Conceptio › Archive › USGS Publications

USGS Publicationspublic full text

Applying Intrinsic Potential Models to Evaluate Salmon ( Oncorhynchus spp.) Introduction into Main-Stem and Tributary Habitats Upstream from the Skagit River Hydroelectric Project, Northern Washington

Jeffrey J. Duda, Jill M. Hardiman, L. Benda, D. Miller, K. Andras, P. Bigelow · U.S. Geological Survey

USGS Publications · Papers · License: Public Domain

filereport habitats introduction models northern potential river usgsopen

usgs, geological survey, united states, science, government publication, usgs open-file report, 2023, 1077, applying, intrinsic, potential, models, evaluate, salmon, oncorhynchus, spp, introduction, main, stem, tributary, habitats, upstream, skagit, river, hydroelectric, project, northern, washington

Prepared in cooperation with Seattle City Light

Applying Intrinsic Potential Models to Evaluate Salmon (Oncorhynchus spp.) Introduction into Main-Stem and Tributary Habitats Upstream from the Skagit River Hydroelectric Project, Northern Washington

Open-File Report 2023–1077

U.S. Department of the Interior U.S. Geological Survey

Cover. View of Ross Lake, Nohokomeen Glacier, and the Cascade Mountain Range from Desolation Peak, North Cascades National Park. Photograph taken August 28, 2020, by Jeffrey J. Duda, U.S. Geological Survey

Applying Intrinsic Potential Models to Evaluate Salmon (Oncorhynchus spp.) Introduction into Main-Stem and Tributary Habitats Upstream from the Skagit River Hydroelectric Project, Northern Washington By Jeffrey J. Duda and Jill M. Hardiman

Prepared in cooperation with Seattle City Light

Open-File Report 2023–1077

U.S. Department of the Interior U.S. Geological Survey

U.S. Geological Survey, Reston, Virginia: 2023

For more information on the USGS—the Federal source for science about the Earth, its natural and living resources, natural hazards, and the environment—visit https://www.usgs.gov or call 1–888–392–8545. For an overview of USGS information products, including maps, imagery, and publications, visit https://store.usgs.gov/ or contact the store at 1–888–275–8747. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government. Although this information product, for the most part, is in the public domain, it also may contain copyrighted materials as noted in the text. Permission to reproduce copyrighted items must be secured from the copyright owner. Suggested citation: Duda, J.J., and Hardiman, J.M., 2023, Applying intrinsic potential models to evaluate salmon (Oncorhynchus spp.) introduction into main-stem and tributary habitats upstream from the Skagit River Hydroelectric Project, northern Washington: U.S. Geological Survey Open-File Report 2023–1077, 44 p., https://doi.org/10.3133/ofr20231077. Associated data for this publication: Duda, J.J., and Hardiman, J.M., 2023, Upper Skagit River intrinsic potential results: U.S. Geological Survey data release, https://doi.org/10.5066/P9MKQ2UK ISSN 2331-1258 (online)

iii

Acknowledgments This project was informed by input from several biologists and stakeholders involved in the Skagit River Hydroelectric Project relicensing process. We would like to thank Kevin Andras from Terrainworks for assistance with the NetMap software and its underlying datasets.

Contents Acknowledgments��iii Abstract��1 Introduction��1 Purpose and Scope��4 Methodology��4 Results and Interpretations��7 Potential Fish Distribution Upstream from the Skagit River Hydroelectric Project Dams��7 Chinook Salmon Intrinsic Potential Model Results��7 Chinook Salmon Intrinsic Potential Based on the Connor Model��15 Chinook Salmon Intrinsic Potential Based on the Busch Model��15 Chinook Salmon Intrinsic Potential Based on the Cooney and Holzer Model��18 Coho Salmon Intrinsic Potential Model Results��18 Coho Salmon Intrinsic Potential Based on the Burnett Model��18 Coho Salmon Intrinsic Potential Based on the Agrawal Model��18 Coho Salmon Intrinsic Potential Based on the Romey Model��18 Steelhead Intrinsic Potential Model Results��31 Steelhead Intrinsic Potential Based on the Puget Sound Technical Recovery Team Model��31 Steelhead Intrinsic Potential Based on the Agrawal Model��31 Steelhead Intrinsic Potential Based on the Burnett Model��31 Summary��43 References Cited��43

Figures 1.

Map of upper Skagit River watershed showing targeted and non-targeted tributary habitat with potential accessibility for anadromous salmonids upstream from three currently impassable dams, in Whatcom County, northern Washington��2 Graphs showing length of high, medium, and low intrinsic potential habitat for Chinook salmon based on three different models for target tributaries upstream from the three upper Skagit River dams, in northern Washington��8 Graphs showing length of high, medium, and low intrinsic potential habitat for Chinook salmon based on three different models for non-target tributaries upstream from the three upper Skagit River dams, in northern Washington��15 Map showing high, medium, and low intrinsic potential scores for Chinook salmon derived from the Connor intrinsic potential model applied to tributary and main-stem habitat upstream from the Skagit River Hydroelectric Project dams, in northern Washington��16 Map showing high, medium, and low intrinsic potential scores for Chinook salmon derived from the Busch intrinsic potential model applied to tributary and main-stem habitat upstream from the Skagit River Hydroelectric Project dams, in northern Washington��17 Map showing high, medium, and low intrinsic potential scores for Chinook salmon derived from the Cooney and Holzer intrinsic potential model applied to tributary and main-stem habitat upstream from the Skagit River Hydroelectric Project dams, in northern Washington��19

10.

11.

12.

13.

14.

15.

16.

Graphs showing length in kilometers of high, medium, or low habitat intrinsic potential for coho salmon based on three different models for target tributaries upstream from the three upper Skagit River dams, in northern Washington��26 Graphs showing length in kilometers of high, medium, or low habitat intrinsic potential for coho salmon based on three different models for non-target tributaries upstream from the three upper Skagit River dams, in northern Washington��27 Map showing high, medium, and low intrinsic potential scores for coho salmon derived from the Burnett intrinsic potential model applied to tributary and main-stem habitat upstream from the Skagit River Hydroelectric Project dams, in northern Washington��28 Map showing high, medium, and low intrinsic potential scores for coho salmon derived from the Agrawal intrinsic potential model applied to tributary and main-stem habitat upstream from the Skagit River Hydroelectric Project dams, in northern Washington��29 Map showing high, medium, and low intrinsic potential scores for coho salmon derived from the Romey intrinsic potential model applied to tributary and main-stem habitat upstream from the Skagit River Hydroelectric Project dams, in northern Washington��30 Graphs showing length in kilometers of high, medium, and low habitat intrinsic potential for steelhead based on three different models for target tributaries upstream from the three upper Skagit River dams, in northern Washington��32 Graphs showing length in kilometers of high, medium, and low habitat intrinsic potential for steelhead based on three different models for non-target tributaries upstream from the three upper Skagit River dams, in northern Washington��39 Map showing high, medium, and low intrinsic potential scores for steelhead derived from the Puget Sound Technical Recovery Team intrinsic potential model applied to tributary and main-stem habitat upstream from the Skagit River Hydroelectric Project dams, in northern Washington��40 Map showing high, medium, and low intrinsic potential scores for steelhead derived from the Agrawal intrinsic potential model applied to tributary and main-stem habitat upstream from the Skagit River Hydroelectric Project dams, in northern Washington��41 Map showing high, medium, and low intrinsic potential scores for steelhead derived from the Burnett intrinsic potential model applied to tributary and main-stem habitat upstream from the Skagit River Hydroelectric Project dams, in northern Washington��42

Tables 1.

Tributary and main-stem reaches upstream from Ross Lake reservoir identified by stakeholders as having accessible fish habitat that could potentially be used by introduced Pacific salmon and steelhead, with estimates of the amount and average percent gradient of the accessible sections, in northern Washington��5 2. Ross Lake reservoir tributary reaches identified by stakeholders as having accessible fish habitat that could potentially be used by introduced Pacific salmon and steelhead, with estimates of the amount and average percent gradient of the accessible sections, in northern Washington��5

vii

3. 4.

Comparison of intrinsic potential model parameters used for Pacific salmon and steelhead intrinsic potential models��7 Reach average values of gradient, valley width index, bankfull width, and mean annual flow for targeted upper Skagit River streams in the United States by high, medium, or low intrinsic potential rank for Chinook salmon��9 Reach average values of gradient, valley width index, bankfull width, and mean annual flow for targeted upper Skagit River streams in Canada by high, medium, or low intrinsic potential rank for Chinook salmon��13 Reach average values of gradient, valley width index, bankfull width, and mean annual flow for targeted upper Skagit River streams in United States by high, medium, or low intrinsic potential rank for coho salmon��20 Reach average values of gradient, valley width index, bankfull width, and mean annual flow for targeted upper Skagit River streams in Canada by high, medium, or low intrinsic potential rank for coho salmon��24 Reach average values of gradient, valley width index, bankfull width, and mean annual flow for targeted upper Skagit River streams in the United States by high, medium, or low intrinsic potential rank for steelhead��33 Reach average values of gradient, valley width index, bankfull width, and mean annual flow for targeted upper Skagit River streams in Canada by high, medium, and low intrinsic potential rank for steelhead��37

Conversion Factors International System of Units to U.S. customary units

Multiply

To obtain

Length meter (m)

3.281

foot (ft)

kilometer (km)

0.6214

mile (mi)

square meter (m2)

0.0002471

acre

square meter (m2)

10.76

square foot (ft2)

square kilometer (km2)

0.3861

square mile (mi2)

Area

Datums Vertical coordinate information is referenced to the North American Vertical Datum of 1983 (NAVD 83). Horizontal coordinate information is referenced to the North American Datum of 1983 (NAD 83).

viii

Abbreviations CnH

Cooney and Holzer

DEM

digital elevation model

FERC

Federal Energy Regulatory Commission

GIS

geographic information system

intrinsic potential

licensing participants

rkm

river kilometer

PS-TRT

Puget Sound Technical Recovery Team

USGS

U.S. Geological Survey

Abstract

Introduction

We assessed habitat suitability for salmonids across selected tributaries upstream from three hydroelectric dams on the upper Skagit River in Whatcom County, northern Washington. We used NetMap, a commercial toolset within the ArcMap geographic information system (GIS), to analyze stream attributes based upon a synthetic stream channel network derived from digital elevation models. The GIS-derived stream attributes—including gradient, bankfull width, valley width index, elevation, and stream flow— allowed us to examine the spatial distribution and relative quality of spawning and rearing habitat for salmonids based on existing intrinsic potential (IP) models. As a first step, we created maps of potential anadromous fish distribution by identifying potential migration barriers within the synthetic stream network. Next, we applied a suite of existing IP models for steelhead, coho, and Chinook salmon (Oncorhynchus mykiss, O. kisutch, and O. tshawytscha, respectively) to estimate low, medium, and high IP habitat for each species. Three different IP models were used for each species, based on species preference curves from populations from coastal Oregon, northern California, Alaska, and western Washington. We found that at least 25 tributaries that were greater than third order and contained habitat with the potential for anadromous fish, totaling about 470 river kilometers in 4,453 synthetic stream reaches averaging about 100 meters (m) in length. The IP of each of these reaches was calculated and placed into low, medium, and high IP categories. For Chinook salmon, the only stream with significantly (in other words, greater than 1 kilometer [km]) high IP reaches was the upper Skagit River upstream from Ross Lake reservoir in Canada, upstream from the third dam in the hydroelectric system. There were differences among the three models evaluated, with the model derived for the lower Skagit River showing more high and medium IP habitat than the other two models that were designed for the Columbia River Basin. For coho salmon, all three models showed similar results favoring medium IP over low and high IP habitat. Of the 3 species examined with existing IP models, steelhead had the most habitat rated as high IP with 19 targeted tributaries showing greater than 1 km of high intrinsic potential habitat.

Seattle City Light (hereinafter City Light) is one of the Nation’s largest municipally owned utilities in terms of the number of customers served. City Light owns and operates the Skagit River Hydroelectric Project in Whatcom County, northern Washington, which produces about 20 percent of the power supply for the City of Seattle. The operating license issued by the Federal Energy Regulatory Commission (FERC) for the three dams in the project expires in 2025 and the process is underway to renew the license. A component of this relicensing process is the implementation of an integrated study plan to provide additional information on a suite of topics related to recreation, fisheries, cultural resources, wildlife, and vegetation. Included in the Revised Study Plan submitted to FERC was an interrelated set of studies (FA07– Reservoir Tributary Habitat; Seattle City Light, 2023) to assess habitat, environmental conditions, and species interactions that affect the availability and production potential of habitat for native salmonids (Oncorhynchus spp.) in reservoirs and major tributaries associated with Ross, Diablo, and Gorge dams on the Skagit River. The Skagit River Hydroelectric Project is in the upper Skagit River watershed (fig. 1) on the main stem of the Skagit River. The three hydroelectric dams and their reservoirs are located between river kilometer (rkm) 151 and 204 and work together through hydrological coordination to generate about 700 megawatts of power. Currently, the three hydroelectric dams block upstream and downstream migration of fish in the Skagit River. The dams and their reservoirs are contained within the Ross Lake National Recreation Area, which is administered by the National Park Service (NPS) as part of North Cascades National Park. The upstream-most 1.5 km section of the upper reservoir (Ross Lake) is in Canada and is contained within the Skagit Valley Provincial Park. Several major tributaries and about 50 km of the main stem of the Skagit River are located upstream from Ross Lake in Canada.

2 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington 121°30' W

121° W

120°30' W

EXPLANATION

it R ag

Cr.

HUC12 watershed boundary

Sna ss

o R.

Hydroelectric dam Potential anadromous fish distribution

r. ile C ysix M Twent

i l k wa R.

Mas elpa nik C

Kles

Reservoir

rg Fe

Su ma ll

Cr. on us

u po p e k Ne

Canada U.S. Cr.

Seattle 49° N

or k

. Cas

Cr ver

Cr.

tle F

Three Fool s Cr

e ittle B

Lightning C

49° N

Lu Big

Devil s Cr.

B eaver C r.

. Cr na . an Cr ill

t tl

Cr .

ta Ste

Cr . ROSS DAM

Ruby C r.

DIABLO DAM

Can

GORGE DAM

Cr .

t he r

te C ani Gr

C r. der Thu n

Skagit River

48°30' N 121°30' W

Base modified from U.S. Geological Survey data, vaious scales and dates. Projection: NAD1983, NAVD 1988.

121° W

120°30' W

0 0

10 10

20 MILES 20 KILOMETERS

Figure 1. Upper Skagit River watershed showing targeted and non-targeted tributary habitat with potential accessibility for anadromous salmonids (Oncorhynchus spp.) upstream from three currently impassable dams, in Whatcom County, northern Washington. USGS Hydrological Unit Code (HUC) is a classification hierarchy to delimit watershed boundaries in successively smaller hydrologic units. HUC12 indicates boundaries of local sub-watershed scale that represents tributary systems.

Introduction 3 Over the past several decades, City Light, NPS, and fisheries biologists from other agencies have gathered data on aquatic resources in the project reservoirs, including bull trout (Salvelinus confluentus) movement, fish species assemblages, sport fishing catches (creel surveys in Ross Lake, for example), water quality, and hydrodynamic modeling of the reservoirs. Although these efforts provide valuable baseline information, a comprehensive analysis of trophic interactions, environmental conditions, and the hydroelectric project operations that affect reservoir and tributary fish populations was identified by City Light to be a part of the Revised Study Plan for the relicensing application process. Such data could provide information necessary to evaluate the status of fish populations upstream from the dams, current effects that project operations have on these fish populations, and specific fishery monitoring and management activities into the future, if necessary, beyond the needs of the relicensing process. To address this information need, a multi-element study was initiated by the U.S. Geological Survey’s Western Fisheries Research Center in Seattle, Washington. The goal of the ongoing study was to identify and quantify factors that might limit recruitment or production of native adfluvial salmonids, which populate the reservoirs and associated tributaries upstream from the three main-stem dams on the Skagit River (Ross, Diablo, and Gorge dams). From 2018 to 2020, several related projects were underway. These projects focused on seasonal and size-structured food web interactions in the reservoirs, ontogenetic connections of adfluvial salmonids to tributary habitats, presence and geographic extent of native and non-native fishes in targeted reservoir tributaries in the basin, levels of hybridization among native and introduced species, and habitat suitability and bioenergetic growth potential of select tributaries. Food web interactions, distribution, and growth of native salmonids will ultimately be linked to environmental conditions, and these will be evaluated within the context of projected changes in climate or dam operations.

Subsequently, introduction of salmon or steelhead upstream from the dams was posed as a study topic by several parties to the relicensing project (licensing participants, hereinafter LPs). Because of the widespread installation of dams across the range of Pacific salmon, a significant percentage of historical spawning and rearing habitat for anadromous salmonids is inaccessible at locations where dams lack fish passage structures and block migratory fish access to historical spawning and rearing habitats upstream (Ward and Stanford, 1979; Waldman and Quinn, 2022). Where installation of fish passage structures such as fishways or fish ladders is not possible because of engineering or economic constraints, so called trap-and-haul operations have been used to (1) collect migrating adults and move them upstream (Clay and Eng, 2017) or downstream (Piper and others, 2020) past the dam, (2) collect migrating juveniles and move them past the dam (Johnson and others, 2005), or (3) move both adults and juveniles (referred to as two-way trap-and-haul; Lusardi and Moyle, 2017; Kock and others, 2021). Repatriating anadromous fish to formerly disconnected upstream areas could be useful in situations where habitat is limiting population growth in downstream areas while production potential exists upstream. It is also seen as a possible mitigation to future warming water temperatures in the face of climate change, given that colder temperature waters are expected to shift farther upstream under climate change scenarios (Naughton and others, 2018). To date, the efficacy of these approaches has seen mixed results (Lusardi and Moyle, 2017; Kock and others, 2021). Given the recent increase in projects to introduce or reintroduce anadromous salmonids upstream from passable barriers, it was determined that identifying the amount of suitable habitat in streams that could be used by Pacific salmon upstream from Gorge, Diablo, and Ross dams would be useful information to consider as part of the relicensing process.

4 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington

Purpose and Scope The purpose of this study was to (1) use watershed attributes estimated from synthetic stream networks derived from digital elevation models (DEMs) in a geographic information system (GIS) to identify the upstream limit of tributary habitat that could be accessed by introduced Pacific salmon and steelhead and (2) to use several existing intrinsic potential (IP) models for three species of anadromous salmonids (Chinook, coho, and steelhead [Oncorhynchus tshawytscha, O. kisutch, and O. mykiss, respectively]) to identify suitable habitat in tributaries and main-stem reaches upstream from the three Skagit River dams. Where IP models do not exist for a given species, use of results from other species could be used heuristically to infer the potential for a species’ performance.

Methodology We used a synthetic watershed approach (that is, a watershed with stream segments composed within a GIS based on DEM data; Benda and others, 2016) to assess the potential distribution of anadromous salmon in the Skagit River Basin upstream from the dams. NetMap (Terrainworks, 2023) is a commercial software system available as an add-on analysis toolkit for ArcGIS (ESRI, 2023) for use in watershed analyses (Benda and others, 2007). NetMap is an integrated set of numerical algorithms and base parameters that are used to create a synthetic stream network based on DEMs (Benda and others, 2007, 2011). Effectively, NetMap derives stream networks consisting of reaches ranging from 10 to 200 meters (m; average = 102 m ± 12 m; number of reaches [n] = 133,808 reaches to represent the study area of the upper Skagit River Basin). Within each reach, geomorphic and hydrological characteristics are derived based upon the surrounding DEM-derived landscape (such as drainage area and elevation), flow accumulation, and channel delineation processing. Geomorphic and hydrological habitat variables for each reach are estimated throughout the channel network using modeled parameters and empirically based relationships available from published region-specific models related to valley geometry, river network structure and hydrology, climate, and channel elevation profiles (Benda and others, 2011). At the start of the project, an existing NetMap model for the upper Skagit River watershed up to the Canadian border was based on a 10-m DEM. Terrainworks extended the model into the remaining watershed in Canada using a 20-m DEM, which was the highest resolution publicly available DEM. Terrainworks merged the two DEMs together to create a seamless DEM that could be packaged into NetMap as a dataset for further analysis and estimates of fish distribution and IP modeling. As part of the study plan review, LPs proposed a list of 21 streams with potential salmonid habitat that could potentially be used by introduced anadromous salmonids. On

the United States side of the border (table 1), these included Big Beaver Creek, Luna Creek, Little Beaver Creek, Canyon Creek, North Fork Canyon Creek, Granite Creek, Lightning Creek, Three Fools Creek, Castle Fork Creek, Ruby Creek, Cabinet Creek, and East Creek. Subsequent analyses and discussions with LPs added additional target tributaries, including a tributary to Gorge Lake reservoir (Stetattle Creek) and five tributaries to Ross Lake reservoir (Hozomeen Creek, Devils Creek, McMillan Creek, Panther Creek, and Slate Creek). In the Canadian part of the watershed upstream from the Ross Lake reservoir, the Skagit River main stem and several tributaries, including the Klesilkwa River, Sumallo River, Ferguson Creek, Nepopekum Creek, Maselpanik Creek, Snass Creek, Twentysix Mile Creek, and Marmotte Creek, were proposed for assessment (table 2). As a first analysis step, we used the hydrographic representation of the upper Skagit River study area in NetMap to identify the upper extent of potential anadromous salmonid distribution in each tributary upstream from the dams. We did this by applying two criteria related to fish passage: changes to gradient that would limit anadromous fish passage, and the presence of natural waterfalls and other water surface drops that could serve as anadromous fish barriers. Definitions of both criteria were obtained from the anadromous fish passage assessment protocols for the State of Washington (Washington Department of Fish and Wildlife, 2019). The first criterion to assess the extent of potential habitat was a maximum of 20-percent gradient over a 160-m reach, which studies have shown to be impassable for upstream migrating salmonids (Washington Department of Fish and Wildlife, 2019). Although some studies have used lower gradient thresholds (such as, 16 percent over 300 m for Fraser River salmonids; Finn and others, 2021), we used the standard for upstream migration barriers as set by the State of Washington. The second criterion was waterfalls or geomorphic features with vertical drops greater than or equal to 3.7 m. We assessed such drops using the NetMap node-based stream layer, which represents the synthetic stream network at the grain of the DEM that is 10 m for reaches in the United States and 20 m for reaches in Canada. NetMap identifies a node drop value for a reach based on differences in elevation between adjacent nodes. Potential upstream distribution was stopped at nodes that equaled or exceeded 3.7 m. In most cases, because of elevation changes, several nodes would exceed the drop height threshold and be obvious areas of steep gradient and vertical waterfalls or cascades. We cross-referenced known barriers and waterfalls with the node drop values in NetMap as verification that the tool was accurately depicting known barriers. Because of the remote nature of the watershed and few existing on-the-ground surveys of potential barriers, most of the potential barriers stopping upstream anadromous fish distribution were identified only from the NetMap environment. However, there were a few cases where single or small clusters of adjacent nodes that exceeded the height threshold were surrounded by several hundred meters of habitat upstream and downstream from the impacted reach.

Methodology 5 Table 1. Tributary and main-stem reaches upstream from Ross Lake reservoir identified by stakeholders as having accessible fish habitat that could potentially be used by introduced Pacific salmon and steelhead (Oncorhynchus spp.), with estimates of the amount and average percent gradient of the accessible sections, in northern Washington. [km, kilometer; <, less than]

Reach estimates River/stream name Skagit River Klesilkwa River Sumallo River Ferguson Creek Nepopekum Creek

Reach description

Length (km)

Gradient (percent)

Ross Lake to Klesilkwa

17.5

Klesilkwa to barrier falls near Snass Creek

17.1

Skagit River to Silverhope Divide

14.3

Silverhope Divide to end of third-order stream

3.7

Skagit River to Ferguson Creek

16.6

Ferguson Creek to end of third-order stream

12.1

Sumallo River to Highway 3

3.9

Highway 3 to end of third-order stream

3.7

Skagit River to start of canyon

2.7

Start of canyon to Poland Creek

9.3

Maselpanik Creek

Klesilkwa River to end of third-order stream

12.2

Snass Creek

Skagit River to Dry Lake

3.9

Twentysix Mile Creek

Skagit River to end of third-order stream

5.8

Marmotte Creek

Skagit River to end of third-order stream

4.3

Table 2. Ross Lake reservoir tributary reaches identified by stakeholders as having accessible fish habitat that could potentially be used by introduced Pacific salmon and steelhead (Oncorhynchus spp.), with estimates of the amount (in kilometers) and average percent gradient of the accessible sections, in northern Washington. [km, kilometer; <, less than]

Reach estimates River/stream name

Reach description

Length (km)

Gradient (percent)

Big Beaver Creek

Ross Lake to McMillan Creek

14.6

Little Beaver Creek

Ross Lake to end of third-order stream

24.2

Ruby Creek

Ross Lake to confluence of Canyon and Granite Creeks

5.5

Canyon Creek

Ruby Creek to Slate Creek

11.9

Slate Creek to barrier

4.2

North Fork Canyon Creek

Canyon Creek to barrier

1.0

Granite Creek

Ruby Creek to barrier

8.6

Lightning Creek

Ross Lake to Three Fools Creek

3.5

Three Fools Creek to Freezeout Creek

8.8

Freezeout Creek to Boundary Creek

6.3

Luna Creek

Big Beaver Creek to end of third-order stream

4.5

Three Fools Creek

Lightning Creek to Castle Fork Creek

10.1

Castle Fork Creek

Three Fools Creek to Rustle Creek

5.8

East Creek

Granite Creek to end of third-order stream

6.9

Cabinet Creek

Granite Creek to end of third-order stream

3.2

6 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington To be conservative with identifying potential barriers, in these cases we by-passed such “spurious nodes,” which had drops ranging from 3.7 to 8.8 m, and assumed that fish passage was possible. There were only a few such cases in Canyon Creek (one node), Castle Fork Creek (six nodes in three reaches), Devils Creek (eight nodes in four reaches), Granite Creek (three nodes in two reaches), Lightning Creek (two nodes in one reach), McMillan Creek (two nodes in one reach), Nepopekum Creek (eight nodes in five reaches), and Snass Creek (eight nodes in five reaches). Given the above two decision rules, we identified the potential areas of anadromous fish distribution within GIS for use in IP modeling (fig 1). For the next step, we used several existing species-specific IP models to estimate the geomorphically based habitat suitability for Chinook, coho, and steelhead. The concept of IP was originally developed in the Pacific Northwest for use as a broad-scale assessment tool to estimate the potential for stream reaches to support salmonids based on species habitat preferences related to persistent landscape features that are not easily modified by human disturbances (Burnett and others, 2007; Sheer and others, 2009). Features, such as hydrology, gradient, and channel width, create the template upon which other transient features, like the presence of riffle and pool channel units and large woody debris, interact to create habitat that support salmonid populations. The influence of these landscape features creating the template are determined by the geomorphic context of landforms interacting with other habitat forming processes, such as hydrology and wood and sediment recruitment and retention, to create habitat conditions that are suitable for spawning and rearing of salmon (Benda and others, 2004; Burnett and others, 2007; Bidlack and others, 2014). As detailed below, there were two approaches to calculate the IP depending on the model and species. The first approach was the original one described by Burnett and others (2007) where IP for each reach is calculated based upon species suitability curves for each of three habitat variables so that for each reach, habitat scores for the species would range from 1 (highest preference) to 0 (not suitable). An IP score based on the geometric mean from the three variables of interest (for example, gradient, valley confinement, and width) is calculated for each reach, ranging from 1 to 0. The

second approach was a categorically based one, where low, medium, and high IP values were assigned to each reach according to ranges of values for the 3 variables of interest. For example, reaches with a bankfull width between 3.8 and 25 m, a gradient from 0 to 0.5 percent, and a moderate valley width ratio would receive a relative IP score of “medium.” To compare model-derived IP scores and maps across species, we assigned reaches from numerically derived models as low (score less than or equal to 0.25), medium (greater than 0.25 or less than or equal to 0.75), and high (greater than 0.75) IP. All calculations were made in GIS using the NetMap tool applied to the fish distribution described above. For each target species of salmonid (Chinook salmon, coho salmon, steelhead), we used three existing models to score the IP of each stream reach in the areas identified as having potential for introduced anadromous salmonids (fig. 1). For Chinook salmon, we used existing models from Connor and others (2015; hereinafter Connor model), Busch and others (2013; hereinafter Busch model), and Cooney and Holzer (2006; hereinafter Cooney and Holzer model) that were developed from populations in the interior Columbia River Basin, the lower Columbia River Basin, and the Skagit River downstream from the Skagit River Hydroelectric Project, respectively. The Connor model used elevation as a factor in place of bankfull width. For coho salmon, existing models by Burnett and others (2007; hereinafter Burnett model), Agrawal and others (2005; hereinafter Agrawal model), and Romey (Terrainworks, 2023; hereinafter Romey model) were based on species preference curves developed from populations from coastal Oregon, northern California, and Alaska, respectively. All three models used the same three physical habitat variables (percent gradient, mean annual flow, and valley width index; table 3), with differences among the models in the shape of the habitat suitability curves that assigned slightly different IP ranking scores to each attribute. For steelhead, we used the Burnett model, the Agrawal model, and the model from the Puget Sound Technical Recovery Team (Hard and others, 2015; hereinafter Puget Sound Technical Recovery Team model) based on populations in coastal Oregon, northern California, and Puget Sound, respectively.

Results and Interpretations 7 Table 3. Comparison of intrinsic potential model parameters used for Pacific salmon and steelhead (Oncorhynchus spp.) intrinsic potential models. Variables Model1

Gradient

Bankfull

Valley width index

Elevation

Streamflow

Yes

Chinook Connor

Yes

Cooney and Holzer1

Yes

Busch

Yes

Agrawal

Yes

Romey

Yes

Burnett

Yes

Coho

Steelhead Burnett

Yes

Agrawal

Yes

Puget Sound Technical Recovery Team2

Yes

1Named models include Connor model (Connor and others, 2015), Cooney and Holzer model (Cooney and Holzer, 2006), Busch model (Busch and others 2013), Agrawal model (Agrawal and others, 2005), Romey model (Terrainworks, 2023), Burnett model (Burnett and others, 2007), and Puget Sound Technical Recovery Team model (Hard and others, 2015). 2Categorical based IP scoring (none, low, medium, or high for Cooney and Holzer model; low, medium, or high for Puget Sound Technical Recovery Team model) based on combined ranges of parameter values.

Results and Interpretations Potential Fish Distribution Upstream from the Skagit River Hydroelectric Project Dams Of the 27 proposed targeted streams, 25 of them were included in the final analysis based on potential fish distribution. Two proposed streams, East and Cabinet Creeks, were identified to have several reach nodes exceeding the fish passage thresholds within about 200 m or less from the confluence with their respective tributaries and were not considered for further analysis. Based upon gradient threshold and waterfall drop values for the remaining 25 target streams, we identified 4,453 stream reaches totaling an estimated length of 470 rkm of potential fish habitat for migratory fish species (fig. 1). The average length of reaches was 105 m (standard deviation; SD ± 13.8 m, range 9–227) and the median stream order of these reaches being was 4. These included 2,838 reaches within 25 target streams (tables 1 and 2 plus an additional four streams) totaling 306 km with a median stream order of 5. There were 1,615 reaches totaling 164 km that were found in other accessible tributaries that were often smaller (median stream order was 2) named or unnamed tributaries to the targeted streams (hereinafter referred to as non-targeted streams or non-targeted reaches). The targeted tributary and

main-stem Skagit River reaches identified were larger streams, and the other non-targeted reaches were smaller and generally shorter reaches of accessible habitat.

Chinook Salmon Intrinsic Potential Model Results There were similarities and differences among the results for the three Chinook salmon IP models (fig. 2). The Chinook salmon IP model proposed by Connor and others (2015; Connor model), which was parameterized based on other Skagit River Basin populations downstream from the Skagit River Hydroelectric Project, was different from the Busch model (Busch and others, 2013) and the Cooney and Holzer model (Cooney and Holzer, 2006) developed for Chinook salmon populations in the Columbia River Basin. All three models used gradient and stream bankfull width habitat preference curves, whereas the Connor IP model replaced valley width ratio used by the other two models with elevation as the third parameter (table 3). In total, the Connor model rated more reaches as having high IP habitat than the other two models. Summaries of habitat parameter values and the number of reaches scored as low, medium, and high IP sections are provided for main-stem and tributary habitat in the United States (table 4) and Canada (table 5).

8 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington Skagit River Sumallo River Little Beaver Creek Big Beaver Creek Granite Creek Lightning Creek Panther Creek Klesilkwa River Canyon Creek Devils Creek Nepopekum Creek Three Fools Creek Castle Fork Creek Stetattle Creek McMillan Creek Ruby Creek Ferguson Creek Snass Creek Luna Creek Maselpanik Creek Slate Creek Twentysix Mile Creek North Fork Canyon Creek Marmotte Creek Hozomeen Creek Skagit River Sumallo River Little Beaver Creek Big Beaver Creek Granite Creek Lightning Creek Panther Creek Klesilkwa River Canyon Creek Devils Creek Nepopekum Creek Three Fools Creek Castle Fork Creek Stetattle Creek McMillan Creek Ruby Creek Ferguson Creek Snass Creek Luna Creek Maselpanik Creek Slate Creek Twentysix Mile Creek North Fork Canyon Creek Marmotte Creek Hozomeen Creek

EXPLANATION High intrinsic potential Medium intrinsic potential Low intrinsic potential Species: Chinook salmon Model: Connor

EXPLANATION High intrinsic potential Medium intrinsic potential Low intrinsic potential Species: Chinook salmon Model: Busch

Skagit River Sumallo River Little Beaver Creek Big Beaver Creek Granite Creek Lightning Creek Panther Creek Klesilkwa River Canyon Creek Devils Creek Nepopekum Creek Three Fools Creek Castle Fork Creek Stetattle Creek McMillan Creek Ruby Creek Ferguson Creek Snass Creek Luna Creek Maselpanik Creek Slate Creek Twentysix Mile Creek North Fork Canyon Creek Marmotte Creek Hozomeen Creek

EXPLANATION High intrinsic potential Medium intrinsic potential Low intrinsic potential Species: Chinook salmon Model: Cooney and Holzer

15 20 25 30 Chinook salmon intrinsic potential, in river kilometers

Figure 2. Length of high, medium, and low intrinsic potential habitat for Chinook salmon (Oncorhynchus tshawytscha) based on three different models for target tributaries upstream from the three upper Skagit River dams, in northern Washington.

Results and Interpretations 9 Table 4. Reach average values of gradient, valley width index (ratio), bankfull width, and mean annual flow for targeted upper Skagit River streams in the United States by high, medium, or low intrinsic potential (IP) rank for Chinook salmon (Oncorhynchus tshawytscha). [Plus or minus (±) standard deviations are shown in parentheses. CnH, Cooney and Holzer model; NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Model

Number of reaches

Big Beaver Creek

Connor

Big Beaver Creek

Busch

Big Beaver Creek

CnH

High

Big Beaver Creek

Connor

230

Medium

1.3 (±2.1)

0.6 (±0.3)

Big Beaver Creek

Busch

Medium

Big Beaver Creek

CnH

113

Medium

0.1 (±0.1)

0.5 (±0.1)

Big Beaver Creek

Connor

Low

Big Beaver Creek

Busch

230

Low

1.3 (±2.1)

Big Beaver Creek

CnH

117

Low

2.4 (±2.5)

Little Beaver Creek

Connor

High

Little Beaver Creek

Busch

High

Little Beaver Creek

CnH

High

Little Beaver Creek

Connor

231

Medium

1.8 (±1.4)

0.7 (±0.2)

12.1 (±2.5)

3.2 (±1.4)

Little Beaver Creek

Busch

Medium

2.1 (±1.0)

1.0 (±0.0)

8.3 (±1.6)

1.4 (±0.7)

Little Beaver Creek

CnH

Medium

0.2 (±0.2)

0.6 (±0.1)

13.8 (±1.0)

4.1 (±0.7)

Little Beaver Creek

Connor

Low

8.6 (±1.1)

0.9 (±0.3)

8.7 (±3.9)

1.8 (±1.9)

Little Beaver Creek

Busch

230

Low

2.1 (±1.9)

0.7 (±0.2)

12.1 (±2.6)

3.2 (±1.4)

Little Beaver Creek

CnH

202

Low

2.4 (±1.8)

0.7 (±0.2)

11.6 (±2.7)

3.0 (±1.5)

Ruby Creek

Connor

High

Ruby Creek

Busch

High

Ruby Creek

CnH

High

Ruby Creek

Connor

Medium

1.5 (±0.5)

0.5 (±0.2)

19.9 (±1.4)

9.2 (±1.4)

Ruby Creek

Busch

Medium

1.4 (±0.3)

1.2 (±0.2)

19.8 (±0.2)

9.0 (±0.2)

Ruby Creek

CnH

Medium

0.1 (±0.1)

0.3 (±0.1)

22.2 (±0.1)

11.6 (±0.1)

Ruby Creek

Connor

Low

Ruby Creek

Busch

Low

1.5 (±0.5)

0.4 (±0.1)

19.9 (±1.4)

9.2 (±1.5)

Ruby Creek

CnH

Low

1.6 (±0.4)

0.5 (±0.2)

19.8 (±1.3)

9.1 (±1.4)

Canyon Creek

Connor

High

Canyon Creek

Busch

High

Canyon Creek

CnH

High

Canyon Creek

Connor

148

Medium

2.8 (±1.6)

0.8 (±0.4)

12.0 (±2.4)

3.1 (±1.2)

Canyon Creek

Busch

Medium

2.2 (±0.6)

1.1 (±0.0)

10.6 (±2.3)

2.4 (±1.2)

Canyon Creek

CnH

Medium

0.1 (±0.1)

0.7 (±0.0)

12.4 (±1.0)

3.2 (±0.6)

Canyon Creek

Connor

Low

8.3 (±0.4)

1.9 (±0.6)

6.5 (±0.1)

0.8 (±0.0)

Canyon Creek

Busch

149

Low

3.2 (±2.0)

0.8 (±0.5)

11.8 (±2.7)

3.1 (±1.3)

Canyon Creek

CnH

152

Low

3.2 (±2.0)

0.8 (±0.5)

11.7 (±2.7)

3.0 (±1.3)

North Fork Canyon Creek

Connor

High

North Fork Canyon Creek

Busch

High

Stream

Percent gradient

Valley width ratio

Bankfull width (m)

Mean annual flow (m3/s)

High

IP rank

14.1 (±2.7)

4.4 (±1.5)

15.7 (±1.5)

5.5 (±0.8)

0.6 (±0.3)

14.1 (±2.7)

4.4 (±1.5)

0.7 (±0.3)

12.5 (±2.7)

3.5 (±1.4)

10 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington Table 4. Reach average values of gradient, valley width index (ratio), bankfull width, and mean annual flow for targeted upper Skagit River streams in the United States by high, medium, or low intrinsic potential (IP) rank for Chinook salmon (Oncorhynchus tshawytscha).—Continued [Plus or minus (±) standard deviations are shown in parentheses. CnH, Cooney and Holzer model; NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Stream

Model

Number of reaches

IP rank

Percent gradient

Valley width ratio

Bankfull width (m)

Mean annual flow (m3/s)

North Fork Canyon Creek

CnH

High

North Fork Canyon Creek

Connor

Medium

6.4 (NA)

1.0 (NA)

North Fork Canyon Creek

Busch

Medium

North Fork Canyon Creek

CnH

Medium

North Fork Canyon Creek

Connor

Low

9.4 (±1.5)

1.6 (±0.4)

6.3 (±0.0)

0.7 (±0.0)

North Fork Canyon Creek

Busch

Low

9.0 (±1.8)

1.5 (±0.5)

6.4 (±0.0)

0.7 (±0.0)

North Fork Canyon Creek

CnH

Low

9.0 (±1.8)

1.5 (±0.5)

6.4 (±0.0)

0.7 (±0.0)

Granite Creek

Connor

High

Granite Creek

Busch

High

Granite Creek

CnH

High

Granite Creek

Connor

215

Medium

2.9 (±1.5)

0.8 (±0.3)

10.5 (±1.7)

2.3 (±0.8)

Granite Creek

Busch

Medium

1.8 (±0.7)

1.3 (±0.4)

9.1 (±1.2)

1.6 (±0.5)

Granite Creek

CnH

Medium

0.4 (±0.0)

0.9 (±0.2)

8.9 (±0.2)

1.5 (±0.1)

Granite Creek

Connor

Low

Granite Creek

Busch

176

Low

3.1 (±1.5)

0.7 (±0.2)

10.9 (±1.6)

2.5 (±0.8)

Granite Creek

CnH

213

Low

2.9 (±1.5)

0.8 (±0.3)

10.6 (±1.7)

2.3 (±0.8)

Lightning Creek

Connor

High

Lightning Creek

Busch

High

Lightning Creek

CnH

High

Lightning Creek

Connor

190

Medium

2.8 (±1.7)

0.8 (±0.4)

11.3 (±2.8)

2.8 (±1.7)

Lightning Creek

Busch

Medium

1.5 (±1.4)

1.5 (±0.7)

8.7 (±1.6)

1.5 (±0.7)

Lightning Creek

CnH

Medium

0.0 (±0.0)

1.4 (±0.6)

7.6 (±0.4)

1.1 (±0.1)

Lightning Creek

Connor

Low

8.6 (±1.1)

1.3 (±0.2)

7.7 (±0.5)

1.1 (±0.2)

Lightning Creek

Busch

178

Low

3.2 (±1.9)

0.8 (±0.3)

11.5 (±2.9)

2.9 (±1.7)

Lightning Creek

CnH

189

Low

3.1 (±1.9)

0.8 (±0.4)

11.3 (±2.8)

2.8 (±1.7)

Luna Creek

Connor

High

Luna Creek

Busch

High

Luna Creek

CnH

High

Luna Creek

Connor

Medium

2.4 (±1.9)

0.8 (±0.1)

8.9 (±0.3)

1.5 (±0.1)

Luna Creek

Busch

Medium

2.6 (±0.5)

1.0 (±0.0)

8.6 (±0.1)

1.4 (±0.0)

Luna Creek

CnH

Medium

0.2 (±0.2)

0.8 (±0.1)

9.1 (±0.2)

1.6 (±0.1)

Luna Creek

Connor

Low

7.7 (NA)

0.9 (NA)

8.1 (NA)

1.2 (NA)

Luna Creek

Busch

Low

2.5 (±2.1)

0.8 (±0.1)

8.9 (±0.3)

1.5 (±0.1)

Luna Creek

CnH

Low

2.9 (±2.0)

0.9 (±0.1)

8.9 (±0.3)

1.5 (±0.1)

6.4 (NA)

0.7 (NA)

Results and Interpretations 11 Table 4. Reach average values of gradient, valley width index (ratio), bankfull width, and mean annual flow for targeted upper Skagit River streams in the United States by high, medium, or low intrinsic potential (IP) rank for Chinook salmon (Oncorhynchus tshawytscha).—Continued [Plus or minus (±) standard deviations are shown in parentheses. CnH, Cooney and Holzer model; NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Model

Number of reaches

Three Fools Creek

Connor

Three Fools Creek

Busch

Three Fools Creek

CnH

High

Three Fools Creek

Connor

102

Medium

3.7 (±1.6)

0.9 (±0.4)

10.6 (±1.5)

2.3 (±0.7)

Three Fools Creek

Busch

Medium

2.5 (±1.0)

1.2 (±0.2)

9.9 (±2.1)

2.1 (±0.9)

Three Fools Creek

CnH

Medium

0.2 (NA)

0.9 (NA)

9.9 (NA)

1.9 (NA)

Three Fools Creek

Connor

Low

9.7 (±3.8)

1.0 (±0.6)

9.8 (±4.3)

2.2 (±1.6)

Three Fools Creek

Busch

Low

4.0 (±2.0)

0.9 (±0.4)

10.6 (±1.0)

2.3 (±0.7)

Three Fools Creek

CnH

104

Low

3.9 (±1.9)

0.9 (±0.4)

10.6 (±1.6)

2.3 (±0.7)

Castle Fork Creek

Connor

High

Castle Fork Creek

Busch

High

Castle Fork Creek

CnH

High

Castle Fork Creek

Connor

Medium

4.8 (±1.6)

1.9 (±1.0)

5.7 (±1.1)

0.6 (±0.2)

Castle Fork Creek

Busch

Medium

2.1 (±0.4)

1.7 (±0.5)

6.3 (±0.2)

0.7 (±0.0)

Castle Fork Creek

CnH

Medium

Castle Fork Creek

Connor

Low

9.3 (±2.1)

2.3 (±0.6)

4.0 (±1.3)

0.3 (±0.2)

Castle Fork Creek

Busch

Low

7.6 (±2.6)

2.1 (±0.9)

4.7 (±1.4)

0.4 (±0.3)

Castle Fork Creek

CnH

Low

7.1 (±2.9)

2.1 (±0.9)

4.8 (±1.4)

0.4 (±0.3)

Panther Creek

Connor

High

Panther Creek

Busch

High

Panther Creek

CnH

High

Panther Creek

Connor

176

Medium

4.1 (±1.8)

1.2 (±0.9)

8.5 (±2.2)

1.5 (±0.8)

Panther Creek

Busch

Medium

2.7 (±0.5)

1.6 (±0.6)

8.1 (±1.2)

1.3 (±0.4)

Panther Creek

CnH

Medium

0.2 (±0.2)

0.9 (±0.0)

9.0 (±0.0)

1.6 (±0.0)

Panther Creek

Connor

Low

7.9 (±0.7)

1.4 (±0.7)

6.9 (±2.1)

1.0 (±0.7)

Panther Creek

Busch

167

Low

4.6 (±2.0)

1.2 (±0.9)

8.5 (±2.3)

1.5 (±0.8)

Panther Creek

CnH

188

Low

4.4 (±1.9)

1.2 (±0.9)

8.4 (±2.2)

1.5 (±0.8)

Devils Creek

Connor

High

Devils Creek

Busch

High

Devils Creek

CnH

High

Devils Creek

Connor

126

Medium

4.4 (±1.2)

1.2 (±0.5)

7.5 (±1.4)

1.1 (±0.4)

Devils Creek

Busch

Medium

2.6 (±0.2)

1.3 (±0.1)

7.0 (±0.2)

0.9 (±0.1)

Devils Creek

CnH

Medium

0.3 (NA)

0.8 (NA)

9.0 (NA)

1.6 (NA)

Devils Creek

Connor

Low

8.5 (±1.1)

1.4 (±0.6)

6.9 (±2.7)

1.0 (±0.7)

Devils Creek

Busch

121

Low

5.0 (±1.5)

1.3 (±0.5)

7.5 (±1.6)

1.1 (±0.4)

Devils Creek

CnH

135

Low

4.7 (±1.6)

1.3 (±0.5)

7.4 (±1.6)

1.1 (±0.4)

Stetattle Creek

Connor

High

Stetattle Creek

Busch

High

Stetattle Creek

CnH

High

Stetattle Creek

Connor

Medium

4.0 (±1.9)

1.2 (±1.5)

9.8 (±2.3)

2.0 (±0.8)

Stream

Percent gradient

Valley width ratio

Bankfull width (m)

Mean annual flow (m3/s)

High

IP rank

12 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington Table 4. Reach average values of gradient, valley width index (ratio), bankfull width, and mean annual flow for targeted upper Skagit River streams in the United States by high, medium, or low intrinsic potential (IP) rank for Chinook salmon (Oncorhynchus tshawytscha).—Continued [Plus or minus (±) standard deviations are shown in parentheses. CnH, Cooney and Holzer model; NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Stream

Model

Number of reaches

IP rank

Percent gradient

Valley width ratio

Bankfull width (m)

Mean annual flow (m3/s)

Stetattle Creek

Busch

Medium

1.3 (±1.1)

1.2 (±0.2)

10.6 (±0.9)

2.3 (±0.4)

Stetattle Creek

CnH

Medium

0.2 (±0.0)

1.0 (±0.1)

10.3 (±0.0)

2.1 (±0.0)

Stetattle Creek

Connor

Low

8.7 (±2.4)

1.0 (±0.1)

11.2 (±0.6)

2.5 (±0.3)

Stetattle Creek

Busch

Low

4.6 (±2.4)

1.2 (±1.5)

9.9 (±2.2)

2.1 (±0.8)

Stetattle Creek

CnH

Low

4.5 (±2.4)

1.2 (±1.5)

9.9 (±2.2)

2.1 (±0.8)

McMillan Creek

Connor

High

McMillan Creek

Busch

High

McMillan Creek

CnH

High

0.7 (±0.6)

26.3 (±7.7)

6.9 (±0.0)

0.9 (±0.0)

McMillan Creek

Connor

Medium

3.1 (±1.8)

3.7 (±7.3)

8.2 (±1.0)

1.3 (±0.3)

McMillan Creek

Busch

Medium

1.7 (±1.3)

5.6 (±9.3)

8.0 (±1.0)

1.2 (±0.3)

McMillan Creek

CnH

Medium

0.2 (±0.2)

1.1 (±0.2)

7.7 (±0.5)

1.1 (±0.2)

McMillan Creek

Connor

Low

9.5 (±2.4)

2.1 (±1.4)

7.5 (±1.9)

1.1 (±0.6)

McMillan Creek

Busch

Low

4.6 (±2.6)

2.7 (±5.6)

8.2 (±1.1)

1.3 (±0.4)

McMillan Creek

CnH

Low

4.2 (±2.5)

2.8 (±5.3)

8.2 (±1.1)

1.3 (±0.4)

Slate Creek

Connor

High

Slate Creek

Busch

High

Slate Creek

CnH

High

Slate Creek

Connor

Medium

5.7 (±0.8)

1.1 (±0.2)

Slate Creek

Busch

Medium

Slate Creek

CnH

Medium

Slate Creek

Connor

Low

8.2 (±0.9)

Slate Creek

Busch

Low

6.6 (±1.5)

Slate Creek

CnH

Low

Hozomeen Creek

Connor

Hozomeen Creek

Busch

Hozomeen Creek

CnH

Hozomeen Creek

Connor

Hozomeen Creek Hozomeen Creek

8.6 (±0.1)

1.4 (±0.0)

1.3 (±0.3)

8.6 (±0.0)

1.4 (±0.0)

1.2 (±0.2)

8.6 (±0.0)

1.4 (±0.0)

6.6 (±1.5)

1.2 (±0.2)

8.6 (±0.0)

1.4 (±0.0)

High

Medium

Busch

Medium

CnH

Medium

Hozomeen Creek

Connor

Low

12.8 (NA)

1.7 (NA)

4.9 (NA)

0.4 (NA)

Hozomeen Creek

Busch

Low

12.8 (NA)

1.7 (NA)

4.9 (NA)

0.4 (NA)

Hozomeen Creek

CnH

Low

12.8 (NA)

1.7 (NA)

4.9 (NA)

0.4 (NA)

Results and Interpretations 13 Table 5. Reach average values of gradient, valley width index (ratio), bankfull width, and mean annual flow for targeted upper Skagit River streams in Canada by high, medium, or low intrinsic potential (IP) rank for Chinook salmon (Oncorhynchus tshawytscha). [Plus or minus (±) standard deviations are shown in parentheses. CnH, Cooney and Holzer model; NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Stream

Model

Number of reaches

IP rank

Percent gradient

Valley width ratio

Bankfull width (m)

Mean annual flow (m3/s)

Skagit River

Connor

314

High

0.5 (±0.7)

3.2 (±10.5)

23.3 (±2.6)

13.2 (±3.1)

Skagit River

Busch

High

0.3 (±0.4)

12.8 (±24.0)

25.0 (±2.3)

15.4 (±3.0)

Skagit River

CnH

High

0.2 (±0.2)

4.8 (±1.3)

21.3 (±3.3)

11.0 (±3.6)

Skagit River

Connor

Medium

0.6 (±0.5)

1.1 (±0.8)

18.0 (±2.4)

7.5 (±2.1)

Skagit River

Busch

170

Medium

0.5 (±0.7)

1.6 (±0.6)

22.4 (±3.2)

12.3 (±3.5)

Skagit River

CnH

178

Medium

0.1 (±0.2)

4.0 (±13.6)

22.7 (±2.7)

12.5 (±3.1)

Skagit River

Connor

Low

Skagit River

Busch

139

Low

0.6 (±0.8)

0.8 (±0.2)

21.9 (±2.9)

11.6 (±3.2)

Skagit River

CnH

176

Low

0.9 (±0.8)

1.7 (±2.9)

22.5 (±3.5)

12.4 (±3.9)

Klesilkwa River

Connor

High

Klesilkwa River

Busch

High

Klesilkwa River

CnH

High

Klesilkwa River

Connor

166

Medium

0.5 (±1.1)

0.7 (±0.2)

11.0 (±2.7)

2.6 (±1.3)

Klesilkwa River

Busch

Medium

0.4 (±0.7)

1.1 (±0.0)

7.9 (±0.5)

1.2 (±0.2)

Klesilkwa River

CnH

127

Medium

0.1 (±0.1)

0.7 (±0.2)

11.2 (±2.6)

2.8 (±1.3)

Klesilkwa River

Connor

Low

10.5 (±1.9)

1.2 (±0.0)

7.3 (±0.0)

1.0 (±0.0)

Klesilkwa River

Busch

158

Low

0.8 (±2.0)

0.7 (±0.2)

11.1 (±2.7)

2.7 (±1.3)

Klesilkwa River

CnH

Low

2.7 (±3.0)

0.8 (±0.2)

9.7 (±2.8)

2.1 (±1.4)

Sumallo River

Connor

High

Sumallo River

Busch

High

Sumallo River

CnH

High

Sumallo River

Connor

245

Medium

1.3 (±1.6)

0.7 (±0.2)

11.0 (±2.7)

2.7 (±1.2)

Sumallo River

Busch

Medium

2.0 (±0.8)

1.2 (±0.1)

6.8 (±0.5)

0.8 (±0.2)

Sumallo River

CnH

118

Medium

0.1 (±0.1)

0.6 (±0.1)

12.7 (±1.3)

3.4 (±0.8)

Sumallo River

Connor

Low

7.5 (±0.1)

1.4 (±0.1)

5.4 (±0.0)

0.5 (±0.0)

Sumallo River

Busch

224

Low

1.3 (±1.8)

0.7 (±0.2)

11.4 (±2.4)

2.8 (±1.2)

Sumallo River

CnH

129

Low

2.6 (±1.6)

0.9 (±0.3)

9.4 (±2.6)

1.9 (±1.2)

Ferguson Creek

Connor

High

Ferguson Creek

Busch

High

Ferguson Creek

CnH

High

Ferguson Creek

Connor

Medium

0.6 (±1.2)

1.3 (±0.2)

5.6 (±0.6)

0.6 (±0.1)

Ferguson Creek

Busch

Medium

0.2 (±0.4)

1.3 (±0.2)

5.9 (±0.5)

0.6 (±0.1)

Ferguson Creek

CnH

Medium

0.0 (±0.1)

1.3 (±0.2)

5.7 (±0.5)

0.6 (±0.1)

Ferguson Creek

Connor

Low

7.5 (±0.6)

1.6 (±0.1)

4.6 (±0.0)

0.4 (±0.0)

Ferguson Creek

Busch

Low

3.2 (±3.0)

1.5 (±0.3)

4.9 (±0.5)

0.4 (±0.2)

Ferguson Creek

CnH

Low

3.3 (±2.8)

1.5 (±0.3)

5.3 (±0.9)

0.5 (±0.2)

Nepopekum Creek

Connor

High

Nepopekum Creek

Busch

High

Nepopekum Creek

CnH

High

Nepopekum Creek

Connor

107

Medium

2.6 (±2.0)

0.9 (±0.1)

8.7 (±0.2)

1.5 (±0.1)

Nepopekum Creek

Busch

Medium

1.5 (±1.0)

1.1 (±0.1)

8.8 (±0.2)

1.5 (±0.1)

14 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington Table 5. Reach average values of gradient, valley width index (ratio), bankfull width, and mean annual flow for targeted upper Skagit River streams in Canada by high, medium, or low intrinsic potential (IP) rank for Chinook salmon (Oncorhynchus tshawytscha).— Continued [Plus or minus (±) standard deviations are shown in parentheses. CnH, Cooney and Holzer model; NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Stream

Model

Number of reaches

IP rank

Percent gradient

Valley width ratio

Bankfull width (m)

Mean annual flow (m3/s)

Nepopekum Creek

CnH

Medium

0.1 (±0.1)

0.9 (±0.1)

8.9 (±0.2)

1.5 (±0.1)

Nepopekum Creek

Connor

Low

9.0 (±1.4)

0.9 (±0.1)

8.5 (±0.2)

1.4 (±0.1)

Nepopekum Creek

Busch

Low

3.3 (±2.7)

0.9 (±0.1)

8.7 (±0.3)

1.5 (±0.1)

Nepopekum Creek

CnH

Low

3.7 (±2.4)

0.9 (±0.1)

8.7 (±0.2)

1.4 (±0.1)

Maselpanik Creek

Connor

High

Maselpanik Creek

Busch

High

Maselpanik Creek

CnH

High

Maselpanik Creek

Connor

Medium

2.7 (±1.7)

0.7 (±0.1)

10.4 (±0.2)

2.2 (±0.1)

Maselpanik Creek

Busch

Medium

Maselpanik Creek

CnH

Medium

Maselpanik Creek

Connor

Low

8.5 (±1.2)

0.7 (±0.2)

10.5 (±0.2)

2.2 (±0.1)

Maselpanik Creek

Busch

Low

5.1 (±3.3)

0.7 (±0.1)

10.4 (±0.2)

2.2 (±0.1)

Maselpanik Creek

CnH

Low

5.1 (±3.3)

0.7 (±0.1)

10.4 (±0.2)

2.2 (±0.1)

Snass Creek

Connor

High

Snass Creek

Busch

High

Snass Creek

CnH

High

Snass Creek

Connor

Medium

3.7 (±1.9)

1.3 (±0.3)

6.6 (±0.8)

0.8 (±0.2)

Snass Creek

Busch

Medium

1.6 (±0.8)

1.3 (±0.2)

6.7 (±0.8)

0.8 (±0.2)

Snass Creek

CnH

Medium

0.2 (NA)

1.6 (NA)

5.1 (NA)

0.5 (NA)

Snass Creek

Connor

Low

9.6 (±1.6)

1.0 (±0.2)

6.9 (±0.9)

0.9 (±0.2)

Snass Creek

Busch

Low

6.4 (±2.8)

1.2 (±0.3)

6.7 (±0.9)

0.8 (±0.2)

Snass Creek

CnH

Low

5.4 (±3.1)

1.2 (±0.3)

6.8 (±0.8)

0.8 (±0.2)

Twentysix Mile Creek

Connor

High

Twentysix Mile Creek

Busch

High

Twentysix Mile Creek

CnH

High

Twentysix Mile Creek

Connor

Medium

5.5 (±0.2)

1.1 (±0.2)

6.3 (±0.0)

0.7 (±0.0)

Twentysix Mile Creek

Busch

Medium

Twentysix Mile Creek

CnH

Medium

Twentysix Mile Creek

Connor

Low

11.6 (±1.4)

1.2 (±0.2)

6.3 (±0.0)

0.7 (±0.0)

Twentysix Mile Creek

Busch

Low

10.1 (±3.1)

1.1 (±0.2)

6.3 (±0.0)

0.7 (±0.0)

Twentysix Mile Creek

CnH

Low

10.1 (±3.1)

1.1 (±0.2)

6.3 (±0.0)

0.7 (±0.0)

Marmotte Creek

Connor

High

Marmotte Creek

Busch

High

Marmotte Creek

CnH

High

Marmotte Creek

Connor

Medium

6.4 (±0.8)

1.6 (±0.4)

4.8 (±0.0)

0.4 (±0.0)

Marmotte Creek

Busch

Medium

Marmotte Creek

CnH

Medium

Marmotte Creek

Connor

Low

8.6 (NA)

1.7 (NA)

4.7 (NA)

0.4 (NA)

Marmotte Creek

Busch

Low

7.1 (±1.4)

1.6 (±0.3)

4.7 (±0.0)

0.4 (±0.0)

Marmotte Creek

CnH

Low

7.1 (±1.4)

1.6 (±0.3)

4.7 (±0.0)

0.4 (±0.0)

Results and Interpretations 15

Chinook Salmon Intrinsic Potential Based on the Connor Model The Connor model identified 165 low (17.3 km), 2,355 medium (253.6 km), and 314 high (34.3 km) IP reaches across all 25 targeted streams for Chinook salmon (fig. 2), with an additional 852 (86.1 km) low and 763 medium (78.2 km) IP reaches from non-targeted streams (fig. 3). The reaches that scored a high IP (that is, modeled IP scores ≥ 0.75) for Chinook salmon were found in 34.3 km of reaches in the Skagit River main stem upstream from the Ross Lake reservoir in Canada (fig. 4). Most of the other targeted tributaries were dominated by medium IP habitat (scores greater than 0.25 and less than or equal to 0.75; fig. 4). Across all targeted reaches, 5.1 percent was rated as having low (less than 0.25) intrinsic potential habitat. After the main-stem Skagit River, which had about 40 km of main-stem habitat rated as high or medium IP, 11 different targeted tributaries had greater than 10 km of medium-rated IP habitat (fig. 2). Those tributaries with medium IP amounts greater than 10 km were the Sumallo River (26.7 km), Little Beaver Creek (25.3 km), Big Beaver Creek (24.9 km), Granite Creek (23.3 km), Lightning Creek (20.0 km), Panther Creek (18.8 km), Klesilkwa River (18.4 km), Canyon Creek (15.7 km), Devil’s Creek (13.3 km), Nepopekum Creek (11.4 km), and Three Fools Creek (10.7 km). An additional seven tributaries contained at least 1 km of medium IP habitat.

Chinook salmon intrinsic potential, in river kilometers

180

Connor

180

Across all reaches with low IP habitat, 59.8 percent were scored as having zero IP. In non-targeted tributary reaches, there was no high IP habitat identified, with 78.2 km of medium IP habitat and 86.1 km of low IP habitat identified from non-targeted tributaries (fig. 3).

Chinook Salmon Intrinsic Potential Based on the Busch Model The Busch model identified 2,360 low (253.6 km), 423 medium (46.0 km), and 51 high (5.6 km) intrinsic potential reaches within the 25 targeted streams for Chinook salmon (fig. 2), with an additional 25 medium (2.8 km) and 1,590 (161.5 km) low intrinsic potential reaches identified from non-targeted reaches (fig. 3). The only high intrinsic potential reaches for Chinook salmon, totaling 5.6 km, were found in the main-stem Skagit River (fig. 5) upstream from the Ross Lake reservoir in Canada. The other 24 targeted streams, except Ferguson Creek (fig. 2), and non-targeted streams (fig. 3) had most of the habitat identified as low intrinsic potential, a total of 253.9 km. Of the 46.0 km of reaches scored with medium intrinsic potential, the Skagit River (18.8 km), Granite Creek (4.3 km), and Ferguson Creek (3.6 km) had greater than 3 km of habitat, with the other 23 reaches ranging from 0 to 2.4 km. There was only 10 km of non-target tributary reaches identified as medium or low non-zero intrinsic potential (fig. 3). Of the low intrinsic potential habitat across all reaches, 90.4 percent scored as having 0 intrinsic potential (with the other 9.6 percent having scores greater than 0 and less than 0.25).

Busch

180

120

High Medium Low

Cooney and Holzer

High Medium Low

Intrinsic potential category Figure 3. Length of high, medium, and low intrinsic potential habitat for Chinook salmon (Oncorhynchus tshawytscha) based on three different models for non-target tributaries upstream from the three upper Skagit River dams, in northern Washington.

16 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington 121°30' W

121° W

120°30' W

EXPLANATION

it R ag

Cr.

HUC12 watershed boundary

Sna ss

o R.

Hydroelectric dam Chinook Intrinsic Potential Connor model

r. ile C ysix M Twent

i l k wa R.

Mas elpa nik C

Kles

Reservoir

rg Fe

Su ma ll

Cr. on us

po p e k Ne

High (>0.75) Medium (0.25–0.75)

Cr.

Low (<0.25)

49° N

or k

. Cas

Cr ver

Cr.

tle F

Three Fool s Cr

e ittle B

Lightning C

49° N

Lu Big

Devil s Cr.

B eaver C r.

. Cr na . an Cr ill

t tl

Cr .

ta Ste

Cr . ROSS DAM

Ruby C r.

DIABLO DAM

Can

GORGE DAM

Cr .

t he r

te C ani Gr

C r. der Thu n

Skagit River

48°30' N 121°30' W

Base modified from U.S. Geological Survey 1:100,000-scale data. Projection: NAD1983, NAVD 1988.

121° W

120°30' W

0 0

10 10

20 MILES 20 KILOMETERS

Figure 4. High (greater than [>] 0.75), medium (0.25–0.75), and low (less than [<] 0.25) intrinsic potential scores for Chinook salmon (Oncorhynchus tshawytscha) derived from the Connor intrinsic potential model applied to tributary and main-stem habitat upstream from the Skagit River Hydroelectric Project dams, in northern Washington. USGS Hydrological Unit Code (HUC) is a classification hierarchy to delimit watershed boundaries in successively smaller hydrologic units. HUC12 indicates boundaries of local sub-watershed scale that represents tributary systems.

Results and Interpretations 17 121°30' W

121° W

120°30' W

EXPLANATION Reservoir

it R ag

Cr.

HUC12 watershed boundary

Sna ss

Hydroelectric dam Chinook Intrinsic Potential Busch model

r. ile C ysix M Twent

i l k wa R.

Mas elpa nik C

Kles

o R.

rg Fe

Su ma ll

Cr. on us

po p e k Ne

High (>0.75) Medium (0.25–0.75)

Cr.

Low (<0.25)

49° N

or k

. Cas

Cr ver

Cr.

tle F

Three Fool s Cr

e ittle B

Lightning C

49° N

Lu Big

Devil s Cr.

B eaver C r.

. Cr na . an Cr ill

t tl

Cr .

ta Ste

Cr . ROSS DAM

Ruby C r.

DIABLO DAM

Can

GORGE DAM

Cr .

t he r

te C ani Gr

C r. der Thu n

Skagit River

48°30' N 121°30' W

Base modified from U.S. Geological Survey 1:100,000-scale data. Projection: NAD1983, NAVD 1988.

121° W

120°30' W

0 0

10 10

20 MILES 20 KILOMETERS

Figure 5. High (greater than [>] 0.75), medium (0.25–0.75), and low (less than [<] 0.25) intrinsic potential scores for Chinook salmon (Oncorhynchus tshawytscha) derived from the Busch intrinsic potential model applied to tributary and main-stem habitat upstream from the Skagit River Hydroelectric Project dams, in northern Washington. USGS Hydrological Unit Code (HUC) is a classification hierarchy to delimit watershed boundaries in successively smaller hydrologic units. HUC12 indicates boundaries of local subwatershed scale that represents tributary systems.

18 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington

Chinook Salmon Intrinsic Potential Based on the Cooney and Holzer Model The Cooney and Holzer model identified 2,164 low (232.0 km), 661 medium (72.2 km), and 9 high (1.0 km) intrinsic potential reaches within the 25 targeted streams for Chinook salmon (fig. 2), with an additional 35 medium (3.7 km) and 1,580 (160.5 km) low intrinsic potential reaches identified from non-targeted reaches (fig. 3). Of the three IP models, this model scored the lowest amount of high intrinsic potential habitat (totaling 1.0 km) for Chinook salmon, with 661 m found in the main-stem Skagit River and 339 m in McMillan Creek (figs. 5 and 6). There were seven targeted streams with greater than 1 km of medium intrinsic potential habitat for Chinook salmon, including the main-stem Skagit River (19.5 km), Klesilkwa River (14.0), Sumallo River (13.0 km), Big Beaver Creek (12.0 km), Little Beaver Creek (4.1 km), Ferguson Creek (3.5 km), and Nepopekum Creek (2.3). Across all 25 streams, 232.0 km of the reaches were rated as having low intrinsic potential (fig. 2). For non-targeted tributaries, there were no high intrinsic potential reaches, 3.7 km of medium intrinsic potential reaches, and 160.5 km of low intrinsic potential reaches (fig. 3). Across all reaches rated as having low intrinsic potential, 59.8 percent had IP scores of zero.

Coho Salmon Intrinsic Potential Model Results The estimates of IP for coho salmon upstream from the Skagit River Hydroelectric Project dams were similar across the three models. The Burnett model (Burnett and others, 2007), Agrawal model (Agrawal and others, 2005), and Romey model (Terrainworks, 2023) were all based on the same three parameters (gradient, bankfull width, and valley width index), but each model differed in the shape of the habitat preference curves and the IP weights assigned to the curves. All three models showed that most of the habitat upstream was classified as having medium IP habitat for coho salmon. For the 25 target streams, the main difference among the models was the higher amount of low IP reaches identified by the Burnett model and the greater number of streams having high IP by the Romey model. Summaries of habitat parameter values and the number of reaches scored as low, medium, and high IP sections are provided for main-stem and tributary habitat in the United States (table 6) and Canada (table 7).

Coho Salmon Intrinsic Potential Based on the Burnett Model The Burnett model identified 551 low (58.4 km), 2,258 medium (244.0 km), and 25 high (2.8 km) IP reaches across all 25 targeted streams (fig. 7), with additional 998 low (100.0 km), 611 medium (63.7 km), and 6 high (0.6 km) IP reaches identified from non-targeted reaches (fig. 8). The only high IP habitat identified for coho salmon by the Burnett model was 2.4 km in the main-stem Skagit River upstream from the

Ross Lake reservoir and 339 m in McMillan Creek (fig. 9). Of the 25 target streams, 19 had greater than 1 km of medium IP habitat available, totaling 243.2 km. The other six streams were dominated by (Maselpanik Creek and Slate Creek) or had all (Twentysix Mile, North Fork Canyon, Marmotte, and Hozomeen Creeks) low IP for coho salmon. In the non-targeted reaches, most of the habitat was modeled to have low (100.0) or medium (63.7) IP for coho salmon, with only 634 m of habitat identified from unnamed tributary reaches (fig. 8).

Coho Salmon Intrinsic Potential Based on the Agrawal Model Results from the Agrawal model were similar to those found with the Burnett model (fig. 7). The Agrawal model identified 115 low (118.0 km), 2,693 medium (290.5 km), and 26 high (2.4 km) IP reaches within the 25 targeted streams for coho salmon, with an additional 816 low (80.6 km), 790 medium (82.7 km), and 9 high (0.98 km) IP reaches identified from non-targeted reaches. The main difference between the Burnett and Agrawal models was some high intrinsic potential habitat identified in Stetattle Creek (<100 m) by the Agrawal model (fig. 10), which was classified as having medium intrinsic potential by the Burnett model. The amounts of high intrinsic potential habitat found in the main-stem Skagit River (2.4 km) and McMillan Creek (339.4 m) were identical in amount and the identified reaches in the two models. The main difference between the two models was the greater number of reaches classified as having medium intrinsic potential habitat by the Agrawal model (290.5 km), with the difference due to the lesser amount of low intrinsic potential habitat scores for this model compared with the Burnett model. A similar pattern was seen in the non-targeted streams, with more medium intrinsic potential habitat identified by the Agrawal model compared to the Burnett model (82.7 km versus 63.6 km), with the difference due to less total low intrinsic potential habitat (80.6 km versus 100 km; fig. 8).

Coho Salmon Intrinsic Potential Based on the Romey Model Of the three coho salmon IP models, a larger amount of high intrinsic potential habitat was identified by the Romey model (fig. 7). This model identified 219 low (22.8 km), 2,564 medium (276.8 km), and 51 high (5.5 km) IP reaches in the 25 targeted streams, and 655 (64.8 km) low, 620 (64.7 km) medium, and 340 (34.7 km) high IP reaches identified from non-targeted reaches. A total of five streams (Skagit River, McMillan Creek, Panther Creek, Stetattle Creek, and Castle Fork Creek) had 5.5 km of high intrinsic potential habitat (fig. 11), with an additional 34.7 km identified in non-targeted tributaries (fig. 8). As seen in the other two intrinsic potential models for coho, most of the total accessible reaches in targeted streams were classified as having medium intrinsic potential (276.8 km).

Results and Interpretations 19 121°30' W

121° W

120°30' W

EXPLANATION

it R ag

Cr.

HUC12 watershed boundary

Sna ss

o R.

Hydroelectric dam Chinook Intrinsic Potential Cooney and Holzer model

r. ile C ysix M Twent

i l k wa R.

Mas elpa nik C

Kles

Reservoir

rg Fe

Su ma ll

Cr. on us

po p e k Ne

High Medium

Cr.

Low

49° N

or k

. Cas

Cr ver

Cr.

tle F

Three Fool s Cr

e ittle B

Lightning C

49° N

Lu Big

Devil s Cr.

B eaver C r.

. Cr na . an Cr ill

t tl

Cr .

ta Ste

Cr . ROSS DAM

Ruby C r.

DIABLO DAM

Can

GORGE DAM

Cr .

t he r

te C ani Gr

C r. der Thu n

Skagit River

48°30' N 121°30' W

Base modified from U.S. Geological Survey 1:100,000-scale data. Projection: NAD1983, NAVD 1988.

121° W

120°30' W

0 0

10 10

20 MILES 20 KILOMETERS

Figure 6. High, medium, and low intrinsic potential scores for Chinook salmon (Oncorhynchus tshawytscha) derived from the Cooney and Holzer intrinsic potential model applied to tributary and main-stem habitat upstream from the Skagit River Hydroelectric Project dams, in northern Washington. USGS Hydrological Unit Code (HUC) is a classification hierarchy to delimit watershed boundaries in successively smaller hydrologic units. HUC12 indicates boundaries of local sub-watershed scale that represents tributary systems.

20 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington Table 6. Reach average values of gradient, valley width index (ratio), bankfull width, and mean annual flow for targeted upper Skagit River streams in the United States by high, medium, or low intrinsic potential (IP) rank for coho salmon (Oncorhynchus kisutch). [Plus or minus (±) standard deviations are shown in parentheses. NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Stream

Model

Number of reaches

Percent gradient

Valley width ratio

Bankfull width (m)

Mean annual flow (m3/s)

Big Beaver Creek

Burnett

Big Beaver Creek

Agrawal

High

Big Beaver Creek

Romey

High

Big Beaver Creek

Burnett

209

Medium

0.9 (±1.1)

0.6 (±0.3)

14.1 (±2.7)

4.5 (±1.5)

Big Beaver Creek

Agrawal

216

Medium

1.1 (±1.4)

0.6 (±0.3)

14.0 (±2.7)

4.4 (±1.5)

Big Beaver Creek

Romey

215

Medium

1.0 (±1.4)

0.6 (±0.3)

14.0 (±2.7)

4.4 (±1.5)

Big Beaver Creek

Burnett

Low

5.0 (±4.5)

0.6 (±0.1)

13.8 (±2.1)

4.2 (±1.4)

Big Beaver Creek

Agrawal

Low

4.6 (±5.5)

0.5 (±0.1)

14.5 (±2.2)

4.6 (±1.5)

Big Beaver Creek

Romey

Low

4.8 (±5.4)

0.6 (±0.1)

14.3 (±2.2)

4.5 (±1.5)

Little Beaver Creek

Burnett

High

Little Beaver Creek

Agrawal

High

Little Beaver Creek

Romey

High

Little Beaver Creek

Burnett

217

Medium

1.6 (±1.0)

0.7 (±0.2)

12.1 (±2.4)

3.2 (±1.3)

Little Beaver Creek

Agrawal

234

Medium

1.9 (±1.5)

0.7 (±0.2)

12.1 (±2.6)

3.2 (±1.4)

Little Beaver Creek

Romey

230

Medium

1.8 (±1.4)

0.7 (±0.2)

12.0 (±2.5)

3.2 (±1.4)

Little Beaver Creek

Burnett

Low

6.6 (±1.7)

0.8 (±0.3)

10.2 (±3.9)

2.5 (±2.0)

Little Beaver Creek

Agrawal

Low

9.1 (±1.0)

1.1 (±0.1)

6.6 (±0.1)

0.8 (±0.0)

Little Beaver Creek

Romey

Low

8.4 (±1.2)

0.9 (±0.3)

9.5 (±4.3)

2.2 (±2.1)

Ruby Creek

Burnett

High

Ruby Creek

Agrawal

High

Ruby Creek

Romey

High

Ruby Creek

Burnett

Medium

1.5 (±0.5)

0.5 (±0.2)

19.9 (±1.4)

9.2 (±1.4)

Ruby Creek

Agrawal

Medium

1.5 (±0.5)

0.5 (±0.2)

19.9 (±1.4)

9.2 (±1.4)

Ruby Creek

Romey

Medium

1.5 (±0.5)

0.5 (±0.2)

19.9 (±1.4)

9.2 (±1.4)

Ruby Creek

Burnett

Low

Ruby Creek

Agrawal

Low

Ruby Creek

Romey

Low

Canyon Creek

Burnett

High

Canyon Creek

Agrawal

High

Canyon Creek

Romey

High

Canyon Creek

Burnett

130

Medium

2.4 (±1.1)

0.7 (±0.2)

12.5 (±2.0)

3.3 (±1.0)

Canyon Creek

Agrawal

152

Medium

3.0 (±1.8)

0.8 (±0.4)

11.9 (±2.6)

3.1 (±1.2)

Canyon Creek

Romey

144

Medium

2.7 (±1.5)

0.7 (±0.4)

12.1 (±2.4)

3.2 (±1.2)

Canyon Creek

Burnett

Low

6.8 (±1.2)

1.5 (±0.8)

8.1 (±2.7)

1.4 (±1.2)

Canyon Creek

Agrawal

Low

8.6 (±0.2)

2.3 (±0.7)

6.5 (±0.1)

0.8 (±0.0)

Canyon Creek

Romey

Low

7.9 (±0.7)

1.6 (±0.7)

7.6 (±2.4)

1.2 (±1.0)

North Fork Canyon Creek Burnett

High

North Fork Canyon Creek Agrawal

High

North Fork Canyon Creek Romey

High

North Fork Canyon Creek Burnett

Medium

North Fork Canyon Creek Agrawal

Medium

7.2 (±0.9)

IP rank

1.2 (±0.3)

6.4 (±0.0)

0.7 (±0.0)

Results and Interpretations 21 Table 6. Reach average values of gradient, valley width index (ratio), bankfull width, and mean annual flow for targeted upper Skagit River streams in United States by high, medium, or low intrinsic potential (IP) rank for coho salmon (Oncorhynchus kisutch).—Continued [Plus or minus (±) standard deviations are shown in parentheses. NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Number of reaches

IP rank

Valley width ratio

Bankfull width (m)

Mean annual flow (m3/s)

North Fork Canyon Creek Romey

Medium

North Fork Canyon Creek Burnett

Low

6.4 (NA)

1.0 (NA)

6.4 (NA)

0.7 (NA)

9.0 (±1.8)

1.5 (±0.5)

6.4 (±0.0)

0.7 (±0.0)

North Fork Canyon Creek Agrawal

North Fork Canyon Creek Romey

Low

10.1 (±1.0)

1.7 (±0.5)

6.3 (±0.0)

0.7 (±0.0)

Low

9.4 (±1.5)

1.6 (±0.4)

6.3 (±0.0)

0.7 (±0.0)

Granite Creek

Burnett

Granite Creek

Agrawal

High

Granite Creek

Romey

High

Granite Creek

Burnett

184

Medium

2.4 (±1.0)

0.9 (±0.3)

10.5 (±1.8)

2.3 (±0.9)

Granite Creek

Agrawal

215

Medium

2.9 (±1.5)

0.8 (±0.3)

10.5 (±1.7)

2.3 (±0.8)

Granite Creek

Romey

213

Medium

2.8 (±1.4)

0.8 (±0.3)

10.5 (±1.7)

2.3 (±0.8)

Granite Creek

Burnett

Low

5.7 (±0.7)

0.7 (±0.2)

11.0 (±0.7)

2.5 (±0.4)

Granite Creek

Agrawal

Low

Granite Creek

Romey

Low

7.8 (±0.4)

0.6 (±0.1)

10.7 (±0.5)

2.3 (±0.2)

Lightning Creek

Burnett

High

0.0 (±0.0)

Lightning Creek

Agrawal

High

0.0 (±0.0)

Lightning Creek

Romey

High

0.0 (±0.0)

Lightning Creek

Burnett

159

Medium

2.3 (±1.2)

0.8 (±0.4)

11.6 (±2.9)

3.0 (±1.7)

Lightning Creek

Agrawal

193

Medium

2.9 (±1.8)

0.8 (±0.4)

11.3 (±2.9)

2.8 (±1.7)

Lightning Creek

Romey

188

Medium

2.8 (±1.6)

0.8 (±0.4)

11.3 (±2.9)

2.8 (±1.7)

Lightning Creek

Burnett

Low

6.1 (±1.4)

1.0 (±0.4)

9.4 (±2.1)

1.8 (±1.2)

Lightning Creek

Agrawal

Low

9.5 (±0.6)

1.2 (±0.3)

7.9 (±0.6)

1.2 (±0.2)

Lightning Creek

Romey

Low

8.3 (±1.0)

1.1 (±0.3)

8.5 (±1.6)

1.4 (±0.6)

Luna Creek

Burnett

High

Luna Creek

Agrawal

High

Luna Creek

Romey

High

Luna Creek

Burnett

Medium

1.8 (±1.4)

0.8 (±0.1)

9.0 (±0.3)

1.6 (±0.1)

Luna Creek

Agrawal

Medium

2.5 (±2.1)

0.9 (±0.1)

8.9 (±0.3)

1.5 (±0.1)

Luna Creek

Romey

Medium

2.4 (±1.9)

0.8 (±0.1)

8.9 (±0.3)

1.5 (±0.1)

Luna Creek

Burnett

Low

6.0 (±0.8)

0.9 (±0.2)

8.6 (±0.4)

1.4 (±0.2)

Luna Creek

Agrawal

Low

Luna Creek

Romey

Low

7.5 (NA)

0.9 (NA)

8.1 (NA)

1.2 (NA)

Three Fools Creek

Burnett

High

Three Fools Creek

Agrawal

High

Three Fools Creek

Romey

High

Three Fools Creek

Burnett

Medium

3.1 (±1.0)

0.9 (±0.3)

10.6 (±1.5)

2.3 (±0.7)

Three Fools Creek

Agrawal

102

Medium

3.7 (±1.5)

0.9 (±0.4)

10.6 (±1.5)

2.3 (±0.7)

Three Fools Creek

Romey

Medium

3.4 (±1.2)

0.9 (±0.4)

10.5 (±1.5)

2.3 (±0.7)

Three Fools Creek

Burnett

Low

6.7 (±2.2)

0.9 (±0.6)

10.5 (±2.0)

2.3 (±0.8)

Three Fools Creek

Agrawal

Low

10.7 (±2.9)

1.0 (±0.6)

9.8 (±4.2)

2.2 (±1.5)

Three Fools Creek

Romey

Low

8.5 (±2.1)

0.8 (±0.3)

11.1 (±2.3)

2.6 (±0.9)

Castle Fork Creek

Burnett

High

Stream

Model

Percent gradient

22 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington Table 6. Reach average values of gradient, valley width index (ratio), bankfull width, and mean annual flow for targeted upper Skagit River streams in United States by high, medium, or low intrinsic potential (IP) rank for coho salmon (Oncorhynchus kisutch).—Continued [Plus or minus (±) standard deviations are shown in parentheses. NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Stream

Model

Number of reaches

Percent gradient

Valley width ratio

Bankfull width (m)

Mean annual flow (m3/s)

Castle Fork Creek

Agrawal

Castle Fork Creek

Romey

High

Castle Fork Creek

Burnett

Castle Fork Creek

Agrawal

Medium

3.1 (±1.1)

2.0 (±1.2)

5.8 (±0.9)

0.6 (±0.2)

Medium

5.7 (±1.8)

1.9 (±0.9)

5.3 (±1.2)

0.5 (±0.2)

Castle Fork Creek Castle Fork Creek

Romey

Medium

4.9 (±1.6)

1.7 (±0.8)

5.7 (±1.1)

0.6 (±0.2)

Burnett

Low

8.0 (±2.4)

2.1 (±0.8)

4.6 (±1.4)

0.4 (±0.3)

Castle Fork Creek

Agrawal

Low

10.6 (±0.9)

2.5 (±0.7)

3.5 (±1.1)

0.2 (±0.1)

Castle Fork Creek

Romey

Low

9.3 (±2.1)

2.3 (±0.6)

4.0 (±1.2)

0.3 (±0.2)

Panther Creek

Burnett

High

Panther Creek

Agrawal

High

Panther Creek

Romey

High

Panther Creek

Burnett

115

Medium

3.0 (±1.0)

1.3 (±1.0)

8.3 (±1.7)

1.4 (±0.6)

Panther Creek

Agrawal

184

Medium

4.2 (±1.8)

1.3 (±0.9)

8.4 (±2.2)

1.5 (±0.8)

Panther Creek

Romey

160

Medium

3.8 (±1.5)

1.1 (±0.4)

8.5 (±2.1)

1.5 (±0.7)

Panther Creek

Burnett

Low

6.4 (±1.2)

1.2 (±0.5)

8.6 (±2.9)

1.6 (±1.0)

Panther Creek

Agrawal

Low

8.8 (±0.4)

0.9 (±0.3)

9.1 (±2.4)

1.7 (±1.0)

Panther Creek

Romey

Low

7.8 (±0.7)

1.2 (±0.6)

8.5 (±2.8)

1.6 (±1.0)

Devils Creek

Burnett

High

Devils Creek

Agrawal

High

Devils Creek

Romey

High

Devils Creek

Burnett

Medium

3.5 (±0.9)

1.4 (±0.5)

7.1 (±1.4)

1.0 (±0.4)

Devils Creek

Agrawal

131

Medium

4.5 (±1.4)

1.3 (±0.5)

7.4 (±1.5)

1.1 (±0.4)

Devils Creek

Romey

126

Medium

4.4 (±1.2)

1.2 (±0.5)

7.5 (±1.4)

1.1 (±0.4)

Devils Creek

Burnett

Low

6.0 (±0.3)

1.2 (±0.4)

7.8 (±1.6)

1.2 (±0.4)

Devils Creek

Agrawal

Low

9.5 (±0.4)

1.4 (±0.7)

6.9 (±2.9)

1.0 (±0.8)

Devils Creek

Romey

Low

8.5 (±1.1)

1.4 (±0.6)

6.9 (±2.7)

1.0 (±0.7)

Stetattle Creek

Burnett

High

Stetattle Creek

Agrawal

High

Stetattle Creek

Romey

High

Stetattle Creek

Burnett

Medium

2.7 (±1.2)

1.3 (±1.9)

10.4 (±2.3)

2.3 (±0.7)

Stetattle Creek

Agrawal

Medium

4.2 (±2.1)

1.1 (±1.3)

10.0 (±2.0)

2.1 (±0.7)

Stetattle Creek

Romey

Medium

3.8 (±1.9)

0.9 (±0.2)

10.2 (±1.6)

2.1 (±0.7)

Stetattle Creek

Burnett

Low

6.6 (±1.6)

1.0 (±0.2)

9.5 (±1.9)

1.8 (±0.8)

Stetattle Creek

Agrawal

Low

10.7 (±3.3)

1.1 (±0.0)

11.4 (±0.5)

2.7 (±0.2)

Stetattle Creek

Romey

Low

8.2 (±2.2)

1.0 (±0.1)

10.7 (±1.5)

2.4 (±0.6)

McMillan Creek

Burnett

High

McMillan Creek

Agrawal

High

McMillan Creek

Romey

High

McMillan Creek

Burnett

Medium

2.6 (±1.3)

2.9 (±5.9)

8.3 (±0.9)

1.3 (±0.3)

McMillan Creek

Agrawal

Medium

3.3 (±1.8)

2.6 (±5.3)

8.3 (±0.9)

1.3 (±0.3)

McMillan Creek

Romey

Medium

3.1 (±1.7)

1.8 (±4.7)

8.4 (±0.9)

1.4 (±0.3)

IP rank

Results and Interpretations 23 Table 6. Reach average values of gradient, valley width index (ratio), bankfull width, and mean annual flow for targeted upper Skagit River streams in United States by high, medium, or low intrinsic potential (IP) rank for coho salmon (Oncorhynchus kisutch).—Continued [Plus or minus (±) standard deviations are shown in parentheses. NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Stream

Model

Number of reaches

Bankfull width (m)

Mean annual flow (m3/s)

McMillan Creek

Burnett

Low

7.1 (±2.4)

McMillan Creek

Agrawal

Low

10.7 (±1.8)

2.0 (±1.9)

7.9 (±1.5)

1.2 (±0.4)

2.8 (±1.2)

7.0 (±2.1)

1.0 (±0.6)

McMillan Creek

Romey

Low

9.4 (±2.2)

Slate Creek

Burnett

High

2.1 (±1.3)

7.8 (±1.9)

1.2 (±0.6)

Slate Creek

Agrawal

High

Slate Creek

Romey

High

Slate Creek

Burnett

Medium

4.6 (NA)

0.8 (NA)

8.5 (NA)

1.4 (NA)

Slate Creek

Agrawal

Slate Creek

Romey

Medium

6.1 (±1.1)

1.1 (±0.3)

8.6 (±0.1)

1.4 (±0.0)

Medium

5.5 (±0.6)

1.1 (±0.2)

8.6 (±0.1)

1.4 (±0.0)

Slate Creek

Burnett

Slate Creek

Agrawal

Low

6.9 (±1.4)

1.2 (±0.2)

8.6 (±0.0)

1.4 (±0.0)

Low

8.9 (±0.7)

1.3 (±0.1)

8.6 (±0.0)

1.4 (±0.0)

Slate Creek Hozomeen Creek

Romey

Low

8.0 (±1.0)

1.3 (±0.3)

8.6 (±0.0)

1.4 (±0.0)

Burnett

High

Hozomeen Creek

Agrawal

High

Hozomeen Creek

Romey

High

Hozomeen Creek

Burnett

Medium

Hozomeen Creek

Agrawal

Medium

Hozomeen Creek

Romey

Medium

Hozomeen Creek

Burnett

Low

12.8 (NA)

1.7 (NA)

4.9 (NA)

0.4 (NA)

Hozomeen Creek

Agrawal

Low

12.8 (NA)

1.7 (NA)

4.9 (NA)

0.4 (NA)

Hozomeen Creek

Romey

Low

12.8 (NA)

1.7 (NA)

4.9 (NA)

0.4 (NA)

IP rank

Percent gradient

Valley width ratio

24 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington Table 7. Reach average values of gradient, valley width index (ratio), bankfull width, and mean annual flow for targeted upper Skagit River streams in Canada by high, medium, or low intrinsic potential rank (IP) for coho salmon (Oncorhynchus kisutch). [Plus or minus (±) standard deviations are shown in parentheses. NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Stream

Model

Number of reaches

IP rank

Percent gradient

Valley width ratio

Bankfull width (m)

Mean annual flow (m3/s)

Skagit River

Burnett

High

0.3 (±0.5)

24.6 (±33.5)

26.4 (±1.1)

17.3 (±1.5)

Skagit River

Agrawal

High

0.3 (±0.5)

24.6 (±33.5)

26.4.0 (±1.1)

17.3 (±1.5)

Skagit River

Romey

High

0.4 (±0.5)

9.4 (±7.8)

24.9 (±3.2)

15.5 (±3.8)

Skagit River

Burnett

336

Medium

0.5 (±0.6)

1.5 (±1.0)

22.4 (±3.0)

12.2 (±3.4)

Skagit River

Agrawal

338

Medium

0.5 (±0.7)

1.5 (±1.0)

22.3 (±3.0)

12.1 (±3.4)

Skagit River

Romey

326

Medium

0.5 (±0.7)

2.2 (±9.8)

22.3 (±3.0)

12.1 (±3.4)

Skagit River

Burnett

Low

4.8 (±0.0)

2.3 (±0.7)

14.5 (±0.7)

4.6 (±0.5)

Skagit River

Agrawal

Low

Skagit River

Romey

Low

Klesilkwa River

Burnett

High

Klesilkwa River

Agrawal

High

Klesilkwa River

Romey

High

Klesilkwa River

Burnett

162

Medium

0.4 (±0.7)

0.7 (±0.2)

11.0 (±2.7)

2.7 (±1.3)

Klesilkwa River

Agrawal

166

Medium

0.5 (±1.1)

0.7 (±0.2)

11.0 (±2.7)

2.6 (±1.3)

Klesilkwa River

Romey

165

Medium

0.5 (±1.0)

0.7 (±0.2)

11.0 (±2.6)

2.6 (±1.3)

Klesilkwa River

Burnett

Low

8.3 (±2.7)

1.1 (±0.1)

7.3 (±0.0)

1.0 (±0.0)

Klesilkwa River

Agrawal

Low

10.5 (±1.9)

1.2 (±0.0)

7.3 (±0.0)

1.0 (±0.0)

Klesilkwa River

Romey

Low

9.8 (±2.3)

1.1 (±0.0)

7.3 (±0.0)

1.0 (±0.0)

Sumallo River

Burnett

High

Sumallo River

Agrawal

High

Sumallo River

Romey

High

Sumallo River

Burnett

232

Medium

1.1 (±1.4)

0.7 (±0.2)

11.2 (±2.6)

2.7 (±1.2)

Sumallo River

Agrawal

246

Medium

1.4 (±1.7)

0.7 (±0.2)

11.0 (±2.7)

2.6 (±1.3)

Sumallo River

Romey

243

Medium

1.3 (±1.6)

0.7 (±0.2)

11.1 (±2.6)

2.7 (±1.2)

Sumallo River

Burnett

Low

5.4 (±1.7)

1.0 (±0.2)

7.6 (±1.5)

1.1 (±0.5)

Sumallo River

Agrawal

Low

0.2 (NA)

07 (NA)

9.8 (±NA)

1.9 (±NA)

Sumallo River

Romey

Low

5.5 (±3.6)

1.2 (±0.3)

6.5 (±2.2)

0.9 (±0.7)

Ferguson Creek

Burnett

High

Ferguson Creek

Agrawal

High

Ferguson Creek

Romey

High

Ferguson Creek

Burnett

Medium

0.5 (±0.9)

1.3 (±0.2)

5.6 (±0.6)

0.6 (±0.1)

Ferguson Creek

Agrawal

Medium

1.0 (±2.0)

1.4 (±0.2)

5.6 (±0.7)

0.5 (±0.1)

Ferguson Creek

Romey

Medium

0.6 (±1.2)

1.3 (±0.2)

5.6 (±0.6)

0.6 (±0.1)

Ferguson Creek

Burnett

Low

6.0 (±3.0)

1.5 (±0.2)

4.9 (±0.5)

0.4 (±0.1)

Ferguson Creek

Agrawal

Low

4.2 (±5.8)

1.4 (±0.2)

5.3 (±1.0)

0.5 (±0.2)

Ferguson Creek

Romey

Low

6.0 (±3.3)

1.5 (±0.2)

4.9 (±0.6)

0.4 (±0.1)

Nepopekum Creek

Burnett

High

Nepopekum Creek

Agrawal

High

Nepopekum Creek

Romey

High

Nepopekum Creek

Burnett

Medium

1.9 (±1.5)

0.9 (±0.1)

8.8 (±0.2)

1.5 (±0.1)

Nepopekum Creek

Agrawal

109

Medium

2.7 (±2.1)

0.9 (±0.1)

8.7 (±0.3)

1.5 (±0.1)

Results and Interpretations 25 Table 7. Reach average values of gradient, valley width index (ratio), bankfull width, and mean annual flow for targeted upper Skagit River streams in Canada by high, medium, or low intrinsic potential rank (IP) for coho salmon (Oncorhynchus kisutch).—Continued [Plus or minus (±) standard deviations are shown in parentheses. NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Stream

Model

Number of reaches

IP rank

Percent gradient

Valley width ratio

Bankfull width (m)

Mean annual flow (m3/s)

Nepopekum Creek

Romey

107

Medium

2.6 (±2.0)

0.9 (±0.1)

8.7 (±0.2)

1.5 (±0.0)

Nepopekum Creek

Burnett

Low

6.8 (±1.7)

0.9 (±0.1)

8.5 (±0.2)

1.4 (±0.1)

Nepopekum Creek

Agrawal

Low

9.4 (±1.2)

0.8 (±0.1)

8.5 (±0.1)

1.4 (±0.0)

Nepopekum Creek

Romey

Low

9.0 (±1.4)

0.9 (±0.1)

8.5 (±0.2)

1.4 (±0.1)

Maselpanik Creek

Burnett

High

Maselpanik Creek

Agrawal

High

Maselpanik Creek

Romey

High

Maselpanik Creek

Burnett

Medium

2.1 (±0.9)

0.7 (±0.1)

10.4 (±0.2)

2.2 (±0.1)

Maselpanik Creek

Agrawal

Medium

3.7 (±2.5)

0.7 (±0.1)

10.4 (±0.1)

2.2 (±0.1)

Maselpanik Creek

Romey

Medium

2.7 (±1.7)

0.7 (±0.1)

10.4 (±0.2)

2.2 (±0.1)

Maselpanik Creek

Burnett

Low

8.0 (±1.5)

0.7 (±0.2)

10.5 (±0.2)

2.2 (±0.1)

Maselpanik Creek

Agrawal

Low

9.3 (±0.6)

0.8 (±0.2)

10.0 (±0.3)

2.2 (±0.1)

Maselpanik Creek

Romey

Low

8.5 (±1.2)

0.7 (±0.2)

10.5 (±0.2)

2.2 (±0.1)

Snass Creek

Burnett

High

Snass Creek

Agrawal

High

Snass Creek

Romey

High

Snass Creek

Burnett

Medium

2.6 (±1.4)

1.3 (±0.2)

6.6 (±0.9)

0.8 (±0.2)

Snass Creek

Agrawal

Medium

4.0 (±2.1)

1.3 (±0.3)

6.6 (±0.8)

0.8 (±0.2)

Snass Creek

Romey

Medium

3.6 (±1.9)

1.3 (±0.3)

6.6 (±0.8)

0.8 (±0.2)

Snass Creek

Burnett

Low

7.7 (±2.3)

1.2 (±0.3)

6.8 (±0.8)

0.9 (±0.2)

Snass Creek

Agrawal

Low

10.1 (±1.3)

1.0 (±0.2)

7.1 (±0.08)

0.9 (±0.2)

Snass Creek

Romey

Low

9.4 (±0.7)

1.0 (±0.2)

6.9 (±0.9)

0.9 (±0.2)

Twentysix Mile Creek

Burnett

High

Twentysix Mile Creek

Agrawal

High

Twentysix Mile Creek

Romey

High

Twentysix Mile Creek

Burnett

Medium

Twentysix Mile Creek

Agrawal

Medium

5.5 (±0.2)

1.1 (±0.2)

6.3 (±0.0)

0.7 (±0.0)

Twentysix Mile Creek

Romey

Medium

5.5 (±0.2)

1.1 (±0.2)

6.3 (±0.0)

0.7 (±0.0)

Twentysix Mile Creek

Burnett

Low

10.1 (±3.1)

1.1 (±0.2)

6.3 (±0.0)

0.7 (±0.0)

Twentysix Mile Creek

Agrawal

Low

11.6 (±1.4)

1.2 (±0.2)

6.3 (±0.0)

0.7 (±0.0)

Twentysix Mile Creek

Romey

Low

11.6 (±1.4)

1.2 (±0.2)

6.3 (±0.0)

0.7 (±0.0)

Marmotte Creek

Burnett

High

Marmotte Creek

Agrawal

High

Marmotte Creek

Romey

High

Marmotte Creek

Burnett

Medium

Marmotte Creek

Agrawal

Medium

6.4 (±0.8)

1.6 (±0.4)

4.8 (±0.0)

0.4 (±0.0)

Marmotte Creek

Romey

Medium

5.9 (NA)

1.9 (NA)

4.8 (NA)

0.4 (NA)

Marmotte Creek

Burnett

Low

7.1 (±1.4)

1.6 (±0.3)

4.7 (±0.0)

0.4 (±0.0)

Marmotte Creek

Agrawal

Low

8.6 (NA)

1.7 (NA)

4.7 (NA)

0.4 (NA)

Marmotte Creek

Romey

Low

7.8 (±1.2)

1.5 (±0.3)

4.7 (±0.0)

0.4 (±0.0)

26 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington Skagit River Sumallo River Little Beaver Creek Big Beaver Creek Granite Creek Lightning Creek Panther Creek Klesilkwa River Canyon Creek Devils Creek Nepopekum Creek Three Fools Creek Castle Fork Creek Stetattle Creek McMillan Creek Ruby Creek Ferguson Creek Snass Creek Luna Creek Maselpanik Creek Slate Creek Twentysix Mile Creek North Fork Canyon Creek Marmotte Creek Hozomeen Creek Skagit River Sumallo River Little Beaver Creek Big Beaver Creek Granite Creek Lightning Creek Panther Creek Klesilkwa River Canyon Creek Devils Creek Nepopekum Creek Three Fools Creek Castle Fork Creek Stetattle Creek McMillan Creek Ruby Creek Ferguson Creek Snass Creek Luna Creek Maselpanik Creek Slate Creek Twentysix Mile Creek North Fork Canyon Creek Marmotte Creek Hozomeen Creek

EXPLANATION High intrinsic potential Medium intrinsic potential Low intrinsic potential Species: Coho salmon Model: Burnett

EXPLANATION High intrinsic potential Medium intrinsic potential Low intrinsic potential Species: Coho salmon Model: Agrawal

EXPLANATION High intrinsic potential Medium intrinsic potential Low intrinsic potential Species: Coho salmon Model: Romey

15 20 25 Coho salmon intrinsic potential, in river kilometers

Figure 7. Length in kilometers of high, medium, or low habitat intrinsic potential for coho salmon (Oncorhynchus kisutch) based on three different models for target tributaries upstream from the three upper Skagit River dams, in northern Washington.

Coho salmon intrinsic potential, in river kilometers

Results and Interpretations 27

100 80 60 40 20 0

Burnett

High Medium Low

100 80 60 40 20 0

Agrawal

High Medium Low

100 80 60 40 20 0

Romey

High Medium Low

Intrinsic potential category Figure 8. Length in kilometers of high, medium, or low habitat intrinsic potential for coho salmon (Oncorhynchus kisutch) based on three different models for non-target tributaries upstream from the three upper Skagit River dams, in northern Washington.

28 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington 121°30' W

121° W

120°30' W

EXPLANATION

it R ag

Cr.

HUC12 watershed boundary

Sna ss

o R.

Hydroelectric dam Coho Intrinsic Potential Burnett model

r. ile C ysix M Twent

i l k wa R.

Mas elpa nik C

Kles

Reservoir

rg Fe

Su ma ll

Cr. on us

po p e k Ne

High (>0.75) Medium (0.25–0.75)

Cr.

Low (<0.25)

49° N

or k

. Cas

Cr ver

Cr.

tle F

Three Fool s Cr

e ittle B

Lightning C

49° N

Lu Big

Devil s Cr.

B eaver C r.

. Cr na . an Cr ill

t tl

Cr .

ta Ste

Cr . ROSS DAM

Ruby C r.

DIABLO DAM

Can

GORGE DAM

Cr .

t he r

te C ani Gr

C r. der Thu n

Skagit River

48°30' N 121°30' W

Base modified from U.S. Geological Survey 1:100,000-scale data. Projection: NAD1983, NAVD 1988.

121° W

120°30' W

0 0

10 10

20 MILES 20 KILOMETERS

Figure 9. High (greater than [>] 0.75), medium (0.25–0.75), and low (less than [<] 0.25) intrinsic potential scores for coho salmon (Oncorhynchus kisutch) derived from the Burnett intrinsic potential model applied to tributary and main-stem habitat upstream from the Skagit River Hydroelectric Project dams, in northern Washington. USGS Hydrological Unit Code (HUC) is a classification hierarchy to delimit watershed boundaries in successively smaller hydrologic units. HUC12 indicates boundaries of local subwatershed scale that represents tributary systems.

Results and Interpretations 29 121°30' W

121° W

120°30' W

EXPLANATION

it R ag

Cr.

HUC12 watershed boundary

Sna ss

o R.

Hydroelectric dam Coho Intrinsic Potential Agrawal model

r. ile C ysix M Twent

i l k wa R.

Mas elpa nik C

Kles

Reservoir

rg Fe

Su ma ll

Cr. on us

po p e k Ne

High (>0.75) Medium (0.25–0.75)

Cr.

Low (<0.25)

49° N

or k

. Cas

Cr ver

Cr.

tle F

Three Fool s Cr

e ittle B

Lightning C

49° N

Lu Big

Devil s Cr.

B eaver C r.

. Cr na . an Cr ill

t tl

Cr .

ta Ste

Cr . ROSS DAM

Ruby C r.

DIABLO DAM

Can

GORGE DAM

Cr .

t he r

te C ani Gr

C r. der Thu n

Skagit River

48°30' N 121°30' W

Base modified from U.S. Geological Survey 1:100,000-scale data. Projection: NAD1983, NAVD 1988.

121° W

120°30' W

0 0

10 10

20 MILES 20 KILOMETERS

Figure 10. High (greater than [>] 0.75), medium (0.25–0.75), and low (less than [<] 0.25) intrinsic potential scores for coho salmon (Oncorhynchus kisutch) derived from the Agrawal intrinsic potential model applied to tributary and main-stem habitat upstream from the Skagit River Hydroelectric Project dams, in northern Washington. USGS Hydrological Unit Code (HUC) is a classification hierarchy to delimit watershed boundaries in successively smaller hydrologic units. HUC12 indicates boundaries of local subwatershed scale that represents tributary systems.

30 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington 121°30' W

121° W

120°30' W

EXPLANATION

it R ag

Cr.

HUC12 watershed boundary

Sna ss

o R.

Hydroelectric dam Coho Intrinsic Potential Romey model

r. ile C ysix M Twent

i l k wa R.

Mas elpa nik C

Kles

Reservoir

rg Fe

Su ma ll

Cr. on us

po p e k Ne

High (>0.75) Medium (0.25–0.75)

Cr.

Low (<0.25)

49° N

or k

. Cas

Cr ver

Cr.

tle F

Three Fool s Cr

e ittle B

Lightning C

49° N

Lu Big

Devil s Cr.

B eaver C r.

. Cr na . an Cr ill

t tl

Cr .

ta Ste

Cr . ROSS DAM

Ruby C r.

DIABLO DAM

Can

GORGE DAM

Cr .

t he r

te C ani Gr

C r. der Thu n

Skagit River

48°30' N 121°30' W

Base modified from U.S. Geological Survey 1:100,000-scale data. Projection: NAD1983, NAVD 1988.

121° W

120°30' W

0 0

10 10

20 MILES 20 KILOMETERS

Figure 11. High (greater than [>] 0.75), medium (0.25–0.75), and low (less than [<] 0.25) intrinsic potential scores for coho salmon (Oncorhynchus kisutch) derived from the Romey intrinsic potential model applied to tributary and main-stem habitat upstream from the Skagit River Hydroelectric Project dams, in northern Washington. USGS Hydrological Unit Code (HUC) is a classification hierarchy to delimit watershed boundaries in successively smaller hydrologic units. HUC12 indicates boundaries of local subwatershed scale that represents tributary systems.

Results and Interpretations 31

Steelhead Intrinsic Potential Model Results The IP estimates for steelhead upstream from the Skagit River Hydroelectric Project dams were similar across the three models (fig. 12). The Agrawal and Burnett models were calculated using the same three parameters (gradient, bankfull width, and valley width index) but with different weights assigned to the habitat preference curves, whereas the Puget Sound Technical Recovery Team model was based on combinations of gradient and bankfull width assigned as low, medium, and high IP. In general, steelhead showed greater amounts of high IP of the three species examined. The three models differed in the amount of high IP in the Skagit River main stem and the total amount of low IP habitat: the Puget Sound TRT model estimated greater amounts of these habitats than the other two models. Summaries of habitat parameter values and the number of reaches scored as low, medium, and high IP sections are provided for main-stem and tributary habitat in the United States (table 8) and Canada (table 9).

Steelhead Intrinsic Potential Based on the Puget Sound Technical Recovery Team Model The Puget Sound Technical Recovery Team model identified 880 low (94.4 km), 573 medium (62.1 km), and 1382 high (149.0 km) IP reaches for steelhead across all 25 targeted streams (fig. 12), with additional 919 low (92.6 km), 428 medium (44.2 km), and 268 high (27.5 km) IP reaches identified from non-target reaches (fig. 13). Most of the high IP steelhead habitat was found in tributaries (145.6 km), with fewer high IP steelhead reaches (totaling 3.3.km) found in the main-stem Skagit River (fig. 14). There were 19 targeted tributaries with greater than 1 km of high IP habitat, all targeted streams except Slate Creek, Twentysix Mile Creek, North Fork Canyon Creek, Maselpanik Creek, Marmotte Creek, and Hozomeen Creek (with five of these six having only low IP habitat for steelhead).

Steelhead Intrinsic Potential Based on the Agrawal Model The Agrawal model identified 28 low (2.5 km), 842 medium (92.3 km), and 1964 high (210.4 km) intrinsic potential reaches for steelhead across all 25 targeted streams (fig. 12), with additional 326 low (30.9 km), 970 medium (99.5 km), and 319 high (33.8 km) IP reaches identified from non-target reaches (fig. 13). As seen in the Puget Sound TRT steelhead model IP results, most of the high IP habitat was found in tributaries (195.5 km) compared with the 14.9 km in the main-stem Skagit River (fig. 15). There were 19 targeted streams that had greater than 1 km of high IP habitat, with 5 of the 6 remaining targeted streams having 329–952 m of high IP habitat. Of the non-targeted streams with names, Thunder Creek (6.8 km) and Cinnamon Creek (1.8 km) had greater than 1 km of high IP habitat, in addition to 22.7 km of high IP habitat found in unnamed streams.

Steelhead Intrinsic Potential Based on the Burnett Model The Burnett model identified 205 low (21.3 km), 658 medium (71.4 km), and 1971 high (212.5 km) IP reaches for steelhead across all 25 targeted streams (fig. 12), with an additional 858 low (84.9 km), 387 medium (40.2 km), and 370 high (39.2 km) IP reaches identified from non-target reaches (fig. 13). As seen with the other two steelhead IP models, most of the high IP habitat was found in tributaries (186.1 km) compared with the 26.4 km of high IP habitats in the main-stem Skagit River (fig. 16). There were 19 targeted streams that had greater than 1 km of high IP habitat, with 3 of the 6 having 111–703 m of high IP habitat, in addition to 28.9 km of high IP habitat found in unnamed streams.

32 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington Skagit River Sumallo River Little Beaver Creek Big Beaver Creek Granite Creek Lightning Creek Panther Creek Klesilkwa River Canyon Creek Devils Creek Nepopekum Creek Three Fools Creek Castle Fork Creek Stetattle Creek McMillan Creek Ruby Creek Ferguson Creek Snass Creek Luna Creek Maselpanik Creek Slate Creek Twentysix Mile Creek North Fork Canyon Creek Marmotte Creek Hozomeen Creek Skagit River Sumallo River Little Beaver Creek Big Beaver Creek Granite Creek Lightning Creek Panther Creek Klesilkwa River Canyon Creek Devils Creek Nepopekum Creek Three Fools Creek Castle Fork Creek Stetattle Creek McMillan Creek Ruby Creek Ferguson Creek Snass Creek Luna Creek Maselpanik Creek Slate Creek Twentysix Mile Creek North Fork Canyon Creek Marmotte Creek Hozomeen Creek

Explanation High intrinsic potential Medium intrinsic potential Low intrinsic potential Species: Steelhead Model: Puget Sound TRT

Explanation High intrinsic potential Medium intrinsic potential Low intrinsic potential Species: Steelhead Model: Agrawal

Explanation High intrinsic potential Medium intrinsic potential Low intrinsic potential Species: Steelhead Model: Burnett

15 20 25 Steelhead intrinsic potential, in river kilometers

Figure 12. Length in kilometers of high, medium, and low habitat intrinsic potential for steelhead (Oncorhynchus mykiss) based on three different models for target tributaries upstream from the three upper Skagit River dams, in northern Washington.

Results and Interpretations 33 Table 8. Reach average values of gradient, valley width index (ratio), bankfull width, and mean annual flow for targeted upper Skagit River streams in the United States by high, medium, or low intrinsic potential (IP) rank for steelhead (Oncorhynchus mykiss). [Plus or minus (±) standard deviations are shown in parentheses. PS_TRT, Puget Sound Technical Recovery Team model; NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Stream

Model

Number of reaches

Valley width ratio

Bankfull width (m)

Mean annual flow (m3/s)

Big Beaver Creek

PS_TRT

120

High

Big Beaver Creek

Agrawal

102

High

1.3 (±1.0)

0.7 (±0.3)

12.8 (±3.0)

3.7 (±1.5)

1.9 (±1.7)

0.6 (±0.1)

13.2 (±2.1)

3.8 (±1.3)

Big Beaver Creek

Burnett

150

Big Beaver Creek

PS_TRT

High

1.3 (±1.2)

0.6 (±0.3)

13.3 (±2.9)

4.0 (±1.5)

Medium

0.1 (±0.1)

0.5 (±0.0)

16.0 (±0.5)

5.6 (±0.4)

Big Beaver Creek

Agrawal

119

Medium

0.8 (±2.1)

0.6 (±0.3)

14.7 (±3.0)

4.9 (±1.5)

Big Beaver Creek Big Beaver Creek

Burnett

Medium

0.4 (±1.5)

0.5 (±0.1)

15.8 (±1.1)

5.5 (±0.8)

PS_TRT

Low

7.0 (±2.9)

0.6 (±0.1)

12.4 (±0.9)

3.3 (±0.6)

Big Beaver Creek Big Beaver Creek

Agrawal

Low

1.4 (±4.1)

0.5 (±0.1)

15.7 (±1.5)

5.5 (±1.1)

Burnett

Low

4.8 (±5.4)

0.6 (±0.1)

14.3 (±2.2)

4.5 (±1.5)

Little Beaver Creek

PS_TRT

189

High

1.7 (±0.8)

0.7 (±0.2)

12.0 (±2.4)

3.1 (±1.3)

Little Beaver Creek

Agrawal

209

High

2.3 (±1.7)

0.7 (±0.2)

11.7 (±2.7)

3.0 (±1.5)

Little Beaver Creek

Burnett

209

High

1.8 (±1.1)

0.7 (±0.2)

11.9 (±2.5)

3.1 (±1.4)

Little Beaver Creek

PS_TRT

Medium

0.1 (±0.1)

0.6 (±0.1)

13.9 (±0.9)

4.2 (±0.6)

Little Beaver Creek

Agrawal

Medium

0.5 (±1.9)

0.6 (±0.1)

13.5 (±1.6)

4.0 (±0.9)

Little Beaver Creek

Burnett

Medium

1.9 (±2.9)

0.6 (±0.2)

13.4 (±2.7)

4.0 (±1.4)

Little Beaver Creek

PS_TRT

Low

6.1 (±1.8)

0.8 (±0.3)

10.4 (±3.8)

2.6 (±1.9)

Little Beaver Creek

Agrawal

Low

Little Beaver Creek

Burnett

Low

8.6 (±1.1)

0.9 (±0.3)

8.7 (±3.9)

1.8 (±1.9)

Ruby Creek

PS_TRT

High

1.5 (±0.3)

0.5 (±0.2)

19.1 (±0.4)

8.4 (±0.4)

Ruby Creek

Agrawal

High

1.6 (±0.4)

0.5 (±0.2)

19.8 (±1.3)

9.1 (±1.4)

Ruby Creek

Burnett

High

1.5 (±0.4)

0.5 (±0.2)

19.8 (±1.3)

9.1 (±1.4)

Ruby Creek

PS_TRT

Medium

1.7 (±0.6)

0.4 (±0.1)

21.8 (±1.0)

11.3 (±1.1)

Ruby Creek

Agrawal

Medium

0.1 (±0.1)

0.3 (±0.1)

22.2 (±0.1)

11.6 (±0.1)

Ruby Creek

Burnett

Medium

0.0 (NA)

0.3 (NA)

22.2 (NA)

11.7 (NA)

Ruby Creek

PS_TRT

Low

0.1 (±0.1)

0.3 (±0.1)

22.2 (±0.1)

11.6 (±0.1)

Ruby Creek

Agrawal

Low

Ruby Creek

Burnett

Low

Canyon Creek

PS_TRT

115

High

2.3 (±0.8)

0.7 (±0.2)

12.6 (±1.8)

3.4 (±1.0)

Canyon Creek

Agrawal

152

High

3.2 (±2.0)

0.8 (±0.5)

11.7 (±2.7)

3.0 (±1.3)

Canyon Creek

Burnett

132

High

2.5 (±1.1)

0.7 (±0.3)

12.4 (±2.1)

3.3 (±1.1)

Canyon Creek

PS_TRT

Medium

0.1 (±0.1)

0.7 (±0.0)

12.4 (±1.0)

3.2 (±0.6)

Canyon Creek

Agrawal

Medium

0.1 (±0.1)

0.7 (±0.0)

12.4 (±1.0)

3.2 (±0.6)

Canyon Creek

Burnett

Medium

5.4 (±2.3)

1.3 (±08)

8.4 (±2.8)

1.5 (±1.2)

Canyon Creek

PS_TRT

Low

6.1 (±1.5)

1.3 (±0.8)

8.9 (±3.2)

1.8 (±1.4)

Canyon Creek

Agrawal

Low

Canyon Creek

Burnett

Low

8.1 (±0.6)

North Fork Canyon Creek PS_TRT

High

North Fork Canyon Creek Agrawal

High

7.6 (±1.1)

North Fork Canyon Creek Burnett

High

North Fork Canyon Creek PS_TRT

Medium

North Fork Canyon Creek Agrawal

Medium

IP rank

Percent gradient

NA 10.5 (±0.7)

1.7 (±0.8)

7.7 (±2.6)

1.3 (±1.1)

1.4 (±0.4)

6.4 (±0.0)

0.7 (±0.0)

1.6 (±0.6)

6.3 (±0.0)

0.7 (±0.0)

34 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington Table 8. Reach average values of gradient, valley width index (ratio), bankfull width, and mean annual flow for targeted upper Skagit River streams in the United States by high, medium, or low intrinsic potential (IP) rank for steelhead (Oncorhynchus mykiss).—Continued [Plus or minus (±) standard deviations are shown in parentheses. PS_TRT, Puget Sound Technical Recovery Team model; NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Number of reaches

IP rank

Percent gradient

North Fork Canyon Creek Burnett

Medium

North Fork Canyon Creek PS_TRT

Low

North Fork Canyon Creek Agrawal

Low

North Fork Canyon Creek Burnett

Low

9.4 (±1.5)

1.6 (±0.4)

6.3 (±0.0)

0.7 (±0.0)

Stream

Model

Valley width ratio

Bankfull width (m)

Mean annual flow (m3/s)

6.4 (NA)

1.0 (NA)

6.4 (NA)

0.7 (NA)

9.0 (±1.8)

1.5 (±0.5)

6.4 (±0.0)

0.7 (±0.0)

Granite Creek

PS_TRT

168

High

2.2 (±0.8)

0.9 (±0.3)

10.4 (±1.8)

2.3 (±0.9)

Granite Creek

Agrawal

184

High

3.1 (±1.5)

0.8 (±0.3)

10.8 (±1.7)

2.4 (±0.8)

Granite Creek

Burnett

198

High

2.6 (±1.2)

0.9 (±0.3)

10.5 (±1.8)

2.3 (±0.8)

Granite Creek

PS_TRT

Medium

Granite Creek

Agrawal

Medium

1.6 (±0.5)

1.1 (±0.4)

9.1 (±0.6)

1.6 (±0.2)

Granite Creek

Burnett

Medium

5.9 (±0.4)

0.6 (±0.1)

11.2 (±0.6)

2.5 (±0.3)

Granite Creek

PS_TRT

Low

5.2 (±0.9)

0.7 (±0.2)

10.9 (±1.0)

2.4 (±0.4)

Granite Creek

Agrawal

Low

Granite Creek

Burnett

Low

7.8 (±0.4)

0.6 (±0.1)

10.7 (±0.5)

2.3 (±0.2)

Lightning Creek

PS_TRT

136

High

2.2 (±0.9)

0.7 (±0.4)

12.1 (±2.7)

3.2 (±1.7)

Lightning Creek

Agrawal

185

High

3.1 (±1.8)

0.8 (±0.4)

11.4 (±2.9)

2.9 (±1.7)

Lightning Creek

Burnett

169

High

2.7 (±1.3)

0.8 (±0.4)

11.6 (±2.8)

3.0 (±1.7)

Lightning Creek

PS_TRT

Medium

0.0 (±0.0)

1.4 (±0.6)

7.6 (±0.4)

1.1 (±0.1)

Lightning Creek

Agrawal

Medium

1.4 (±3.0)

1.2 (±0.6)

8.7 (±1.7)

1.5 (±0.7)

Lightning Creek

Burnett

Medium

3.9 (±3.1)

1.0 (±0.5)

8.9 (±2.2)

1.7 (±1.2)

Lightning Creek

PS_TRT

Low

5.6 (±1.4)

1.0 (±0.4)

9.4 (±2.1)

1.9 (±1.2)

Lightning Creek

Agrawal

Low

Lightning Creek

Burnett

Low

8.3 (±1.0)

1.1 (±0.3)

8.5 (±1.6)

1.4 (±0.6)

Luna Creek

PS_TRT

High

1.6 (±1.0)

0.8 (±0.1)

9.0 (±0.3)

1.6 (±0.1)

Luna Creek

Agrawal

High

3.5 (±2.0)

0.9 (±0.1)

8.7 (±0.3)

1.5 (±0.1)

Luna Creek

Burnett

High

2.0 (±1.5)

0.8 (±0.1)

9.0 (±0.3)

1.6 (±0.1)

Luna Creek

PS_TRT

Medium

0.1 (±0.1)

0.9 (±0.1)

9.0 (±0.1)

1.6 (±0.0)

Luna Creek

Agrawal

Medium

1.3 (±1.5)

0.8 (±0.1)

9.1 (±0.3)

1.6 (±0.1)

Luna Creek

Burnett

Medium

3.8 (±3.0)

0.9 (±0.1)

8.1 (±0.4)

1.5 (±0.2)

Luna Creek

PS_TRT

Low

5.3 (±1.1)

0.9 (±0.1)

8.6 (±0.4)

1.4 (±0.1)

Luna Creek

Agrawal

Low

Luna Creek

Burnett

Low

7.5 (NA)

0.9 (NA)

8.1 (NA)

1.2 (NA)

Three Fools Creek

PS_TRT

High

2.9 (±0.8)

0.8 (±0.2)

10.8 (±1.4)

2.4 (±0.6)

Three Fools Creek

Agrawal

102

High

3.7 (±1.6)

0.9 (±0.4)

10.6 (±1.5)

2.3 (±0.7)

Three Fools Creek

Burnett

High

3.3 (±1.1)

0.9 (±0.4)

10.5 (±1.5)

2.3 (±0.7)

Three Fools Creek

PS_TRT

Medium

0.2 (NA)

0.9 (NA)

9.9 (NA)

1.9 (NA)

Three Fools Creek

Agrawal

Medium

4.9 (±6.6)

0.8 (±0.1)

11.0 (±1.6)

2.5 (±0.8)

Three Fools Creek

Burnett

Medium

6.1 (±0.7)

0.8 (±0.2)

10.4 (±1.0)

2.2 (±0.5)

Three Fools Creek

PS_TRT

Low

5.8 (±2.0)

1.0 (±0.6)

10.1 (±1.9)

2.1 (±0.8)

Three Fools Creek

Agrawal

Low

14.0 (NA)

1.7 (NA)

4.9 (NA)

0.4 (NA)

Three Fools Creek

Burnett

Low

8.6 (±2.1)

0.8 (±0.3)

11.0 (±2.4)

2.6 (±0.9)

Castle Fork Creek

PS_TRT

High

2.6 (±0.8)

2.1 (±1.4)

5.9 (±0.7)

0.6 (±0.1)

Results and Interpretations 35 Table 8. Reach average values of gradient, valley width index (ratio), bankfull width, and mean annual flow for targeted upper Skagit River streams in the United States by high, medium, or low intrinsic potential (IP) rank for steelhead (Oncorhynchus mykiss).—Continued [Plus or minus (±) standard deviations are shown in parentheses. PS_TRT, Puget Sound Technical Recovery Team model; NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Stream

Model

Number of reaches

Castle Fork Creek

Agrawal

High

6.2 (±2.1)

Castle Fork Creek

Burnett

High

3.7 (±1.3)

Castle Fork Creek

PS_TRT

Medium

Castle Fork Creek

Agrawal

Medium

10.5 (±0.6)

2.5 (±0.4)

3.9 (±0.1)

0.2 (±0.1)

Castle Fork Creek

Burnett

Medium

6.2 (±0.5)

2.0 (±1.1)

5.1 (±1.3)

0.5 (±0.2)

Castle Fork Creek

PS_TRT

Low

7.8 (±2.5)

2.1 (±0.8)

4.6 (±1.5)

0.4 (±0.3)

Castle Fork Creek

Agrawal

Low

13.3 (±1.4)

2.6 (±0.7)

3.1 (±1.0)

0.2 (±0.1)

Castle Fork Creek

Burnett

Low

9.4 (±2.1)

2.3 (±0.7)

4.0 (±1.3)

0.3 (±0.2)

Panther Creek

PS_TRT

High

2.7 (±0.8)

1.1 (±0.5)

8.6 (±1.3)

1.5 (±0.4)

Panther Creek

Agrawal

187

High

4.3 (±1.9)

1.2 (±0.9)

8.4 (±2.2)

1.5 (±0.8)

Panther Creek

Burnett

131

High

3.3 (±1.2)

1.2 (±0.8)

8.4 (±1.9)

1.4 (±0.6)

Panther Creek

PS_TRT

Medium

0.1 (NA)

0.9 (NA)

9.0 (NA)

1.6 (NA)

Panther Creek

Agrawal

Medium

6.1 (±3.0)

1.2 (±0.4)

6.9 (±3.9)

1.1 (±1.3)

Panther Creek

Burnett

Medium

5.9 (±0.5)

1.4 (±1.2)

8.3 (±3.0)

1.6 (±1.0)

Panther Creek

PS_TRT

Low

5.9 (±1.3)

1.4 (±1.1)

8.2 (±2.9)

1.5 (±1.0)

Panther Creek

Agrawal

Low

Panther Creek

Burnett

Low

7.9 (±0.7)

1.2 (±0.6)

8.8 (±2.7)

1.7 (±1.0)

Devils Creek

PS_TRT

High

3.0 (±0.7)

1.2(±0.4)

7.3 (±1.3)

1.0 (±0.4)

Devils Creek

Agrawal

132

High

4.6 (±1.4)

1.3 (±0.5)

7.4 (±1.5)

1.1 (±0.4)

Devils Creek

Burnett

High

3.9 (±1.0)

1.3 (±0.5)

7.3 (±1.4)

1.0 (±0.4)

Devils Creek

PS_TRT

Medium

Devils Creek

Agrawal

Medium

7.4 (±4.7)

1.2 (±0.7)

7.7 (±2.7)

1.2 (±0.7)

Devils Creek

Burnett

Medium

5.9 (±0.4)

1.1 (±0.4)

8.0 (±1.4)

1.2 (±0.4)

Devils Creek

PS_TRT

Low

5.5 (±1.3)

1.3 (±0.5)

7.5 (±1.7)

1.1 (±0.4)

Devils Creek

Agrawal

Low

Devils Creek

Burnett

Low

8.5 (±1.1)

1.4 (±0.6)

6.9 (±2.7)

1.0 (±0.7)

Stetattle Creek

PS_TRT

High

2.5 (±1.0)

0.8 (±0.2)

11.1 (±0.7)

2.5 (±0.3)

Stetattle Creek

Agrawal

High

4.4 (±2.1)

0.9 (±0.2)

10.2 (±1.6)

2.2 (±0.7)

Stetattle Creek

Burnett

High

2.8 (±1.3)

1.0 (±1.0)

10.7 (±1.5)

2.3 (±0.5)

Stetattle Creek

PS_TRT

Medium

1.7 (±1.7)

4.3 (±3.7)

6.3 (±4.6)

1.1 (±1.2)

Stetattle Creek

Agrawal

Medium

2.6 (±1.9)

4.6 (±4.2)

6.2 (±4.3)

1.1 (±1.1)

Stetattle Creek

Burnett

Medium

6.0 (±0.8)

1.7 (±2.2)

8.2 (±2.5)

1.4 (±0.8)

Stetattle Creek

PS_TRT

Low

6.3 (±1.7)

1.2 (±1.4)

9.4 (±2.1)

1.8 (±0.8)

Stetattle Creek

Agrawal

Low

14.4 (NA)

1.1 (NA)

11.6 (NA)

2.8 (NA)

Stetattle Creek

Burnett

Low

8.5 (±2.3)

1.0 (±0.1)

11.2 (±0.6)

2.6 (±0.3)

McMillan Creek

PS_TRT

High

2.5 (±1.1)

3.1 (±6.6)

8.3 (±0.9)

1.3 (±0.3)

McMillan Creek

Agrawal

High

3.9 (±1.9)

3.9 (±7.5)

8.3 (±1.0)

1.3 (±0.3)

McMillan Creek

Burnett

High

2.7 (±1.3)

1.1 (±0.8)

8.5 (±0.8)

1.4 (±0.3)

McMillan Creek

PS_TRT

Medium

0.1 (±0.1)

5.1 (±8.4)

7.7 (±0.6)

1.1 (±0.2)

McMillan Creek

Agrawal

Medium

2.0 (±3.0)

2.6 (±4.5)

7.6 (±1.0)

1.1 (±0.3)

McMillan Creek

Burnett

Medium

3.8 (±2.3)

10.4 (±11.5)

7.4 (±0.9)

1.0 (±0.3)

IP rank

Percent gradient

Bankfull width (m)

Mean annual flow (m3/s)

2.0 (±0.9)

5.1 (±1.3)

0.5 (±0.2)

1.8 (±1.0)

6.0 (±0.8)

0.6 (±0.2)

Valley width ratio

36 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington Table 8. Reach average values of gradient, valley width index (ratio), bankfull width, and mean annual flow for targeted upper Skagit River streams in the United States by high, medium, or low intrinsic potential (IP) rank for steelhead (Oncorhynchus mykiss).—Continued [Plus or minus (±) standard deviations are shown in parentheses. PS_TRT, Puget Sound Technical Recovery Team model; NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Stream

Model

Number of reaches

IP rank

Percent gradient

Valley width ratio

Bankfull width (m)

Mean annual flow (m3/s)

McMillan Creek

PS_TRT

Low

6.5 (±2.4)

4.1 (±7.6)

7.8 (±1.3)

1.2 (±0.4)

McMillan Creek

Agrawal

Low

12.4 (±0.5)

2.6 (±0.7)

5.5 (±0.0)

0.5 (±0.0)

McMillan Creek

Burnett

Low

9.4 (±2.2)

2.1 (±1.3)

7.8 (±1.9)

1.2 (±0.6)

Slate Creek

PS_TRT

High

Slate Creek

Agrawal

High

6.4 (±1.2)

1.2 (±0.3)

8.6 (±0.1)

1.4 (±0.0)

Slate Creek

Burnett

High

5.0 (±0.4)

1.0 (±0.2)

8.5 (±0.1)

1.4 (±0.0)

Slate Creek

PS_TRT

Medium

Slate Creek

Agrawal

Medium

9.4 (NA)

1.2 (NA)

8.6 (NA)

1.4 (NA)

Slate Creek

Burnett

Medium

6.3 (±0.4)

1.1 (±0.1)

8.6 (±0.0)

1.4 (±0.0)

Slate Creek

PS_TRT

Low

6.6 (±1.5)

1.2 (±0.2)

8.6 (±0.0)

1.4 (±0.0)

Slate Creek

Agrawal

Low

Slate Creek

Burnett

Low

8.2 (±0.9)

1.3 (±0.3)

8.6 (±0.0)

1.4 (±0.0)

Hozomeen Creek

PS_TRT

High

Hozomeen Creek

Agrawal

High

Hozomeen Creek

Burnett

High

Hozomeen Creek

PS_TRT

Medium

Hozomeen Creek

Agrawal

Medium

Hozomeen Creek

Burnett

Medium

Hozomeen Creek

PS_TRT

Low

12.8 (NA)

1.7 (NA)

4.9 (NA)

0.4 (NA)

Hozomeen Creek

Agrawal

Low

12.8 (NA)

1.7 (NA)

4.9 (NA)

0.4 (NA)

Hozomeen Creek

Burnett

Low

12.8 (NA)

1.7 (NA)

4.9 (NA)

0.4 (NA)

Results and Interpretations 37 Table 9. Reach average values of gradient, valley width index, bankfull width (ratio), and mean annual flow for targeted upper Skagit River streams in Canada by high, medium, and low intrinsic potential rank (IP) for steelhead (Oncorhynchus mykiss). [Plus or minus (±) standard deviations are shown in parentheses. PS_TRT, Puget Sound Technical Recovery Team model; NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Number of reaches

Percent gradient

Valley width ratio

Bankfull width (m)

High

1.4 (±0.9)

1.4 (±0.8)

16.8 (±2.4)

6.4 (±2.0)

High

1.0 (±0.9)

1.5 (±1.0)

21.0 (±3.3)

10.7 (±3.5)

239

High

0.7 (±0.7)

1.5 (±1.0)

22.0 (±3.1)

11.8 (±3.4)

167

Medium

0.7 (±0.6)

2.8 (±12.8)

22.4 (±2.6)

12.1 (±3.1)

Agrawal

225

Medium

0.2 (±0.4)

3.7 (±12.4)

23.5 (±2.6)

13.5 (±3.1)

Skagit River

Burnett

121

Medium

0.1 (±0.2)

5.6 (±16.6)

23.7 (±2.8)

13.8 (±3.4)

Skagit River

PS_TRT

163

Low

0.1 (±0.5)

3.3 (±6.9)

23.9 (±2.3)

13.9 (±2.9)

Skagit River

Agrawal

Low

Skagit River

Burnett

Low

Stream

Model

Skagit River

PS_TRT

Skagit River

Agrawal

135

Skagit River

Burnett

Skagit River

PS_TRT

Skagit River

IP rank

Mean annual flow (m3/s)

Klesilkwa River

PS_TRT

High

1.0 (±0.9)

0.8 (±0.2)

9.8 (±2.5)

2.1 (±1.3)

Klesilkwa River

Agrawal

High

0.9 (±0.9)

0.7 (±0.2)

11.9 (±3.0)

3.1 (±1.5)

Klesilkwa River

Burnett

High

0.9 (±1.0)

0.8 (±0.2)

9.9 (±2.6)

2.1 (±1.3)

Klesilkwa River

PS_TRT

104

Medium

0.0 (±0.0)

0.7 (±0.2)

11.7 (±2.5)

3.0 (±1.3)

Klesilkwa River

Agrawal

150

Medium

0.6 (±1.7)

0.8 (±0.2)

10.8 (±2.7)

2.5 (±1.3)

Klesilkwa River

Burnett

Medium

0.2 (±1.1)

0.7 (±0.2)

11.7 (±2.6)

3.0 (±1.3)

Klesilkwa River

PS_TRT

Low

8.3 (±2.7)

1.1 (±0.1)

7.3 (±0.0)

1.0 (±0.0)

Klesilkwa River

Agrawal

Low

12.9 (NA)

1.1 (NA)

7.2 (NA)

1.0 (NA)

Klesilkwa River

Burnett

Low

10.5 (±1.9)

1.2 (±0.0)

7.3 (±0.0)

1.0 (±0.0)

Sumallo River

PS_TRT

116

High

1.7 (±1.1)

0.8 (±0.2)

10.2 (±2.7)

2.3 (±1.2)

Sumallo River

Agrawal

135

High

2.4 (±1.7)

0.8 (±0.3)

9.8 (±2.8)

2.1 (±1.3)

Sumallo River

Burnett

154

High

1.9 (±1.5)

0.8 (±0.2)

10.2 (±2.7)

2.3 (±1.2)

Sumallo River

PS_TRT

101

Medium

0.0 (±0.1)

0.6 (±0.1)

12.9 (±1.2)

3.5 (±0.7)

Sumallo River

Agrawal

111

Medium

0.2 (±0.6)

0.6 (±0.1)

12.5 (±1.6)

3.3 (±0.9)

Sumallo River

Burnett

Medium

0.3 (±1.4)

0.6 (±0.1)

12.5 (±1.7)

3.4 (±0.9)

Sumallo River

PS_TRT

Low

5.0 (±1.0)

1.0 (±0.2)

7.8 (±1.3)

1.2 (±0.4)

Sumallo River

Agrawal

Low

0.2 (NA)

0.7 (NA)

9.8 (NA)

1.9 (NA)

Sumallo River

Burnett

Low

5.1 (±4.2)

1.1 (±0.4)

6.9 (±2.5)

1.0 (±0.8)

Ferguson Creek

PS_TRT

High

1.3 (±0.7)

1.4 (±0.3)

5.7 (±0.9)

0.6 (±0.2)

Ferguson Creek

Agrawal

High

4.6 (±3.0)

1.6 (±0.3)

4.9 (±0.6)

0.4 (±0.1)

Ferguson Creek

Burnett

High

1.1 (±1.2)

1.4 (±0.3)

5.7 (±0.9)

0.6 (±0.2)

Ferguson Creek

PS_TRT

Medium

0.0 (±0.1)

1.3 (±0.2)

5.7 (±0.5)

0.6 (±0.1)

Ferguson Creek

Agrawal

Medium

0.3 (±0.5)

1.3 (±0.2)

5.7 (±0.6)

0.6 (±0.1)

Ferguson Creek

Burnett

Medium

0.2 (±1.1)

1.3 (±0.2)

5.6 (±0.4)

0.5 (±0.1)

Ferguson Creek

PS_TRT

Low

6.7 (±1.4)

1.6 (±0.1)

4.6 (±0.0)

0.4 (±0.0)

Ferguson Creek

Agrawal

Low

0.1 (NA)

1.2 (NA)

6.0 (NA)

0.6 (NA)

Ferguson Creek

Burnett

Low

6.0 (±3.3)

1.5 (±0.2)

4.9 (±0.6)

0.4 (±0.1)

Nepopekum Creek

PS_TRT

High

2.0 (±1.0)

1.0 (±0.1)

8.8 (±0.2)

1.5 (±0.1)

Nepopekum Creek

Agrawal

High

3.9 (±2.0)

0.9 (±0.1)

8.6 (±0.2)

1.4 (±0.1)

38 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington Table 9. Reach average values of gradient, valley width index, bankfull width (ratio), and mean annual flow for targeted upper Skagit River streams in Canada by high, medium, and low intrinsic potential rank (IP) for steelhead (Oncorhynchus mykiss).—Continued [Plus or minus (±) standard deviations are shown in parentheses. PS_TRT, Puget Sound Technical Recovery Team model; NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Number of reaches

Percent gradient

Valley width ratio

Bankfull width (m)

Mean annual flow (m3/s)

High

2.4 (±1.4)

0.9 (±0.1)

8.8 (±0.2)

1.5 (±0.1)

Medium

0.0 (±0.1)

0.9 (±0.1)

8.9 (±0.2)

1.5 (±0.1)

Agrawal

Medium

1.4 (±2.7)

0.9 (±0.1)

8.9 (±0.2)

1.5 (±0.1)

Nepopekum Creek

Burnett

Medium

3.0 (±3.1)

0.9 (±0.1)

8.7 (±0.3)

1.5 (±0.1)

Nepopekum Creek

PS_TRT

Low

6.0 (±1.9)

0.9 (±0.1)

8.5 (±0.2)

1.4 (±0.1)

Nepopekum Creek

Agrawal

Low

Nepopekum Creek

Burnett

Low

9.0 (±1.4)

0.9 (±0.1)

8.5 (±0.2)

1.4 (±0.1)

Maselpanik Creek PS_TRT

High

2.1 (±0.9)

0.7 (±0.1)

10.4 (±0.2)

2.2 (±0.1)

Maselpanik Creek Agrawal

High

7.1 (±3.5)

0.8 (±0.2)

10.4 (±0.2)

2.2 (±0.1)

Maselpanik Creek Burnett

High

2.1 (±0.9)

0.7 (±0.1)

10.4 (±0.2)

2.2 (±0.1)

Maselpanik Creek PS_TRT

Medium

Maselpanik Creek Agrawal

Medium

4.4 (±3.1)

0.7 (±0.1)

10.4 (±0.2)

2.2 (±0.1)

Maselpanik Creek Burnett

Medium

6.0 (NA)

0.7 (NA)

10.3 (NA)

2.1 (NA)

Maselpanik Creek PS_TRT

Low

8.0 (±1.5)

0.7 (±0.2)

10.5 (±0.2)

2.2 (±0.1)

Maselpanik Creek Agrawal

Low

Maselpanik Creek Burnett

Low

8.5 (±1.2)

0.7 (±0.2)

10.5 (±0.2)

2.2 (±0.1)

Snass Creek

PS_TRT

High

2.3 (±1.1)

1.3 (±0.2)

6.7 (±0.8)

0.8 (±0.2)

Snass Creek

Agrawal

High

4.4 (±2.4)

1.2 (±0.3)

6.7 (±0.8)

0.8 (±0.2)

Snass Creek

Burnett

High

3.0 (±1.6)

1.3 (±0.3)

6.7 (±0.8)

0.8 (±0.2)

Snass Creek

PS_TRT

Medium

0.2 (NA)

1.6 (NA)

5.1 (NA)

0.5 (NA)

Snass Creek

Agrawal

Medium

7.9 (±4.1)

1.1 (±0.2)

6.8 (±1.0)

0.8 (±0.2)

Snass Creek

Burnett

Medium

6.3 (±0.5)

1.2 (±0.3)

6.5 (±0.9)

0.8 (±0.2)

Snass Creek

PS_TRT

Low

7.2 (±2.4)

1.2 (±0.3)

6.8 (±0.8)

0.9 (±0.2)

Snass Creek

Agrawal

Low

12.0 (NA)

0.9 (NA)

7.2 (NA)

0.9 (NA)

Snass Creek

Burnett

Low

9.6 (±1.6)

1.0 (±0.2)

6.9 (±0.9)

0.9 (±0.2)

Twentysix Mile Creek

PS_TRT

High

Twentysix Mile Creek

Agrawal

High

5.5 (±0.2)

1.1 (±0.2)

6.3 (±0.0)

0.7 (±0.0)

Twentysix Mile Creek

Burnett

High

5.3 (NA)

0.9 (NA)

6.3 (NA)

0.7 (NA)

Twentysix Mile Creek

PS_TRT

Medium

Twentysix Mile Creek

Agrawal

Medium

10.5 (NA)

1.1 (NA)

6.3 (NA)

0.7 (NA)

Twentysix Mile Creek

Burnett

Medium

5.6 (±0.0)

1.2 (±0.0)

6.3 (±0.0)

0.7 (±0.0)

Stream

Model

Nepopekum Creek

Burnett

Nepopekum Creek

PS_TRT

Nepopekum Creek

IP rank

Results and Interpretations 39 Table 9. Reach average values of gradient, valley width index, bankfull width (ratio), and mean annual flow for targeted upper Skagit River streams in Canada by high, medium, and low intrinsic potential rank (IP) for steelhead (Oncorhynchus mykiss).—Continued [Plus or minus (±) standard deviations are shown in parentheses. PS_TRT, Puget Sound Technical Recovery Team model; NA, no reaches were identified for a given stream, model, and intrinsic potential category; m, meter; m3/s, cubic meter per second]

Number of reaches

Percent gradient

Valley width ratio

Bankfull width (m)

Mean annual flow (m3/s)

Low

10.1 (±3.1)

1.1 (±0.2)

6.3 (±0.0)

0.7 (±0.0)

Low

12.7 (±0.5)

1.2 (±0.2)

6.3 (±0.0)

0.7 (±0.0)

Burnett

Low

11.6 (±1.4)

1.2 (±0.2)

6.3 (±0.0)

0.7 (±0.0)

Marmotte Creek

PS_TRT

High

Marmotte Creek

Agrawal

High

7.1 (±1.4)

1.6 (±0.3)

4.7 (±0.0)

0.4 (±0.0)

Marmotte Creek

Burnett

High

Marmotte Creek

PS_TRT

Medium

Marmotte Creek

Agrawal

Medium

Marmotte Creek

Burnett

Medium

6.4 (±0.8)

0.6 (±0.4)

4.8 (±0.0)

0.4 (±0.0)

Marmotte Creek

PS_TRT

Low

7.1 (±1.4)

1.6 (±0.3)

4.7 (±0.0)

0.4 (±0.0)

Marmotte Creek

Agrawal

Low

Marmotte Creek

Burnett

Low

8.6 (NA)

1.7 (NA)

4.7 (NA)

0.4 (NA)

Model

Twentysix Mile Creek

PS_TRT

Twentysix Mile Creek

Agrawal

Twentysix Mile Creek

Steelhead intrinsic potential, in river kilometers

Stream

100 80 60 40 20 0

IP rank

Puget Sound TRT

High Medium Low

100 80 60 40 20 0

Agrawal

High Medium Low

100 80 60 40 20 0

Burnett

High Medium Low

Intrinsic potential category Figure 13. Length in kilometers of high, medium, and low habitat intrinsic potential for steelhead (Oncorhynchus mykiss) based on three different models for non-target tributaries upstream from the three upper Skagit River dams, in northern Washington.

40 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington 121°30' W

121° W

120°30' W

EXPLANATION

it R ag

Cr.

HUC12 watershed boundary

Sna ss

o R.

Hydroelectric dam Steelhead Intrinsic Potential Puget Sound TRT Model

r. ile C ysix M Twent

i l k wa R.

Mas elpa nik C

Kles

Reservoir

rg Fe

Su ma ll

Cr. on us

po p e k Ne

High Medium

Cr.

Low

49° N

or k

. Cas

Cr ver

Cr.

tle F

Three Fool s Cr

e ittle B

Lightning C

49° N

Lu Big

Devil s Cr.

B eaver C r.

. Cr na . an Cr ill

t tl

Cr .

ta Ste

Cr . ROSS DAM

Ruby C r.

DIABLO DAM

Can

GORGE DAM

Cr .

t he r

te C ani Gr

C r. der Thu n

Skagit River

48°30' N 121°30' W

Base modified from U.S. Geological Survey 1:100,000-scale data. Projection: NAD1983, NAVD 1988.

121° W

120°30' W

0 0

10 10

20 MILES 20 KILOMETERS

Figure 14. High, medium, and low intrinsic potential scores for steelhead (Oncorhynchus mykiss) derived from the Puget Sound Technical Recovery Team intrinsic potential model applied to tributary and main-stem habitat upstream from the Skagit River Hydroelectric Project dams, in northern Washington. USGS Hydrological Unit Code (HUC) is a classification hierarchy to delimit watershed boundaries in successively smaller hydrologic units. HUC12 indicates boundaries of local sub-watershed scale that represents tributary systems.

Results and Interpretations 41 121°30' W

121° W

120°30' W

EXPLANATION

it R ag

Cr.

HUC12 watershed boundary

Sna ss

o R.

Hydroelectric dam Steelhead Intrinsic Potential Agrawal model

r. ile C ysix M Twent

i l k wa R.

Mas elpa nik C

Kles

Reservoir

rg Fe

Su ma ll

Cr. on us

po p e k Ne

High (>0.75) Medium (0.25–0.75)

Cr.

Low (<0.25)

49° N

or k

. Cas

Cr ver

Cr.

tle F

Three Fool s Cr

e ittle B

Lightning C

49° N

Lu Big

Devil s Cr.

B eaver C r.

. Cr na . an Cr ill

t tl

Cr .

ta Ste

Cr . ROSS DAM

Ruby C r.

DIABLO DAM

Can

GORGE DAM

Cr .

t he r

te C ani Gr

C r. der Thu n

Skagit River

48°30' N 121°30' W

Base modified from U.S. Geological Survey 1:100,000-scale data. Projection: NAD1983, NAVD 1988.

121° W

120°30' W

0 0

10 10

20 MILES 20 KILOMETERS

Figure 15. High (greater than [>] 0.75), medium (0.25–0.75), and low (less than [<] 0.25) intrinsic potential scores for steelhead (Oncorhynchus mykiss) derived from the Agrawal intrinsic potential model applied to tributary and main-stem habitat upstream from the Skagit River Hydroelectric Project dams, in northern Washington. USGS Hydrological Unit Code (HUC) is a classification hierarchy to delimit watershed boundaries in successively smaller hydrologic units. HUC12 indicates boundaries of local subwatershed scale that represents tributary systems.

42 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington 121°30' W

121° W

120°30' W

EXPLANATION

it R ag

Cr.

HUC12 watershed boundary

Sna ss

o R.

Hydroelectric dam Steelhead Intrinsic Potential Burnett model

r. ile C ysix M Twent

i l k wa R.

Mas elpa nik C

Kles

Reservoir

rg Fe

Su ma ll

Cr. on us

po p e k Ne

High (>0.75) Medium (0.25–0.75)

Cr.

Low (<0.25)

49° N

or k

. Cas

Cr ver

Cr.

tle F

Three Fool s Cr

e ittle B

Lightning C

49° N

Lu Big

Devil s Cr.

B eaver C r.

. Cr na . an Cr ill

t tl

Cr .

ta Ste

Cr . ROSS DAM

Ruby C r.

DIABLO DAM

Can

GORGE DAM

Cr .

t he r

te C ani Gr

C r. der Thu n

Skagit River

48°30' N 121°30' W

Base modified from U.S. Geological Survey 1:100,000-scale data. Projection: NAD1983, NAVD 1988.

121° W

120°30' W

0 0

10 10

20 MILES 20 KILOMETERS

Figure 16. High (greater than [>] 0.75), medium (0.25–0.75), and low (less than [<] 0.25) intrinsic potential scores for steelhead (Oncorhynchus mykiss) derived from the Burnett intrinsic potential model applied to tributary and main-stem habitat upstream from the Skagit River Hydroelectric Project dams, in northern Washington. USGS Hydrological Unit Code (HUC) is a classification hierarchy to delimit watershed boundaries in successively smaller hydrologic units. HUC12 indicates boundaries of local subwatershed scale that represents tributary systems.

References Cited 43

Summary

References Cited

We used intrinsic potential (IP) models to evaluate the potential habitat for anadromous salmonids upstream from three impassable dams in the upper Skagit River in Whatcom County, northern Washington. This was based on a two-step process within a synthetic geographic information system-based representation of the upper Skagit River Basin derived from 10-meter (m; on United States side of the international border) or 20-m (on the Canadian side) digital elevation models. The first step was estimating the upper extent of potential anadromous fish distribution within select tributaries. We used existing fish passage criteria from the State of Washington (Washington Department of Fish and Wildlife, 2019) based on waterfall height and stream gradient to filter continuous reaches of stream habitat from the confluence of tributary with a reservoir in the upper Skagit River to the first likely upstream barrier, thus defining the likely upstream extent of fish upstream migration. The length estimates derived from the resulting maps were similar to assessments of targeted streams compiled by stakeholders to the Federal Energy Regulatory Commission relicensing process. The second step was to apply existing IP models for target species (coho, Chinook, steelhead [Oncorhynchus kisutch, O. tshawytscha, O. mykiss, respectively]). Intrinsic potential models are general models that were created to estimate habitat potential based on generalized and broad-scale patterns of each species. Our use of three different IP models for each species was an attempt to obtain a consensus view of the amount and quality of habitat available in select tributaries that would be likely for anadromous fish to move into should an introduction program be initiated. With habitat preference curves derived from different populations and regions, the IP models showed similarities and differences in the amount of habitat in each IP category. For coho and steelhead, IP models with habitat suitability curves derived for Skagit River Chinook (Connor and others, 2015) or Puget Sound steelhead behaved differently than the other IP models based on other regions. This is an important point to consider when comparing the differences among the IP models for these species. When applied to the same “end of fish distribution” maps describing potential tributary habitat in the upper Skagit River watershed, the IP models found that the most high IP habitat existed for steelhead, whereas coho and Chinook IP models found that most stream sections were rated as having medium or low IP. These assessments, when coupled with other habitat models to estimate juvenile rearing capacity (for example, Cramer and Ackerman, 2009; Cooper and others 2020; Ramos and Ward, 2022) and bioenergetic growth potential (Weber and others, 2014; Thompson and Beauchamp, 2016) are a useful first step in evaluating the feasibility of potential anadromous salmonid introduction programs.

Agrawal, A., Schick, R.S., Bjorkstedt, E.P., Szerlong, R.G., Goslin, M.N., Spence, B.C., Williams, T.H., and Burnett, K.M., 2005, Predicting the potential for historical coho, Chinook, and steelhead habitat in northern California: Santa Cruz, California, National Oceanic and Atmospheric Administration, Technical Memorandum NOAA-TM-NMFS-SWFSC-379, 34 p. Benda, L., Miller, D., Andras, K., Bigelow, P., Reeves, G., and Michael, D., 2007, NetMap—A new tool in support of watershed science and resource management: Forest Science, v. 53, no. 2, p. 206–219. Benda, L., Miller, D., and Barquin, J., 2011, Creating a catchment scale perspective for river restoration: Hydrology and Earth System Sciences, v. 15, no. 9, p. 2995–3015. Benda, L., Miller, D., Barquin, J., McCleary, R., Cai, T., and Ji, Y., 2016, Building virtual watersheds—A global opportunity to strengthen resource management and conservation: Environmental Management, v. 57, no. 3, p. 722–739. Benda, L., Poff, N.L., Miller, D., Dunne, T., Reeves, G.H., Pess, G., and Pollock, M., 2004, The network dynamics hypothesis—How channel networks structure riverine habitats: Bioscience, v. 54, no. 5, p. 413–427. Bidlack, A.L., Benda, L.E., Miewald, T., Reeves, G.H., and McMahan, G., 2014, Identifying suitable habitat for Chinook salmon across a large, glaciated watershed: Transactions of the American Fisheries Society, v. 143, no. 3, p. 689–699. Burnett, K.M., Reeves, G.H., Miller, D.J., Clarke, S., Vance-Borland, K., and Christiansen, K., 2007, Distribution of salmon-habitat potential relative to landscape characteristics and implications for conservation: Ecological Applications, v. 17, no. 1, p. 66–80. Busch, D.S., Sheer, M., Burnett, K., McElhany, P., and Cooney, T., 2013, Landscape-level model to predict spawning habitat for lower Columbia River fall Chinook salmon (Oncorhynchus tshawytscha): River Research and Applications, v. 29, no. 3, p. 297–312. Clay, C.H., and Eng, P., 2017, Design of fishways and other fish facilities (2d ed.): Boca Raton, Florida, CRC Press. [Also available at https://doi.org/10.1201/9781315141046.]

44 Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington Connor, E., Lowry, E., Smith, D., Ramsden, K., Barkdull, B., Warinner, B., Hartson, R., and Brocksmith, R., 2015, Tributary assessment for potential Chinook salmon rearing habitat and recommendations for prioritizing habitat protection and restoration: Skagit Watershed Council Report, accessed November 22, 2021, at https://www. skagitwatershed.org/wp-content/uploads/Tier-2-Tributary- composite-2015-02-04_final-draft-2.pdf. Cooney, T.D., and Holzer, D.M., 2006, Appendix C—Interior Columbia basin stream type Chinook salmon and steelhead populations—Habitat intrinsic potential analysis: Interior Columbia Basin Technical Recovery Team (ICTRT), Viability criteria review draft: Seattle, Washington, Northwest Fisheries Science Center, 17 p. [Also available at https://www.salmonrecovery.gov/Files/BiologicalOpinions/ 2000/XAppendix_C.pdf.] Cooper, E.J., O’Dowd, A.P., Graham, J.R., Mierau, D.W., Trush, W.J., and Taylor, R., 2020, Salmonid habitat and population capacity estimates for steelhead trout and Chinook salmon upstream of Scott Dam in the Eel River, California: Northwest Science, v. 94, no. 1, p. 70–96. Cramer, S.P., and Ackerman, N.K., 2009, Linking stream carrying capacity for salmonids to habitat features: American Fisheries Society Symposium, v. 71, p. 225–254. ESRI, 2023, ArcGIS Pro version 3.1, accessed April 14, 2023, at https://www.esri.com/en-us/arcgis/products/arcgis-pro/ overview. Finn, R.J.R., Chalifour, L., Gergel, S.E., Hinch, S.G., Scott, D.C., and Martin, T.G., 2021, Quantifying lost and inaccessible habitat for Pacific salmon in Canada’s lower Fraser River: Ecosphere, v. 12, no. 7, p. e03646. Hard, J.J., Myers, J.M., Connor, E.J., Hayman, R.A., Kope, R.G., Lucchetti, G., Marshall, A.R., Pess, G.R., and Thompson, B.E., 2015, Viability criteria for steelhead within the Puget Sound distinct population segment: National Oceanic and Atmospheric Administration Technical Memorandum NMFS-NWFSC-129, 367 p. [Also available at https://doi.org/10.7289/V5/TM-NWFSC-129.] Johnson, G.E., Anglea, S.M., Adams, N.S., and Wik, T.O., 2005, Evaluation of a prototype surface flow bypass for juvenile salmon and steelhead at the powerhouse of Lower Granite Dam, Snake River, Washington, 1996–2000: North American Journal of Fisheries Management, v. 25, no. 1, p. 138–151. Kock, T.J., Ferguson, J.W., Keefer, M.L., and Schreck, C.B., 2021, Review of trap-and-haul for managing Pacific salmonids (Oncorhynchus spp.) in impounded river systems: Reviews in Fish Biology and Fisheries, v. 31, p. 53–94.

Lusardi, R.A., and Moyle, P.B., 2017, Two-way trap and haul as a conservation strategy for anadromous salmonids: Bethesda, Maryland, Fisheries, v. 42, no. 9, p. 478–487. Naughton, G.P., Keefer, M.L., Clabough, T.S., Knoff, M.J., Blubaugh, T.J., Sharpe, C., and Caudill, C.C., 2018, Reservoir provides cool-water refuge for adult Chinook salmon in a trap-and-haul reintroduction program: Marine and Freshwater Research, v. 69, no. 12, p. 1995–2007. Piper, A.T., Rosewarne, P.J., Wright, R.M., and Kemp, P.S., 2020, Using ‘trap and transport’to facilitate seaward migration of landlocked European eel (Anguilla anguilla) from lakes and reservoirs: Fisheries Research, v. 228, p. 105567. Ramos, M., and Ward, D.W., 2022, Modelling the reestablishment of coho salmon (Oncorhynchus kisutch) in Klamath River tributaries after dam removal: Ecology Freshwater Fish, v. 32, no. 1, p. 133–146. Seattle City Light, 2023, Skagit relicensing public documents library: accessed April 14, 2023, at https://www.seattle.gov/ light/skagit/Relicensing/default.htm. Sheer, M.B., Busch, D.S., Gilbert, E., Bayer, J.M., Lanigan, S., Schei, J.L., Burnett, K.M., and Miller, D., 2009, Development and management of fish intrinsic potential data and methodologies—State of the IP 2008 Summary report: Portland, Oregon, Pacific Northwest Aquatic Monitoring Partnership Series 2009–004, 56 p. Terrainworks, 2023, NetMap tools, accessed April 14, 2023, at https://terrainworks.com/get-netmap-virtual-watersheds- and-tools. Thompson, J.N., and Beauchamp, D.A., 2016, Growth of juvenile steelhead Oncorhynchus mykiss under size-selective pressure limited by seasonal bioenergetic and environmental constraints: Journal of Fish Biology, v. 89, no. 3, p. 1720–1739. Washington Department of Fish and Wildlife, 2019, Fish passage inventory, assessment, and prioritization manual: Olympia, Washington, Washington Department of Fish and Wildlife, 284 p. [Also available at https://wdfw.wa.gov/ publications/02061.] Waldman, J.R., and Quinn, T.P., 2022, North American diadromous fishes—Drivers of decline and potential for recovery in the Anthropocene: Science Advances, v. 8, no. 4, p. eabl5486. Ward, J.V., and Stanford, J.A., eds., 1979, The ecology of regulated streams: New York, Plenum Press, 398 p. Weber, N., Bouwes, N., and Jordan, C.E., 2014, Estimation of salmonid habitat growth potential through measurements of invertebrate food abundance and temperature: Canadian Journal of Fisheries and Aquatic Sciences, v. 71, no. 8, p. 1158–1170.

For information about the research in this report, contact Director, Western Fisheries Research Center U.S. Geological Survey 6505 NE 65th Street Seattle, Washington 98115-5016 https://www.usgs.gov/centers/wfrc Manuscript approved on September 20, 2023 Publishing support provided by the U.S. Geological Survey Science Publishing Network, Tacoma Publishing Service Center

Duda and Hardiman—Applying Models to Evaluate Salmon Introduction, Skagit River Hydroelectric Project, Northern Washington—OFR 2023–1077

ISSN 2331-1258 (online) https://doi.org/10.3133/ofr20231077

Applying Intrinsic Potential Models to Evaluate Salmon ( Oncorhynchus spp.) Introduction into Main-Stem and Tributary Habitats Upstream from the Skagit River Hydroelectric Project, Northern Washington

Related documents