily erodible soil, 0%–5% slope unless on side slopes 2.5 design Dense sod, fair condition growing in moderately cohesive soil 0.35 limit Austin and Theisen 1994 12.5 cm of excellent growth of grass/woody veg on outside bend 1 limit Parsons 1963 Flume trials, fabric reinforced veg failed after 50 hrs 5 limit Theisen 1992 Flume trials, fabric reinforced veg failed after 8 hrs 8 limit Sod revetment, short period of attack 0.41 design Schoklitsch 1937 Wattles (coarse sand between) 0.2 design Wattles (gravel between) 0.31 design Wattles (parallel or oblique to current) 1 design Fascine revetment 1.4 design Cribs with stone 30 design Reed plantings (immediately after construction) 0.10 limit Schiechtl and Stern 1994 Reed plantings (after 3-4 seasons) 0.61 limit Reed roll (immediately after construction) 0.61 limit Reed roll (after 3-4 seasons) 1.22 limit Wattle fence (immediately after construction) 0.20 limit Wattle fence (after 3-4 seasons) 1.02 limit Live fascine (immediately after construction) 1.22 limit Live fascine (after 3-4 seasons) 1.63 limit Willow brush layer (immediately after construction) 0.41 limit Table 2.2. Permissible shear and velocity data compiled in the late 1990s by Fripp.

Findings 19 and the states of Alaska (Walter et al. 2005), Arizona (Arizona Department of Environmental Quality 2005), Georgia (Georgia Department of Natural Resources 2007 and 2011), Iowa (Iowa Department of Natural Resources 2006), Maryland (Maryland Department of the Environment 2000), Nebraska (Admiraal et al. 2007), New Jersey (New Jersey Department of Agriculture 2012), New York (Glath et al. 2003), Ohio (Kush 2007, Baker 2007), Oregon (Oregon DOT 2011), Texas (Landphair and Li 2001), Washington (Cramer et al. 2003), and Wisconsin (Wisconsin DOT 2013). In addition, a Canadian province (Donat 1995) and the federal government have produced handbooks for regional (Bentrup and Hoag 1998, Hoag et al. 2001, Hoag and Fripp 2002 and 2005, Yochum 2013) and national [Biedenharn et al. 1997, Fischenich and Allen 2000, Lewis 2000, Federal Interagency Stream Restoration Working Group (FISRWG) 2001, Eubanks and Meadows 2002, Wells 2002a and b] application. Additional guides have been published by nongovernmental organizations (The River Restoration Centre 2002, Schueler and Brown 2004) and the governments of Scotland (Scottish Environmental Protection Agency 2008) and Australia Bank Material/Protection Shear Stress (lb/ft2) Velocity (ft/s) Type of Criteria Source Willow Brush layer (after 3-4 seasons) 2.86 limit Schiechtl and Stern 1994 (cont) Willow mat (immediately after construction) 1.02 limit Willow mat (after 3-4 seasons) 6.12 limit Deciduous tree plantings (immediately after construction) 0.41 limit Deciduous tree planting (after 3-4 seasons) 2.45 limit Live stakes in riprap (immediately after construction) 2.04 limit Live stakes in riprap (after 3-4 seasons) 6.12 limit Coarse gravel and stone cover with live cuttings (immediately after construction) 1.02 limit Coarse gravel and stone cover with live cuttings (after 3-4 seasons) 5.10 limit Coir fiber roll, single stake, <1:3 slope 0.2 - .8 5 design Bitterroot Restoration Product Literature Coir fiber roll, double stake, with brush mat 0.8 - 3.0 8 design Turf reinforcement mat, permanent 8 20 design Rolanka Product Literature Straw reinforcement mat, temporary 0.45 8 design Jute mat 0.45 design Chen and Cotton 1988 Straw with net 1.45 design Curled wood net 1.55 design Synthetic mat 2 design Rootwads 8.7 observation Allen and Leech 1997 Rootwads 12 observation Willow posts 3.1 observation Herbaceous and woody 8 design Soil cement 25 limit Portland Cement Association Brush mattress w/willows 6.5 limit Gerstgraser 1999 Wattle fence 1 limit Fascine 2.1 9.8 limit Cuttings of willows/willow stakes 2.1 9.8 limit Articulated concrete mats, unvegetated, USACE block, 40% open 4.3 13.2 limit Lipscomb et al. 2001 Articulated concrete mats, vegetated, COE block, 40% open 6.1 13.8 limit Table 2.2. (Continued).

20 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures (Torre 2001), although some of these guides cover all types of stream restoration measures as well as bed and bank treatments. Several writers cover slope stabilization using bioengineering approaches with stream bank measures as a subset (e.g., Lewis 2000, Holanda and da Rocha 2011). Almost all reviews present some sort of taxonomy that they use to classify various tech- niques, and some (e.g., Li and Eddleman 2002, Hagen et al. 2002) emphasize the relative cost- effectiveness of selected measures, but hydraulic data are generally lacking. Two of the most important national handbooks are the ones produced by the Federal Inter- agency Stream Restoration Working Group (1998, revised 2001) and the NRCS (2007a, b, and c). The latter contains more engineering design guidance and includes a compilation of allow- able velocities and shear stresses (Table 2.4) based upon, but slightly different from the earlier ones by Fripp (Table 2.2) and by Fischenich (2001b) (Table 2.3). The text that accompanies Table 2.4 is worth noting: Boundary Category Boundary Type Shear Stress lb/ft2 Velocity ft/s Source Vegetation Class A turf 3.7 6-8 Gray and Sotir 1996, unpub data Fischenich Class B turf 2.1 4-7 Class C turf 1.0 3.5 Long native grasses 1.2-1.7 4-6 Kouwen, Li, and Simons 1980, Norman 1975, Temple 1980, unpub data Fischenich Short native and bunch grass 0.7-0.95 3-4 Reed plantings 0.1-0.6 N/A Gray and Sotir 1996, unpub data Fischenich Hardwood tree plantings 0.41-2.5 N/A Temporary degradable rolled erosion control products Jute net 0.45 1-2.5 Gray and Sotir 1996, Norman 1975, TXDOT 1999 Straw with net 1.5-1.65 1-3 Coconut fiber with net 2.25 3-4 Gray and Sotir 1996, Texas Department of Transportation (TXDOT) 1999 Fiberglass roving 2.00 2.5-7 Gray and Sotir 1996, Norman 1975, TXDOT 1999 Non-degradable rolled erosion control products Unvegetated 3.00 5-7 Gray and Sotir 1996, Kouwen, Li, and Simons 1980,TXDOT 1999 Partially established 4.0-6.0 7.5-15 Fully vegetated 8.00 8-21 Julien 1995, Temple 1980, TXDOT 1999 Soil bioengineering Wattles 0.2-1.0 3 Gerstgraser 1998, Schiechtl and Stern 1994, Schoklitsch 1937, unpub data Fischenich Reed fascine 0.6-1.25 5 Gray and Sotir 1996 Coir roll 3-5 8 Gray and Sotir 1996, TXDOT 1999, unpub data Fischenich Vegetated coir mat 4-8 9.5 Live brush mattress (initial) 0.4-4.1 4 Forineth 1982, Gray and Sotir 1996, Schiechtl and Stern 1996 Live brush mattress (grown) 3.90-8.2 12 Forineth 1982, Gerstgraser 1998, Gray and Sotir 1996, Schiechtl and Stern 1996, unpub data Fischenich Brush layering (initial/grown) 0.4-6.25 12 Gray and Sotir 1996, Schiechtl and Stern 1996, Fischenich 2001 Live fascine 1.25-3.10 6-8 Gerstgraser 1998, Gray and Sotir 1996, Schiechtl and Stern 1996, Schoklitsch 1937 Live willow stakes 2.10-3.10 3-10 Gray and Sotir 1996, Allen and Leech 1997, unpub data Fischenich Table 2.3. Permissible shear and velocity criteria presented by Fischenich (2001b).

Findings 21 Recommendations for limiting velocity and shear vary widely. . . . The designer should proceed cautiously and not rely too heavily on these values. Judgment and experience should be weighed with the use of this information. The recommendations in [Table 2.4] were empirically determined and, therefore, are most applicable to the conditions in which they were derived. The recommendations must be scrutinized and modified according to site-specific conditions such as duration of flow, soils, temperature, debris and ice load in the stream, plant species, as well as channel shape, slope and planform. Specific cautions are also noted in the table. However, there are anecdotal reports that mature and established practices can with- stand larger forces than those indicated in this table. (NRCS 2007b) Although not a handbook or review, work by Niezgoda and Johnson (2012) should be men- tioned here because it lays a foundation for an important new aspect of applied practice: risk analysis. These authors present a quantitative approach for weighing risks of failure for specific practices against economic, environmental, and social benefits. A weakness of the approach is the lack of objective criteria for quantifying failure risk and benefits. Guidelines for a Specific Measure Another class of documents consists of guidelines for designing or constructing a single type of environmentally sensitive bank-protection technique. Although guidance may be gleaned from case studies that feature a specific technique, single-measure guidelines are primarily direc- tive in nature and do not relate experience at a single site or group of sites. Several local, state, Practice Permissible Shear Stress (lb/ft2) Permissible Velocity (ft/s) Live poles (depends on the length of the poles and nature of the soil) Initial: 0.5 to 2 Established: 2 to 5+ Initial: 1 to 2.5 Established: 3 to 10 Live poles in woven coir (depends on the installation and anchoring of coir) Initial: 2 to 2.5 Established: 3 to 5+ Initial: 3 to 5 Established: 3 to 10 Live poles in riprap (joint planting) (depends on riprap stability) Initial: 3+ Established: 6 to 8+ Initial: 5 to 10+ Established: 12+ Live brush sills with rock (depends on riprap stability) Initial: 3+ Established: 6+ Initial: 5 to 10+ Established: 12+ Brush mattress (depends on soil conditions and anchoring) Initial: 0.4 to 4.2 Established: 2.8 to 8+ Initial: 3 to 4 Established: 10+ Live fascine (very dependent on anchoring) Initial: 1.2 to 3.1 Established: 1.4 to 3+ Initial: 5 to 8 Established: 8 to 10+ Brush layer/branch packing (depends on soil conditions) Initial: 0.2 to 1 Established: 2.9 to 6+ Initial: 2 to 4 Established: 10+ Live cribwall [depends on nature of the fill (rock or earth), compaction and anchoring] Initial: 2 to 4+ Established: 5 to 6+ Initial: 3 to 6 Established: 10 to 12 Vegetated reinforced soil slopes (VRSS) (depends on soil conditions and anchoring) Initial: 3 to 5 Established: 7+ Initial: 4 to 9 Established: 10+ Grass turf—bermuda grass excellent stand (depends on vegetation type and condition) Established: 3.2 Established: 3 to 8 Live brush wattle fence (depends on soil conditions and depth of stakes) Initial: 0.2 to 2 Established: 1.0 to 5+ Initial: 1 to 2.5 Established: 3 to 10 Vertical bundles (depends on bank conditions, anchoring, and vegetation) Initial: 1.2 to 3 Established: 1.4 to 3+ Initial: 5 to 8 Established: 6 to 10+ Sources: NRCS (1996), Hoag and Fripp (2002), Fischenich (2001b), Gerstgraser (1999), Nunnally and Sotir (1997), Gray and Sotir (1996), Schiechtl and Stern (1994), Allen and Leech (1997), Forienth (1982), and Schoklitsch (1937). Table 2.4. Permissible hydraulic loadings for bioengineering measures from NRCS (2007b).

22 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures and federal agencies have published series of these documents in print or on the web. McCullah and Gray (2005) note several examples of websites that disseminated such guidelines. Since con- clusion of their research project, additional guidelines have appeared. For example, the USACE’s Institute for Water Resources maintains a website with guidelines for 30 techniques, while a smaller number of techniques are available from the Ohio Department of Natural Resources (Table 2.5) and numerous other sources. Typically, such guidelines focus on construction details: how to prepare the site and materi- als and assemble or plant them to produce a finished project. Hydraulic design criteria familiar to designers who work with riprap revetment (e.g., Lagasse et al. 2006) or even similar to those produced for grassed channels (e.g., Killgore and Cotton 2005) are usually absent. Printed guid- ance documents have also been published. Examples of measure-specific guidelines, most either from the USACE ERDC or the state of Ohio Department of Natural Resources, are provided in Table 2.6. Case Studies The literature contains a large number of reports of the success or failure of bank-protection measures applied to a given site (a segment of bankline or a stream reach), and even more have been published recently (Goldsmith et al. 2014). Many case studies describe projects Institute for Water Resources, USACE Ohio Department of Natural Resources* 1. Bank Cover and Current Deflector with Sand Bag and Cellular Confinement System 07 Restoring Streambanks with Vegetation 2. Bank Crib with Cover Log 3. Bank Shaping and Vegetation 11 Tree Kickers 4. Bioengineering and Bioengineering Techniques 5. Branch packing and Brush layering 12 Evergreen Revetments 6. Cable Concrete 7. Coconut Fiber Roll, Coir Rolls, Coir Mats and Coir Netting 13 Forested Buffer Strips 8. Dormant Posts or Dormant Cuttings 9. Erosion Control Blanket 14 Live Fascines 10. Grout-filled Mattress 11. Gabion and Gabion Mattresses 17 Live Cribwalls 12. Grass Rolls 13. Hedge-Brush Layering 19 Deflectors 14. Joint Planting/Vegetative Riprap 15. Live Cribwalls 20 Eddy Rocks 16. Live Fascines and Wattlings 17. Live Siltation 22 Gravel Riffles 18. Live Staking 19. Log and Brush Shelter 20. Log Cribbing 21. Native Material Revetment (Log, Rootwad, and Boulder Revetment) 22. Overhanging Bank Cover 23. Placement of Boulders 24. Riprap 25. Rootball or Rootwad Placement 26. Straw Rolls 27. Stream Bank Debrushing, Brush Bundles, and Brush Mats 28. Training Fences 29. Vegetated Geogrids 30. Vegetation/Revegetation www.pmcl.com/mmdl/MM.asp?ID=1 www.dnr.state.oh.us/tabid/4178/default.aspx *The numbers preceding the Ohio Department of Natural Resources treatment types are the guide numbers assigned by the department. Table 2.5. List of selected bank treatment guidelines available from two web sources.

Findings 23 that employ several types of measures in adjacent bankline segments or combined along the same bank. To date, case studies have provided a higher level of reality than model studies or laboratory experiments, but are difficult to generalize for application to other sites due to site-specific conditions, short periods of observation, or insufficient data to fully characterize the hydraulic and geotechnical processes operating on the constructed site. Most common are studies that describe the appearance or geometry of an eroding site before and after treatment and provide details on design, construction, and costs, but little if any hydraulic data. For example, only 10 of the 35 case studies presented by Goldsmith et al. (2014) are for sites for which discharge data are available, and near field measurements of flow depths and velocities are much less com- mon. In some cases, practitioners infer shear stresses acting on installed biotechnical measures using 1-D or 2-D models. Key facts regarding recently published case studies are summarized in Table 2.7. Handling Plant Materials Most of the handbooks and several of the guidelines for individual measures contain sections about selecting, harvesting, and handling plant materials (e.g., NRCS 2007b). Guidelines that Measure Hydraulic Design Criteria Content/Remarks Reference Willow spilling yes Summary of performance of 140 projects. Criteria modified from Fischenich (2001b) and Sotir and Fischenich (2007). Anstead and Boar 2010 Brush mattress yes Guidance for fabrication and installation. Hydraulic criteria from Fischenich (2001b) supplemented with Schietchl and Stern (1997) and experience by Fischenich. Allen and Fishenich 2001, American Society for Testing and Materials (ASTM) 2003 Live fascines, "vertical bundles" yes Hydraulic criteria from Fischenich (2001b). Limits are "empirical information collected from constructed projects." Guidance for fabrication and installation. Sotir and Fishenich 2001, Ervin 2007, ASTM 2014 Live stakes and joint planting yes Example contract specifications. Guidance for preparation and installation. Hydraulic loading limits based on empirical information collected from constructed projects. Shafer and Lee 2003, Sotir and Fishenich 2007, ASTM 2013, Hoag 2009b Rootwads yes Design methodology for use with a computer spreadsheet and/or a family of empirical curves that can quantify the amount of ballast required to stabilize a rootwad for a variety of load conditions. Safety factor computations. Sylte and Fishenich 2000, Wood and Jarrett 2004 Tree revetments no Guidance for site selection, fabrication, and installation. Bishop et al. 2007 Live cribwalls no Guidance for site selection, fabrication and installation. Ervin and Fulmer 2007 Willow and cottonwood no Guidance for using cuttings or clumps in a variety of ways and for constructing stable banks by placing poles or bundles under riprap. Hoag 2007, Hoag and Sampson 2007 Coir logs yes Allowable velocity and shear from vendors or constructed projects. Allen and Fischenich 2000 VMSE yes Limits based on empirical information collected from constructed projects. Sotir and Fischenich 2003 Flexible channel linings Qualitative Guidance on use of manufacturer's reported allowable velocity and shear stress for rolled erosion control products, turf reinforcement mats, erosion control blankets, etc. Miller et al. (2012) Table 2.6. Typical guidance documents for individual bank-protection measures.

24 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures Measure(s) Locale Remarks Reference Willow spilling Two sites, East Anglia, U.K. One yr observation, successful. Anstead et al. 2012 Live poles, stakes, fascines Mill Creek, Cincinnati, OH Five yr observation, successful. Protecting landfill Barrett et al. 2006 Large wood structures, brush mattress, VMSE Redwood and Corte Madera Creeks, California Three and six yrs of observation, mostly successful Blomberg et al. 2006 Rootwads, brush layers, coir logs, transplanted vegetation mats 19 sites on flowing waters or lakes in Matanuska- Susitna Borough, Alaska, 10 of which involved bank restoration A few evaluations of plant survival, structure condition and bank stability within one to three yrs of construction. Several opportunities for improved practice noted Davis and Davis 2005 and 2007 Slope flattening, boulder grade control structures, VMSE, coir logs at toe, brush layering, turf reinforcement mat, fascines, live siltation Willow and Sand Creeks, Denver, CO Two yr observation, deemed successful Denver Urban Drainage 2001a and 2001b Various combinations of wood, planting cuttings, coir, rock 10 sites, Western WA Various periods of observation. All reported as successful Federal Emergency Management Agency (FEMA) 2011 Brush layers with stone toe recommended, but not constructed since this is a planning-level report Four reaches of Cazenovia Creek, NY Reach-mean velocity and shear stress computed using uniform flow equations and compared to Fishenich 2001b Frothingham 2008 Wattle fence Huaijiu River, Beijing, Salix alba recommended. Gu et al. 2012 Rootwads, brush layering Ketnai River, Alaska Root wads experienced ice damage, brush layers protected upper banks well Karle 2007 Rootwads, live staking, brush layers, coir logs 11 sites, Alaska 1-D model used to compute average bed and bank shear stresses under 50-year and 100- year events. Bioengineering measures not reliable in channels with high shear stresses Karle et al. 2003 Vegetated cribwall Two southern Ontario watersheds, 12 cribwalls in all. Age of structures not specified Cribwall and vegetation characteristics, sediment sampling, erosion pin monitoring and computer generated stream power analysis. Vague about performance Krymer and Robert 2013 VMSE—with geogrids and willow cuttings. Dormant willow posts— combined with stone toe and protection of surface between posts with reinforced turf Cottonwood Creek, Hutchins, TX Three yr observation. Successful stabilization with 90% survival of cuttings. Questionable approach used to monitor velocity at a single point. Li 2006 Reinforced turf VMSE Vegetated riprap Jute netting and large wood Four sites in Oregon Four yr observation. Quantitative measurements of stream stage, velocity, and discharge. No bank erosion observed. Vertically averaged velocity continuously logged using acoustic instruments at two points. Mabey 2009 China Table 2.7. Recently published case studies of environmentally sensitive bank-protection projects.

Findings 25 focus specifically on handling plant materials include Fischenich (2001a), Luna et al. (2006), Darris (2006), Bergdorf (2007), Hoag (2007) and Balch (2008). Additional guidance is available on irrigation (Fischenich 2000a), soil compaction (Goldsmith et al. 2001), and soil amendments (Fischer 2004). Other researchers have experimented with special equipment and techniques to make woody cutting planting on banks, typically a labor-intensive task, more efficient (Hoag et al. 2001, Hoag 2009a, Hoag and Ogle 2011). Furthermore, the literature contains results of applied research on techniques for handling and installing plant materials to ensure acceptable levels of survival and growth vigor. Many of these deal with propagation of woody species from cuttings. Greer et al. (2006a and 2006b) found that the diameter of Salix nigra cuttings interacted with soil moisture regime in a complex fashion; they recommended planting both large (10 cm) and small (1 cm) diameter cuttings. Although willow (Salix sp.) and cottonwood (Populus sp.) species are most often used due to ease of rooting from cuttings, Hunolt (2012) and Hunolt et al. (2013) compared performance of silky dogwood (Cornus amomum) and Virginia sweet- spire (Itea virginica) as live stakes to two willow species and found they were able to survive and establish if harvested during dormancy. Several practitioners have tested the effect of soaking woody cuttings before they are planted on survival and growth. Soaking cuttings before planting for periods as long as 14 days was beneficial if cuttings were dormant (Schaff et al. 2002, Martin et al. 2004, Tilley and Hoag 2008). Measure(s) Locale Remarks Reference Vegetated riprap South central and interior Alaska "…specific hydrologic and hydraulic characteristics need to exist for a riprap armored stream bank to allow and sustain vegetative growth." Maniaci and Nolen 2005 Vegetated cribwalls Nicaragua Three of the four native species evaluated are recommended for future use. Economics attractive. Petrone and Preti 2008 12 types of structures including vegetated crib walls, debris jams, log bank structures Guadalupe Creek, San Jose, CA Five yr observation. Crib walls and debris jams successful; log bank structures stabilized banks but did not maintain undercut habitats Seville and MacKay 2006 Boulders at toe with VMSE on upper bank Three reaches of Fort Branch Creek, Austin, TX Period of observation short and unspecified, successful. Byars and Renfro 2012 Vegetated gabions Manchester River at golf course, Manchester, NH Six months observation. After only one growing season, the vegetation is generally well established where it was planted. Some areas where the plants were too small or not properly placed during the installation have some problems with establishment in the structure. Brunet and Shuey 2005 Reinforced turf with anchors Anthony Creek and Cow Creek, WV No post-construction observations reported. Merritt et al. 2010 Large wood structures, plantings of willow poles and stiff grasses Topashaw Creek, MS Five yr observation. Measurements of flow, depth and velocity. Unsuccessful due to ongoing channel incision. Shields et al. 2008 Coir logs, erosion control blankets Nineveh Creek, Edinburgh, IN Very brief period of observation USACE 2006 VMSE: Lifts of soil are encapsulated in geotextile that is wrapped over the exposed face of each lift, creating a "stair-step" appearance. Vegetation is planted between lifts. This is a well-established bioengineering technique widely employed with variations in the types of vegetation and geotextile used. Other terms used for this technique include "vegetated reinforced soil slope," "vegetated soil lifts," and "vegetated geogrids." Table 2.7. (Continued).

26 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures Results for nondormant cuttings were less favorable (Pezeshki et al. 2005). Martin et al. (2004) showed cuttings responded positively to soaking water oxygen concentration, giving credence to anecdotal guidelines for soaking plant materials in moving waters. Hunolt (2012) and Hunolt et al. (2013) found soaking cuttings of four woody species for 48 hours before planting had mini- mal effects, but dormancy was important. Survival rates were 100% for all species when stakes were harvested and planted during the dormant season, but three of the four had 0% survival when harvested during the growing season. Tilley and Hoag (2008) found fall willow cutting plantings did better than spring plantings. Because some regions have short dormant seasons, Landphair and Li (2002) and Li et al. (2005) experimented with extending dormancy by stor- ing cuttings in refrigerated chambers and achieved field survival rates for Salix nigra cuttings in Texas of 44% to 81%. 2.2.5 Advances in Fundamental Science Subsequent to NCHRP Project 24-19 Plant Characteristics Considerable research has been done in recent years to characterize aspects of woody riparian species that are important in bioengineering applications. Species tolerances to various environ- mental stresses and habitat preferences have been better defined for some species (Evette et al. 2012, Liu et al. 2010, Pezeshki and Shields 2006). Work on Salix nigra (Schaff et al. 2003, Li et al. 2004, Pezeshki et al. 2007) and S. exigua (Caplan et al. 2012) has shown that soil texture and soil moisture are key characteristics, with best performance within rather narrow ranges of these two interrelated variables. A few researchers have attempted to define plant characteristics (e.g., root biomass per unit soil volume) that correlate with erosion rates (Gyssels et al. 2005). Work by Beasley (2011) and Beasley et al. (2010) on levee slopes subjected to overtopping is note- worthy in this regard. Slope stability models that account for contributions of plant roots to soil strength require information about root density, morphology, and tensile strength, which must be painstakingly measured (Adhikari et al. 2013). These characteristics vary by species and with environmental variables, but at least one report indicates that engineered slopes can produce superior plant characteristics (biomass and root tensile strength) to natural slopes (Ying et al. 2011). Fluid Mechanics In order to design and manage systems with bank-protection measures that feature vegeta- tion, engineers need robust tools to simulate channel flow conveyance, bank shear stresses, and habitats. Riparian vegetation interacts with channel flow in complex ways, such as modifying flow resistance locally and at the reach scale, governing near-bank turbulent structure and shear stresses, momentum exchange, and, in some cases, concentrating flows by displacing threads of higher velocities either away from the bank or underneath the canopy and closer to the bank (Kean and Smith 2004, McBride et al. 2007, Czarnomski et al. 2012). Fluid mechanics for flows in channels with vegetated banks may be categorized based on flexibility of vegetation (flexible or rigid) and relative flow depth (vegetation submerged or emergent) (Rahmeyer and Werth 1996, McKay and Fischenich 2011, Aberle and Järvellä 2013). Rigid plants that protrude above the free surface, for example, offer much greater flow resistance than flexible ones that are fully submerged. The density of the plant stems and branches is also quite important as the flow field responds to plant density in a highly nonlinear fashion (Kean and Smith 2004). For deciduous plants, the presence or absence of leaves is important (Freeman et al. 2000 and 2004, Wunder et al. 2011, Czarnomski et al. 2012, Aberle and Järvelä 2013). Research on interactions between rivers and terrestrial vegetation has been advancing rapidly over the past 10–15 years, and a range of approaches has been used. Both floodplain and stream

Findings 27 bank vegetation have been studied; the latter is of primary interest here. Some workers have used dowels or other manmade objects as artificial plants in laboratory flumes (e.g., Czarnomski 2010, Czarnomski et al. 2012), but artificial plants do not deform in flows as real ones do (Järvelä 2006, Wunder et al. 2011). Others have used small, real plants in laboratory flumes or wind tunnels (Freeman et al. 2000, Wunder et al. 2011). Laboratory flumes allow high-frequency, spatially-detailed measurements of the velocity field in prismatic channels under steady flow. Most of these flume experiments (and associated numerical modeling) have assumed a uniform stand of bank vegetation that does not vary its characteristics in the lateral direction (e.g., Kean and Smith 2004, Bledsoe et al. 2011) even though natural riparian vegetation and biotechnical bank-protection measures usually feature different types of protection for bank toe, mid bank, and upper bank. Interactions between riparian vegetation and stream channel morphology have been studied at larger scales using statistical approaches by fluvial geomorphologists (e.g., Bledsoe et al. 2011). Increasingly sophisticated numerical models (e.g., computational fluid dynamics) have been used to extend and complement field and laboratory results across a range of spatial scales (e.g., Wilson et al. 2006, Jahra et al. 2011). Some of the most interesting work in this area has been directed toward forested floodplains: stands of woody vegetation on flat surfaces. For example, Chen et al. (2009) conducted a series of flume studies examining the flow resistance and soil erosion rates for flat surfaces composed of bare soil, and for soil surfaces planted with uniform stands of four California native floodplain plant species. For vegetated surfaces, flow resistance declined as a linear function of Reynolds’ number, with linear slopes and intercepts varying according to plant characteristics. When mean flow velocities exceeded 3 ft/s, erosion rates under plant canopy were less than 30% of those observed for bare soil. The following explanation was offered: Observations demonstrate that these floodplain-adapted species lay over under higher flows creating a laminar break in the velocity profile where the soil/water interface is essentially insulated and protected from the scour forces of higher velocities. The bare soil vertical velocity gradient was quite uniform whereas the velocity gradients in the plant tests were S shaped, being slow at the soil surface and accelerat- ing over the top of the plant canopy as the plants bent over. In another example, Rahmeyer and Werth (1996) and Freeman et al. (2000 and 2004) reported results of over 220 flume experiments involving 27 different real plant types and groupings. In- channel (not stream bank) vegetative flow resistance was found to decrease with velocity and depth for submerged plants but to increase with depth for emergent plants. Flexible, submerged plants with leaves formed a streamlined (teardrop) shape that reduced the flow forces on the plants, protecting leaves and smaller stems from breakage. Minimum plant velocity limits of 3 to 4 ft/s were observed for leaf failure, and most of the leaf and stem failures were the result of impact with bed material and debris. Stands with more or less uniform canopy geometry con- centrated flow underneath the canopy, resulting in general scour. In a field (rather than flume) study, Manners et al. (2013) used terrestrial laser scan images of clumps of tamarisk shrubs to characterize stage dependence of hydraulic roughness. The result- ing models were extrapolated to reach-scale, two-dimensional hydraulic models built with aerial LiDAR data. The studies cited above on floodplain vegetation and similar works have made important contributions, but key differences exist between flows across vegetated floodplains and flows in channels with vegetated banks (e.g., turbulence peaks at the interface between the main channel and riparian zone and near the toe of sloping banks) (McBride et al. 2007, Czarnomski et al. 2012). Bank slope influences spatial distributions of Reynolds stresses (Czarnomski et al. 2012). Despite considerable effort devoted to studying flow forces on vegetation (Fischenich and Dudley 2000, Wunder et al. 2011) and associated effects of vegetation on flow resistance (e.g., Fischenich 2000b, McKay and Fischenich 2011, Aberle and Järvelä 2013), almost no work has been done

28 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures since that reported by McCullah and Gray (2005) to define the ability of biotechnical measures to withstand flow forces. Criteria for selecting and designing environmentally sensitive measures are usually qualitative or based on experience of a few individuals. Much of the hard data that does exist is derived from smaller-scale flume experiments in prismatic channels with highly uniform vegetation or artificial vegetation. Slope Stability Stream banks erode or retreat due to fluvial erosion processes, sliding or mass wasting, and combinations of these processes. Scientists have long realized the links between vegetation and bank slope stability (Gray and Leiser 1982, Gray and Sotir 1996), but since publication of McCullah and Gray (2005), several important advances have been made in development of analytical techniques. Both beneficial and detrimental effects of vegetation on bank stability have been observed: plant roots reinforce soils and remove moisture by evapotranspiration, increasing matric suction; however, very large trees impose weight loading on banks and woody vegetation may also enhance infiltration of precipitation (Simon and Collison 2002, Gray and Barker 2004, Pollen et al. 2004). Earlier models of root reinforcement of soils ignored the effects of soil type and moisture on root-soil bonds; root tensile strength was simply added to soil strength and roots were assumed to be oriented perpendicular to failure planes. Work by Pollen (2007) has advanced understanding of temporal and spatial variability in root reinforcement due to variations in soil type and moisture, and work by her team (Pollen et al. 2004, Pollen and Simon 2005) has shown that soil-root matrices are better simulated by fiber-bundle models that allow progressive failure of roots (weakest first) rather than the “all-roots-break-instantaneously” assumption of the older perpendicular models. The fiber-bundle model algorithm has been incorporated into the Bank Stability and Toe-Erosion Model (BSTEM), a spreadsheet tool used to simulate stream bank stability. The user supplies bank geometry, soil properties, vegetation cover, and bank water-table levels, and the model outputs factors of safety. If banks fail or erode, new bank geometry is generated. Fluvial erosion of material at the bank toe is simulated using a simple excess-shear stress approach (Simon et al. 2011). Fluvial Geomorphology At the river reach scale, geomorphologists have produced stronger explanations for the mutual adjustment of bank vegetation and channel width, depth, and slope including physically based models that account for vegetation effects on slope stability, sediment transport, and flow resis- tance (Millar and Eaton 2011, Bledsoe et al. 2011). Some effects are non-intuitive: Trimble (2004) presented data from a Wisconsin watershed that showed grass on banks promotes a smaller bankfull cross-sectional area than forested banks. Opposite effects were reported by Bledsoe et al. (2011), who found that channels narrower than 65.6 ft (20 m) have shear stress distributions quite sensitive to bank vegetation. Streams with banks covered with woody vegeta- tion should be narrower and deeper. Environmental Effects Research is emerging on ecological effectiveness of a range of ecosystem rehabilitation and restoration strategies and measures. Roni et al. (2008) present a global review of literature on the effects of stream habitat rehabilitation measures; 142 of the 345 papers analyzed report effects of instream habitat measures that include various bank treatments. However, few of the bank treat- ments can be classified as erosion controls: riparian silviculture, fencing, and instream habitat structures are the focus. Frequently the literature reports a lack of data on restoration success or projects that have no measurable ecological benefit (Doyle and Shields 2012). Within this literature there is a small subset relevant to environmentally sensitive bank-protection measures.

Findings 29 For example, Everaert et al. (2012) reported on analysis of 82 (macroinvertebrate) and 112 (macrophyte) “records” from Dutch projects involving “ecologically sound banks.” In general, the projects contributed to diverse macroinvertebrate and macrophyte communities, but site selection, design, and maintenance were crucial to success. Conversely, Cooperman et al. (2007) were unable to detect differences in habitat or macroinvertebrate abundance between stream bank restoration sites and untreated reference sites along salmon streams in southern British Columbia. A recent textbook notes that while bank protection can be a legitimate component of an ecological restoration project, bank stabilization constructed with the main objective of protecting infrastructure is unlikely to achieve habitat improvement since natural channel pro- cesses require some lateral channel migration (Roni and Beechie 2013). 2.3 Survey of Current Practice 2.3.1 Survey Form and Distribution A copy of the final survey instrument for Task 2 is included in Appendix B. In early November 2013, the survey questionnaire was distributed by email to 319 individuals representing all 50 state highway agencies (DOTs) and other federal, state, and local government agencies. Selected Native American nations, consultants, and academic institutions involved in stream bank protection measures were also included in the survey outreach effort. In addition to the DOTs, the distribution list included: • FHWA: – Resource Centers – Federal Lands Divisions • USDA: – NRCS Plant Materials Centers – Forest Service – ARS • U.S. Department of the Interior (USDI): – USGS – Bureau of Land Management – Bureau of Reclamation – Fish and Wildlife Service • USACE • National Oceanic and Atmospheric Administration (NOAA) – National Marine Fisheries Service (NMFS) • AASHTO Standing Committee on the Environment (SCOE) • Muckleshoot Tribe • Lower Elwha Klallam Tribe • Quinault Indian Nation • Selected universities, counties, and cities In terms of response, many recipients/respondents indicated that they had not only received the request for information, but had distributed the survey to many of their col- leagues. For example, a single Pennsylvania Department of Transportation (PennDOT) contact on the initial email list solicited, and received, responses to the survey from all 11 of the PennDOT districts. From PennDOT District 11 alone, 140 files of photos and site reconnaissance/monitoring documents were received. The California Department of Trans- portation (Caltrans) also provided a wealth of valuable information on stream bank protec- tion sites that are currently being evaluated through a cooperative study with California State University, Humbolt (CSU-H).

30 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures As a direct result of the survey a number of sites in the Ann Arbor, MI, area were suggested for field work by a co-author of NCHRP Report 544 in his response. Follow-up discussion revealed that most of these sites would be included in a text on bioengineering treatments and, as such, are very well documented (Goldsmith et al. 2014). Many of these stream bank projects are located within a mile of the University of Michigan campus in Ann Arbor and are (1) easily accessible, (2) have USGS stream gaging information, and 3) are well documented. Based on these factors it was proposed under Task 3 that sites in the Ann Arbor area be included in the Task 6 field evaluation program as the upper Midwest region (see Section 2.4). 2.3.2 Summary of Survey Results The purpose of the survey was to gather information on the current state of practice with respect to a variety of topics related to environmentally sensitive stream bank protection, and included detailed performance-related issues/questions for 16 specific protection treatments (see Appendix A). In total, 35 completed questionnaires were received. Eighteen states, the District of Columbia, and British Columbia are represented by the completed questionnaires, with eight states located west of the Mississippi River (12 responses) and 10 states located east of the Mississippi (23 responses). Most of the individuals responding were from state DOTs (30 responses). Several federal and state agencies responded (one each from FHWA, NOAA, U.S. Forest Service, and California Department of Fish and Wildlife), as well as one consulting firm. The distribution of and number of responses by state are shown in Figure 2.1. It should be noted that 12 of the 35 responses were received from PennDOT, with each of its districts providing a response across that state’s diverse physiography. Also note that the responses from South Dakota and Missouri are labeled “n/a” because the respondents from these two DOTs indicated that they did not have any experience with any of the 16 treatments listed in the survey. The physiographic regions included in the responses across the U.S. are indicated in Figure 2.2 by a check mark next to the legend. Sixteen of the 25 physiographic regions defined by the USGS (Fenneman and Johnson 1946) are represented by the survey responses. Because of the many responses received from PennDOT, physiographic regions 2 (Appala- chian Plateau) and 24 (Valley and Ridge) are well represented. Also well represented is region 6 Figure 2.1. Survey responses by state.

Findings 31 (Central Lowland) owing to the number of responses received from Illinois, Kansas, Minnesota, and New York. It should be noted that many responses represent multiple physiographic regions corresponding to a particular agency’s area of jurisdiction. A detailed listing of responses by state is provided in table format and shows the responses and frequency of use reported for the 16 treatments listed in the survey. Also shown are the physiographic regions associated with each response. The above-mentioned tabled and detailed response spreadsheets for each of the 16 treatments are included in the Compendium that accompanies this report. These spreadsheets tabulate performance, monitoring, and failure mode information for each of the 16 treatments. The survey was constructed to elicit specific information regarding the relative frequency of use of the various types of environmentally sensitive bank-protection treatments, and their performance. Table 2.8 provides a numeric tabulation of the frequency of use of the 16 bank- protection measures for which specific information was requested. Figure 2.2. Survey responses by physiographic region. Live brush layering 7 VMSE 9 Large woody debris 12 Vegetated gabions 6 Live staking 27 Willow posts/poles 11 Live siltation 5 Rootwad revetment 16 Live brush mattress 7 Vegetated articulating concrete blocks (ACBs) 6 Vegetated riprap 17 Vegetated gabion mattress 2 Soil/grass-covered riprap 10 Live fascines 13 Coconut fiber roll 12 Turf reinforcement mats 12 Other (describe): 0 Table 2.8. Bank-protection treatments: frequency of use.

32 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures By far, the most popular treatment currently in use is live staking (27 out of 35 responses). The survey responses indicated that vegetated riprap (17 responses) and rootwad revetments (16 responses) were the second-most popular treatments in current practice. In terms of performance, respondents were asked to qualitatively assess their experience with a particular treatment using the following scale: 1 = Failure 2 = Satisfactory 3 = Good 4 = Excellent This ranking scheme revealed that, in general, environmentally sensitive bank-protection measures typically result in satisfactory to good performance at most installation sites. The aver- age value of the numerical rankings ranged from a low of 2.2 to a high of 3.0, indicating that there is not much differentiation between treatments. Table 2.9 provides the averaged rankings along with a summary of the comments and observations received from the surveys. Only two treatments received more than one response indicating “excellent” performance: soil/grass-covered riprap (3 “excellent” ratings) and coconut fiber rolls (2 “excellent” ratings). On the other hand, none of the treatments received more than one response indicating “failure.” The two most common causes of substandard performance were drought and erosion of the bank slope. These factors were a recurring theme across nearly all of the 16 bank-protection measures listed in the survey. Protection measures incorporating live cuttings tended to indicate that improper plant selection or improper installation resulted in poor performance. Protection measures that incorporated a “hard” or “engineered” component such as rootwad revetment, large woody debris, articulated blocks, gabions, etc. exhibited failure modes associated with toe or edge scour, undermining, flanking, or geotechnical slope instability. 2.4 Field Site Evaluations 2.4.1 Site Selection Criteria Initially, 31 sites where environmentally sensitive bank-protection measures have been installed across the country were identified. Each of those sites had either been designed, con- structed, and/or monitored by one or more of this project’s research team members. Based on the information obtained during Phase I (Tasks 1 and 2), additional candidate sites for field investigation during Task 6 (Phase II) were identified and screened. The final “short list” of sites included a diversity of protection measures and geographic locations. The schedule and budget allowed site visits to 16 field evaluation sites in three geographic regions: • Southeast (Mississippi) • Upper Midwest (Michigan) • West Coast (Northern California) Priorities were to visit sites where measurements have or can be made and where those mea- surements can be correlated to original design intent and hydrologic/hydraulic history. The following section discusses the screening process and the schedule for field investigations completed during Task 6. The “short list” of field sites considered site groupings to maximize the return on Task 6 activities and also considered the desire to strike a balance between investment in field visits and Task 7 laboratory studies.

Findings 33 Protection Measure Avg. Score (1 = Failure, 4 = Excellent) Comments 1. Live brush layering 2.8 Seven respondents indicated experience with multiple sites and reported generally good performance. Montana reported some poor performance due to improper installation, and California noted both slope erosion and drought as factors leading to instances of substandard performance. 2. VMSE 2.7 Generally good performance reported by nine respondents. Montana reported some poor performance due to improper installation, while California and D.C. noted both slope erosion and drought as factors leading to substandard performance. 3. Large woody debris 2.9 Generally good performance. California noted failure modes associated with undermining and/or flanking as well as slope erosion and drought. Pennsylvania reported damage associated with herbivory leading to poor condition of the vegetation. 4. Vegetated gabions 2.8 Only six responses, generally reporting good performance. California noted toe scour, geotechnical slope instability, flanking, slope erosion, and drought as factors leading to failure. 5. Live staking 2.5 Because of the large number of responses for this treatment, performance ranged from "failure" to "excellent." Most respondents indicated experience with multiple sites and reported generally satisfactory performance. When mortality was indicated, the cause was typically lack of moisture, poor soil conditions, poor quality of cuttings, or a combination of these factors. California indicated that they do not use this as a standalone measure but incorporate a hard armor toe. 6. Willow posts/poles 2.5 Generally satisfactory performance. Wyoming reported failure due to improper plant selection. Drought was mentioned by five respondents as a factor leading to substandard performance. 7. Live siltation 2.5 Generally satisfactory performance, but not much usage reported (only five responses). 8. Rootwad revetment 2.4 Generally satisfactory performance, with toe scour/undermining being the leading cause of performance problems. 9. Live brush mattress 3.0 Generally good performance, but not much usage reported (seven responses). California noted that slope erosion, drought, and beaver damage resulted in some performance problems. 10. Vegetated ACBs 2.2 Satisfactory performance, but not much usage reported (six responses). Slope erosion and geotechnical slope stability problems were noted as factors leading to substandard performance. 11. Vegetated riprap 2.8 Generally good performance. Slope erosion and drought were the most commonly reported causes of substandard performance. 12. Vegetated gabion mattress 3.0 Only two responses for this treatment. South Carolina reported "excellent" performance, while Arizona reported "satisfactory" performance. 13. Soil/grass-covered riprap 3.0 Good performance. Slope erosion and drought were noted as factors causing less than desired performance. 14. Live fascines 2.7 Generally satisfactory to good performance. Bank and toe scour were noted as failure mechanisms, along with slope erosion and drought. 15. Coconut fiber roll 2.8 Generally good performance. Toe scour was reported as the most common cause of problems. Montana noted that this treatment is "expensive, but always works." 16. Turf reinforcement mats 2.8 Generally good performance. Bank scour, toe scour, and drought most often mentioned as factors leading to substandard performance. Table 2.9. Bank-protection treatments: qualitative performance assessments.

34 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures 2.4.2 Screening, Schedule, and Initial Observations—Field Site Visits Screening of the field evaluation sites included 31 field sites identified initially, review of responses to the Task 2 survey and follow-up discussion with several survey respondents, and the experience of research team members with specific sites in various geographic regions. The regional coverage listed in Section 2.4.1 provided the desired geographic diversity to obtain relevant information on a national basis from both humid and less humid regions and regions that experience significantly different climatic conditions (e.g., the upper Midwest sites experience a freeze/thaw cycle, while the Southeast sites do not). Regional selection was also influenced by the grouping of sites in reason- able proximity so that a maximum number of sites could be visited in a two- or three-day field trip. Additional considerations in site selection included: • Well-documented sites with a detailed photographic record. • Availability of detailed site design and installation data. • Access to a wide range of treatment types as classified by NCHRP Report 544 (see Appendix A). • Providing specific information on research needs identified in NCHRP Report 544 as listed in Appendix A. • Flood history, monitoring history, and availability of quantitative hydrologic and hydraulic data. • Availability of a person involved in design, construction, and monitoring of a site to guide the team during the site visit(s) and provide access to historic and current site data. The screening process and field evaluation sites recommended to the panel are discussed briefly below. Site photographs, construction date, and details of the treatment components for each site are provided in the following sections. The schedule for site visits during Task 6 and initial observations from each region are also summarized. Southeast (Mississippi) Site MS1: Buttahatchee River, near Columbus, MS (1.5 miles northeast) • Stone toe. • Erosion control blanket. • Rootwads. • Willow trenches—live staking. • VMSE in some locations. Site MS2: Buttahatchee River, near Columbus, MS (0.5 miles due north) • Willow posts and poles. • Bendway weirs with embedded large wood “locked” logs. Site MS3: Goodwin Creek near Batesville, Mississippi (northern Mississippi) • Stone toe. • Bendway weirs. • Rock riffles. • Live staking. • Live willow posts. • Containerized plantings. Site MS4: Harland Creek, near Lexington, MS (west-central Mississippi) • Stone toe. • Bendway weirs. • Live willow posts.

Findings 35 Site MS5: Hotophia Creek near Batesville, Mississippi (northern Mississippi) • Stone toe. • Stone spurs. • Live willow posts. • Live siltation. Sites MS1 through MS5 were suggested as typical of biotechnical stream bank treatments installed by many agencies throughout the Southeast (see Figure 2.3). A member of the research team has a unique historical perspective on these installations and notes that there are a number of open literature reports available covering design, installation, and monitoring of several of these sites. The research team member noted the following with reference to his knowledge and proximity to these sites: Buttahatchee River. 600 ft reach, 85 miles away from Oxford, MS. Construction 2012. Ero- sion control blanket, stone toe, willow stakes, rootwads, VMSE. No transportation infrastructure. Limited description of project design details available. Continuous stage and discharge record at USGS gage about 20 miles upstream. Goodwin Creek. 330 ft reach, 20 miles away from Oxford, MS. Construction 2007. Bendway weirs, stone-toe protection, cuttings (poles and stakes), and container plants. No transportation infrastructure. Design details and performance described in open literature. Heavy geotechni- cal analysis in design. Continuous discharge record from gage at nearby meteorological station, immediately downstream. Harland Creek. 11,000 ft reach, 125 miles away from Oxford, MS. Construction 1994. Bendway weirs, willow posts, stone toe. Bridge downstream. Design details and performance described in open literature. USGS stage, discharge, and suspended sediment records through 2000 for gage about 4 miles downstream. Hotophia Creek. 3,000 ft reach, 30 miles away from Oxford, MS. Construction 1992. Stone toe, stone spurs, post plantings, and post plantings above stone toe in a fashion similar to “live siltation.” Centered on highway bridge. Design details and performance described in open litera- ture. USGS stage and discharge record through 1997. Sites representative of the Southeastern U.S. were visited September 9–10, 2014. Typical site photos are shown in Figure 2.3. Value was added to these site visits by the presence of the original project designers and planners within the visit party at all sites. One site included two projects in the same reach of the Buttahatchee River, so a total of five projects were inspected, with construction dates ranging from 1992 to 2012. Treatments included stone toe, bendway weirs, rootwads, erosion control blanket, willow posts and poles (planted in a variety of ways), live staking, rootstock plantings, VMSE, and bank resloping. All projects were subjected to significant high flows within two years of construction. Banks at all project sites were stable with little damage or outstanding maintenance observed. However, one project (Site MS1 on the Buttahatchee) exhibited very poor performance of planted vegetation due to infertile, droughty soils exposed by bank grading. Older projects tended to have lush bank vegetation and a high level of ecological functionality within the riparian zone, but planted willow (pri- marily Salix nigra) had given way to planted or volunteer river birch (Betula nigra) and syca- more (Platanus occidentalis). The dominance of sycamore was impressive. Success of woody vegetation at one site was severely limited by the exotic vine, kudzu (Pueraria lobata). Design and construction best practices were noted at all sites. For more details on a typical Southeast site see Section 2.5.

36 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures Site Typical Before Project Typical 2014 Conditions Buttahatchee River Site MS1 Buttahatchee River Site MS2 Goodwin Creek Site MS3 Harland Creek Site MS4 Hotophia Creek Site MS5 Figure 2.3. Task 6 field site visits—southeastern field sites.

Findings 37 Upper Midwest (Ann Arbor, Michigan) Site MI1: Huron River (Nichols Drive), Ann Arbor, Michigan • Rock vanes. • VMSE. • Chimney drains. • Revegetation. Site MI2: Huron River (Nichols Arboretum/River Landing) Ann Arbor, Michigan • Vegetated riprap. • Live fascine. • ECB (erosion control blanket). • Revegetation. Site MI3: Huron River (Argo Cascades) Ann Arbor, Michigan • ECB w/grass seeding. • Boulder spurs. • Stepped pools. Site MI4: Fleming Creek, Northfield, Michigan • Rock Vanes. • Willow Pole planting. Site MI5: Malletts Creek, Ann Arbor, Michigan • Rock vanes. • Coir logs. • ECB w/live stakes. The upper Midwest sites were suggested by a co-author of NCHRP Report 544 in his response to the Task 2 survey. Follow-up phone conversations revealed that most of these sites are included in a text on bioengineering treatments and, as such, are very well documented. The text is entitled Bioengineering Case Studies: Sustainable Stream Bank and Slope Stabilization (Goldsmith, Gray, and McCullah 2014). The co-author noted that there are a total of 35 project case studies that are described and evaluated in the book from a retrospective point of view. Over a third of the case studies are stream bank protection or repair projects. At least five of these stream bank projects are located within a mile of the University of Michigan campus in Ann Arbor, and are (1) easily accessible, (2) have USGS stream gaging information, and (3) are well documented. For example: Fleming Creek has substantial stream bathymetry data. The velocity field around the rock vanes was measured using an acoustic Doppler velocimeter. Malletts Creek has a treated reach that is 1-½ miles long and consists of both conventional and bioengineered measures. River Landing and Nichols Drive are both large projects located on the Huron River, a major river that runs through the city of Ann Arbor. Both sites can be accessed by a road that runs parallel and adjacent to the river. Argo Cascades is also on the Huron River near Ann Arbor and provides the opportunity to document in-channel structures including boulder spurs and stepped pools. One of the NCHRP Report 544 coauthors volunteered to assist the research team in visiting and inspecting the Ann Arbor sites. Representative photographs from each site are shown in Figure 2.4.

38 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures Site MI1: Huron River (Nichols Drive) Ann Arbor, Michigan Constructed 2008. Rock vanes VMSE Chimney drains Revegetation Site MI2: Huron River (Nichols Arboretum/River Landing) Ann Arbor, Michigan Constructed 2005. Vegetated riprap Live fascine ECB Revegetation Site MI3: Huron River (Argo Cascades) Ann Arbor, Michigan Constructed 2013. ECB w/grass seeding Boulder spurs Stepped pools Figure 2.4. Task 6 field site visits—upper Midwest, Michigan field sites.

Findings 39 Five site visits were completed during the period October 1–2, 2014, for locations representa- tive of the upper Midwest. All sites are in Ann Arbor, MI. Conditions at Site MI2 on the Huron River are typical of those encountered during the Michi- gan site visits (see Figure 2.4). This site is also referenced as River Landing at Nichols Arboretum. The Nichols Arboretum project was intended to replace concrete rubble on the bank with a natural, less visually intrusive treatment with the help of trainees and volunteers. Another goal was to provide visitor access to the river at the site. Concrete rubble was removed from the site and the bank graded back to a stable 2H:1V slope. This project involved design of an environ- mentally sensitive, stream bank repair approach that also provides access to the river. The project provided hands-on opportunities for volunteer participation and training. Steps built into the bank at two locations in the protected reach now provide ready access to water’s edge. For more details on the upper Midwest sites see Sections 2.5 and 4.4. West Coast (Northern California) Site CA1: Sacramento River at Sacramento, California (RM 47.0L and RM 62.5R) • Large woody debris. • Willow posts and poles. Site MI4: Fleming Creek, Northfield Township, Michigan Constructed 2006. Rock vanes Willow pole planting Site MI5: Malletts Creek, Ann Arbor, Michigan Constructed 2011. Rock vanes Coir logs ECB w/live stakes Figure 2.4. (Continued).

40 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures • Vegetated riprap. • Soil and grass-covered riprap. Site CA2: Lower American River (LAR) at Sacramento, California (RM 2.0L) • Large woody debris. • Willow posts and poles. • Vegetated riprap. • Soil and grass-covered riprap. Site CA3: LAR at Sacramento, California (RM 6.9L) • Willow posts and poles. • Vegetated riprap. • Soil and grass-covered riprap. Site CA4: Alamitos Creek near San Jose, California • Willow posts and poles. • Live siltation. • TRMs (turf reinforcement mats). Site CA5: Guadalupe River near San Jose, California • Willow posts and poles. • Live siltation. • Coconut fiber roll. • Longitudinal logs at toe. Site CA6: Russian River at Highway 128 (Geyserville Bridge) • Large woody debris. • Willow posts and poles. • Live siltation. • Longitudinal stone toe. • Rock vanes. The CA1 sites were implemented as part of the Sacramento River Bank-Protection Project on the Sacramento River and the LAR, California. Specific guidance was also pro- vided for erosion sites located on the LAR, a tributary to the Sacramento River (Sites CA2 and CA3). A special report on the “Guadalupe River Restoration Project: Biotechnical Channel Sta- bilization Solutions” (McCullah and Dettman 2004) was published and thus, the Guadalupe River site (CA5) is very well documented. In addition, the evolution of treatments at this site includes initial construction of treatment alternatives in 2002, failure of the initial approach, and reconstruction in 2003 using different techniques (which survived a large flood in 2004). The Alamitos Creek and Russian River sites (Sites CA4 and 6), while some distance away from the Guadalupe River sites, offered the opportunity to evaluate alternative treatments (rock toe and rock barbs with willow siltation and staking), and TRMs in a different geomorphic setting. Representative photographs from each site are shown in Figure 2.5. Six site visits were completed during the period October 21–22, 2014 for sites representative of the West Coast (Northern California).

Findings 41 The Sacramento River and LAR sites represent a unique design that balances the need to upgrade protection at critical erosion sites on the Sacramento levee system with the requirement to enhance environmental/ecological values. Regarding the Guadalupe River site, it was concluded in McCullah and Dettman (2004) that the restoration work performed in 2003 was extremely successful in stabilizing the slope. In April of 2004, the bank was becoming vegetated and the willows were successfully established at the toe. Combining techniques such as cobble revetment, live staking, pole planting, live siltation, hydroseeding, and coir blankets provided the stability, habitat, and aesthetic value needed for the site. For more details on a typical California site see Section 2.5. Site(s) CA1: Sacramento River, California (two priority erosion sites downstream of Sacramento) Constructed 2006. Rock toe with scalloped bench Instream woody material Live staking Container plantings Beaver fence (temporary) Sites CA2 and CA3: LAR, Carmichael, California (two sites) Constructed 1998. Rock toe with scalloped bench Instream woody material Live staking Container plantings Beaver fence (temporary) Site CA4: Alamitos Creek near San Jose, California Constructed 2010. Willow posts and poles Live siltation TRM Coir netting Figure 2.5. Task 6 field site visits—West Coast (Northern California) field sites. (continued on next page)

42 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures 2.5 Compendium of Field Data, Documentation, and Photographs 2.5.1 Introduction Based on the data acquired from the Task 2 survey, and photographic and case history docu- mentation from the Task 6 field investigations, a Compendium of biotechnical treatments in a searchable database format was developed. The Compendium represents a collection of the 16 field data forms from the site visits described in Section 2.4. The Compendium provides access to the full site visit report from each of the 16 site visits in addition to a variety of supporting data and documentation gathered during the site visits and could include (where available for a specific site): • Design information, • Presentations, • Reports and published articles, • Specifications, Site CA5: Guadalupe River at Alma Road, San Jose, California Constructed 2002. Log toe Coir wattle ECB Native grass vegetation Failed in 2002, redesigned and rebuilt Reconstructed 2003. Rock toe Live siltation ECB Native grass vegetation Coir wattles and cutoff trenches Survived large flood in 2004 Site CA6: Russian River at Geyserville Bridge, SR 128, Geyserville, California Constructed 2010. Rock toe Rock barbs Live siltation Live staking Figure 2.5. (Continued).

Findings 43 • Spreadsheet files, • Cost data, and • Hydraulic and/or hydrologic data. The Compendium has a graphical user interface (map) which directs the user to a specific site location and is presented in a searchable database format. Tables 2.10 and 2.11 illustrate the topics available for searching the Compendium database by both treatment type (Table 2.10) and site information available (Table 2.11). The Compendium is provided with this Final Report as a companion to the detailed design guidelines in Chapter 4. As an introduction to the Compendium, the following section provides extracts of the information contained in a typical site visit report from each of the three regions referenced in Section 2.4. In addition, data and documentation from the Mallets Creek site visit in the upper Midwest region (see Figure 2.4, Site MI5) have been expanded to provide one of two application examples/case studies in Section 4.4. 2.5.2 Selected Findings on Environmentally Sensitive Treatments Upper Midwest Region (Michigan) Huron River, Nichols Arboretum Near Ann Arbor, Michigan 1. Purpose and Selection The Huron River, which drains into Lake Erie, has a drainage area of 730 sq. miles, a daily mean stream flow of 380 cfs, and maximum recorded peak discharge at the site of 5,840 cfs (1918). Dams immediately upstream and downstream of the site tend to regulate flow and sup- press peaks in this reach of stream. The river is also quite wide at the site, which tends to decrease velocity and depth of flow. The Nichols Arboretum site (see Figure 2.4, Site MI2) is also referenced as the River Landing site. The site in question was one of the most heavily used and degraded sites in the University of Michigan Arboretum. It traditionally provides one of the main access and viewing points on the river. Bank erosion was a perennial problem at this site. Concrete rubble was placed on the bank in the 1970s in an attempt to control erosion. The rubble was unsightly, had exposed rebar, and was largely ineffective. Effective repair and restoration required removal of the rubble and replacement with a more natural bank-protection system. The bank-protection project was completed with amenities that create a gathering place. Stone steps provide access to the water’s edge. A rail fence along the top bank separates the heavily vegetated bank from a grassed picnic area with tables and shade trees. The local USDA NRCS was looking for sites on streams that could be used as training exercises for county road department personnel. Cost for this site was reduced by this program and by use of project construction as an exercise for a training course offered by the arboretum. The project was completed in October 2005. The project was tested by a flow of about 4,000 cfs in May 2011, which was the 14-year event based on the annual series. The project was intended to replace the concrete rubble on the bank with a natural, less visually intrusive treatment. Another goal was to provide visitor access to the river at the site. Measures selected for implementation were compatible with construction (planting) by training course participants. 2. Design of System Components The concrete rubble was removed and the bank graded back to a stable 2H:1V slope (see Figure 2.6).

Site Number Site Name Short Name Table 2.10. Treatment Types. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Live brush layering VMSE Large woody debris Veg. gabions Live staking Willow posts and poles Live siltaon Rootwad revetment Live brush maress Veg. ACBs Veg. riprap Veg. gabion maress Soil and grass covered riprap Live fascines Coconut fiber roll TRMs & ECBs 1 Alamitos Creek - San Jose CA Alamitos CA4    2 Buahatchee River - US HWY45 - Columbus MS Buahachee MS1      3 Buahatchee River - US HWY373 - Columbus MS Buahachee MS2   4 Fleming Creek - Dixboro MI Fleming MI4  5 Goodwin Creek - BatesvilleMS Goodwin MS3   6 Guadalupe River - San JoseCA Guadalupe CA5     7 Harland Creek - Howard MS Harland MS4  8 Hotophia Creek - BatesvilleMS Hotophia MS5  9 Huron River - Argo Cascades- Ann Arbor MI Huron MI3  10 Huron River - NicholsArboretum - Ann Arbor MI Huron MI2     11 Huron River - Nichols Dr - Ann Arbor MI Huron MI1   12 Lower American River - RM2.0L - Sacramento CA LAR CA2     13 Lower American River - RM6.9L - Sacramento CA LAR CA3    14 Malles Creek - Ann ArborMI Malles MI5    15 Russian River - GeyservilleCA Russian CA6    16 Sacramento River - RM47.0L& RM62.5R - Sacramento CA Sacramento CA1       Table 2.10. Treatment types.

Site Number Name Short Name Table 2.11. Site Informaon Available. 1 2 3 4 5 6 7 Design Info PowerPoint Reports Specificaons Excel file Cost Data Hydraulics and Hydrology(in PDF format)(in PDF format) 1 Alamitos Creek - San Jose CA Alamitos CA4 2 BuŽahatchee River - US HWY 45 - Columbus MS BuŽahachee MS1 3 Buahatchee River - US HWY 373 - Columbus MS Buahachee MS2 4 Fleming Creek - Dixboro MI Fleming MI4 5 Goodwin Creek - Batesville MS Goodwin MS3 6 Guadalupe River - San Jose CA Guadalupe CA5 7 Harland Creek - Howard MS Harland MS4 8 Hotophia Creek - Batesville MS Hotophia MS5 9 Huron River - Argo Cascades - Ann Arbor MI Huron MI3 10 Huron River - Nichols Arboretum - Ann Arbor MI Huron MI2 11 Huron River - Nichols Dr - Ann Arbor MI Huron MI1 12 Lower American River - RM2.0L - Sacramento CA LAR CA2 13 Lower American River - RM6.9L - Sacramento CA LAR CA3 14 Malles Creek - Ann Arbor MI Malles MI5 15 Russian River - Geyserville CA Russian CA6 16 Sacramento River - RM47.0L & RM62.5R - Sacramento CA Sacramento CA1 Table 2.11. Site information available.

46 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures Figure 2.6. Huron River at Nichols Arboretum.

Findings 47 Vegetated riprap. The bank toe was armored with riprap. The riprap was vegetated using the joint-planting method, i.e., live stakes representing six to eight woody tree and shrub species were inserted between openings or interstices in the riprap: sandbar willow, red osier dogwood, ninebark, silky dogwood, and native roses. Not all stakes were dormant when planted. Some non-native white willow was also used, but the project coordinator said he would like to avoid using non-native species in the future. Live fascine. A single fascine was placed above the riprap toe, running longitudinally. The bottom of the ECB was run into the fascine trench, and the fascine was anchored in the trench using both live stakes and lumber construction stakes. ECB. An ECB was placed on the bank and up over the top. The blanket was anchored to the bank using staples and pegs. Topsoil was placed and spread over the bank before placing the ECB. Revegetation. A seed mix of native grasses and plants was spread on the bank before placing the ECB. 3. Structural and Installation Guidelines: Plant materials were harvested from nearby parks and soaked in barrels of water for about one week. Except for bank resloping and rock placement, all construction was completed in one day. Compost was spread on the new bank slope to fill the trench that was excavated to receive the live fascines. Teams of workers trimmed the plant materials, bound fascines on a rack made from lumber (like a sawhorse), and carried the material down the bank for installation. Consid- erable plant mortality occurred due to driving stakes through riprap. Many stakes were damaged (skinned, split) in that way. After planting, no irrigation was provided for plantings. Fascines and stakes were on a lower bank, where soil moisture was higher. Plantings resulting from live fascines have required some cutting to maintain the view of the river from the recreation area on the top bank. 4. Construction and Maintenance Best Practices: Some concern was expressed about the bank vegetation limiting access to the river. This con- cerned was allayed by constructing a set of stairs consisting of granite blocks from the top of the bank to water’s edge. In addition, the top of the bank was planted with grass, and picnic tables were provided so visitors could enjoy riverine views. Plants growing on this site have been used as a source of cuttings for other work. The vegetated bank has been maintained by some selective removal of large individual and invasive species to maintain plant diversity and views of the river from the picnic area on the top bank. 5. Performance, Failure Mechanisms, and Longevity Vegetative establishment beneath the ECB and in the fascine has generally been excellent. The bank is now stable. The live stakes in the riprap have not fared as well (see Table 2.12). Many stakes were damaged when driven through the rock blanket to plant them in interstices. A year after installation only GENUS RIPRAP (%) No. of Surviving Stakes (% of Planted Stakes) FASCINE (%) No. of Surviving Stakes (% of Planted Stakes) Cornus (dogwood) 26 (48) 14 (78) Salix (willow) 13 (24) 3 (16) Sambucus (elderberry) 15 (28) 1 ( 6) Table 2.12. Vegetative survival rates for 2006.

48 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures about 30% of the live stakes survived and sent out shoots. The bent willow pole or willow bundle method, which eliminates many of the problems associated with the joint-planting method, should be considered in the future (see Section 4.3.3). Red osier dogwood was the most successful species planted and was quite prolific when the site was inspected in 2014. The riprap toe has survived several high-water events but has not been tested by floods approaching peak flows of record. This project is a good example of an environmentally sensitive, stream bank repair approach that also provides access to the river. The project has provided hands-on opportunities for vol- unteer participation and training. Steps built into the bank at two locations in the protected reach now provide ready access to water’s edge. The project resulted from planning and cooperation on the part of personnel from the Uni- versity of Michigan, the city of Ann Arbor, Washtenaw County Road Commission, a private consulting firm, and the NRCS. 6. Cost and Availability Only ~$30,000 due to the magnitude of in-kind contributions and NRCS assistance. 7. Ecological Issues Exotic European alder becoming a nuisance. 8. Assessment of Functionality During October 2014 Site Visit Aquatic ecosystem functionality and recovery is limited by a dam located up- and downstream from this site. However, these dams ensure vertical channel stability and sharply limit water- level fluctuations, which greatly facilitates revegetation of treated stream bank. The river level fluctuates only about 3 ft. Minor sediment deposition was noted where a large gully system entered the channel. Invasive exotic woody species (buckhorn, honey suckle, European alder, black locust) compete with planted native species. The site was evaluated as “functional at risk” with an upward trend (improving) in functionality. 9. Published Documentation/Sources/Citation Goldsmith, W., Gray, D.H., and McCullah, J., 2014. Bioengineering Case Studies: Sustainable Stream Bank and Slope Stabilization, Springer Verlag, New York. Lawson, R., Bouma, D., Olsson, K., Rubin, L., and Powell, E., 2008. Watershed Management Plan for the Huron River in the Ann Arbor—Ypsilanti Metropolitan Area, prepared on behalf of and with funding support from Janis A. Bobrin, Washtenaw County Drain Commissioner, Huron River Watershed Council, Ann Arbor, MI. Southeast Region (Mississippi) Buttahatchee River Near Columbus, Mississippi 1. Purpose and Selection At this site, bendway weirs were selected as a robust but cost-effective way to protect the erod- ing bank along a very narrow strip of land dividing the channel of the Buttahatchee River, which provides valuable habitat, from a large, abandoned gravel pit (see Figure 2.3 Site MS2). Several old floodplain gravel pits have captured portions of the river in this reach, to the detriment of habitat quality and overall channel stability. The project was constructed under a United States Fish and Wildlife Service (USFWS) program in 2012.

Findings 49 2. Design of System Components The functionality of the bendway weirs at this site was enhanced by the addition of embedded large wood “locked” logs (Figure 2.7). • Weir crest width = 4 D100 (8–10 ft), where D100 is the maximum stone size. Some of the crests here were even wider to allow machine access down the crest for stone placement during construction. • Gage records show water near the top bank during January, March, and April 2011 (Figure 2.8). The January 18, 2013, peak flow of 15,100 cfs was slightly greater than one-year discharge from USGS Streamstats of 11,600 cfs. 3. Structural and Installation Guidelines Start and end protection at stable points—these usually are located at the extreme ends of the bendway. Figure 2.7. Bendway weirs with locked logs at Buttahatchee River site. Figure 2.8. Stage hydrograph for Buttahatchee River near Aberdeen. Mean daily stage for inspection of date of September 9, 2014 was 5.44 ft.

50 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures 4. Construction and Maintenance Best Practices • Construction of weirs should proceed by constructing the weirs located at the upstream and downstream ends and the one in the middle to establish “line of control” riverward tips. Remaining weirs are then constructed from upstream to downstream. • Avoid wood (locked logs) projecting above waterline that will trap more debris. • “Locked logs” are constructed by placing logs with rootwad near the location for the bendway key and placing logs pointing downstream at an angle of between 20 and 70 degrees from the bank depending on local conditions. Locked logs were also placed with upstream angles between 20 and 40 degrees (see Figure 2.7). Then stone for the bendway weir is placed on top of the rootwad and lower part of the tree trunk to anchor them securely. These locked logs resist ice removal in northern rivers—ice freezes around them first. 5. Performance, Failure Mechanisms, and Longevity • Most of the 2012 willow planting did not succeed. Willow cutting survival has been slightly better on the upper and middle banks. • Hardwood tree rootstock planted in early 2014 with plastic tree protectors. These are cylindri- cal plastic enclosures about four inches in diameter with perforations in the sides near the top or bottom of the tube. Haul road planted with hardwoods and native grass plugs (mossy oak). • Rock volume required to complete the project was twice the estimate due to sinking and erosion. 6. Cost and Availability • Contractor: $102,000. • Rock; $89,000. • Design; $13,000. • Planting; $12,500. • LiDAR; $22,000. • Planning, preliminary designs, permits, etc.; $48,160. TOTAL: $287,520 7. Ecological Issues The USFWS re-surveyed mussels (earlier done in 1989) finding strong recovery with threat- ened and endangered species in a riffle just downstream from this site. 8. Assessment of Functionality During September 2014 Site Visit Bendway weirs and locked logs have provided aquatic habitat diversity and cover, and appear to be stable and functioning well. Bank revegetation is proceeding slowly. A few of the willow cuttings have survived and are growing slowly, perhaps due to herbivory. Planting stock (see Walter et al. 2013 below) on the top bank and haul road are surviving at present, but are less than one year old. The site was evaluated as “functional at risk” with an upward trend (improving) in functionality. 9. Published Documentation/Sources/Citation Pollen, N., Simon, A. Klimetz, L., and Klimetz, D., 2005. “Stability analysis of the Buttahatchee River Basin, Mississippi and Alabama,” Watershed Physical Process Research Unit, National Sedimentation Laboratory, Oxford, MS. Prepared for Mississippi Department of Environmental Quality. Walter, W. D., Godsey, L. D., Garrett, H. E., Dwyer, J. P., Van Sambeek, J. W., and Ellersieck, M. R., 2013. “Survival and 14-Year Growth of Black, White, and Swamp White Oaks Established as Bareroot and RPM®-Containerized Planting Stock,” Northern Journal of Applied Forestry, 30(1), 43–46.

Findings 51 West Coast (Northern California) Lower American River Near Sacramento, California 1. Purpose and Selection This project is part of the Sacramento River Bank-Protection Project on the Sacramento River and the LAR, California. This project for the USACE, Sacramento District provided design guid- ance for bank (levee) protection measures at critical erosion sites to be constructed on both the Sacramento River and the LAR, a tributary to the Sacramento River. This site on the LAR served as a basis for the design of future bank-protection sites along the LAR (Site RM 2.0L—see Figure 2.5 Sites CA2 and CA3). Hydraulic analyses were conducted to provide information necessary for the design of bank-protection measures (velocities, shear stresses) and to analyze the impact of the designs on hydraulic conditions and flood conveyance. A two-dimensional model was cre- ated representing the lower 15 km of the LAR. The model was used to analyze both existing and project conditions. The bank-protection and mitigation design concepts developed represented the state of the art in balancing flood control and environmental needs along the LAR. Innovative features include water-side low berms for the establishment of native terrestrial vegetation and irregular shorelines providing hydraulic variability for fisheries habitat. The result is a highly complex riprap design. This site is approximately 2,200 ft. in total length and the bank-protection and mitigation features were constructed in 1999. The site is bound by four bridges: N 12th Street/Hwy 160, Lincoln Hwy, Pedestrian, and R/R. Mitigation features include instream woody material (IWM), an undulating, cobble-lined, low-berm soil trench, and plantings on the low-berm, middle- berm, and upper-slope planting surfaces (see Figure 2.9). The onsite design includes a variety of surfaces capable of supporting vegetation; low river- side berms (small constructed floodplains) with varying berm-surface elevations and shoreline configuration as well as woody materials submerged in constructed embayments or smaller bank scallops (see Figure 2.10). Native woody and herbaceous riparian vegetation was planted on an engineered soil trench placed in the revetment at the site (including a low-berm face, low berm, lower slope, upper slope, and middle berm) with the goal of creating a self-sustaining, mixed canopy riparian forest and riparian scrub habitat, shaded riparian aquatic (SRA) habitat, and valley elderberry longhorn beetle (VELB) habitat. Project goals included: • Geotechnical stability for bank and levee, • Environmentally self-mitigating, Figure 2.9. 2002 aerial photo of LAR reach RM 2.0L.

52 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures • No impact on flood capacity, and • Erosion protection (high-energy environment). 2. Design of System Components • 2-D Hydraulic Model (pre- and post-project), • Topographic and bathymetric surveys, • Slope stability analysis, • Multiple benches for habitat enhancement, • Lower benches submerged at specific environmental flows, and • Embayments along lower bench for fish habitat (Figure 2.11). 3. Structural and Installation Guidelines USACE Standard Specifications (included in data and documents folder for this site in the Compendium). Figure 2.10. 2001 Aerial photo—close-up of embayments with woody material embedded in riprap at LAR Reach RM 2.0L. Figure 2.11. 2001 photo looking downstream at LAR reach RM 2.0L.

Findings 53 4. Performance, Failure Mechanisms, and Longevity Excellent performance—strong local sponsor [Sacramento Area Flood Control Agency (SAFCA)] ensured adequate follow-up and vegetation maintenance. Reducing beaver pruning of planted trees (in particular cottonwood) and shrubs on the berm face and low berm remains the greatest challenge to meeting SRA habitat goals. Due to the age of the beaver fence, the fence is becoming brittle, providing easier access to the site by beavers. During 2010, linear beaver fence maintenance and cage installation and relocation continued to be important management actions to minimize beaver damage to vegetation. This strategy proved successful and limited further tree losses or damage at these sites, as well as other sites where cages were previously installed. 5. Cost and Availability Unknown 6. Ecological Issues Bank-protection construction resulted in the loss of riparian, SRA, and special-status species habitat. Several mitigation measures were incorporated into project designs, including a vari- ety of surfaces capable of supporting vegetation, low riverside berms (small constructed flood- plains) with varying berm-surface elevations and shoreline configuration, and woody materials submerged in constructed embayments or smaller bank scallops. Native woody and herbaceous riparian vegetation was planted on revetment (including a low berm face, low berm, lower slope, upper slope, and middle berm) with the goal of creating a self-sustaining, mixed canopy riparian forest and riparian scrub habitat, SRA habitat, and VELB habitat. 7. Assessment of Functionality During October 2014 Site Visit: Vegetation at this site continued to grow well and many native plants, including cottonwood, alder, box elder, rose, and sandbar willow are naturally regenerating on the low berm (see Fig- ure 2.12). Protecting the vegetation at this site from beavers has been an ongoing challenge and has required intensive beaver fence maintenance. Maintenance actions include repairing and replacing sections of fence that are in disrepair due to both natural degradation of the fencing over time and vandalism. This site has not received supplemental irrigation since 2004 and yearly monitoring confirms that the plants at this site have successfully established. Coloniza- tion by non-native woody vegetation is minimal at this site but was removed when found. This site was evaluated as “functional” during the 2014 site visit. Figure 2.12. 2011 photo looking downstream at LAR reach RM 2.0L.

54 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures 8. Published Documentation/Sources/Citation SAFCA 2010 Annual Report (December 2010)—Lower American River Sites Year 11, Mitiga- tion Monitoring Report, Bank-Protection Sites: 1–5, Offsite Mitigation Areas: RM 0.9R, RM 3.3R, 11.6R, Sacramento Area Flood Control Agency, Sacramento, CA. 2.5.3 User Guide for the Compendium A brief step-by-step guide to using the Compendium is provided below: 1. Make sure you have an active connection to the Internet. 2. Insert the accompanying disk into your computer. 3. Open the SiteBrowser by clicking on the Field Site Map (NCHRP-Oct202015.htm) button from the disk menu screen. A map of the United States will pop up with pointers showing the location of each of the 16 field sites. 4. You can use your mouse to pan and zoom in on a group of sites and click on a pointer to see which site it is pointing to. Alternatively, if you have perused Tables 2.10 and 2.11 of this document and know which site you would like to navigate to, you can click on the site name in the pane on the right-hand side of the map and it will take you to the appropriate pointer. 5. Once a site pointer is selected, click on the Data Folders link to gain access that site’s compre- hensive data. Then, by clicking on the “Docs & Supporting Data” folder, you’ll have access to all the available information for that site. By clicking on the “Site Visit Field Form” folder, you can examine the detailed field visit notes, sketches, and assessments performed by the research team for the site visit performed at this site as part of this research project. 6. Alternatively, to access the site data folders directly, click on the Field Site Date Folder (NCHRP_Data) button from the disk menu screen. You’ll have access to the “Docs & Sup- porting Data” folders and the “Site Visit Field Form” folders from here. To access the Compendium database in Microsoft Access™, click the Field Site Searchable Database (NCHRP_Oct20-2015) button. The database contains the following four tables: • Bank-Protection Treatments: This table simply contains a cross-reference to abbreviations for the 16 different types of treatments used as column headings in other tables. • Field Sites: This table contains the site name, a short name, and the latitude-longitude coor- dinates of each of the 16 field sites. • Site Treatment Types: This table contains the various types of bank-protection treatments found at any particular field site. Many sites consist of more than one treatment. This table is identical to Table 2.10 in Section 2.5.1. • Site Information Available: This table contains the various types of information available for each site (e.g., design information, PowerPoint presentations, documents from the literature, etc.). This table is identical to Table 2.11 in Section 2.5.1. By using the database’s structured query language (SQL), you can create a specific query to develop a list of sites that have certain attributes of interest. For example, if you wanted a list of sites where live staking AND vegetated riprap was used, AND for which design information AND specifications are available, an appropriately structured query would provide a short list containing the following sites: Site Number Name Short Name 2 Buttahatchee River—US HWY 45—Columbus, MS Buttahatchee MS1 16 Sacramento River—RM47.0L & RM62.4L—Sacramento, CA Sacramento CA1

Findings 55 2.6 Summary of Findings and Observations from Current Practice 2.6.1 Literature Review Since publication of the results of NCHRP Project 24-19 by McCullah and Gray (2005), numerous reviews, handbooks, and measure-specific guidance documents for environmentally sensitive bank protection have been published by federal, state, and local agencies. These docu- ments emphasize the steps describing various measures and provide construction or installation guidance. Similarly, large numbers of case studies of biotechnical bank-protection projects have appeared, but few contain hydraulic data. Hydraulic design criteria are scarce and with few exceptions rely on the literature that was summarized within the NCHRP Project 24-19 prod- ucts. The data underlying the hydraulic criteria are drawn from a variety of sources and vary in quality from qualitative anecdotal rules of thumb to detailed laboratory measurements. Many are based on isolated spot measurements of velocity. Considerable progress has been made in recent years in understanding the biology of ripar- ian plants and developing technology for improving planting success. Limited but important advances have been made in understanding and simulating the complex fluid mechanics of open- channel flows adjacent to vegetated banks. Slope stability models have been modified to include contributions of roots to soil strength, and such models are becoming more widely employed. Over the past 10 years, publications dealing with stream erosion control, riparian zone and floodplain restoration, and interactions among plants, slope stability, and stream hydraulics have multiplied prolifically. Applied literature includes how-to guides, handbooks, case studies, and other studies and documents focused on the performance of a specific measure (e.g., willow stakes or rootwads). A subset of these documents is composed of guidelines and research reports on techniques for handling plant materials (e.g., soaking or refrigeration) to improve survival and performance. Most of the applied literature has been published by governmental agencies and not in peer-reviewed scientific or engineering journals. Much of the content in these docu- ments is duplicative or redundant. Another class of documents consists of guidelines for designing or constructing a single type of environmentally sensitive bank-protection technique. Although guidance may be gleaned from case studies that feature a specific technique, single-measure guidelines are primarily direc- tive in nature and do not relate experience at a single site or group of sites. Typically such guide- lines focus on construction details: how to prepare the site and materials and assemble or plant them to produce a finished project. To date, case studies have provided a higher level of reality than model studies or labora- tory experiments, but are difficult to generalize for application to other sites due to site-specific conditions, short periods of observation or insufficient data to fully characterize the erosional processes. Most of the flume experiments (and associated numerical modeling) have assumed a uniform stand of bank vegetation that does not vary its characteristics in the lateral direction even though natural riparian vegetation and biotechnical bank-protection measures usually feature different types of protection for the bank toe, mid bank, and upper bank. Interactions between riparian vegetation and stream channel morphology have been studied at larger scales using statistical approaches by fluvial geomorphologists (e.g., Bledsoe et al. 2011). Increasingly sophisticated numerical models (e.g., computational fluid dynamics) have been used to extend and complement field and laboratory results across a range of spatial scales. In a particularly relevant example, Rahmeyer and Werth (1996) and Freeman et al. (2000 and 2004) reported results of over 220 flume experiments involving 27 different real plant types and groupings. In-channel (not stream bank) vegetative flow resistance was found to decrease with

56 Evaluation and Assessment of Environmentally Sensitive Stream Bank Protection Measures velocity and depth for submerged plants but to increase with depth for emergent plants. Flex- ible, submerged plants with leaves formed a streamlined (teardrop) shape that reduced the flow forces on the plants, protecting leaves and smaller stems from breakage. Minimum plant velocity limits of 3 to 4 fps (0.9 to 1.2 mps) were observed for leaf failure, and most of the leaf and stem failures were the result of impact with bed material and debris. Stands with more or less uniform canopy geometry concentrated flow underneath the canopy, resulting in general scour. Stream banks erode or retreat due to fluvial erosion processes, sliding or mass wasting, and combinations of these processes. Scientists have long realized the links between vegetation and bank slope stability, but since publication of McCullah and Gray (2005), several important advances have been made in development of analytical techniques. For example, work by Pollen (2007) has advanced understanding of temporal and spatial variability in root reinforcement due to variations in soil type and moisture, and work by her team has shown that soil-root matri- ces are better simulated by fiber-bundle models that allow progressive failure of roots (weakest first) rather than the “all-roots-break-instantaneously” assumption of the older perpendicular models. The fiber-bundle model algorithm has been incorporated into the BSTEM, a spread- sheet tool used to simulate stream bank stability (see Section 4.2.4). 2.6.2 Observations from Current Practice Both the survey of practitioners (Task 2) and the site visits conducted under Task 6 of this study provide fertile ground for insights on current practice for environmentally sensitive treat- ments. For the treatment components tested under Task 7 (live staking, live siltation, rock toe, VMSE) observations on current practice are summarized in Sections 4.5.2 and 4.5.3 in relation to the design guidelines presented in Section 4.3. Additional examples of current practice can be found in the application case studies presented in Section 4.4. Both a humid region example and an arid region example were chosen to illustrate the integration of hydraulic engineering analysis and design with the multidisciplinary approach necessary to achieving success with a river restoration project. The examples involve the cooperative efforts of multiple agencies and the challenges of meeting stakeholder goals and expectations. In both cases monitoring and maintenance issues are addressed and lessons learned are summarized.