Attachment L - Burrowing Mammal Plant Levee Study-Final-31Jan2023.pdf

Final

Chinook Landing Levee Burrowing Mammal and Woody Vegetation Intrusion Study

Prepared for January 2023

Multnomah County Drainage District

Attachment L Burrowing Mammal Plant Levee Study

TABLE OF CONTENTS

Page

1.0 PROJECT UNDERSTANDING AND BACKGROUND

2.0 METHODS

2.1 Species of Interest

2.2 Field Survey Methods

2.3 Geospatial Analysis

3.0 FINDINGS AND RESULTS

4.0 CONCLUSIONS

5.0 REFERENCES

Appendices A. Figures .......................................................................................................................... A-1 B. Photographs ................................................................................................................. B-1

List of Tables

Table 1. Burrowing mammal and woody-plant Species with the Potential to Occur in the Project Area; Includes Characteristics Contributing to Levee Structural Damage; Species Listed in Alphabetical Order by Common Name

1.0 EXECUTIVE SUMMARY

Chinook Landing Levee may be compromised by burrowing mammal excavations and intrusion of roots from woody plants (Figure 1). This levee is an essential piece of infrastructure, preventing inland flows of the Columbia River and protecting a variety of public and private structures. Aquatic mammals excavate burrows at or near the waterline during stable flow periods, typically in spring and fall. Plant roots of trees penetrate levees and facilitate piping of water. The presences of large root systems create potential for wide and deep cratering within the levee if trees topple due to storm events and/or disease. Of particular concern are the density and size of burrows and large trees rooted between seasonal low-water and high-water marks. This study documents burrowing mammal excavations, specifically those of large aquatic mammals (i.e., American beavers), as well as large trees growing on the waterward levee face, to establish the potential of levee failure.

This potential was based on known biological characteristics of these influences, allowing the failure potential and significance to be quantified.

Large trees are present throughout the levee and have caused extensive erosion from their root systems being pulled out of the levee surface when toppled. This potential is substantially increased when undermined by large beaver tunnels. Allowing cottonwoods to grow to maturity on the levee has the potential to create structural problems when they topple, and allowing ivy to expand into their crowns substantially increases the likelihood of toppling. One large cottonwood had toppled spring of 2022. This tree had ivy growing into its 50 ft (15 m) diameter crown. The rootward crater left in the levee prism was 50 ft (15 m) in diameter and over 5 ft (1.5 m) deep.

This tree had been undermined by extensive burrowing under its root-wad. It is important to consider whether the erosive process of beaver tunneling and root-wad cratering is sufficient to structurally damage the levee, or facilitate its structural damage indirectly by making it susceptible to high flows.

Figure 1. Potential Failure Scenario Present at Chinook Landing Levee; x-axis represents Linear Feet

2.0 PROJECT UNDERSTANDING AND BACKGROUND

Discussions with the Multnomah County Drainage District (MCDD) identified the objective of this work being to inform the design of levee repairs through supporting the prioritization process established in the DRAFT IGA6: MCDD Animal Burrow Remediation on Salmon Creek (Metro and Multnomah County Drainage District 2021). This will be accomplished by providing a biologically based assessment of the extent that the levee is currently compromised by mammal burrows and tree roots. Based on the findings from this study, recommendations are provided to further guide management practices of the levee to mitigate the potential for significant degradation.

Approximately 1,200 linear ft (366 m) of levee at Chinook Landing Marine Park, in Fairview, Oregon, may be compromised by burrowing mammal excavations and intrusion of roots from woody plants. This levee is an essential piece of infrastructure, preventing inland flows of the Columbia River and protecting a variety of public and private structures, including Marine Drive.

The study area for this analysis consists of the entire 1,200-foot Chinook Landing Levee, including its crown, as well as the river-facing and land-facing portions of its prism (Appendix A, Figure 1).

The Salmon Creek Slough that runs along the toe of the Chinook Landing Levee hosts relatively high-value natural resources. It contains off-channel habitat on the Columbia River, which is relatively rare and important to native aquatic species. The natural resources in this area are managed by Oregon Metro, who is a close partner with MCDD for implementing natural function into flood protection obligations. It is for these reasons that maintaining natural features along the Levee are of interest as long as flood protection is not fundamentally compromised.

A diversity of mammals that excavate into the substrate are found in our region and can be generally classified as those who primarily live in either aquatic or terrestrial environments.

Aquatic mammals excavate burrows at or near the waterline during stable flow periods, typically in spring and fall (DWR 2022). In contrast, terrestrial mammals typically excavate elsewhere on the levee prism, including both the river-facing and land-facing slopes, as well as on the crown.

These mammal types additionally vary in their extent of excavation with some generally intruding only a short distance or creating small diameter excavations, while others can intrude several feet and/or create wide excavations (Ordeñana 2012). Surface erosion by the downslope biogenic movement of soil ejected from the animal burrows can also eliminate plant cover that provides erosion protection, and structural integrity losses occur when a large number of burrows or dens collapse, generating a localized loss of freeboard or flood protection elevation, or slope instability in the vicinity of the collapse (FEMA 2005).

Plant roots of certain woody tree species are known to be able to penetrate levees and facilitate piping of water. Further, the presences of large trees with large root systems creates potential for wide and deep cratering within the levee if the tree gets toppled by storm events and/or disease.

Woody vegetation has been documented to affect soil permeability on both sides of a levee and reduce the distance between the phreatic line and the inner slope of the levee (URS 2011). This contributes to water seeping out on the protected side of the levee, causing it or its foundation to lose soil particles, eventually deforming and weakening the levee structure and causing breaching. Dead tree roots decay, leaving defects in the levee and may even attract burrowing mammals (Cobos-Roa 2015). Woody vegetation also adds weight to the slope and may transfer the force of wind to the slip plane, increasing or decreasing the ability to withstand downward force along the slip plane (Shields Jr 1992; Tuel 2017). In general, the diameter of a tree’s root-wad is roughly equivalent to the diameter of its crown, resulting in mature trees of large species having the greatest potential to damage a levee they grow upon.

Burrowing mammal excavations and root intrusions can have varying impacts on the degradation of levee function (Diego et al. 2013; Saghaee et al. 2017; Central Valley Flood Protection Board 2019; Zhongming et al. 2021; Department of Water Resources 2022). The extent of this impact would be influenced by the diversity of burrowing mammals and woody trees present, as well as their distributions and densities across the levee. High density areas of burrows or trees would be of greater concern than low density areas, and larger burrows and trees would exacerbate the potential for degradation. Where the burrows and trees are vertically co-located relative to the surface water elevation, risk of failure is magnified.

Of particular concern for the Chinook Landing Levee are the density and size of burrows and large trees rooted between or near the seasonal low-water and high-water marks, as they are topographically positioned where soil saturation and stream flows can trigger erosional degradation through the piping of water into the levee through burrows and root linings, or by cratering in the event that a large tree topples, exposing the root-wad. However, any burrows or root craters at waterline, regardless of surface elevation, can pose a threat to the levee stability.

Surface erosion by the downslope biogenic movement of soil ejected from the animal burrows eliminates vegetal cover that provides erosion protection, and structural integrity losses occur when a large number of burrows or dens collapse, generating a localized loss of freeboard, flood protection elevation, and/or slope stability in the vicinity of the collapse (FEMA 2005). Mammal burrows in levee embankments tend to form tunnels, galleries, and dens that can form complex interconnected systems extending for hundreds of feet. These voids then pose a threat to the integrity of a levee in several ways, including those listed below (DWR 2009):

• The large voids can provide a preferential seepage path during high stages on adjacent water bodies, altering the saturation front through the embankment and potentially generating higher water tables and/or through seepage that triggers surface erosion of the landside levee slope. Higher phreatic surfaces represent higher pore pressures across the levee prism, that in turn reduce the effective stress of the soil and can contribute to slope stability failures.

• If the voids are large enough and several of them are connected, localized collapses of the embankment can occur and generate loss of protection elevation, leading to possible overtopping.

• If the embankment is composed of erodible materials, large burrow systems that extend from one side, deep into the levee can become conduits for piping and rapid erosion of the levee materials.

3.0 METHODS

3.1 Species of Interest

A list of burrowing mammals and woody-plant species with the potential to occur in the study area was generated based on distribution and species data from the West Multnomah Soil and Water Conservation District (2022) and the Oregon Department of Fish and Wildlife (ODFW) Burrowing Mammal species list (2022), as well as through coordination with MCDD and direct knowledge of the site (Table 1). This list of species provided two important functions: 1) it was used as a basis for the field survey to inform optimal search patterns (spatial and temporal), as well as prime field staff for accurate field identification, and 2) it provided an essential framework for incorporating the biological context of how these species can potentially contribute to degrading the Chinook Landing Levee. Each burrowing mammal species excavates relatively unique subterranean diggings – in both structure and size – with some larger species creating burrow systems that are particularly detrimental to levees.

Woody plant species also have differing potential for compromising levees, with large tree species (once mature) being the most susceptible to toppling while having the largest root-wads. In addition, some species are known to grow particularly quickly to a large size and have soft wood and/or diffuse root systems that make them susceptible to toppling (i.e., cottonwood). Surface water elevation relative to the location of burrows and woody plants is a significant influence on the potential for a levee to become compromised, and the wide-ranging tidal- and stream-flow-driven hydrology of this reach of the Columbia River, exacerbates this potential. From the perspective of levee performance, the importance of quantifying the effects of vegetation and animal activity on the hydrogeologic performance of embankments resides in the potential for internal erosion or piping due to concentrated flow along potential preferential seepage paths within the embankment as described by Abe and Ziemer (1991). Environmental Science Associates (ESA) reviewed historical data (1993-2006) and performance reports from levees in the United States focusing specifically on where negative impacts from burrowing mammal and large wood intrusion were documented (Bayoumi 2011; URS 2011; Van Vuren 2011). Sources of information were based on the literature review initiated by the USACE Engineering Research and Development Center (Cobos-Roa 2015) and USACE levee performance reports from 1993- 2009 (Cobos-Roa 2015), in addition to other applicable sources.

TABLE 1.

BURROWING MAMMAL AND WOODY-PLANT SPECIES WITH THE POTENTIAL TO OCCUR IN THE PROJECT AREA; INCLUDES

CHARACTERISTICS CONTRIBUTING TO LEVEE STRUCTURAL DAMAGE; SPECIES LISTED IN ALPHABETICAL ORDER BY

COMMON NAME

Common Name Characteristics Contributing to Levee Structural Damage

MAMMALS 1

American Beaver

Aquatic species. On large streams with flows too great to be dammed, lodges are built against a bank into which tunnels are dug. At low water levels, they work on lodges to ensure entrances remain under water.

They are known to sometimes dig burrows into the banks of ponds, rivers, and drainage ditches instead of building a lodge. Males generally dig burrows, which may be as long as 160 ft (49 m), but most are much shorter. Beavers are strong and expert burrowers, and they will create long large burrows with sizeable chambers to around 10 ft (3 m) in height. The impact of burrows can extend to 66-98 ft (20-30 m) from the riverbank.

By weakening the structure of the bank, the burrows can result in increased erosion. Bank collapse can result in a river channel that is wider than it should be in relation to the available flow of water, resulting in warm and shallow water that some species (such as trout and salmon) will not find hospitable. Bank erosion will also release higher levels of sediment into the river.

California Ground Squirrel Upland species. Burrows are constructed on slopes or other well-drained sites, sometimes in considerable density. Burrows are 5-200 ft (1.5-61 m) long and 3-6 in (8-15 cm) in diameter. They are >3 ft (0.9 m). in depth below soil surface and are often branched and have multiple chambers.

Muskrat Aquatic species. Burrows usually 4-6 in (10-15 cm) in diameter located 6-8 in (15-20 cm) below the surface of the water. They are 15-50 ft (4.6-15 m) long. They do not rely on only burrows for lodging.

Nutria Aquatic species. Burrow into banks and levees creating complex dens that span as far as 19.5 ft (6 m) deep-vertically, 164 ft (50 m) into the bank. This can cause severe streambank erosion, increased sedimentation, levee failures, and roadbed collapses.

Opossum Upland species. Use burrows of other mammals of the same size or larger.

Camas pocket gopher

Endemic to the Willamette Valley, its geographic range matches almost the Bretz Flood of 13,000 years ago. They are associated with early seral plant communities or cultivated croplands that mimic early seral situations such as weedy lawns and waste-ground areas. In most occupied sites, soils are heavy clays.

The burrow system can cover from 200-2,000 square-ft with the main runway situated parallel to the surface and about 6-18 in (15-46 cm) below. Nesting and food storage areas will be located as deep as 6 ft (1.8 m).

Shrew Mole Upland species. Construct networks of runways about 1-2 in (2.5-5 cm) deep in the soft soil beneath duff.

Burrows are dug as deep as 12 in (31 cm).

Townsend Mole Upland species. Digs shallow foraging tunnels as well as deeper permanent tunnels. May vacate flooded area and return to home area when waters recede. A spherical breeding nest cavity is excavated 7.5-20 in (19-51 cm) below ground level. A network of tunnels may exist but are superficial.

WOODY PLANTS 2

Black Cottonwood Pioneer species specifically adapted to colonizing and dominating areas disturbed by floodwaters. It is generally perpetuated in such environments by recurring floods. Mature cottonwoods attain heights of 125- 150 ft (38-46 m) (225 ft [66 m] maximum) and diameters at breast height of 48 to 60 in (122-152 cm) (108 in [274 cm] maximum). Cottonwoods mature as early as 60 years.

Big-Leaf Maple Has a shallow but wide-spreading root system. Roots are associated with vesicular-arbuscular mycorrhizae. Damaging agents include wind and snow.

Blue Elderberry Grows up to 20-30 ft (6-9 m) tall and wide in canopy. Spreads by root suckers to form colonies.

Cascara A deciduous tree that can attain a mature height of 30 ft (9 m) and canopy width of 15 ft (9 m) It has a shallow root system with many lateral branches.

Oregon Ash

Develop short, narrow crowns with small branches on long, straight boles. Open-grown trees have broad, rounded crowns with large limbs on short boles. The root system of ash is moderately shallow, wide-spreading, and densely fibrous. Mature trees attain heights of 60-80 ft (18-24 m) and DBH of 16-30 in (41- 76 cm). They may live at least 250 years.

Common Name Characteristics Contributing to Levee Structural Damage

Pacific Willow Typically found at low- to mid-elevations next to water bodies. A high-water table will encourage excellent growth for Pacific willow. It has a very resistant, hard root system that in some cases develops out of the ground. Grow to a height of 50 ft (15 m) and a crown width of 30 ft (9 m).

Red-Osier Dogwood

Fast growing short shrub with moderate to high suckering and has large, strong fibrous roots. While it does not tolerate drought, it is remarkably adaptable to a range of moisture regimes and soil types: from the banks of water bodies (where it tolerates annual flooding) to upland areas in full sun or almost complete shade.

1 Verts and Carraway 1998; Eder 2002; ODFW 2022 2 Oregon State University 2022

3.2 Field Survey Methods

To support effective levee design through assessing the extent of mammal burrowing and root intrusion in the levee, ESA conducted targeted field surveys within the study area to document existing conditions and gain a quantitative understanding of scope of degradation. This was performed by identifying and inventorying mammalian species responsible for observed diggings, and well as documenting all trees growing on the levee. The pre-survey assessment of burrowing species potentially at the levee was completed with a desk-top review of species in our region (Table 1), as well as acquiring information about their burrowing activities and characteristics.

This information provided species-specific details such as where burrowing typically occurs, as well as typical burrow structure (length, width, orientation, branching pattern, etc.). A parallel analysis was focused on woody plant species known to grow on levees in the region, and their sub-surface growth characteristics (root/root-wad density, depth, and size by age class). To maximize the utility of this data, two surveys were performed – one during the first seasonal low-water period in late spring and one during the second seasonal low-water period in the fall – both accounting for tidal influence. These surveys occurred from 16-17 May, 2022, and 26-27 September, 2022, respectively.

The ESA field biologists walked a systematic meandering transect across the entire study area, visually scanning the ground, floodplain, and banks for burrows and other physical signs of burrowing mammals. When a burrow was encountered, it was geospatially documented and the species was determined, along with the active status of the diggings and any physical evidence of recent mammal activity. The biologists counted all entrances, measured their diameters, and recorded other pertinent characteristics to aid in identifying the species that excavated the burrow.

Physical parameters and dimensions of burrows were evaluated in the field and documented to the nearest 0.1 ft. Burrows were documented using two parameter classes: width and depth, and three activity classes: ongoing, recent, and legacy, which correspond with California standard methods (Tuel 2017).

While walking the transect, all trees encountered were identified and documented. Included in the collected data was tree height, canopy width, and diameter at breast height (DBH), which were all measured or estimated and recorded to the nearest foot. All trees and burrows were geospatially documented at a spatial tolerance of less than three inches (horizontally and vertically), and georectified photos were taken of most burrows and trees documented. Other pertinent information and data was collected, including sign of infrastructure deterioration and sign of tree stress or structural problems. Burrowing mammal and woody-plant species occurrences were documented using a combination of point data and centerlines mapped in the field as well as “cluster” polygons of high-densities of burrows that were digitized in the office based on field observations. Each burrowing mammal and woody-plant species occurrence was assigned a unique identification (ID). Esri ArcGIS was used to prepare the maps depicting the survey area and Field Maps was used in the field to collect data.

Surface water elevations of the Columbia River at the time of the spring field survey, were unseasonably high due to extensive dam releases (ranging from a gage height of +9.36 ft to +10.21 ft) accommodating extensive runoff. Although the survey was strategically scheduled during low tide, the high flows caused the surface water elevation to obscure all areas below the high-water mark (MHWM) of 6.56 ft (Columbia River Datum; 9440083 Vancouver, WA = -0.050 m [-1.969 in]) (i.e., the mark established by the combination of high tide and high stream flow) (NOAA 2022). As a result, surveying for burrows of aquatic mammal species was largely not possible during spring. The ESA biologists returned in September to perform the fall survey during late-summer/early fall low-flows, which ranged from a gauge height of 0.3 ft to 2 ft, which allowed the biologists to access all areas below the top-of-bank to visually assess and document burrows and other sign of aquatic mammal activity below the mean high-water mark.

3.3 Geospatial Analysis

Incorporating guidance from other jurisdictions, this quantitative approach is intended to provide a large-scale levee structural analysis that documents potential impacts to the Chinook Landing Levee from animal burrowing and root penetration of woody plants. The ESA team applied this novel approach to document the location, type, and hazard significance of animal burrows and tree root-wads on the levee. They then overlaid those findings on a three-dimensional lidar-based spatial model of the levee and drainage system it fortifies against to provide a clear spatial representation of current conditions. The model is intended to fully support and inform alternative analysis for repair options appropriate for informing engineering design.

The modeling exercise produced two types of figures: aerial-view figures and cross-section figures. Aerial view figures show the horizontal distribution of burrows and root-wads across the backdrop of current vegetation conditions and topography. The cross-section figures show the vertical distribution and interaction between burrows, root-wads, and topography at three locations chosen because of the extent of burrowing activity and tree growth observed. The aerial view figures are intended to convey the extent and distribution of burrows and trees across the entire levee system, while the cross-sections are intended to show how the important physical aspects of the levee interact, and form a potential structural threat, at three specific locations that are important for alternative decision making.

The production of the cross-section figures was inherently more complex than the aerial view figures because they required spatial information that was not readily observable in the field. The levee profiles at the three selected locations were derived from the most recent LIDAR available (DOGAMI 2014), specifically to depict an accurate profile. On each profile, burrows and root-wads were accurately configured in their true location across the levee surface and shown in actual size relative to the scale of the levee (based on field measurements). The subsurface dimensions of burrows and root-wads were determined using available data and drawn on the cross-sections in a manner depicting accurate diameter, length, depth, and position/location. For example, the published “typical” burrow length for an adult male beaver (females don’t burrow) in tidally influenced systems is 98 ft (30 m) long (Dugan 2015; Palladino 2022), and therefore, all beaver burrows shown on the cross-sections are depicted to extend 98 ft (30 m) into the levee prism. For deciduous trees, it is well established that root-wads extend out laterally at least to the dripline of the tree canopy (AAIS 1995), therefore, all root-wads were depicted at the same diameter as the tree crown. Root depth for mature deciduous trees has been shown to average 5 ft (1.5 m) (AAIS 1995), therefore root-wads were depicted to extend into the substrate to that depth. During the field survey, opportunities were sought to ground-truth this approach and the projected spatial dimensions, and in all cases, proved accurate.

4.0 FINDINGS AND RESULTS

Burrowing Mammals

The survey of Chinook Landing Levee documented a total of 72 burrows attributable to three taxa of burrowing mammals: American beaver (N = 29), American shrew mole (Neurotrichus gibbsii) (N = 42), nutria (Myocastor coypus) (N = 4), and camas pocket gopher (Thomomys bulbivorus) (N = 1) (Appendix A, Figure 2). American beavers and shrew moles both were widely distributed along the levee structure, but only American beaver burrows were observed below the 6.56 foot MHWM (Columbia River Datum; 9440083 Vancouver, WA = -0.050 m [-1.969 in]) (NOAA 2022).

Burrows of all species were excavated in soil or beneath tree roots, and many American beaver burrows were observed under and through tree root systems as though the roots provided a “roof” for the burrows (Appendix B, Photos 1, 3, 4, 5, 7, 11).

The beaver burrow systems observed have caused the internal levee structure to become hollowed, and as such, much of the waterward facing bank has become eroded. The openings of many burrows were large enough to allow limited human entry, and on examination, it was confirmed that all extended deep into the levee. Burrowing was observed vertically from just below the MHWM (Columbia River Datum; 9440083 Vancouver, WA = -0.050 m [-1.969 in]), down to the thalweg of Salmon Creek, with those extending to the thalweg resulting in increased scour incision surrounding the burrow and the creation of the lowest elevation water entrances in the study area (Appendix B, Photos 2, 3, 10, 11, 12). This poses a significant threat to not only the levee structure but also water quality and fish habitat.

Beaver burrows were observed throughout the majority of the waterward facing bank, but tended to be clustered, resulting in several large burrow openings concentrated in discrete areas (Appendix B, Photos 10, 11, 12). This distribution appears to increase the potential for failure, as each concentration of burrows focuses degrading characteristics at each location, as opposed to a more even distribution, which would spread out degrading characteristics across the levee.

These burrow clusters form the highest risk locations for levee failure, and result in heavy erosion (Appendix B, Photo 5). The most severe instance was observed at section 1 on the northwest end of the levee where burrowing activity caused approximately 20 linear ft (6 m) of heavy bank erosion, exposing the root-wad of a large cottonwood and causing it to topple into the Creek, leaving a rootward crater 50 ft (15 m) in diameter and 5 ft (1.5 m) in depth just above MHWM (Columbia River Datum; 9440083 Vancouver, WA = -0.050 m [-1.969 in]) (Appendix A, Figures 2, 4; Appendix B, Photos 7, 8, 14). The potential for this large cottonwood to topple was likely to have been significantly increased by an English ivy (Hedera helix) infestation in its crown, making it top-heavy and increasing its surface area for wind impacts (see Findings and Results section, Woody Plants, for further discussion). Burrow entrances were present at approximately 4 ft (1 m) below the MHWM (Columbia River Datum; 9440083 Vancouver, WA = -0.050 m [-1.969 in]) and have caused large amounts of bank below the MHWM (Columbia River Datum; 9440083 Vancouver, WA = -0.050 m [-1.969 in]) to become hollow and erode. Surface erosion, at this location, by the downslope biogenic movement of the soil has generated a regular loss of flood protection elevation and slope instability in the vicinity of the collapse. Most of the bank material present in the vicinity of these burrows is compacted mud that appears moderately erodible (Appendix B, Photos 11, 12). The two other cross-sections: section 2 (Appendix A, Figures 2, 5;

Appendix B, Photo 10) and section 3 (Appendix A, Figures 2, 6), show similar characteristics with a system of large burrows underling large tree roots, thus having similar failure and damage potential. Additional areas where this potential exists also occurs across the levee face.

Mole burrows were observed in the highest density on the landward-facing side of the levee and were generally small in size and diffuse in distribution (Appendix A, Figure 2; Appendix B, Photo 6). Of the mole burrows observed and documented in the study area, none pose a significant threat to the structure of the levee because of their small size and location above the MHWM (Columbia River Datum; 9440083 Vancouver, WA = -0.050 m [-1.969 in]).

Habitat characteristics surrounding burrow entrances varied among species. Beaver and gopher activity was mostly observed in confined segments of the study area, and exclusively on the waterward facing side of the levee. Beaver burrows were strongly associated where cottonwoods were present and the levee bank was within 32 ft (10 m) of the thalweg. Beaver activity halted in areas where the floodplain was wide and the levee bank was beyond 32 ft (10 m) from the thalweg. Mole burrows were widely distributed and mostly found on the landward facing side of the levee in open grassland on the upper and lower slopes where large tree root systems were present. Interestingly, riparian areas that mostly consisted of woody shrubs such as red osier dogwood, willow, and cascara, and had little to no cottonwood, did not have burrowing activity by either species. Recent chew marks on surrounding woody vegetation and freshly chewed bark were observed directly next to several burrows (Appendix B, Photo 9) indicating that beaver use is ongoing.

Woody Plants

Several woody plant species with the potential to cause damage to Chinook Landing Levee were observed growing in mature states at high densities throughout the levee. A total of 92 woody plants were documented, including 45 black cottonwood (Populus trichocarpa), 32 Oregon ash (Fraxinus latifolia), 10 Pacific willow (Salix lucida), 4 cascara (Frangula purshiana), 2 red osier dogwood (Cornus sericea), and 1 big-leaf maple (Acer macrophyllum) (Appendix A, Figure 3).

The most common woody plants were large black cottonwoods that ranged in size from 1 to 6 ft (<1 to 2 m) DBH (Appendix B, Photo 15), with the most being 3 ft (1 m) DBH. These trees all had canopies larger than 20 ft (6 m) in diameter, with the average being around 30 ft (9 m).

Ivy was observed growing in high density in the northwest portion of the study area where it covered the surface of the levee in the understory and was growing on several large cottonwood trees (Appendix B, Photo 13). Ivy, given its nature of climbing trees and spreading into canopies, greatly accelerates the likelihood of trees toppling because of it weakening the tree, substantially increasing the top heaviness of the tree, and also increasing the surface area of the crown, making it susceptible to windthrow (Princeton 2022). Ivy, in this area, has clearly pressed its influence on some of the large cottonwoods on the levee. Observations of several mature trees with ivy in their crowns that have succumb and toppled, were made on the levee. Of these, one specific tree that appeared to have just toppled spring of 2022 was noted and documented (see tree 26 in Appendix A, Figures 2, 4) (Appendix B, Photo 14; Photo 13 shows standing tree infested with ivy). This tree had ivy growing into its crown, which was 50 ft (15 m) in diameter. Its rootward crater left in the levee prism was 50 ft (15 m) in diameter and over 5 ft (1.5 m) deep. This tree had been additionally disturbed by extensive beaver burrowing under its root-wad, likely undermining the tree.

The extensive distribution of mature cottonwood trees present on the waterward-facing side of the levee, together, equate to much of the levee surface being underlaying by root-wads – many of which would be larger than 30 ft (9 m) in diameter and extend into the levee an estimated 5 ft (1.5 m) in depth. Although not all these trees are susceptible to toppling in the near future, even if a subset topples – such as those currently infested by ivy – they could damage the levee significantly.

5.0 CONCLUSIONS

This report summarizes findings intended to inform effective levee design through the assessment of mammal burrowing and root intrusion in the Chinook Landing Levee. Although four species of burrowing mammals were documented having excavations on the levee, only the American beaver has created burrows large enough, and positioned appropriately, to cause potential for substantial damage to the levee. Coupled with the large cottonwoods growing across the waterward face of the levee, this potential has increased substantially.

Large trees with thick, integrated, root systems are present throughout most all the levee structure and were observed to have caused extensive erosion from their root systems being pulled out of the levee surface when toppled. This potential is substantially increased when these trees are underling with large beaver tunnels, which are common. Ivy was observed growing in high density in the northwest portion of the study area where it carpeted the understory and extended up into the canopies of several large cottonwoods. Allowing trees to grow to maturity on the levee has the potential to create structural problems when they topple but allowing ivy to expand into their crowns substantially increases the likelihood of toppling occurring. Important to consider, however, is whether the erosive process of beaver tunneling and root-wad cratering – its collective scope and location – is sufficient to structurally damage the levee directly, or facilitate its structural damage indirectly, by making it susceptible to high flows.

Examples of levee failures attributable to the activity of animal burrows are numerous (Bayoumi 2011; Van Vuren 2011), and specifically have shown animal burrows increasing the vulnerability of levees through concentrated erosion (Menduni 2021). By weakening the structure of the bank, the burrows can hasten erosion, causing even overbuilt levees to collapse when exposed to flows below their maximum design capacity (Menduni 2021; Wild Trout Trust 2022). Given the dimensions of the Chinook Landing Levee and the horizontal location of the MHWM (Columbia River Datum; 9440083 Vancouver, WA = -0.050 m [-1.969 in]), a “typical” beaver burrow would likely extend into the prism 98 ft (30 m), placing the tunnel and den (assumed to be 10 ft [3 m] in height) (Verts 1998) below Marine Drive. The relatively large diameters of the beaver burrows, their propensity to be clustered in discrete areas, and the undermining of large cottonwoods that overlay them, all suggest the potential damage to the levee could extend deep into the prism.

The immediate environmental outcome of a levee collapse is the release of sediment into the adjoining river (USACE 2011), but it will also allow for natural processes to return, including the creation of meanders and side channels. Beaver activity and input of large wood from upslope sources would create a complex mosaic of habitats that are good for natural biodiversity.

However, rivers have been confined for viable reasons, such as for flood control (Ceccato 2022), which is of paramount importance for the Chinook Landing Levee because of historic events and current protection it provides to greater Portland.

Several standards for levee burrow deterrent designs are summarized in FEMA 2020, including grouting, mud filling with stable soil types (i.e., sand, clay, etc.) like that of the surrounding embankment, and sheet piling. A sheet pile barrier constructed in the levee to prevent intrusive burrowing is a practical solution and is a widely used standard in levee management (Palladino 2020). This method is well described in Appendix A of Impact of Animal Burrows on Earthen Levee Body Vulnerability to Seepage (Palladino 2020). The depth that the sheet piles would be driven would be largely determined by the depth of mammal burrowing at the toe of the levee on the waterward side, which would be roughly 12 ft (4 m) in elevation, mean sea level.

Potential solutions have been presented in the DRAFT IGA6: MCDD Animal Burrow Remediation on Salmon Creek document developed for Levee Ready Columbia (Metro and Multnomah County Drainage District 2021). Of those that pertain to mitigating animal burrowing, installing a sheet pile wall that will prevent burrows from penetrating deeply into the levee would be preferred.

Current burrowing penetration is likely deep and may reach to under Marine Drive, and removing animals would only result in a temporary reduction of burrowing and natural processes. The sheet piles would lead to a long-term solution with little maintenance costs needed. Being cognizant of impacting trees through root pruning would be important for minimizing tree mortality and toppling after the sheet piles are installed.

Root pruning may be needed to install sheet piles into the levee for burrowing mammal protection.

However, to prevent damaging or killing trees growing along the levee, guidelines have been developed to minimize this risk. If these guidelines are not able to be followed because of trees growing too close to the site of sheet pile placement, removal of the trees should be considered to prevent their weakening and toppling:

• Cut tree roots outside of the tree’s drip line only. This is the area outside the range of the outer branches, above.

• Do not cut roots larger than 2 in (5 cm) wide. Larger roots are primary roots that that carry all the nutrients gathered by the smaller roots into the tree and provide the structural support of the tree.

• Make clean cuts – ragged cuts or cuts that peel back the outer layer of the roots takes longer for the tree to heal. Rough cuts also invite pests and bacteria into the tree’s weak inner layers.

• Prune when the tree is dormant. Late winter and early spring are the best times to prune any part of the tree because the tree isn’t actively taking up nutrients or producing energy.

Pruning during the growing season will stress the tree by impairing its ability to grow when it’s most active.

• Give the tree plenty of time to recover from root pruning. The tree will grow new fibrous root mass in 3 to 4 months, but it’s best to give it up to two years after root pruning before doing any other major root work.

• Don’t remove more than 25% of the tree’s roots at once. Any more than this will stress or kill the tree.

• Take good care of cut roots. After cutting roots, be sure to water the area very well to prevent them from drying out. You can also cover them with mulch or burlap and soak the topping well.

• Watch and protect against pests that attack a stressed or exposed tree. Apply organic compost around the tree to encourage new fibrous root growth and help it maintain its health while stressed.

Most of the woody plant species located on the levee are large, mature trees. However, such trees are tall and have large root-wads, posing a significant threat to Chinook Landing Levee by way of piping of water and/or toppling. Most of the trees documented are considered Facultative Wet (FACW) species that grow in close association with wetland or stream hydrology. It is for this reason that the zone between the low-water mark and high-water mark will always attract the native tree species documented during the survey, highlighting that they will pose a continuous management issue. Current guidelines for management of vegetation on earthen levees calls for the removal of woody vegetation, especially trees, and their replacement with grassland vegetation (Palladino 2020). Trees and other woody vegetation, such as shrubs and vines, can create both structural and seepage instabilities, prevent adequate inspection, and create obstacles to maintenance and flood-fighting/flood-control activities. Further, controlling the establishment of ivy infestations in tree canopies would further aid in reducing the potential for tree toppling. Vegetation must be controlled for the following reasons (from Palladino 2020):

• To allow proper inspection, surveillance, and monitoring of all structures and adjacent areas for seepage, cracking, sinkholes, settlement, displacement, and other signs of distress.

• To allow access for normal and emergency operations and maintenance (O&M) activities.

• To prevent root-related damage to structures, such as shortened seepage paths through embankments and/or foundations; voids in embankments and/or foundations due to decayed roots or fallen trees; clogged seepage collector systems; and expansion of cracks or joints in concrete walls, spillway floors, and canal linings.

• To limit those habitat characteristics that encourage the creation of animal burrows.

• To allow full design-discharge capability of waterways, spillway inlet and outlet channels, outlet-works discharge channels, and other open conveyance channels.

• To avoid any incidental growth and subsequent presence of endangered species that might prohibit activities necessary for O&M, or access.

Dense groundcover and thick trees and shrubs not only hinder inspections but can also obscure indicators of potential performance problems such as animal burrows, settlement, depressions, cracks, and similar issues. If vegetation is thick, animal burrows can go undiscovered and proper animal intrusion mitigation may not occur (FEMA 2020). This same dense vegetation, however, is important for natural process including providing wildlife habitat and surface water shading, which are uncommon in this reach of the Columbia River.

The three cross-sections developed from the spatial data collected during this study consist of three main spatial components: 1) the area of impact in relation to the MHWM (Columbia River Datum; 9440083 Vancouver, WA = -0.050 m [-1.969 in]), 2) the toe of levee embankment, and 3) the channel thalweg. Taken together, the context they provide demonstrates the current interior state of Chinook Landing Levee structure. The “hot spots” these cross-sections represent are heavily impacted by mammal burrows and large tree root systems, which each pose their own threats to the integrity of the structure, but when combined, have created a mosaic of hollow systems and tunnels throughout the levee, resulting in extensive bank erosion and surface cave-ins. Because these tunnel systems are all below the MHWM (Columbia River Datum; 9440083 Vancouver, WA = -0.050 m [-1.969 in]), they likely allow water to enter the levee prism, creating a “sand-boil” like effect throughout the interior of the structure, making it increasingly susceptible to damage and failure (Shiro 2014). Large tree root systems embedded within the levee structure create a problem over time because of piping potential, but when the trees are infested with ivy, they can become an immediate problem by increasing the likelihood of toppling and creating a root crater. Because these degrading sources co-occur in space, they magnify the potential for damage as well as its severity.

Landscape plantings to existing flood damage reduction systems, such as the Chinook Landing Levee, must comply with the appropriate O&M manual. New plantings may not be approved without an appropriately detailed and documented engineering evaluation to ensure that design intent and safety criteria are maintained as originally authorized. Levees must have effective and reliable erosion protection, and as a mitigation effort, the use of suitable vegetation, such as shrub forms placed riverward of the vegetation-free zone, is encouraged as an environmentally beneficial means to moderate the erosive potential of water currents and wave action while also hastening the recruitment of large tree species (Shiro 2014), such as cottonwoods.

6.0 CONTRIBUTING SCIENTISTS

Jeff Barna, MS, PWS – Senior Ecologist, Mammal and Tree Specialist

Mike Higgins, BS – Ecologist / Aquatic Scientist

Kevin Coulton, PE, CFM – Principal Associate / Senior Engineer

7.0 REFERENCES

Abe, K., & Ziemer, R. R. 1991. Effect of tree roots on a shear zone: modeling reinforced shear stress. Canadian Journal of Forest Research, 21(7), 1012-1019.

Arboricultural Advisory and Information Service (AAIS). 1995. Arboriculture Research and Information Note; Tree Root Systems. Published by Arboricultural Advisory and Information Service, Alice Holt Lodge, Surry. July 1995. Available:

https://www.trees.org.uk/Trees.org.uk/files/61/6181f2b7-e35d-4075-832f- 5e230d16aa9e.pdf.

Bayoumi, A.; Meguid, M.A. 2011. Wildlife and Safety of Earthen Structures: A Review. J. Fail.

Prev. 2011, 11, 295–319.

Ceccato, F., Malvestio, S., & Simonini, P. 2022. Effect of Animal Burrows on the Vulnerability of Levees to Concentrated Erosion. Water, 14(18), 2777.

Central Valley Flood Protection Board. 2019. State of California Central Valley Flood Protection Board California Code of Regulations, Title 23. Waters. Division 1. Central Valley Flood Protection Board Chapter 1. Organization, Powers, and Standards Article 2. Definitions.

Cobos-Roa, D. A. 2015. Transient Seepage Through Levees and the Influence of Roots and Animal Burrows. University of California, Berkeley.

Department of Water Resources. 2022. Proposed Mitigated Negative Declaration and Initial Study for the Delta Dams Rodent Burrow Remediation Project. State of California – California Natural Resources Agency.

Diego, R., M. Shriro, N. Sitar, and J. Bray. 2013. Influence of Tree Roots and Mammal Burrowing Activity on Levee Performance: Volume 1. – Review of Literature and Case Histories.

10.13140/RG.2.2.21036.26240.

Dugan, S. 2015. Report of the Scottish Beaver Salmonid Working Group.

Eder, T. 2002. Mammals of Washington and Oregon. Lone Pine Publishing. Edmonton, AB, Canada and Renton, WA, United States.

Menduni, G.; Cocchi, R.; Manselli, L.; Simonini, P. Commissione Tecnico-Scientifica per La Valutazione Delle Cause All’origine Della Rotta Arginale Lungo Il Fiume Panaro in Località Gaggio Di Castelfranco Emilia-Relazione Di Dettaglio. 2021. Int. J. Disaster Risk Reduct. 2021, 74, 102926.

Metro and Multnomah County Drainage District. 2021. DRAFT IGA6: MCDD Animal Burrow Remediation on Salmon Creek. Project Overview/ Business Case. Levee Ready Columbia.

19 April 2021.

Municipality of Princeton New Jersey (Princeton). 2022. Negative Effects of English Ivy.

Government Website. Accessed: 18 November 2022. Available:

https://www.princetonnj.gov/1151/Negative-Effects-of-English-Ivy.

National Oceanic and Atmospheric Administration (NOAA). 2022. NOAA Tides & Currents.

Vancouver Station. Water Levels. Accessed: 2 October 2022.

Available.https://tidesandcurrents.noaa.gov/waterlevels.html?id=9440083&units=standard &bdate=20121001&edate=20221002&timezone=GMT&datum=STND&interval=m&action= data

Ordeñana, M. A., Van Vuren, D. H., & Draper, J. P. 2012. Habitat associations of California ground squirrels and Botta's pocket gophers on levees in California. The Journal of Wildlife Management, 76(8), 1712-1717.

Oregon Department of Fish and Wildlife (ODFW). 2022. Burrowing Mammal Species List.

Accessed: 8 May 2022. Available: myodfw.com.

Oregon State University. 2022. Oregon Wood Innovation Center (website). College of Forestry, Corvallis, OR. Accessed: 20 June 2022. Available: https://owic.oregonstate.edu/homepage.

Palladino, M. R., Barbetta, S., Camici, S., Claps, P., & Moramarco, T. 2020. Impact of animal burrows on earthen levee body vulnerability to seepage. Journal of Flood Risk Management, 13, e12559.

Saghaee, G., A. Mousa, and M. Meguid. 2017. Plausible Failure Mechanisms of Wildlife-damaged Earth Levees: Insights from Centrifuge Modeling and Numerical Analysis. Canadian geotechnical journal, 54(10), 1496-1508.

Shields Jr, F. D., & Gray, D. H. 1992. Effects of woody vegetation on sandy levee integrity. Water resources bulletin (USA).

Shriro, M., Cobos-Roa, D., Bray, J., & Sitar, N. 2014. Effects of woody vegetation on levee integrity during full scale testing at twitchell island in Rio Vista, California. In Association of State Dam Safety Officials Annual Conference 2014, Dam Safety 2014 (Vol. 2, pp. 651- 663).

Tuel, A. L. B. 2017. Levee Vegetation Management in California: An Overview of Law, Policy and Science, and Recommendations for Addressing Vegetation Management Challenges.

Environs: Envtl. L. & Pol'y J., 41, 369.

Van Vuren, D. H., & Ordeñana, M. A. 2011. California Levee Vegetation Research Program.

Verts, B.J. and L.N. Carraway. 1998. Land Mammals of Oregon. University of California Press, Ltd. Berkeley and Los Angeles, California and London, England.

https://www.princetonnj.gov/1151/Negative-Effects-of-English-Ivy https://tidesandcurrents.noaa.gov/waterlevels.html?id=9440083&units=standard&bdate=20121001&edate=20221002&timezone=GMT&datum=STND&interval=m&action=data

West Multnomah Soil and Water Conservation District. 2022.