US-191/MT-64 
WILDLIFE & TRANSPORTATION 
ASSESSMENT 
Prepared by: 
Elizabeth Fairbank, Kristeen Penrod, Matt Blank, Matthew Bell,              
Anna Wearn, Abigail Breuer,           Marcel Huijser, Robert Ament 
Braden Hance 
Western Transportation Institute 
Center for Large Landscape Conservation Montana State University 
Bozeman, Montana Bozeman, Montana 
largelandscapes.org westerntransportationinstitute.org 
Photo by Holly Pippel  
Photo by Mat Ludin  
US-191/MT-64 Wildlife & Transportation 
Assessment 
Prepared by: 
Elizabeth Fairbank1, Kristeen Penrod1, Anna Wearn1, Matt Blank2, Matthew Bell2, 
Marcel Huijser2, Rob Ament2, Damon Fick2, Abigail Breuer1, and Braden Hance1
Figure 1 Elk on roadway in path of oncoming car. Photo by Holly Pippel. 
1 Center for Large Landscape Conservation 
2 Western Transportation Institute - Montana State University 
Disclaimer 
The contents of this report reflect the views of the authors, who are responsible for the facts and accuracy 
of the data included. The contents do not necessarily reflect the official policies of the Center for Large 
Landscape Conservation, Western Transportation Institute (WTI) or Montana State University (MSU). 
This report does not constitute a standard, specification, or regulation. 
A reasonable effort has been undertaken to make this document compliant with Section 508 of the 
Rehabilitation Act. If you would like assistance, please contact: info@largelandscapes.org. 
Acknowledgements 
Funders: Big Sky Resort Area District, Moonlight Community Foundation, Yellowstone Club Community 
Foundation, and The Volgenau Foundation 
Other Organizations: Big Sky Chamber of Commerce, Big Sky Hikers, Federal Highway Administration, 
Gallatin County Commission, Gallatin Ranger District, Gallatin River Task Force, Hebgen Ranger District, 
Interagency Grizzly Bear Study Team, Dr. Kate Eisele and students, Lone Peak High School, Montana 
Department of Transportation, Montana Fish, Wildlife and Parks, West Yellowstone Town Council, West 
Yellowstone Chamber of Commerce, Yellowstone National Park 
Individuals: Melissa Alder, Danny Bierschwale, Brent Brock, Emily O’Connor, Renee Callahan, Julie 
Cunningham, Maren Dunn, Pat Flowers, Tut Fuentevilla, Dave Gates, Kristen Gardner, Josh Hemingway, 
Kris Inman, Adam Johnson, David Kack, Anna Johnson, Meg O’Leary, Cole Oshiro-Levitt, Kyle Meakins, 
Mike McGrath, Christy Mikulich, Jeremy Mitchell, Erik Morrison, Jenny Muscat, Brad Niva, Rich 
Noonan, Sean O’Callaghan, Duncan & Eva Patten, Jeff Patten, Sara Pearce, Holly Pippel, Caitlin 
Quisenberry, Satoshi Saito, Randy Scarlett, Mike Scholz, Anna Shipley, Dan Sholly, Joe Skinner, Frank 
van Mannen, Dan Walker, Deb Wambach, Jessica Westbrook, Jessie Wiese, Zach Wurtzebach, Rumsey 
Young 
Please note: the parties above have provided support, assistance, or input that have helped shape this report, 
and the authors are grateful for their assistance. Their inclusion is not intended to suggest endorsement 
of this report or its recommendations. 
DOI: https://doi.org/10.15788/1706207536
i 
List of Abbreviations 
AADT Average Annual Daily Traffic 
Assessment US-191/MT-64 Wildlife & Transportation Assessment 
BIL Bipartisan Infrastructure Law 
CGNF Custer Gallatin National Forest 
FHWA Federal Highway Administration 
FWP Montana Fish, Wildlife & Parks 
GPS Global Positioning System 
GYE Greater Yellowstone Ecosystem 
IGBST Interagency Grizzly Bear Study Team 
IIJA Infrastructure Investment and Jobs Act 
km Kilometer 
mi Mile 
MT-64 Montana State Highway 64 (Lone Mountain Trail) 
MDT Montana Department of Transportation 
MHP Montana Highway Patrol 
MSU-WTI Montana State University-Western Transportation Institute 
MWTP Montana Wildlife and Transportation Partnership 
MWA Moving Window Average 
NCDE Northern Continental Divide Ecosystem 
NDVI Normalized Difference Vegetation Index 
RM Road Mile Post 
ROaDS  Roadkill Observation and Data System Tool 
SNOTEL Snow Telemetry Network 
TAC Technical Advisory Committee 
NPS National Park Service 
USDA United States Department of Agriculture 
USFS United States Forest Service 
USFWS United States Fish and Wildlife Service 
WTI Montana State University-Western Transportation Institute 
WVC Wildlife-Vehicle Collision 
US-191  U.S. Highway 191 
ii 
Table of Contents 
Report Summary .......................................................................................................................................... xi 
1. Introduction .......................................................................................................................................... 1 
1.1. Project Overview .......................................................................................................................... 2 
1.2. Wildlife-Vehicle Interactions and Ecological Connectivity ......................................................... 3 
1.3. Earlier Research and Regional Priorities ...................................................................................... 4 
1.4. Study Area Overview .................................................................................................................... 6 
1.4.1. Road Use and Area Visitation ............................................................................................... 6 
1.4.2. Study Area Context ............................................................................................................... 6 
2. Methodology ...................................................................................................................................... 10 
2.1. Study area .................................................................................................................................... 10 
2.2. Data sources ................................................................................................................................ 10 
2.3. Initial Screening of Road Segments ............................................................................................ 14 
2.3.1. Indices of Road Segment Importance for each Prioritization Characteristic ...................... 16 
2.3.2. Composite Indices of Road Segment Importance ............................................................... 18 
2.4. Data Gaps, Limitations, and General Assumptions .................................................................... 19 
3. Results ................................................................................................................................................ 21 
3.1. Wildlife Road Crossings Importance Index ................................................................................ 21 
3.2. Wildlife Observation Importance Index...................................................................................... 24 
3.3. Wildlife-Vehicle Collision Risk Importance Index .................................................................... 26 
3.4. Habitat Suitability/Connectivity (Regional Conservation Value) Importance Index ................. 28 
3.5. Priority Composite Index ............................................................................................................ 32 
3.6. Identifying Areas for Field Evaluation ....................................................................................... 38 
4. Wildlife Accommodation Measures ................................................................................................... 42 
4.1. Measures to Influence Driver Behavior ...................................................................................... 43 
4.1.1. Permanent or Temporary Warning Signs ............................................................................ 43 
4.1.2. Animal Detection Systems .................................................................................................. 43 
4.1.3. Reduced Speed Limit .......................................................................................................... 43 
4.1.4. Highway Lighting ............................................................................................................... 44 
4.2. Measures to Separate Wildlife from Traffic ............................................................................... 44 
4.2.1. Wildlife Crossing Structures ............................................................................................... 44 
4.2.2. Wildlife Underpasses .......................................................................................................... 44 
4.2.3. Wildlife Overpasses ............................................................................................................ 45 
4.2.4. Wildlife Fencing (in combination with Wildlife Crossing Structures) ............................... 45 
5. Field Evaluation and Ranking of Priority Sites .................................................................................. 46 
5.1. Field Evaluation Matrix Form..................................................................................................... 49 
iii 
6. Priority Sites and Recommendations ................................................................................................. 50 
6.1. Teepee Creek .............................................................................................................................. 50 
6.2. Specimen Creek to Bacon Rind Creek ........................................................................................ 54 
6.3. Taylor Fork ................................................................................................................................. 59 
6.4 Porcupine Creek ................................................................................................................................ 63 
6.4. North of Big Sky Entrance .......................................................................................................... 68 
6.5. Gallatin Gateway to Spanish Creek ............................................................................................ 71 
6.6. Four Corners to Gallatin Gateway .............................................................................................. 77 
6.7. West Fork Gallatin ...................................................................................................................... 80 
6.8. Upper Big Sky Connectivity Area .............................................................................................. 83 
6.9. Considerations for Additional Priority Sites ............................................................................... 84 
6.10. Discussion ............................................................................................................................... 86 
7. Renderings of Potential Wildlife Accommodation Options ............................................................... 87 
7.1. Retrofit of an Existing Structure ................................................................................................. 87 
7.2. Replacing and Upsizing an Existing Structure............................................................................ 89 
7.3. New Wildlife Overpass ............................................................................................................... 90 
8. Potential Funding Sources .................................................................................................................. 92 
8.1. Introduction ................................................................................................................................. 92 
8.2. Federal Infrastructure Funding Sources ...................................................................................... 92 
8.2.1. Discretionary Grant Programs ............................................................................................ 94 
8.2.2. Wildlife Crossings Pilot Program ....................................................................................... 94 
8.2.2.1. Rural Surface Transportation Grant Program ............................................................. 94 
8.2.2.2. Promoting Resilient Operations for Transformative, Efficient, and Cost-saving 
Transportation (PROTECT) ............................................................................................................ 95 
8.2.2.3. Bridge Investment Program ........................................................................................ 95 
8.2.2.4. National Culvert Removal, Replacement and Restoration Program ........................... 95 
8.2.2.5. Local and Regional Project Assistance Grants: Rebuilding American Infrastructure 
Sustainably and Equitably (RAISE) ............................................................................................... 95 
8.2.2.6. Nationally Significant Multimodal Freight and Highway Projects Program (INFRA)
96 
8.2.2.7. Nationally Significant Federal Lands and Tribal Projects .......................................... 96 
8.2.2.8. National Infrastructure Project Assistance (“Mega”) .................................................. 96 
8.2.3. Formula Funding Programs ................................................................................................ 96 
8.2.3.1. Bridge Formula Program ............................................................................................ 96 
8.2.3.2. Surface Transportation Block Grant Program ............................................................. 97 
8.2.3.3. Highway Safety Improvement Program ...................................................................... 97 
iv 
8.2.3.4. Federal Lands Transportation Program ....................................................................... 97 
8.2.3.5. Federal Lands Access Program ................................................................................... 98 
8.3. Federal Conservation Funding Sources ...................................................................................... 98 
8.3.1. Collaborative-based, Aquatic-focused, Landscape-scale Restoration Program .................. 98 
8.3.2. Forest Service Legacy Road and Trail Remediation Program ............................................ 98 
8.3.3. America the Beautiful Challenge ........................................................................................ 98 
8.3.4. Proposed Recovering America’s Wildlife Act .................................................................... 99 
8.3.5. U.S. Fish & Wildlife Service Funding for Imperiled Species ............................................. 99
8.3.6. Western Big Game Seasonal Habitat and Migration Corridors Fund ................................. 99 
8.4. Innovative State Funding Mechanisms ....................................................................................... 99 
8.4.1. Montana Wildlife and Transportation Partnership Project Program ................................... 99 
8.4.2. Other State Funding Models ............................................................................................. 100 
8.5. Innovative Local Funding Mechanisms .................................................................................... 100 
8.6. Public-Private Partnerships ....................................................................................................... 100 
8.7. Conclusion ................................................................................................................................ 101 
9. Aquatic Organism Passage Assessment ........................................................................................... 102 
9.1. Introduction ............................................................................................................................... 102 
9.1.1. Objectives ......................................................................................................................... 102 
9.1.2. Literature Review .............................................................................................................. 102 
9.1.2.1. Fish Passage and Approaches to Assess Culvert Barriers ......................................... 102 
9.1.2.1. Other Impacts of Roads on Aquatic Habitat ............................................................. 104 
9.2. Methods..................................................................................................................................... 104 
9.2.1. Data Collection ................................................................................................................. 105 
9.2.2. Initial Passage Assessment ................................................................................................ 105 
9.3. Results ....................................................................................................................................... 106 
9.3.1. Physical Data .................................................................................................................... 106 
9.3.2. Initial Passage Assessment ................................................................................................ 111 
9.4. Recommendations ..................................................................................................................... 115 
9.4.1. Mitigation for Aquatic Connectivity ................................................................................. 115 
9.4.2. Other Benefits to Large Structures (Resilience to Climate Change) ................................ 117 
9.4.3. Possible Next Steps ........................................................................................................... 118 
9.5. Sample Culvert Data Collection Sheet ...................................................................................... 120 
10. Spatio-temporal Hotspot Analysis ................................................................................................ 121 
10.1. Background ........................................................................................................................... 121 
10.2. Case Study Framework ......................................................................................................... 122 
10.3. Methods ................................................................................................................................. 122 
v 
10.4. Results ................................................................................................................................... 123 
10.4.1. Crash Data Analysis .......................................................................................................... 124 
10.4.1.1. 10-year Crash Data Analysis ..................................................................................... 125
10.4.1.2. 5-year Crash Data Analysis ....................................................................................... 125
10.4.1.3. 3-year Crash Data Analysis ....................................................................................... 126
10.4.2. Carcass Data Analysis ....................................................................................................... 127 
10.4.2.1. 10-year Carcass Data Analysis.................................................................................. 128
10.4.2.2. 5-year Carcass Data Analysis.................................................................................... 129
10.4.2.3. 3-year Carcass Data Analysis.................................................................................... 130
10.4.3. Results Summary .............................................................................................................. 131 
10.5. Discussion ............................................................................................................................. 131 
10.6. Conclusions ........................................................................................................................... 132 
11. Cost Benefit Analysis ................................................................................................................... 134 
11.1. Introduction ........................................................................................................................... 134 
11.2. Methods ................................................................................................................................. 134 
11.2.1. Period ................................................................................................................................ 134 
11.2.2. Carcass Removal Data ...................................................................................................... 134 
11.2.3. Costs of Wildlife-Vehicle Collisions ................................................................................ 135 
11.2.4. Costs of Mitigation Measures ........................................................................................... 136 
11.3. Results ................................................................................................................................... 136 
11.3.1. Cost-Benefit of Wildlife-Vehicle Collisions along US-191 and MT-64 .......................... 136 
11.4. Discussion ............................................................................................................................. 138 
12. Conclusion .................................................................................................................................... 139 
13. Literature Cited ............................................................................................................................. 141 
vi 
List Of Tables 
Table 1 Description of Data Sets Used in Analyses .................................................................................................... 12 
Table 2 Within-Category Weights for Each Prioritization Characteristic ................................................................... 15 
Table 3 Weights for Composite Importance Index ..................................................................................................... 16 
Table 4 Wildlife Road Crossing Importance Index: Top Ten 0.10-mile Segments .................................................... 21 
Table 5 Wildlife Observation Importance Index: Top Ten 0.10-mile segments ......................................................... 24 
Table 6 Wildlife Vehicle Collision Risk Importance Index: Top Ten 0.10-mile Segments ....................................... 28 
Table 7 Habitat Suitability/Connectivity (Regional Conservation Value) Importance Index: Top Ten 0.10-mile 
Segments..................................................................................................................................................................... 29 
Table 8 Summary Statistics on Normalized Values for Habitat Suitability/Connectivity (Regional Conservation Value) 
Model Inputs ............................................................................................................................................................... 32 
Table 9 Top 10% 0.10-mile Road Segments on US-191 for the Priority Composite Index ....................................... 34 
Table 10 Top 10% 0.10-mile Road Segments on MT-64 for the Priority Composite Index ...................................... 37 
Table 11 Effectiveness of Mitigation Measures .......................................................................................................... 42 
Table 12 Categories Used to Score Criteria 1-4 based on the Spatial Data Analysis. ................................................ 47 
Table 13 Summary of Scores and Ranking of each Priority Site following Field Evaluation. ................................... 48 
Table 14 Maximum, Minimum, and Average Index Values for each Prioritization Characteristic and Composite Index 
of all 0.10-mile Road Segments within the Teepee Creek Priority Site ..................................................................... 50 
Table 15 Teepee Creek Field Evaluation Scores and Priority Ranking ...................................................................... 50 
Table 16 Index Values of all 0.10-mile Segments within the Teepee Creek Priority Site .......................................... 52 
Table 17 Maximum, Minimum, and Average Index Values for each Prioritization Characteristic and Composite Index 
of all 0.10-mile Road Segments within the Specimen Creek to Bacon Rind Creek Priority Site ............................... 54 
Table 18 Specimen Creek to Bacon Rind Creek Field Evaluation Scores and Priority Ranking ................................ 54 
Table 19 Index Values of all 0.10-mile Segments within the Specimen Creek to Bacon Rind Creek Priority Site ... 56 
Table 20 Maximum, Minimum, and Average Index Values for each Prioritization Characteristic and Composite Index 
of all 0.10-mile Road Segments within the Taylor Fork Priority Site ........................................................................ 59 
Table 21 Taylor Fork Field Evaluation Scores and Priority Ranking ......................................................................... 59 
Table 22 Index Values of all 0.10-mile segments within the Taylor Fork Priority Site .............................................. 60 
Table 23 Maximum, Minimum, and Average Index Values for each Prioritization Characteristic and Composite Index 
of all 0.10-mile Road Segments within the Porcupine Creek Priority Site ................................................................. 63 
Table 24 Porcupine Creek Field Evaluation Scores and Priority Ranking.................................................................. 63 
Table 25 Index Values of all 0.10-mile Segments within the Porcupine Creek Priority Site ..................................... 65 
Table 26 Maximum, Minimum, and Average Index Values for each Prioritization Characteristic and Composite Index 
of all 0.10-mile Road Segments within the North of Big Sky Entrance Priority Site ................................................. 68 
Table 27 North of Big Sky Entrance Field Evaluation Scores and Priority Ranking ................................................. 68 
Table 28 Index Values of all 0.10-mile Segments within the North of Big Sky Entrance Priority Site ..................... 69 
Table 29 Maximum, Minimum, and Average Index Values for each Prioritization Characteristic and Composite Index 
of all 0.10-mile Road Segments within the Gallatin Gateway to Spanish Creek Priority Site ................................... 71 
Table 30 Gallatin Gateway to Spanish Creek Field Evaluation Scores and Priority Ranking .................................... 71 
vii 
Table 31 Index Values of all 0.10-mile Segments within the Gallatin Gateway to Spanish Creek Priority Site ........ 73 
Table 32 Maximum, Minimum, and Average Index Values for each Prioritization Characteristic and Composite Index 
of all 0.10-mile Road Segments within the Four Corners to Gallatin Gateway Priority Site ..................................... 77 
Table 33 Four Corners to Gallatin Gateway Field Evaluation Scores and Priority Ranking ...................................... 77 
Table 34 Maximum, Minimum, and Average Index Values for each Prioritization Characteristic and Composite Index 
of all 0.10-mile Road Segments within the West Fork Gallatin Priority Site ............................................................. 80 
Table 35 West Fork Gallatin Field Evaluation Scores and Priority Ranking .............................................................. 80 
Table 36 Maximum, Minimum, and Average Index Values for each Prioritization Characteristic and Composite Index 
of all 0.10-mile Road Segments within the Upper Big Sky Connectivity Area Priority Site ..................................... 83 
Table 37 Upper Big Sky Connectivity Area Field Evaluation Scores and Priority Ranking ...................................... 83 
Table 38 Eligible recipients for transportation funding programs under the Infrastructure Investment and Jobs Act of 
2021 ............................................................................................................................................................................ 93 
Table 39 Aquatic Passage Assessment: The table summarizes the criteria, thresholds, and passage assessment 
outcome. ................................................................................................................................................................... 106 
Table 40 Aquatic Passage Assessment: Summary of Key Physical Characteristics at Each Crossing ..................... 106 
Table 41 Aquatic Passage Assessment: Summary of Initial Passage Assessment at Each Crossing ........................ 111 
Table 42 Comparison of the top 10 identified WVC hotspots using 10 years of crash data ..................................... 125 
Table 43 Comparison of the Top 10 Identified WVC Hotspots using 5 years of Crash Data ................................... 126 
Table 44 Comparison of the Top 10 Identified WVC Hotspots using 3 years of Crash Data ................................... 127 
Table 45 Comparison of the Top 10 Identified WVC Hotspots using 10 years of Carcass Data .............................. 129 
Table 46 Comparison of the Top 10 Identified WVC Hotspots using 5 years of Carcass Data ................................ 130 
Table 47 Comparison of the top 10 identified WVC hotspots using 3 years of carcass data .................................... 131 
Table 48 Summary of Results from the KDE+ Hotspot Analyses ............................................................................ 131 
Table 49 Species, Number of Carcasses Recorded, and Percentage of Each Species of the Total Number of Recorded 
Carcasses .................................................................................................................................................................. 134 
Table 50 Total Costs Associated with Wildlife-Vehicle Collisions (in 2020 US$) .................................................. 135 
Table 51 Species Reported in Carcass Data as Categorized for Cost-benefit Analyses ........................................... 135 
 
 
  
 
viii 
 
List of Figures 
Figure 1 Elk on roadway in path of oncoming car. Photo by Holly Pippel.................................................................... i 
Figure 2 Archway about a mile north of Spanish Creek (circa 1927-30) and the same location along US-191 in 2016. 
Credits: Historic photo-National Park Service; Recent photo-Duncan Patten .............................................................. 1 
Figure 3 Roadway into West Yellowstone (circa 1910-1914) and US-191 (2016) in the same location (Road Mile Post 
6). Credits: Historic photo-Yellowstone Heritage Center; Recent photo-Duncan Patten ............................................. 3 
Figure 4 Aftermath of a wildlife-vehicle collision on US-191 just outside the mouth of the Gallatin Canyon. Credit: 
Holly Pippel .................................................................................................................................................................. 4 
Figure 5 Reported Crashes and Carcasses Collected in Montana, 2008-2020, and Average Percent of Wildlife-Vehicle 
Collisions by Species. Compiled by Matthew Bell, MSU-WTI. .................................................................................. 5 
Figure 6 The Gallatin Way to Yellowstone (Road Mile Post 61.1) around 1900 and the same stretch of US-191 through 
Gallatin Canyon in 2016. Credits: Historic photo-Gallatin History Museum; Recent photo-Duncan Patten ............... 6 
Figure 7 Study Area Overview Map ............................................................................................................................. 9 
Figure 8 The Relationship between Average Daily Traffic, WVCs, and the Barrier Effect of a Road (from Seiler 2003).
 .................................................................................................................................................................................... 20 
Figure 9 Intensity of Wildlife Crossing Roads in the Study Area ............................................................................... 22 
Figure 10 Grizzly Bear Habitat Suitability, Crossing Frequency and Roadkill .......................................................... 23 
Figure 11 Intensity of Wildlife Use of Roadside Environments in the Study Area .................................................... 25 
Figure 12 Frequency of Collisions with Wildlife in the Study Area based on Crash and Carcass Data ..................... 27 
Figure 13 Regional Conservation Value ..................................................................................................................... 30 
Figure 14 Species-neutral Connectivity Showing General Permeability of the Landscape to Ecological Flows and 
Riparian Corridors Important for Climate Resilience ................................................................................................. 31 
Figure 15 Priority Composite Indices for US-191 and MT-64 ................................................................................... 33 
Figure 16 A view of early road configuration within the Gallatin Gateway to Spanish Creek Priority Site looking south 
to Gallatin Canyon at the beginning of the 1900s, and in 2016. Credits: Historic photo-Museum of the Rockies; Recent 
photo- Duncan Patten. ................................................................................................................................................ 38 
Figure 17 Priority Road Areas for Field Evaluations .................................................................................................. 40 
Figure 18 Double Culverts Facilitating Water Flow from Teepee Creek ................................................................... 51 
Figure 19 Teepee Creek Priority Site Map ................................................................................................................. 53 
Figure 20 Bridge over the Gallatin River (between RM 23-24) ................................................................................. 56 
Figure 21 Specimen Creek to Bacon Rind Creek Priority Site Map ........................................................................... 58 
Figure 22 Bridge over Taylor Fork (RM 33.9) ........................................................................................................... 60 
Figure 23 Taylor Fork Priority Site Map .................................................................................................................... 62 
Figure 24 View north toward the Porcupine Creek Priority Site, just south of Big Sky in 1922 (left) and 2016 (right). 
Credit: Historic photo-Gamel Family; Recent photo-Duncan Patten. ........................................................................ 65 
Figure 25 Porcupine Creek Priority Site Map ............................................................................................................. 67 
Figure 26 North of Big Sky Priority Site Map ............................................................................................................ 70 
Figure 27 Gallatin River Bridge (RM 70.5) ................................................................................................................ 72 
Figure 28 Spanish Creek Bridge (RM 68.1) ............................................................................................................... 73 
ix 
Figure 29 Gallatin Gateway to Spanish Creek Priority Site Map ............................................................................... 76 
Figure 30 Four Corners to Gallatin Gateway Priority Site Map ................................................................................. 79 
Figure 31 Bridge over the West Fork of the Gallatin River along MT-64 (Lone Mountain Trail) RM 0.2 ................ 81 
Figure 32 West Fork Gallatin Priority Site Map ......................................................................................................... 82 
Figure 33 Map of Carcass Data along US-191 in Areas within and adjacent to Yellowstone National Park ............. 85 
Figure 34 (Top image) Existing Gallatin Gateway Bridge; (bottom image) Conceptual Bridge Retrofit (for illustrative 
purposes only). ........................................................................................................................................................... 88 
Figure 35 (Top image) Existing Spanish Creek Bridge; (bottom image) Conceptual Bridge Replacement (for 
illustrative purposes only)........................................................................................................................................... 90 
Figure 36 (Top image) US-191 north of Big Sky; (bottom image) Conceptual Wildlife Overpass (for illustrative 
purposes only) ............................................................................................................................................................ 91 
Figure 37 Funding Flow for Federal Reauthorization Funds (Montana Department of Transportation, 2022-2026 
Statewide Transportation Improvement Program) ...................................................................................................... 94 
Figure 38 Aquatic Passage Assessment: Pie Chart Showing the Number of Culverts by Type ............................... 109 
Figure 39 Aquatic Passage Assessment: Cumulative Frequency and Histogram of Culvert Slope .......................... 110 
Figure 40 Aquatic Passage Assessment: Cumulative Frequency and Histogram of Outlet Drop ............................. 110 
Figure 41 Aquatic Passage Assessment of Culverts Along US-191 and MT-64 ...................................................... 114 
Figure 42 Aquatic Passage Assessment: Images of a “Green” Crossing (left), a “Gray” Crossing (middle), and a “Red” 
Crossing (right) ......................................................................................................................................................... 115 
Figure 43 Aquatic Passage Assessment: illustration of a bridge spanning a mountain stream ................................. 116 
Figure 44 Aquatic Passage Assessment: Hydrographs for the Gardner River and Yellowstone River at Corwin Springs 
showing the June 13/June 14, 2022, historic flood event ......................................................................................... 118 
Figure 45 Aquatic Passage Assessment: Picture shows a culvert designed as a fish barrier .................................... 119 
Figure 46 Distribution of WVC data with Large Mammals within the US-191 Study Area .................................... 124 
Figure 47 Proportion of WVC Data per kilometer of road segments equal to using 13 years of Crash Data ........... 124 
Figure 48 Comparison of WVC Hotspots per kilometer of road segments using 10 years of Crash Data................ 125 
Figure 49 Comparison of WVC Hotspots per kilometer of road segments using 5 years of Crash Data.................. 126 
Figure 50 Comparison of WVC Hotspots per kilometer of road segments using 3 years of Crash Data.................. 127 
Figure 51 Proportion of WVC Data per kilometer of road segments equal to using 13 years of Carcass Data ........ 128 
Figure 52 Comparison of WVC Hotspots per kilometer of road segments using 10 years of Carcass Data ............ 128 
Figure 53 Comparison of WVC Hotspots per kilometer of Road Segments using 5 years of Carcass Data ............ 129 
Figure 54 Comparison of WVC Hotspots per kilometer of Road Segments using 3 years of Carcass Data ............ 130 
Figure 55 US-191 from West Yellowstone (left side of graph) to Four Corners (right side of graph). .................... 137 
Figure 56 MT-64 from the junction with US-191 (left side of graph) to the end of the road up the mountain (right side 
of graph). .................................................................................................................................................................. 137 
 
 
  
 
x 
 
Report Summary 
The US Highway 191 (US-191)/Montana Highway 64 (MT-64) Wildlife & Transportation Assessment (the 
“Assessment”) improves understanding of the issues affecting driver safety, wildlife mortality, and wildlife 
movement along the major routes that connect Yellowstone National Park, the Custer Gallatin National 
Forest, and other public lands to the growing population centers of Bozeman, Big Sky, and nearby 
communities in Southwest Montana. By engaging personnel from multiple federal, state, and local agencies 
along with key stakeholders to examine problems and possibilities through the lens of spatial ecology, the 
US-191/MT-64 Wildlife & Transportation Assessment brings new insight into the impact of two major 
roads that unite local communities yet divide the landscape and natural habitats. 
The information included in this report should inform and support area communities and agency decision-
makers to select and pursue wildlife accommodation options. With the passage of the Infrastructure 
Investment and Jobs Act of 2021, significant funds for wildlife accommodation measures are available 
nationwide on a competitive basis. The US-191/MT-64 Wildlife & Transportation Assessment better 
equips part of Southwest Montana’s gateway to Yellowstone National Park to take advantage of new 
funding opportunities.  
Methodology 
As detailed in Chapter 2, the Assessment compiles, overlays, and evaluates wildlife-vehicle collision and 
wildlife carcass data, wildlife movement and habitat data, and live wildlife observations from aerial 
surveys. It also incorporates wildlife sightings and roadkill information gathered via citizen science. The 
Assessment further draws upon local and expert knowledge gathered through in-person outreach and an 
interactive map. Road areas identified through an in-depth spatial data analysis were evaluated in a field 
review conducted by an interdisciplinary Technical Advisory Committee of county, state, and federal 
planners, along with biologists, engineers, transportation experts, and the Research Team from the Center 
for Large Landscape Conservation (CLLC) and Montana State University’s Western Transportation 
Institute (WTI). 
Spatial Analysis 
The results of the spatial analysis, developed from 26 data sets, are shown in a series of maps in Chapter 3. 
These maps illustrate key characteristics of the US-191 and MT-64 road corridors along each 0.1-mile 
segment within the study area. The characteristics illustrated are Wildlife-Vehicle Collision Risk, Wildlife 
Road Crossings, Wildlife Observations Near Roads (intensity of wildlife use of roadside environments), 
and Regional Conservation Value (see Figures 9-13). The latter of these is based on models of habitat 
connectivity/suitability to ensure that species from bighorn sheep to wolverine to sage grouse, which are 
either not prominent or not accounted for in other data sets, are included. The data sets that make up the 
key characteristics are shown in Table 2.  
In order to identify an initial set of road areas for field evaluation, a Composite Index was developed from 
a weighted combination of the key characteristics (see Section 3.5). The Composite Index value of each 
0.1-mile road segment is shown on the map in Figure 15. Additional details of initial road area selection—
including a smoothing process used to find not only important 0.1-mile segments but also broader areas 
with consistently elevated values—are described in Section 3.6. The results are seven Priority Road Areas 
shown on the map in Figure 17.  
The CLLC-WTI research team also identified three more road areas to include in the field review based on 
the spatial analysis. Two of these areas were included due to their high rates of wildlife-vehicle collisions 
and concerns for human safety, and one due to its high connectivity value in an area where development 
pressure is increasing. While the work to determine these initial areas is robust, Section 2.5 of the report 
xi 
describes data gaps, limitations, and general assumptions that require further research to better assess their 
impact.  
Evaluation of Priority Road Areas and Recommendations 
Descriptions of eleven Priority Sites that are located in areas important areas for wildlife movement and/or 
pose elevated risks to human and wildlife safety, as well as recommendations for sites reviewed in the field, 
are described in Chapter 6. During the field review, the CLLC-WTI research team and Technical Advisory 
Committee evaluated and ranked sites via a consensus process using a site evaluation matrix that includes 
the results of the data analysis along with other factors to address wildlife-vehicle collision risk, 
conservation value, mitigation options, barrier effect of the road corridor, vulnerability to future change, 
and land security (see Chapter 5). The data analysis and field review process led to the definition of nine 
priority sites. Following the spatial analysis and field review, two more priority sites were identified by the 
CLLC-WTI research team upon the availability of additional data and greater insight into potential 
opportunities for wildlife accommodations. These sites are supported by recent and historical data. The 
recommendations also highlight that mitigation along other areas of the study roads may also be warranted 
and should be considered based on the documentation and analysis of the report, especially when highway 
projects are planned. 
Culvert Assessment 
In addition to consideration of terrestrial species, Chapter 9 provides information on an evaluation of the 
impacts of US-191 and MT-64 on aquatic species and aquatic connectivity. The information summarizes 
the potential “barrier effect” of culverts passing under US-191 and MT-64 based on an assessment of 
factors in the field. Fifty-three culverts were coded based on whether substrate is present and continuous 
throughout the structure, outlet drop, culvert slope, and presence/absence of internal baffling structures 
designed for fish passage. The details of each criterion are in Table 39. Of the total structures, 30 ranked 
as green, which implies conditions are adequate for upstream passage of fish. Six of the culverts ranked as 
gray, meaning conditions may or may not be adequate for upstream passage of fish. Lastly, 17 culverts 
ranked as red, indicating conditions may not be adequate for fish passage, and further evaluation should be 
considered. It is important to note that there are some instances where fish passage is not desirable, such as 
where there is a need to protect native fish populations from non-native or invasive species. Local biologists 
should always be consulted prior to initiating any project to remove a potential barrier.  
Recommendations and Cost-benefit Analysis 
The recommendations in Chapter 6 describe appropriate locations for prospective wildlife accommodation 
measures such as culverts, bridges, underpasses and overpasses, and/or animal detection systems—each in 
combination with fencing—and take into account both terrestrial and/or aquatic wildlife passage. Many 
sites include major drainages from surrounding public lands that intersect with US-191 or MT-64 and 
feature existing infrastructure that has the potential to facilitate animal movements, such as a bridge 
spanning a riparian corridor. During the field evaluation, the CLLC-WTI research team and Technical 
Advisory Committee considered means to incorporate existing infrastructure, new structures, and 
additional alternatives to reduce collisions with wildlife (e.g., variable message signs for areas that have 
spatially discreet or seasonal conflicts and traffic calming measures to effectively reduce the design speed 
of the highway). Chapter 4 describes types of wildlife accommodation measures and their effectiveness 
(see Table 11). A range of accommodation strategies have a role to play in helping to reduce collisions and 
maintain wildlife movement in the study area. 
In the case of terrestrial species, despite crossing structures with fencing requiring high initial investment, 
research shows they are cost-effective over the course of their lifetime (generally 75 years or more) due to 
greater efficacy in reducing wildlife-vehicle collisions and lower maintenance costs than other options 
xii 
(Brennan, Chow, and Lamb 2022; Huijser et al. 2009). Further, given that bridges and culverts that are 
upscaled and designed to allow for wildlife passage are usually better able to accommodate stream and 
floodplain function due to larger size and capacity, they may also make infrastructure more resilient to 
“extreme” weather events like flooding.  
Chapter 11 addresses the cost-benefit of a suite of wildlife accommodation measures. Based on available 
carcass data, three locations meet the threshold for structural wildlife accommodation measures; however, 
the chapter also details the well-documented under-reporting of carcass data and the possibility of a 
correction factor to allow for a more accurate cost-benefit examination of potential measures in the study 
area.  
The Montana Department of Transportation has already identified several bridges in need of replacement 
in the study area. Applying the findings of this Assessment to bridges or other priority locations when 
replacement is scheduled offers a best-case scenario for cost-effective implementation.  
Funding 
Chapter 8 overviews federal funding sources that could be used to advance the wildlife accommodation 
measures discussed in this Assessment, along with innovative state and local funding mechanisms. The 
vast majority of Montana’s roads and bridges are funded through federal dollars, and the Montana 
Department of Transportation receives no funding from the state’s general revenue fund (Montana 
Department of Transportation 2022). One of the criteria against which the Federal Highway Administration 
may evaluate project proposals under various programs of the Infrastructure Investment and Jobs Act is the 
extent to which a proposed project leverages other funding sources, including from public-private 
partnerships. Strong partnerships include diverse stakeholders—such as wildlife and transportation 
agencies, counties, academic researchers, non-governmental organizations, and local landowners and 
business owners—working together to advance the common cause of making roads safer for drivers and 
wildlife. 
Spatio-temporal Hotspot Analysis 
Chapter 10 of this report is a spatiotemporal hotspot analysis that uses US-191 as a case study that considers 
how wildlife-vehicle collision hotspots may shift based on varying timescales in order to test the validity 
of using various lengths of data sets (3, 5, and 10 years) to identify locations for prospective wildlife 
mitigation measures. The results of the analysis are not intended as guidance in the same manner as in other 
sections of this Assessment. Rather, they provide additional information intended to further substantiate 
best practices for the preparation of wildlife and transportation assessments.  
Next Steps 
The options for prospective wildlife accommodation measures along key road segments described in this 
report are intended as a guide to inform decision-making processes rather than serve as a prescription for 
specific actions. Implementation of any prospective measure depends upon multiple factors such as public 
support, design, and engineering feasibility, potential agreements with land management agencies and/or 
private landowners, as well as funding availability.  
Wildlife accommodations are carried out as part of planning under Montana’s Five-year Statewide 
Transportation Improvement Plan and a new program of the Montana Wildlife and Transportation 
xiii 
Partnership1 (described in Section 8.4.1). Montana Department of Transportation is the key agency partner 
responsible for actions to change or improve US-191 and MT-64. 
The CLLC-WTI Research Team suggests that making US-191 and MT-64 safer for travelers and wildlife 
is a multi-year, multi-site, multi-stakeholder proposition that will take collective action to bring about. The 
Assessment provides a foundation to allow for discussion about how to reach these goals based on robust 
understanding.
1 The Montana Wildlife and Transportation Partnership consists of representatives of Montana Department of Transportation, 
Montana Fish, Wildlife and Parks, and Montanans for Safe Wildlife Passage. https://www.mdt.mt.gov/pubinvolve/mwt/ 
xiv 
1. Introduction
Figure 2 Archway about a mile north of Spanish Creek (circa 1927-30) and the same location along US-191 in 
2016. Credits: Historic photo-National Park Service; Recent photo-Duncan Patten 
The goal of the US Highway 191 (US-191)/Montana Highway 64 (MT-64) Wildlife & Transportation 
Assessment (Assessment) is to improve understanding of the issues affecting driver safety, wildlife 
mortality, and wildlife movement along the major routes that connect Yellowstone National Park, the 
Custer Gallatin National Forest, and other public lands to the growing population centers of Bozeman, Big 
Sky, and nearby communities in Southwest Montana. A joint project of the Center for Large Landscape 
Conservation (CLLC) and Montana State University's Western Transportation Institute (WTI), this 
Assessment provides robust information on opportunities to improve highway safety and ecological 
connectivity2 for terrestrial and aquatic species. 
Prepared using a landscape approach, the Assessment brings together disciplines and data from ecology to 
engineering, as well as local knowledge, to devise integrated solutions that convey benefits to human safety, 
ecological connectivity, and infrastructure resilience. 
The options for prospective wildlife accommodation measures along key road segments described in this 
report are intended as a guide to inform decision-making processes rather than serve as a prescription for 
specific actions. Implementation of any prospective measure depends upon multiple factors, such as public 
support, design and engineering feasibility, and potential agreements with land management agencies 
and/or private landowners, along with funding availability. Wildlife accommodations are carried out as 
part of planning under Montana’s Five-year Statewide Transportation Improvement Plan (Montana 
Department of Transportation 2022) and a new program of the Montana Wildlife and Transportation 
Partnership3 (described in Section 8.4.1). 
Montana Department of Transportation is the key agency partner in charge of US-191 and MT-64. The 
CLLC-WTI Research Team (see report authors) suggests that making US-191 and MT-64 safer for 
travelers and wildlife is a multi-year, multi-site, multi-stakeholder proposition that will take collective 
action to bring about. 
2 Ecological Connectivity is defined as the unimpeded movement of species and the flow of natural processes that sustain life on 
Earth.  Wildlife need to be able to move to find food, water, and mates and carry out daily and seasonal life needs.  
3 The Montana Wildlife and Transportation Partnership consists of representatives of Montana Department of Transportation, 
Montana Fish, Wildlife and Parks, and Montanans for Safe Wildlife Passage. https://www.mdt.mt.gov/pubinvolve/mwt/ 
1 
1.1. Project Overview 
The US-191/MT-64 Wildlife & Transportation Assessment combines local and expert knowledge, public 
data, citizen science, and engineering expertise to identify feasible sites for a range of potential wildlife 
accommodation options to improve the safety of travelers and wildlife. By engaging personnel from 
multiple federal, state, and local agencies along with key stakeholders to examine problems and 
possibilities through the lens of spatial ecology, the US-191/MT-64 Wildlife & Transportation Assessment 
brings new insight into the impact of two major roads that unite local communities yet divide the landscape 
and natural habitats. 
The Assessment is intended to: 
a) Lay the groundwork for the implementation of best management practices to conserve wildlife,
reduce the barrier effect4 of highways, and improve driver safety in the face of unprecedented
regional traffic growth.
b) Provide residents and officials of communities along US-191 and MT-64 with essential tools to
guide decision-making.
c) Enable public agencies to prioritize win-win designs in future road redevelopment projects.
d) Identify key road segments that might benefit from wildlife accommodation measures.
The information in this report should inform and support area communities and agency decision-makers to 
select and pursue wildlife accommodation options. With the passage of the Infrastructure Investment and 
Jobs Act of 2021, significant funds for wildlife accommodation measures are available nationwide on a 
competitive basis. The information in this Assessment better equips part of Southwest Montana’s gateway 
to Yellowstone National Park to take advantage of new funding opportunities. 
As described in detail in the Methodology (see Chapter 2), the Assessment compiles, overlays, and 
evaluates information from multiple public agencies: wildlife-vehicle collision and wildlife carcass data, 
wildlife movement and habitat data, and live wildlife observations from aerial surveys. It also incorporates 
information on roadkill and wildlife sightings collected by citizen scientists using the ROaDS (Roadkill 
Observation and Data System) Tool, a smartphone application that allows the collection of spatially precise 
information. The Assessment further draws upon local and expert knowledge gathered through in-person 
outreach and an interactive map. The priority road segments identified through spatial data analysis were 
evaluated in a field review conducted by an interdisciplinary Technical Advisory Committee of county, 
state, and federal planners, biologists, engineers, transportation experts, and the joint CLLC-WTI Research 
Team. 
In addition to consideration of terrestrial species, the Assessment includes a chapter on the evaluation of 
the impacts of US-191 and MT-64 on aquatic species and aquatic connectivity (see Chapter 9). This chapter 
summarizes the potential “barrier effect”5 of all culverts passing under US-191 and MT-64 based on a field 
survey carried out in the fall of 2021. 
The Assessment identifies appropriate locations for prospective wildlife accommodation measures such as 
culverts, bridges, underpasses and overpasses, and/or animal detection systems—each in combination with 
fencing—and takes into account both terrestrial and/or aquatic wildlife passage. In the case of terrestrial 
4 The barrier effect of a highway may be due to road width, associated habitat change, or traffic volume, among other factors. 
While high speed roads with large traffic volumes may be the most disruptive to animal movements and population interchange, 
secondary highways and unpaved roads can also impede animal movements (Clevenger and Huijser 2011).
5 Culverts beneath roads can impede aquatic organism passage. The impediments they may pose are described in Chapter 9. 
2 
species, while crossing structures with fencing require high initial investment, research shows they are cost-
effective over the course of their lifetime (generally 75 years or more) due to greater efficacy in reducing 
wildlife-vehicle collisions and lower maintenance costs than other options, like animal detection systems 
(Brennan, Chow, and Lamb 2022; Huijser et al. 2009). Further, given that bridges and culverts that are 
upscaled and designed to allow for wildlife passage are usually better able to accommodate stream and 
floodplain function due to larger size and capacity, they may also make infrastructure more resilient to 
“extreme” weather events like flooding. These types of events are occurring with increased frequency and 
severity (Hoeppe 2016). MDT has already identified several bridges in need of replacement in the study 
area. Applying the findings of this Assessment to bridges or other priority locations when replacement is 
scheduled offers a best-case scenario for cost-effective implementation. As described in the 
recommendations (see Chapter 6), a range of accommodation strategies have a role to play in helping to 
reduce collisions and maintain wildlife movement in the study area. Further, Chapter 4 provides an 
overview of the effectiveness of various wildlife accommodation measures. 
Figure 3 Roadway into West Yellowstone (circa 1910-1914) and US-191 (2016) in the same location (Road Mile 
Post 6). Credits: Historic photo-Yellowstone Heritage Center; Recent photo-Duncan Patten 
1.2. Wildlife-Vehicle Interactions and Ecological Connectivity 
Beyond severing intact landscapes and serving as one of the greatest threats to habitat connectivity, 
increasing traffic volumes through areas of regular wildlife movement poses a growing safety risk to 
motorists and wildlife alike. In the U.S., 200 people, 1-2 million large animals, and countless smaller 
animals die every year in vehicle crashes with wildlife—which also lead to 26,000 human injuries—at an 
annual cost of more than $8 billion (Conover et al. 1995; Huijser et al. 2009). In addition to these impacts, 
roads can act as barriers to wildlife movement, reducing the ability of wildlife to find adequate food, water, 
or mates or to complete their seasonal migrations. For example, one grizzly bear attempted to cross I-90 in 
Montana near Drummond 46 times—over 29 days in Fall 2020 and 25 days in Spring 2021—before finally 
succeeding (Pashby 2022). Another study estimated that US Highway 2 near Glacier National Park is a 
barrier to grizzly bears when traffic exceeds 100 vehicles per hour (Waller and Miller 2015). 
In this Assessment, we analyzed data sources that relate both to direct human and wildlife safety (crashes 
and carcass records) and ecological connectivity (wildlife movement, wildlife observation, and habitat 
suitability) to identify priority locations for potential wildlife accommodation measures. 
3 
Figure 4 Aftermath of a wildlife-vehicle collision on US-191 just outside the mouth of the Gallatin Canyon. 
Credit: Holly Pippel 
1.3. Earlier Research and Regional Priorities 
A 2020 MDT Corridor Study of US-191 (Four Corners to Beaver Creek) found 24% of all crashes in the 
study area are due to collisions with wildlife. Across Montana, 10% of all crashes reported to law 
enforcement involve wildlife (Mat Bell, MSU-WTI, pers. comm; see Figure 5), while nationally, animal-
caused vehicle crashes comprise less than 5% of total crashes (National Center for Statistics and Analysis 
2022). State Farm Insurance consistently rates Montana as second in the nation for risk of collision with 
wildlife based on claims reported (State Farm 2023). Crashes involving wildlife increased by over 50% 
between 2008 and 2020, according to an analysis of Montana Department of Transportation data, and over 
90% of these crashes were with whitetail and mule deer (Mat Bell, MSU-WTI, pers. comm; see Figure 5). 
The US-191 Corridor Study states: “MDT will consider the potential for targeted wildlife study and 
standalone wildlife accommodation projects within the corridor based on [Montana Wildlife and 
Transportation Partnership] efforts or through partnerships with other interested stakeholders resulting in 
the identification of data collection gaps, research needs, and funding opportunities.” This Wildlife & 
Transportation Assessment helps to fulfill the need for additional research to determine potential wildlife 
accommodation measures (Montana Department of Transportation 2020). 
4 
In addition to the existing MDT Corridor Study, the US-191/MT-64 Wildlife & Transportation Assessment 
aligns with a statewide “areas of greatest need” assessment (see MWTP Planning Tool) developed by the 
Montana Wildlife and Transportation Partnership and with the Gallatin County Growth Policy, which calls 
for identification of suitable wildlife crossing areas. It also helps to realize the goals outlined in Our Big 
Sky: Community Strategy and Vision, devised as a strategic plan for the Big Sky community. 
 
Figure 5 Reported Crashes and Carcasses Collected in Montana, 2008-2020, and Average Percent of Wildlife-
Vehicle Collisions by Species. Compiled by Matthew Bell, MSU-WTI. 
 
The Assessment follows upon earlier studies, including a review of high-risk zones for collisions between 
vehicles and large ungulates during Montana’s fall migration season (Creech, McClure, and Callahan 2016) 
and a preliminary assessment of wildlife and transportation issues in the Greater Yellowstone Ecosystem 
(Hardy, Willer, and Williamson 2008). Both efforts describe portions of the present study area as high 
priority. The former found a 10-mile segment of US-191 south of Four Corners to have the second highest 
mean frequency of recorded ungulate carcasses (whitetail deer, mule deer, elk, pronghorn, moose, and/or 
bighorn sheep) per mile in Montana each fall, from 2010-2015. The latter found four areas of US-191 
between Bozeman and West Yellowstone with relatively high numbers of recorded wildlife-vehicle 
collisions based on crash and carcass data from 1998-2002. It also describes a portion of the study area as 
part of the Gallatin River megasite, which it identifies as 4th priority in Greater Yellowstone in part due to 
US-191 bisecting the extensive elk migration corridor that links Yellowstone National Park to the Madison 
Range. One further study is a 2012 assessment of bison-vehicle collisions along 10 miles of US-191 from 
West Yellowstone to the border of Yellowstone National Park. Based on data collected from 1999-2009, 
this study and its findings are described in relation to more recent information in Section 6.10 (Dupree and 
Dimambro 2012). 
 
 
5 
 
1.4. Study Area Overview 
 
Figure 6 The Gallatin Way to Yellowstone (Road Mile Post 61.1) around 1900 and the same stretch of US-191 
through Gallatin Canyon in 2016. Credits: Historic photo-Gallatin History Museum; Recent photo-Duncan 
Patten 
 
1.4.1. Road Use and Area Visitation 
US Highway 191 (US-191) facilitates vehicle travel from Grand Teton National Park to Yellowstone 
National Park and through the expansive national forest lands that join them to the expanding population 
centers of Big Sky and Bozeman, Montana. Visitation to Yellowstone National Park increased by 20% 
from 2014-2017, and over 1 million trips on US-191 occur each year to enter Yellowstone (National Park 
Service 2022). West Yellowstone, Montana, is the busiest of Yellowstone’s gateways, hosting over 4 
million visitors per year (National Park Service 2022). As traffic increases along this route, the movement 
of bighorn sheep, bison, elk, moose, mule, and whitetail deer herds, as well as black and grizzly bears, is 
progressively threatened. 
The study area (see Figure 7), which extends 82 miles along US-191 from Four Corners to West 
Yellowstone and includes the 10-mile extent of MT-64, is busy due to additional attributes associated with 
the region’s growth: 2,500 construction and service workers commute daily along the highway from 
Bozeman and nearby communities to Big Sky as part of the area’s ongoing development boom (Future 
West 2019). Big Sky Ski Resort has grown to over 500,000 skier visits per year (Max 2017). The area is 
part of the fastest-growing region in the United States. Gallatin County’s population increased by nearly a 
third from 2010-2020, with Bozeman ranking among the fastest-growing areas in the nation each year since 
2018. Maintaining the connectivity of critical wildlife routes outside of Yellowstone National Park along 
US-191 and those that cross the sole public access to Big Sky, named Lone Mountain Trail (MT-64), 
requires intentional action (Dietrich 2022; Policom 2023; US Census Bureau 2023). 
1.4.2. Study Area Context 
High traffic volumes along US-191 in its current form are a relatively recent phenomenon. In The Gallatin 
Way to Yellowstone, Bozeman author Duncan Patten describes trailing 120 horses 35 miles along US-191 
from Spanish Creek at the mouth of the Gallatin Canyon to Elkhorn Ranch near Sage Creek during several 
springs in the 1950s without encountering issues with traffic. The book also documents changes in the 
highway from its initial construction in 1911 to the first paving in the 1930s to redesign with alterations to 
the Gallatin River and reconstruction into a modern highway in the 1950s and 1960s. 
 
 
6 
 
In Southwest Montana, three valleys—the Madison, Gallatin, and Paradise—are crossed by highways that 
lead through the surrounding landscape to Yellowstone’s North and West entrances. In this region, a mix 
of public and private land serves as the “Gateway to Yellowstone,” where busier roads and growing 
subdivisions have the potential to fragment the Greater Yellowstone Ecosystem6 that is home to the largest 
concentration of wildlife species in the lower 48 states (National Park Service 2023). In the study area, US-
191 leads from Yellowstone National Park through the Gallatin National Forest, which is also crossed by 
MT-64. At lower elevations along the US-191 road corridor, privately held ranch land provides habitat, 
including critical winter range for wildlife, along with livestock. 
The Environmental Scan developed for the MDT Corridor Study of US-191 describes the study area as 
follows: 
Vegetation below the tree line consists of coniferous forests, grasslands, shrublands, and willow 
and aspen groves in the riparian areas. The coniferous forest community is dominated by conifers 
such as lodgepole pine and Douglas fir but also contains Engelmann spruce and subalpine fir. Big 
sagebrush dominates the grassland shrubland community, with other co-dominant shrubs including 
silver sagebrush, antelope bitterbrush, three-tip sagebrush, Idaho fescue, spike fescue, and poverty 
oatgrass. The riparian community is dominated by black cottonwood, aspen, snowberry, Wood’s 
rose, white spirea, red-osier dogwood, pacific willow, sandbar willow, reed canarygrass, and 
smooth scouring rush. Areas of cultivated cropland and developed lands are also present in the 
study area, primarily from Four Corners to the mouth of Gallatin Canyon (Montana Department of 
Transportation 2020). 
Most of the land adjacent to the road corridor in the study area is managed by the U.S. Forest Service, with 
numerous recreational accesses within the Custer Gallatin National Forest, including the Lee Metcalf 
Wilderness Area and indirect access to the Beaverhead-Deerlodge National Forest. The state-owned 
Gallatin Wildlife Management Area encompasses 8,611 acres along the eastern side of US-191, just south 
of MT-64. Just over 20 miles of the study area pass through Yellowstone National Park. US-191 serves the 
communities of Four Corners, Gallatin Gateway, and Big Sky and the town of West Yellowstone, with 
developed areas experiencing unprecedented growth in recent years. 
As described in the MDT Environmental Scan (2020), much of the area along US-191 provides significant 
wildlife habitat: 
The Gallatin River riparian corridor provides important wildlife habitat in areas that have been less 
impacted by adjacent development. Common wildlife includes those found in the adjacent 
shrub/woodlands and grasslands and species frequenting riparian areas like bats, martens, weasels, 
and raccoons. The riparian zone provides habitat for ducks, geese, herons, eagles, and other raptors, 
as well as migratory songbird species found in the adjacent non-riparian areas. Common riparian 
species include painted turtles, northern leopard frogs, and western toads. In addition, the Gallatin 
Canyon provides forested and riverine habitat for a variety of Montana wildlife species, including 
large ungulates, carnivores, small mammals, raptors, amphibians, reptiles, and aquatic species. 
The Gallatin Range provides suitable habitat for elk, moose, mountain goats, and bighorn sheep 
because of its relatively large size, its relatively diverse and high-quality vegetative communities, 
and elevational relief, its geographic location and connectivity to other habitats, and its relatively 
low level of human development. In addition to providing habitat for resident wildlife, the Gallatin 
 
6 The National Park Service describes the Greater Yellowstone Ecosystem in the following link: 
https://www.nps.gov/yell/learn/nature/greater-yellowstone-ecosystem.htm 
 
 
7 
 
Range plays a role in maintaining habitat connectivity for wide-ranging wildlife species such as 
wolverine, lynx, grizzly bear, mountain lion, and wolf. The Gallatin Range represents the northern 
reaches of core wildlife habitat within the Greater Yellowstone Ecosystem, and the northern end 
of the range forms a possible linkage to a wildlife corridor that may connect the Greater 
Yellowstone Ecosystem to the Northern Continental Divide Ecosystem (Gehman 2010).  Grizzly 
bears currently occupy the entire Gallatin and Madison Ranges, which are the boundaries of the 
Gallatin Canyon. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8 
 
 
Figure 7 Study Area Overview Map 
 
9 
 
2. Methodology 
2.1. Study area 
The study focused on two roads within the Greater Yellowstone Ecosystem, US Highway 191 (US-191) 
and Montana Highway 64 (MT-64, also known as Lone Mountain Trail), for which adequate data were 
available for identifying priority sites and potential wildlife accommodations to reduce wildlife-vehicle 
collisions and maintain or improve habitat connectivity. The study area includes 82 miles of US-191, from 
road milepost (RM) 0 through RM 81, and 10.1 miles of MT-64, from RM 0 through RM 9. 
In order to engage with communities along US-191 and MT-64, project members reached out to businesses, 
homeowner associations, community groups, and nonprofit partners to invite staff, residents, and 
commuters to use the Roadkill Observation and Data System (ROaDS) Tool to record wildlife sightings 
and gain local insight through an interactive map. Users logged more than 2,225 sightings of live and road-
killed animals and provided 100 additional comments. Further, project members have spoken at public 
gatherings, engaged with local Chambers of Commerce, spent time at community events, and provided 
information to local organizations to help raise awareness of the Assessment and invite participation. 
2.2. Data sources 
The CLLC-WTI Research Team acquired a diversity of data sets to inform the analyses, which can be 
classified into five categories: (1) law enforcement records of wildlife-vehicle collisions (WVCs); (2) 
records of wildlife carcasses observed along roads; (3) observations of live wildlife on or near roads; (4) 
observations of wildlife crossing roads; and (5) other ecological data such as wildlife telemetry locations, 
habitat models, and connectivity models. Available data emphasize large mammals, which tend to be 
species of management or conservation concern (e.g., grizzly bears) or greater human safety risks 
associated with collisions (e.g., elk, moose, deer). Data sources vary with respect to temporal extent, spatial 
extent, locational accuracy, and number of observations. Brief descriptions of the data sets in each category 
follow, with the characteristics of each data set summarized in Table 1. 
Wildlife-Vehicle Collision Data: We obtained Montana Highway Patrol (MHP) records for wildlife-
vehicle collisions (WVCs) occurring during 2012-2020 via a data request submitted to the Montana 
Department of Transportation (MDT). These records included the date, time, and location of all WVCs to 
which MHP responded. 
Carcass data: We obtained data on animal carcasses observed along study area roads from four sources: 
(1) MDT maintenance personnel records from 2012-2020; (2) National Park Service (NPS) records of 
wildlife carcasses along study area roads from 1989-20217; (3) Interagency Grizzly Bear Study Team 
(IGBST) records of grizzly carcasses along study area roads from 1977-2021; and (4) records of wildlife 
carcasses recorded by citizen scientists using the Roadkill Observation and Data System (ROaDS) 
smartphone application from March 6, 2021 through March 8, 2022. We reviewed these data for possible 
duplication of records by date and distance and removed any we suspected, as described in Table 1 below. 
Wildlife Observation data: We obtained records of observations of live animals on or near roads from 
four sources: (1) Montana Fish, Wildlife and Parks (FWP) flight monitoring data from 2002-2022; (2) FWP 
Madison Elk Herd telemetry data* from 2006-2020; (3) FWP Gallatin Elk Herd telemetry data* from 2002-
 
7 Note that NPS records were not used in spatial analyses but were used to supplement data in priority sites identified, 
when available, and in developing recommendations. 
*Elk collar data do not cover the entire study area, including the area north of RM 51, where high rates of collisions 
with elk exist. For more information on elk collar data limitations, please see Section 2.4. 
 
 
10 
 
2005; and (4) observations of live wildlife recorded by citizen scientists using the ROaDS smartphone 
application during 2021-2022. 
Wildlife Movement data: We received data on wildlife movement from three sources: (1) Interagency 
Grizzly Bear Study Team (IGBST) road crossing locations of 34 GPS-collared grizzly bears and crossing 
frequency records during 2000-2020; (2) observations of wildlife crossings of study area roads recorded 
by citizen scientists using the ROaDS smartphone application during 2021-2022; and (3) Montana Fish, 
Wildlife and Parks (FWP) elk telemetry data* for the Madison (2006-2020) and Gallatin (2002-2005) elk 
herds. As noted in the prior paragraph, elk telemetry data were used to infer locations of road crossings, as 
detailed below in Section 2.3.1. 
Habitat Suitability & Connectivity Models: We obtained habitat suitability and connectivity data for 
multiple species, including habitat specialists and species that do not disperse to new areas readily, to 
capture the habitat and movement needs of a diversity of wildlife. The models, which are described below 
and in Table 1, are derived from several sources: 
We used geospatial data from A Wildlife Conservation Assessment of the Madison Valley, Montana (Brock 
et al. 2006) on potential habitat suitability for several focal species, including wolverine, bighorn sheep, 
elk, and boreal toad. These data incorporate relevant human influences and predict the distribution of 
habitat quality. Potential winter, nesting, and brood habitat for sage grouse and potential habitat for grizzly 
bears are also included. 
We used geospatial data from Northern Rockies Black Bear Connectivity (Cushman, Lewis, and Landguth 
2013) generated by applying cumulative factorial least-cost path modeling coupled with resistant kernel 
analysis to predict a movement corridor network associated with locations of actual bear highway 
crossings. 
We used geospatial data from Peck et al. (2017) on potential grizzly bear road crossings, which include 
raster values converted to point features 300 m apart at intersections with major transportation corridors 
for grizzly bear movement between the Greater Yellowstone Ecosystem and the Northern Continental 
Divide Ecosystem based on Randomized Shortest Path analysis. 
We used geospatial data from Brock (unpublished) on range-wide habitat suitability for bison in winter and 
summer based on Shamon et al. (2022). 
We used geospatial data from Krosby et al. (2018) on riparian climate corridors to determine where roads 
intersect with riparian zones that are likely to facilitate climate-induced species range shifts. This data set 
includes a resiliency index for each riparian zone based on its ability to facilitate range shifts and serve as 
a climate micro-refugium (estimated based on the temperature gradient along its length and degree of 
canopy cover, solar insolation, and human modification). 
We used geospatial data from Dickson et al. (2016) on ecological connectivity to determine where roads 
intersect with major dispersal corridors. This data set is the product of a Circuitscape analysis of species-
neutral connectivity among large, protected areas within the western U.S. and contains a connectivity value 
for each landscape pixel reflecting its estimated contribution to west-wide connectivity. Movement is 
assumed to be more difficult through areas with more rugged topography and a higher degree of human 
modification. 
 
 
*Elk collar data do not cover the entire study area, including the area north of RM 51, where high rates of collisions 
with elk exist. For more information on elk collar data limitations, please see Section 2.4. 
 
11 
 
Table 1 Description of Data Sets Used in Analyses 
Dataset Description Sampling Spatial 
Period Extent Sample Size1 Precision2 
Wildlife Crash 
(MDT) Coordinates, date, time 2012-2020 US-191 
and MT-64 328 Good 
1,077 Removed 
Wildlife Carcass Coordinates of wildlife US-191 2 grizzly bear 
(MDT) carcasses, date, species, sex 2012-2020 and MT-64 records Good 
included in 
IGBST data 
Grizzly Bear Coordinates, ID, age class, 
Roadkill (IGBST) sex, date, location details of 1977-2021 US-191 12 Good 
loss 
62 minus 7 that 
Records of wildlife were off study 
ROaDS Tool: carcasses submitted by local 3/2021- US-191 routes = 55 
Wildlife Carcass residents using the ROaDS 3/2022 and MT-64 minus 1 Good 
smartphone application duplicate dated 
12/2/21 = 54 
US-191 Flight Coordinates, date, and 
Monitoring Data species from aerial 2002-2022 US-191 3041 Good 
(FWP) monitoring flights 
Records of live wildlife on 
ROaDS Tool: or near roads submitted by US-191 
Wildlife Alive on local residents using the 3/2021-
3/2022 and MT-64 122 Good 
Road ROaDS smartphone  
application 
Gallatin Elk Herd 
Telemetry Data Coordinates, Animal ID, Within 5 
(FWP) Device ID, date, time 2002-2005 miles of 33,772 Good 
US-191 
Madison Elk Within 5 
Herd Telemetry Coordinates, Animal ID, 
Device ID, date, time 2006-2020 miles of 102,152 Good 
Data (FWP) US-191 
Grizzly Bear 
Crossings  US-191 34 individuals Good 
Telemetry Data Animal ID, age, sex, cohort 2000-2020 and MT-64 236 events  
(IGBST) 
Generated from Madison Gallatin 
Valley and Gallatin Herd Herd 2002-
Elk Crossings data (approach described in 2005; Within 5 37 individuals Good 
(FWP) Section 2.3.1); coordinates, Madison miles of 
Animal ID, Device ID, date, Valley Herd US-191 2,189 events  
time 2006-2020 
ROaDS Tool: Records of wildlife 
Wildlife crossings submitted by local 3/2021- US-191 
Crossings residents using the ROaDS 3/2022 and MT-64 62 Good 
smartphone tool  
Potential Grizzly Point features 300 m apart 
Bear Passage identifying indices of GYE to 3
Along Major potential passages at 2018 NCDE 339 on US-191 Moderate  
Road Corridors intersections with major 
 
12 
 
Dataset Description Sampling Spatial 
Period Extent Sample Size1 Precision2 
in Northwest transportation corridors for 
Montana grizzly bear movement 
between GYE and NCDE 
based on Randomized 
Shortest Path (Peck et al. 
2016). 
Grizzly Bear 
Habitat Functional habitat of a 
Suitability minimum quality, size, and 
distance from major core N/A Study Area N/A Moderate3 
(Craighead et al. 
2006) areas 
Cumulative factorial least 
Northern Rockies cost path modeling coupled 
Black Bear with resistant kernel 
Connectivity analysis to predict Montana, 
movement corridor 2013 Idaho N/A Moderate3 
(Cushman et al. 
2013) network; associated with 
locations of actual bear 
highway crossings. 
Wolverine 
Habitat Potential habitat mapped 
Effectiveness using logistic regression N/A Study Area N/A Moderate3 
(Brock et al. from wolverine telemetry 
2006) and GPS locations 
Preferred summer habitat 
Bison Summer types and estimates of plant 
Habitat biomass needed to support 
Suitability Index bison populations; restricted Range- 3
(Shamon et al. to slopes below 35 percent; N/A wide N/A Moderate  
2022) further constrained to 
remove roads and cropland. 
 
Potential year-round (a.k.a. 
Bison Winter winter range) for bison in 
Habitat the mountainous West; 
Suitability Index generally a subset of Rocky 3
(Brock summer habitat where snow N/A Mountains N/A Moderate  
unpublished) depth remains low through 
the winter, allowing for 
year-round use. 
Bighorn Sheep Preferred habitat types 
Potential Habitat within a certain distance of 
Effectiveness escape terrain, adjusted to 3
(Brock et al. eliminate areas impacted by N/A Study Area N/A Moderate  
2006) domestic sheep grazing 
allotments and road salting 
Elk Habitat Range of habitats with 
Effectiveness emphasis on winter range, 3
(Brock et al. riparian areas, mountains, N/A Study Area N/A Moderate  
2006) and valleys 
 
13 
 
Dataset Description Sampling Spatial Sample Size1
Period Extent  Precision2 
Modeled habitat 
Boreal Toad components (i.e., wetlands, 
Habitat landcover, edge, and soils) 
Effectiveness within 300 m of lakes, 
(Brock et al. ponds, and springs and N/A Study Area N/A Moderate3 
2006) adjusted to address threats 
 (dewatering, fish stocking, 
floodplain loss, pollution, 
roads) 
Sage Grouse Modeled stands of 
Potential Nesting sagebrush with 15-31% 
Habitat (Brock et cover using 30 m GAP N/A Study Area N/A Moderate3 
al. 2006) Land Cover and late May 
NDVI 
Fall greenness calculated 
Sage Grouse from Landsat imagery, 
Potential Brood masked to include only 
Habitat (Brock et areas within potential brood N/A Study Area N/A Moderate3 
al. 2006) habitat landcover 
 classes (e.g., sagebrush, 
shrub/steppe, grassland) 
Preferred habitat types 
Sage Grouse refined with slope, aspect, 
Winter Habitat and NDVI calculated from 
Suitability (Brock spring Landsat Thematic N/A Study Area N/A Moderate3 
et al. 2006) Mapper imagery to detect 
 areas where sagebrush 
protrudes above snow cover 
Ecological Estimated value for 
Connectivity facilitating ecological flows 
(Dickson et. al. (e.g., wildlife movement) NA Study Area NA Moderate3 
2016) between protected areas in 
the Western U.S. 
Riparian Climate Estimated value of riparian 
Corridors corridors for facilitating 3
(Krosby et al. climate-induced species NA Study Area NA Moderate  
2018) range shifts 
1 Sample size is the total number of observations (e.g., carcasses, crashes, or live animal sightings) along study roads. 
2 A rough, categorical estimate of locational error associated with the data set. Good: location recorded at the time of observation 
with GPS coordinates. Moderate: location recorded at time of observation with road reference marker to the nearest mile or 
detailed location description based on local landmarks. Poor: location estimated based on memory of past events. 
3 Data set consists of connectivity model output rather than locations of specific events, assigned to the moderate precision 
category based on the spatial resolution of the model. 
 
2.3. Initial Screening of Road Segments 
To identify an initial set of road segments for potential wildlife accommodation measures, we categorized 
the data sets described in Section 2.2 and Table 1 into four Prioritization Characteristics, each of which 
represents a specific mitigation rationale: 1) wildlife-vehicle collision risk, 2) live wildlife observations 
along roads, 3) evidence of wildlife crossing roads, and 4) modeled wildlife habitat suitability or 
connectivity value. Each Prioritization Characteristic is informed by a distinct subset of the 26 available 
data sets, which were weighted according to the sample size, spatial precision, duration, and diversity or 
 
14 
 
conservation importance of wildlife species represented (see Table 2), and then applied to each road 
segment. Finally, based on a weighted average of the four Prioritization Characteristics (see Table 3), we 
calculated a Composite Importance Index for each road segment. 
To identify potential locations for wildlife crossing improvements, we used 0.10-mile road segments as our 
unit of analysis. We determined these segments as follows: (1) we used the Montana Department of 
Transportation (MDT) reference markers point shapefile to divide polyline shapefiles for US-191 and MT-
64 into 1-mile segments between mile markers using the Split Line at Point tool in ArcGIS. [N.B. The 
starting mile MDT reference marker was included in each segment]; (2) each 1-mile segment was then 
divided into 0.10-mile segments using the Split Command and given a unique ID using the MDT reference 
marker as a prefix (e.g., RM 81.0 through RM 81.9). Overall, the study area includes 921 0.10-mile 
segments. 
 
 
Table 2 Within-Category Weights for Each Prioritization Characteristic 
Within-
Dataset category Weight justifications 
weight 
Wildlife-Vehicle Collision Risk Prioritization Characteristic 
Wildlife Crash (MDT) 0.4 Large, precise dataset 
Wildlife Carcass (MDT) 0.4 Large, precise dataset 
Grizzly Bear Roadkill (IGBST) 0.1 Very small number of observations, but high 
conservation importance 
Small number of observations and short study duration; 
ROaDS Tool: Wildlife Carcass 0.1 probably poorer spatial precision than other WVC 
datasets 
Wildlife Observation Near Roads Prioritization Characteristic 
Flight Monitoring Data: US-191 
(FWP) 0.3 Large, long-duration, multi-species dataset 
ROaDS Tool: Wildlife Alive on 0.1 Small number of observations and short study duration; 
Road likely poorer spatial precision than other WVC datasets 
Gallatin Elk Herd (FWP) 0.3 Large, precise, single-species dataset, only partial spatial 
coverage 
Madison Elk Herd (FWP) 0.3 Large, precise, single-species dataset, only partial spatial 
coverage 
Wildlife Road Crossings Prioritization Characteristic 
Grizzly Bear Crossings (IGBST) 0.45 Moderate number of observations and high conservation 
importance 
Elk Crossings (FWP) 0.45 Large number of observations but lower conservation 
importance, only partial spatial coverage 
ROaDS Tool: Wildlife Crossings 0.1 Small number of observations and likely poorer spatial 
precision than other movement datasets 
 
15 
 
Habitat Suitability/Connectivity Prioritization Characteristic 
Potential grizzly bear passage along Each of the data sets in this category were weighted 
major road corridors in northwest 0.05 equally. For species with more than one model, weights 
Montana (Peck et al. 2016) were divided equally among the multiple models (e.g., 
each of the 2 grizzly bear models weighted at 0.05). 
Grizzly Bear Habitat Suitability 
(Craighead et al. 2006) 0.05  
Northern Rockies Black Bear 
Connectivity (Cushman et al. 2013) 0.1  
Wolverine Habitat Effectiveness 
(Brock et al. 2006) 0.1  
Bison Summer Habitat Suitability 
Index (Shamon et al. 2022) 0.05  
Bison Winter Habitat Suitability 
Index (Brock unpublished) 0.05  
Bighorn Sheep Potential Habitat 
Effectiveness (Brock et al. 2006) 0.1  
Elk Habitat Effectiveness (Brock et 
al. 2006) 0.1  
Sage Grouse Potential Nesting 
Habitat (Brock et al. 2006) 0.033  
Sage Grouse Potential Brood 
Habitat (Brock et al. 2006) 0.033  
Sage Grouse Winter Habitat 
Suitability (Brock et al. 2006) 0.034  
Boreal Toad Habitat Effectiveness 
(Brock et al. 2006) 0.1 
Ecological Connectivity (Dickson  
et al. 2016) 0.1 
Riparian Climate Corridors (Krosby 
et al. 2018) 0.1 
 
Table 3 Weights for Composite Importance Index 
Prioritization Characteristic Weight 
Wildlife-Vehicle Collision Risk 0.30 
Wildlife Observations Near Roads 0.10 
Wildlife Road Crossings 0.30 
Habitat Suitability/Connectivity 0.30 
 
2.3.1. Indices of Road Segment Importance for each Prioritization Characteristic 
Establishing a segment-level importance index for each Prioritization Characteristic required combining 
results across data sets. First, we developed segment-level indices for each data set within each 
characteristic. Indices for the WVC Risk, Wildlife Observation, and Wildlife Road Crossing Prioritization 
Characteristics are based on counts of events (e.g., crashes, carcasses, pathways intersecting the road) or 
numbers of individuals (e.g., live animals) within each 0.10-mile road segment. We generated indices for 
the Habitat Suitability/Connectivity Prioritization Characteristic by extracting the maximum value (e.g., 
 
16 
 
highest suitability, maximum connectivity value) of the pixels intersecting each road segment. The Index 
values generated for each data set were rescaled to range from 0 to 1 to allow for comparison. We used a 
weighted averaging approach for the contributing data sets to each Prioritization Characteristic to derive 
segment-level importance indices. 
We also generated general summary statistics for each data set and Prioritization Characteristic. 
Wildlife-Vehicle Collision Risk Prioritization Characteristic: We calculated the number of recorded 
collisions with wildlife and the number of carcasses within each 0.10-mile road segment, as described in 
Table 1, as an index of WVC risk using 4 data sets: 
• Wildlife Crash (MDT) 
• Wildlife Carcass (MDT) 
• Grizzly Bear Roadkill (IGBST) 
• ROaDS Tool: Wildlife Carcass 
We calculated the rate of collisions or carcasses (x per 0.10-mile) for each segment in each data set. We 
then normalized the native values from each data set (i.e., rescaled from 0 to 1). Using the within-category 
weights in Table 2, we developed a Wildlife-Vehicle Collision Importance Index calculated from the 
weighted mean of the normalized values across the data sets. 
Wildlife Observation Near Roads Prioritization Characteristic: We calculated the number of recorded 
wildlife observations within 500 meters of each 0.10-mile road segment as an index of animal use intensity 
using 4 data sets: 
• Flight Monitoring Data US-191 Clip (FWP) 
• ROaDS Tool: Wildlife Alive on Road 
• Gallatin Elk Herd (FWP)* 
• Madison Elk Herd (FWP)* 
We calculated the number of recorded wildlife observations within 500 m of the road for each segment 
within each data set. We then normalized the native values from each data set (i.e., rescaled from 0 to 1). 
Using the within-category weights in Table 2, we developed a Wildlife Observation Importance Index 
calculated from the weighted mean of the normalized values across the data sets. 
Wildlife Road Crossings Prioritization Characteristic: We calculated the number of grizzly bear or elk 
paths intersecting each 0.10-mile road segment (i.e., number of inferred road crossings by bears or elk) or 
live observations of road crossings from the ROaDS Tool as an index of safe wildlife passage for each of 
the 3 data sets: 
• Grizzly Bear Crossings (IGBST) 
• Elk Crossings (FWP) * 
• Roads Tool: Wildlife Crossings 
Using the elk telemetry data, we converted locations into movement paths by assuming straight-line travel 
between consecutive telemetry fixes, limiting our analysis to fixes separated by less than 8 hours to 
minimize potential deviation from assumed straight-line paths. We then determined where inferred 
movement paths intersected the road network (i.e., approximate locations of elk crossings). 
 
* Elk collar data do not cover the entire study area, including the area north of RM 51, where high rates of collisions 
with elk exist. For more information on elk collar data limitations, please see Section 2.4. 
 
17 
 
For each data set, we calculated the number of paths intersecting the road or recorded the number of wildlife 
crossings for each segment. The native values from each data set were then normalized (i.e., rescaled from 
0 to 1). Using the within-category weights in Table 2, we developed a Wildlife Movement Importance 
Index calculated from the weighted mean of the normalized values across the data sets. 
Habitat Suitability/Connectivity Prioritization Characteristic: We generated indices for the Habitat 
Suitability/Connectivity prioritization characteristic by extracting the maximum value (e.g., highest 
suitability, maximum connectivity value) of the pixels overlapping each road segment in each data set: 
• Potential grizzly bear passage: We calculated the maximum index value of potential passage rates 
for grizzly bears overlapping each road segment (Peck et al. 2017) as an index of importance for 
facilitating regional connectivity for grizzly bears. 
• Grizzly bear habitat (Craighead 2006): We calculated the maximum suitability value of the pixels 
overlapping each road segment as an index of importance for grizzly bear habitat. 
• Northern Rockies black bear connectivity (Cushman, Lewis, and Landguth 2013): We calculated 
the maximum connectivity value of the pixels overlapping each road segment from Cushman, 
Lewis, and Landguth (2013) as an index of importance for black bear connectivity. 
• Bison summer and winter habitat suitability (2 data sets): We calculated the maximum habitat 
suitability value of the pixels overlapping each road segment as an index of importance for bison 
winter and summer habitat from Brock (unpublished) and Shamon et al. (2022), respectively 
• Wolverine habitat effectiveness: We calculated the maximum value from Brock et al. (2006) of the 
pixels overlapping each road segment as an index of importance for Wolverine habitat. 
• Bighorn sheep habitat effectiveness: We calculated the maximum value from Brock et al. (2006) 
of the pixels overlapping each road segment as an index of importance for bighorn sheep habitat. 
• Elk habitat effectiveness: We calculated the maximum value from Brock et al. (2006) of the pixels 
overlapping each road segment as an index of importance for elk habitat. 
• Sage grouse winter, nesting, and brood habitat (3 data sets): We calculated the maximum value 
from Brock et al. (2006) of the pixels overlapping each road segment as an index of importance for 
sage grouse winter, nesting, and brood habitat. 
• Boreal toad habitat effectiveness: We calculated the maximum value from Brock et al. (2006) of 
the pixels overlapping each road segment as an index of importance for boreal toad habitat. 
• Riparian climate corridors: We calculated the maximum resiliency index value from Krosby et al. 
(2018) of any riparian zones intersecting each road segment as an index of potential importance 
for climate change adaptation. 
• Ecological connectivity: We calculated the maximum connectivity value from Dickson et al. 
(2016) of the landscape pixels overlapping each road segment as an index of importance for multi-
species connectivity. 
We then normalized the values from each data set (i.e., rescaled from 0 to 1). Using the within-category 
weights in Table 2, we developed a Habitat Suitability/Connectivity Importance Index calculated from the 
weighted mean of the normalized values across the data sets. 
2.3.2. Composite Indices of Road Segment Importance 
We weighted each Prioritization Characteristic, as shown in Table 3, to develop a Composite Importance 
Index for each road segment: 
(WVC Importance Index x 30%) + (Wildlife Observation Importance Index x 10%) + (Wildlife Movement 
Importance Index x 30%) + (Suitability/Connectivity Importance Index x 30%) = Composite Importance 
Index. Following this calculation, the resulting values were rescaled from 0 to 1. 
 
18 
 
To identify priority sites for field evaluation, we spatially smoothed the results of the Composite Importance 
Index using a Moving Window Average (MWA) to identify road areas with consistently elevated values 
for the prioritization characteristics. We calculated an MWA for each 0.10-mile road segment by taking 
the average of the Composite Importance Index values for each road segment with the values for the five 
segments to either side of each segment (i.e., each MWA value applied to a road section of 1.1 miles). We 
then selected the Top 10% of MWA values for each highway as areas for field evaluation. 
The approaches described above tend to identify road segments with high index values across multiple 
characteristics and may overlook road segments that are extremely important for a single characteristic, 
such as wildlife-vehicle collision risk (that have lower composite index values). In order to include WVC 
outlier segments with high risk, we added 2 areas along US-191 to the priority sites identified by the MWA 
for field evaluation (see Section 3.6 for further discussion). 
2.4. Data Gaps, Limitations, and General Assumptions 
The analysis within this Assessment is based on available data. Not all data sets are comprehensive; some 
are collected opportunistically (including both MDT and ROaDS Tool carcass data), and the data sets are 
skewed towards large mammals and charismatic species such as elk and grizzly bears. Further, some basic 
assumptions have been made throughout the document as information is still being learned about the ways 
in which wildlife interact with roads and transportation. Some data gaps, limitations, and general 
assumptions of this report that require further research to better assess their impact are described below: 
Underreporting: The primary carcass data available for this study are: MDT maintenance personnel 
removal of wildlife carcasses along roadsides and citizen science-collected data. These data sets are 
opportunistically—rather than systematically—collected and are likely to represent only a fraction of the 
animals hit and killed along roads. These data sets are also skewed towards large animals that can be easily 
seen when driving along a highway (> 90% of all MDT carcass data are deer). In addition to not accounting 
for smaller wildlife, these data are likely an undercount of large species such as deer and elk, given that 
many animals hit do not die on a road surface or immediately next to a road. Studies that document the 
underreporting of WVCs have found up to 8.5 times more animals are hit on roads than reported 
(Donaldson 2017). In Montana, while crashes reported to law enforcement have nearly doubled over the 
last 10 years, the number of wildlife carcasses removed from roads has declined (Figure 5). This is more 
likely due to changes in search and reporting procedures than fewer wildlife-vehicle collisions on 
Montana’s roads. This factor is important to consider in reading Chapter 11 on cost-benefit analysis for 
constructing wildlife accommodation measures, as cost-benefit thresholds are based on carcass data. 
Gaps in Existing Wildlife Movement Data: Individual, GPS-collared wildlife provides the primary 
wildlife movement data available for the study. These data were only available for two species: elk from 
the Gallatin and Madison herds and grizzly bears. The elk collar data are not necessarily representative of 
all elk herds or their movement in the study area. Further, the data do not cover the entire study area, 
including the area north of RM 51, where high rates of collisions with elk are documented. This is a limiting 
factor in the Wildlife Crossing Road Importance Index as the two sets of collar data are primary 
contributors. During the site visits, local wildlife biologists scored each priority site based on conservation 
value, including the importance for wildlife movement to offset this limitation. Collar data for an expanded 
group of species and additional elk herds, such as those in the Gallatin Gateway area, could help to better 
inform potential locations for wildlife accommodations. 
Traffic Volumes: Studies of the impact of traffic volumes on wildlife crossing behavior and WVC risk are 
limited in number. For moose, Seiler (2003) found a non-linear relationship between traffic volumes and 
a) WVCs and b) the barrier effect of a highway. The highest rates of WVCs were at moderate traffic 
volumes (2,500-10,000 vehicles per day; Figure 8), with the road posing nearly a complete barrier for safe 
wildlife crossings at traffic volumes between 10,000 and 15,000 vehicles per day. A recent study in 
 
19 
 
Wyoming (Riginos 2022) found traffic volumes above 15,000 AADT pose a complete barrier to wildlife 
movement. 
 
Figure 8 The Relationship between Average Daily Traffic, WVCs, and the Barrier Effect of a Road (from Seiler 
2003). 
 
Some species-specific studies have described hourly traffic thresholds. A study in northwest Montana 
found traffic volumes of over 100 vehicles per hour to be a substantial barrier to grizzly bear movement 
(Waller and Servheen 2005; Waller and Miller 2015). For mule deer, a study in western Wyoming found 
traffic volume over 120 vehicles per hour to be a substantial barrier, with interactions between deer and 
vehicles often unsafe or not allowing passage when less than 60 seconds between vehicles exist (Riginos 
et al. 2018). These finer-scale studies help to show the degree to which a road poses a barrier. For the 
Assessment, the available traffic volume data are Annual Average Daily Traffic (AADT), which is an 
average across the full year that does not account for temporal (day/night) or seasonal (winter/summer) 
variations. Each of the US-191 and MT-64 road corridors is highly influenced by temporal and seasonal 
variations. AADT is not adequate to determine when wildlife movements are likely to be impeded by traffic 
volume. Additional research is necessary to determine whether portions of US-191 and MT-64 have 
become a significant barrier to wildlife movement and when they are likely to become one. 
  
 
20 
 
3. Results 
3.1. Wildlife Road Crossings Importance Index 
In the Wildlife Road Crossings Prioritization Characteristic, 48% (439/921) of segments have recorded safe 
passages, with the top 10% of these segments having index values of 0.463 or greater (Figure 9). The total 
number of wildlife detected crossing roads is 2,484, with a maximum of 39 crossings in any one 0.10-mile 
segment (mean = 2.70, SD = 4.85). All data for the top 10 segments identified in the Wildlife Road Crossing 
Importance Index are shown in Table 4. 
Table 4 Wildlife Road Crossing Importance Index: Top Ten 0.10-mile Segments 
Road Add Weighted N = Wildlife 
Segment Elk Griz ROaDS Elk Crsgs Griz Crsgs ROaDS (ElkCrsN*0.45) + Crossings 
ID Crsgs Crsgs Crsgs Normalized Normalized Crsgs 
Normalized (GrizCrsN*0.45) + Imp 
(ROaDCrsN*0.1) Index 
RM7.7 2 5 0 0.051 1.000 0.000 0.473 1.000 
RM3.1 0 5 0 0.000 1.000 0.000 0.450 0.951 
RM46.4 39 0 0 1.000 0.000 0.000 0.450 0.951 
RM46.3 37 0 0 0.949 0.000 0.000 0.427 0.902 
RM46.0 35 0 0 0.897 0.000 0.000 0.404 0.854 
RM46.5 33 0 0 0.846 0.000 0.000 0.381 0.805 
RM11.1 1 4 0 0.026 0.800 0.000 0.372 0.785 
RM1.4 0 4 0 0.000 0.800 0.000 0.360 0.761 
RM7.9 0 4 0 0.000 0.800 0.000 0.360 0.761 
RM45.9 29 0 0 0.744 0.000 0.000 0.335 0.707 
 
There are 2,189 crossing events from 37 individual collared elk in the Gallatin and Madison herds. These 
collar data have the largest sample size of all data used in this analysis. Thus, it is not surprising that elk 
make up the great majority (88%: 2,189/2,484) of crossings, with a maximum of 39 elk crossings in any 
one segment (RM 46.4) and a mean of 2.38 (SD = 4.70). Elk crossings occur in 353 of the 921 0.10-mile 
segments. Roughly 20% (70/353) of these segments have > 10 elk crossings and 10 have > 20 elk crossings, 
all of which are in RM 45 and 46. About 55% (195/353) of segments have < 5 elk crossings. Of the 37 
collared elk that cross the study roads, seven individuals make > 145 crossings and 3 individuals cross the 
highway more than 200 times, with elk #22164 making the greatest number of crossings at 248 events. 
Nineteen individuals cross between 10 and 88 times, and 11 individuals make <10 crossings. 
 
21 
 
 
F igure 9 Intensity of Wildlife Crossing Roads in the Study Area 
 
22 
 
There are 236 crossing events by 34 individual collared grizzly bears between 2000-2020. Grizzly bears 
make up 9.5% (236/2484) of total crossings, with a maximum of 5 grizzly bear crossings (mean = 0.26, 
SD = 0.66) in 2 segments (i.e., RM 3.1 and RM 7.7). Grizzly bear crossings occur in 17% (159/921) of the 
0.10-mile segments, with one-time crossing events in 68% (108/159) of segments. Five individual bears 
that make >10 crossings contribute roughly 55% (129/236) of crossing events, with GB0653 having the 
most crossings at 60 events, before eventually being struck and killed by a vehicle near RM 23 (Figure 10). 
 
Figure 10 Grizzly Bear Habitat Suitability, Crossing Frequency and Roadkill 
 
23 
 
There are 59 crossing events by various species (e.g., bison, elk, white-tailed deer, mule deer) recorded in 
the ROaDS Tool by citizen scientists in the analysis. ROaDS Tool crossings make up 2.4% of total 
crossings (59/2,484) and are documented in 14 of the 0.10-mile segments, ranging from 1 to 30 crossings 
per event. About 51% (30/59) of the recorded crossings are a herd of 30 bison crossing in segment RM 3.7. 
3.2. Wildlife Observation Importance Index 
In the Wildlife Observations Near Roads Prioritization Characteristic, 54% (498/921) of segments have 
various species of wildlife recorded within 500 m of US-191 or MT-64 in the study area, with the top 10% 
of these segments having importance index values of 0.364 or greater (Figure 11). The total number of 
wildlife observed is 10,727 with a maximum of 285 observations in any one 0.10-mile segment (mean = 
11.65, SD = 24.79). All data for the top 10 segments identified in the Wildlife Observation Importance 
Index are shown in Table 5. 
Table 5 Wildlife Observation Importance Index: Top Ten 0.10-mile segments 
(RDSN*0.10) 
+ 
ROaDS Madison Gallatin Flight (MADN*0.30) 
Road  ROaDS Madison Gallatin Flight Wildlife 
Elk Elk Clip Total # Obs Elk # Elk # Clip # + 
ID # Obs #Obs Normal- Obs Obs Obs (GALN*0.30) Obs 
#Obs #Obs #Obs ized Normal- Normal- Normal- + Imp 
ized ized ized (FLTN*0.30) Index 
= 
Observations 
RM46.7 0 0 28 206 234 0.000 0.000 0.609 0.805 0.424 1.000 
RM24.8 0 83 4 1 88 0.000 1.000 0.0870 0.004 0.327 0.772 
RM72.7 29 0 0 256 285 0.180 0.000 0.000 1.000 0.318 0.750 
RM46.5 0 0 38 56 94 0.000 0.000 0.826 0.219 0.313 0.739 
RM26.3 0 74 5 1 80 0.000 0.892 0.109 0.004 0.301 0.710 
RM46.2 0 0 46 1 47 0.000 0.000 1.000 0.004 0.301 0.710 
RM24.6 42 74 0 1 117 0.261 0.892 0.000 0.004 0.295 0.695 
RM46.1 0 0 33 62 95 0.000 0.000 0.717 0.242 0.288 0.679 
RM46.4 0 0 29 69 98 0.000 0.000 0.630 0.270 0.270 0.637 
RM29.9 0 58 9 0 67 0.000 0.699 0.196 0.000 0.268 0.633 
 
Wildlife observations from roughly 20 years of flight monitoring data make up 35% (3,761/10,727) of the 
total wildlife observations within 500 m of the study roads, ranging from 0 to 256 observations per event 
(mean = 4.08, SD = 18.23). These observations are documented in 24% (217/921) of the 0.10-mile 
segments. Of the 217 events, 3 road segments have over 200 individuals observed in discrete events (RM 
72.6, RM 72.7, and RM 46.7) and 7 road segments have over 100 individuals observed in each event. The 
primary species encountered within 500 m of the study area roads, from the greatest to least number of 
occurrences, include elk, bighorn sheep, moose, bison, black bear, mule deer, white-tailed deer, and wolf. 
 
 
 
24 
 
 
Figure 11 Intensity of Wildlife Use of Roadside Environments in the Study Area 
 
25 
 
 
Data on collared individuals in the Madison Elk Herd make up roughly 31% (3,304/10,727) of the total 
wildlife observations within 500 m of study area roads, with a maximum of 83 individual elk observed in 
any segment (RM 24.8) and a mean of 3.59 (SD = 9.64). Elk observations from the Madison herd occur in 
256 of the 921 0.10-mile segments and all but 6 of these events occur between RM 20.1 and RM 36.7.  
Roughly 3.9% (10/256) of these segments have > 50 elk observed in any one event and 57 of these segments 
have > 20 elk observed. 
Data on collared individuals in the Gallatin Elk Herd make up about 22% (2,333/10,727) of the total 
wildlife observations within 500 m of the study area roads, with a maximum of 46 individual elk observed 
in any one segment (RM 46.2) and a mean of 2.53 (SD = 5.20).  Elk observations from the Gallatin herd 
occur in 38% (353/921) of the 0.10-mile segments. Twenty of the segments have > 20 individuals observed 
and all but 2 of these segments are between RM 44.9 and RM 49.3. 
There are 1,329 wildlife observations of various species (e.g., bighorn sheep, elk, bison, white-tailed deer, 
mule deer) recorded by citizen scientists in the ROaDS Tool included in the analysis. ROaDS Tool data 
make up 12.3% of total observations (1,329/10,727) and are documented in 73 of the 0.10-mile segments, 
ranging from 0 to 161 observations per event (mean = 1.44, SD = 10.64). There are 8 segments with > 50 
individuals recorded and all of these observations are between RM 71.8 and RM 75.4. 
3.3. Wildlife-Vehicle Collision Risk Importance Index 
In the Wildlife-Vehicle Collision Risk Prioritization Characteristic, 56% (512/921) of segments have at 
least one WVC (i.e., crash or carcass), with the top 10% of these segments having importance indices of 
0.411 or greater (Figure 12). The total number of WVCs (i.e., crash and carcass) across the 10 years 
analyzed is 1,852 with a maximum of 25 WVCs in any one 0.10-mile segment (mean = 2.01, SD = 3.62). 
Road Mile segment 79.8 has the highest total count of WVCs. All data for the top 10 segments identified 
in the Wildlife Vehicle Collision Importance Index are included in Table 6. 
In the MDT carcass data, 44% (408/921) of the segments have at least one carcass. The total number of 
carcasses across the 10 years analyzed is 1,347 with a maximum of 22 carcasses in any one 0.10-mile 
segment (mean = 1.46, SD = 3.0). Road Mile segment 77.6 has the highest total count of carcasses. The 
top 10 segments with carcasses are all between RM 73.0 and RM 79.8. 
 
 
26 
 
 
Figure 12 Frequency of Collisions with Wildlife in the Study Area based on Crash and Carcass Data 
 
27 
 
Table 6 Wildlife Vehicle Collision Risk Importance Index: Top Ten 0.10-mile Segments 
(CARC*0.4
# # # # 0) + 
# GRI #ROa Tota Carca # # Griz ROaD (CRASH*0.
Road ID CAR l Crash Kill S 40) + WVC 
CAS CRAS Z DS ss Imp 
S H KIL KILL WV Norm Norm Norm KILL (GRIZ*0.10
L C al-ized al-ized al-ized Norm ) + Index 
al-ized (RDS*0.10) 
= WVC 
RM79.0 19 5 0 0 24 0.864 0.833 0.000 0.000 0.679 1.000 
RM74.3 17 5 0 0 22 0.773 0.833 0.000 0.000 0.642 0.946 
RM79.5 11 6 0 1 18 0.500 1.000 0.000 0.333 0.633 0.933 
RM79.8 20 2 0 3 25 0.909 0.333 0.000 1.000 0.597 0.879 
RM77.5 13 5 0 0 18 0.591 0.833 0.000 0.000 0.570 0.839 
RM72.9 12 5 0 0 17 0.545 0.833 0.000 0.000 0.552 0.813 
RM72.7 5 6 0 1 12 0.227 1.000 0.000 0.333 0.524 0.772 
RM77.6 22 1 0 1 24 1.000 0.167 0.000 0.333 0.500 0.737 
RM79.6 7 4 0 3 14 0.318 0.667 0.000 1.000 0.494 0.728 
RM75.0 8 5 0 0 13 0.364 0.833 0.000 0.000 0.479 0.705 
 
In the MDT crash data, 29% (267/921) of the segments have at least one crash. The total number of crashes 
across the 10 years analyzed is 439 with a maximum of 6 crashes in any one 0.10-mile segment (mean = 
0.48, SD = 0.93). Road Mile segments 48.4, 72.7, and 79.5 have the highest number of crashes. Roughly 
60% (160/267) of the segments have just one crash, while 14% have 3 or more crashes. The great majority 
of crashes are on US-91 north of its intersection with MT-64, with quite a few crashes also within about 
the first mile MT-64. 
The Interagency Grizzly Bear Study Team (IGBST) has recorded a dozen grizzly bears killed on US-191 
over roughly a 40-year period. All recorded roadkills occur in 11 0.10-mile segments, with 2 grizzly bears 
as the maximum killed in any one segment (at RM 19.8). While none of the 11 segments are in top 10 of 
the WVC Importance Index shown in Table 6, four segments with grizzly bear kills (i.e., RM 9.5, RM 10.2, 
RM 10.9 and RM 35.2) occur within the Teepee Creek and Taylor Fork Priority Sites that are described in 
Chapter 6. 
The carcass data collected by citizen scientists using the ROaDS Tool over a one-year period include a total 
of 54 events, with a maximum of 3 individual carcasses recorded in any one event (mean = 0.06, SD = 
0.28). These data are recorded on 46 0.10-mile segments, with 76% of the events recorded north of RM 
50.0 on US-191. Only one event is recorded on MT-64 at RM 0.3. 
3.4. Habitat Suitability/Connectivity (Regional Conservation Value) Importance Index 
Figure 13 shows regional conservation value of all segments in the study area as described by Habitat 
Suitability/Connectivity Importance Index, with the top 10% having index values of 0.804 or greater. Table 
7 highlights the top ten 0.10-mile segments with the highest values of this index, the majority of which are 
along the southern extent of US-191 in the study area within Yellowstone National Park, with the highest 
value at RM 13.3. This is due to the expanse of relatively intact habitat within the park compared to the 
matrix of public and private land under development pressure further north. The average value of this index 
across all 921 0.10-mile segments is 0.53 (SD = 0.21) with a minimum value of <0.001 in the northernmost 
part of the study area. 
 
28 
 
Table 7 Habitat Suitability/Connectivity (Regional Conservation Value) Importance Index: Top Ten 0.10-mile 
Segments 
 
 
The Habitat Suitability/Connectivity Importance Index is derived from 14 different data sets (Table 1), 
most of which are species-specific. Table 8 provides summary statistics for the normalized values of each 
of the habitat suitability/connectivity model inputs. We highlight the results of two of the model inputs that 
are species-neutral here (see Figure 14). The information from Dickson et al. (2016), which is based on 
current ecological flows of connectivity among large, protected areas in the western U.S., shows the 
majority of the study area is relatively permeable south of RM 72.0. The most highly permeable area is 
between Bacon Rind and Teepee creeks, where 60 segments have connectivity values above 0.80. The 
average connectivity value along the study area roads is 0.46 (SD = 0.19). The top 10% of the riparian 
climate corridors delineated in Krosby et al. (2018) that intersect with study area highways are south of 
RM 18.0 and include Grayling, Teepee, Duck, and Cougar Creeks. However, the highest rank is RM 64.3, 
where Hell Roaring Creek flows into the Gallatin River, and RM 53.1-2, where Portal Creek flows into the 
Gallatin River, also scores in the top 10%. 
 
29 
 
 
Figure 13 Regional Conservation Value 
 
30 
 
 
Figure 14 Species-neutral Connectivity Showing General Permeability of the Landscape to Ecological Flows and 
Riparian Corridors Important for Climate Resilience 
 
31 
 
Table 8 Summary Statistics on Normalized Values for Habitat Suitability/Connectivity (Regional Conservation 
Value) Model Inputs 
No. of 
Data Input Min Max Mean SD Segments Highest Value Segment or 
with Values Range of Segments 
Ecological Connectivity 0 1 0.465 0.189 921 RM 13.4-14.0 (US-191) 
Riparian Climate Corridors 0.059 1 0.551 0.118 340 RM 64.3 (US-191) 
Boreal Toad Habitat 
Effectiveness 0.003 1 0.382 0.225 354 RM 11.0 (US-191) 
Sage Grouse Potential Brood 
Habitat 0.018 1 0.435 0.210 435 RM 8.6 (US-191) 
Sage Grouse Winter Habitat 
Suitability 0.287 1 0.615 0.123 144 RM 8.5 (US-191) 
Sage Grouse Potential Nesting (highest) RM 8.6 (US-191); 
Habitat 0.066 1 0.292 0.227 122 top 20 values between RM 
4.1-8.9 (US-191) 
Bighorn Sheep Habitat 
Effectiveness 0.128 1 0.574 0.274 422 RM 63.0-63.9 (US-191) 
Elk Habitat Effectiveness 0.124 1 0.671 0.145 921 RM 4.3 (MT-64) 
Bison Winter Habitat Suitability 0.437 1 0.740 0.115 553 RM 81.8-81.9 (US-191) 
Bison Summer Habitat 
Suitability 0.233 1 0.711 0.175 624 RM 1.0-7.1 (US-191) 
Wolverine Habitat Effectiveness 0.001 1 0.490 0.3052 424 RM 39.4 (US-191) 
Northern Rockies Black Bear 
Connectivity 0.005 1 0.046 0.117 463 RM 66.5-67.0 (US-191) 
(highest) RM 43.7 (US-191); 
154 segments > 0.8, most of 
Grizzly Bear Habitat Suitability 0.555 1 0.751 0.106 581 which are south of RM 44 on 
US-191; and between RM 
4.5-8.5 on MT-64 
Potential Grizzly Bear Passage 0.036 1 0.301 0.258 345 RM 63.6 (US-191) 
 
3.5. Priority Composite Index 
Figure 15 provides an overview of how each of the 921 0.10-mile road segments evaluated ranks in the 
Priority Composite Index, highlighting the top 10% of each highway separately. The top 10% of the 820 
segments analyzed on US-191 have index values between 0.689 and 1, while the top 10% of the 101 
segments evaluated on MT-64 have index values between 0.456 and 0.632. Tables 9 and 10 provide 
detailed rankings of the top 10% of the 0.10-mile road segments along US-191 and MT-64 with the highest 
Priority Composite Index values. The top 10% of 0.10-mile segments overall are along US-191 with the 
highest value at RM 11.1. The average Priority Composite Index value of all 921 0.10-mile segments is 
0.45 (SD = 0.17). 
 
32 
 
 
Figure 15 Priority Composite Indices for US-191 and MT-64 
 
33 
 
 
Table 9 Top 10% 0.10-mile Road Segments on US-191 for the Priority Composite Index 
(WMII*0.30) 
+ 
US-191 Wildlife Wildlife (WOII*0.10) Priority 
Road Movement Observation Suitability / WVC 
Importance Importance Connectivity Importance WMII * WOII * SCII * WVCII * + (SCII*0.30) Composite  
Mile + Index 
Reference Index Index Importance Index 0.30 0.10 0.30 0.30 
Index (SCII) (WVCII) (WVCII*0.30) Normal-
(WMII) (WOII) = Composite ized 
Index 
RM11.1 0.785 0.000 0.808 0.054 0.236 0.000 0.242 0.016 0.494 1.000 
RM23.7 0.576 0.193 0.919 0.000 0.173 0.019 0.276 0.000 0.468 0.945 
RM23.2 0.668 0.191 0.810 0.000 0.200 0.019 0.243 0.000 0.463 0.935 
RM46.4 0.951 0.637 0.309 0.054 0.285 0.064 0.093 0.016 0.458 0.925 
RM46.0 0.854 0.431 0.272 0.223 0.256 0.043 0.082 0.067 0.448 0.904 
RM24.8 0.644 0.772 0.579 0.000 0.193 0.077 0.174 0.000 0.444 0.896 
RM26.3 0.337 0.710 0.847 0.054 0.101 0.071 0.254 0.016 0.442 0.892 
RM7.6 0.595 0.000 0.470 0.393 0.179 0.000 0.141 0.118 0.437 0.882 
RM26.7 0.502 0.312 0.848 0.000 0.151 0.031 0.254 0.000 0.436 0.880 
RM46.3 0.902 0.426 0.351 0.054 0.271 0.043 0.105 0.016 0.435 0.877 
RM19.4 0.337 0.562 0.925 0.000 0.101 0.056 0.277 0.000 0.435 0.876 
RM36.4 0.551 0.433 0.717 0.027 0.165 0.043 0.215 0.008 0.432 0.870 
RM46.5 0.805 0.739 0.350 0.000 0.241 0.074 0.105 0.000 0.421 0.847 
RM48.8 0.507 0.313 0.782 0.000 0.152 0.031 0.235 0.000 0.418 0.842 
RM21.6 0.693 0.123 0.659 0.000 0.208 0.012 0.198 0.000 0.418 0.841 
RM3.1 0.951 0.119 0.389 0.000 0.285 0.012 0.117 0.000 0.414 0.834 
RM48.9 0.463 0.319 0.710 0.098 0.139 0.032 0.213 0.029 0.413 0.832 
RM22.4 0.483 0.530 0.714 0.000 0.145 0.053 0.214 0.000 0.412 0.830 
RM46.1 0.561 0.679 0.354 0.223 0.168 0.068 0.106 0.067 0.409 0.824 
RM34.9 0.434 0.132 0.779 0.098 0.130 0.013 0.234 0.029 0.406 0.818 
RM7.7 1.000 0.090 0.321 0.000 0.300 0.009 0.096 0.000 0.405 0.816 
RM12.4 0.380 0.000 0.970 0.000 0.114 0.000 0.291 0.000 0.405 0.816 
RM10.0 0.380 0.000 0.779 0.179 0.114 0.000 0.234 0.054 0.401 0.808 
RM34.6 0.390 0.138 0.899 0.000 0.117 0.014 0.270 0.000 0.400 0.806 
RM49.1 0.293 0.337 0.607 0.321 0.088 0.034 0.182 0.096 0.400 0.805 
RM28.2 0.644 0.200 0.591 0.027 0.193 0.020 0.177 0.008 0.398 0.801 
RM35.0 0.244 0.116 0.887 0.152 0.073 0.012 0.266 0.046 0.396 0.797 
RM25.8 0.502 0.102 0.783 0.000 0.151 0.010 0.235 0.000 0.396 0.796 
34 
 
(WMII*0.30) 
+ 
US-191 Wildlife Wildlife (WOII*0.10) Priority 
Road Movement Observation Suitability / WVC 
Connectivity Importance WMII * WOII * SCII * WVCII * + (SCII*0.30) Composite  
Mile Importance Importance Importance Index 0.30 0.10 0.30 0.30 + Index 
Reference Index Index Index (SCII) (WVCII) (WVCII*0.30) Normal-
(WMII) (WOII) = Composite ized 
Index 
RM47.8 0.366 0.374 0.729 0.098 0.110 0.037 0.219 0.029 0.395 0.795 
RM45.8 0.585 0.369 0.313 0.295 0.176 0.037 0.094 0.088 0.395 0.794 
RM48.4 0.073 0.157 0.521 0.670 0.022 0.016 0.156 0.201 0.395 0.794 
RM27.5 0.551 0.364 0.610 0.027 0.165 0.036 0.183 0.008 0.393 0.790 
RM49.0 0.220 0.563 0.620 0.277 0.066 0.056 0.186 0.083 0.391 0.786 
RM28.3 0.551 0.351 0.632 0.000 0.165 0.035 0.190 0.000 0.390 0.784 
RM35.3 0.478 0.127 0.769 0.000 0.143 0.013 0.231 0.000 0.387 0.778 
RM13.6 0.405 0.015 0.824 0.054 0.121 0.002 0.247 0.016 0.386 0.776 
RM48.7 0.230 0.431 0.803 0.107 0.069 0.043 0.241 0.032 0.385 0.774 
RM36.2 0.366 0.510 0.690 0.054 0.110 0.051 0.207 0.016 0.384 0.771 
RM7.2 0.571 0.000 0.607 0.098 0.171 0.000 0.182 0.029 0.383 0.769 
RM20.2 0.405 0.119 0.825 0.000 0.121 0.012 0.247 0.000 0.381 0.765 
RM46.7 0.561 1.000 0.375 0.000 0.168 0.100 0.113 0.000 0.381 0.765 
RM31.4 0.580 0.138 0.640 0.000 0.174 0.014 0.192 0.000 0.380 0.763 
RM16.2 0.454 0.171 0.753 0.000 0.136 0.017 0.226 0.000 0.379 0.761 
RM26.9 0.312 0.328 0.841 0.000 0.094 0.033 0.252 0.000 0.379 0.761 
RM16.6 0.478 0.015 0.774 0.000 0.143 0.002 0.232 0.000 0.377 0.757 
RM49.2 0.244 0.396 0.684 0.196 0.073 0.040 0.205 0.059 0.377 0.757 
RM45.3 0.415 0.309 0.583 0.152 0.124 0.031 0.175 0.046 0.376 0.754 
RM26.2 0.454 0.452 0.647 0.000 0.136 0.045 0.194 0.000 0.375 0.754 
RM1.4 0.761 0.051 0.375 0.098 0.228 0.005 0.112 0.029 0.375 0.754 
RM35.1 0.434 0.079 0.762 0.027 0.130 0.008 0.229 0.008 0.375 0.753 
RM23.5 0.361 0.314 0.781 0.000 0.108 0.031 0.234 0.000 0.374 0.751 
RM72.7 0.000 0.750 0.225 0.772 0.000 0.075 0.067 0.232 0.374 0.751 
RM42.0 0.361 0.135 0.786 0.054 0.108 0.013 0.236 0.016 0.374 0.750 
RM74.3 0.000 0.009 0.295 0.946 0.000 0.001 0.088 0.284 0.373 0.749 
RM24.1 0.215 0.241 0.948 0.000 0.064 0.024 0.284 0.000 0.373 0.748 
RM45.9 0.707 0.574 0.316 0.027 0.212 0.057 0.095 0.008 0.373 0.748 
RM7.9 0.761 0.000 0.247 0.223 0.228 0.000 0.074 0.067 0.369 0.741 
 
35 
 
(WMII*0.30) 
+ 
US-191 Wildlife Wildlife 
Movement Observation Suitability / WVC (WOII*0.10) Priority 
Road Connectivity Importance WMII * WOII * SCII * WVCII * + (SCII*0.30) Composite  
Mile Importance Importance 
Index Index Importance Index 0.30 0.10 0.30 0.30 + Index 
Reference Index (SCII) (WVCII) (WVCII*0.30) Normal-
(WMII) (WOII) = Composite ized 
Index 
RM45.5 0.390 0.334 0.545 0.179 0.117 0.033 0.164 0.054 0.368 0.737 
RM49.3 0.341 0.387 0.618 0.125 0.102 0.039 0.186 0.038 0.364 0.730 
RM46.6 0.683 0.390 0.298 0.098 0.205 0.039 0.089 0.029 0.363 0.728 
RM11.8 0.312 0.000 0.892 0.000 0.094 0.000 0.268 0.000 0.361 0.725 
RM46.9 0.366 0.348 0.495 0.223 0.110 0.035 0.148 0.067 0.360 0.722 
RM4.8 0.380 0.008 0.616 0.196 0.114 0.001 0.185 0.059 0.359 0.719 
RM7.0 0.380 0.000 0.562 0.250 0.114 0.000 0.169 0.075 0.358 0.717 
RM45.2 0.341 0.261 0.708 0.054 0.102 0.026 0.212 0.016 0.357 0.715 
RM10.5 0.190 0.000 0.993 0.000 0.057 0.000 0.298 0.000 0.355 0.711 
RM48.6 0.293 0.335 0.627 0.152 0.088 0.033 0.188 0.046 0.355 0.711 
RM34.0 0.532 0.158 0.594 0.000 0.160 0.016 0.178 0.000 0.354 0.708 
RM11.0 0.190 0.000 0.983 0.000 0.057 0.000 0.295 0.000 0.352 0.705 
RM35.4 0.288 0.024 0.876 0.000 0.086 0.002 0.263 0.000 0.352 0.704 
RM6.6 0.571 0.000 0.599 0.000 0.171 0.000 0.180 0.000 0.351 0.703 
RM12.7 0.405 0.015 0.758 0.000 0.121 0.002 0.227 0.000 0.350 0.702 
RM33.9 0.337 0.224 0.755 0.000 0.101 0.022 0.227 0.000 0.350 0.701 
RM48.5 0.268 0.169 0.684 0.152 0.080 0.017 0.205 0.046 0.348 0.697 
RM25.5 0.195 0.400 0.831 0.000 0.059 0.040 0.249 0.000 0.348 0.696 
RM79.0 0.000 0.000 0.158 1.000 0.000 0.000 0.047 0.300 0.347 0.695 
RM17.7 0.246 0.084 0.880 0.000 0.074 0.008 0.264 0.000 0.346 0.693 
RM28.0 0.385 0.171 0.658 0.054 0.116 0.017 0.197 0.016 0.346 0.693 
RM11.3 0.215 0.033 0.901 0.027 0.064 0.003 0.270 0.008 0.346 0.693 
RM29.5 0.268 0.372 0.681 0.076 0.080 0.037 0.204 0.023 0.345 0.690 
RM77.5 0.000 0.022 0.301 0.839 0.000 0.002 0.090 0.252 0.344 0.689 
RM36.5 0.244 0.606 0.674 0.027 0.073 0.061 0.202 0.008 0.344 0.689 
 
 
 
 
36 
 
Table 10 Top 10% 0.10-mile Road Segments on MT-64 for the Priority Composite Index 
Wildlife Wildlife Suitability / (WMII*0.30) + 
MT-64 Road Movement Observation Connectivit WVC (WOII*0.10) + Priority 
Mile Importanc Importance y Importance WMII WOII * SCII * WVCII (SCII*0.30) + Composite 
Reference e Index Index Importance Index * 0.30 0.10 0.30 * 0.30 (WVCII*0.30) Index 
(WMII) (WOII) Index (WVCII) = Composite Normalized 
(SCII) Index 
RM0.5 0.000 0.075 0.781 0.250 0.000 0.007 0.234 0.075 0.317 0.632 
RM0.3 0.000 0.151 0.809 0.196 0.000 0.015 0.243 0.059 0.317 0.632 
RM0.6 0.000 0.041 0.785 0.196 0.000 0.004 0.236 0.059 0.299 0.594 
RM1.0 0.000 0.011 0.673 0.250 0.000 0.001 0.202 0.075 0.278 0.551 
RM0.2 0.000 0.217 0.759 0.027 0.000 0.022 0.228 0.008 0.257 0.509 
RM1.2 0.000 0.000 0.437 0.420 0.000 0.000 0.131 0.126 0.257 0.508 
RM1.1 0.190 0.018 0.441 0.196 0.057 0.002 0.132 0.059 0.250 0.494 
RM1.3 0.000 0.000 0.506 0.321 0.000 0.000 0.152 0.096 0.248 0.490 
RM7.4 0.190 0.000 0.625 0.000 0.057 0.000 0.187 0.000 0.244 0.482 
RM0.9 0.000 0.030 0.649 0.125 0.000 0.003 0.195 0.038 0.235 0.463 
 
 
37 
 
 
3.6. Identifying Areas for Field Evaluation 
 
Figure 16 A view of early road configuration within the Gallatin Gateway to Spanish Creek Priority Site looking 
south to Gallatin Canyon at the beginning of the 1900s, and in 2016. Credits: Historic photo-Museum of the 
Rockies; Recent photo- Duncan Patten. 
 
We used a MWA to smooth the results of the Priority Composite Index described in the previous section 
to identify areas of highway with consistently elevated values for field evaluation. The general rationale 
for using a MWA approach is three-fold: (1) averaging helps to lessen the effects of any spatial error in the 
recording of specific locations (e.g., WVCs); (2) the 0.1-mile segments used as units of analysis are small 
and the counts for wildlife crossings, crash, carcass, or observations for any single segment are more likely 
to be influenced by statistical sampling noise than are the counts for a group of consecutive segments; and 
(3) from a mitigation perspective, it can be useful to identify longer stretches with elevated values (e.g., 
high WVCs) than to identify short individual segments with the highest WVCs (e.g., when considering 
mitigation measures designed to reduce WVCs at more than a point location, such as fencing). 
The top 10% of all 921 0.1-mile segments identified by the MWA on US-191 and MT-64 are clustered in 
seven locations: six on US-191 and one on MT-64 (Figure 17). The top 10% MWA of the 820 segments 
analyzed on US-191 have index values between 0.624 and 0.797, while the top 10% MWA of the 121 
segments evaluated on MT-64 have index values between 0.406 and 0.500. These seven locations are: 
US-191 
 Teepee Creek from RM 10.4 to RM 11.6 
 Bacon Rind Creek from RM 23.2 to RM 24.0 
 Specimen Creek from RM 25.8 to RM 27.1 
 Taylor Fork from RM 34.2 to RM 35.5 + RM 36.4 
 Porcupine Creek from RM 45.4 to RM 47.0 
 North of Big Sky from RM 48.1 to RM 49.4 
 
MT-64 
 West Fork Gallatin on MT-64 from RM 0.2 to RM 1.2 
 
38 
 
Roughly half of the top 10% segments with the highest Priority Composite Index values (see Tables 9 and 
10) are included in the locations identified by the MWA for field evaluation. While many of the highest 
individual 0.10-mile segments are omitted, a number of these segments show localized spikes not 
surrounded by other high-value segments. 
Along with the seven locations identified via the MWA, we identified three additional areas for field 
evaluation. We included two areas on the northern end of the study area with the highest WVC risk to 
capture the important human safety concerns in those locations (Four Corners to Gallatin Gateway and 
Gallatin Gateway to Spanish Creek), as well as one area on MT-64 based on habitat suitability/connectivity. 
These three locations are: 
US-191 
 Gallatin Gateway to Spanish Creek along US-191 from RM 68.1 to RM 73.7 
 Four Corners to Gallatin Gateway along US-191 from RM 74.1 to RM 81.3 
MT-64 
 Upper Big Sky Connectivity Area on MT-64 from RM 7.3 to RM 8.2 
 
 
39 
 
 
Figure 17 Priority Road Areas for Field Evaluations 
 
40 
 
During field visits, the CLLC-WTI research team and Technical Advisory Committee (TAC) expanded 3 
of these locations to include additional segments among the top 10% of Priority Composite Index values. 
The Teepee Creek site was expanded southward to incorporate RM 9.5 to RM 10.3. The Bacon Rind and 
Specimen Creek sites were combined and expanded to incorporate RM 24.1 to RM 25.7. Following the 
field visits, the Porcupine Creek site was extended south to RM 43 based on additional conversations with 
area biologists from MDT and FWP (Deb Wambach, Butte District Biologist, MDT; Julie Cunningham, 
Bozeman Area Biologist, FWP, pers. comms.). Detailed maps and descriptions of each of the expanded 
sites are in Chapter 6. 
  
 
41 
 
4. Wildlife Accommodation Measures 
A wildlife overpass was first installed in France in the 1950s, with Utah installing the first similar structure 
in the U.S. in 1975 on I-15 near Beaver, followed by additional construction of dedicated crossings for 
wildlife in New Jersey and other states. Today, many states have installed wildlife crossing structures, with 
the majority allowing animals to cross beneath the road surface via specially designed culverts, 
underpasses, and bridges rather than more visible wildlife overpasses. Dozens of wildlife crossing 
structures have been installed along US-93 in western Montana, including an overpass. While structures 
usually serve multiple species, they must be located and designed with appropriate characteristics to serve 
target species and based on data about crossing locations. 
The wildlife accommodation measures considered in this report fall into two overarching categories: a) 
measures aimed at influencing driver behavior and b) measures to make roads more permeable to wildlife. 
A discussion of both categories follows. While some measures to influence driver behavior may reduce the 
risk of wildlife-vehicle collisions, with varying degrees of success, this category of measures generally 
fails to address the barrier effect of roads on wildlife movement, a key concern of this study. The latter 
category, in which wildlife is separated from traffic by fencing and crossing structures designed to enable 
safe wildlife passage, achieves the dual objectives of reducing wildlife-vehicle collisions and maintaining 
habitat connectivity. 
A 2021 Pooled Fund Study supported by the Departments of Transportation of nine states in cooperation 
with the Federal Highway Administration reviewed the literature on known accommodation measures for 
both large and small (coyote or smaller) animals to determine effectiveness. The information below and 
summarized in Table 11 is derived from this review by Huijser et al. (2021). 
Table 11 Effectiveness of Mitigation Measures 
Effectiveness in 
Mitigation Measure Reducing Collisions Effectiveness in Reducing the 
with Large Mammals Barrier Effect of Roads and Traffic 
Measures aimed at influencing driver behavior 
Seasonal wildlife warning signs 9-50% None 
Animal detection systems 33-97% None 
Seasonal road closure 100% during closure Reduces barrier effect of traffic, not the 
road, during closure 
Increase visibility for the driver 57-68% None, may increase barrier effect for 
some species 
Reduced speed with traffic 
calming measures Unknown-59% Unknown 
Measures to separate wildlife from the road and traffic 
Barriers: fences, boulders, walls 80-100% None, increases barrier effect 
Underpasses and overpasses Varies greatly based on 
without fencing structural design and Reduces barrier effect 
location 
Underpasses and overpasses 
with fencing 80-100% Reduces barrier effect 
(Adapted from Huijser et al. 2021) 
 
 
42 
 
4.1. Measures to Influence Driver Behavior 
4.1.1. Permanent or Temporary Warning Signs 
While commonly implemented, permanent warning signs have been found to reduce collisions with 
wildlife only for a short period of time after their installation. They are generally considered to be 
ineffective, in spite of the fact that they may increase driver awareness or cause temporary speed reduction 
(Pojar et al. 1975; Coulson 1982; Al-Ghamdi 2004; Sullivan et al. 2004; Meyer 2006; Bullock, Malan, and 
Pretorius 2011). Researchers believe they may help increase awareness of collision danger, yet may also 
falsely suggest that signs alone are an effective tool. 
In contrast, while studies are limited, temporary warning signs may reduce collisions. Effectiveness, 
however, varies substantially (9–50%) (Sullivan et al. 2004; Colorado Department of Transportation 2014). 
It appears that seasonal or other types of enhanced warning signs increase in effectiveness when used in 
increasingly precise locations during specific periods of high risk. However, as enhanced wildlife warning 
signs are often applied over long road sections, and not limited to periods of high risk, their effectiveness 
may be limited in practice (Huijser et al. 2015). 
4.1.2. Animal Detection Systems 
Like temporary warning signs, animal detection systems can reduce wildlife-vehicle collisions with large 
mammals. Collisions may be reduced by as much as 33-97% when sensors are able to reliably detect the 
target species (Mosler-Berger and Romer 2003; Dai, Young, and Vander Giessen 2009; Gagnon et al. 2010; 
Strein 2010; Minnesota Department of Transportation 2011; Sharafsaleh et al. 2012; Huijser, Gunson, and 
Abrams 2006).  It is important to note that animal detection systems may be most appropriate on low-
volume roads (such as less than 5,000 vehicles per day) to limit the likelihood of rear-end collisions when 
vehicles brake suddenly. Further, vehicle speed may have to be reduced substantially (for example, to 35-
40 miles per hour) to reduce the likelihood of a collision (Huijser et al. 2015).  Animal detection systems 
are still considered experimental, especially as a stand-alone measure on open roads over longer distances; 
technology, management, and maintenance issues hamper their effectiveness. Animal detection systems 
that have had the highest efficacy are those that have been used in combination with other measures and 
which have a very discreet detection zone, such as at the end of wildlife exclusion fencing or at an animal 
“crosswalk” or gap in fencing (Gagnon et al. 2010). 
4.1.3. Reduced Speed Limit 
Speed management is often suggested as a strategy to reduce wildlife-vehicle collisions. However, unless 
the design speed of the road is also reduced, posting a lower speed limit may lead to more dangerous 
driving conditions due to a mix of slower and faster drivers who continue to follow a road’s ‘design speed’ 
rather than the speed limit. This phenomenon, called “speed dispersion” (a mix of fast- and slow-moving 
vehicles), is known to increase crashes (Elvik 2014; Huang et al. 2013).  Regardless of the speed limit, 
most drivers operate their vehicles at a rate that is near to or higher than a road’s design speed (Fitzpatrick 
2003; Ouyang, Jiang, and Jadaan 2016; Donnell, Kersavage, and Tierney 2018). 
Further, even with substantial enforcement, lowered speed limits do not change driver behavior sufficiently 
to reduce collisions with wildlife. For example, despite a 43% increase in citations issued over two years 
following the reduction of nighttime speed limits (to 55 miles/hr) in marked wildlife crossing zones in 
Colorado, collisions with wildlife increased in nearly half of the areas studied (Colorado Department of 
Transportation 2014). Drivers continued to exceed the posted limit by 11 miles/hr (Colorado Department 
of Transportation 2014). Similarly, in a Wyoming study in which nighttime speed limits were lowered by 
15 miles/hour during times of the year with the greatest mule deer collision risk, drivers reduced their speed 
by only 3-5 miles/hr (Riginos et al. 2019). 
 
43 
 
Finally, because most collisions occur at dawn and dusk, highway speeds would need to be considerably 
slower than is common for drivers whose headlights have median reach to avoid a collision with a large 
mammal. An operating speed of only 40 miles/hr is still insufficient to allow one-half of drivers to stop a 
vehicle in time to avoid a collision (Huijser, Fairbank, and Abra 2017). To allow nearly all drivers to stop 
in time, the nighttime operating speed of a road needs to be as low as 25-30 miles/hr, far lower than the 
design speed of most roads (Huijser, Fairbank, and Abra 2017; Huijser et al. 2015). 
4.1.4. Highway Lighting 
Highway lighting may reduce wildlife-vehicle collisions by 57-68% (McDonald 1991; Riley and Marcoux 
2006; Wanvik 2009). It is unclear, however, if collision reductions along lighted highways are because of 
increased visibility of the animals to drivers or because animals avoid lighted highways. 
It is likely that highway lighting contributes to the barrier effect of roads for some species. While a 
reasonable tool to consider in high-risk areas to reduce wildlife-vehicle collisions, highway lighting is less 
appropriate where habitat connectivity is an objective. Lights can be expected to increase the barrier effect 
of roads for species that are light-avoidant, such as nocturnal species. At the same time, other species may 
be attracted to lights and experience a higher risk of road mortality. Since highway lighting can affect 
animal physiology and behavior, and even predation rates, its use may require additional measures to 
reduce effects on wildlife (Blackwell, DeVault, and Seamans 2015). Limiting highway lighting to periods 
when vehicles approach can reduce the negative impacts of lighting on wildlife, such as restricting lighted 
areas to those that increase drivers’ sight distance on the road and roadside. 
4.2. Measures to Separate Wildlife from Traffic 
4.2.1. Wildlife Crossing Structures 
The main objective of wildlife crossing structures is to connect wildlife populations or entire ecosystems 
and allow ecosystem processes to continue over or under a road. These structures allow for safe daily, 
seasonal, or dispersal movements between areas on either side of a highway (Dodd et al. 2007; Gagnon et 
al. 2010; Huijser et al. 2015; Huijser, Fairbank, et al. 2016). Wildlife use of crossing structures increases 
when the structures are connected to fencing (see Wildlife Fencing, below), which also serves to reduce 
wildlife-vehicle collisions. 
4.2.2. Wildlife Underpasses 
Wildlife underpasses are designed to allow wildlife to cross safely under the road. When designed for large 
mammals, they can also successfully allow passage of smaller species. Designing an underpass and its 
approaches with appropriate vegetative cover encourages use by target species and also makes smaller 
species feel more secure in its use. Underpasses can also be adapted to support use by amphibian and semi-
aquatic species and to accommodate water flow. 
Often, existing bridges or culverts that span water bodies can be re-designed to be longer and/or wider to 
make space for riparian habitat alongside a river or stream beneath a bridge or within a culvert, allowing 
for use by more species. As these structures are often located in riparian areas that are attractive to terrestrial 
wildlife, appropriately designed culverts and bridges can support the movement of both aquatic and 
terrestrial species. 
Numerous types of underpass structures exist, including open-span bridges, concrete bottomless arches, 
corrugated steel arches, and box culverts. Because these structures are “bottomless,” each allows for the 
natural substrate of the area to be used within the crossing structure. Dimensions vary greatly, depending 
on the target species, local terrain, and other parameters. 
 
44 
 
4.2.3. Wildlife Overpasses 
Due to their size and visibility to drivers and the public, wildlife overpasses may be the most commonly 
known type of wildlife accommodation measure and are typically designed to allow movement by a suite 
of large animals. By including specific design elements, however, they can also attract small- and medium-
sized mammals, as well as amphibians, reptiles, semi-aquatic species, ground-dwelling birds, and 
butterflies. Some species, such as elk, moose, pronghorn, and grizzly bears, have demonstrated a preference 
for overpasses in certain locales. Family groups of species like grizzly bears prefer larger open structures 
like overpasses but may also use large-span bridges as a wildlife underpass (Clevenger and Barrueto 2014; 
Ford, Barrueto, and Clevenger 2017; Sawyer, Rodgers, and Hart 2016). 
4.2.4. Wildlife Fencing (in combination with Wildlife Crossing Structures) 
On average, an 87 percent reduction in wildlife-vehicle collisions can be expected from fencing when 
combined with wildlife crossing structures (Huijser 2008). Because fencing itself creates a barrier, it is not 
a solution to wildlife connectivity but rather is intended to guide animals to crossing structures. 
Wildlife fencing is commonly constructed at a height of 8 ft (2.4 m). Wildlife fencing is typically placed 
at the edge of a department of transportation’s right-of-way, or at least outside of the clear zone of the 
highway, so it does not interfere with operations such as snow plowing. 
Effective fencing may be continuous between safe passage opportunities (Clevenger and Barrueto 2014) 
or partial (i.e., disjunct fenced segments of highway with numerous fence ends) (Gagnon et al. 2010; 
Huijser et al. 2016; Huijser, Camel-Means, et al. 2016). Fencing should include escape ramps or “jump-
outs,” which allow wildlife trapped on the highway side of a fence to jump to safety outside the fenced 
section. The height of a jump-out should be 4-6 ft (1.2-1.8 m) above the outside surface to deter wildlife 
from jumping up and entering a roadway. Fencing may also need to include climbing or digging barriers 
to be effective for species adept at climbing or digging. Fencing usually needs to extend 1.5 miles on either 
side of a crossing structure or adjacent to a series of crossing structures to be most effective. Fence-end 
treatments may be necessary where fencing terminates in an area with wildlife movement potential. Fence-
end treatments can include cattle guards, electrified mats, boulder fields, or natural landscape features, such 
as a cliff or other feature that acts as a barrier. Animal detection systems can also be used at fence ends to 
warn drivers when an animal is approaching or crossing the road. 
 
 
  
 
45 
 
5. Field Evaluation and Ranking of Priority Sites 
Representatives of the Center for Large Landscape Conservation (CLLC) and Western Transportation 
Institute (WTI) research team and an interdisciplinary Technical Advisory Committee (TAC) of biologists, 
transportation ecologists, engineers, and land-use planners visited each of the 10 highway sites identified 
through spatial analysis on October 12-13, 2022, in order to evaluate their potential for wildlife 
accommodations. The following individuals participated in the field evaluation: 
CLLC-WTI Research Team Representatives 
• Rob Ament - Western Transportation Institute, MSU 
• Marcel Huijser - Western Transportation Institute, MSU 
• Damon Fick - Western Transportation Institute, MSU 
• Matthew Bell - Western Transportation Institute, MSU 
• Elizabeth Fairbank - Center for Large Landscape Conservation 
• Abigail Breuer - Center for Large Landscape Conservation 
• Braden Hance - Center for Large Landscape Conservation 
Technical Advisory Committee 
• Deb Wambach - Montana Department of Transportation 
• Dave Gates - Montana Department of Transportation 
• Mike McGrath - U.S. Fish and Wildlife Service 
• Jeff Patten - Federal Highway Administration 
• Randy Scarlett - U.S. Forest Service 
• Kyle Meakins - National Park Service 
• Doug Madsen - National Park Service 
• Sean O’Callaghan - Gallatin County 
• Frank Van Manen - Interagency Grizzly Bear Study Team 
The research team and TAC evaluated and ranked each site using a Field Evaluation Matrix (see Section 
5.1) to assess WVC risk, conservation value, mitigation options, barrier effect of the road corridor, 
vulnerability to future change, and land security. Many sites include major drainages from surrounding 
public lands that intersect with US-191 or MT-64 and feature existing infrastructure that has the potential 
to facilitate animal movements, such as a bridge spanning a riparian corridor. The research team and TAC 
evaluated existing infrastructure for its (1) current ability to function as a wildlife crossing, (2) potential 
for retrofitting to provide safe crossing opportunity, and (3) potential for replacement to provide safe 
crossing opportunity where retrofitting is not possible. The construction of purpose-built structures for 
wildlife crossing in areas of high-quality habitat is important for wildlife movement, as is the feasibility or 
advisability of adding wildlife exclusion fencing to new or existing structures. Wildlife crossing structures 
combined with fencing are the most effective way to reduce collisions with wildlife while maintaining 
habitat connectivity (Huijser et al. 2021). The research team and TAC also considered other alternatives to 
reduce collisions with wildlife that can be effective (see Table 11), including: animal detection systems 
that warn drivers when wildlife is on the road, variable message signs for areas that have spatially discreet 
or seasonal conflicts, and traffic calming measures to effectively reduce the design speed of the highway. 
The Field Evaluation Matrix used to rank each site provides a score from 1 (low priority) to 5 (high priority) 
for each of 9 criteria: 
1. Wildlife-vehicle collision risk: Frequency of collisions with wildlife. 
2. Wildlife crossing roads: Intensity of wildlife crossing roads. 
 
46 
 
3. Live wildlife near roads: Intensity of wildlife use of roadside environments. 
4. Regional conservation value: Contribution to regional conservation (if mitigated) by serving as 
a movement corridor or high-quality wildlife habitat at the regional scale. 
5. Land security: Presence/absence of “secured” land (e.g., state or federal land or private holdings 
with voluntary conservation easements) on both sides of the road to allow for effective crossing 
structures. 
6. Local conservation value: Contribution to regional conservation (if mitigated) by serving as a 
movement corridor or high-quality wildlife habitat at the local scale. 
7. Mitigation options: Type and engineering feasibility of mitigation measures that could be 
implemented. 
8. Barrier effect: Degree of negative impact on wildlife movement potential due to high traffic 
volume, non-wildlife-friendly fencing, or other adjacent linear features. 
9. Vulnerability: Potential for future increase in WVC risk or negative impact on wildlife due to 
increased speed limit, traffic volume, road width, or number of lanes 
Criteria 1-5 were scored based on the spatial data analysis, with the research team representatives and TAC 
confirming or adjusting the scores by consensus based on local conditions during field evaluation.  Criteria 
6-9 were scored by consensus in the field. At the conclusion of the site visits, each site received an overall 
score (from 1 to 5) by averaging the scores of each of the 9 criteria. 
Criteria 1-4 were assigned a value of 1 (low priority) to 5 (high priority) based on mean percentiles 
calculated from the road segments within each priority site. These percentile categories are shown in Table 
12. 
Table 12 Categories Used to Score Criteria 1-4 based on the Spatial Data Analysis. 
Score Percentile Categories Description 
5 95-100% The 5% of road segments with the highest index values 
4 75-94.9% The next 20% of road segments with the highest index values 
3 50-74.9% The next 25% of road segments with the highest index values 
2 25-49.9% The next 25% of road segments with the highest index values 
1 <24.9% The 25% of road segments with the lowest index values 
 
Using the following rubric, Criteria 5 on land security was also scored prior to field evaluation, with the 
Research team and TAC adjusting the score if conditions encountered during the field visits substantiated 
doing so: 
1. Housing or industrial/commercial development on both sides of site (or adjacent/nearby) 
2. Housing or industrial/commercial development on one side of the site, privately owned open space 
on the other side (no conservation easement). 
3. Privately owned open space lands on both sides (no conservation easement). 
4. Public lands (federal, state, or tribal) or private land with a voluntary conservation easement on 
one side of the site, open space on the other side (no conservation easement) 
5. Public lands (federal, state, or tribal) or private lands with conservation easements on both sides 
The research team and TAC assigned Criteria 6-9 a value from 1 (low priority) to 5 (high priority) during 
field visits based on collective, expert understanding through a consensus process. Scores for all criteria 
(Criteria 1-9) evaluated are shown in Table 13 below. 
 
47 
 
Table 13 Summary of Scores and Ranking of each Priority Site following Field Evaluation. 
Live 
WVC Wildlife Wildlife Regional Land Local Wildlife 
Priority Site Barrier Overall Priority 
Risk Crossing Near Conservation 
Road Value Security Conservation Accommodation Effect Vulnerability Average Site 
Road Value Options Score Rank 
Four Corners to Gallatin 
Gateway 4 1 1 1 1 3 2 5 4 2.40 6 
Gallatin Gateway to 
Spanish Creek 3 1 1 2 4 5 4 4 4 3.11 1 
N. of Big Sky Entrance 1 2 2 3 3 3 4 4 3 2.78 2 
Porcupine Creek 1 3 2 2 2 2 4 4 5 2.78 2 
Taylor Fork 1 2 1 4 3 3 2 3 3 2.44 5 
Specimen Creek to Bacon 
Rind Creek 1 2 2 4 5 5 2 1 2 2.67 3 
Teepee Creek 1 1 1 4 5 4 3 1 2 2.44 5 
West Fork Gallatin 1 1 1 3 4 3 4 3 3 2.56 4 
Upper Big Sky 1 1 1 3 2 4 1 1 3 1.89 7 
         Note: Two sites tied at each of Rank 2 and Rank 5
 
48 
 
 
5.1.  Field Evaluation Matrix Form 
Highway 191: Field Evaluation Matrix 
Site ID:  
AADT: 
Site Description:  _____________________________________________________________________  
 ___________________________________________________________________________________  
 ___________________________________________________________________________________  
 Overall Composite Score: Max                , Min                 , Avg                . 
 Individual Evaluation Criteria Scores: 
WVC risk: Max                , Min                 , Avg                . 
Wildlife crossing road: Max                , Min                 , Avg                . 
 Live wildlife near road: Max                , Min                 , Avg                . 
Scenario Description Score Evaluation method 
WVC risk Frequency of collisions with wildlife  Data analysis 
Wildlife Crossing 
Road Intensity of wildlife crossing the road  Data analysis 
Live wildlife near 
roads Intensity of wildlife use of roadside environments  Data analysis 
Regional Contribution to regional conservation (if mitigated) by 
conservation serving as a movement corridor or high-quality wildlife  Data analysis 
value habitat at the regional scale 
Presence/absence of “secured” land (e.g., state, federal, 
Land security private conservation easement) on both sides of road to  Data Analysis/Field 
allow for effective crossing structures 
Local Contribution to regional conservation (if mitigated) by 
conservation serving as a movement corridor or high-quality wildlife  Field 
value habitat at the local scale 
Wildlife 
accommodation Type and engineering feasibility of wildlife 
options accommodations that could be implemented  Field 
Degree of negative impact on wildlife movement potential 
Barrier effect due to high traffic volume, non-wildlife-friendly fencing,  Field 
adjacent linear features, etc. 
Vulnerability Potential for future increase in WVC risk or negative  
impact to wildlife due to increased speed limit, traffic 
volume, road width, number of lanes, or other development Field 
pressure 
Overall Average Take the average from all scenarios  
Score Field 
 Regional conservation value: Max                , Min                 , Avg                . 
 
6. Priority Sites and Recommendations 
6.1. Teepee Creek 
 TEEPEE CREEK: US Highway 191 (US-191), RM 9.5-11.6 
AADT: 2,509 
Priority Rank: 5 (tied with Taylor Fork Site) 
Table 14 Maximum, Minimum, and Average Index Values for each Prioritization Characteristic and Composite 
Index of all 0.10-mile Road Segments within the Teepee Creek Priority Site 
Prioritization Characteristic Maximum Index 
Value Minimum Index Value Average Index Value 
Composite (overall) 1.0 0.545 0.661 
WVC Risk 0.074 0.0 0.016 
Wildlife Crossing Road 0.785 0.0 0.205 
Live Wildlife Near Road 0.043 0.0 0.006 
Regional Conservation Value 0.993 0.731 0.886 
 
Table 15 Teepee Creek Field Evaluation Scores and Priority Ranking 
Live 
Priority WVC Wildlife Regional 
Crossing Wildlife Conservation Land Local Wildlife Overall 
Site Risk Conservation Accommodation Barrier Vulnerability Average Priority 
Road Near Security 
Road Value Value Options Effect Score Rank 
Teepee 
Creek 1 1 1 4 5 4 3 1 2 2.44 5 
 
The Teepee Creek Priority Site is surrounded by public lands, including Yellowstone National Park (YNP) 
and Custer Gallatin National Forest (CGNF). The site includes a mix of forest and wetlands and provides 
important habitat for a variety of species, including elk, moose, and grizzly bears. Upon visiting the site, 
the research team and TAC extended it south to include the Grayling Creek drainage as well as Fir Ridge, 
which are important for wildlife movement between YNP and CGNF. A highway bridge runs over the 
Grayling Creek drainage, along with a snowmobile bridge that runs parallel just west of the road. The 
highway bridge provides some opportunity for terrestrial wildlife passage, especially during periods of low 
flow. To allow for terrestrial wildlife passage most of the year, the highway bridge should be expanded to 
include the stream banks on either side and sufficient vertical clearance for large mammals (>15 ft or 4.6 
m) (Clevenger and Huijser 2011). This expansion could occur at the end of its service life when 
replacement is scheduled or earlier. The snowmobile bridge would need to be expanded at the same time, 
especially if located within the fenced road corridor. 
The existing double-pipe culvert at Teepee Creek is perched (Figure 18) above the surrounding landscape 
and is not passable by most wildlife. Teepee Creek is surrounded by wetlands and non-wetland riparian 
habitat in this area (RM 11-11.1) and has some of the highest values of the composite and habitat 
connectivity/suitability indices for elk, grizzly bears, bighorn sheep, wolverine, and boreal toad, and 
wildlife movement for elk and grizzly bears (Table 16). Three grizzly bear mortalities have been 
documented within this site, in addition to moose, elk, and deer mortalities, and many smaller species, 
including coyotes, foxes, pine martens, beaver, and porcupines. Few of the wildlife mortalities recorded 
by the National Park Service have been reported as vehicle crashes. Especially in the case of larger-bodied 
species, this may indicate that the mortalities were caused by large vehicles like semi-trucks, which are 
less likely to sustain damage or report accidents with wildlife (Huijser and Begley 2019; Abra et al., 2019). 
 
50 
 
While reported WVC numbers are not high in this location compared to some others, as traffic volumes 
increase, WVCs may increase. Present traffic volume is relatively low, at just over 2,500 AADT, such that 
wildlife can cross the highway safely during some parts of the day or year by taking advantage of times 
when traffic volume is lesser (Riginos 2022; Riginos et al. 2018). However, as traffic volume continues to 
increase due to YNP visitation, local population growth, and recreation on public lands, this is an important 
area to implement measures to mitigate WVCs and to maintain habitat connectivity for multiple species 
between YNP and CGNF. 
Near the Teepee Creek culvert, the area has the highest composite score (1.0) of any of the 0.10-mile 
segments across all priority sites (see Figure 17). Ultimately, the culvert at Teepee Creek should be replaced 
with a structure that has sufficient vertical clearance to provide suitable safe crossing opportunities for elk, 
moose, and grizzly bears; this would require >15 ft (4.6 m) vertical clearance above stream banks 
(Clevenger and Huijser 2011). An optimized structure design would span all, or the majority of, the 
wetlands and riparian habitat surrounding Teepee Creek. Nonetheless, a structure of this type is unlikely 
to accommodate movements by grizzly bear family groups (i.e., sows with cubs), which strongly prefer 
overpasses or large-span bridges (Ford, Barrueto, and Clevenger 2017). If a goal is to reduce WVCs in the 
area while maximizing crossing structure use, wildlife-exclusionary fencing should be used and extended 
north and south of the structure. Such fencing may be tied into the Grayling Creek bridge (Dodd et al. 
2007; Gagnon et al. 2010). 
Key Next Steps: 
• Grayling Creek Bridge should be expanded to include dry terrestrial passage opportunities beneath 
it along both sides of the creek year-round when the bridge reaches the end of its service life or 
earlier. The adjacent snowmobile bridge should be expanded similarly. 
• The double-pipe culverts at Teepee Creek should be replaced with a structure large enough for use 
by elk, moose, and grizzly bears and should span the majority of the wetland and riparian area. 
• Once these structures have been upgraded to meet the needs of the target species, they could be 
connected by fencing that may extend as far south as Fir Ridge in order to reduce WVCs while 
providing connectivity. 
 
Figure 18 Double Culverts Facilitating Water Flow from Teepee Creek 
 
 
51 
 
Table 16 Index Values of all 0.10-mile Segments within the Teepee Creek Priority Site 
US-191 Wildlife Wildlife Suitability / 
Movement Observation Connectivity WVC Risk Composite  Moving 
Road Mile 
Reference Importance Importance Importance Importance Importance Window 
Index Index Index Index Index Average 
9.5 0.571 0.024 0.448 0.100 0.676 0.413 
9.6 0.214 0 0.526 0.026 0.453 0.423 
9.7 0.190 0 0.517 0 0.414 0.450 
9.8 0 0 0.439 0 0.248 0.485 
9.9 0.570 0.002 0.518 0.026 0.670 0.518 
10 0.380 0 0.778 0.178 0.807 0.565 
10.1 0 0 0.688 0 0.403 0.561 
10.2 0.190 0 0.655 0.073 0.547 0.577 
10.3 0.190 0.005 0.833 0.026 0.629 0.589 
10.4 0.190 0 0.901 0 0.654 0.625 
10.5 0.190 0 0.993 0 0.711 0.628 
10.6 0.190 0 0.882 0 0.642 0.646 
10.7 0.214 0 0.828 0 0.624 0.662 
10.8 0 0.005 0.923 0 0.550 0.676 
10.9 0.263 0 0.730 0.073 0.639 0.668 
11 0.190 0 0.983 0 0.705 0.663 
11.1 0.785 0 0.807 0.053 1 0.661 
11.2 0 0.043 0.916 0.053 0.587 0.651 
11.3 0.214 0.033 0.900 0.026 0.692 0.660 
11.4 0.024 0 0.891 0 0.544 0.665 
11.5 0.190 0 0.820 0 0.603 0.657 
11.6 0.214 0 0.932 0 0.688 0.641 
  
 
52 
 
 
Figure 19 Teepee Creek Priority Site Map 
 
53 
 
6.2. Specimen Creek to Bacon Rind Creek 
 SPECIMEN CREEK TO BACON RIND CREEK: US Highway 191 (US-191), RM 23.2-27.1 
AADT: 2509 
Priority Rank: 3 
Table 17 Maximum, Minimum, and Average Index Values for each Prioritization Characteristic and Composite 
Index of all 0.10-mile Road Segments within the Specimen Creek to Bacon Rind Creek Priority Site 
Prioritization Maximum Index Minimum Index Average Index 
Characteristic Value Value Value 
Composite (overall) 0.945 0.38 0.630 
WVC Risk 0.054 0 0.010 
Wildlife Crossing Road 0.668 0.024 0.240 
Live Wildlife Near Road 0.772 0.009 0.261 
Regional Conservation Value 0.948 0.473 0.714 
 
Table 18 Specimen Creek to Bacon Rind Creek Field Evaluation Scores and Priority Ranking 
WVC Wildlife Live 
Priority Site Crossing Wildlife Regional Local Wildlife Overall 
Risk Near Conservation Land Conservation Accommodation Barrier Vulnerability Average Priority 
Road Security Effect Rank 
Road Value Value Options Score 
Specimen 
Creek-
Bacon 1 2 2 4 5 5 2 1 2 2.67 3 
Rind 
Creek 
 
The Specimen Creek to Bacon Rind Creek Priority Site was initially identified as two shorter stretches of 
highway that fell into the top 10% MWA overall. During site visits, the research team and TAC decided to 
combine them into a single site due to their proximity and the fact that they follow continuous riparian 
meadows along the Gallatin River and US-191 highway corridor. This site occurs within the boundary of 
Yellowstone National Park (YNP), and the public lands on either side of the highway connect YNP to the 
Custer Gallatin National Forest (CGNF) and Lee Metcalf Wilderness Area. It is characterized by open 
riparian meadows along the roadside and river corridor, which provide forage and water for wildlife; lower 
elevation areas are surrounded by forested slopes. This area is very important from both a regional and 
local conservation perspective, with frequent crossings by elk and grizzly bears documented by GPS. In 
recent years, numerous elk have been hit in this section of highway, along with moose, bighorn sheep, deer, 
black bears, wolves, coyotes, foxes, pine martens, and beavers. Grizzly bear road mortalities have also 
been documented approximately one mile to the south. Almost none of these mortalities were reported as 
crashes. Especially in the case of larger-bodied species, this may indicate that the mortalities were caused 
by large vehicles like semi-trucks, which are less likely to sustain damage or report collisions with wildlife 
(Huijser and Begley 2019; Abra et al. 2019). 
While reported WVC numbers are not high in this location compared to some others, as traffic volumes 
increase, WVCs may increase. Present traffic volume is relatively low, at just over 2,500 AADT, such that 
wildlife can cross the highway safely during some parts of the day or year by taking advantage of times 
when traffic volume is lesser (Riginos 2022; Riginos et al. 2018). However, as traffic volume increases 
due to YNP visitation, local population growth, and recreation on public lands, this is an important area for 
measures to mitigate WVCs in order to maintain habitat connectivity between YNP and CGNF for multiple 
species. 
 
54 
 
There are two bridges within this priority site: one over the Gallatin River (between RM 23 and 24) and 
one over Specimen Creek (between RM 26 and 27). These bridges have low vertical clearance and are not 
suitable for species like elk, moose, or grizzly bears. Both bridges also have little streambank (horizontal) 
clearance, with abutments and riprap making passage by ungulates difficult, especially during higher flows 
(see Figure 20). These bridges should be replaced by larger structures with higher vertical clearance (>15 
ft or 4.6 m above stream banks) and span the full width of the stream banks along Specimen Creek and the 
Gallatin River to allow for safe passage (Clevenger and Huijser 2011). Replacement could happen at the 
end of their service life. However, with an increase in traffic volume or WVCs, the bridges may need to be 
replaced to accommodate wildlife movement before that time. 
Because this priority site is within YNP, concerns exist over the aesthetics of fencing. While appropriately 
sized crossing structures without fencing may be used opportunistically by wildlife, without fencing, 
structures will not prevent animals from crossing the road at-grade. If WVCs continue to increase, it would 
likely be necessary to add fencing in addition to improved bridges to achieve the dual goals of maintaining 
connectivity and reducing WVCs. Fencing becomes a more urgent consideration in areas where threatened 
or endangered species, such as grizzly bears, are killed by vehicles. Fencing might also become more urgent 
if WVCs involving large ungulates (e.g., moose, bison, elk) become more frequent, increasing property 
damage, risk of human injury or fatality, and wildlife mortality. If fencing remains undesirable due to 
aesthetic reasons, alternative options that may aid in reducing WVCs are: a) changing the management of 
the road and b) installing an animal detection system. 
Despite passing through Yellowstone National Park, US-191 in this area is managed as a through-road. 
Alternately, US-191 could be managed as a “park road” in this section and either not permit through traffic 
or be managed in a hybrid fashion in which through traffic is allowed during the day but not at night 
(especially for semi-trucks) when collisions are most likely to occur (i.e., a partial night-time closure). 
Another potential measure to reduce WVCs might be the installation of an animal detection system to warn 
drivers of animals’ presence. However, this alternative does not reduce the potential barrier effect of the 
highway (Huijser et al. 2015). 
There are four smaller culverts within this priority site (see Chapter 9: Wetland #1, Wetland #2, Terminal 
Monument Creek, Wickup Creek), and one larger culvert south of the priority site at Bacon Rind Creek. If 
highway projects are proposed, these structures should be evaluated for potential use by aquatic, semi-
aquatic and terrestrial species. At least two of the structures are small pipe culverts built into deep fill. 
These structures could be upsized to better accommodate aquatic organism passage and to provide safe 
crossing opportunities for small terrestrial wildlife. The culvert at Bacon Rind Creek is much larger and 
could be upsized to accommodate small- to medium-sized terrestrial or semi-aquatic wildlife, with dry 
crossing opportunities provided along the stream bank. 
Key Next Steps: 
• Continue to evaluate traffic volume, wildlife movement, and WVCs. 
• Replace existing bridges over the Gallatin River and Specimen Creek with structures that fully 
span the stream banks, allow for dry terrestrial passage year-round, and have sufficient vertical 
clearance (a minimum of 15 ft or 4.6 m) for use by species such as elk, moose, and grizzly bears. 
• If a highway improvement or stand-alone wildlife accommodation project is proposed in the area, 
evaluate the five existing culverts to determine the feasibility of upsizing them to allow for suitable 
safe passage by both aquatic and small- to medium-sized terrestrial species. See Chapter 9 for more 
information on these culverts and aquatic organism passage at this site. 
• Once the bridges and culverts have been replaced with structures suitable for the full range of target 
species, these structures could be tied together with fencing to further reduce WVCs and to guide 
wildlife to the structures. Alternatively, an animal-detection system could be implemented. 
 
55 
 
• Another potential measure to reduce WVCs could be a change in the designation and management 
of the highway in this section through Yellowstone National Park. 
 
 
Figure 20 Bridge over the Gallatin River (between RM 23-24) 
 
Table 19 Index Values of all 0.10-mile Segments within the Specimen Creek to Bacon Rind Creek Priority Site 
US-191 Wildlife Wildlife Suitability / WVC Risk Composite Moving 
Road Mile Movement Observation Connectivity 
Reference Importance Importance Importance Importance Importance Window 
Index Index Index Index Index Average 
RM23.2 0.668 0.191 0.810 0.000 0.935 0.640 
RM23.3 0.361 0.079 0.722 0.000 0.665 0.648 
RM23.4 0.073 0.159 0.778 0.000 0.538 0.654 
RM23.5 0.361 0.314 0.781 0.000 0.751 0.663 
RM23.6 0.146 0.130 0.902 0.000 0.654 0.679 
RM23.7 0.576 0.193 0.919 0.000 0.945 0.680 
RM23.8 0.122 0.317 0.758 0.000 0.589 0.654 
RM23.9 0.171 0.152 0.487 0.027 0.432 0.631 
RM24.0 0.361 0.009 0.654 0.027 0.625 0.625 
RM24.1 0.215 0.241 0.948 0.000 0.748 0.606 
RM24.2 0.146 0.173 0.737 0.054 0.594 0.594 
RM24.3 0.263 0.232 0.718 0.027 0.651 0.589 
RM24.4 0.073 0.264 0.540 0.000 0.411 0.583 
RM24.5 0.215 0.225 0.484 0.027 0.473 0.578 
RM24.6 0.171 0.695 0.490 0.027 0.547 0.562 
RM24.7 0.073 0.496 0.634 0.000 0.518 0.533 
RM24.8 0.644 0.772 0.579 0.000 0.896 0.531 
RM24.9 0.337 0.194 0.473 0.000 0.519 0.527 
RM25.0 0.098 0.165 0.498 0.000 0.380 0.553 
RM25.1 0.073 0.123 0.642 0.000 0.445 0.552 
 
56 
 
US-191 Wildlife Wildlife Suitability / 
Road Mile Movement Observation Connectivity WVC Risk Composite Moving 
Reference Importance Importance Importance Importance Importance Window 
Index Index Index Index Index Average 
RM25.2 0.049 0.153 0.639 0.000 0.435 0.550 
RM25.3 0.098 0.197 0.796 0.000 0.572 0.575 
RM25.4 0.146 0.147 0.787 0.027 0.603 0.556 
RM25.5 0.195 0.400 0.831 0.000 0.696 0.554 
RM25.6 0.098 0.231 0.586 0.027 0.465 0.569 
RM25.7 0.171 0.159 0.656 0.000 0.522 0.597 
RM25.8 0.502 0.102 0.783 0.000 0.796 0.639 
RM25.9 0.293 0.179 0.791 0.000 0.687 0.642 
RM26.0 0.024 0.329 0.644 0.054 0.493 0.648 
RM26.1 0.122 0.145 0.747 0.000 0.546 0.643 
RM26.2 0.454 0.452 0.647 0.000 0.754 0.681 
RM26.3 0.337 0.710 0.847 0.054 0.892 0.694 
RM26.4 0.024 0.252 0.892 0.027 0.614 0.691 
RM26.5 0.454 0.170 0.576 0.027 0.668 0.683 
RM26.6 0.312 0.261 0.664 0.000 0.637 0.691 
RM26.7 0.502 0.312 0.848 0.000 0.880 0.685 
RM26.8 0.171 0.423 0.780 0.027 0.671 0.672 
RM26.9 0.312 0.328 0.841 0.000 0.761 0.629 
RM27.0 0.146 0.145 0.809 0.000 0.600 0.645 
RM27.1 0.073 0.201 0.830 0.000 0.579 0.624 
 
 
 
 
57 
 
 
Figure 21 Specimen Creek to Bacon Rind Creek Priority Site Map 
 
58 
 
6.3. Taylor Fork 
 TAYLOR FORK: US Highway 191 (US-191), RM 34.2-36.4 
AADT: 2,239 
Priority Rank: 5 (tied with Teepee Creek Site) 
Table 20 Maximum, Minimum, and Average Index Values for each Prioritization Characteristic and Composite 
Index of all 0.10-mile Road Segments within the Taylor Fork Priority Site 
Prioritization Characteristic Maximum Index Minimum Index Average Index 
Value Value Value 
Composite (overall) 0.87 0.518 0.694 
WVC Risk 0.152 0 0.040 
Wildlife Crossing Road 0.551 0.098 0.311 
Live Wildlife Near Road 0.433 0.024 0.147 
Regional Conservation Value 0.899 0.567 0.755 
 
Table 21 Taylor Fork Field Evaluation Scores and Priority Ranking 
WVC Wildlife Live Regional Local Wildlife Overall 
Priority Site Risk Crossing Wildlife 
Near Conservation Land Conservation Accommodation Barrier Vulnerability Average Priority 
Road Road Value Security Value Options Effect Score Rank 
Porcupine 
Creek 1 3 2 2 2 2 4 4 5 2.78 2 
 
The Taylor Fork drainage is an important wildlife movement corridor between the Madison and Gallatin 
Mountain Ranges and Valleys. The Taylor Fork Priority Site is primarily surrounded by the Custer Gallatin 
National Forest (CGNF) on both sides of the highway with a private land inholding between RM 35-36. 
This site was identified due to its high regional conservation value and high composite scores. Elk crossings 
were documented in each of the 0.10-mile segments and grizzly bear crossings documented in 5 out of 15 
0.10-mile segments. The site also has high habitat connectivity/suitability values for grizzly bears, elk, 
bighorn sheep, and wolverines. One grizzly bear mortality has been documented, along with moose, elk, 
deer, and pine marten mortalities. Only about one in three of these wildlife mortalities were reported as 
crashes. Especially in the case of larger-bodied species, this may indicate that the mortalities were caused 
by large vehicles like semi-trucks, which are less likely to sustain damage or report collisions with wildlife 
(Huijser and Begley 2019; Abra et al. 2019) 
While reported WVC numbers are not high in this location compared to some others, as traffic volume 
increases, WVCs may increase. Present traffic volume is relatively low, at just over 2,200 AADT, such 
that wildlife can cross the highway safely during some parts of the day or year by taking advantage of times 
when traffic volume is lesser (Riginos 2022; Riginos et al. 2018). However, as traffic volume continues to 
increase due to YNP visitation, local population growth, and recreation on public lands, this is an important 
area to implement measures to mitigate WVCs and to maintain habitat connectivity between YNP and 
CGNF for multiple species. 
A bridge exists just south of the priority area (RM 33.9) over Taylor Fork. It has low vertical clearance as 
well as riprap extending nearly to the water’s edge, which makes it unsuitable for use by large ungulates 
(Figure 22). Although the existing structure is beyond the priority area based on the top 10% MWA scores, 
it should be replaced to provide safe wildlife passage. The bridge would need to be expanded in length and 
height to span the stream banks and to provide greater clearance in order to accommodate wildlife 
movement. This could occur at the end of its service life or earlier. Once it has been made suitable for 
 
59 
 
passage by large mammals [i.e., minimum 15 ft (4.6 m) vertical clearance and natural stream banks free of 
riprap, allowing for dry, unobstructed passage (Clevenger and Huijser 2011)] the bridge should be 
combined with fencing to reduce WVCs and to guide animals to safe crossing. A small pipe culvert just 
south of RM 36 (Chapter 9: Flints Creek, 320 Ranch) should be evaluated for upsizing to accommodate 
aquatic and potential small mammal passage. See Chapter 9 for more information on this culvert and 
aquatic organism passage at this site. 
Key Next Steps: 
• Replace the Taylor Fork bridge with a larger structure to provide a safe crossing opportunity for 
target species such as elk, moose, and grizzly bears, as well as a variety of smaller species. This 
structure should be wide enough to fully span the stream banks and provide year-round dry passage, 
with >15 ft (4.6 m) of vertical clearance above the stream banks and terrestrial walkways. 
• If a highway improvement project is planned in the area, the culvert immediately south of RM 36 
should be evaluated and potentially upsized to better accommodate movement across the highway 
by aquatic, semi-aquatic, and small terrestrial species. 
 
 
Figure 22 Bridge over Taylor Fork (RM 33.9) 
 
Table 22 Index Values of all 0.10-mile segments within the Taylor Fork Priority Site 
US-191 Wildlife Wildlife Suitability / 
Road Mile Movement Observation Connectivity WVC Risk Composite Moving 
Importance Importance Importance Importance Importance Window 
Reference Index Index Index Index Index Average 
RM34.2 0.366 0.173 0.596 0.000 0.609 0.635 
RM34.3 0.293 0.224 0.567 0.000 0.556 0.643 
RM34.4 0.439 0.233 0.602 0.000 0.671 0.667 
RM34.5 0.122 0.092 0.751 0.125 0.615 0.676 
RM34.6 0.390 0.138 0.899 0.000 0.806 0.680 
 
60 
 
US-191 Wildlife Wildlife Suitability / 
Road Mile Movement Observation Connectivity WVC Risk Composite Moving 
Importance Importance Importance Importance Importance Window 
Reference Index Index Index Index Index Average 
RM34.7 0.171 0.158 0.853 0.000 0.645 0.686 
RM34.8 0.195 0.085 0.763 0.049 0.619 0.702 
RM34.9 0.434 0.132 0.779 0.098 0.818 0.715 
RM35.0 0.244 0.116 0.886 0.152 0.797 0.701 
RM35.1 0.434 0.079 0.762 0.027 0.753 0.700 
RM35.2 0.171 0.116 0.766 0.127 0.661 0.662 
RM35.3 0.478 0.127 0.769 0.000 0.778 0.654 
RM35.4 0.288 0.024 0.876 0.000 0.704 0.647 
RM35.5 0.098 0.087 0.746 0.000 0.518 0.625 
RM36.4 0.551 0.433 0.717 0.027 0.870 0.626 
 
 
61 
 
 
Figure 23 Taylor Fork Priority Site Map 
 
62 
 
6.4 Porcupine Creek 
 PORCUPINE CREEK: US Highway 191 (US-191), RM 43.0-47.0 
AADT: 7,348 
Priority Rank: 2 (tied with N. of Big Sky Entrance Site) 
Table 23 Maximum, Minimum, and Average Index Values for each Prioritization Characteristic and Composite 
Index of all 0.10-mile Road Segments within the Porcupine Creek Priority Site 
Prioritization Characteristic Maximum Index Minimum Index Average Index 
Value Value Value 
Composite (overall) 0.925 0.546 0.740 
WVC Risk 0.295 0 0.100 
Wildlife Crossing Road 0.951 0.195 0.556 
Live Wildlife Near Road 1.0 0.193 0.491 
Regional Conservation Value 0.638 0.272 0.392 
 
Table 24 Porcupine Creek Field Evaluation Scores and Priority Ranking 
Wildlife Live 
WVC Wildlife Regional Local Wildlife Overall 
Priority Site Risk Crossing Near Conservation Land Conservation Accommodation Barrier Vulnerability Average Priority 
Road Road Value Security Value Options Effect Score Rank 
Porcupine 
Creek 1 3 2 2 2 2 4 4 5 2.78 2 
 
The Porcupine Creek Priority Site is located just over one mile south from the intersection of US-191 and 
MT-64 (Lone Mountain Trail). This site borders the Gallatin Wildlife Management Area and Custer 
Gallatin National Forest (CGNF) on the east side of the highway and has high concentrations of elk and 
other wildlife moving through the area. The research team extended the priority site south to RM 43 
following the site visits, due to the existence of WVCs and documented wildlife movement in the extended 
area and absence of feasible options for safe crossing structures within the initial site location. Local 
biologists emphasize the significance of the large, migratory elk herd that moves from Yellowstone 
National Park and CGNF west into the Madison Valley (Deb Wambach, Butte District Biologist, MDT; 
Julie Cunningham, Bozeman Area Biologist, FWP, pers. comms.). 
The east side of the highway is primarily public land covered by riparian meadows along the Gallatin River 
that lead to forested slopes in the Gallatin Mountains. This area provides important winter habitat for elk. 
Private land holdings exist along the west side of the highway, with dense development from RM 45-47, 
and primarily larger parcels south of RM 45. 
Porcupine Creek and Beaver Creek, which border the site on the west and east, respectively, serve as 
movement corridors for wildlife traveling between the Gallatin and Madison Ranges. The site scores very 
high for wildlife crossing the road (especially elk, from GPS collar data), regional conservation value, and 
composite importance. This site has relatively high rates of WVCs with elk and deer compared to other 
areas along US-191 and is tied for the second-highest priority ranking within the study area. Two grizzly 
bear mortalities have been documented within 1.5 miles south of the priority site. While the 0.10-mile 
segments with the highest scores (RM 45.5-47) might have been an ideal location to construct an overpass 
and develop fencing to allow wildlife to maintain habitat connectivity, existing levels of development 
increases the potential human-wildlife conflict, especially in areas west of the highway. 
 
63 
 
By extending the priority area south to RM 43, a wildlife overpass (preferred by elk, moose, and grizzly 
bears) could be built away from the densest levels of development and connected by fencing to guide 
animals to an overpass. The fence end locations would need to be strategically located and include end 
treatments such as an animal detection system to warn drivers of animals trying to cross. In addition to 
structural improvements such as an overpass and fencing with fence end treatments, additional measures 
should be considered in the northern portion of the priority site to reduce WVCs. 
During the site visit, the research team and TAC observed an injured and dying elk that had been hit by a 
vehicle near RM 45. To reduce WVCs with elk and other species from RM 45-47, mitigation measures 
that reflect the increasingly suburban nature of this zone are appropriate. Adding traffic-calming measures 
within the roadway to slow vehicle speeds and night-time lighting to make wildlife on the road more visible 
to drivers are options, even though the latter may increase the barrier effect of the road for some species. 
Potential physical changes to slow traffic include rumble strips, new pavement marking strategies, and 
roundabouts and other intersection improvements that also improve safety for vehicles entering and exiting 
the roadway. Roundabouts could feature elk or bighorn sheep sculptures to serve the dual purpose of 
slowing traffic and announcing the gateway to Big Sky. Similar measures have been developed in Radium 
Hotsprings, British Columbia. Physical changes, in addition to roadway lighting, may help to reduce 
WVCs, but would not provide habitat connectivity (Huijser et al. 2021). It is important to note that solely 
lowering the posted speed limit is generally not effective in changing driver behavior (Riginos 2022) and 
can lead to more frequent and severe crashes by introducing a phenomenon known as “speed dispersion” 
in which some drivers observe the posted limit while others drive the design speed of the road (Elvik 2014). 
Also at this site, a culvert at Beaver Creek (just north of RM 45) should be evaluated to provide safe 
crossing opportunity for small to medium-sized aquatic, semi-aquatic, and terrestrial wildlife. See Chapter 
9 for more information on this culvert and aquatic organism passage at this site. 
Key Next Steps: 
• Determine the potential to achieve land security through voluntary conservation easements on the 
west side of the road (RM 43-45). If land security is possible, then determine the engineering and 
design feasibility of an overpass and fencing, along with animal detection systems at fence ends. 
• Examine designs and feasibility for traffic-calming measures to reduce vehicle speed significantly, 
improve intersection safety, and increase the visibility of wildlife on and near roads from RM 45-
47. 
• Evaluate and potentially upsize the Beaver Creek culvert to accommodate passage by aquatic, 
semi-aquatic, and terrestrial wildlife. See Chapter 9 for more information on this culvert and 
aquatic organism passage at this site. 
  
 
64 
 
 
Figure 24 View north toward the Porcupine Creek Priority Site, just south of Big Sky in 1922 (left) and 2016 
(right). Credit: Historic photo-Gamel Family; Recent photo-Duncan Patten. 
 
Table 25 Index Values of all 0.10-mile Segments within the Porcupine Creek Priority Site 
US-191 Wildlife Wildlife Suitability / 
Road Movement Observation Connectivity WVC Risk  Composite  Moving 
Mile Importance Importance Importance Importance Importance Window 
Reference Index Index Index Index Index Average 
RM43.0 0.0244 0.042 0.492 0.134 0.388 0.421 
RM43.1 0.2146 0.135 0.548 0.125 0.555 0.438 
RM43.2 0.0244 0.107 0.614 0.000 0.394 0.453 
RM43.3 0.0000 0.068 0.537 0.000 0.323 0.458 
RM43.4 0.0488 0.052 0.752 0.223 0.623 0.462 
RM43.5 0.0000 0.201 0.526 0.054 0.377 0.480 
RM43.6 0.0244 0.039 0.652 0.107 0.471 0.481 
RM43.7 0.0244 0.089 0.540 0.348 0.561 0.465 
RM43.8 0.0488 0.098 0.356 0.422 0.510 0.477 
RM43.9 0.0244 0.073 0.374 0.429 0.505 0.494 
RM44.0 0.0000 0.070 0.512 0.429 0.575 0.484 
RM44.1 0.0244 0.032 0.548 0.098 0.399 0.500 
RM44.2 0.0732 0.377 0.452 0.000 0.380 0.504 
RM44.3 0.0488 0.123 0.515 0.277 0.524 0.494 
RM44.4 0.1463 0.189 0.423 0.223 0.507 0.482 
RM44.5 0.1707 0.464 0.448 0.098 0.517 0.495 
RM44.6 0.1463 0.333 0.519 0.152 0.553 0.502 
RM44.7 0.2683 0.252 0.320 0.196 0.515 0.531 
RM44.8 0.1707 0.175 0.335 0.196 0.448 0.565 
RM44.9 0.1707 0.308 0.374 0.000 0.377 0.568 
RM45.0 0.1951 0.231 0.458 0.350 0.648 0.589 
RM45.1 0.5366 0.218 0.456 0.027 0.654 0.592 
RM45.2 0.3415 0.261 0.708 0.054 0.715 0.593 
RM45.3 0.4146 0.309 0.583 0.152 0.754 0.618 
 
65 
 
US-191 Wildlife Wildlife Suitability / 
Road Movement Observation Connectivity WVC Risk  Composite  Moving 
Mile Importance Importance Importance Importance Importance Window 
Reference Index Index Index Index Index Average 
RM45.4 0.2439 0.318 0.589 0.000 0.559 0.646 
RM45.5 0.3902 0.334 0.545 0.179 0.737 0.693 
RM45.6 0.3415 0.354 0.361 0.098 0.546 0.700 
RM45.7 0.4390 0.307 0.308 0.103 0.567 0.716 
RM45.8 0.5854 0.369 0.313 0.295 0.794 0.708 
RM45.9 0.7073 0.574 0.316 0.027 0.748 0.719 
RM46.0 0.8537 0.431 0.272 0.223 0.904 0.752 
RM46.1 0.5610 0.679 0.354 0.223 0.824 0.779 
RM46.2 0.4634 0.710 0.351 0.000 0.629 0.797 
RM46.3 0.9024 0.426 0.351 0.054 0.877 0.781 
RM46.4 0.9512 0.637 0.309 0.054 0.925 0.769 
RM46.5 0.8049 0.739 0.350 0.000 0.847 0.720 
RM46.6 0.6829 0.390 0.298 0.098 0.728 0.707 
RM46.7 0.5610 1.000 0.375 0.000 0.765 0.693 
RM46.8 0.4146 0.545 0.452 0.000 0.627 0.685 
RM46.9 0.3659 0.348 0.495 0.223 0.722 0.762 
RM47.0 0.1951 0.193 0.638 0.134 0.617 0.642 
RM45.4 0.244 0.318 0.589 0.000 0.559 0.646 
RM45.5 0.390 0.334 0.545 0.179 0.737 0.693 
RM45.6 0.341 0.354 0.361 0.098 0.546 0.700 
RM45.7 0.439 0.307 0.308 0.103 0.567 0.716 
RM45.8 0.585 0.369 0.313 0.295 0.794 0.708 
RM45.9 0.707 0.574 0.316 0.027 0.748 0.719 
RM46.0 0.854 0.431 0.272 0.223 0.904 0.752 
RM46.9 0.366 0.348 0.495 0.223 0.722 0.762 
RM46.1 0.561 0.679 0.354 0.223 0.824 0.779 
RM46.2 0.463 0.710 0.351 0.000 0.629 0.797 
RM46.3 0.902 0.426 0.351 0.054 0.877 0.781 
RM46.4 0.951 0.637 0.309 0.054 0.925 0.769 
RM46.5 0.805 0.739 0.350 0.000 0.847 0.720 
RM46.6 0.683 0.390 0.298 0.098 0.728 0.707 
RM46.7 0.561 1.000 0.375 0.000 0.765 0.693 
RM46.8 0.415 0.545 0.452 0.000 0.627 0.685 
RM47.0 0.195 0.193 0.638 0.134 0.617 0.642 
 
 
66 
 
 
Figure 25 Porcupine Creek Priority Site Map 
 
67 
 
6.4. North of Big Sky Entrance 
 NORTH OF BIG SKY ENTRANCE: US Highway 191 (US-191), RM 48.1-49.4 
AADT: 8,421 
Priority Rank: 2 (tied with Porcupine Creek Site) 
Table 26 Maximum, Minimum, and Average Index Values for each Prioritization Characteristic and Composite 
Index of all 0.10-mile Road Segments within the North of Big Sky Entrance Priority Site 
Prioritization Characteristic Maximum Index Minimum Index Average Index 
Value Value Value 
Composite (overall) 0.842 0.593 0.7319 
WVC Risk 0.67 0 0.186 
Wildlife Crossing Road 0.507 0.073 0.263 
Live Wildlife Near Road 0.563 0.088 0.2796 
Regional Conservation Value 0.803 0.521 0.6736 
 
Table 27 North of Big Sky Entrance Field Evaluation Scores and Priority Ranking 
Wildlife Live Regional Local Wildlife Overall 
Priority Site WVC 
Risk Crossing Wildlife 
Near Conservation Land 
Security Conservation Accommodation Barrier Priority 
Road Value Value Options Effect Vulnerability Average 
Road Score Rank 
N. of Big 
Sky 1 2 2 3 3 3 4 4 3 2.78 2 
Entrance 
 
The North of Big Sky Entrance Priority Site was selected due to WVC risk as well as importance for 
wildlife movement and regional conservation value. Elk, deer, and bighorn sheep have been hit in this 
location. Elk frequently cross the highway along each of the 0.10-mile segments within the site, and 
bighorn sheep are often observed using the road to access water as well as to lick salt residue on the 
highway. With traffic volume nearing 8,500 AADT and rising, this area is high-risk for wildlife-vehicle 
conflict both in terms of WVCs and barrier effect to wildlife movement. 
Just north of the site, a bridge over the Gallatin River exists that is not suitable for use by most ungulates 
due to low vertical clearance and riprap that extends to the water’s edge, disrupting terrestrial passage. This 
bridge should be upsized to accommodate safe passage for ungulates and other wildlife. As an interim 
measure, fill could be added on top of the riprap along the north side of the bridge to create a pathway for 
terrestrial species. A purpose-built overpass structure in combination with fencing is recommended to 
accommodate safe passage across the highway for elk, which move between steep slopes on the west side 
down to the Gallatin River and riparian meadows (See conceptual rendering in Section 7.3). A wildlife 
overpass could also accommodate bighorn sheep with appropriate management. According to local 
biologists, bighorn sheep reside primarily in steep habitat west of the road (Julie Cunningham, Bozeman 
Area Biologist, FWP; Randy Scarlett, Wildlife Biologist, USFS, pers. comms.). During the site visits, the 
research team and TAC identified an area with heavy wildlife use where a steep slope west of the highway 
leads to a riparian meadow on the east, providing an ideal location to construct an overpass for wildlife 
(just north of Dudley Creek/RM 49). An overpass is a preferred structure for elk and bighorn sheep, as well 
as grizzly bears, which also use the area. The structure should include fencing to keep wildlife off the road 
and to guide them towards safe crossing. Fencing could connect to the bridge to the north, especially 
following its reconstruction to accommodate large wildlife such as elk [i.e. minimum 15 ft (4.6 m) vertical 
clearance and natural stream banks free of riprap, allowing for dry, unobstructed passage (Clevenger and 
 
68 
 
Huijser 2011)]. This location is largely surrounded by the Custer Gallatin National Forest and leads to the 
Lee Metcalf Wilderness Area northeast of the highway; however, several private parcels are located 
immediately adjacent to the highway and one or more of the parcels would have to hold a voluntary 
conservation easement for the development of a purpose-built structure at this site. A parcel along the east 
side already holds a voluntary conservation easement. Determining where the exclusionary fencing might 
tie into on the landscape toward the south for effectiveness is also a challenge of the location, given the 
nearby intersection of US-191 and MT-64 (Lone Mountain Trail), along with access roads and driveways. 
Key Next Steps: 
• Determine the potential to achieve land security through voluntary conservation easements on the 
west side of the road. If land security is possible, then consider the engineering and design 
feasibility of an overpass near RM 49 and fencing. 
• Evaluate retrofitting opportunities to render the bridge over the Gallatin River (between RM 49-
50) passable by wildlife species. Following shorter-term retrofitting and longer-term replacement 
to accommodate safe passage, consider the addition of fencing, especially if connection to an 
overpass is possible. 
 
Table 28 Index Values of all 0.10-mile Segments within the North of Big Sky Entrance Priority Site 
US-191 Wildlife Wildlife Suitability / 
Road Mile Movement Observation Connectivity WVC Risk Composite  Moving 
Reference Importance Importance Importance Importance Importance Window 
Index Index Index Index Index Average 
RM48.1 0.146 0.088 0.665 0.152 0.593 0.641 
RM48.2 0.195 0.119 0.676 0.107 0.608 0.675 
RM48.3 0.122 0.160 0.743 0.223 0.685 0.696 
RM48.4 0.073 0.157 0.521 0.670 0.794 0.700 
RM48.5 0.268 0.169 0.684 0.152 0.697 0.722 
RM48.6 0.293 0.335 0.627 0.152 0.711 0.739 
RM48.7 0.230 0.431 0.803 0.107 0.774 0.754 
RM48.8 0.507 0.313 0.782 0.000 0.842 0.765 
RM48.9 0.463 0.319 0.710 0.098 0.832 0.760 
RM49.0 0.220 0.563 0.620 0.277 0.786 0.746 
RM49.1 0.293 0.337 0.607 0.321 0.805 0.736 
RM49.2 0.244 0.396 0.684 0.196 0.757 0.710 
RM49.3 0.341 0.387 0.618 0.125 0.730 0.678 
RM49.4 0.293 0.141 0.690 0.027 0.632 0.639 
 
 
 
 
69 
 
 
Figure 26 North of Big Sky Priority Site Map 
 
 
70 
 
6.5. Gallatin Gateway to Spanish Creek 
 GALLATIN GATEWAY TO SPANISH CREEK: US Highway 191 (US-191), RM 68.1-73.7 
AADT: 10,047 
Priority Rank: 1 - Highest Priority 
Table 29 Maximum, Minimum, and Average Index Values for each Prioritization Characteristic and Composite 
Index of all 0.10-mile Road Segments within the Gallatin Gateway to Spanish Creek Priority Site 
Prioritization Characteristic Maximum Index Minimum Index Average Index 
Value Value Value 
Composite (overall) 0.541 0.256 0.414 
WVC Risk 0.812 0 0.231 
Wildlife Crossing Road* 0 0 0 
Live Wildlife Near Road 0.750 0 0.058 
Regional Conservation Value 0.896 0.148 0.463 
*See Section 2.4 on data gaps and limitations 
Table 30 Gallatin Gateway to Spanish Creek Field Evaluation Scores and Priority Ranking 
Wildlife Live 
WVC Wildlife Regional Land Local Wildlife Barrier Overall 
Priority Site Priority 
Risk Crossing Conservation Conservation Accommodation Vulnerability Average 
Road Near Value Security 
Road Value Options Effect Score Rank 
Gallatin 
Gateway 
to 3 1 1 2 4 5 4 4 4 3.11 1 
Spanish 
Creek 
 
The Gallatin Gateway to Spanish Creek Priority Site ranked as the Assessment’s highest priority primarily 
due to WVC risk as well as due to high regional and local conservation value in localized areas, especially 
from Big Bear Creek to the south. This area is one of the longest priority sites, covering over five miles, 
and has a variety of land-use and habitat types. The northern end is characterized by open grasslands, while 
south of RM 70, the highway begins to enter Gallatin Canyon. The area north of RM 70.5 is significantly 
fragmented and includes both residential and agricultural uses, especially east of the highway. Many 
smaller roads along with access areas and driveways complicate the potential for use of wildlife-
exclusionary fencing. The Custer Gallatin National Forest is set back from the highway east of the site. To 
the west, a single property under a voluntary conservation easement borders the highway. 
Because of land ownership patterns, there is potential to build an overpass with fencing to reduce WVCs 
and to connect the remaining grassland ecosystem between Gallatin Gateway and the Gallatin River bridge 
(RM 71-73), providing safe passage for elk and other species. The connection to the grassland ecosystem 
has been lost further north due to development. Given the current rate of subdivision and development, the 
opportunity to conserve habitat connectivity at the site may be lost without near-term action to conserve 
land. An overpass would be the most effective measure to reduce WVCs and to accommodate safe passage 
for elk; it is also the primary option available given that the roadbed is largely at-grade with the surrounding 
landscape. To do so would require ensuring land security long-term through voluntary measures east of the 
highway. 
Two bridges are within this site: one across the Gallatin River (RM 70.5, Figure 27) and another at the 
southern extent of the site across Spanish Creek (Figure 28). Neither of these structures are suitable for 
wildlife use due to the lack of passable stream banks along both sides. The Spanish Creek bridge also has 
 
71 
 
low vertical clearance. The Gallatin River bridge should be retrofitted in the near-term to provide a 
terrestrial pathway along the north side by adding fill to allow ungulates and other species to pass (See 
conceptual rendering in Section 7.1). To do so may require securing a voluntary conservation easement on 
a property that lies east of the river and highway. When replaced, this bridge should be increased to provide 
natural streambanks beneath it along both sides of the river to accommodate terrestrial wildlife movement. 
In addition to facilitating safe passage, an expanded bridge would increase infrastructure resilience to 
extreme weather events and lateral stream channel migration. 
  
Figure 27 Gallatin River Bridge (RM 70.5) 
 
On the southern end of the site, replacement of the bridge across Spanish Creek (Figure 28) is included in 
the MDT 2022-2026 Statewide Transportation Improvement Program. When this structure is replaced, it 
should be expanded to allow for greater vertical clearance beneath the bridge (>15 ft or 4.6 m from stream 
banks) and to provide natural passable stream banks for terrestrial wildlife movement under the bridge 
through the riparian area (Clevenger and Huijser 2011). In addition to facilitating safe passage, an expanded 
bridge would increase infrastructure resilience to extreme weather events and lateral stream channel 
migration. The area around Spanish Creek is an important area for local and regional conservation that 
serves as a movement corridor for wildlife. This area also has a localized spike in WVCs. When the bridge 
is expanded, wildlife-exclusionary fencing should be added along the highway both to the north and to the 
south in order to keep wildlife off the highway and guide them toward the underpass. Other areas south of 
Spanish Creek have the potential for purpose-built crossings for elk and grizzly bears as part of a larger 
project combining overpasses, underpasses, and exclusionary fencing for comprehensive mitigation of this 
stretch of highway. 
Key Next Steps: 
• Determine the potential to achieve land security through voluntary conservation easements from 
RM 70.5-73. If land security is possible, then engineering and design feasibility for an overpass 
and fencing in this area should also be pursued. 
• Continue to deploy variable message signs seasonally in the area of high WVCs from RM 70-73 
to warn drivers of elk and other wildlife on the road. 
• Determine if land conservation is necessary to retrofit the bridge over the Gallatin River at RM 
70.5. Based on feasibility, retrofit the structure to allow for wildlife to pass underneath by 
reconfiguring the riprap and providing a pathway for wildlife. Upon replacement, use a larger 
structure that allows for wildlife to pass along natural dry streambanks of both sides of the river. 
 
72 
 
• Upon replacement (currently in the planning phase), expand the Spanish Creek bridge at RM 68.1 
to allow passage along natural stream banks for terrestrial wildlife along both sides and vertical 
clearance of >15 ft (4.6 m) from the stream banks for species like elk, moose, and grizzly bears. 
Consider opportunities for additional wildlife crossings further to the south and connect safe 
crossing locations with fencing. 
 
 
Figure 28 Spanish Creek Bridge (RM 68.1) 
 
Table 31 Index Values of all 0.10-mile Segments within the Gallatin Gateway to Spanish Creek Priority Site 
US-191 Wildlife Wildlife Suitability / 
Movement Observation Connectivity WVC Risk Composite  Moving 
Road Mile Importance Importance Importance Importance Importance Window 
Reference Index * Index Index Index Index Average 
68.1 0 0 0.631816 0.026786 0.384627 0.417188 
68.2 0 0 0.536896 0.276786 0.481187 0.430612 
68.3 0 0 0.695195 0.053571 0.440765 0.451993 
68.4 0 0 0.710097 0.053571 0.450044 0.475578 
68.5 0 0 0.874782 0.026786 0.535909 0.487866 
68.6 0 0 0.882024 0.223214 0.662721 0.495209 
68.7 0 0 0.859868 0.080358 0.559976 0.498273 
 
73 
 
US-191 Wildlife Wildlife Suitability / 
Movement Observation Connectivity WVC Risk Composite  Moving 
Road Mile Importance Importance Importance Importance Importance Window 
Reference Index * Index Index Index Index Average 
68.8 0 0.001465 0.896951 0.026786 0.550014 0.486031 
68.9 0 0 0.76874 0.098215 0.514355 0.492418 
69 0 0 0.647574 0.075893 0.425014 0.506814 
69.1 0 0 0.751853 0 0.442689 0.503197 
69.2 0 0 0.560951 0.151785 0.418333 0.488862 
69.3 0 0 0.597409 0 0.346526 0.498014 
69.4 0 0 0.763372 0.098215 0.511014 0.494244 
69.5 0 0.008291 0.836662 0.178571 0.608401 0.4974 
69.6 0 0 0.810894 0.026786 0.496128 0.492264 
69.7 0 0 0.726984 0.125 0.505034 0.496598 
69.8 0 0 0.771544 0.330357 0.660643 0.489033 
69.9 0 0 0.804065 0.053571 0.508552 0.501921 
70 0 0 0.601271 0.321429 0.549065 0.489273 
70.1 0 0 0.605947 0.026786 0.36852 0.45492 
70.2 0 0 0.64985 0.178571 0.490363 0.440397 
70.3 0 0 0.480874 0.098215 0.335117 0.417249 
70.4 0 0 0.601893 0.223214 0.488299 0.390681 
70.5 0 0 0.48635 0.151785 0.371884 0.36643 
70.6 0 0.00293 0.28511 0.125 0.230513 0.348113 
70.7 0 0.157534 0.350024 0.178571 0.336375 0.336577 
70.8 0 0 0.344826 0.098215 0.250408 0.311813 
70.9 0 0.54446 0.424255 0.026786 0.368393 0.295239 
71 0 0 0.277426 0.151785 0.241797 0.275629 
71.1 0 0.030762 0.267412 0.321429 0.347574 0.263464 
71.2 0 0.037588 0.362829 0.053571 0.241623 0.264772 
71.3 0 0 0.364132 0.026786 0.217956 0.262537 
71.4 0 0.030762 0.177823 0.098215 0.15281 0.265249 
71.5 0 0.035929 0.189894 0.276786 0.272585 0.256519 
 
74 
 
US-191 Wildlife Wildlife Suitability / 
Road Mile Movement Observation Connectivity WVC Risk Composite  Moving 
Importance Importance Importance Importance Importance Window 
Reference Index * Index Index Index Index Average 
71.6 0 0.021973 0.362324 0.053571 0.238068 0.277964 
71.7 0 0.036621 0.395205 0.026786 0.244902 0.285405 
71.8 0 0.073243 0.436855 0.080358 0.311793 0.32074 
71.9 0 0.014649 0.262847 0.223214 0.280235 0.358229 
72 0 0 0.201522 0.276786 0.272367 0.371918 
72.1 0 0 0.343788 0.464286 0.477694 0.396472 
72.2 0 0.146485 0.172794 0.508928 0.429424 0.443101 
72.3 0 0.235841 0.331697 0.642859 0.630302 0.465904 
72.4 0 0.219727 0.444272 0.535714 0.630339 0.489622 
72.5 0 0.124512 0.258954 0.22768 0.303393 0.5117 
72.6 0 0.574862 0.167242 0.553573 0.542675 0.541669 
72.7 0 0.750002 0.22468 0.772323 0.750991 0.53182 
72.8 0 0 0.22096 0.616072 0.495726 0.514368 
72.9 0 0 0.148154 0.812501 0.572697 0.48334 
73 0 0.004395 0.227724 0.651786 0.523087 0.456238 
73.1 0 0.228923 0.266276 0.665179 0.602029 0.449093 
73.2 0 0.046875 0.265771 0.352679 0.369356 0.433039 
73.3 0 0.030762 0.179833 0.232144 0.237451 0.375323 
73.4 0 0 0.183585 0.321429 0.288996 0.347676 
73.5 0 0 0.217281 0.357143 0.332213 0.330764 
73.6 0 0 0.29478 0.107143 0.224807 0.316673 
73.7 0 0 0.195762 0.433036 0.366071 0.302376 
*See section 2.4 on data gaps and limitations 
 
75 
 
 
Figure 29 Gallatin Gateway to Spanish Creek Priority Site Map 
 
76 
 
6.6. Four Corners to Gallatin Gateway 
 FOUR CORNERS TO GALLATIN GATEWAY: US Highway 191 (US-191), RM 74.1-81-3 
AADT: 14,607 
Priority Rank: 6 
Table 32 Maximum, Minimum, and Average Index Values for each Prioritization Characteristic and Composite 
Index of all 0.10-mile Road Segments within the Four Corners to Gallatin Gateway Priority Site 
Prioritization Characteristic Maximum Index Minimum Index Average Index 
Value Value Value 
Composite (overall) 0.455 0.220 0.360 
WVC Risk 1.0 0.080 0.403 
Wildlife Crossing Road* 0.140 0 0 
Live Wildlife Near Road 0.183 0 0.005 
Regional Conservation Value 0.403 0.090 0.220 
*See section 2.4 on data gaps and limitations 
 
Table 33 Four Corners to Gallatin Gateway Field Evaluation Scores and Priority Ranking 
WVC Wildlife Live Regional 
Priority Site Crossing Wildlife Conservation Land Local Wildlife 
Conservation Accommodation Barrier Overall 
Risk Near Security Effect Vulnerability Average Priority 
Road Value Value Options Score Rank 
Road 
Four 
Corners 
to 4 1 1 1 1 3 2 5 4 2.40 6 
Gallatin 
Gateway 
 
Located between Four Corners and Gallatin Gateway, this priority site was selected due to the number of 
WVCs over the last 10 years. The area is heavily developed, with new construction ongoing. It has the 
highest traffic volume of any location within the study area and traffic volume has also increased rapidly 
(>38%, MDT 2020). Studies from Wyoming have found that a road becomes an absolute barrier for wildlife 
movement at traffic volumes above 15,000 AADT; most animals will stop trying to cross, and the few 
which try have a high likelihood of being struck by a vehicle (Riginos 2022). This site is likely to approach 
this threshold in the near future, resulting in severed connectivity and decreased frequency of wildlife 
attempting to cross the highway.  
Due to intensive development and the presence of many secondary roads, access roads, and driveways, few 
mitigation measures can be implemented. Neither highly nor even moderately effective options (i.e., 
wildlife crossings with fencing, fencing, animal detection systems) are feasible due to high traffic volumes 
and the number of access points along the highway. Seasonal signage, novel signage, alternative modes of 
transportation (e.g., shared transit options), and intelligent transportation system (ITS) features could help 
to incrementally reduce collisions but are unlikely to result in major reductions in WVCs. Lack of spatial 
planning for conservation has severely limited the available mitigation options. A medium box culvert at 
South Cottonwood Creek just south of RM 77 could provide safe passage opportunity for small- to 
medium-sized terrestrial wildlife species when water levels are low. 
 
 
 
77 
 
Key Next Steps: 
• Consider the use and potential for seasonal and/or novel signage, alternative modes of 
transportation (shared transit options and intelligent transportation system features), with the 
understanding that none of these are likely to result in major reductions in WVCs. 
• Evaluate the culvert at South Cottonwood Creek for use by small and medium terrestrial wildlife 
species. Consider the potential for the addition of a dry shelf for use by small to medium terrestrial 
species during higher flows. 
 
 
 
78 
 
 
Figure 30 Four Corners to Gallatin Gateway Priority Site Map 
 
79 
 
6.7. West Fork Gallatin 
 WEST FORK GALLATIN: MT-64 (Lone Mountain Trail), RM 0.2-1.2 
AADT: 10,513 
Priority Rank: 4 
Table 34 Maximum, Minimum, and Average Index Values for each Prioritization Characteristic and Composite 
Index of all 0.10-mile Road Segments within the West Fork Gallatin Priority Site 
Prioritization Characteristic Maximum Index Minimum Index Average Index 
Value Value Value 
Composite (overall) 0.632 0.354 0.500 
WVC Risk 0.420 0 0.150 
Wildlife Crossing Road 0.19 0 0.017 
Live Wildlife Near Road 0.217 0 0.056 
Regional Conservation Value 0.809 0.437 0.657 
 
Table 35 West Fork Gallatin Field Evaluation Scores and Priority Ranking 
Live 
Priority Site WVC Wildlife Wildlife Regional Land Local Wildlife 
Crossing Conservation Conservation Accommodation Barrier Overall Site 
Risk Road Near Value Security Vulnerability Average 
Value Options Effect Score Rank 
Road 
West 
Fork 1 1 1 3 4 3 4 3 3 2.56 4 
Gallatin 
 
The West Fork Gallatin Priority Site is located on MT-64 (Lone Mountain Trail) just west of the 
intersection with US-191. This area has relatively high traffic volumes (>10,000 AADT) and provides the 
sole access to the rapidly growing community of Big Sky. Residents and commuters report frequent 
sightings of bighorn sheep and elk. Bighorn sheep are primarily on steep slopes of the Custer Gallatin 
National Forest located on the north side of the road, while elk are often grazing in the meadows on mostly 
private land on the south side. Although bighorn sheep regularly approach the road to lick salt, the area is 
not considered critical for connectivity for either bighorn sheep or elk (Randy Scarlett, Wildlife Biologist, 
USFS, pers comm). The site has two existing structures: a small bridge at RM 0.2 (Figure 31) and a culvert 
at RM 1.1. While the east side of the underpass below the existing bridge is likely passable by bighorn 
sheep, black bears, mountain lions and small- to medium-bodied mammals such as coyotes, bobcats, lynx 
and wolverines, it is not passable by elk on the west side (Figure 31). Similarly, the culvert at RM 1.1 is 
potentially passable by small- and medium-sized species but not by large mammals. 
The research team and TAC believe the area could be treated with wildlife-exclusionary fencing between 
the two existing structures to keep wildlife, especially bighorn sheep, off of MT-64, while the smaller 
structures could still provide connectivity for other species. The fencing could extend beyond the priority 
site to the west and connect on the north side to fencing associated with the proposed overpass in the North 
of Big Sky Priority Site along the Custer Gallatin National Forest boundary. Fence-end treatments would 
need to be incorporated to keep bighorn sheep from accessing the road to lick salt. Traffic calming features 
such as transverse rumble strips or intersection control improvements could be used to slow vehicles where 
they are most likely to encounter bighorn sheep, along with an animal-detection system to alert drivers 
when wildlife is on the road. In addition, we recommend replacement of the culvert at RM 1.1 at the end 
of its service life with a larger structure that accommodates natural substrate or by a small span bridge to 
allow for movement by a broader suite of wildlife species including deer and black bears. 
 
80 
 
Key Next Steps: 
• Evaluate the feasibility of a combination of fencing, traffic calming measures, and animal detection 
systems to reduce WVCs. 
• When the culvert at RM 1.1 needs to be replaced, consider expansion to a larger structure or span 
bridge that could be used by species such as deer and black bears. 
 
 
Figure 31 Bridge over the West Fork of the Gallatin River along MT-64 (Lone Mountain Trail) RM 0.2 
 
81 
 
 
Figure 32 West Fork Gallatin Priority Site Map 
 
82 
 
6.8. Upper Big Sky Connectivity Area 
 UPPER BIG SKY CONNECTIVITY AREA: MT-64 (Lone Mountain Trail), RM 7.3-8.2 
AADT: 2,891 
Priority Rank: 7 
Table 36 Maximum, Minimum, and Average Index Values for each Prioritization Characteristic and Composite 
Index of all 0.10-mile Road Segments within the Upper Big Sky Connectivity Area Priority Site 
Prioritization Characteristic Maximum Index Minimum Index Average Index 
Value Value Value 
Composite (overall) 0.482 0.307 0.397 
WVC Risk 0.027 0 0.005 
Wildlife Crossing Road 0.190 0 0.038 
Live Wildlife Near Road 0 0 0 
Regional Conservation Value 0.758 0.533 0.636 
 
Table 37 Upper Big Sky Connectivity Area Field Evaluation Scores and Priority Ranking 
Live 
Priority Site WVC Wildlife Regional Local Wildlife Overall 
Risk Crossing Wildlife 
Near Conservation Land 
Security Conservation Accommodation Barrier Vulnerability Average Site 
Road Road Value Value Options Effect Score Rank 
Upper 
Big Sky 1 1 1 3 2 4 1 1 3 1.89 7 
 
The Upper Big Sky Connectivity Area Priority Site was selected due to its habitat value for higher elevation 
species including wolverines. This area sits high on steep slopes with low traffic and lower speed limits 
compared to other sites in the study area. Due to the currently low barrier effect of the road and low WVC 
rates, along with limited wildlife accommodation options due to topography and steep roadway grades, this 
area does not require immediate attention. Rather, as development and traffic volumes increase, it should 
be monitored for opportunities to help maintain regional connectivity and offset potential increases in 
WVCs. 
Key Next Steps: 
• None at this time. Continue to monitor and evaluate based on development and traffic pressures. 
  
 
83 
 
6.9.   Considerations for Additional Priority Sites 
In addition to the areas ranking in the highest 10% of the MWA identified in the Assessment and described 
in detail in Sections 6.1-6.9, other areas highlighted by the study remain important to consider in light of 
WVCs and the opportunity to maintain or improve habitat connectivity. Additional road segments and 
areas should be considered on a case-by-case basis, especially when planned highway projects align with 
nearby areas of wildlife-vehicle conflict. For example, areas adjacent to Yellowstone National Park 
regularly experience WVCs. When bridges in these areas are scheduled for replacement, standard practice 
should include development of structures with span lengths and clear heights suitable for safe passage by 
all species known to use an area, especially those for which wildlife-vehicle conflict is documented (Figure 
33). 
The information in Figure 33 also provides insight into localized considerations for bison. All recently 
recorded bison-vehicle collisions on US-191 have occurred south of RM 20. Within Yellowstone National 
Park, 15 bison carcasses were recorded by the National Park Service in the 1990s, along with single 
carcasses in both 2001 and 2004 (from a data set covering 1989-2021). The majority of these carcasses 
were recorded between RM 13-14, with a few in the vicinity of RM 20. From 2011-2020, the Montana 
Department of Transportation collected data on 29 bison carcasses south of RM 9.3. This total rose steeply 
in late 2022 due to the collision of a herd of 13 bison with a semi-truck near RM 4, increasing the total of 
recently recorded bison killed along this stretch of US-191 to 42. 
Based on these data, the vast majority [74% (42/57)]  of bison-vehicle collisions on US-191 beyond 
Yellowstone National Park in Montana have occurred south of RM 10. Including the data from the 
December 2022 incident, 22 bison crashes and carcasses have been recorded between RM 3-4 (Madison 
River), with 6 recorded between RM 7-9 (Cougar Creek and Duck Creek). This finding is consistent with 
a 2012 Western Transportation Institute assessment of bison-vehicle collisions based on data collected 
from 1999-2009, which identified two priority mitigation sites at these locations. The 2012 assessment 
states: “The road segment between 2.5 to 4.5 at the Madison River mitigation site includes 2 miles (3.22 
km) of very high BVC rates, representing only 20% of the total highway segment under study and 51% of 
the total BVCs (Dupree and Dimambro 2012).” The area near Cougar Creek also has very high composite 
scores in the present analysis, with five of the ten 0.1-mile segments scoring in the top 10% of the composite 
index values. Thus, the Cougar/Duck Creek and Madison River sites should be considered as potential 
additional priority sites. 
Mitigation options are complicated due to bison life-history characteristics, limited existing safe-crossing 
opportunities, and the need to provide connectivity for other wildlife species. Moose is the only other 
carcass data recorded in these two stretches. Crossing structures or other wildlife accommodation measures 
located at or near the Madison River bridge and Duck Creek bridge would require at least one mile of 
fencing, north and south along US-191, on both sides of the highway, to keep bison and other wildlife off 
of the road and direct them to crossings. Careful consideration should be given to where the fencing starts 
and ends, and fence end treatments will likely be necessary to discourage animals from entering the fenced 
road corridor. Other measures, such as animal detection systems, may be suitable as well.   
 
 
84 
 
 
Figure 33 Map of Carcass Data along US-191 in Areas within and adjacent to Yellowstone National Park 
 
 
85 
 
6.10.    Discussion 
US-191 and MT-64 bisect important wildlife habitat and movement corridors between Yellowstone 
National Park, the Custer Gallatin National Forest, and the Gallatin and Madison Ranges and their 
surrounding valleys. Rapidly expanding development, visitation, recreation, and traffic volumes are 
increasing the risk of human-wildlife conflict both in terms of WVCs and habitat connectivity for wildlife. 
In some areas, lack of spatial planning for conservation has limited effective wildlife accommodation 
measures. Long-term land conservation, such as through voluntary conservation easements, is critical to 
siting and implementing measures to reduce WVCs and to maintain or improve habitat connectivity for 
wildlife in priority sites from Porcupine Creek north to Four Corners. 
Traffic volume is a major factor influencing WVC risk and the barrier effect of a highway to wildlife 
movement. Traffic on US-191 from Porcupine Creek north to Four Corners already represents a significant 
impediment to wildlife (Riginos et al. 2018; Waller and Miller 2015). South of Porcupine Creek, traffic 
volumes are lower but may still be a partial barrier to wildlife movement depending on time of year or time 
of day. For this Assessment, the research team had access to Average Annual Daily Traffic (AADT) as a 
metric of traffic volume. However, this annual average is not representative of traffic patterns, especially 
in areas with large fluctuations in seasonal use (see Section 2.4). Conducting additional traffic studies to 
determine locations where seasonal traffic volumes present a substantial barrier to wildlife movement 
would provide more information about the potential effect of accommodation measures. 
The US-191/MT-64 Wildlife and Transportation Assessment identifies and prioritizes highway locations 
with the highest rank based on a combination of human safety and biological conservation factors. 
Mitigation along other areas of the study roads may be warranted and should also be considered based 
upon the documentation and analysis in this report, especially when highway projects are planned. 
This report includes recommendations for mitigation measures along priority road locations. The 
recommendations are based on the dual objectives of reducing collisions with larger-bodied wild mammals 
and providing safe crossing opportunities for wildlife given parameters such as local traffic volume and 
topography. However, the recommended measures should not be considered actions that must happen 
specifically as described. Rather, the human safety, biological conservation, and economic information 
summarized should be used to understand why certain highway sections are important to mitigate for 
wildlife, target species by location, and type of measures that may be effective. The exact location, type, 
and dimensions of any prospective measure depends upon public support, design and engineering 
feasibility, potential agreements with land management agencies and/or private landowners, and funding 
availability. 
  
 
86 
 
7. Renderings of Potential Wildlife Accommodation Options 
7.1. Retrofit of an Existing Structure 
In some locations in the study area, it may be possible to retrofit existing bridges to improve terrestrial 
habitat connectivity as a relatively low-cost improvement. The image below illustrates a conceptual bridge 
retrofit based upon an existing bridge over the Gallatin River (Figure 34; RM 70.5) within the Gallatin 
Gateway to Spanish Creek Priority Site. Currently, the area adjacent to the bridge abutments is covered in 
riprap and steeply angled, such that deer, elk, and moose are limited in their ability to move along the 
riparian corridor beneath the highway. 
By incorporating an elevated, gravel path into the area below the bridge, a retrofit would provide a secure 
option for terrestrial species to cross beneath the bridge year-round. The example protects the existing 
bridge abutment with a concrete retaining wall in order to enable development of the built-up, gravel path. 
Preferably, the path would have a minimum vertical clearance of 15 ft (4.6 m) to accommodate elk and 
moose, in addition to smaller-bodied species. The path also needs to be as wide as possible to provide a 
clear crossing area (Clevenger and Huijser 2011). In the example, a sloped rock embankment wall extends 
from the path to the riverbank (which is covered in snow in the image) to reduce erosion during high-water 
events. 
 
 
 
 
 
 
 
 
 
 
87 
 
 
 
 
Figure 34 (Top image) Existing Gallatin Gateway Bridge; (bottom image) Conceptual Bridge Retrofit (for 
illustrative purposes only). 
Note: Rendering is an artist interpretation to help visualize possible solutions to accommodate wildlife. 
Renderings are not drawn to scale, nor do they incorporate FHWA or MDT bridge specifications or other 
required elements. 
 
 
88 
 
7.2. Replacing and Upsizing an Existing Structure 
Replacement of existing bridges prior to or at the end of their service life with structures that can 
accommodate terrestrial wildlife passage offers another option to improve connectivity significantly. In 
this example, the existing bridge over Spanish Creek (Figure 35) is replaced with a upscaled version to 
accommodate terrestrial species, including moose and elk. The current bridge has low vertical clearance 
and sits just a few feet above the stream banks. It also barely spans the stream channel, such that no suitable 
dry passage exists for large wildlife to travel along the riparian corridor beneath the highway. The greater 
span length and height of an upscaled bridge provides dry footing and sufficient clearance to allow 
terrestrial wildlife to move freely beneath the bridge year-round, including during high water. At least 15 
ft (3.7 m) of vertical clearance is preferable to accommodate elk and moose (Clevenger and Huijser 2011). 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
89 
 
 
 
Figure 35 (Top image) Existing Spanish Creek Bridge; (bottom image) Conceptual Bridge Replacement (for 
illustrative purposes only) 
Note: Rendering is an artist interpretation to help visualize possible solutions to accommodate wildlife. 
Renderings are not drawn to scale, nor do they incorporate FHWA or MDT bridge specifications or other 
required elements. 
 
7.3. New Wildlife Overpass 
A purpose-built structure such as a wildlife overpass, in conjunction with fencing, is a highly effective 
option to maintain wildlife movement while reducing wildlife-vehicle collisions. This example depicts a 
conceptual wildlife overpass (Figure 36) in the North of Big Sky Entrance Priority Site, one of three 
locations where a wildlife overpass is among the recommendations of this report. In order to accommodate 
elk as a target species, an overpass would need a minimum width of 164 ft (50 m) and provide a clear line 
of sight across the entire structure. At this and other sites, an overpass would need to be combined with 8 
ft (2.43 m) high wildlife exclusionary fencing along the highway in both directions to guide wildlife onto 
the structure in order to reduce WVCs (Huijser, Fairbank, et al. 2016). 
 
90 
 
This example depicts an arched structure with a minimum vertical clearance of 16.5 ft (5.03 m) over a 32 
ft (9.75 m) roadway. Greater vertical clearance may be necessary in areas with high volumes of truck 
traffic. For this reason, the CLLC-WTI research team recommends increasing the minimum clearance by 
1-2 ft (0.3-0.61 m) to 17.5-18.5 ft (5.3-5.6 m) decrease the risk of a truck strike. 
 
 
 
 
 
 
Figure 36 (Top image) US-191 north of Big Sky; (bottom image) Conceptual Wildlife Overpass (for illustrative 
purposes only) 
  
 
91 
 
8. Potential Funding Sources 
8.1. Introduction 
According to Montana Department of Transportation (MDT) surveys, safety concerns and “wildlife 
crossings and barriers” are both in the top three or four priorities that stakeholders and the public would 
like to see the agency address in order to improve our state’s transportation system. However, the number 
one obstacle to advancing more wildlife crossing projects in the U.S. is a lack of available funding, 
according to a recent survey of state Department of Transportation staff across the country (Cramer 2022). 
Fortunately, growing support for wildlife crossings nationwide, across all levels of government, has 
established unprecedented new public funding opportunities for this work. Additionally, there is increasing 
interest from the philanthropic community in investing in projects that reduce wildlife-vehicle collisions 
while improving habitat connectivity. These wildlife accommodations can be “stand-alone” projects or 
incorporated into a variety of transportation improvement projects, including new construction and 
reconstruction; resurfacing; restoration, rehabilitation, or replacement; “spot improvements” and safety 
projects; and preservation of roads and bridges. To learn more, see the definition of “construction” under 
23 USC § 101(a)(4) and the improvement types listed on pages 8-9 of  Montana’s 2022-2026 Statewide 
Transportation Improvement Program. 
This chapter begins with an overview of existing federal funding sources that could be used to advance the 
wildlife accommodation measures discussed in this Assessment. The majority of public funding available 
comes from these federal sources, given that 88.5% of Montana’s roads and bridges are funded through 
federal dollars, and the Montana Department of Transportation receives no funding from the state’s general 
revenue fund (Montana Department of Transportation 2022). The majority of non-federal matching funds 
come from Montana’s fuel tax and gross vehicle weight fees (Montana Department of Transportation 
2022). 
After reviewing federal funding sources, this chapter offers examples from across the country of innovative 
state and local funding mechanisms, as well as public-private partnerships. These examples could serve as 
models for establishing additional funding sources to implement the recommendations of the Assessment. 
Many wildlife accommodation projects throughout the U.S. are funded through a diverse array of funding 
sources, including local, state, and federal funding, as well as private contributions from individuals, 
foundations, and corporations. Leveraging all existing, eligible funding sources—in addition to 
establishing new funding streams and attracting private investments—is paramount for garnering sufficient 
funds to complete successful wildlife accommodation projects. A diversified portfolio approach is 
important for advancing both dedicated wildlife accommodations and wildlife accommodation measures 
as components of larger transportation projects. 
8.2. Federal Infrastructure Funding Sources 
In 2021, the federal government enacted the Infrastructure Investment and Jobs Act (“IIJA,” Public Law 
117-58), which established the first-ever dedicated federal funding for wildlife crossings. In addition to 
this dedicated funding, projects to reduce wildlife-vehicle collisions and/or improve habitat connectivity 
are eligible for funding from at least fourteen other federal infrastructure programs. While there are billions 
of dollars available under the latter programs, they do not have funding set aside specifically for wildlife 
crossings. However, examples of wildlife crossing projects that have received federal funding under these 
broader transportation programs are provided in this section. Thus, wildlife components of larger capital 
improvement projects can receive funding from a variety of sources. 
The funding amounts listed in the descriptions below are for fiscal years 2022-2026. Eligible entities for 
each of these programs are listed in Table 38. Additional information and resources on the programs 
 
92 
 
described below can be found on the Center for Large Landscape Conservation’s Wildlife and the 
Bipartisan Infrastructure Law webpage, Animal Road Crossings Solutions’ webpage, the Federal Highway 
Administration’s Bipartisan Infrastructure Law webpage, and the US Department of Transportation’s 
notice of funding opportunities webpage. For further information about wildlife crossing projects that have 
taken advantage of federal transportation dollars, including many of those described in this chapter, visit 
the Wonderful World of Wildlife Crossings Story Map. 
 
Table 38 Eligible recipients for transportation funding programs under the Infrastructure Investment and Jobs 
Act of 2021 
Program State Metro/Regional Local Tribal Federal 
Bridge Formula Program x     
Bridge Investment Program x x x x x 
Federal Lands Access Program x  x x x 
Federal Lands Transportation Program x    x 
Forest Service Legacy Road and Trail     x 
Remediation Program 
Highway Safety Improvement Program x     
National Infrastructure Project Assistance x x x x  
(Mega) Program 
Local and Regional Project Assistance Grants x x x x x 
(RAISE) 
National Culvert Removal, Replacement, and x  x x  
Restoration Program 
Nationally Significant Federal Lands and Tribal x  x x x 
Projects 
Nationally Significant Freight and Highway x x x x x 
Projects (INFRA) 
PROTECT (competitive) x x x x x 
PROTECT (formula) x     
Rural Surface Transportation Grant x x x x  
Surface Transportation Block Grant x     
Wildlife Crossings Pilot Program x x x x x 
 
Categories of eligible entities listed in the header include state departments of transportation (“State”), 
metropolitan planning organizations or regional transportation authorities (“Metro/Regional”), local 
transportation authorities (“Local”), tribal departments of transportation (“Tribal”), and federal natural 
resource management agencies (“Federal”). Programs are listed in alphabetical order. 
Montana's funding flow for federal-aid funding, MDT’s non-federal matching funds, state-funded 
programs, and other federal allocations and transfers is illustrated in Figure 37. 
 
93 
 
 
Figure 37 Funding Flow for Federal Reauthorization Funds (Montana Department of Transportation, 2022-
2026 Statewide Transportation Improvement Program) 
 
8.2.1. Discretionary Grant Programs 
The following Infrastructure Investment and Jobs Act (IIJA) programs provide discretionary grants 
distributed through competitive application processes at the national level. 
8.2.2. Wildlife Crossings Pilot Program 
This new $350-million pilot grant program establishes dedicated funding for projects to reduce wildlife-
vehicle collisions and improve habitat connectivity (23 USC § 17). Proposals for funding under this 
program will also be evaluated based on the extent by which the project leverages non-federal funding, 
supports local economies and tourism, integrates innovative technology and design techniques, provides 
educational and outreach opportunities, and evaluates the efficacy of the project. Eligible entities include 
state departments of transportation, Tribes, federal land management agencies, metropolitan planning 
organizations, local governments, regional transportation authorities, and special districts. Eligible partners 
include foundations, non-governmental organizations, universities, and other government agencies. 
Additionally, 60% of funds will go to projects in rural areas (Federal Highway Administration 2022). 
8.2.2.1. Rural Surface Transportation Grant Program 
This new $2-billion program is for highway, bridge, or tunnel projects to improve and expand the surface 
transportation infrastructure in rural areas (those with a population of fewer than 200,000 residents) to 
increase connectivity, improve safety, enhance reliability, generate regional economic growth, and improve 
quality of life (23 USC § 173). Wildlife infrastructure projects, including tunnels and detection systems, 
are eligible for funding. Projects must have completed preliminary engineering studies and begin 
construction within 18 months. The federal cost share is generally 80%, and other sources of federal 
assistance can be used for match (US Department of Transportation 2023). 
 
94 
 
8.2.2.2. Promoting Resilient Operations for Transformative, Efficient, and Cost-saving 
Transportation (PROTECT) 
This new program includes $7.3 billion in formula allocation funding and $1.4 billion in discretionary grant 
funding for states and communities to make transportation resilience improvements to mitigate the risk of 
recurring damage from extreme weather events, flooding, or other natural disasters (23 USC § 176). 
Eligible projects include natural infrastructure and “protective features” that increase the size or number of 
drainage structures, replace culverts with bridges, lengthen or raise bridges, or upsize culverts for climate 
resilience and ecosystem benefits (Federal Highway Administration 2022). The federal cost share for 
projects under this program in Montana is 86.58% (Montana Department of Transportation 2022). 
8.2.2.3. Bridge Investment Program 
This new $12.5-billion program is for projects to replace, rehabilitate, preserve, or protect one or more 
bridges on the National Bridge Inventory; replace or rehabilitate culverts for purposes of improving flood 
control; and to improve habitat connectivity for aquatic species (23 USC § 124). The program includes 
$100 million in grants for planning, feasibility analysis, and revenue forecasting. This set-aside is for 
measures to improve bridge and culvert condition, safety, efficiency, and reliability and to replace, 
rehabilitate, preserve, or protect one or more bridges on the National Bridge Inventory. More specifically, 
projects that replace or rehabilitate culverts to improve flood control and habitat connectivity for aquatic 
species are eligible for funding. The federal cost share is 50% for large bridge projects, 80% for other 
bridge projects, and up to 90% for “off-system” bridges. 
Example: In 2022, Flathead County (Montana) secured $240,000 in Bridge Investment Program funding 
for bridge improvements, including wildlife connectivity improvements (Federal Highway Administration 
2022). 
8.2.2.4. National Culvert Removal, Replacement and Restoration Program 
This new $1-billion program is for states, local governments, and Tribes to facilitate fish passage by 
removing, replacing, or repairing culverts or weirs (23 USC § 6703). The federal cost share is up to 80% 
(Federal Highway Administration 2023). 
8.2.2.5. Local and Regional Project Assistance Grants: Rebuilding American 
Infrastructure Sustainably and Equitably (RAISE) 
This $7.5-billion program is for road, rail, transit, and other surface transportation of local and/or regional 
significance (49 USC § 6702). Selection criteria include safety, sustainability, equity, economic 
competitiveness, mobility, and community connectivity. Eligible projects include those that replace or 
rehabilitate a culvert or prevent stormwater runoff for the purpose of improving habitat for aquatic species. 
The maximum award is $25 million per project. The federal cost share is generally 80% but could be higher 
for projects located in rural, disadvantaged, or impoverished areas (U.S. Department of Transportation 
2022). 
Example: In 2019, the Wyoming Department of Transportation received $14.5 million under a previous 
iteration of this program (Better Utilizing Investments to Leverage Development, or “BUILD”) to construct 
wildlife crossings (including underpasses, fencing, and jump-outs) along U.S. Route 189 (Wyoming 
Department of Transportation 2019). 
 
 
95 
 
8.2.2.6. Nationally Significant Multimodal Freight and Highway Projects Program 
(INFRA) 
This $8-billion program is for multimodal freight and highway projects of national or regional significance 
that improve the safety, efficiency, and reliability of the movement of freight and people (23 USC § 117). 
Projects that increase safety on freight corridors where wildlife frequently cross the road are now eligible 
for funding. At least 30% of funding for small projects ($5 million minimum) goes towards rural areas, 
and 25% of funding for large projects ($25 million) goes towards rural areas. Large projects must have 
completed preliminary engineering studies, begin construction within 18 months, demonstrate a need for 
federal funding, and demonstrate available non-federal funding. The federal cost share is generally 60% 
but could be up to 80% for small projects (U.S. Department of Transportation 2022). 
Example: In 2022, the Colorado Department of Transportation (CDOT) received a $100-million INFRA 
grant for highway improvements on Interstate 70 (U.S. Department of Transportation 2022). The project 
includes the construction of a wildlife underpass and directional fencing, “the first major wildlife crossing 
to be constructed along the I-70 Mountain Corridor, and it will allow wildlife to safely cross underneath 
the interstate at a location which has historically been a hotspot for wildlife related crashes,” according to 
CDOT (Colorado Department of Transportation 2022) 
8.2.2.7. Nationally Significant Federal Lands and Tribal Projects 
This $275-million program is for the construction, reconstruction, and rehabilitation of nationally 
significant federal lands and tribal transportation projects (23 USC § § 203). Federal entities and state, 
county, or local governments may apply if sponsored by an eligible federal land management agency or 
Tribe. Eligible projects include measures to mitigate wildlife-vehicle collisions and mitigate damage to 
habitat connectivity and aquatic organism passage, including constructing, maintaining, replacing, or 
removing culverts and bridges. The federal cost share is up to 90% for projects on non-tribal transportation 
facilities (Federal Highway Administration 2022). 
8.2.2.8. National Infrastructure Project Assistance (“Mega”) 
This new $5-billion program is for large and complex transportation projects (including highway, bridge, 
freight, railway, and certain public transportation projects) with national or regional economic, mobility, 
or safety benefits. Half of the funds are allocated to projects greater than $500 million, and the other 50% 
are allocated to projects between $100-500 million. The federal cost share is 60%, but total federal 
assistance for a project receiving a grant under this program could be up to 80% (U.S. Department of 
Transportation 2023). 
Example: In 2023, the North Carolina Department of Transportation was awarded $110M to replace the 
Alligator River Bridge on U.S. Highway 64. The project includes wildlife crossing structures and 
directional fencing to improve habitat connectivity between the north and south areas of the roadway and 
reduce wildlife-vehicle collisions (U.S. Department of Transportation 2022). 
8.2.3. Formula Funding Programs 
The following IIJA programs distribute funding directly to states via formula allocations or formula grants. 
8.2.3.1. Bridge Formula Program 
This new $27.5-billion program is for states to complete bridge replacement, rehabilitation, preservation, 
protection, or construction projects on public roads (23 USC § 124). Funds can be used to plan, design, 
engineer, or construct bridges, to replace and rehabilitate bridges, and to improve bridges in poor condition. 
Given that “improvements that reduce the number of wildlife-vehicle collisions, such as wildlife crossing 
 
96 
 
structures” are included in the definition of “construction” under IIJA (23 USC § 101(a)(4)(H)), wildlife 
crossings are presumably an eligible component of construction projects funded under this program 
(Federal Highway Administration 2022). The federal cost share for projects under this program in Montana 
is 86.58% (Montana Department of Transportation, 2022-2026 Statewide Transportation Improvement 
Program). 
8.2.3.2. Surface Transportation Block Grant Program 
This $72-billion program is for state and local government projects on federal-aid highways and bridges 
on any public road, as well as transit capital projects (23 USC § 133). It includes the design, construction, 
monitoring, and maintenance of wildlife crossing structures or other measures designed to reduce wildlife-
vehicle collisions. The $7.2 billion for the Transportation Alternatives (TA) set-aside has funding available 
for local transportation projects, including environmental mitigation activities, such as reducing wildlife 
mortality caused by roads or improving terrestrial or aquatic connectivity. Local governments and other 
eligible entities (such as metropolitan planning organizations and non-profits) can apply for this funding 
through a competitive grant process developed by the state (Federal Highway Administration 2022). The 
federal cost share in Montana for projects under this program is 86.58% (Montana Department of 
Transportation, 2022-2026 Statewide Transportation Improvement Program). 
Example: In 2010, a public-private partnership secured a Transportation Alternatives grant that—when 
matched with a State Wildlife Grant and from the U.S. Fish and Wildlife Service and private dollars—
funded the construction of underpasses and fencing to provide safe passage for amphibians during their 
seasonal migration across the Monkton-Vergennes Road in Vermont. 
8.2.3.3. Highway Safety Improvement Program 
This $15.6-billion program provides states with funding to save lives and prevent serious injuries on public 
roads in accordance with their highway safety plans (23 USC § 148). Projects that improve safety by 
reducing wildlife-vehicle collisions are eligible. The federal cost share is 90% (Federal Highway 
Administration 2022). 
Example: In 2015, the Colorado Department of Transportation used Highway Safety Improvement 
Program dollars to construct a series of wildlife underpasses along State Highway 160 (Colorado 
Department of Transportation 2015). 
8.2.3.4. Federal Lands Transportation Program 
This $2.2-billion program is for projects that improve multimodal transportation on roads, bridges, trails, 
transit systems, and other transportation facilities (23 USC § 204). It includes $130 million in direct federal 
spending for the U.S. Forest Service, $180 million for the U.S. Fish & Wildlife Service, $1.7 billion for 
the National Park Service, and $154 million in competitive grant funding for other federal land 
management agencies. IIJA doubles the previous cap of $10 million to a current cap of $20 million per 
year for funds under this program that can be used for projects that reduce wildlife mortality due to roads 
while maintaining habitat connectivity (Federal Highway Administration 2022). 
Example: In recent years, the Texas Department of Transportation, in coordination with the US. Fish & 
Wildlife Service constructed a series of wildlife underpasses, primarily for ocelots, around the Laguna 
Atascosa National Wildlife Refuge using Federal Lands Transportation Program funds (Federal Highway 
Administration 2023). 
 
97 
 
8.2.3.5. Federal Lands Access Program 
This $1.5-billion program is for improving multimodal transportation on roads, bridges, trails, transit 
systems, and other transportation facilities that access the federal estate on infrastructure owned or 
maintained by states and local governments (23 USC § 204). It places an emphasis on high-use federal 
recreation sites and federal economic generators. It includes environmental mitigation during planning, 
engineering, and construction phases on or adjacent to federal lands to reduce wildlife mortality due to 
roads and maintain habitat connectivity (Federal Highway Administration 2023). 
Example: In 2017, the Idaho Department of Transportation received $2.8 million in Federal Lands Access 
Program funds to build a wildlife overpass on State Highway 21, with $220,000 in matching funds from a 
public-private partnership including Idaho Department of Fish and Game, the U.S. Forest Service, Army 
Corps of Engineers, the Western Federal Lands Highway Division, non-governmental organizations, and 
local cities and counties (Idaho Department of Transportation). 
8.3. Federal Conservation Funding Sources 
8.3.1. Collaborative-based, Aquatic-focused, Landscape-scale Restoration 
Program 
IIJA established this new $80-million competitive funding program for projects to restore water quality or 
fish passage on federal and non-federal lands (23 USC § 204). Priority is given to a proposal resulting in 
the most miles of streams being restored for the lowest amount of federal funding. Projects should contain 
proposed non-federal funding and request no more than $5 million (U.S. Forest Service 2023). 
8.3.2. Forest Service Legacy Road and Trail Remediation Program 
IIJA provided $250 million for direct federal spending on capital improvement and maintenance under this 
existing program (established under 16 USC 532). Eligible projects include decommissioning and repairing 
roads and trails to mitigate detrimental impacts on sensitive ecosystems and watersheds. Additionally, 
funding can be used to replace or install bridges and culverts (or low-water trail crossings), address public 
safety of roads and trails, restore unneeded roads and trails to a more natural state, address storm-damaged 
areas, and remove or replace pipes and other structures that impede aquatic habitat connectivity (U.S. 
Forest Service 2023). 
8.3.3. America the Beautiful Challenge 
The America the Beautiful Challenge combines funding from federal agencies and the private sector for a 
total of $1 billion over the next five years to support the implementation of large-scale ecosystem 
conservation and restoration across public and private lands. It provides funding for projects to address 
five objectives, one of which is “connecting and reconnecting wildlife corridors, large landscapes, 
watersheds, and seascapes” (National Fish and Wildlife Foundation 2022). The first funding cycle was 
completed in November 2022 and awarded $91 million. The 2023 Request for Proposals was issued and 
closed on April 20, 2023. Implementation grants are $1 million to $5 million in size over four years (with 
landscape-scale restoration projects potentially eligible to receive more), and planning grants are $200,000 
to $2 million in size over two or three years. State government agencies, U.S. territories, Indian Tribes, 
non-profit 501(c) organizations, local governments, municipal governments, and educational institutions 
are eligible to apply for the grants. 
 
98 
 
8.3.4. Proposed Recovering America’s Wildlife Act 
Recovering America’s Wildlife Act is a proposed bipartisan, popular piece of federal legislation (S. 1149) 
that would provide significant financial and technical assistance to states and Tribes to recover listed 
species, including implementation of State Wildlife Action Plans. This funding could be used to conserve 
or restore wildlife and plant species of greatest conservation need, support state wildlife conservation 
strategies, or enable wildlife conservation education and recreation. Montana’s State Wildlife Action Plan 
identifies habitat fragmentation as a specific threat to the state’s wildlife and habitat and further identifies 
highways as a major source of fragmentation (Montana Fish, Wildlife & Parks, 2015). 
8.3.5. U.S. Fish & Wildlife Service Funding for Imperiled Species 
Wildlife accommodation projects that conserve species that are listed as threatened or endangered under 
the Endangered Species Act and/or conserve the habitat of those species could potentially receive funding 
from existing U.S. Fish and Wildlife Service (USFWS) grant programs such as the Cooperative 
Endangered Species Conservation Fund and the Partners for Fish and Wildlife Program. Additional 
USFWS financial assistance opportunities can be found here. 
8.3.6. Western Big Game Seasonal Habitat and Migration Corridors Fund 
The National Fish and Wildlife Foundation administers the Western Big Game Seasonal Habitat and 
Migration Corridors Fund, which provides grants for projects to conserve the winter range and migration 
routes of pronghorn, elk, and mule deer in 11 western states, including Montana (National Fish and Wildlife 
Foundation 2022(a)). The fund has around $3 million available annually, awards approximately six to ten 
grants per cycle, and requires a 1:1 non-federal match of in-kind or cash contributions. One of the four 
strategies of the program is to work “cooperatively with tribal nations, private landowners and state 
highway departments to improve fencing, including modifying, removing, installing if serving to direct big 
game movement out of harm’s way, or seasonally adapting fencing if proven to impede movement of big 
game through priority migration corridors” (emphasis added, National Fish and Wildlife Foundation 
2022(b)). The program has funded a number of projects to reduce wildlife-vehicle collisions and improve 
habitat connectivity across highways. For instance, in 2022, the program funded the following initiatives: 
• A project in Montana to “address habitat connectivity on the Blackfeet Nation through removal 
and upgrade of fences, [and] reduction of animal-vehicle collisions along key sections of highway.” 
• A project in Idaho to “improve big game passage success rates across US-95 ... and reduce wildlife-
vehicle collisions by repairing and extending 2.8 miles of wildlife funnel fencing along wildlife 
crossing infrastructure.” 
• A project in Oregon to install 10 miles of fencing that will “direct mule deer and elk to a newly 
constructed wildlife underpass along U.S. Highway 97,” thereby reducing wildlife-vehicle 
collisions and reconnecting 75 miles of a wildlife migration corridor. 
8.4. Innovative State Funding Mechanisms 
8.4.1. Montana Wildlife and Transportation Partnership Project Program 
In 2018, the Montana Department of Transportation (MDT), Montana Fish, Wildlife, & Parks, and 
Montanans for Safe Wildlife Passage formed the Montana Wildlife and Transportation Partnership 
(MWTP) to address wildlife-vehicle conflicts. In 2023, MWTP launched a new planning tool and program 
to identify areas of greatest need of wildlife accommodations on highways throughout the state and 
strategically pursue effective solutions. This includes a new MWTP Project Program to solicit, evaluate, 
and advance public-private partnership proposals for transportation projects specifically aimed at 
addressing wildlife-vehicle conflicts and improving habitat connectivity. Under this program, MDT 
 
99 
 
funding may be available for feasibility studies of selected projects. However, applicants should identify 
additional sources of funding to design and build the proposed project if it is determined feasible by the 
preliminary analyses (Montana Department of Transportation, 2023). 
8.4.2. Other State Funding Models 
Additionally, other states are establishing innovative funding mechanisms for wildlife crossing projects 
rather than drawing resources from existing state transportation budgets. The following examples could 
serve as potential models for a new, state-level funding mechanism for wildlife crossings: 
• In 2005, the Wyoming Legislature created the Wyoming Wildlife & Natural Resource Trust to 
fund habitat and natural resource conservation across the state. Eligible projects include those that 
mitigate adverse consequences to wildlife habitat or mitigate wildlife conflicts, including wildlife 
crossings (Wyoming Wildlife & Natural Resource Trust 2023). The fund allows individuals and 
organizations to donate to these projects. 
• Both Wyoming and Oregon have created specialty wildlife conservation license plates. Money 
from the sale of these license plates and the renewal fees go towards established state funds that 
support habitat connectivity and wildlife crossing projects (Wyoming Game & Fish Department 
2023, Oregon Wildlife Foundation 2022). 
• Wildlife connectivity legislation enacted in California in 2021 sets up a compensatory mitigation 
credit scheme that allows the California Department of Fish and Wildlife to grant the California 
Department of Transportation credits for wildlife crossings that can be used for future 
transportation projects requiring environmental mitigation (California Legislative Information 
2021). 
8.5. Innovative Local Funding Mechanisms 
One tool that local governments have employed to fund wildlife crossing projects is the establishment of 
special purpose taxes. For instance, in 2019, Teton County (Wyoming) passed a $10-million Special 
Purpose Exercise Tax to support the construction of priority wildlife crossing structures identified in the 
county-wide Wildlife Crossing Master Plan (Jackson Hole Conservation Alliance 2020). Similarly, Pima 
County, Arizona, dedicated $45 million of its local sales tax revenues to conserve and restore “critical 
wildlife linkages” through measures such as building wildlife crossings, including underpasses on State 
Route 86 (Regional Transportation Authority, Pima County 2022). 
8.6. Public-Private Partnerships 
Fully covering all the costs associated with designing, building, monitoring, and maintaining wildlife 
crossing infrastructure is more feasible when partnerships are able to secure private contributions to match 
public funding available for a project. Indeed, one of the criteria against which the Federal Highway 
Administration will evaluate proposals under the Wildlife Crossings Pilot Program is the extent to which 
the proposed project leverages other funding sources, including from public-private partnerships (as stated 
explicitly in the statute, 23 USC § 171(e)(2)(A)). Identifying or establishing a non-profit organization that 
can receive tax-deductible donations for the project is an important consideration. Private contributions 
often come from individuals, charitable foundations, or corporations. Additionally, land trusts across the 
country are contributing to wildlife crossing projects by ensuring land on both sides of the road remains 
protected as viable habitat for the species attempting to cross (Center for Large Landscape Conservation 
2022). 
Below are some examples of public-private partnerships that have funded wildlife crossing projects: 
 
100 
 
• The Western Big Game Seasonal Habitat and Migration Corridors Fund is a public-private 
partnership between the National Fish and Wildlife Foundation, the Department of the Interior, the 
Department of Agriculture, Bezos Earth Fund, and several corporations: Burlington North Santa 
Fe Railway, ConocoPhillips, Altria Group, and Microsoft. The program has awarded $11.7 million 
across 52 projects, leveraging $57.5 million in matching contributions (National Fish and Wildlife 
Foundation 2022(c)). 
• A large portion of the Liberty Canyon wildlife crossing on US Highway 101 in California has been 
funded through private contributions. The project has received donations from thousands of 
private, philanthropic, and corporate institutions globally, including $26 million from the 
Annenberg Foundation and Wallis Annenberg, who is now the namesake of the overpass 
(Annenberg Foundation 2022). 
• The Land Trust of Santa Cruz County protected three properties on both sides of Highway 17, 
collected data, raised funds, and is partnering with the California Department of Transportation, 
Santa Cruz County Regional Transportation Commission, and others to construct a wildlife 
underpass (Land Trust of Santa Cruz County 2022). 
• In 2017, the Adirondack Chapter of The Nature Conservancy (TNC), Adirondack Land Trust, and 
the New York State Department of Transportation partnered to install the state’s first “critter shelf” 
inside a large culvert to help wildlife safely cross underneath the road (Civil & Structural Engineer 
Media 2017). The culvert is located between private lands protected by conservation easements, 
and TNC monitored the culvert with wildlife cameras (The Nature Conservancy 2019). 
• The WYldlife Fund, a partner foundation of the Wyoming Game and Fish Department, has 
prioritized funding wildlife crossing projects, such as building 15 miles of exclusionary fencing to 
direct wildlife to underpasses on Interstate 25 (WYldlife Fund 2022). 
In Montana, public-private partnerships have long been a critical source of support for conserving the 
state’s natural heritage. For instance, Montana’s Outdoor Legacy Foundation (MOLF) is the primary non-
profit partner of Montana Fish, Wildlife, and Parks, and the two entities have partnered on funding 
proposals to reconnect habitats and facilitate wildlife movement. MOLF leverages private funding for 
conservation (including “wildlife/wildlands management and care”) and manages the Montana Fish and 
Wildlife Conservation Trust (Montana Outdoor Legacy Foundation 2023). Thus, Montana has an existing 
entity that could potentially play a similar role to the WYldlife Fund and the Oregon Wildlife Foundation 
in funneling private funding toward wildlife accommodation projects to leverage public dollars. 
8.7. Conclusion 
In summary, there are various public and private transportation and conservation funding sources currently 
available to support the implementation of the recommendations of the Assessment. In addition, a number 
of state and local funding programs could be established to provide financial support. Ultimately, 
cultivating public-private partnerships will be important for securing robust and reliable funding streams 
for all the phases and elements of successful wildlife crossing projects. The strongest project partnerships 
include diverse stakeholders—such as wildlife and transportation agencies, academic researchers, non-
governmental organizations, and local landowners and business owners—working together to advance the 
common cause of making roads safer for drivers and wildlife. 
  
 
101 
 
9. Aquatic Organism Passage Assessment 
9.1. Introduction 
Fish and other aquatic species move throughout their range from streams and rivers to connected tributaries, 
lakes, and wetlands. If improperly designed, maintained, or constructed, road crossings can impede or 
prevent the upstream movement of aquatic species such as fish. In some cases, establishing or maintaining 
a structure as a barrier may be desired to protect a native species from introgressive hybridization or 
competition with non-native species.  
The Gallatin River on the northern end of the study area and the streams and rivers near Hebgen Lake at 
its southern extent support world-class fisheries and native fish, including Westslope cutthroat trout 
(Oncorhynchus clarkii lewisi). Other fish species along the corridor in the Gallatin River include brook 
trout (Salvelinus fontinalis), brown trout (Salmo trutta), longnose dace (Rhinichthys cataractae), mottled 
sculpin (Cottus Bairdii), mountain whitefish (Prosopium williamsoni), rainbow trout (O. mykiss), white 
sucker (Catostomus commersonii), and Yellowstone cutthroat trout (O. clarkii bouvieri) (MDT, 2020). 
Tributaries have similar species assemblages; however, in many cases, only a few species may be present. 
Westslope cutthroat trout are a species of concern in Montana and are often considered the target species 
for habitat restoration or connectivity efforts. Arctic grayling (Thymallus arcticus), native to the upper 
Missouri River, were present in portions of the study area but are now absent.  
Aquatic connectivity refers to the movement of all aquatic organisms and ecological flows and processes. 
Ecological flows include the movement of water, nutrients, woody debris, sediment, and other materials. 
The degree to which existing US-191 infrastructure prevents, limits, or provides aquatic passage for native 
fish species is the focus of this chapter. 
This information in this chapter is based upon an initial field evaluation of road-stream crossings along US-
191 and MT-64 in the study area carried out over the course of September-December 2021. The 
information is intended to provide a baseline and initial ranking of crossings in terms of aquatic organism 
passage, with preliminary recommendations from which more detailed studies could be initiated or 
mitigation measures, such as new crossing designs, could be implemented when sections of this highway 
are under reconstruction. 
9.1.1. Objectives 
The objectives of this aquatic organism assessment are: 
• Collect baseline data of existing road-stream crossings along the corridor, 
• Perform an initial assessment of culverts along the highway corridor to identify the level of 
connectivity through each structure, and 
• Identify culverts that may be limiting the unimpeded movement of fish species and opportunities 
to improve connectivity. 
9.1.2. Literature Review 
9.1.2.1. Fish Passage and Approaches to Assess Culvert Barriers  
Culverts are a common and often cost-effective means of providing transportation intersections with 
naturally occurring streams or rivers but have been identified as potential barriers to fish mobility. Bridges, 
as compared to culverts, are typically used to cross larger waterways and generally do not impede fish 
movements. Fish passage presents a complex challenge to engineers, hydrologists, and biologists due in 
part to the dynamic nature of streams and rivers, both physically and biologically. The interactions between 
 
102 
 
the physical and biological elements further complicate the problem. There are many physical factors that 
determine whether a fish can or cannot pass through a culvert; insufficient water depth, large outlet drop 
height, and excessive water velocity comprise the most common physical factors limiting passage (Baker 
and Votapka 1990; Burford et al. 2009; Votapka 1991; Fitch 1995). Biological factors such as a fish’s 
swimming ability, motivation, and behavior play an equally important role in passage. 
Although no comprehensive inventory of the number of culverts on fish-bearing streams in North America 
is available, there are an estimated 5+ million stream-road crossings in the United States. Examples of the 
number of culvert barriers from various parts of North America highlight the problem that improperly 
designed, constructed, or maintained culverts can pose to fish movement. Sixty-one percent of culvert 
crossings in the Notikewin watershed and 74% of culvert crossings in the Swan River watershed, both in 
Alberta, likely impede fish movement (Tchir, Hvenegaard, and Scrimgeour 2004). In Whatcom County, 
Washington, researchers assessed the passage status of culvert crossings on 1,673 crossings; they believe 
837 (50%) are barriers to fish passage (Whatcom County Public Works 2006). An analysis of fish passage 
across road-stream crossings identified 2,900 culverts on 50,000 miles of forest roads in Montana, northern 
Idaho, and western North and South Dakota. The analysis showed that about 80% of the culverts are 
barriers to Westslope or Yellowstone cutthroat trout at some life stage. Of the total surveyed, 576 (about 
20%) were classified as total barriers that completely isolate upstream fish populations (USDA National 
Technology Development Program 2008). An evaluation of four bridges and 47 culverts along a 210-km 
segment of the Trans Labrador Highway in Canada identified that 53% of the culverts posed fish passage 
problems due to poor design or construction (Gibson, Haedrich, and Wernerheim 2005). In Alberta’s 
Kakwa River watershed, 57% of culvert crossings are perched, thus blocking fish access to an estimated 
98 km of upstream habitats (Johns and Ernst 2007). 
There are many different methods to analyze the barrier status of culverts, each with distinct advantages 
and disadvantages. For this discussion, these methods are split into direct and indirect assessments. Direct 
assessments measure the amount of movement by fish in the field with an experiment such as a mark-
recapture study (Cahoon et al. 2005; Belford and Gould 1989; Warren Jr and Pardew 1998). Another 
method that can directly measure passage and also allows for the ability to analyze fish movement through 
a range of flow conditions is the use of PIT (passive integrated transponder) tags and antennae placed at 
the upstream and downstream ends of a culvert (Cahoon et al. 2005). These approaches can provide detailed 
information concerning both the passage status of a culvert and the hydraulic environment within and 
adjacent to the culvert that allows or prevents passage; however, they can be labor-intensive and are only 
practical for assessing a smaller number of culverts. 
Indirect methods generally approximate fish movement potential by comparing the culvert's physical 
conditions to those the fish are known to be able to overcome. FishXing is a software program that 
combines culvert characteristics (slope, length, roughness, etc.) and stream hydrology to model the 
hydraulic conditions in and near the culvert (Six Rivers National Forest 2012). These hydraulic conditions 
are then compared to the swimming ability of the fish species of interest to determine a passage status. 
Although this method of analysis may be useful for assessing a large number of culverts with a relatively 
small amount of field data collection, caution must be used when interpreting the results, as research shows 
that this method can provide a conservative (i.e. more barriers to movement are predicted when compared 
to direct assessment results) estimate of the barrier status of culverts (Cahoon et al. 2005; Karle 2005). 
HEC 26, developed by the Federal Highway Administration, also provides guidance on passage 
assessment, inventory, and design of culverts for fish and aquatic organism passage (Kilgore, Bergendahl, 
and Hotchkiss 2010). 
A common indirect approach—the one used in this study—to provide an initial assessment of how well a 
culvert allows, limits, or impedes upstream passage is to evaluate the culvert slope, outlet drop, and level 
of continuous substrate in the bottom. As the culvert slope increases, the upstream passage becomes more 
 
103 
 
challenging for fish because water velocity increases with increasing culvert slope. Outlet drop is the 
difference in vertical elevation between the water surface in the culvert outlet (downstream end) and the 
water surface in the pool downstream. Some fish species do not jump, and in all cases, fish of a given size 
and species can only jump a certain height. Thus, outlet drops may present a barrier to upstream movement. 
Continuous substrate in a culvert usually means the culvert was designed in this manner, often following a 
stream-simulation approach where the conditions within the culvert are constructed to match the conditions 
of the natural channel. The assumption is that if a culvert “mimics” the natural channel, passage through it 
should be no more or less challenging than along the stream itself. Finally, a culvert may impede passage 
by having a somewhat steep slope combined with an outlet drop. This condition can be viewed as a 
combination barrier where slope or outlet drop alone may not create issues; yet, the combined effect of 
slope with outlet drop makes for a challenging passage. 
Species abundance, size structure, and genetic differentiation can also be used as an indirect approach to 
evaluate fish passage. Typically, such studies compare the results of fish samples taken from locations 
upstream and downstream of a culvert. For example, population surveys performed upstream and 
downstream of a perched culvert indicated that cutthroat trout density was 64% lower upstream than 
downstream, and size structure was skewed to a higher proportion of large fish downstream of the culvert, 
suggesting the culvert was functioning at least as a partial barrier to upstream movement  (US Fish and 
Wildlife Service 2002). This upstream and downstream approach can provide valuable information about 
how culverts affect the abundance, size structure, and distribution of fish populations; however, results 
from these types of studies may be inconclusive regarding the barrier status of a culvert. There may not be 
significant differences between upstream and downstream samples, even when a culvert is a barrier. 
Inconclusive results may indicate either recent genetic isolation or that a culvert allows partial movement 
of a species of interest (Knaepkens et al. 2004). 
9.1.2.1. Other Impacts of Roads on Aquatic Habitat 
Although the focus of this assessment is on aquatic connectivity, a brief discussion of other impacts of 
roads on aquatic habitats is warranted. Roads and crossing structures can alter flow regimes, surface and 
groundwater hydrology, local geomorphology, and nutrient cycling (McKay et al. 2013). Some impacts 
may include the channelization of streams to accommodate road infrastructure, changes to flood dynamics, 
and changes to groundwater-surface water interactions. Roads can penetrate soil horizons that are conduits 
for water flows from groundwater to surface water and disrupt pathways. Groundwater temperatures do 
not fluctuate as much as surface water temperatures and typically maintain more constant temperatures 
throughout the year; therefore, they can help regulate stream temperatures, an especially important process 
during summer low flow periods (Edwards 1998; Giller 1998). For example, bull trout seek habitat with 
groundwater upwelling or downwelling for spawning sites (Baxter and Hauer 2000). 
Debris and sediment movement across landscapes can be altered by roads (Jones et al. 2000). Road-stream 
crossings can increase the accumulation of fine sediment downstream of crossings with potential impacts 
to salmonids, such as brook trout (Lachance et al. 2008). However, if proper best management practices 
(BMPs) are used at crossings, sediment load can be minimized (Morris et al. 2016; USDA National 
Technology Development Program 2008) 
9.2. Methods 
Study methods involved gathering basic physical data at road-stream crossings. The baseline data was then 
used to assess passage at each crossing (an indirect assessment method described in greater detail in the 
Literature Review, section 9.1.2, above) and identify opportunities for improving connectivity. 
 
104 
 
9.2.1. Data Collection 
Crossing locations were initially identified using the national hydrography dataset (NHD) to identify 
intersections between US-191, MT-64 and streams. A few additional crossings were identified in the course 
of assessing other structures in the field. 
Baseline field data was collected over several months during Fall 2021 at each crossing location. Monthly 
average stream flows for September, October, November, and December of 2021, as measured at Gallatin 
Gateway (USGS Gage No. 06043500) compared to the 20-year average for those same months, indicated 
2021 had lower average flows than the past 20 years. For comparison, 2021 average flow for September 
was 368 cubic feet per second (cfs) compared to the 20-year average of 442 cfs. For October, the 2021 
average was 393 cfs compared to the 20-year average of 424 cfs. For November, the 2021 average was 350 
cfs compared to the 20-year average of 364 cfs. For December, the 2021 average was 280 cfs compared to 
the 20-year average of 295 cfs. Section 9.5 exhibits the sample data sheet used for data collection activities. 
The physical data collected includes: 
• Shape of culvert, 
• Culvert dimensions, including length, width (span), and height, 
• Culvert material, 
• Culvert length and slope, 
• Presence of internal structures such as baffles, 
• Amount of substrate (if any) within the culvert barrel, and 
• Outlet drop height (perch). 
In addition, field personnel made observations of any structural defects, such as exposed rebar or separated 
culvert sections, to aid in operation and maintenance (O&M). Field personnel also collected photographs 
of each road-stream crossing for potential future use. 
9.2.2. Initial Passage Assessment 
We used an indirect assessment approach to provide an initial assessment of passage at all road-stream 
crossings utilizing culverts. This approach assumes all bridge structures provide unimpeded fish passage; 
therefore, it was not applied to bridge structures. The approach is a blend of approaches used in the 
Northern Region of the USDA Forest Service for juvenile Yellowstone or Westslope cutthroat trout (USDA 
National Technology Development Program 2008), the Alaska Region of the USDA Forest Service, 
Region 10 for juvenile coho salmon, and California’s Assessment Screen for juvenile salmonids (Taylor 
and Love 2002). 
The primary criteria include whether the substrate is present and continuous throughout the structure, outlet 
drop, culvert slope, and presence/absence of internal baffling structures designed for fish passage. Table 
39 provides the details for each criterion and describes outcomes as one of three possibilities: 
• Green: indicates conditions through the structure are assumed adequate for passage of fish. 
• Gray: indicates conditions may or may not be adequate for unimpeded passage of fish. 
• Red: indicates conditions may not be adequate for unimpeded passage of fish. 
 
 
 
 
 
105 
 
Table 39 Aquatic Passage Assessment: The table summarizes the criteria, thresholds, and passage assessment 
outcome. 
Criteria Assessment Passage Assessment Outcome 
Continuous Substrate Throughout 
Structure or Conditions assumed adequate for 
No Outlet Drop Combined with passage of fish Green 
Slope < 1.0% 
Outlet Drop > 0.34 ft or Slope > 0.01 Conditions may or may not be 
ft/ft (1.0%) adequate for unimpeded passage of Gray 
fish 
Outlet Drop > 0.34 ft and Slope > 0.01 Conditions may not be adequate for 
ft/ft (1.0%) and No Internal Structures unimpeded passage of fish Red 
Slope > 0.03 ft/ft (3%) or Outlet Drop Conditions may not be adequate for 
> 2.0 ft and No Internal Structures unimpeded passage of fish Red 
 
9.3. Results 
9.3.1. Physical Data 
A total of 41 road-stream crossings with culverts were surveyed. Nine sites included multiple structures, 
resulting in a total of 53 different culverts surveyed. Sixteen structures had either continuous substrate 
throughout them or a natural channel bottom. Internal baffles were not observed in any structure. Table 40 
summarizes the key physical characteristics of each culvert. Road-stream crossings with multiple structures 
(such as Big Bear Creek 2.0 A and B) are identified by the number of structures on site. 
 
Table 40 Aquatic Passage Assessment: Summary of Key Physical Characteristics at Each Crossing 
Number 
Crossing of Shape of Structure Slope Width Height Structure Outlet 
Name Structures (%) (ft) (ft) Material Drop Latitude Longitude 
at this Site (ft) 
South 
Cottonwood 1 Box 0.14% 18.75 10.00 Concrete 0.00 -111.196 45.597 
Creek 
Big Bear Creek 1 Pipe_Arch,Open 
1.0 Bottom_Arch 2.19% 12.83 6.33 Annular_CMP 0.00 -111.204 45.556 
Big Bear Creek 
2.0 b 1 of 2 Pipe_Arch 1.08% 7.17 3.17 Concrete 0.00 -111.205 45.555 
Big Bear Creek 
2.0 a 2 of 2 Pipe_Arch 1.11% 7.33 4.17 Concrete 0.00 -111.205 45.555 
Wilson Creek b 1 of 2 Pipe_Arch 0.85% 5.96 2.54 Concrete 0.00 -111.216 45.547 
Wilson Creek a 2 of 2 Pipe_Arch 0.39% 4.92 2.58 Concrete 0.00 -111.216 45.547 
Logger Creek 1 Box 2.84% 4.00 4.17 Concrete 2.50 -111.244 45.455 
Hellroaring 
Creek 1 Circular 2.76% 11.50 13.40 Annular_CMP 1.00 -111.237 45.449 
 
106 
 
Number 
Crossing of Shape of Structure Slope Width Height Structure Outlet 
Name Structures (%) (ft) (ft) Material Drop Latitude Longitude 
at this Site (ft) 
Pioneer Lakes 1 Circular 6.58% 1.50 1.50 Annular_CMP 0.00 -111.199 45.390 
Greek Creek 1 Pipe_Arch 7.57% 4.92 3.00 Concrete 0.00 -111.177 45.380 
Moose Creek 1 Box 5.98% 8.00 6.00 Concrete 0.00 -111.172 45.352 
Tamphery 
Creek 1 Circular 4.22% 3.00 3.00 Annular_CMP 0.00 -111.175 45.324 
Portal Creek 1 Box 3.88% 18.00 8.17 Concrete 0.00 -111.186 45.317 
Goose Creek 1 Circular 2.87% 3.00 3.00 Annular_CMP 0.00 -111.199 45.301 
Beehive Basin 
Creek 1 of 2 Circular 5.62% 3.88 3.88 Annular_CMP 0.50 -111.387 45.290 
Upstream 
Beehive Basin 
Creek 2 2 of 2 Circular 6.40% 6.17 6.17 Annular_CMP 0.00 -111.387 45.290 
Dudley Creek 1 Box 8.91% 13.08 4.00 Concrete 0.00 -111.242 45.272 
West Fork 
Gallatin River 1 Pipe_Arch 2.93% 15.50 10.50 Annular_CMP 0.00 -111.274 45.268 
North Fork 
Gallatin River 1 Open_Bottom_Arch 2.71% 20.00 12.16 Annular_CMP 0.00 -111.321 45.268 
Beaver Creek 1 Pipe_Arch 3.82% 11.25 7.42 Annular_CMP 0.00 -111.251 45.226 
Corral Creek 1 Circular 0.67% 3.00 3.00 Annular_CMP 0.00 -111.238 45.194 
Creek Across 
from Rainbow 1 Circular 1.22% 2.00 2.00 Annular_CMP 0.00 -111.238 45.193 
Ranch 
Buck Creek b 1 of 2 Pipe_Arch 2.65% 8.42 6.17 Annular_CMP 0.67 -111.245 45.168 
Buck Creek a 2 of 2 Pipe_Arch 0.76% 8.42 6.17 Annular_CMP 0.00 -111.245 45.168 
Cinnamon 
Creek 1 Circular 2.85% 3.00 3.00 Annular_CMP 0.00 -111.225 45.113 
Flints Creek 
(320 Ranch) 1 Circular 2.67% 2.58 2.58 Annular_CMP 0.00 -111.218 45.100 
Sage Creek 1 Pipe_Arch 1.02% 13.00 7.25 Annular_CMP 0.00 -111.188 45.067 
Snowflake 
Springs/Teepee 1 Circular 2.65% 4.00 4.00 Annular_CMP 0.00 -111.168 45.062 
Creek 
Dailey Creek 1 Circular 1.40% 8.00 5.33 Annular_CMP 0.00 -111.140 45.048 
Black Butte 
Creek 1 Circular 4.79% 4.00 4.00 Annular_CMP 0.00 -111.114 45.034 
Wickup Creek 1 Circular -3.78% 2.00 2.00 Annular_CMP 0.00 -111.091 45.020 
 
107 
 
Number 
Crossing of Shape of Structure Slope Width Height Structure Outlet 
Name Structures (%) (ft) (ft) Material Drop Latitude Longitude 
at this Site (ft) 
Terminal 
Monument 1 Circular 0.44% 4.00 7.75 Annular_CMP 0.00 -111.079 44.986 
Creek 
Wetland #1 1 Circular 0.15% 2.67 1.83 Annular_CMP 0.00 -111.079 44.981 
Wetland #2 1 Circular 1.26% 3.00 2.75 Annular_CMP 0.00 -111.079 44.978 
Bacon Rind 
Creek 1 Circular 0.67% 11.60 12.16 Annular_CMP 0.00 -111.067 44.958 
Tributary 1 of 
Grayling Creek 1 Circular 1.11% 6.92 7.00 Annular_CMP 0.50 -111.054 44.894 
Tributary 2 to 
Grayling Creek 1 Circular 5.79% 2.00 2.00 Annular_CMP 1.00 -111.054 44.887 
Grayling Creek 
1c 1 of 3 Pipe_Arch -0.13% 10.00 7.00 Annular_CMP 0.00 -111.022 44.882 
Grayling Creek 
1b 2 of 3 Pipe_Arch -0.06% 10.00 7.00 Annular_CMP 0.00 -111.045 44.872 
Grayling Creek 
1a 3 of 3 Pipe_Arch 0.24% 7.00 10.00 Annular_CMP 0.00 -111.045 44.872 
Tributary 3 to 
Grayling Creek 1 Circular 0.24% 3.00 1.00 Annular_CMP 0.00 -111.047 44.862 
Grayling Creek 
2b 1 of 3 Pipe_Arch 0.66% 10.00 7.00 Annular_CMP 0.17 -111.054 44.857 
Grayling Creek 
2c 2 of 3 Pipe_Arch 0.37% 10.00 7.00 Annular_CMP 0.33 -111.054 44.857 
Grayling Creek 
2a 3 of 3 Pipe_Arch 0.72% 10.00 7.00 Annular_CMP 0.08 -111.054 44.857 
Grayling Creek 
3b 1 of 3 Pipe_Arch 0.95% 10.00 7.00 Annular_CMP 0.00 -111.058 44.855 
Grayling Creek 
3a 2 of 3 Pipe_Arch 0.29% 10.00 7.00 Annular_CMP 0.00 -111.058 44.855 
Grayling Creek 
3c 3 of 3 Pipe_Arch 0.31% 10.00 7.00 Annular_CMP 0.00 -111.058 44.855 
Grayling Creek 
4a 1 of 3 Pipe_Arch 0.67% 10.00 6.75 Annular_CMP 0.00 -111.063 44.853 
Grayling Creek 
4b 2 of 3 Pipe_Arch 1.25% 10.00 6.75 Annular_CMP 0.67 -111.063 44.853 
Grayling Creek 
4c 3 of 3 Pipe_Arch 0.85% 10.00 6.75 Annular_CMP 0.00 -111.062 44.853 
Teepee Creek a 1 of 2 Pipe_Arch 1.77% 11.75 7.71 Annular_CMP 1.17 -111.095 44.810 
 
108 
 
Number 
Crossing of Outlet 
Name Structures Shape of Structure Slope Width Height Structure 
(%) (ft) (ft) Material Drop Latitude Longitude 
at this Site (ft) 
Teepee Creek b 2 of 2 Pipe_Arch 1.53% 11.42 7.75 Annular_CMP 0.67 -111.095  44.810 
Duck Creek 1 Pipe_Arch 1.47% 15.25 9.13 Annular_CMP 0.00 -111.113 44.780 
Table Note: “Annular CMP” refers to a corrugated metal pipe (CMP) with annular (circular) corrugations. 
 
Four different culvert types were identified: (1) pipe arch, (2) circular, (3) box, and (4) open bottom arch. 
Figure 38 includes a pie chart summarizing the number of structures by culvert type. The dominant culvert 
types were pipe arch at 25 and circular at 21, respectively. Pipe slopes ranged from adverse (negative) 
slopes up to approximately 0.1 ft/ft (or 10%). Figure 39 highlights the number of culverts in each slope 
range and the cumulative frequency distribution. The majority of culverts (n=29) had slopes ranging 
between 0.0 and 0.02 ft/ft (or 2%). Twelve of 53 structures had outlet drops ranging from 0.08 ft to a 
maximum of 2.5 ft. Figure 40 presents the number of culverts in each outlet drop range and the cumulative 
frequency distribution. 
2
5
25
21
Pipe Arch Circular Box Open Bottom Arch
 
Figure 38 Aquatic Passage Assessment: Pie Chart Showing the Number of Culverts by Type 
 
 
109 
 
35 100%
90%
30
80%
25 70%
20 60%
50% Frequency
15 40% Cumulative %
10 30%
20%
5
10%
0 0%
-0.04 -0.02 0 0.02 0.04 0.06 0.08 0.1 More
Culvert Slope (ft/ft)
 
Figure 39 Aquatic Passage Assessment: Cumulative Frequency and Histogram of Culvert Slope 
 
45 100%
40 90%
35 80%
30 70%
25 60%
50% Frequency
20 40% Cumulative %
15 30%
10 20%
5 10%
0 0%
Outlet Drop (ft)
 
Figure 40 Aquatic Passage Assessment: Cumulative Frequency and Histogram of Outlet Drop 
 
 
110 
 
Number of Culverts Number of Culverts
There were 13 bridges within the study area. Bridges were not assessed for fish passage as they are assumed 
to provide unimpeded passage to resident fish at all flows. 
9.3.2. Initial Passage Assessment 
Table 41 provides a summary of the initial passage assessment of road-stream crossings along the study 
corridor and Figure 41 provides a map of the crossing locations and their passage assessment. In addition, 
bridges are also identified in addition to the culverts. 
Of the total structures, thirty ranked as green, which implies conditions are adequate for the upstream 
passage of fish. Many of the green structures had substrate throughout, a natural channel bottom, or gentle 
slopes of less than 1.0%. Six of the culverts ranked as gray, meaning conditions may or may not be adequate 
for upstream passage of fish. A gray ranking does not necessarily mean the culvert is creating passage 
issues but should be viewed as a structure that may necessitate further evaluation. Lastly, 17 culverts ranked 
as red, indicating conditions may not be adequate for fish passage, and further evaluation should be 
considered. The numbers in the “ID for Map” column of the table correspond to the numbers on Figure 41. 
Table 41 Aquatic Passage Assessment: Summary of Initial Passage Assessment at Each Crossing 
Passage Assessment 
Rank: Green, Gray, or 
Continuous Red 
ID for Outlet Substrate (Sub.) in 
Crossing Name Map Slope (S) Drop (O) Structure?  
   Yes (Y) or No (N) Reason: Slope (S), Outlet 
Drop (O), or Substrate 
(Sub.) 
 
% ft 
    
South Cottonwood 
Creek 1 0.14% 0.00 Y 
 
Big Bear Creek 1.0 2 2.19% 0.00 Y 
 
Big Bear Creek 2.0 b 3 1.08% 0.00 Y 
 
Big Bear Creek 2.0 a 4 1.11% 0.00 Y 
 
Wilson Creek b 5 0.85% 0.00 Y 
 
Wilson Creek a 6 0.39% 0.00 Y 
 
Logger Creek 7 2.84% 2.50 N S, O 
Hellroaring Creek 8 2.76% 1.00 N S, O 
Pioneer Lakes 9 6.58% 0.00 N S 
Greek Creek 10 7.57% 0.00 N S 
Moose Creek 11 5.98% 0.00 N S 
Tamphery Creek 12 4.22% 0.00 N S 
 
111 
 
Passage Assessment 
Rank: Green, Gray, or 
Continuous Red 
ID for Outlet Substrate (Sub.) in 
Crossing Name Map Slope (S) 
 Drop (O) Structure?  
  Yes (Y) or No (N) Reason: Slope (S), Outlet 
Drop (O), or Substrate 
(Sub.) 
 
Portal Creek 13 3.88% 0.00 Y 
 
Goose Creek 14 2.87% 0.00 Y 
 
Beehive Basin Creek 
Upstream 15 5.62% 0.50 N S, O 
Beehive Basin Creek 
2 16 6.40% 0.00 N S 
Dudley Creek 17 8.91% 0.00 N S 
West Fork Gallatin 
River 18 2.93% 0.00 N S 
North Fork Gallatin 
River 19 2.71% 0.00 Y 
 
Beaver Creek 20 3.82% 0.00 N S 
Corral Creek 21 0.67% 0.00 Y 
 
Creek Across from 
Rainbow Ranch 22 1.22% 0.00 N S 
Buck Creek b 23 2.65% 0.67 N S, O 
Buck Creek a 24 0.76% 0.00 N 
 
Cinnamon Creek 25 2.85% 0.00 N S 
Flints Creek (320 
Ranch) 26 2.67% 0.00 N S 
Sage Creek 27 1.02% 0.00 Y 
 
Snowflake 
Springs/Teepee 
Creek 28 2.65% 0.00 N S 
Dailey Creek 29 1.40% 0.00 Y 
 
Black Butte Creek 30 4.79% 0.00 N S 
Wickup Creek 31 -3.78% 0.00 N 
 
Terminal Monument 
Creek 32 0.44% 0.00 N 
 
 
112 
 
Passage Assessment 
Rank: Green, Gray, or 
Continuous Red 
ID for Outlet Substrate (Sub.) in 
Crossing Name Map Slope (S) 
 Drop (O) Structure?  
  Yes (Y) or No (N) Reason: Slope (S), Outlet 
Drop (O), or Substrate 
(Sub.) 
 
Wetland #1 33 0.15% 0.00 Y 
 
Wetland #2 34 1.26% 0.00 Y 
 
Bacon Rind Creek 35 0.67% 0.00 Y 
 
Tributary 1 of 
Grayling Creek 36 1.11% 0.50 N S, O 
Tributary 2 to 
Grayling Creek 37 5.79% 1.00 N S, O 
Grayling Creek 1c 38 -0.13% 0.00 N 
 
Grayling Creek 1b 39 -0.06% 0.00 N 
 
Grayling Creek 1a 40 0.24% 0.00 N 
 
Tributary 3 to 
Grayling Creek 41 0.24% 0.00 Y 
 
Grayling Creek 2b 42 0.66% 0.17 N 
 
Grayling Creek 2c 43 0.37% 0.33 N 
 
Grayling Creek 2a 44 0.72% 0.08 N 
 
Grayling Creek 3b 45 0.95% 0.00 N 
 
Grayling Creek 3a 46 0.29% 0.00 N 
 
Grayling Creek 3c 47 0.31% 0.00 N 
 
Grayling Creek 4a 48 0.67% 0.00 N 
 
Grayling Creek 4b 49 1.25% 0.67 N S, O 
Grayling Creek 4c 50 0.85% 0.00 N 
 
Teepee Creek a 51 1.77% 1.17 N S, O 
Teepee Creek b 52 1.53% 0.67 N S, O 
Duck Creek 53 1.47% 0.00 N S 
 
 
113 
 
 
Figure 41 Aquatic Passage Assessment of Culverts Along US-191 and MT-64 
 
114 
 
Figure 42 shows three photos that are representative of the three possible passage rankings: “green,” “gray,” 
and “red”: 
• The left photo shows a “green” culvert, which indicates conditions are adequate for fish passage.  
This is because the continuous substrate throughout the structure mimics natural channel 
conditions. The substrate provides a range of different water depths and velocities, in addition to 
interstitial areas, for both large and small fish to move upstream through the structure as well as to 
rest or hide within it. 
• The middle photo shows a “gray” culvert, which indicates conditions may or may not provide 
adequate passage of fish. Potential passage issues are due to the culvert restricting channel flow 
(i.e., the culvert width is less than the stream channel width; this is often referred to as a 
“constriction ratio”), and there is an outlet drop. 
• The right photo shows a “red” culvert, which indicates conditions may not be adequate for 
unimpeded passage. The culvert creates shallow water depths and high velocity because it is set at 
a steep slope and may be undersized. 
 
 
Figure 42 Aquatic Passage Assessment: Images of a “Green” Crossing (left), a “Gray” Crossing (middle), and a 
“Red” Crossing (right) 
 
9.4.  Recommendations 
These recommendations describe general mitigation strategies for improving aquatic connectivity and 
include information on the benefits of larger structures for resiliency to climate change. They also 
summarize possible next steps to implement in order to improve aquatic connectivity along US-191. 
9.4.1. Mitigation for Aquatic Connectivity 
From the standpoint of aquatic species, the goal of a new crossing should be to ensure passable conditions 
are maintained throughout the engineering life of a crossing. This goal implies that “best” approaches are 
crossings that prevent passage problems—such as outlet drops or elevated velocities—from developing 
once a crossing is in operation. In some cases, passage problems are created when crossings restrict or 
 
115 
 
impede natural channel function or stream continuity, which then degrades the stream channel bed near 
them. Or in other cases, crossings that are undersized or set on a shallower slope than the natural stream 
slope may promote the aggradation of sediment and/or woody debris and create a damming effect on the 
upstream end of the crossing. This situation may not only reduce passage but can also create conditions 
that result in a structure failing. 
There are a range of options available for the design and construction of road-stream crossings specifically 
for aquatic organism passage and many good design manuals such as the FHWA document: HEC 26 -
Culvert Design for Aquatic Organism Passage, the USDA manual Stream Simulations: An Ecological 
Approach to Providing Passage for Aquatic Organisms at Road-Stream Crossings, and state guidance like 
the Washington Department of Fish and Wildlife design manual: Water Crossing Design Guidelines 
(2013). When possible, using bridges that are wide enough to allow for full floodplain function through 
them allows for the natural function of the stream or river and riparian area. Figure 43 is an illustration, 
taken from Huijser et al. 2018, of a bridge spanning a high-gradient mountain stream similar to those in 
the Gallatin River corridor (Huijser et al. 2018). These types of structures can also provide connectivity for 
both terrestrial and aquatic species. However, bridges are typically a more expensive option for road-stream 
crossings, which in some cases makes them infeasible. 
 
 
Figure 43 Aquatic Passage Assessment: illustration of a bridge spanning a mountain stream 
Other types of smaller crossing structures that fully span a stream or river and its banks provide the next-
best solution to ensure long-term passage of fish and aquatic species and are typically less expensive than 
 
116 
 
bridges (USDA National Technology Development Program 2008). The stream-simulation approach relies 
upon the following basic principle: 
 
“…designing crossing structures (usually culverts) that create a structure that is as similar as possible 
to the natural channel. When channel dimensions, slope, and stream bed structure are similar 
between the crossing and the natural channel, water velocities and depths will also be similar. 
Thus, the simulated channel should present no more of an obstacle to aquatic animals than the 
natural channel (USDA National Technology Development Program 2008). 
 
Stream-simulation culverts can be designed and constructed using a variety of different culvert 
types and shapes, including round, bottomless arches, box, and squash (or elliptical) shapes, 
amongst others (USDA National Technology Development Program 2008; Barnard et al. 2015). 
In addition, these designs can utilize single channel and additional structures to convey higher 
flows that activate side-channels or floodplain areas.” 
 
Recent studies investigated the effects of “stream simulation” on aquatic species movement, habitat, and 
channel form. In Alaska, crossings on a salmon stream in Anchorage were replaced with new crossings 
designed to mimic natural channel conditions. The study used a pre-post monitoring scheme to evaluate 
the response of salmon in the system to the new crossings. Monitoring showed a 300% increase in coho 
salmon escapement from pre-restoration to post-restoration conditions, i.e., from 481 adults in 2008 to 
approximately 1,500 adults on average, 2009-2013 (Myers and Nieraeth 2016). A study in Washington 
State evaluated 50 culverts designed following stream-simulation approaches by comparing channel bed 
and hydraulic conditions between culvert (treatment) reach and reference reaches. Sediment size and 
gradation, the velocity at the 2-year recurrence interval (RI) flow, and flow widths (i.e., the width of water 
at the surface) during the 2-year RI flow were similar in both the culvert and reference reaches. The culvert 
reaches, however, were not as likely to maintain channel complexity due to the difficulty of 
accommodating natural channel-forming features (such as woody debris or vegetative processes) within 
them. The stream-simulation culverts did not deteriorate from large flood events observed during the study, 
implying that such an approach maintains flood resiliency (Barnard et al. 2015). 
9.4.2. Other Benefits to Large Structures (Resilience to Climate Change) 
Designing larger spans to better accommodate stream and floodplain function, along with fish and wildlife 
passage, makes a crossing more resilient to future large flood events. One of the most promising mitigation 
techniques for aquatic species under changing climates is to increase habitat connectivity (Lawler et al. 
2009). Climate forecasts for the Greater Yellowstone Region indicate a shift towards higher temperatures, 
which could result in more precipitation as rain instead of snow and changes to the amount and timing of 
spring runoff (National Park Service 2022). Based on the last 50 years of recorded climate data in 
Yellowstone National Park, there are 80 more days with above-freezing temperatures at the northeast 
entrance and approximately 30 fewer days per year with snow on the ground than in the 1960s. 
Road crossing design requires the estimation of a flood flow to determine the size of crossing structures 
and road surface elevations, amongst other road features. In the United States, the design flood flow for 
many smaller county roads is the 50-year recurrence interval (RI) flood; for interstates and larger state 
highways, the design flood flow is often the 100-year flood flow. This value is typically determined based 
on empirical data from the past climate record using either records of gaged data or regional regression 
equations. With changing climates affecting future flood size and frequency, some countries have already 
begun to change their design flood RI, or their road design practices, based upon how the future climate 
 
117 
 
may look versus relying upon past climate records. For example, the Norwegian Public Roads 
Administration is requiring design elevation for road surfaces based on a 200-year RI. 
Recent historic flooding in Yellowstone National Park and surrounding areas highlights the need to build 
new highway infrastructure that will be resilient to large flow events. Figure 44 shows the hydrograph for 
the Gardner River (USGS gage #06191000) near Gardiner, Montana, and the Yellowstone River at Corwin 
Springs (USGS gage #06191500) downstream of Gardiner. The 2022 flood event was caused by a 
combination of excessive rainfall on top of a wet snowpack. Natural Resources Conservation Service 
SNOTEL sites in the area recorded up to 8.7 inches of combined rainfall and snow water equivalent (SWE) 
leading up to this flood. From June 10-13, 2022, the Fisher Creek gage station in the North Beartooth 
Wilderness recorded 5.1 inches of rain and 3.6 inches of SWE melt. Katherine Chase, a hydrologist with 
the United States Geological Survey, stated in The Livingston Enterprise (Jun 18, 2022) that the flood peak 
in the Yellowstone River near Livingston, about 50 miles downstream of Gardiner, was a 1 in 500-year 
event or larger. A 1 in 500-year event has a 0.2% chance of occurring every year. 
 
 
Figure 44 Aquatic Passage Assessment: Hydrographs for the Gardner River and Yellowstone River at Corwin 
Springs showing the June 13/June 14, 2022, historic flood event 
 
9.4.3. Possible Next Steps 
Possible next steps are described below. 
• Confirm the presence of fish species in crossings assessed as gray or red prior to initiating 
mitigation. Stream habitat upstream of each crossing should be evaluated to determine if natural 
barriers exist. Natural barriers close to a culvert may make re-construction of a crossing a poor 
investment should habitat be limited regardless of the degree of passage through the crossing. 
• Carry out additional evaluation at sites identified as red or gray to better understand the degree to 
which passage may be limited. In some cases, the culverts may already provide sufficient passage 
without significant impact on fish species. One option for red and grey sites would be to perform 
a more detailed hydraulic analysis as a relatively easy and cost-effective way to gauge what flows 
and how often a structure might be limiting upstream passage. 
• Design new structures following the approaches identified in section 9.4.1 as possible when 
sections of US-191 or MT-64 are upgraded. 
 
118 
 
• Evaluate aquatic passage for species other than fish. This study focused on fish passage; other 
aquatic species and semi-aquatic species would also benefit from improved connectivity along US-
191 and MT-64. 
• Consult with Montana Fish, Wildlife and Parks, the US Forest Service (USFS), and the National 
Park Service to identify opportunities to protect native species. In some cases, such as streams with 
pure cutthroat populations in upper reaches or headwaters, it may be desirable to leave a barrier 
culvert in place or to design the new crossing as a complete barrier to protect native species. For 
example, Figure 45, below, shows a culvert designed as an intentional barrier to protect Westslope 
cutthroat trout on a USFS road over Leverich Creek, south of Bozeman. 
 
 
Figure 45 Aquatic Passage Assessment: The picture shows a culvert designed as a fish barrier 
 
 
  
 
119 
 
9.5. Sample Culvert Data Collection Sheet 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
120 
 
10. Spatio-temporal Hotspot Analysis 
10.1. Background 
The following theoretical analysis based on crash and carcass data is included in this report to test the 
validity of using various lengths of data sets (3, 5, and 10 years) to identify locations for prospective 
wildlife mitigation measures. It uses US-191 as a case study. The results are not intended as guidance in 
the same manner as in other sections of this Assessment. Rather, this analysis provides additional 
information to further substantiate best practices for preparing wildlife and transportation assessments as 
part of an ongoing research program of MSU-Western Transportation Institute (WTI) and the Small Urban 
and Rural Transportation Center on Mobility (SURTCOM). This analysis adds to the body of road ecology 
knowledge drawn upon to prepare the US-191/MT-64 Wildlife & Transportation Assessment and provides 
additional insight into best practices in the field. 
Many transportation agencies use wildlife-vehicle collision (WVC) crash and carcass data for planning and 
project development under the Federal Highway Administration’s (FHWA’s) Highway Safety 
Improvement Program as well as other federal and state highway programs. Working with WVC data can 
be challenging because it is often not systematically collected and has been susceptible to under-reporting 
(Bíl et al. 2019; Donaldson 2017). The difference between the number of reported crash and carcass 
locations for the same section of road demonstrates the variability between these two types of WVC data 
and highlights the importance of robust mechanisms to predict high-risk locations (Creech, McClure, and 
Callahan 2016; Wang et al. 2010). The two types of WVC data also produce data sets with different 
strengths. Carcass data can be better for identifying animal characteristics (e.g., species, sex, age), while 
crashes reported by law enforcement can capture collision scenarios more accurately (e.g., weather, time, 
day, and driver characteristics). The spatial accuracy of WVC locations also varies greatly depending on 
the methods used by the individuals collecting and reporting data and the resources they have available 
(Creech et al. 2019). 
There are different WVC analysis methods that researchers generally agree upon, and many continue to 
develop innovative methods to increase the accuracy of models (Creech et al. 2019). Something that does 
not appear to be agreed upon is the amount of data required to get accurate results (Hou, Huo, and Leng 
2020; Santos et al. 2017). WVC data sets are highly variable in both the amount and the types of 
information collected. Some published results are based on analyses that use as little as three years of data 
(Favilli et al. 2018), while others take advantage of more than 20 years of data (L. McDonald, Messmer, 
and Guttery 2019). Most studies, regardless of the cumulative years of data used, identify patterns and 
landscape characteristics associated with road segments with high rates of WVCs. Such locations are often 
referred to as “hot spots.” Most models evaluate WVCs and their associated variables over the longest time 
frame that data has been consistently collected. Rarely do models evaluate subsets of data for shorter time 
frames to determine a data set’s variability over time or to identify changes in the locations of hot spots 
over time. Thus, variations in hot spot locations over time and environmental factors that are temporal are 
often masked in the longer-term data sets. 
This analysis aims to explore WVC data sets to determine if WVC hotspots are highly variable or 
temporally invariant for the same road segments based on the number of years of crash and carcass data 
incorporated into analyses. This question has important consequences for transportation agency investment 
in mitigation measures, such as the siting of wildlife overpasses and underpasses with life expectancies of 
50 to 100 years. If hotspot locations migrate over time, and the magnitude of the change in distance of a 
hotspot location exceeds the spatial extent of crossing structures and associated wildlife fencing, the long-
term effectiveness of mitigation measures could potentially be greatly diminished. Ultimately, mitigation 
efforts are based on the number of years of WVC data used and when an analysis is conducted. Results 
may vary if an agency uses five years of data in 2012 compared to the same number of years of data in 
 
121 
 
2017. Hotspot locations may change between the time of an initial analysis and when the installation of 
mitigation elements may be completed. This analysis evaluates and compares multiple scenarios using 
different types (e.g., crash and carcass) and amounts (e.g., number of years) of data analyzed over time. 
10.2. Case Study Framework 
Identifying the amount and type of WVC data required to conduct a hotspot analysis is a complex problem. 
Numerous approaches exist to compare the different amounts of data used in a specific analysis and 
resulting hotspots. For this reason, we developed a case study to examine results based on actual mitigation 
practices, using US-191 as an example. The intent of this case study is to decipher the degree to which 
initial mitigation efforts capture WVC hotspots compared to a future period when new crash and carcass 
data are available. 
Three “year-groups” (i.e., 3, 5, and 10 years) of data were compared during the period from 2008 to 2020. 
Initial analyses were conducted to identify significant WVC hotspot locations along US-191 for each 3-, 
5-, and 10-year group. The initial hotspot analysis for the 3-year group is based on the years 2008-2010, 
the initial 5-year analysis is based on the years 2008-2012, and the initial 10-year analysis is based on the 
years 2008-2017. We assumed the top ten hotspot locations identified in the initial analyses would be 
mitigated to reduce WVCs. To simplify, we focused on fencing practices as a mitigation measure and 
assumed 1.86 mi (3 km) of fencing would be installed in each direction in addition to the 0.62 mi (1 km) 
road segment with the identified hotspot (Huijser et al. 2015). Therefore, if a hotspot in a future analysis 
was identified within 1.86 mi (3 km) of an initial hotspot, it was assumed to be captured by the initial 
mitigation effort. A threshold of 80% similarity was used to assess whether the year-group comparisons 
were similar or different. A hotspot analysis was conducted in each following year to see how hotspots 
move based upon when an analysis was conducted. The average of each year group of data analyzed was 
used to determine how well subsequent years compared with the initial year examined. 
10.3. Methods 
We obtained 13 years (2008-2020) of WVC data crash and carcass data along US-191 from Four Corners 
to West Yellowstone. Crash data was obtained from the Montana Highway Patrol (MHP) and included all 
records involving collisions with animals resulting in a human injury or fatality or where property damage 
exceeds $1,000 (Montana Uniform Accident Reporting Act, Section 61-7-109). Only reported crashes 
where a collision with an animal was identified as the first harmful event were included in the data 
analyzed. Crash data are less spatially biased than carcass data because MHP responds to a collision 
regardless of location, which is recorded with GPS accuracy. Carcass data were obtained from the Montana 
Department of Transportation (MDT) and include all carcass locations picked up by MDT maintenance 
staff, US Forest Service (USFS), Montana Fish, Wildlife and Parks (FWP), or other local agencies, and are 
recorded to the nearest tenth-mile. All state roads are not surveyed equally, with bias to roads that are 
monitored more frequently. US-191 may show some bias in some locations due to the multiple agencies 
that collect carcasses and some areas not conducive to locating carcasses (e.g., thick vegetation, steep 
embankments, seasonal weather, etc.). This fact may cause issues in comparing analyses statewide and 
even along US-191, even though carcass data is abundant. Such inconsistencies within data sets should be 
considered when interpreting the results of any WVC analysis. For both crash and carcass data sets, all 
domestic (e.g., livestock, horses, dogs, sheep, etc.) and small animals (e.g., coyote, bobcat, badger, beaver, 
etc.) were removed so that data sets analyzed only include larger-bodied wildlife. 
To understand the amount of WVC data required to capture the hotspots along a road, we adapted a 
nonparametric Monte Carlo approach to compare two point-location data sets to see how they 
proportionately differ from each other (Andresen 2009; Creech et al. 2019). This is a practical approach to 
compare two data sets where there are large differences between the groups. Following Creech et al. (2019), 
 
122 
 
we: (1) assigned a year group as the reference data and calculated the proportion of WVCs observed in 
each 0.62 mi (1 km) road segment along the US-191 study area; (2) randomly selected 85% of test data for 
1,000 simulations; (3) calculated the percentage of WVCs within each km road segment for each of the 
random samples; (4) calculated a 95% confidence interval for each of the km road segments from the test 
data; (5) identified if the reference data set proportions were within a 95% confidence interval for each 
road segment created from the test data. All statistical analyses were performed using R version 4.2.1. 
To identify and compare WVC hotspots along US-191, we used a KDE+ clustering method (Bíl, Andrášik, 
and Janoška, 2013). This clustering analysis is commonly used in transportation hotspot identification 
analyses (Bíl et al. 2019; Favilli et al. 2018; Andrášik and Bíl 2015). The KDE+ is based on the Kernel 
Density Estimate (KDE), which estimates the expected density of the input data. It uses Monte Carlo 
simulations to identify only statistically significant clusters and ranks them based on cluster strength (Bíl, 
Andrášik, and Janoška 2013). This hotspot identification method is applicable to this analysis because the 
cluster strength is based on the total number of collisions (i.e., crash or carcass) that happen along a road 
segment and can identify locations along the roadway with the highest probability of collisions regardless 
of the number of collisions along a roadway. The results from the hotspot analysis were overlaid upon 0.62 
mi (1 km) road segments to compare differences along US-191. 
The same US-191 study area analyzed throughout this report was divided into 132 segments, with the first 
segment located in West Yellowstone and the 132nd segment ending at Four Corners. The KDE+ analyses 
were performed using the KDE+ toolbox version 3.3 for ArcGIS Desktop version 10.8 using a bandwidth 
of 150 with 800 Monte Carlo simulations per analysis. 
10.4. Results 
The two WVC data sets were subset to include only records within 328 ft (100 m) of US-191 from West 
Yellowstone in the south to Four Corners in the north. Crash data contained 451 collision locations, while 
the carcass data had 1,685 locations (Figure 46). It is a common trend across the entire state of Montana to 
have more carcass data than crash data along a road segment. There is a decreasing trend in the number of 
carcasses along US-191 over the 13-year period examined, while the opposite, a slight upward trend, exists 
for the crash data. Both data sets show a large decrease in the number of collisions in 2020, which is likely 
influenced by the COVID-19 pandemic shutdown and lower traffic volume numbers. 
 
 
123 
 
 
Figure 46 Distribution of WVC data with Large Mammals within the US-191 Study Area 
 
10.4.1. Crash Data Analysis 
The proportion of crash locations shows that using fewer years of data in the analysis leads to less similar 
results than using 10 years of data (Figure 47). Based on the Monte Carlo simulations, using 7 to 9 years 
of data shows over 80% similarity in the proportion of crashes for the 0.62 mi (1 km) road segments along 
US-191 when looking at the averages of the four analyses. This comparison shows that, on average, using 
one year of data is approximately 52% proportionately similar to using 10 years of data. 
 
Figure 47 Proportion of WVC Data per kilometer of road segments equal to using 13 years of Crash Data 
 
 
124 
 
10.4.1.1. 10-year Crash Data Analysis 
A total of 32 significant hotspots were identified in the initial 10-year crash data analysis. The top 10 
hotspots identified result in a total of 46 km of fencing with five distinct mitigated areas: 1-5 km, 66-72 
km, 76-83 km, 91-97 km, and 114-132 km. A comparison of each of the four time periods analyzed (2008-
2017, 2009-2018, 2010-2019, and 2011-2020 data years) and the initial mitigation is shown in Figure 48.  
In three analyses (2009-2018, 2010-2019, and 2011-2020 data years) following the initial analysis (2008-
2017 data years), the top 10 hotspots identified were located within the initial mitigation efforts 83.3% of 
the time (Table 42). The hotspots that were not captured by the mitigation efforts were 1.24-5.6 mi (2-9 
km) away from the fence ends established under the initial analysis. 
 
Figure 48 Comparison of WVC Hotspots per kilometer of road segments using 10 years of Crash Data 
 
Table 42 Comparison of the top 10 identified WVC hotspots using 10 years of crash data 
Top 10 Road Segment Hotspots (km)
Hotspot Rank
2008-2017 2009-2018 2010-2019 2011-2020
1 79 79 79 79
2 80 69 118 118
3 94 118 70 70
4 118 80 69 69
5 69 94 121 121
6 2 2 94 7
7 120 125 89 80
8 129 126 7 129
9 117 7 129 122
10 125 117 80 14
% Covered by Initial Mitigation 90 80 80  
 
10.4.1.2. 5-year Crash Data Analysis 
A total of 20 significant hotspots were identified in the initial 5-year crash data analysis. The top 10 hotspots 
result in a total of 29.2 mi (47 km) of fencing used in four distinct mitigated areas: segments 4-16, 66-82, 
114-123, and 125-132. A comparison of each of the nine 5-year periods analyzed with the initial mitigation 
is shown in Figure 49. 
 
125 
 
In the eight analyses following the initial period (2008-2012 data years), the top 10 hotspots identified were 
located within the initial mitigation efforts 71.3% of the time (Table 43). The hotspots not captured were 
1.86-9.32 mi (3-15 km) away from the fence ends established under the initial analysis. 
 
Figure 49 Comparison of WVC Hotspots per kilometer of road segments using 5 years of Crash Data 
 
Table 43 Comparison of the Top 10 Identified WVC Hotspots using 5 Years of Crash Data 
Hotspot Top 10 Road Segment Hotspots (km)
Rank 2008-2012 2009-2013 2010-2014 2011-2015 2012-2016 2013-2017 2014-2018 2015-2019 2016-2020
1 79 79 80 80 79 80 118 79 79
2 120 80 79 79 80 118 80 70 82
3 7 7 7 110 14 79 111 118 70
4 8 8 8 111 118 110 79 89 89
5 69 94 92 94 110 111 98 121 80
6 117 69 94 7 111 94 12 80 12
7 13 77 117 8 94 98 75 12 122
8 128 120 118 92 121 12 110 75 7
9 129 118 77 117 12 2 7 108 81
10 76 129 132 118 122 91 108 7 118
% Covered by Initial 
Mitigation 90 80 60 70 40 60 80 90
 
 
10.4.1.3. 3-year Crash Data Analysis 
A total of nine significant hotspots were identified in the initial 3-year crash data analysis. These hotspots 
result in a total of 16.78 mi (27 km) of fencing with two distinct mitigated areas: segments 91-97, and 113-
132. A comparison of each of eleven 3-year analyses with the initial mitigation is shown in Figure 50. 
In the 10 analyses conducted following the initial period (2008-2010 data years), the top 10 hotspots 
identified were located within the mitigation efforts 40% of the time (Table 44). The hotspots that were 
not captured in the mitigation efforts were 1.24-54.06 mi (2-87 km) away from the fence ends established 
under the initial analysis (2008-2010 data years). Only the second analysis (2009-2011 data years) 
 
126 
 
following the initial mitigation hotspot analysis captured greater than 50% of the top 10 hotspots previously 
identified. 
 
Figure 50 Comparison of WVC Hotspots per kilometer of road segments using 3 years of Crash Data. 
 
Table 44 Comparison of the Top 10 Identified WVC Hotspots using 3 years of Crash Data 
Hotspot Top 10 Road Segment Hotspots (km)
Rank 2008-2010 2009-2011 2010-2012 2011-2013 2012-2014 2013-2015 2014-2016 2015-2017 2016-2018 2017-2019 2018-2020
1 116 13 117 79 80 80 80 74 80 7 79
2 120 117 79 7 79 79 110 75 12 118 122
3 132 74 7 8 76 118 111 8 108 70 7
4 94 69 8 77 77 110 79 118 118 79 118
5 118 129 74 117 14 111 8 80 72 108 72
6 128 120 120 120 118 77 118 12 8 120 127
7 129 132 129 13 121 94 3 132 110 129 81
8 126 128 13 76 120 127 122 131 7 75 75
9 124 116 14 129 132 125 121 7 98 110 121
10 - 124 128 14 66 126 12 98 125 121 129
% Covered by Initial 70 40 30 40 50 30 30 20 40 50
Mitigation  
 
10.4.2. Carcass Data Analysis 
Similar to the crash data, the proportion of carcass locations shows that using fewer years of data in the 
analysis leads to less similar results than using 10 years of data (Figure 51). Based on the Monte Carlo 
simulations, using 8 to 9 years of data shows over 80% similarity in the proportion of crashes for the 0.62 
mi (1 km) road segments along US-191 when looking at the averages of the four analyses. This comparison 
shows that, on average, using a single year of data is approximately 32% proportionately similar to using 
10 years of data. 
 
127 
 
 
Figure 51 Proportion of WVC Data per kilometer of road segments equal to using 13 years of Carcass Data 
 
10.4.2.1. 10-year Carcass Data Analysis 
A total of 52 significant hotspots were identified in the initial 10-year carcass data analysis. The top 10 
hotspots result in a total of 28.58 mi (46 km) of fencing with five distinct mitigated areas: segments 1-4, 
6-22, 65-73, 101-107, and 115-123. A comparison of the four analysis periods with the initial mitigation 
areas is shown in Figure 52. 
In the three analyses conducted following the initial analysis period (2008-2017 data years), the top 10 
hotspots identified were located within the mitigation efforts 96.7% of the time (Table 45). The one hotspot 
not captured in mitigation efforts was 4.97 mi (8 km) away from the fence ends established under the initial 
analysis. 
 
Figure 52 Comparison of WVC Hotspots per kilometer of road segments using 10 years of Carcass Data 
 
 
128 
 
 
Table 45 Comparison of the Top 10 Identified WVC Hotspots using 10 years of Carcass Data 
Top 10 Road Segment Hotspots (km)
Hotspot Rank
2008-2017 2009-2018 2010-2019 2011-2020
1 9 9 9 9
2 70 70 70 104
3 118 104 120 70
4 104 118 104 120
5 68 120 118 72
6 120 1 1 69
7 19 14 7 19
8 1 15 14 1
9 15 81 15 14
10 17 19 19 118
% Covered by Initial Mitigation 90 100 100  
 
10.4.2.2. 5-year Carcass Data Analysis 
A total of 43 significant hotspots were identified in the initial 5-year carcass data analysis. The top 10 
hotspots result in a total of 32.93 mi (53 km) of fencing with seven distinct mitigated areas: segments 3-9, 
11-18, 23-29, 58-64, 66-73, 101-107, and 115-123. A comparison of the nine analysis periods and initial 
mitigation areas is shown in Figure 53. 
In the eight analyses conducted following the initial analysis period (2008-2012 data years), the top 10 
hotspots identified were within the mitigation efforts 76.3% of the time (Table 46). The hotspots that were 
not captured in the mitigation efforts were 0.62-2.49 mi (1-4 km) away from the fence ends established 
under the initial analysis. 
 
Figure 53 Comparison of WVC Hotspots per kilometer of Road Segments using 5 years of Carcass Data 
 
 
 
 
129 
 
 
 
Table 46 Comparison of the Top 10 Identified WVC Hotspots using 5 years of Carcass Data 
Hotspot Top 10 Road Segment Hotspots (km)
Rank 2008-2012 2009-2013 2010-2014 2011-2015 2012-2016 2013-2017 2014-2018 2015-2019 2016-2020
1 6 9 9 9 9 9 70 7 7
2 14 14 14 104 70 70 14 104 70
3 15 15 72 15 118 14 104 70 120
4 70 61 69 19 110 118 7 120 23
5 120 69 15 72 1 111 118 118 111
6 104 6 104 69 66 110 23 23 28
7 69 104 70 1 67 1 77 77 125
8 118 120 61 118 104 104 76 76 126
9 61 1 66 16 7 7 30 14 103
10 26 118 67 17 15 15 78 125 43
% Covered by Initial 
Mitigation 90 100 80 80 70 60 70 60
 
 
10.4.2.3. 3-year Carcass Data Analysis 
A total of 33 significant hotspots were identified in the initial 3-year carcass data analysis. The top 10 
hotspots lead to a total of 29.2 mi (47 km) of fencing with four distinct mitigated areas: segments 3-18, 66-
73, 75-81, and 108-123. A comparison of the 11 analyses with the initial mitigation areas is shown in 
Figure 54. 
In the 10 analyses conducted following the initial analysis period (2008-2012 data years), the top 10 
hotspots identified were located within the mitigation efforts 61% of the time (Table 47). The hotspots that 
were not captured in the mitigation efforts were 1.24-14.29 mi (2-23 km) away from the fence ends 
established under the initial analysis. 
 
Figure 54 Comparison of WVC Hotspots per kilometer of Road Segments using 3 years of Carcass Data 
 
130 
 
 
Table 47 Comparison of the top 10 identified WVC hotspots using 3 years of carcass data 
Hotspot Top 10 Road Segment Hotspots (km)
Rank 2008-2010 2009-2011 2010-2012 2011-2013 2012-2014 2013-2015 2014-2016 2015-2017 2016-2018 2017-2019 2018-2020
1 6 14 14 9 9 9 14 7 7 103 43
2 14 15 15 72 14 82 70 118 70 66 66
3 15 6 117 69 79 104 118 104 108 108 40
4 70 80 61 104 80 118 7 70 103 104 85
5 118 81 104 61 66 14 104 121 23 118 84
6 111 104 89 14 67 70 66 14 76 57 103
7 78 120 26 1 15 66 67 77 120 58 80
8 120 118 27 57 70 67 46 76 86 76 81
9 69 61 79 26 69 69 49 111 111 125 28
10 11 69 72 27 110 57 30 85 126 126 36
% Covered by Initial 
Mitigation 80 50 40 100 70 60 80 60 40 30
 
 
10.4.3. Results Summary 
Table 48 summarizes the results of all the hotspot analyses conducted using the two data types (i.e., crash 
and carcass), including the mitigation fence length (km), the number of mitigated highway sections, and 
the degree to which hotspots covered by the initial mitigation efforts were captured in subsequent data 
years analyzed (Table 48). 
 
Table 48 Summary of Results from the KDE+ Hotspot Analyses 
Distance to 
Number of Mitigation Number of Hotspots within 
Number of Hotspots 
Data Type Significant Fence Length Mitigated Initial 
Data Years Outside of 
Hotspots (km) Sections Mitigation (%)
Mitigation (km)
Crash 10 32 46 5 83.3 2-9
Crash 5 20 47 4 71.3 3-15
Crash 3 9 27 2 40 2-87
Carcass 10 52 46 5 96.7 8
Carcass 5 43 53 7 76.3 1-4
Carcass 3 33 47 4 61 2-23  
 
10.5.  Discussion 
This case study was designed to investigate the variation in WVC hotspots by conducting sequential 
analyses each year and comparing them to an initial analysis used to identify where a road should be 
mitigated for WVCs to increase human safety. The hotspot locations are based on the type of data (i.e., 
crash vs. carcass), the number of years of data used, and the year in which an analysis is conducted. There 
may be large fluctuations in the number of collisions along a road segment each year identified by either 
data type (Figure 47). For each of the analyses using crash and carcass data sets, this case study suggests 
that using more data is supported by the proportions of WVCs along road segments (Figures 48 and 52). It 
shows that using data from shorter time periods (i.e., 3- and 5-year intervals) leads to results that are less 
similar to long-term trends found over 10-year intervals. This means there is higher variability in hotspot 
locations identified in analyses conducted using fewer years of data. More years of data capture hotspots 
 
131 
 
that are persistent over a longer time period and de-emphasize locations where hotspots may only appear 
for a short timeframe. 
The 10-year crash data analysis captures similar hotspots regardless of the year in which an analysis is 
conducted. There are higher densities of hotspots in similar locations year after year, and “new” top 10 
hotspots not captured under initial mitigation efforts are only short distances away from initial fence ends. 
The 10-year carcass data analysis is even more successful at capturing subsequent top 10 hotspots in initial 
mitigation efforts. The carcass data sets also capture weaker hotspots that are more variable from one year 
to the next. There are many more carcass observations than there are crash data locations; crash data focus 
on WVCs that involve reports of property damage or injuries to vehicle passengers. Both data sets capture 
over 80% of the top 10 hotspot locations in subsequent year analyses and are within the priority areas 
identified for US-191 via the more in-depth analysis carried out in the US-191 and MT-64 Wildlife & 
Transportation Assessment. 
The 5-year crash data analysis also identifies the top 10 hotspots in its initial year but only captures 71.3% 
of the top 10 hotspots in subsequent years. This is beneath the 80% similarity target set as a target in the 
case study. Hotspots not captured under initial mitigation measures were 1.24-9.32 mi (2-15 km) away 
from fence ends and, even if 3.1 mi (5 km) mitigation extensions were used (rather than the minimum 1.86 
mi (3 km) used in this case study), would still not capture all hotspots. However, longer mitigation sections 
would surpass the 80% similarity threshold. The initial 5-year carcass data analysis is similar to the crash 
data but captures only 76.3% of subsequent hotspots. Using longer mitigation sections would also lead the 
5-year carcass analysis to meet the 80% threshold. However, so doing would still not capture 100% of 
subsequent year hotspots identified in the 5-year analyses. 
The 3-year analysis suggests how much variation exists in the movement of hotspots from year to year and 
shows how a single year of data can change the distribution of hotspots dramatically. In the initial 3-year 
crash data analysis (2008-2010 data years), hotspots are clustered at the northern end of US-191 (between 
segments 94-132). In the following four analysis years that use crash data from 2009-2014, some top 10 
hotspots are clustered at the southern end (at segments 7-8 and 13-14). The southern hotspots disappear in 
the fifth analysis (2013-2015 data years) but reappear in each additional subsequent year analysis. The 
carcass data follows a similar pattern but captures hotspots more successfully. The initial 3-year carcass 
data analysis captures the southern hotspots along US-191 (at segments 6, 11, 14, and 15). The following 
eight analyses using carcass data between 2009 and 2018 also capture hotspots in the general area of the 
southern section of the road. However, these are not captured in the final two analyses conducted in 2019 
and 2020. This highlights how a large number of collisions recorded in a single year can change hotspots 
identified using a 3-year time frame. Areas with high collision rates within a single year have less impact 
when data sets include longer time frames (i.e., 5-10 years of data). 
10.6. Conclusions 
This analysis shows that the top 10 hotspots along a section of the road vary over time based on the number 
of years of data and the type of data used in the analysis. Crash and carcass data both capture the strongest 
hotspot locations when using 10 years of data; however, because the amount of carcass data is three times 
greater than the amount of crash data, crash data does not capture weaker and more variable hotspots 
located outside of the top 10 hotspot locations. Using carcass data captures the top 10 hotspots and also 
identifies weaker hotspots that are more variable from year to year. 
The quantity of data used in an analysis should reflect desired outcomes. If a goal is the construction of 
WVC mitigation infrastructure to increase human safety long-term, then 10 years of crash or carcass data 
should be analyzed. If identification of changes in hotspot locations due to changes in the landscape, traffic 
conditions, land use, or other factors is a goal, then smaller, 3-year data sets are appropriate. 
 
132 
 
None of the mitigation efforts identified in the initial analyses captured 100% of the top 10 hotspots 
identified in subsequent year timeframes. The best results came from using 10 years of carcass data. The 
lowest percentage of hotspots captured under initial mitigation efforts occurred using 3-year crash data 
sets. Due to high fluctuations in collisions along a road segment from year to year, use of 3-year data sets 
to identify mitigation areas for improving human safety does not capture hotspots that are prevalent over a 
longer time period. Five-year data sets are acceptable and the use of multiple 5-year analyses over a longer 
timeframe ensures that the hotspot locations are captured. Extending the length of fencing in areas 
identified using 5-year data sets increases the similarity to mitigation efforts identified with 10-year data 
sets but requires more fencing to be installed and maintained, increasing overall project cost. 
Based on the 80% similarity target of this case study, using ten years of carcass data in the analysis does 
the best job of capturing the variation of future hotspots through initial mitigation efforts. Using ten years 
of WVC data to identify locations where mitigation efforts should be implemented to increase driver safety 
smooths out the random noise of the weaker temporal hotspots and shows stronger, more robust hotspot 
locations over the long term. 
  
 
133 
 
11. Cost Benefit Analysis 
11.1. Introduction 
This chapter provides a cost-benefit analysis based on public agency data on wildlife carcasses collected 
over a ten-year period from 2011-2020. However, the available data are likely to represent only a fraction 
of the actual number of animals killed by vehicles on US-191 and MT-64 over this time period. In general, 
the underreporting of wildlife crash and carcass data remains a major obstacle in accurately accounting for 
the number of crashes with wildlife and the impact of wildlife-vehicle collisions on wildlife mortality. 
Underreporting is also a complicating factor in the preparation of cost-benefit analyses. Snow et al. (2015) 
summarize the reasons for underreporting identified by academic literature. These include: duration of 
intervals between carcass collection activities; injured animals moving away from roads following 
collisions; carcasses scavenged, dragged off the road, decomposed, or removed for salvage; carcasses out 
of sight (hidden in vegetation, etc.) and not detected, or not a species of concern to departments of 
transportation. The present analysis shows areas where the benefits of implementing measures to reduce 
wildlife-vehicle collisions exceed the costs of their construction based on the available data. However, 
more accurate data may increase the number of locations where the cost-benefit threshold is met. 
Additional information, such as a correction factor for underreporting described in Section 11.4, below, 
could help to highlight additional sites.  
11.2. Methods 
Highway Sections 
US-191: mile marker 0.0 (West Yellowstone) - 81.9 (Four Corners) 
MT-64: mile marker 0.0 (Junction with US-191) – 10.0 (up the mountain) 
11.2.1. Period 
1 Jan 2011-31 Dec 2020 (10 years). 
11.2.2. Carcass Removal Data 
Carcass removal database (Montana Department of Transportation), supplemented by grizzly bear road 
mortalities (Interagency Grizzly Bear Study Team). 
Table 49 Species, Number of Carcasses Recorded, and Percentage of Each Species of the Total Number of 
Recorded Carcasses 
Species Number % 
White-tailed deer 895 67.70 
Mule deer 181 13.69 
Elk 161 12.18 
Moose 29 2.19 
Bison 29 2.19 
Bighorn sheep 17 1.29 
Grizzly bear 5 0.38 
 
134 
 
Species Number % 
Black bear 4 0.30 
Gray wolf 1 0.08 
Total 1322 100.00 
 
11.2.3. Costs of Wildlife-Vehicle Collisions 
The costs of collisions with wildlife used in this analysis are based on those described in Huijser et al. 
(2022) and shown in Table 50. These costs include direct costs associated with vehicle repair, human 
injuries, and human fatalities, as well as passive use values. Passive use or non-use values are the values 
individual people place on the existence of a given animal species or population, as well as the bequest 
value of knowing that future generations will also benefit from preserving the species (Duffield and Neher 
2019). 
Table 50 Total Costs Associated with Wildlife-Vehicle Collisions (in 2020 US$) 
 Costs per collision 
Cost category Deer Elk Moose Gray Grizzly 
wolf bear 
Direct costs      
Vehicle repair $4,418 $7,666 $9,435 $4,418 $4,418 
Human injuries $6,116 $14,579 $26,811 $6,116 $6,116 
Human fatalities $3,480 $23,200 $46,400 $3,480 $3,480 
Subtotal $14,014 $45,445 $82,646 $14,014 $14,014 
Passive use value $5,075 $27,751 $27,751 $40,342 $4,235,770 
Total $19,089 $73,196 $110,397 $54,356 $4,249,784 
 
In order to carry out cost-benefit analyses of potential wildlife accommodation measures along US-191 and 
MT-64 within the study area, the carcasses recorded by the Montana Department of Transportation and 
Interagency Grizzly Bear Study Team (see Table 49), were categorized based on their size and weight to 
match species for which economic data are available. The species for which economic data are available 
are shown in Table 50. In order to take advantage of this information, some species were recategorized; 
for example, bighorn sheep were categorized as deer, as shown in Table 51. As a result, the potential 
passive use values for several species, including bison, bighorn sheep, and black bears, are likely severely 
underestimated. 
Table 51 Species Reported in Carcass Data as Categorized for Cost-benefit Analyses 
Species Reported in Carcass Data “Species” categorized in Cost-benefit Analyses 
White-tailed deer Deer 
Mule deer Deer 
 
135 
 
Species Reported in Carcass Data “Species” categorized in Cost-benefit Analyses 
Elk Elk 
Moose Moose 
Bison Moose 
Bighorn sheep Deer 
Grizzly bear Grizzly bear 
Black bear Deer 
Gray wolf Gray wolf 
 
11.2.4. Costs of Mitigation Measures 
The cost of a large mammal fence (with buried apron), an underpass once every 2 km, and jump-outs is 
$25,388 per km per year ($40,858 per mile per year) at a discount rate of 3% (Huijser et al. 2022). The 
total cost of including an additional wildlife overpass once every 24 km instead of an underpass is $32,030 
per km per year ($51,547 per mile per year) (Huijser et al. 2022). 
11.3. Results 
11.3.1. Cost-Benefit of Wildlife-Vehicle Collisions along US-191 and MT-64 
The number of wildlife-vehicle collisions along US-191 and MT-64 per species was calculated for each 
0.1-mile road segment. However, the number was averaged over a 1.1-mile road segment (including five 
0.1-mile segments just before and five 0.1-mile segments just after each 0.1-mile segment as a moving 
window). Since the carcass data related to a 10-year period, the number of carcasses per species was then 
divided by 10 to obtain the number of roadkilled animals per mile per year for each tenth of a mile road 
segment. The number of roadkilled animals per species was then multiplied by the average costs for a 
collision with that species (see Table 50). Finally, the costs for the different species were summed into the 
total costs per mile per year for each 0.1-mile road segment. 
In Figure 55, which shows US-191 in the study area, the costs of wildlife-vehicle collisions based on carcass 
counts are represented by the jagged line (in 2020 US$). The two horizontal lines are the thresholds for 
two different types and combinations of mitigation measures at which benefits exceed cost over a 75-year 
period with a discount rate of 3%. 
In Figure 56, which shows MT-64 in the study area, the costs of wildlife-vehicle collisions based on carcass 
counts are represented by the jagged line (in 2020 US$). The two horizontal lines are the thresholds for 
two different types and combinations of mitigation measures at which benefits exceed cost over a 75-year 
period with a discount rate of 3%. 
 
 
136 
 
 
Figure 55 US-191 from West Yellowstone (left side of graph) to Four Corners (right side of graph). 
 
 
Figure 56 MT-64 from the junction with US-191 (left side of graph) to the end of the road up the mountain (right 
side of graph). 
 
137 
 
11.4. Discussion 
In the study area, the costs associated with wildlife-vehicle collisions spike where grizzly bear road 
mortalities occur. However, the total number of carcasses of all species is low compared to the quantity 
that would meet or exceed the thresholds for the different types and combinations of mitigation measures 
in most locations. However, it is important to remember that reported carcass removal data are likely to be 
an undercount, such that the total number of animals that have been hit may be significantly higher (see 
Section 2.4 for more information on underreporting of carcass data).  
It is important to note that just one or a few more grizzly bears hit along a road section would substantially 
affect results. Moreover, while passive use values for wildlife were integrated into the analyses, the model 
does not necessarily capture the full economic value of wildlife and maintaining viable populations. 
Nevertheless, three areas along US-191 meet or exceed the threshold. Two of these areas are located within 
priority areas identified in this report: the Porcupine and the Taylor Fork Priority Sites. The Porcupine site, 
for which an overpass is recommended, exceeds the economic threshold for implementing this type of 
structure. The Taylor Fork site exceeds the threshold for the construction of underpasses and fencing. 
Because grizzly bears are among the target species in these locations, special consideration should be given 
to structures that accommodate their movements. Grizzly bear family groups require open structures such 
as overpasses and large-span bridges (Ford et al. 2017). The third area is located at RM 0.8, adjacent to 
West Yellowstone, and was identified primarily due to a recorded grizzly bear mortality. As stated earlier 
in Section 6.10, mitigation along other areas of the study roads may be warranted. It should also be 
considered based on the documentation and analysis in this report, especially when highway projects are 
planned. 
To prepare a more thorough cost-benefit metric, the CLLC-WTI research team could take a “correction 
factor” based on a study of how available carcass data may relate to the actual number of wildlife struck 
and killed by vehicles (Lee et al. 2021). This type of study is now underway in Montana, and it may be 
appropriate to use the correction factors determined once available, especially given the difficulty of data 
collection along some sections of US-191 and MT-64 (where there are few opportunities to pull over 
safely). Once appropriate correction factors have been determined, this chapter's cost-benefit analyses may 
be updated to provide greater insight into the potential economic benefits of implementing measures to 
reduce wildlife-vehicle collisions.  
  
 
138 
 
12. Conclusion 
The information included in this report should inform and support area communities and agency decision-
makers to select and pursue wildlife accommodation options. With the passage of the Infrastructure 
Investment and Jobs Act of 2021, significant funds for wildlife accommodation measures are available 
nationwide on a competitive basis. The US-191/MT-64 Wildlife & Transportation Assessment better 
equips part of Southwest Montana’s gateway to Yellowstone National Park to take advantage of new 
funding opportunities. 
The Assessment compiled, overlaid, and evaluated wildlife-vehicle collision and wildlife carcass data, 
wildlife movement and habitat data, and live wildlife observations from aerial surveys. It also incorporated 
wildlife sightings and roadkill information gathered via citizen science. The Assessment further drew upon 
local and expert knowledge gathered through in-person outreach and an interactive map. Road areas 
identified through an in-depth spatial data analysis were evaluated in a field review conducted by an 
interdisciplinary Technical Advisory Committee of county, state, and federal planners, along with 
biologists, engineers, transportation experts, and the Research Team from the Center for Large Landscape 
Conservation (CLLC) and Montana State University’s Western Transportation Institute (WTI). 
The report describes eleven Priority Sites located in important areas for wildlife movement and/or that pose 
elevated risks to human and wildlife safety and provides recommendations for potential wildlife 
accommodation measures. Mitigation along other areas of the study roads may also be warranted based on 
the documentation and analysis in this report, especially when highway projects are planned. 
In addition to the terrestrial Priority Sites, the report summarizes the potential “barrier effect” of culverts 
passing under US-191 and MT-64 based on an assessment of factors in the field, finding that over 40% of 
existing culverts may not allow for the unimpeded passage of fish. While retaining a barrier to passage 
may be desirable in some cases to protect the integrity of native species, further evaluation is suggested to 
assess the status of these culverts more fully. 
The report's recommendations describe appropriate locations for prospective wildlife accommodation 
measures such as culverts, bridges, underpasses and overpasses, and/or animal detection systems—each in 
combination with fencing—that consider terrestrial and/or aquatic wildlife passage. Many sites include 
major drainages from surrounding public lands that intersect with US-191 or MT-64 and feature existing 
infrastructure that has the potential to facilitate animal movements, such as a bridge spanning a riparian 
corridor. During the field evaluation, the CLLC-WTI research team and Technical Advisory Committee 
considered means to incorporate existing infrastructure, new structures, and additional alternatives to 
reduce collisions with wildlife (e.g., variable message signs for areas that have spatially discreet or seasonal 
conflicts and traffic calming measures to effectively reduce the design speed of the highway). A range of 
accommodation strategies have a role to play in helping to reduce collisions and maintain wildlife 
movement in the study area. 
In the case of terrestrial species, despite crossing structures with fencing requiring high initial investment, 
research shows they are cost-effective over the course of their lifetime (generally 75 years or more) due to 
greater efficacy in reducing wildlife-vehicle collisions and lower maintenance costs than other options 
(Brennan, Chow, and Lamb 2022; Huijser et al. 2009). Further, given that bridges and culverts that are 
upscaled and designed to allow for wildlife passage are usually better able to accommodate stream and 
floodplain function due to larger size and capacity, they may also make infrastructure more resilient to 
“extreme” weather events like flooding. The Montana Department of Transportation has already identified 
several bridges needing replacement in the study area. Applying the findings of this Assessment to bridges 
or other priority locations when replacement is scheduled offers a best-case scenario for cost-effective 
implementation.  
 
139 
 
The analysis within this report is based on available data. Not all data sets are comprehensive; some are 
collected opportunistically (including both Montana Department of Transportation and citizen science 
carcass data), and the data sets are skewed towards large mammals and charismatic species such as elk and 
grizzly bears. Some data gaps and limitations that could be aided by further research are mentioned in the 
discussion of Priority Sites.  
The options for prospective wildlife accommodation measures along key road segments described in this 
report are intended as a guide to inform decision-making processes rather than serve as a prescription for 
specific actions. Implementation of any prospective measure depends on factors such as public support, 
design, engineering feasibility, potential agreements with land management agencies and/or private 
landowners, and funding availability.  
The CLLC-WTI Research Team suggests that making US-191 and MT-64 safer for travelers and wildlife 
is a multi-year, multi-site, multi-stakeholder proposition that will take collective action to bring about. The 
Assessment provides a foundation to allow for discussion about how to reach these goals based on robust 
understanding. 
 
  
 
140 
 
13. Literature Cited 
Abra, F, B Granziera, M Huijser, K Ferraz, C Haddad, and R Paolino. 2019. “Pay or Prevent? Human 
Safety, Costs to Society and Legal Perspectives on Animal-Vehicle Collisions in São Paulo State, 
Brazil.” PLOS ONE 14 (4): e0215152. https://doi.org/10.1371/journal.pone.0215152. 
Al-Ghamdi, A. 2004. “Warning Signs as Countermeasures to Camel–Vehicle Collisions in Saudi 
Arabia.” Elsevier 36: 749–60. https://doi.org/10.1016/j.aap.2003.05.006. 
Andrášik, R, and M Bíl. 2015. “Traffic Accidents: Random or Pattern Occurrence?” Safety and 2 
Reliability of Complex Engineered Systems (ESREL 2015), 7–10. 
Andresen, M. 2009. “Testing for Similarity in Area-Based Spatial Patterns: A Nonparametric Monte 
Carlo Approach.” Applied Geography 29 (3): 333–45. 
Baker, C, and F Votapka. 1990. “Fish Passage through Culverts.” FHWA-FL-09-006. US Department 
of Transportation. Federal Highway Administration. Washington, DC. 
Barnard, R, S Yokers, A Nagygyor, and T Quinn. 2015. “An Evaluation of the Stream Simulation 
Culvert Design Method in Washington State.” River Research and Applications 31 (10): 1376–87. 
Baxter, C, and F Hauer. 2000. “Geomorphology, Hyporheic Exchange, and Selection of Spawning 
Habitat by Bull Trout (Salvelinus Confluentus).” Canadian Journal of Fisheries and Aquatic Sciences 
57 (7): 1470–81. 
Belford, D, and W Gould. 1989. “An Evaluation of Trout Passage through Six Highway Culverts in 
Montana.” North American Journal of Fisheries Management 9 (4): 437–45. 
Bíl, M, R Andrášik, M Duľa, and J Sedoník. 2019. “On Reliable Identification of Factors Influencing 
Wildlife-Vehicle Collisions along Roads.” Journal of Environmental Management 237: 297–304. 
Bíl, M, R Andrášik, and Z Janoška. 2013. “Identification of Hazardous Road Locations of Traffic 
Accidents by Means of Kernel Density Estimation and Cluster Significance Evaluation.” Accident 
Analysis & Prevention 55: 265–73. 
Blackwell, B, T DeVault, and T Seamans. 2015. “Understanding and Mitigating the Negative Effects 
of Road Lighting on Ecosystems.” Van der Ree, R, D Smith, and C Grilo (eds). Handbook of Road 
Ecology, 143–50. 
Brennan, L, E Chow, and C Lamb. 2022. “Wildlife Overpass Structure Size, Distribution, 
Effectiveness, and Adherence to Expert Design Recommendations.” PeerJ 10: e14371. 
Brock, B. “Bison Winter Habitat Suitability Index.” Unpublished. 
Brock, B, E Atkinson, C Groves, A Toivola, T Olenicki, and L Craighead. 2006. “A Wildlife 
Conservation Assessment of the Madison Valley, Montana.” Wildlife Conservation Society. 
Bozeman, MT. 
Bullock, K, G Malan, and M Pretorius. 2011. “Mammal and Bird Road Mortalities on the Upington 
to Twee Rivieren Main Road in the Southern Kalahari, South Africa.” African Zoology 46 (1): 60–
71. 
Burford, D, T McMahon, J Cahoon, and M Blank. 2009. “Assessment of Trout Passage through 
Culverts in a Large Montana Drainage during Summer Low Flow.” North American Journal of 
Fisheries Management 29 (3): 739–52. 
 
141 
 
Cahoon, J, T McMahon, O Stein, M Blank, and D Burford. 2005. “Fish Passage at Road Crossings in 
a Montana Watershed.” FHWA/MT-05-002/8160. Montana Department of Transportation. Helena, 
MT. 
Clevenger, A, and M Barrueto. 2014. “Trans-Canada Highway Wildlife and Monitoring Research, 
Final Report. Part B: Research.” Prepared for Parks Canada Agency, Radium Hot Springs, BC. 
Clevenger, A, and M Huijser. 2011. “Wildlife Crossing Structure Handbook, Design and Evaluation 
in North America.” Report to US Department of Transportation, Federal Highway Administration. 
Washington, DC. 
Colorado Department of Transportation. 2014. “Final Report – HB 1238 Wildlife Crossing Zones.” 
Conover, M, W Pitt, K Kessler, T DuBow, and W Sanborn. 1995. “Review of Human Injuries, 
Illnesses, and Economic Losses Caused by Wildlife in the United States.” Wildlife Society Bulletin, 
407–14. 
Coulson, G. 1982. “Road-Kills of Macropods on a Section of Highway in Central Victoria.” Wildlife 
Research 9 (1): 21–26. 
Craighead, L. 2006. “Grizzly Bear Habitat Suitability Model.” 
Cramer, P. 2022. “The Strategic Integration of Wildlife Mitigation into Transportation Procedures.” 
Transportation Pooled Fund Study, TPF-5(358). Nevada Department of Transportation. Carson City, 
NV.  
Creech, T, E Fairbank, A Clevenger, R Callahan, and R Ament. 2019. “Differences in Spatiotemporal 
Patterns of Vehicle Collisions with Wildlife and Livestock.” Environmental Management 64: 736–
45. 
Creech, T, M McClure, and R Callahan. 2016. “High-Risk Zones for Ungulate-Vehicle Collisions 
during Montana’s Fall Migration Season.” The Center for Large Landscape Conservation. Bozeman, 
MT. 
Cushman, S, J Lewis, and E Landguth. 2013. “Evaluating the Intersection of a Regional Wildlife 
Connectivity Network with Highways.” Movement Ecology 1: 1–11. 
Dai, Q, R Young, and S Vander Giessen. 2009. “Evaluation of an Active Wildlife-Sensing and Driver 
Warning System at Trapper’s Point.” Wyoming Department of Transportation. Cheyenne, WY. 
Dickson, B, C Albano, B McRae, J Anderson, D Theobald, L Zachmann, T Sisk, and M Dombeck. 
2017. “Informing Strategic Efforts to Expand and Connect Protected Areas Using a Model of 
Ecological Flow, with Application to the Western United States.” Conservation Letters 10 (5): 564–
71. https://doi.org/10.1111/conl.12322. 
Dietrich, E. 2022. “Montana’s Fastest Growing City Last Year? It Wasn’t Bozeman.” Montana Free 
Press, March 3, 2022. 
Dodd, N, J Gagnon, S Boe, and R Schweinsburg. 2007. “Role of Fencing in Promoting Wildlife 
Underpass Use and Highway Permeability.” Proceedings of the 2007 International Conference on 
Ecology and Transportation. Irwin, C, D Nelson, and K McDermott (eds). Center for Transportation 
and the Environment, North Carolina State University. Raleigh, NC. 
 
142 
 
Donaldson, B. 2017. “Improving Animal-Vehicle Collision Data for the Strategic Application of 
Mitigation.” Virginia Transportation Research Council/Virgina Department of Transportation. 
Charlotte, VA. 
Donnell, E, K Kersavage, and L Tierney. 2018. “Self-Enforcing Roadways: A Guidance Report.” US 
Department of Transportation, Federal Highway Administration. McLean, VA. 
Duffield, J, and C Neher. 2019. “Cost Effective Solutions: Incorporating Wildlife Passive Use Values 
in Collision Mitigation Benefit-Cost Calculations.” Transportation Pooled Fund Study, TPF-5(358). 
Nevada Department of Transportation. Carson City, NV. 
Dupree, A, and I Dimambro. 2012. “Development of Mitigation Options for Bison-Vehicle Collisions 
on US 191.” MSU-Western Transportation Institute. Bozeman, MT. 
Edwards, R. 1998. “The Hyporheic Zone.” Naiman, R and  R Bilby (eds). River Ecology and 
Management: Lessons from the Pacific Coastal Ecoregion, 399–429.  
Elvik, R. 2014. “Speed and Road Safety-New Models.” TØI Report, no. 1296/2014. 
Favilli, F, M Bíl, J Sedoník, R Andrášik, P Kasal, A Agreiter, and T Streifeneder. 2018. “Application 
of KDE+ Software to Identify Collective Risk Hotspots of Ungulate-Vehicle Collisions in South 
Tyrol, Northern Italy.” European Journal of Wildlife Research 64: 1–12. 
Federal Highway Administration. 2022. “Building a Better America Guidebook.”  
Fitch, G. 1995. “Nonanadromous Fish Passage in Highway Culverts.” Virginia Transportation 
Research Council/Virginia Department of Transportation. Charlotte, VA. 
Fitzpatrick, K. 2003. Design Speed, Operating Speed, and Posted Speed Practices. Vol. 38. 
Transportation Research Board. Washington, DC. 
Ford, A, M Barrueto, and A Clevenger. 2017. “Road Mitigation Is a Demographic Filter for Grizzly 
Bears.” Wildlife Society Bulletin 41 (4): 712–19. 
Future West. 2019. “Sustaining the New West II: Bold Visions, Inspiring Actions Conference.” 
Bozeman MT. 
Gagnon, J, N Dodd, S Sprague, K Ogren, and R Schweinsburg. 2010. “Preacher Canyon Wildlife 
Fence and Crosswalk Enhancement Project Evaluation: State Route 260.” Arizona Game and Fish 
Department. Phoenix, AZ. 
Gehman, S. 2010. “Wildlife of the Gallatin Mountains, Southcentral Montana.” The Wildlife Society, 
2010. 
Gibson, R, R Haedrich, and C Wernerheim. 2005. “Loss of Fish Habitat as a Consequence of 
Inappropriately Constructed Stream Crossings.” Fisheries 30 (1): 10–17. 
Giller, P. 1998. The Biology of Streams and Rivers. Oxford University Press. 
Hardy, A, S Willer, and E Williamson. 2008. “A Preliminary Assessment of Wildlife-Transportation 
Issues in the Greater Yellowstone Ecosystem.” Prepared for the Greater Yellowstone Coalition. 
Bozeman, MT. 
Hoeppe, P. 2016. “Trends in Weather Related Disasters – Consequences for Insurers and Society.” 
Weather and Climate Extremes 11: 70–79. https://doi.org/10.1016/j.wace.2015.10.002. 
 
143 
 
Hou, Q, X Huo, and J Leng. 2020. “A Correlated Random Parameters Tobit Model to Analyze the 
Safety Effects and Temporal Instability of Factors Affecting Crash Rates.” Accident Analysis & 
Prevention 134: 105326. 
Huang, B, Y Zhang, J Lu, and L Lu. 2013. “A Simulation Study for Minimizing Operating Speed 
Variation of Multilane Highways by Controlling Access.” Procedia-Social and Behavioral Sciences 
96: 2767–81. 
Huijser, M, K Gunson, and C Abrams. 2006. “Animal-Vehicle Collisions and Habitat Connectivity 
along Montana Highway 83 in the Seeley-Swan Valley, Montana: A Reconnaissance.” Montana 
Department of Transportation. Helena, MT. 
Huijser, M. 2008. “Wildlife Vehicle Collision Reduction Study: Report to Congress.” US Department 
of Transportation, Federal Highway Administration. McLean, VA. 
Huijser, M, J Duffield, A Clevenger, R Ament, and P McGowen. 2009. “Cost–Benefit Analyses of 
Mitigation Measures Aimed at Reducing Collisions with Large Ungulates in the United States and 
Canada: A Decision Support Tool.” Ecology and Society 14 (2). 
Huijser, M, E Fairbank, W Camel-Means, J Graham, V Watson, P Basting, and D Becker. 2016. 
“Effectiveness of Short Sections of Wildlife Fencing and Crossing Structures along Highways in 
Reducing Wildlife–Vehicle Collisions and Providing Safe Crossing Opportunities for Large 
Mammals.” Biological Conservation 197: 61–68. https://doi.org/10.1016/j.biocon.2016.02.002. 
Huijser, M, C Mosler-Berger, M Olsson, and M Strein. 2015. “Wildlife Warning Signs and Animal 
Detection Systems Aimed at Reducing Wildlife‐vehicle Collisions.” Van der Ree, R, D Smith, and C 
Grilo (eds). Handbook of Road Ecology, 198–212.  
Huijser, M, W Camel-Means, E Fairbank, J.P Purdum, T Allen, A Hardy, J Graham, J Begley, P 
Basting, and D Becker. 2016. “US 93 North Post-Construction Wildlife-Vehicle Collision and 
Wildlife Crossing Monitoring on the Flathead Indian Reservation between Evaro and Polson, 
Montana.” FHWA/MT-16-009/8208. Montana Department of Transportation. Helena, MT. 
Huijser, M, E Fairbank, W Camel-Means, J Graham, V Watson, P Basting, and D Becker. 2016. 
“Effectiveness of Short Sections of Wildlife Fencing and Crossing Structures along Highways in 
Reducing Wildlife–Vehicle Collisions and Providing Safe Crossing Opportunities for Large 
Mammals.” Biological Conservation 197: 61–68. 
Huijser, M, E Fairbank, and F Abra. 2017. “The Reliability and Effectiveness of a Radar-Based 
Animal Detection System.” US Department of Transportation. Washington, DC. 
Huijser, M, M Blank, R Ament, J.S. Begley, and E.R. Jenne. 2018. “Teton County Wildlife Crossings 
Master Plan- Final Report.” Teton County, WY. 
Huijser, M, and J.S. Begley. 2019. “Large Mammal-Vehicle Collision Hot Spot Analyses, California, 
USA. Report 4W6693.” California Department of Transportation. Sacramento, CA. 
Huijser, M, R Ament, M Bell, A Clevenger, E Fairbank, K Gunson, and T McGuire. 2021. “Animal 
Vehicle Collision Reduction and Habitat Connectivity Pooled Fund Study: Literature Review.” 
Nevada Department of Transportation. Carson City, NV. 
Huijser, M, E Fairbank, and K Paul. 2022. “Best Practices Manual to Reduce Animal-Vehicle 
Collisions and Provide Habitat Connectivity for Wildlife.” Transportation Pooled Fund Study, TPF-
5(358). Nevada Department of Transportation. Carson City, NV.  
 
144 
 
Huijser M, Duffield J, Neher C, Clevenger A, Mcguire T (eds). Final Report 2022: Update and 
expansion of the WVC mitigation measures and their cost-benefit model. Transportation Pooled Fund 
Study, TPF-5(358). Nevada Department of Transportation. Carson City, NV. 
Johns, T, and T Ernst. 2007. “Culvert Crossings as Potential Barriers to Fish Movement in the Kakwa 
River Watershed, Alberta─ Data Report.” D-2007-001. Alberta Conservation Association, Peace 
River, AB. 
Jones, J, F Swanson, B Wemple, and K Snyder. 2000. “Effects of Roads on Hydrology, 
Geomorphology, and Disturbance Patches in Stream Networks.” Conservation Biology 14 (1): 76–
85. 
Karle, K. 2005. “Analysis of an Efficient Fish Barrier Assessment Protocol for Highway Culverts.” 
US Department of Transportation, Washington, DC. 
Kilgore, R, B Bergendahl, and R Hotchkiss. 2010. “Culvert Design for Aquatic Organism Passage: 
Hydraulic Engineering Circular Number 26.” Federal Highway Administration. Central Federal 
Lands Highway Division, Lakewood CO. 
Knaepkens, G, E Verheyen, P Galbusera, and M Eens. 2004. “The Use of Genetic Tools for the 
Evaluation of a Potential Migration Barrier for the Bullhead.” Journal of Fish Biology 64 (6): 1737–
44. 
Krosby, M, D Theobald, R Norheim, and B McRae. 2018. “Identifying Riparian Climate Corridors 
to Inform Climate Adaptation Planning.” PLoS One 13 (11): e0205156. 
Lachance, S, M Dubé, R Dostie, and P Bérubé. 2008. “Temporal and Spatial Quantification of Fine-
Sediment Accumulation Downstream of Culverts in Brook Trout Habitat.” Transactions of The 
American Fisheries Society 137: 1826–38. 
Lawler, J, S Shafer, D White, P Kareiva, E Maurer, A Blaustein, and P Bartlein. 2009. “Projected 
Climate‐induced Faunal Change in the Western Hemisphere.” Ecology 90 (3): 588–97. 
https://doi.org/10.1890/08-0823.1. 
Lee, T, Rondeau, K, Schaufele, R, Clevenger, A, and D Duke. 2021. Developing a correction factor 
to apply to animal–vehicle collision data for improved road mitigation measures. Wildlife Research 
48 (6): 501-510. 
Max, S. 2017. “A Ski Resort Grows Under an Expansive Montana Sky.” The New York Times. 
January 10, 2017. 
McDonald, L, T Messmer, and M Guttery. 2019. “Temporal Variation of Moose–Vehicle Collisions 
in Alaska.” Human–Wildlife Interactions 13 (3): 8. 
McDonald, M. 1991. “Moose Movement and Mortality Associated with the Glenn Highway 
Expansion, Anchorage, Alaska.” Alces 27: 208–19. 
McKay, S, J Schramski, J Conyngham, and J Fischenich. 2013. “Assessing Upstream Fish Passage 
Connectivity with Network Analysis.” Ecological Applications 23 (6): 1396–1409. 
Meyer, E. 2006. “Assessing the Effectiveness of Deer Warning Signs.” Kansas Department of 
Transportation. Topeka, KS. 
Minnesota Department of Transportation. 2011. “Deer Detection and Warning System.” St. Paul, 
MN. 
 
145 
 
Montana Department of Transportation. 2020. “US-191 Corridor Study (Four Corners to Beaver 
Creek).” Helena, MT. 
———. 2022. “Statewide Transportation Improvement Program.” Helena, MT. 
Morris, B, M Bolding, W Aust, K McGuire, E Schilling, and J Sullivan. 2016. “Differing Levels of 
Forestry Best Management Practices at Stream Crossing Structures Affect Sediment Delivery and 
Installation Costs.” Water 8 (3): 92. 
Mosler-Berger, C, and J Romer. 2003. “Wildwarnsystem CALSTROM.” Wildbiologie 3: 1–12. 
Myers, R, and S Nieraeth. 2016. “Restoration and Coho Salmon Carrying Capacity of Chester Creek, 
an Urban Stream in Anchorage, Alaska.” Northwest Science 90 (2): 146–58. 
https://doi.org/10.3955/046.090.0207. 
National Center for Statistics and Analysis. 2022. “Traffic Safety Facts 2020.” 
National Park Service. 2022. “NPS Statistics, Integrated Resource Management Applications 
(IRMA).”  
———. 2023. “Greater Yellowstone Ecosystem.”  
Ouyang, Y, Z Jiang, and K Jadaan. 2016. “Speed Harmonization-Design Speed vs. Operating Speed.” 
Illinois Department of Transportation. Springfield, IL. 
Pashby, J. 2022. “Stymied.” Montana Outdoors, March 2022. 
Peck, C, F Van Manen, C Costello, M Haroldson, L Landenburger, L Roberts, D Bjornlie, and R 
Mace. 2017. “Potential Paths for Male‐mediated Gene Flow to and from an Isolated Grizzly Bear 
Population.” Ecosphere 8 (10): e01969. 
Pojar, T, R Prosence, D Reed, and T Woodard. 1975. “Effectiveness of a Lighted, Animated Deer 
Crossing Sign.” The Journal of Wildlife Management, 87–91. 
Policom. 2023. “Economic Strength Rankings.” 2023. 
Riginos, C. 2022. “Impact of Roadways on Wildlife in Wyoming: Long-Term and Recent Trends.” 
The Nature Conservancy. Lander, WY. 
Riginos, C, E Fairbank, E Hansen, J Kolek, and M Ham. 2019. “Effectiveness of Night-Time Speed 
Limit Reduction in Reducing Wildlife-Vehicle Collisions.” Wyoming Department of Transportation. 
Cheyenne, WY. 
Riginos, C, C Smith, E Fairbank, E Hansen, and P Hallsten. 2018. “Traffic Thresholds in Deer Road-
Crossing Behavior.” Wyoming Dept. of Transportation. Cheyenne, WY. 
Riley, S, and A Marcoux. 2006. “Deer-Vehicle Collisions: An Understanding of Accident 
Characteristics and Drivers’ Attitudes, Awareness and Involvement.” Michigan Department of 
Transportation. Lansing, MI. 
Santos, R, F Ascensão, M Ribeiro, A Bager, M Santos-Reis, and L Aguiar. 2017. “Assessing the 
Consistency of Hotspot and Hot-Moment Patterns of Wildlife Road Mortality over Time.” 
Perspectives in Ecology and Conservation 15 (1): 56–60. 
Sawyer, H, P Rodgers, and T Hart. 2016. “Pronghorn and Mule Deer Use of Underpasses and 
Overpasses along US Highway 191.” Wildlife Society Bulletin 40 (2): 211–16. 
 
146 
 
Seiler, A. 2003. “The Toll of the Automobile - Wildlife and Roads in Sweden.” Doctoral Thesis. 
Swedish University of Agricultural Sciences. Uppsala. 
Shamon, H, O Cosby, C Andersen, H Augare, J BearCub Stiffarm, C Bresnan, B Brock, et al. 2022. 
“The Potential of Bison Restoration as an Ecological Approach to Future Tribal Food Sovereignty on 
the Northern Great Plains.” Frontiers in Ecology and Evolution 10. 
https://doi.org/10.3389/fevo.2022.826282. 
Sharafsaleh, M, M Huijser, C Nowakowski, M Greenwood, L Hayden, J Felder, and M Wang. 2012. 
“Evaluation of an Animal Warning System Effectiveness Phase Two- Final Report.” California 
Department of Transportation. Sacramento, CA. 
Six Rivers National Forest. 2012. “FishXing Software v2.2.” USDA Forest Service.  
Snow, N, Porter, W, and Williams, D. 2015. Underreporting of wildlife-vehicle collisions does not 
hinder predictive models for large ungulates. Biological Conservation (181) 44-53. 
State Farm. 2023. “How Likely Are You to Have an Animal Collision?”  
Strein, M. 2010. “Restoring Permeability of Roads for Wildlife: Wildlife Warning Systems in 
Practice.” Programme and Book of Abstracts. IENE International Conference on Ecology and 
Transportation: Improving Connections in a Changing Environment, 27:77. 
Sullivan, T, A Williams, T Messmer, L Hellinga, and S Kyrychenko. 2004. “Effectiveness of 
Temporary Warning Signs in Reducing Deer‐vehicle Collisions during Mule Deer Migrations.” 
Wildlife Society Bulletin 32 (3): 907–15. 
Taylor, R, and M Love. 2002. “Fish Passage Evaluation at Stream Crossings.” 
Tchir, J, P Hvenegaard, and G Scrimgeour. 2004. “Stream Crossing Inventories in the Swan and 
Notikewin River Basins of Northwest Alberta: Resolution at the Watershed Scale.” Proceedings of 
the Forest Land-Fish Conference II. Alberta Conservation Association, Edmonton, Canada, 53–62. 
US Census Bureau. 2023. “Quick Facts.”  
US Department of Transportation. 2023. “The Rural Surface Transportation Grant.” 
US Fish and Wildlife Service. 2002. “Determination of Endangered Status for the Carson Wandering 
Skipper.” Federal Register 67. 
USDA National Technology Development Program. 2008. “Stream Simulation: An Ecological 
Approach to Providing Passage for Aquatic Organisms at Road-Stream Crossings.” San Dimas, CA. 
Votapka, F. 1991. “Considerations for Fish Passage through Culverts.” Transportation Research 
Record 1291. 
Waller, J, and C Miller. 2015. “Decadal Growth of Traffic Volume on US Highway 2 in Northwestern 
Montana.” Intermountain Journal of Sciences 21 (1–4): 29–37. 
Waller, J, and C Servheen. 2005. “Effects of Transportation Infrastructure on Grizzly Bears in 
Northwestern Montana.” The Journal of Wildlife Management 69 (3): 985–1000.  
Wang, Y, Y Lao, Y Wu, and J Corey. 2010. “Identifying High Risk Locations of Animal-Vehicle 
Collisions on Washington State Highways.” WSDOT Research Rep. WA-RD 752. 
 
147 
 
Wanvik, P. 2009. “Effects of Road Lighting: An Analysis Based on Dutch Accident Statistics 1987–
2006.” Accident Analysis & Prevention 41 (1): 123–28. 
Warren Jr, M, and M Pardew. 1998. “Road Crossings as Barriers to Small-Stream Fish Movement.” 
Transactions of the American Fisheries Society 127 (4): 637–44. 
Whatcom County Public Works. 2006. “Whatcom County Fish Passage Barrier Inventory Final 
Report.” Bellingham, Washington. 
 
 
 
 
 
 
148