Geomorphic Assessment of Thirty Miles of Railroad Infrastructure Along the Klickitat River and Swale Creek, Klickitat County, WA Preliminary Report

Home , Klickitat River

Geomorphic Assessment of Thirty Miles of Railroad Infrastructure along the Klickitat River and Swale Creek, Klickitat County, WA Preliminary Report

Prepared by:

Will Conley, Hydrologist/Geomorphologist Yakama Nation Fisheries Program Klickitat Field Office Wahkiacus, WA

Prepared for:

United States Department of Energy Washington State Recreation and Bonneville Power Administration Conservation Office Environment, Fish, and Wildlife Program Salmon Recovery Funding Board Portland, Oregon Olympia, WA Project Number: 1997-056-00 Project Number: 10-1741

May 31, 2015

THIS PAGE INTENTIONALLY LEFT BLANK

ACKNOWLEDGEMENTS

Project funding was provided by Bonneville Power Administration project number 1997-056-00 (Klickitat Watershed Enhancement Project) and Salmon Recovery Funding Board project number 10-1741. David Lindley (Habitat Biologist, YNFP) provided valuable assistance inventorying crossing structures and reviewing report drafts. The success of the field-based portion of this study was greatly assisted by coordination with Andrew Kallinen, Ranger, Columbia Hills State Park Complex.

THIS PAGE INTENTIONALLY LEFT BLANK

TABLE OF CONTENTS

INTRODUCTION ...... 7

Study Area and Geographic Scope ...... 8 Fisheries Significance ...... 9 Study Purpose ...... 10

METHODS ...... 10

Report Conventions ...... 10 Remote-Sensing / GIS...... 11 High Resolution Topography, Aerial Photos, and Derived Products ...... 12 Aerial Photograph Interpretation ...... 12 Field Methods ...... 12

RESULTS AND DISCUSSION ...... 14

Features and Indicators ...... 17 Channelization / Increased Confinement ...... 18 Stream Crossings ...... 30 Wash-outs ...... 32 Hillslope Processes ...... 34 Runoff ...... 35 Vegetation Management ...... 36 Existing Railbed Alignment Modifications ...... 38 Deliberate Alterations ...... 38 Natural Alterations ...... 40

MANAGEMENT IMPLICATIONS ...... 44

Geographic Priorities ...... 44 Stream Restoration Enhancement Strategies ...... 45 Channelization and Confinement ...... 45 Water Crossings ...... 46 Wash-outs ...... 47 Hillslope Processes ...... 48 Runoff ...... 48 Vegetation Management ...... 49

CONCLUSION ...... 49

LITERATURE AND DATA CITED ...... 51

APPENDIX A Chronology of Railroad Embankments in the Klickitat Subbasin ...... A-1 APPENDIX B Klickitat Trail Stream Continuity Project: Phase 1 ...... B-1 APPENDIX C Glossary of Selected Terms ...... C-1

THIS PAGE INTENTIONALLY LEFT BLANK

INTRODUCTION A railroad embankment was constructed between Lyle and Goldendale, Washington in 1902 and 1903. The lower 30 miles runs through the Swale Creek and Klickitat River canyons. Both waterbodies experienced floodplain filling and grading, bank armoring, and channel realignment associated with railroad construction, re-construction (following various flood events), and maintenance activities. A variety of effects that generally interrupt geomorphic processes and adversely affected fish and wildlife habitat persist.

The Klickitat River and Swale Creek provide habitat to a variety of native fish including steelhead (Oncorhynchus mykiss), the anadromous (ocean-going) form of rainbow trout which are ESA-listed as “Threatened”. Coho (Oncorhynchus kisutch) and Chinook (Oncorhynchus tshawytscha) salmon also inhabit the Klickitat River and have been observed using the lower portions of Swale Creek. Bridgelip suckers (Catostomus columbianus), an important “First Food” to the Yakama People, are also found in both the Klickitat River and Swale Creek.

Figure 1. Looking north into Swale Canyon from vicinity of Stacker Creek. The railway was abandoned and the former railroad holdings were “rail-banked” in 1992 to preserve the right-of-way for potential future railroad operation. A recreational trail was approved for interim use and the railway was decommissioned in 1993. Washington State Parks and Recreation Commission (WSPRC) became owner in 1994 and manages the corridor which is now known as the “Klickitat Trail”. Since 2003, WSPRC has co-managed the lower 14 miles (downstream of Suburbia) with the United States Forest Service (USFS). APPENDIX A has a chronology related to valley-bottom railroad embankments in the Klickitat River subbasin.

Over twenty-seven miles of the corridor lie at the bottom of two canyons that range from 200 to 1,500 feet deep (Figure 1). This dynamic landscape position involves interaction with a variety of geologic processes including: rockfall, hillslope movements, debris flows, and direct reworking by the Klickitat River, Swale Creek, and tributaries. This study was undertaken with the goal of identifying outcomes that decrease conflict between railroad infrastructure (e.g. embankment and structures) and historic actions (e.g. floodplain grading) with natural processes. Implementation of such outcomes will result in more favorable conditions for fish and wildlife as well as reduce maintenance frequency and cost.

Study Area and Geographic Scope The study area is the lower 30 miles of the former embankment of the Goldendale Branch of the Spokane, Portland, and Seattle (SP&S) railway in Klickitat County in south-central Washington State (Figure 2). The downstream end is located at SR14 in Lyle, WA. The upstream end is where WSPRC ownership terminates at Uecker Road, approximately 0.5 miles east of Warwick, WA and 4.5 miles WSW of Centerville, WA. The downstream portion of the study area parallels 16 miles of the mainstem Klickitat River from its confluence with the Columbia River to the mouth of Swale Creek at Wahkiacus, WA. From Wahkiacus to the upstream end of the study area, the trail parallels Swale Creek for an additional 14 miles.

Figure 2. Vicinity map of Klickitat Trail Geomorphic Assessment.

The study area is within the southern domain of the Yakima fold belt (Watters, 1989), a highly faulted zone whose prevailing topographic expression is a series of east-west trending anticlines. Basal geology is basaltic and associated with several different units of the Columbia River Basalt Group (WDNR 2013). Modeling of inundation elevations of late-Pleistocene glacial outburst floods indicates a maximum inundation elevation of approximately 980 feet at the mouth of the Klickitat River (Benito and O’Connor, 2003). Assuming a flat water surface profile, the upstream extent of inundation within the study area is approximately trail mile 21.7 (~5.6 miles up Swale Canyon).

Fisheries Significance The Klickitat River and Swale Creek provide habitat to a variety of native fish including steelhead (Oncorhynchus mykiss), the anadromous (ocean-going) form of rainbow trout which are ESA-listed as “Threatened”. Coho (Oncorhynchus kisutch) and Chinook (Oncorhynchus tshawytscha) salmon also inhabit the Klickitat River and have been observed using the lower portions of Swale Creek. Bridgelip suckers (Catostomus columbianus), an important “First Food” to the Yakama People, are also found in Swale Creek. Though not much is known about their population status, adults of spawner size have been observed. Speckled dace (Rhinichthys osculus) are observed throughout Swale Creek and upstream as far as a perennial reach at RM 9.4 (Conley, unpublished data).

The Klickitat River through the study area is a migration and rearing corridor for nearly 100% of all migratory fish in the Klickitat watershed. It has also accounted, on average, for 10% of observed basin wide steelhead spawning. Fill materials from railroad and highway embankments encroach on the active channel for at least (cumulatively) 9.8 miles of bank in the lower 20 miles of the Klickitat River and associated with lower channel complexity, riparian, and instream habitat quality than upstream reaches (Conley 2005). This artificial confinement compounds the habitat effects of increasing natural confinement as the river approaches Lyle Falls. Floodplain access is restricted by road and railroad embankments as well as localized levies. Some modern channel incision seems likely to have occurred but is currently unconfirmed as interpretation is complicated by incision into older slackwater and landslide deposits. Large woody debris (LWD) is much less abundant than in upstream reaches (Conley 2005), related to reduced potential for recruitment and retention, likely associated with channelization. While the most immediately noticeable effect of the embankments is their limitation on riparian cover, they also displace both waterbodies from former channel alignments and isolate floodplains and valley margins.

Swale Creek is a tributary of the Klickitat River and paralleled by the embankment for its lower 14 miles. In many areas the railroad embankment reduces stream channel width and accessible floodplain. Swale Creek provides both spawning and rearing habitat for steelhead. Habitat conditions are generally poor with pools being particularly limited. Swale Creek within the study area averages 7 pools per mile or about 4.2% of the habitat by channel length (YNFP,

unpublished data). The YNFP is currently conducting a multi-year evaluation of habitat and population status (via PIT tags) in Swale Creek.

Study Purpose The purpose of the project is to evaluate geomorphic effects of thirty miles of infrastructure associated with former SP&S railroad embankment between Lyle, WA and Warwick, WA. This study involved 1) remote sensing of valley topography and ground cover, 2) field delineation, mapping, and characterization of railbed segments, and 3) development of geographic priorities and conceptual enhancement recommendations to diminish adverse effects of the embankment on adjacent water bodies, 4) identify areas needing further investigation, and 5) development of a follow-up project.

This preliminary report provides a mostly descriptive overview of findings associated with field inventory including examples of features and effects. General recommendations are provided at a coarse geographic scale (miles to tens of miles) and one sample project is presented (APPENDIX B).

Analysis at finer spatial resolution, including integration with historic hydraulic analysis and fisheries habitat information is ongoing and will be presented subsequently in a more detailed technical report. This will facilitate a more robust prioritization of individual railroad embankment segments within a broader ecological context. Discussions regarding future management intent and potential for stream enhancement and restoration actions are ongoing between the YNFP and WSPRC.

Note: it is important to note that this report differentiates between physical infrastructure (e.g. the railbed) and a “trail” as a path or walkway. This investigation is focused on the former and makes no judgments about the concept of use of the corridor as a “trail” as such use has no inherent dependency on fill originally deposited and graded for the purposes of heavy industry.

METHODS Methods used for this report are categorized as remote sensing / GIS and field inventory. Conventions used throughout the document are also described.

Report Conventions While this report is of a technical origin, effort has been made to organize and illustrate content to be accessible to non-geomorphologists. A glossary is provided in APPENDIX C. Additional conventions used in this report are described below.

With regard to geomorphic features, relative references (e.g. “left” or “right”) used imply facing downstream/downvalley. For example, “left bank” (LB) is the stream bank appearing on the left when facing downstream. Similarly, “right valley toe” (RVT) is the interface of valley floor and valley wall appearing on the right when facing down-valley. 10

Use of the term “bankfull” is generally avoided due to ambiguities that have developed across a wide range of users and tendency to imply flow frequency and/or effective discharge correlations. Where the term is used in this report, it refers solely to a topographic breakpoint (e.g. top-of-bank) where a unit change in discharge begins to inundate proportionately more area and produces a diminishing incremental change in hydraulic forces. Use of “bankfull” in this report does not imply flow-frequency, stability, or some idealized channel condition.

The prefix “Q” followed by a number is used as short hand for flow event recurrence interval. Table 1 cross-references some common terminology and usage related to event frequency. For example, an event with a magnitude of 100-year recurrence (Q100) has a probability of 0.01 (= 1% chance) of being equaled or exceeded in any given year. A good discussion of recurrence intervals can be found at: http://water.usgs.gov/edu/100yearflood.html

Table 1. Cross-reference of event frequency notation and common usage.

Shorthand Probability Recurrence Percent Colloquial Name* Equaled or Interval Chance of Exceeded in a (years) Occurrence in Given Year Any Given Year Q1 0.99 1.01 99 yearly flood Q2 0.5 2 50 2-year flood Q10 0.1 10 10 10-year flood Q100 0.01 100 1 100-year flood * Colloquial usage is avoided in this report as it has misleading implications for real-world recurrence.

“Railbed” and “embankment” are used interchangeably. “Traveled way” refers specifically to the flattened portion of the embankment that carries traffic.

“River Mile” (RM) and “Mile Marker” (MM) indicate a relative horizontal location along a corridor from some starting point. Mile 0.0 for the river mile references are the railroad bridge at Lyle and mouth of Swale Creek for the Klickitat River and Swale Creek, respectively. Mile marker references are based on the 1:1,200 digitized embankment alignment with a 0.0 point at the “0” marker post at the Lyle trailhead (at transition from concrete to gravel surface).

Remote-Sensing / GIS All geospatial data were managed in a Geographic Information System (GIS). ArcGIS v10.0 (ESRI 2012) and ArcGIS v10.2.2 (ESRI 2014) with a full ArcInfo license, Spatial Analyst, and 3D Analyst, and GPSAnalyst extensions were the primary software used for data management, spatial analyses, and cartography. Geospatial data used in this study were projected in Washington State Plane South (Zone 5626), NAD83, Feet, GRS1980. Data using a vertical datum reference NAVD88. Field data collected in an ArcPad environment and post-processed (e.g. differentially corrected) with GPSAnalyst were managed in an ArcGIS v10.0 personal geodatabase environment. Once corrected and screened, field data, LiDAR data, and derived

data were managed in an ArcGIS v10.2.2 file geodatabase environment. Raster files were maintained in their native formats outside of the geodatabases.

High Resolution Topography, Aerial Photos, and Derived Products Aerial Light Detection and Ranging (LiDAR) and true-color photography were collected for the Klickitat River in April 2009 (WSI 2009) and Swale Creek in October and November 2011 (GeoDigital 2012). Average ground point density was 0.47 and 1.0 points/ft2 for the 2009 and 2011 data, respectively. Accuracy for both flights equaled or exceeded NMAS 1’ contour equivalent. Aerially-acquired deliverables for both flights included:  aircraft trajectories  bare-earth DEM  all-return point cloud  ground point list  model key points  intensity image  breakline vectors  orthophotos  area flown / tile index

Combined with field work, these data provided the foundation for much of this study. A variety of data products were derived from these data, including bare earth terrain model and embankment centerlines. Unless otherwise specified, the mapping scale for manually-digitized products presented in this report is 1:1,200.

Bare earth terrain models were built from both “key-point” and all ground points LAS files. In general, the key point terrain was used for most visualization and interpretation as the smaller number of nodes improved computer performance.

Embankment centerline was manually digitized using terrains, field GPS’d polylines, and current and historic photography.

Aerial Photograph Interpretation Aerial photography was used to review historic channel and vegetation patterns as well as railbed alignment. Available photography was of varying scales and resolutions and mostly unregistered (not georeferenced). High resolution (1.5 foot pixels) imagery was acquired in 2009 and 2011 associated with LiDAR collection. Historic aerial imagery downloaded from USGS (2015) has a variety of scales, resolution, and spatial extents. Imagery from 1996, 2003, 2006, and 2011 was georeferenced and orthocorrected. Post-2000 imagery was collected in true-color. Free imagery from 1947, 1952, 1955, 1973, 1981, 1990, and 1994 was also available, but generally only of medium resolution unsuitable for fine-scale interpretation. Georectification was outside the scope of the current project but higher resolution, large scale imagery from 1947, 1952, and 1973 was helpful for qualitative and descriptive review.

Field Methods Field inventory involved foot, bicycle, and ATV-based surveys and focused primarily on railbed/trail segments as well as drainage structures. Location and attribute data were collected

12 using a Floodlight-enabled Trimble GeoExplorer 6000XT GPS receiver running Windows Mobile 6.5.7 Professional, GNSS firmware v3.00.5, ArcPad (v10.0 R4), and GPSCorrect (v3.40).

All pre- and post-processing of data was conducted in ArcInfo (v10.0, service pack 4). Pre- processing consisted of creating a feature class within a personal geodatabase and related attribute fields. Domains were created for categorical and range fields as a means of quality control. Feature classes were checked-out of the personal geodatabase for field editing in ArcPad using the Trimble GPSAnalyst (v2.40) extension for ArcInfo. The check-out process generates drop-down boxes for fields with domains which reduces field data entry time and provides a means of quality control and quality assurance.

Field inventory of segments occurred during June and July 2012. In general, a minimum mapping unit of approximately 50 lineal feet was applied, though exceptions were made for smaller segments where isolated bedrock exposures were present. Segments were delineated based on six primary delineative criteria:

 embankment is in some stage of wash-out  presence of floodplain landward of embankment  bedrock presence immediately landward of embankment  bedrock projecting from toe of embankment  embankment contact with a stream channel  presence of cutslopes on both sides of the embankment

In addition to delineative criteria, attribution of segments included:  active channel contact  presence of water landward of embankment  presence of rockfall  estimated Q recurrence of floodplain encroachment  presence of LWD  % length natural hillslope projects beyond embankment  deposition on traveled way  % length of traveled way erosion  historic repair  % length of embankment face erosion  bedrock modification  % length that slope of embankment face exceeds 1:1  vegetative cover  % length that slope of embankment face is less than 2:1  seep/spring presence  estimated erosion recurrence

Water crossings were inventoried between Suburbia and Wahkiacus in June 2012. Remaining locations and attributes were collected, as well as verification of previously collected points, in the March of 2013. Attributes collected for drainage crossings included:  % inlet obstructed  nature of channel  structure span  % outlet obstructed  structure type  structure rise  mid-span channel contacts  structure material  structure shape

 deposition on traveled way  upstream deposition  structure damage  elevation relative to traveled way  upstream channel width  height of cover  scour pool presence  LWD retention  water present

RESULTS AND DISCUSSION Cumulative length of field-mapped segments was 30.66 miles compared to 30.34 miles for linework digitized from LiDAR topography and high-resolution photography. Greater length of field data is due primarily to additional length of detours, physical departures from true centerline caused by obstructions, and GPS accuracy. The difference of 0.32 miles was 1.1% of total length and considered negligible for the purposes of this report. The corridor was field- mapped as 509 segments, but results are summarized by seven major geographic units described in Table 2. Figure 3 notes and presents geographic context of unit endpoints.

Table 2. Major geographic units used to summarize findings in this report (Unit ID = downstream mile marker) Unit ID Name Description 0.0 Lyle Trailhead Railbed is ~100’ (vertical) on hillside above left-bank of Klickitat River; SR to Fisher Hill 142 is adjacent, immediately upslope Bridge 1.6 Fisher Hill Railbed is along right-bank of and 10-12’ (vertical) above Klickitat River for Bridge to Pitt most of its length. Downstream-most ~1 mile is upstream end of Lyle Gorge with continuous bedrock toe and railbed increasingly elevated above river in the downstream direction. Floodplain area ~0.9 miles long in vicinity of Logging Camp Creek and several others between 0.1 and 0.4 miles long. Partial washouts and fluvial sculpting in numerous places. SR 142 occupies opposite side of river. 9.9 Pitt to Klickitat Railbed is along right-bank of and 8-20’ (vertical) above Klickitat River; Wastewater except for 0.7 miles in vicinity of Skookum Flat where more extensive Plant floodplain and intermediate infrastructure is present. Several washouts of varying intensity, one of which embankment was completely removed. SR 142 occupies same side of river, upslope of railbed. 12.3 Klickitat Infrastructure (including town of Klickitat) continuously between railbed Wastewater and Klickitat River. No trestle where railbed crosses Snyder Creek flume. Plant to SR 142 is between railbed and river for most of its length. Suburbia 13.9 Suburbia to Railbed adjacent to and 6-12’ (vertical) along left-bank of the Klickitat Wahkiacus River, with several small (<0.2 mile long) floodplain patches, including one ~0.5 mile long. Several partial washouts and fluvial sculpting in numerous places. SR 142 and railbed occupy opposite sides of the river. 16.4 Wahkiacus to Swale Canyon. Moderate to high confinement. 53% of embankment (by Harms Rd length) contacts active channel. At least nine probable stream realignments from 1,700’ to 6,000’ long with extensive channelization from armoring and floodplain filling elsewhere. Railbed crosses Swale Creek five times over trestles, each having 7 to 12 mid-span abutments and multiple points of contact within the active channel. Most aggressive channelization and floodplain grading is between miles 20 and 27. 28.2 Harms Rd to Railbed along right bank (or valley margin) of Swale Creek. Moderate to Uecker Rd unconfined valley. Generally fine-textured stream banks. Active channel generally not in contact with embankment.

Figure 3. Overview map of study area, including mile indices and major landmarks. The major delineative criteria are fundamental to underlying fluvial processes, particularly during peak flow events. Of the cumulative six miles of landward bedrock (Table 3), approximately 85% also lacks a bedrock toe, effectively obscuring fluvial contact and diminished potential for energy dissipation and habitat formation (e.g. pool scour).

Table 3. Cumulative length (ft.) of several delineative criteria by major railbed unit.

Total Length (feet) Landward Landward Landward Unit ID Of Unit Bedrock Valley Floor Water Bedrock Toe 0.0 8,495 93 0 0 313 1.6 44,868 5,917 13,735 0 10,320 9.9 12,642 1,875 5,649 0 567 12.3 8,548 1,742 4,858 182 0 13.9 13,080 2,916 6,230 1,034 300 16.4 62,824 19,988 25,067 1,701 1,643 28.2 11,418 462 4,989 0 462 Total 161,875 32,993 60,529 2,917 13,606

While the Klickitat and Swale Creek valley bottoms are mostly moderately confined naturally, the embankment increases confinement by limiting floodplain access for more than one-third of its length (Table 3). The embankment contacts the active channel of either the Klickitat River or Swale Creek for 41% and the 100-year floodplain for 82% of the total length between Lyle and Uecker Road (Table 4). Roughly 63% of total embankment length encroaches into the valley bottom beyond what would be expected from adjacent undisturbed slopes (Table 5).

Table 4. Length (ft.) of embankment encroachment into estimated zones of recurrence frequency by major unit. Active Unit ID Channel Q5 Q10 Q25 Q50 Q100 >Q100 None 0.0 0 0 0 0 0 313 0 8,182 1.6 19,095 3,129 3,197 3,837 4,494 3,570 3,988 3,559 9.9 5,266 799 2,003 40 0 2,990 1,545 0 12.3 389 0 0 0 0 4,034 859 3,240 13.9 6,745 1,767 1,303 1,024 0 1,599 528 114 16.4 33,230 6,741 8,784 1,372 5,571 1,979 4,747 399 28.2 1,970 1,121 397 597 312 4,982 2,039 0 Total 66,696 13,558 15,683 6,871 10,377 19,466 13,705 15,493

Table 5. Total length (ft.) by hillslope projection category for major units. Smaller categories (e.g. 0-5%) indicate greater embankment encroachment.

Unit ID 0-5% 5-20% 21-40% 40-60% 60-80% 80-95% 95-100% 0.0 4,938 0 0 2,218 0 778 561 1.6 20,509 3,330 3,372 2,833 1,128 2,925 10,771 9.9 8,058 259 0 130 149 499 3,548 12.3 3,132 0 0 0 3,111 0 2,304 13.9 7,676 1,036 168 385 120 1,307 2,388 16.4 33,218 5,630 4,120 4,745 1,202 1,850 12,060 28.2 4,260 0 0 462 312 304 6,080 Total 81,792 10,254 7,660 10,774 6,021 7,661 37,711 Though over 25% (by length) of the embankment has some degree of face erosion (Table 6), it is largely intact (Table 7) with relatively little (<15%, cumulative) intrusion into the traveled way (Table 8). Table 6. Total length (ft.) by face erosion by category for major units.

Unit ID 0-5% 5-20% 21-40% 40-60% 60-80% 80-95% 95-100% 0.0 8,495 0 0 0 0 0 0 1.6 15,783 9,562 2,613 2,648 3,811 3,927 6,524 9.9 6,254 1,180 215 175 167 829 3,822 12.3 8,387 0 0 0 0 0 161 13.9 3,640 583 702 454 1,212 4,186 2,302 16.4 29,057 12,987 6,146 2,927 6,109 2,242 3,355 28.2 9,782 523 1,113 0 0 0 0 Total 81,399 24,835 10,790 6,204 11,300 11,185 16,163

Table 7. Total length (ft.) of traveled-way segment status by major geographic unit. Partial Wash-out: Partial Wash-out: Intentionally Unit ID Intact Drivable Not Drivable Removed Washed-out 0.0 8,495 0 0 0 0 1.6 27,050 16,952 666 0 201 9.9 7,404 2,945 2,294 0 0 12.3 8,038 167 161 182 0 13.9 5,007 4,698 3,375 0 0 16.4 58,335 3,393 1,096 0 0 28.2 11,418 0 0 0 0 Total 125,746 28,155 7,591 182 201

Table 8. Total length (ft.) of traveled-way erosion by category for major units.

Unit ID 0-5% 5-20% 21-40% 40-60% 60-80% 80-95% 95-100% 0.0 8,495 0 0 0 0 0 0 1.6 25,254 7,346 2,275 3,297 177 3,017 3,502 9.9 7,959 3,222 388 330 0 291 453 12.3 7,346 553 170 0 0 0 479 13.9 2,672 3,251 799 423 813 515 4,608 16.4 56,999 3,992 780 169 0 527 357 28.2 11,418 0 0 0 0 0 0 Total 120,143 18,364 4,412 4,218 989 4,350 9,398

A total of 156 water crossing structures were mapped. Of these, 76 were incidental surface drainage points in the absence of an engineered structure (e.g. culvert or trestle) and classified as “fords”. Half of these, have some form of earthen and/or woody debris accumulation on the traveled way (Table 9). Of the 79 trestles or culverts, 22 are damaged and 48 have some amount of inlet obstruction (Table 9) including 30 that are over 50% obstructed. Of the trestles in the Swale Canyon unit (“16.4”) all but one have at least one mid-span abutment. The five trestles that cross Swale Creek have between 7 and 12 mid-span abutments, usually 25-45% of which contact the active channel.

Features and Indicators Field evaluation encompassed a variety of ecological and geo-fluvial elements including hillslopes, valley floor, high stage indicators, woody debris, vegetation, and sediment. Most of the physical features of trail infrastructure (railbed, crossings, etc.) and legacy of railroad construction and maintenance actions can be termed as hydromodifications. In general, hydromodifications are features resulting from human actions that alter hydraulic patterns and/or hydrologic pathways. Categories of hydromodifications observed included: floodplain fill and grading, channel cleaning / enlargement, multiple embankments, floodplain isolation, channel realignment, valley margin isolation, off-alignment levees, floodplain stockpiles, undersized tributary crossing structures, stream crossings with mid-span abutments. Drainage and

vegetation patterns were also noted as well as hillslope processes. The most intensive hydromodifications in terms of spatial extent and magnitude are in Swale Canyon. Many hydromodifications along the Klickitat River have already begun to be reclaimed by the 1996 peak flow event (maximum event of record) and subsequent peak flow events.

Table 9. Count of water crossing structures with selected deficiencies by type and major geographic unit. Traveled Unit Structure Upstream Inlet Outlet Way Structure Mid-Channel ID Type Deposition Obstruction Obstruction Debris Damage Contact 0 Culvert 0 2 1 0 1 0 Trestle 1 1 1 1 1 0 1.6 Culvert 12 20 6 3 6 0 Ford 2 0 0 12 0 0 Culvert 2 1 0 1 1 0 9.9 Ford 2 0 0 7 0 0 Trestle 0 0 0 0 1 1 12.3 Culvert 2 5 3 0 3 0 Ford 1 0 0 2 0 0 Culvert 6 5 4 0 7 0 13.9 Ford 10 0 0 7 11 0 Trestle 1 12 3 1 2 9 16.4 Culvert 13 0 0 0 0 0 Ford 7 0 0 10 0 0 28.2 Culvert 2 2 0 0 0 0 Total* 61 48 18 44 33 10 * An individual structure may have more than one deficiency and counted in multiple columns.

Channelization / Increased Confinement Hydromodifications involving channelization contribute to stream channel instability during normal high water and floods as the stream becomes over-energized compared to the resistance of boundary materials (bed, banks, and vegetation) that developed under a lower-energy state. This increased strain on boundary materials erodes the bed and banks (including railroad embankment in places). Eroded sediments transported downstream were deposited in lower- energy reaches, contributing to erosion in these areas as lateral channel shifts become more prevalent and/or aggradation caused overtopping. Flood control and response actions by the railroad were extensive and left Swale Creek in particular in a severely degraded state.

The Christmas Floods of 1964 did extensive damage to the railbed and “most of the line, particularly the segment between Wahkiacus and Warwick [Swale Canyon], had to be rebuilt” (Grande 1997). Goldendale was without rail service for 4.5 months during Swale Canyon reconstruction (Smeltzer 2015). Decreases in riparian vegetation and pool structure are apparent in the air photo record post-1964. The photo sequence suggests that the stream has crossed a

geomorphic threshold and is currently in a dis-climax cycle. Even though post-1964 riparian vegetation conditions appear to be cyclical, pool structure has not recovered to pre-1964 levels.

Floodplain Fill and Grading, Bank Armoring, and/or Channel Displacement In Swale Canyon, grading and filling commonly extends 2-3 times the width of the railbed proper toward valley centerline with a heavily armored face (Figure 4, Figure 5). Half of the valley width is occupied by ballast materials in places (Figure 6 and Figure 7).

Figure 4. Armored banks protect grading of fill several times the width of the railbed (along valley toe).

Figure 5. Fill several times the width of railbed is protected by armor, occupies over half of valley bottom width. Resulting hydraulic forcing is associated with erosion of left valley toe.

Figure 6. Fill several times the width of railbed is protected by armor and occupies over half of valley bottom resulting in highly channelized condition.

Figure 7. Railbed fill occupies over half of valley bottom width near upstream end of Swale Canyon.

Floodplain Isolation Though the overall volume of material is less than that occupied by direct filling of floodplain and valley bottom, the hydraulic effect of mid-valley embankments is similar. Cross-sectional area available for streamflow is reduced (Figure 8 and Figure 9) and peak flows are unable to access isolated floodplain areas (Figure 10), resulting in greater stream energy.

Figure 8. Examples of embankment that both occupy and isolate floodplain from active channel along Swale Creek vicinity of RM 12 (left) and 9.8 (right).

Figure 9. Mid-valley embankment isolates floodplain (right) vicinity RM 9 on Swale Creek.

Figure 10. Railbed isolates floodplain in vicinity of Swale Creek RM 1.9 during relatively frequent (~Q4) high flow event. Channel Cleaning / Enlargement One of the more dramatic examples of regrading channel and floodplain materials is in the vicinity of Swale Creek RM 6.1. Banks are fairly uniform and have an unnaturally consistent slope (Figure 11) as well as rocky ridges that run parallel to one another, but perpendicular to stream channel (Figure 12) characteristic of sidecast or spill from a dozer blade. A steeper, rubbly backslope with a sharp transition into better vegetated valley bottom with more surface fines (Figure 13) also suggests human grading. This is a reach where the valley width more than doubles after many miles of greater confinement (natural but further confined by railroad) and would be expected to experience considerable deposition during peak flow events. Independent of field indicators, the geomorphic setting and era of railroad operations make this a location where earthwork to increase channel capacity would be expected.

Figure 11. Low, uniform foreslope, ridges perpendicular to stream channel (arrows), and steep backslope indicative of dozer operations.

Figure 12. Ridges perpendicular to stream channel orientation and uniformly sloped surface suggest dozer operations.

Figure 13. Steep backslope with unfilled relic floodplain area (left) and low, uniform foreslope of coarse materials indicative of dozer operations.

Multiple Embankments Several reaches in Swale Canyon are occupied by multiple embankments in varying states of repair which reduce cross-sectional area available for flow (Figure 14), displace floodplain, and straighten flow paths (Figure 15) increasing stream energy (discussed previously).

Figure 14. Multiple embankments effectively isolate over half of valley width from overbank flow in vicinity of Swale Creek RM 5.9.

Figure 15. Abandoned railbed (left) combined with embankment 2.5-3 times railbed width (right) associated with highly channelized condition and scour to bedrock in vicnity of Swale Creek RM 9.9.

Channel Realignment There are at least nine probable locations where extensive stream channel realignment occurred in Swale Canyon. Motivation for realignment was likely to improve hydraulic efficiency and/or, reduce the number of stream crossings, and /or improve railbed geometry to accommodate faster travel. One of the more dramatic realignments seems to have occurred in the vicinity of RM 7.7 where a very large amount of rock appears to have been moved (Figure 16), including substantial cutting of the left valley toe (camera-right in Figure 17) to accommodate the stream channel on the inside corner of a valley bend. A native, indurated layer with sub-rounded clasts appears low-bank with a fresher appearance (Figure 18) that suggests human excavation. Fresh facies in diamict deposits is one of several indicators common to areas of stream realignment (Figure 19).

Figure 16. Vicinity of Swale Cr. RM 7.7 realignment: mid-valley berm of angular material some of which display drill-holes (left) and looking downstream at incised channel with less-weathered cutface along left valley toe (right).

Figure 17. Looking upstream at present Swale Creek alignment vicinity of RM 7.7; circled area is less-weathered.

Figure 18. Bank profile in realigned Swale Creek reach, vicinity RM 7.7. Native, moderately indurated layer of sub-rounded clastes with fine matrix is several feet thick and has fresher face.

Figure 19. Fresher facies (circled) along valley toe indicate recent erosion where embankment confines realigned stream channel to approximately one-third of valley width: vicinity of Swale Creek RM 8.9. Bedrock Isolation Valley margin bedrock contacts are a key element of pool formation for the Klickitat River and Swale Creek. Embankment fill commonly obscures stream access (Figure 20 and Figure 21) and creates a more uniform channel boundary resulting in simpler habitat conditions.

Bedrock

Railbed / Fill

River

Figure 20. Bedrock isolated from Klickitat River by intact embankment.

Figure 21. Bedrock exposures obscured from alluvial contact at most flows by railbed fill materials. Off-Alignment Levees Levees of coarse material exist in intermediate floodplain locations in several places in Swale Canyon (Figure 22). These features reduce effective width for dispersing energy during peak flow events.

Figure 22. Levee of coarse material in floodplain between railbed and stream channel.

Stockpiles Stockpiles of alluvium exist in a number of places through Swale Canyon, frequently immediately adjacent to railbed and with an armored face on the stream side (Figure 23). These are generally linear features oriented parallel to valley bottom. Some are over ten feet high with a toe width of over 50 feet (Figure 24). Stockpiles displace floodplain and may be a result of

27 historic excavation to increase channel capacity in the immediate vicinity or as disposal sites from other source areas.

Railbed Armor Stockpile

Figure 23. Looking downstream at stockpile of alluvial rock in Swale Canyon (~RM 7.9).

Stockpile Railbed

Swale Cr.

Figure 24. Cross-section of Swale Creek valley bottom in alluvial stockpile area.

Signs of Increased Stream Energy A primary effect of the various forms of channelization, confinement, and armoring is increased stream energy. This is commonly manifested in a coarsened bed texture and/or scour to bedrock (Figure 25) as well as erosion of opposite channel margins (Figure 26). In some places it appears that opposite margins (e.g. valley toe, Figure 27) were also armored following some initial erosion.

Figure 25. Where alluvial substrates exists in Swale Creek, texture is usually coarse (large cobble to medium boulder; left) and/or has scoured to bedrock (right).

Figure 26. Fresh facies along valley toe indicate erosion opposite of floodplain fill and bank armor.

Figure 27. Opposite valley toe appears to have been armored in highly channelized reach.

Stream Crossings The corridor crosses many tributaries via a variety of structures. With some exceptions, most are undersized and interrupt continuity of debris and/or sediment. Crossing structures tend to have issues with plugging as well as interrupting transport of sediment and woody debris (Table 9).

Plugged Tributary Crossings Structures on smaller, unnamed tributaries tend to be culverts, most of which are undersized and many of which are plugged. However, trestles across tributaries have plugged (Figure 28) and, in some cases failed (Figure 29).

Figure 28. Low-head trestle is plugged on upstream side by debris and interrupts sediment delivery by tributary.

Figure 29. Partly washed-out trestle crosses tributary. Maple tree (circled) is rooted in alluvium deposited on trestle deck.

Woody Debris Continuity Woody debris is defined as any dead, woody plant material, including logs, branches, down and dead trees, and root wads. The scientific literature has an array of definitions of large woody debris (LWD), but a typical definition is any woody debris at least 6.6’ long and 4” diameter at the small end. Woody debris provides important geomorphic and habitat functions. Because LWD is less abundant in Swale Creek than in more forested watersheds, interception or loss of the LWD that is present may be even more critical. Nearly all of the trestles have one, and most have multiple mid-channel abutments. There are five trestles that cross Swale Creek. These effectively trap debris on the upstream side and interrupt the continuity of debris transport (Figure 30).

Figure 30. Trestles with mid-span abutments trap and interrupt LWD transport on Swale Creek.

Sediment Continuity and Tributary Fans Alluvial fans are a common geomorphic feature where smaller, steeper tributaries meet larger valley bottoms (usually in the vicinity of the trail). Many of these features are associated with intermittent and non-fish bearing tributaries. However, the sediments they provide to fish- bearing streams are important for stream channel maintenance, riparian structure, and spawning material. The pattern most commonly observed along the corridor is undersized structures that plug and get overtopped during a transport event, resulting in sediment being deposited on the embankment (Figure 31).

Figure 31. Tributary debris flow that overtopped embankment resulting in minor re-route. Undersized culvert is still in-place beneath bed.

Many tributary confluences also have some cutting/incision of the outboard (Swale Creek or Klickitat River side) shoulder as the tributary stream attempts to re-establish a more stable profile. A subtle function of these small tributary confluences is near-bank hydraulic shadowing and enrichment of bank sediment. These processes create floodplain pockets that diversify riparian habitats along the trunk stream downstream of the confluence (Figure 32).

Figure 32. Riparian vegetation growing on floodplain pocket created by sediments from tributary debris flow immediately upstream.

Wash‐outs There are a number of locations where natural processes have removed embankment materials. For the most part, this is associated with larger (> 5-10 year recurrence) peak flow events. Embankment erosion occurs via both lateral and overtopping processes.

Lateral Erosion In some locations, the armor layer on the embankment has failed and either Swale Creek (Figure 33) or, more commonly, the Klickitat River have reclaimed portions of the valley bottom. In some cases, some armor and /or other infrastructure (e.g. culverts) is left behind (Figure 34).

Figure 33. Lateral erosion of embankment (left) and combination of overtopping and lateral erosion (right).

Figure 34. Examples of relic infrastructure (culverts) in washout areas due to lateral erosion.

Overtopping and/or Breach There are a number of places where the embankment is located on former floodplain away from the valley toe leaving a topographically low area on the hillslope side. Several of these areas have been overtopped by flood events resulting in full or partial reconnection of the stream with overbank areas (Figure 35).

Figure 35. Segments where peak flows have overtopped and mostly removed embankment.

Hillslope Processes The embankment creates a topographic irregularity that intercepts mineral and organic materials. Areas interrupting hillslope processes occur over longer stretches of trail (i.e. not isolated locations like stream crossings).

Rockfall The embankment routinely intercepts rockfall from cutfaces and adjacent hillslopes. Deposits range from single and scattered colluvial particles to more concentrated rockfalls that generate scree or talus slopes (Figure 36, right).

Figure 36. Deviations of vertical alignment due to slumping (left) and horizontal alignment due to rockfall (right). Landslides and Slumps There are a couple of segments where intercepted materials have very low rock content and are primarily composed of fine (<2mm) sediments. Deposition on these segments occurs due to the finer composition of parent materials via several processes, but is generally associated with lack of toe reinforcement and oversteepened cutface (Figure 36, left).

Blowdown / Deadfall The embankment interrupts the recruitment of large woody debris from adjacent hillsides. The slope disruption generally intercepts LWD from landward locations and can increase likelihood of breakage during initial fall (Figure 37). LWD that falls across the traveled way is more likely to be cut into smaller pieces that are less effective in terms of geomorphic and habitat functions (Figure 38).

Figure 37. Trees fallen from hillside that would likely have contacted river in absence of railbed.

Figure 38. Multiple floodplain LWD pieces bucked where they cross railbed.

Runoff There are several locations where surface and/or ground water are delivered as a point source channeled by ruts along the traveled way. In other locations water is more broadly intercepted from hillslopes by the ditch or traveled way and retained by soil separates on railbed (Figure 39).

Figure 39. Surface saturation over an extended portion of traveled way.

Vegetation Management In many places, vegetation growth is limited on the embankment. The embankment’s general substrate composition (well-drained) and configuration (sharp subsurface hydraulic gradients) make establishment and persistence of native vegetation difficult (Figure 40 and Figure 41).

Figure 40. Native woody vegetation is sparse on surfaces historically mechanically graded and/or armored.

Figure 41. Typical poor woody vegetation establishment along embankment. Noxious Weeds The trail corridor hosts a variety of exotic plant species including, but not limited to: tree of Heaven (Ailanthus altissima), fiddleneck tarweed (Amsinckia lycopsoides), bur chervil (Anthriscus caucalis), cheatgrass (Bromus tectorum), diffuse knapweed (Centaurea diffusa), spotted knapweed (Centaurea maculosa), starthistle (Centaurea solstitialis), Canada thistle (Cirsium arvense, Figure 42), bull thistle (Cirsium vulgare), houndstongue (Cynoglossum officinale), toadflax (Linaria spp.), sulfur cinquefoil (Potentilla recta), Himalayan blackberry (Rubus armeniacus), and medusahead (Taeniatherum caput-medusae). In many places, higher densities along the trail relative to adjacent areas suggests the corridor is acting as a conduit spreading many of these species. Most notably, over the last fifteen years, starthistle has spread up the Klickitat River bottom from Lyle beyond Pitt and is now advancing onto adjacent hillsides where it was not previously observed (Conley, personal observation).

Figure 42. Canada thistle growing along Klickitat Trail.

Existing Railbed Alignment Modifications Numerous locations were found where trail traffic departs from the horizontal and/or vertical alignments used during railroad operations. A variety of mechanisms have caused these departures, but can be generally classified into “deliberate” and “natural” and provide precedent for modifications to enhance stream conditions.

Deliberate Alterations The embankment has been intentionally modified from horizontal and/or vertical alignments or otherwise created travel discontinuities from conditions during railroad operations. These involve water crossings, road crossings, voluntary detours, and trail infrastructure including, but not limited to:

 Lyle Trailhead – approximately 600 lineal feet (l.f.) of railbed lowered two to six vertical feet to accommodate parking lot and trailhead facilities (Figure 43)  Logging Camp Creek – concrete arch installed in 2001 to replace undersized trestle is elevated three to four feet above railbed vertical alignment (Figure 44)  Pitt – SR142 roadway is elevated six to twelve inches above adjacent railbed vertical alignment  Skookum Flats Road – is elevated four to five feet above adjacent railbed vertical alignment (Figure 45)  Klickitat –SR142 elevated one to two feet above adjacent railbed vertical alignment  Klickitat Mill – levee elevated four to five feet above historic railbed vertical alignment  Snyder Creek - trestle has been removed (Figure 46)  Suburbia - elevated levee/armor layer over former railbed and trestle across Klickitat River removed (Figure 46)  Icehouse vicinity – railbed alignments appear intact, but trail detours into shadier and sandier floodplain area (Figure 47).  Harms Rd - elevated five to eight feet above adjacent railbed vertical alignment  Centerville Hwy - elevated 1.5 to 2 feet above adjacent railbed vertical alignment

Figure 43. Construction of trail infrastructure has changed travel along the corridor.

Figure 44. Replacement box arch installed at Logging Camp Creek to reconnect fish passage is elevated 3 to 4 feet above vertical alignment of adjacent railbed.

Figure 45. Skookum Flats Road is elevated 4 to 5 feet above vertical alignment of adjacent railbed.

Figure 46. Trestles have been removed across the Klickitat River and Snyder Creek requiring detours.

Figure 47. Voluntary detour has developed around intact alignment into shadier area with sandier substrate.

Natural Alterations Various earth processes have modified the embankment and cause trail traffic to depart from historic horizontal and/or vertical alignments. These include overtopping and lateral erosion by the Klickitat River and Swale Creek (Figure 48, Figure 49, Figure 50), tributary debris flows and alluvial fan deposits (Figure 31 and Figure 51), vegetation (Figure 51), hillslope processes including creep, slumping, and rockfall (Figure 36), and various states of wash-out due to riverine (Figure 52 and Figure 53) and tributary flow (Figure 54).

Figure 48. Examples of historic overtopping that removed overlying embankment materials.

Figure 49. Alluvial deposition of sand and gravel (left) and organic debris (right) on railbed by overtopping flow.

Figure 50. Trail segment (opposite bank) overtopped during relatively frequent (~Q4) high flow event March 30, 2012.

Figure 51. Trail detours where tributary debris flow (left) and vegetation (right) have obscurred former railbed alignments.

Figure 52. Fill materials have been stripped by peak flows making bedrock fluvially accessible.

Figure 53. Active lateral erosion (left) at approximately MM 8.5 where horizontal alignment of railbed has been totally reclaimed by Klickitat River. Vegetation and boulders in right photo indicate former embankment toe (box).

Figure 54. Localized re-working of embankment by tributaries restores sediment connectivity to trunk streams.

MANAGEMENT IMPLICATIONS

Geographic Priorities Major geographic units are listed in Table 10 with relative priority and recommended strategic approach. The three lowest priority units have either no embankment contact with an active stream channel or relatively minor increased confinement. The sub-section from Klickitat to Suburbia would be a fairly high priority for reconnecting bedrock upstream of town, but due to the presence of SR 142 greatly diminishes feasibility (Figure 55). Moderate priorities have 1) general signs of recovery as a result of repeated exposure to peak flow events (without subsequent embankment repair/armoring) and/or 2) areas where active enhancement (e.g. increase fluvial access to bedrock and floodplains) would be beneficial are more spatially discrete. Swale Canyon is listed as a “High” priority as the scale and magnitude of disturbance will require active intervention to achieve results on any human timescale.

Table 10. Relative priority for stream restoration and recommended strategy by major geographic units. Unit ID Name Priority Strategy 0.0 Lyle Trailhead to Fisher Hill Bridge Very Low Passive Primarily passive with localized active Fisher Hill Bridge to Pitt Moderate 1.6 treatments Primarily passive with localized active Pitt to Klickitat Wastewater Plant Moderate 9.9 treatments 12.3 Klickitat Wastewater Plant to Suburbia Low Passive Primarily passive with localized active Suburbia to Wahkiacus Moderate 13.9 treatments 16.4 Wahkiacus to Harms Rd High Active (primary) and passive 28.2 Harms Rd to Uecker Rd Low Passive

Figure 55. Infrastructure (SR 142) between railbed and river diminishes priority of railbed treatment.

It should be noted that these are preliminary and coarse groupings. Ongoing analysis of topographic data at finer resolution and integration with hydraulic information and pending fisheries data will produce more robust guidance for project-specific work. For example, one of the best opportunities to reconnect floodplain along the Klickitat River occurs in a major unit (Pitt to Wastewater Plant) with an overall rank of “moderate”.

Stream Restoration Enhancement Strategies Strategies to restore or enhance adjacent waterbodies and floodplains fall into two categories: passive and active. A passive strategy relies on non-human processes to achieve an outcome and is desirable due to lower cost and lower likelihood of offending certain recreational interests. The many wash-outs and partial wash-outs between Fisher Hill and Pitt that have not been rebuilt are examples of passive restoration to which recreationists have adapted. An active strategy involves physical human intervention such as use of earth moving equipment. Active approaches are necessary where the scale and/or magnitude of persistent human disturbance are so great that passive restoration is unforeseeable or would take an unacceptably long period of time. Over the entire length of the corridor, a combination of passive and active restoration is appropriate with localities where one approach prevails over the other. Both active and passive strategies are consistent with Alternative 3a selected in the 2003 Record of Decision for the Environmental Assessment (USFS 2003) process.

Channelization and Confinement Active enhancement will provide the most expedient results in addressing channelization and confinement as well as stream crossings. While highly desirable results can be achieved passively (Figure 56), flow events capable of generating such response are often of a 50-100 year (or greater) recurrence interval.

Figure 56. Trail detour through floodplain where Klickitat River reclaimed embankment during 1996 flood. 45

Active approaches (e.g. that uses earth moving equipment) are most timely, but involve more up- front expense. Up-front expense can be amortized by consideration of longevity of ecological services and will generally be less expensive than life-cycle costs of repair and maintenance of the status quo. More refined spatial analysis, hydraulic data, and fisheries data are underway to better inform recommendations for active restoration. Results and recommendations will be detailed in a subsequent report once technical analysis is completed.

Water Crossings Structures, culverts in particular, on smaller, unnamed tributaries should generally be removed with open channels graded commensurate with areas immediately upstream and downstream (along tributary profile). Addressing woody debris continuity will also require an active approach as trestles will need to be physically modified. If debris must be removed from trestles, it should be relocated to a comparable hydraulic environment downstream of the trestle. In other words, if it’s located within the active channel upstream of the trestle, it will be placed into the active channel downstream of the trestle. All woody debris should be left intact in its entire length and not be cut, bucked, or otherwise shortened (e.g. rootwads, if present, should not be removed).

Traditional maintenance practices that clear materials off the trail, re-grade a uniform (square) traveled way, and end-haul spoils to a disposal area impede sediment continuity. Such clearing, diverts materials from the fluvial system, re-sets the sediment “trap”, and ensures the need for future maintenance. An alternative to this is to accept the irregularity (e.g. hump) in the trail profile with minor grading to facilitate trail traffic. YNFP personnel have discussed such alternatives with local Parks’ staff over the past decade and, where implemented (Figure 57, left photo), this seems to be a viable approach that is both less costly and less obstructive of natural processes.

Figure 57. Two examples of grading of fan materials. Example on left maintains outsloped profile and is preferable to stepped profile on right.

There are at least half a dozen examples of this along the corridor that seem to have both improved fluvial function and compatibility with recreational and administrative trail use (Figure 57). Pre-treatment of tributary crossings (e.g. by locally lowering the outboard shoulder of the embankment) could reduce the overall quantity of material retained on the prism and diminish post-deposition maintenance effort. Material displaced during re-grading activities should be spread along the stream side of the trail to replicate natural alluvial dispersal.

Wash‐outs Acceptance of the prevailing post-event landform will generally be preferable with either relocation of the traveled way or minor grading to facilitate traffic. Discretion should be exercised in determining a design standard for trail traffic. While it might be nice to have accessibility for full-sized vehicles along the corridor, it is not necessary as it is the “Klickitat Trail”, not the “Klickitat Road”. There is ample precedence for segments that are only accessible by foot, horse, bike, and ATV (see Existing Railbed Alignment Modifications section). While not ideal from a stream restoration perspective, an ATV/UTV design standard could be largely compatible with reestablishing connectivity of geomorphic processes.

Where wash-outs occur or are anticipated to occur, the trail should be re-routed. Embankment reconstruction will cause greater harm and set-back the longer term trajectory where geologic process prevails. A good example of this is in the vicinity of trail mile-marker 8 where the trail was re-routed to the toe of the valley wall (Figure 58).

Person

Figure 58. Lateral erosion by the Klickitat River would be very expensive to inhibit and work counter to recovery trends. Such solutions are generally win-win as they are usually less costly than reconstruction and “protection” measures (e.g. armoring). They also have lower subsequent maintenance requirements and costs. Lateral stream movements that result in bedrock contact along the stream margin (Figure 59) are particularly valuable as they frequently result in pool-scour and bank irregularities that improve stream energy distribution, create pocket habitat, and sort sediment.

Figure 59. Two locations where Klickitat River has re-worked embankment materials to reestablish bedrock contact along valley toe.

Hillslope Processes Considered in valley cross-section, the embankment is a topographic irregularity that intercepts mineral and organic materials over long horizontal distances (i.e. not isolated locations like stream crossings). With regard to mineral material (i.e. sediment), the interruption can be diminished by grading materials in-situ in a manner that mimics landforms that would have been present pre-railroad, for example, tributary fans. If rockfall must be cleared, then rock should be swept or placed onto the face or along the toe where it can continue its downslope movement and eventual recruitment into the adjacent stream or river though judgment should be exercised (and a geomorphologist consulted) if the end result would be something resembling rip-rap. End- hauling of sediment should be a last resort.

Intercepted materials in some locations have very low rock content and are primarily composed of fine (<2mm) sediments due to composition of parent materials in valley toe. These tend to slump and/or rill due to steep, unreinforced, bare cut-faces. These locations will pose an ongoing challenge until the cross-section across the railbed achieves angle of repose and/or becomes vegetated. As with other colluvial materials, sediment should be stabilized in-situ if feasible (e.g. via revegetation) and runoff should be routed away from these areas.

The embankment interrupts the recruitment of large woody debris from adjacent hillsides and landward floodplains. If trees and snags must be cleared, then they should be left intact in their entire length (with rootwad, if present) and placed along the toe or on the face where it can be recruited into the adjacent stream or river.

Runoff Where surface and/or ground water are intercepted by the ditch or ruts and are channeled along the traveled-way in fine-textured areas, effort should be made to expediently route flow across the embankment. Turnpike construction (Figure 43, right photo) appears to be the preferred approach in lower portions of the trail in recent years. However, these structures create additional maintenance needs and are restrictive for some vehicle and equipment types. Where possible, grade dips or French drains would be preferable.

Vegetation Management Where native vegetation becomes established, it should be encouraged and clearing, brushing, and grubbing should be kept to a minimum. Removal of vegetation should be limited, except in the case of exotic species. Where necessary, cleared woody material from native species should be kept in-tact and placed along the streamside toe or face where it can be recruited into the fluvial system. In the case of exotic species, such as “Tree of Heaven”, those plant materials should be disposed of as recommended by weed and pest experts. Effort should be made to check the spread of weeds and eliminate where possible. The follow-up technical report will include a map of weed locations encountered during segment inventory.

CONCLUSION Railroad infrastructure between Lyle and Warwick has altered geomorphic processes along 13.5 miles of the lower Klickitat River and 14 miles of Swale Creek since 1902. Though the railroad was decommissioned in 1993, remaining infrastructure has a variety of persistent effects including: floodplain displacement and isolation, channel cleaning / enlargement, channel realignment, valley margin isolation, poor riparian vegetation, and interruption of sediment and debris transport. Railroad embankments contact active channels of the Klickitat River or Swale Creek for 41% (~12.4 miles) and the 100-year floodplain for 82% (~24.8 miles) of the total length. Floodplain access to high flows is limited to some degree (flow-dependent) along approximately one-third of the embankment. Approximately 5.3 miles (cumulative) of embankment obscure fluvial access to bedrock. Approximately 6.8 miles (cumulative) of embankment showed some sign of partial washout (e.g. overtopping, breach, or face erosion), though only about 1.4 miles was not drivable by an ATV. One hundred fifty-six water crossing points were identified, of which, 79 were culverts or trestles; the remainder were fords. Sixty percent (n=48) of structures (culverts and trestles only) had some degree of inlet obstruction (thirty structures were obstructed >50%) and 28% (n=22) of structures had some form of damage. A restoration approach that blends active and passive strategies is appropriate over the course of the study area and consistent with both interim use of the corridor as a recreational trail and the 2003 Record of Decision (Alternative 3a; USFS 2003). High flows since the railroad was decommissioned (1996 peak flow, in particular) have been partially effective at reclaiming fluvial function along portions of the Klickitat River. A prevailingly passive strategy is recommended along the Klickitat River, though some discrete localities will benefit from active restoration. The scale and magnitude of hydromodifications in Swale Canyon (Wahkiacus to Harms Road) is extensive and has fundamentally altered geologic processes. Consequently, a passive strategy is unlikely to result in restored habitat or fluvial function. Active restoration will be necessary, particularly between trail miles 20 and 27, to recover stream and floodplain function that resembles pre-railroad conditions. Ongoing analysis and integration with hydraulic

49 and habitat information is occurring at a finer spatial resolution and will be detailed in a future report.

LITERATURE AND DATA CITED Benito, G. and J. O’Connor. 2003. Number and size of last-glacial Missoula floods in the Columbia River valley between the Pasco Basin, Washington and Portland, Oregon. GSA Bulletin 115(5):624-638. Brierly, G.J. and K.A. Fryirs. 2005. Geomorphology and River Management. Blackwell Publishing. Malden, MA. 398p. Conley, W. 2005. Klickitat Watershed Enhancement Project: Annual Report for Sept 1, 2003 – October 31, 2004. Prepared for U.S. Department of Energy, Bonneville Power Administration, Portland, OR. Project #1997-056-00. Cooper, A.B. and M. Mangel. 1999. The dangers of ignoring metapopulation structure for the conservation of salmonids. Fish. Bull. 97:213-226. Durr, G. 2001. “Goldendale Branch: Spokane, Portland, and Seattle Ry. Fourth Subdivision”. http://www.teleport.com/~amacha/gldndale.htm Environmental Systems Research Institute (ESRI). 2012. ArcGIS Desktop: Release 10.0.4.4000. ArcInfo, Spatial Analyst, and 3D Analyst licenses. Environmental Systems Research Institute Redlands, CA. Environmental Systems Research Institute (ESRI). 2014. ArcGIS Desktop: Release 10.2.2.3552. ArcInfo, Spatial Analyst, and 3D Analyst licenses. Environmental Systems Research Institute Redlands, CA. Frissell, C.A., W.J. Liss, C.E. Warren, and M.D. Hurley. 1986. A hierarchical framework for stream habitat classification: Viewing streams in a watershed context. Environmental Management. 10: 199-214.

Gaertner, J.T. 1992. North Bank Road: The Spokane, Portland, and Seattle Railway. Washington State University Press, Pullman, WA.

GeoDigital, Inc. 2012. Aerial LiDAR and high-resolution photography. Data acquisition: November 2011.

Grande, W.R. 1997. The Northwest’s Own Railway: Spokane, Portland, and Seattle Railway and its Subsidiaries. Vol. 2. Grande Press, Portland, Oregon.

Kellerhals, R., M. Church, and D.I. Bray. 1976. Classification and analysis of river processes. Journal of the Hydraulics Division, Proc. Of Am. Soc. Civil Eng. 102(HY7): 813-829. Knighton, D. 1998. Fluvial Forms and Processes. Oxford Univ. Press, New York. 383p. Microsoft Corp. 2013. Microsoft Office Professional Plus 2013. v15.0.4709.1000. Redmond, WA.

Neuendorf, K., J. Mehl, and J. Jackson. 2005. Glossary of Geology. American Geological Institute. Alexandria, VA. 779p. National Water Information System (NWIS). 2014. Historic peak flow data and gage information from USGS-operated gages. http://nwis.waterdata.usgs.gov/wa/nwis/peak National Marine Fisheries Service (NMFS). 2009a. Recovery Plan for the Rock Creek Population of the Middle Columbia River Steelhead Distinct Population Segment. October 2009. 137p.

Osterkamp, W. 2008. Annotated Definitions of Selected Geomorphic Terms and Related Terms of Hydrology, Sedimentology, Soil Science, and Ecology. Reston, VA. Open File Report 2008-1217. 49p. Pierson, T. 2005. Distinguishing between Debris Flows and Floods from Field Evidence in Small Watersheds. USGS Fact Sheet 2004-3142. 4p. Smeltzer, J.A. 2015. “Looking Back: 50 Years Ago”. The Goldendale Sentinel, Wednesday, May 13, 2015. Vol 136(19):3.

USDA Forest Service (USFS). 1997. Environmental Assessment for the Klickitat Rails-To- Trails. Columbia River Gorge National Scenic Area, Hood River, OR. 103p. USDA Forest Service (USFS). 2003. Environmental Assessment for the Klickitat Rails-To- Trails. Columbia River Gorge National Scenic Area, Hood River, OR. 210p. USGS. 2015. EarthExplorer. http://earthexplorer.usgs.gov/ Washington Department of Natural Resources (WDNR). 2013. Surface Geology 1:100,000. File geodatabase. Downloaded by W. Conley August 2013. http://www.dnr.wa.gov/ResearchScience/Topics/GeosciencesData/Pages/gis_data.aspx Watershed Sciences, Inc. (WSI). 2009. Remote Sensing Data Collection: Klickitat River, WA. Prepared for Yakama Nation Fisheries Program. Data acquisition: April 2009. Corvallis, OR. 34p. Watters, T.R. 1989. Periodically spaced anticlines of the Columbia Plateau in Reidel, S.P. and Hooper P.R., eds., Volcanism and tectonism on the Columbia River flood-basalt province: Boulder, Colorado, Geological Society of America, Special Paper 239, Pages 283-292.

APPENDIX A Chronology of Valley Bottom Railroad Embankments in the Klickitat Subbasin

Will Conley, Hydrologist/Geomorphologist Yakama Nation Fisheries Program Klickitat Field Office, Wahkiacus, WA

(updated from December 2003 version)

A-1

Date(s) Description 1889 to 1899 Early efforts to develop rail service from the Columbia River to Goldendale:  group of Goldendale citizens forms Columbia Valley and Goldendale (CV&G) Railway ,  CV&G merges with a company from Pasco to form Pasco, Goldendale, & Columbia Valley (PG&CV) Railroad Company. Surveys made for several years, but no construction undertaken.  Goldendale residents persuade E.E. Lytle (builder of the Columbia and Southern Railway in Oregon) to form the Columbia & Klickitat (C&K) Railroad. Jan. 1902 A group of Portland, OR entrepreneurs incorporates Columbia River & Northern (CR&N) Railway to construct line from Lyle to Goldendale Mar. 1902 Surveys begin in Klickitat Valley from Centerville and work begins on a terminal and dock in Lyle May 1902 CR&N acquires The Dalles, Portland, and Astoria Navigation (DP&AN) Company May 1902 CR&NR awards grading contract to Axtell Anderson for upper 15 miles of line; grading begins June 1902 CR&NR awards grading contract to Corey Bros. and Alden for lower 27.5 miles (head of Swale Canyon to Lyle) Sept. 1902 Grading completed, except 8 mile section in Swale Canyon 11/25/1902 supplemental Articles of Incorporation approved to extend the line from Goldendale to the Columbia River via “Luny [Luna] Gulch” & Rock Creek and to extend the line to Bickleton Nov/Dec  One work party taken off construction to survey to Bickleton. 1902  Another crew followed Rock Cr. down to the Columbia R.  Survey activity aroused suspicion of Northern Pacific Railroad which authorized construction of a parallel line if necessary to protect NPR interests. Surveys performed. Dec. 1902 Grading completed 1903 & 1904  NPR negotiates to purchase CR&N, without success  CR&N not profitable because of dependence on boat traffic…lumber and passengers could travel to Columbia by horse team for less cost Feb. 1903 CR&N purchases boats and interests from Columbia River & Puget Sound, a DP&AN rival April 1903 CR&N track machine reaches Goldendale 5/1/1903 Operation of daily freight begins over entire line, though not completely ballasted or surfaced; depots at Lyle, Centerville, and Goldendale Dec. 1903 Construction completed late-1904 Principle stockholders in CR&N proceed with Bickleton extension early-1905 Principle stockholders in CR&N put Bickleton extension on hold 2/28/1905 Northwestern Improvement Company (NIC) acquires stock and bonds of the CR&N 10/1/1906 NIC purchases underlying collateral and CR&NR settled for the notes. Portland and Seattle Railway begins operating line

A-2

1/15/0908 P&S track from Cliffs to Lyle completed, providing CR&N direct outside rail connection 3/30/1908 Spokane, Portland, and Seattle (SP&S) Railway purchases line from NIC 4/1/1908 SP&S takes over operations 6/1/1908 SP&S acquires stock of DP&AN 1909 Western Pine Lumber (WPL) Company formed; builds a dry log mill, planing mill, and logging roads at town of Klickitat 1911 Both WPL mills burn 1914 Klickitat and Northern (K&N) Railway organized to extend 15 miles northwest of Klickitat and serve Camas Prairie [Glenwood] area WPL constructs inclined railway up Snyder Canyon to access K&N railway and to transport logs from plateau to WPL mills as well as service six mills and several sheep and cattle ranches on the Plateau 4/25/1915 SP&S forced to sell DP&AN because Panama Canal Act (1912) prohibits any railroad from owning a water line that competes with a railroad 7/26/1917 Contract signed between SP&S and K&N allowing K&N to use SP&S line between Klickitat and Pitt 1918 K&N forced into receivership. Acquired by WPL and renamed Western Pine Lumber Company Railroad 6/21/1922 J. Neils Lumber Company (of Minnesota) acquires WPL. Renames local operations to Klickitat Log & Lumber Company (KLLC) 1939 18 miles of line along Klickitat River replaces the Snyder Canyon line 1/1/1957 St. Regis Paper Company acquires J. Neils Lumber Co. 4/4/1964 KLLC logging railroad ceases operations. Trucks begin hauling logs over the railroad KLLC grade 1960s Two families build homes in lower Swale Canyon (~3.5 miles from Wahkiacus). SP&S continues operations from Lyle to Goldendale  Most cars left at Klickitat, remaining cars generally empty  Depots at Klickitat, Centerville, and Goldendale  Shelter sheds at Wahkiacus and Warwick December “Christmas Floods”: 1964  Entire Goldendale line is severely damaged by floods and out of service for many weeks.  “Most of the line, particularly the segment between Wahkiacus and Warwick [Swale Canyon], had to be completely re-built” 5/8/1965 After 4.5 months without service, the first train arrives in Goldendale after testing track laid on new railbed in Swale Canyon 1988 Rail service between Klickitat and Goldendale ceases 1989 Lumber mill at Klickitat closes 1991 Lower Klickitat Wild and Scenic River Management Plan notes potential for railway abandonment late-1992 Burlington Northern (BN) files abandonment notice with Interstate Commerce Commission (ICC) for Lyle to Goldendale route 1993 railway is dismantled

A-3

Aug. 1993 BN sells rights in the Lyle to Warwick section to the Rails-To-Trails Conservancy (RTC) which “banks” the 31-mile line with approval by ICC March 1994 RTC takes title to rail corridor April 1994 RTC donates rail corridor title to WSPRC Feb. 1996 Railbed suffers flood “damage”, mostly Fisher Hill to Wahkiacus; Wahkiacus to Warwick suffers relatively minor damage 1996 WSPRC removes damaged trestle across Klickitat River at Suburbia and develops flood repair proposal for FEMA 1997 Environmental Assessment (EA) published by US Forest Service. No decision made by Regional Forester. Jan. 2003 Memorandum of Understanding signed between WSP and USFS Aug. 2003 Environmental Assessment (EA) published by US Forest Service. 2003 USFS Record of Decision selects Alternative 3a.

Durr (2001), Gaertner (1992), Grande (1997), Smeltzer (2015), USFS (1997), USFS (2003)

A-4

APPENDIX B Klickitat Trail Stream Continuity Project: Phase 1

Will Conley, Hydrologist/Geomorphologist Yakama Nation Fisheries Program Klickitat Field Office, Wahkiacus, WA

B-1

The Klickitat Trail Stream Continuity Project (Phase 1) is intended to reduce impacts of the Klickitat Trail embankment and tributary crossing structures between Schilling Road and Uecker Road.

Project goals are to improve fisheries habitat conditions in Swale Creek and reduce maintenance needs. Primary beneficiaries are expected to be steelhead (Oncorhynchus mykiss; ESA-listed mid-Columbia DPS), bridgelip suckers (Catostomus columbianus), and speckled dace (Rhinichthys osculus), and peamouth minnows (Mylocheilus caurinus). Coho (Oncorhynchus kisutch) and Chinook (Oncorhynchus tshawytscha) salmon in the lower portions of Swale Creek would be secondary beneficiaries of project actions.

Objectives are to:

1) Restore streamflow, bedload, and woody debris connectivity for 29 Swale Creek tributaries or distributaries, and 2) Improve trail surface drainage at 11 locations to reduce flow along trail surface and excessive mobilization of fines by trail traffic, and 3) Evaluate alternatives that reduce stream channel intrusion by five trestles that span Swale Creek. Objectives 1 and 2 involve removal and/or re-grading of 40 embankment crossings that impede streamflow, bedload, and debris delivery along 14 miles of Swale Creek, including 4 tributary trestles, 22 culverts, and 14 fords.

Objective 3 involves evaluating alternatives to existing trestles that have multiple, closely-spaced points of channel contact that prevent downstream transport of large woody debris (LWD). The preferred alternative would involve removal of some (to be determined) number of abutments that results in approximately fifty to sixty feet of unobstructed channel width under each crossing. Key to this effort will be a field evaluation of the suitability for and load-bearing capacity of the various mid-span abutments to support an iterative analysis of the maximum permissible span for a single-lane, light-duty bridge. The design vehicle for bridges would be a Utility Task Vehicle (UTV; GVW ~3,000 lbs.) which would accommodate 99%+ of trail use as well as permit routine patrol and light-duty maintenance. Each crossing has graded approaches for ford crossings already, though in varying states of functionality. Implementation of a light- duty bridge option would also involve improvement of the fords to accommodate over-sized traffic.

B-2

B-3

B-4

APPENDIX C Glossary of Selected Terms

Compiled by:

Will Conley, Hydrologist/Geomorphologist Yakama Nation Fisheries Program Klickitat Field Office, Wahkiacus, WA

C-1

Abundance - In the context of salmon recovery, unless otherwise qualified, abundance refers to the number of adult fish returning to spawn. Active channel – for a given location in a valley, the channel or channels that regularly transmit flow of water, sediment, and debris; inclusive of primary and some secondary channels. Alignment – the horizontal (planform) or vertical (profile) geometry of a feature in two- dimensional space. Alluvial – of or associated with a stream or running water Alluvium – general term for materials deposited by running water Anadromous fish: Species that are hatched in freshwater, migrate to and mature in saltwater, and return to freshwater to spawn. Bar – accumulation of alluvial sediment formed in the channel, along the banks, or at the mouth of a stream where a decrease in velocity during transport conditions induces deposition. Capacity (sediment) - the ability of a current to transport a quantity of sediment, measured as the amount (e.g. mass) at a given point per unit time. Characteristic form time – the time over which the state is expected to persist Clast – an individual constituent, grain, or fragment of a sediment or rock produced by the mechanical or chemical disintegration of a larger rock mass. For the purposes of this report, the term is used generally to indicate a particle larger than sand (>2mm). Colluvial – of or pertaining to colluvium. Colluvium – general term applied to any loose, heterogeneous, and incoherent mass of soil material and/or rock fragments deposited by non-channelized gravitational movement, usually collecting at the base of slopes or hillsides. Competence (sediment) – the ability of a current to transport sediment of a particular size, measured as the largest particle transported. Confinement – the degree to which valley margins encroach horizontally on active channel margins. Cross-section (XS) - diagram or drawing showing configuration or slope (e.g. ground or water surface) along a given line as it would appear if it were intersected by a vertical plane oriented cross-wise to the long axis of a feature (e.g. oriented from one streambank to the other). Distinct population segment (DPS) - A listable entity under the ESA that meets tests of discreteness and significance according to USFWS and NMFS policy. A population is considered distinct (and hence a “species” for purposes of conservation under the ESA) if it is discrete from and significant to the remainder of its species based on factors such as

C-2

physical, behavioral, or genetic characteristics, it occupies an unusual or unique ecological setting, or its loss would represent a significant gap in the species’ range. Debris flow – sediment-water mixture where flow behavior is controlled by entrained sediment, typically exceeding 50% by volume and frequently includes some clasts in suspension. Diversity - All the genetic and phenotypic (life history, behavioral, and morphological) variation within a population. In the context of salmon recovery, variations could include anadromy vs. lifelong residence in freshwater, fecundity, run timing, spawn timing, juvenile behavior, age at smolting, age at maturity, egg size, developmental rate, ocean distribution patterns, male and female spawning behavior, physiology, molecular genetic characteristics, etc. Ecogeomorphic – of or pertaining to the interaction of organisms and landforms. Embankment – a linear structure, usually of earth or gravel, constructed so as to extend above the natural ground surface and designed to hold back water, from overflowing a level tract of land, to retain water in a reservoir, tailings in a pond, or a stream in its channel, or to carry a roadway or railroad. Entrenchment – the degree to which channel(s) are countersunk into valley fill. Equant – with regard to watershed shape, where the width and length are approximately the same (definitely within a factor of 1.5). Evolutionarily significant unit (ESU) - a group of Pacific salmon or steelhead trout that is (1) substantially reproductively isolated from other conspecific units and (2) represents an important component of the evolutionary legacy of the species. Fines – general term to describe a particles <2mm in size; inclusive of sands, silts, and clays. Flood – a streamflow event of sufficient magnitude to exceed capacity of the active channel(s) and inundates overbank areas; unless otherwise specified, assumed to be water flow. Geomorphic – of or pertaining processes that affect to the shape of the earth’s surface. Hydraulics – the science of fluids in motion. For the purposes of this report, movement or action caused by water. Hydrology – the science of water properties, circulation, and distribution in space and time from delivery to a landscape from the atmosphere until it’s return to the atmosphere or delivery to the ocean. Hydromodification – any human action or result that alters natural hydraulics and/or hydrology for a site or watershed. For the mapping in this report, its use is limited to locations where earth-moving has occurred that otherwise alters or obstructs surface or groundwater flow patterns (and does not include rip-rapped banks, etc.) Hydrophyte – plant adapted to habitats of water or very wet conditions.

C-3

Hyperconcentrated flow – intermediate condition between water flow and debris flow where suspended sediment may compose 5 to 60% of the mixture by volume and is typically composed of particles sand-sized or smaller. Induration – the hardening of a soil horizon by heat, pressure, and/or chemical action to form a hardpan; results in layer that is more resistant to erosion and with lower permeability than would otherwise be expected of a layer with similar particle composition. Limiting factor - physical, biological, or chemical features (e.g., inadequate spawning habitat, high water temperature, insufficient prey resources) experienced by the fish that result in reductions in viable salmonid population (VSP) parameters (abundance, productivity, spatial structure, and diversity). Key limiting factors are those with the greatest impacts on a population’s ability to reach a desired status. Longitudinal Profile - diagram or drawing that shows configuration or slope of a feature (e.g. ground or water surface) along a given line as it would appear if it were intersected by a vertical plane oriented parallel to the long axis of a feature (e.g. oriented upstream to downstream). Major population group (MPG) - a group of salmonid populations that are geographically and genetically cohesive. The MPG is a level of organization between demographically independent populations and the ESU or DPS. Natural-origin fish - fish that were spawned and reared in the wild, regardless of parental origin. Planform – the shape of a feature in two dimensions (horizontally), as viewed from above. Primary channel – for a given location in a valley, the one stream channel that most frequently transmits the greatest flow; often, but not necessarily, the last channel to dry-up. Productivity - the average number of surviving offspring per parent. Productivity is used as an indicator of a population’s ability to sustain itself or its ability to rebound from low numbers. The terms “population growth rate” and “population productivity” are interchangeable when referring to measures of population production over an entire life cycle. Can be expressed as the number of recruits (adults) per spawner or the number of smolts per spawner. Reach – section of river along which boundary conditions are sufficiently uniform such that the river maintains a near consistent structure. Reaction time – the time taken for a system to react to a change in conditions. Relaxation time – the time taken for the system to attain a characteristic (equilibrium) state Salmonid - fish of the family Salmonidae, including salmon, trout, chars, grayling, and whitefish. In general usage, the term usually refers to salmon, trout, and chars. Secondary channel - for a given location in a valley, any channel that is not the primary channel; may be “active” or not. C-4

Segment – alternating patterns of reach-scale river behavior. Smolt - a juvenile salmonid that is undergoing physiological and behavioral changes to adapt from freshwater to saltwater as it migrates toward the ocean. Spatial structure - characteristics of a fish population’s geographic distribution. Current spatial structure depends upon the presence of fish, not merely the potential for fish to occupy an area. Thalweg – the line connecting the lowest or deepest points along a stream bed or valley. Trail – a path generally used for foot, bicycle, or horse traffic. May or may not be maintained. Railbed - the ballast layer of earthen materials supporting a railway track. Riprap – a layer of large, durable fragments of broken rock, specially selected and graded, thrown together irregularly or fitted together. Its purpose is to prevent erosion by waves or currents and thereby preserve the shape of a surface, slope or underlying earthen structure. Valley fill – unconsolidated sediments that occupy the valley floor and lie on top of bedrock; inclusive of alluvium and colluvium. Viability criteria - criteria defined by NMFS-appointed Technical Recovery Teams to describe a viable salmonid population, based on the biological parameters of abundance, productivity, spatial structure, and diversity. These criteria are used as technical input into the recovery planning process and provide a technical foundation for development of biological delisting criteria. Viable salmonid population (VSP) - an independent population of Pacific salmon or steelhead trout that has a negligible risk of extinction over a 100-year time frame. VSP parameters - abundance, productivity, spatial structure, and diversity. These describe characteristics of salmonid populations that are useful in evaluating population viability. See NOAA Technical Memorandum NMFS-NWFSC-42, Viable salmonid populations and the recovery of evolutionarily significant units (McElhany et al. 2000). Water flow – condition where the properties of water dictate flow behavior; generally composed of less than 5-10% suspended-sediment by volume. For the purposes of this report, “flow” and “water flow” are used interchangeably.

Glossary References: Brierly and Fryirs (2005), Frissell et al. (1986), Kellerhals et al. (1976), Knighton (1998), Neuendorf et al. (2005), NMFS (2009), Osterkamp (2008), Pierson (2005)

C-5