FINAL REPORT
Prepared for Meg and Bert Raynes Wildlife Fund
January 2021
EXECUTIVE SUMMARY
Outdoor recreation helps connect people to nature, promotes healthy lifestyles, and contributes to thriving local economies. However, recreation can also have negative effects on wildlife if it is not well-managed. In the Jackson Hole area, recreation is expanding rapidly, and managers need accurate and localized information about recreation and wildlife trends to plan for current and future use. The Nature Conservancy (TNC) in Wyoming’s project objectives are to monitor wildlife and human activity in the Cache-Game area of the Bridger-Teton National Forest to establish a baseline for trends, and to involve volunteers and citizen scientists to increase public appreciation of nature and awareness of the issue of recreation and wildlife.
Using funds from the Meg and Bert Raynes Wildlife Fund, we
purchased 10 trail cameras and deployed them, along with other trail cameras, along trails in the Cache-Game area. The cameras were placed strategically to capture variability in human use, including trails with different intensities of use and areas where we expect human use will increase in the future. The cameras collected over 500,000 images in 2020 and included pictures of a broad diversity of wildlife species as well as mountain bikers, hikers, and skiers. To process the images, we uploaded them to the citizen science platform Zooniverse and recruited and trained at least 66 volunteers from the Jackson Hole Community School, Nature Mapping program, and outside the Jackson area to tag the images. To date, over 200,000 classifications have been made on the Zooniverse platform.
Ongoing work on this project involves continuing to process and sort images from the trail cameras with the help of citizen scientists and Artificial Intelligence (AI) machine learning processing. Data garnered from image processing will inform our analyses of wildlife habitat use and activity patterns in relation to recreation. We are also gearing up for the 2021 data collection season.
Photo: Elk captured in a trail camera image from May 2020
Monitoring wildlife
and recreation use
on the Bridger-Teton
National Forest
PRINCIPAL
INVESTIGATORS
COURTNEY LARSON &
TREVOR BLOOM, THE NATURE
CONSERVANCY IN WYOMING
PROJECT
PARTNERS
KATE GERSH, JACKSON HOLE
WILDLIFE FOUNDATION
TIM FERRIS & LINDA
MERIGLIANO, U.S. FOREST SERVICE CHRIS OWENS, FRIENDS OF
PATHWAYS
Photo: Black bear captured in a trail camera image
INTRODUCTION
Outdoor recreation is a tremendous force for good. It encourages healthy lifestyles, increases understanding and connection to nature, brings economic opportunities, and many other benefits (Cooper et al. 2015, Outdoor Industry Association 2017). However, recreation can also have detrimental effects on ecosystems if it is not well-managed. Studies have documented harmful effects of recreation on many wildlife species across a wide range of ecosystems and contexts (Larson et al. 2016, 2019), including in the Greater Yellowstone Ecosystem (Borkowski et al. 2006, Courtemanch 2014). In the Bridger-Teton National Forest (BTNF), recreation is expanding rapidly, and the Cache-Game area of the BTNF is one of the
most frequently visited regions in Jackson Hole. At the same time, the Jackson Hole area is part of the Greater Yellowstone Ecosystem (GYE) and is important habitat to many wildlife species, including key migration routes and important seasonal range for several ungulate species. Managers in the BTNF need accurate and localized information about the status of recreation and wildlife habitat use to be able to manage this area successfully for humans and wildlife.
Our work seeks to improve the coexistence of people and wildlife in the Jackson Hole area by measuring baseline conditions for wildlife and people, understanding wildlife responses to human disturbance, and increasing public knowledge and appreciation of nature. Our study is among the first to quantify the direct impact of recreation on wildlife in this region. Specifically, our objectives are to:
• Monitor wildlife habitat use and recreation activity within the Cache-Game area with sufficient field effort to paint an accurate picture about existing conditions, establish a baseline for trends, and begin to assess potential impacts of recreation on wildlife.
• Increase public knowledge and appreciation of nature throughout Teton County. • Improve public knowledge about trends in wildlife habitat use and recreation activity.
METHODS
Trail camera monitoring
Using funds from the Meg and Bert Raynes Wildlife Fund, we purchased 10 Browning Strike Force HD Pro X trail cameras and associated equipment (cases, batteries, SD cards, locks) to add to our existing stock of trail cameras. We selected placements for the majority of the cameras using a randomized gradient study design, in which cameras were located at randomly selected locations along the trail network in the Cache-Game area (Fig 1). The locations were stratified by estimated level of trail use (low, medium, high). We generated random locations within the human use strata using a spatially-balanced random points algorithm that ensures balanced geographic coverage. We generated more random points than needed, allowing us to remove locations that are unsuitable due to a lack of trees for camera placement, difficult access, or other logistical considerations. Additionally, we placed several cameras in areas where we expect to see increases in recreational trail use. Studies of recreation effects on wildlife with an experimental or quasi-experimental component are rare because it is usually logistically difficult to manipulate the level of recreation on a trail segment. However, an experimental
or before-after-control-impact design can be particularly powerful because the design minimizes variation in other factors that affect wildlife habitat use, such as the vegetation community or proximity to residential development.
We placed 3 cameras in Wilson Canyon and 6 near Snow King Mountain Resort where recreation is expected to increase (Fig 1). Wilson Canyon has been recently developed for sport climbing and recreation is expected to increase with the anticipated release of a new rock-climbing guidebook which will be the first to publicly disclose the area. We placed one camera on the trail below the climbing area (high use), one near the climbing area
(moderate use), and one above the climbing area which still sees very little use. At Snow King, we placed cameras on existing trails that we expect to see increased use given the proposed expansion of the resort, especially the pending mountain bike park and expansion onto the backside of the mountain towards Leeks Canyon road. Cameras were housed in metal boxes, mounted on trees, and secured with a cable lock. When possible, we placed cameras slightly back from the trail, at a diagonal angle to the trail, and with an unobstructed view of the ground surface of the trail to capture images of animals of varying sizes as well as humans. At installation, we tested the field of view using the camera’s built-in photo viewer, removing small branches and leaves when necessary to clear the camera’s view.
Citizen scientist involvement
We involved dozens of citizen scientists in both the field and virtually. In the field, we trained a handful of citizen scientists to help set up and manage the 27 total cameras. Volunteers used GPS to navigate to the camera, unlock the metal box, replace the SD cards and batteries, and verify the correct settings. Our most active volunteer, Leo Harland, is now a senior at Jackson Hole High School. He listed the experience on his college resume and we provided a letter of support for admission. Volunteers from the Nature Mapping program also assisted in the field. Once collected, images were uploaded to Zooniverse—a virtual platform where citizen scientists from anywhere can tag the images for wildlife and human activity. Within Zooniverse we created a Field Guide to assist
volunteers in correct identification of over 20 common and rare wildlife species/focal groups from insects, to moose, wolves, and grizzly bears. Images can also be tagged for human recreation including categories for pedestrians, cyclists, and skiers. Thus far, at least 66 citizen scientists have made over 200,000 classifications on Zooniverse. Each image must be classified by five different individuals to ensure accuracy. Before making the Zooniverse project public, we tested the methods on and received feedback from upper-level high school students at the Jackson Hole Community School.
Photos: Selected wildlife images taken by trail cameras, including cougar and deer.
RESULTS
This is an ongoing project; as such, we are still working through data processing and analysis. We present here some preliminary data representing the sampling effort, data collection, and volunteer involvement.
• We placed 27 cameras in the field during the 2020 season, up from 10 during the 2019 pilot season
• Trail cameras ran from May to October and were left to run into the winter until batteries or SD cards are depleted, thereby collecting data from multiple seasons and across different types of human use (e.g., mountain biking, hiking, skiing) • Trail cameras took 553,800 photos in total during
the spring, summer, and fall of 2020
• Species detected include mule deer, elk, moose, raccoon, black bear, mountain lions, flying squirrels, and others
• 14 Jackson Hole Community School students participated in photo sorting on the Zooniverse platform
• 66 total community volunteers participated in photo sorting on the Zooniverse platform (including Community School students), plus additional volunteers who participated anonymously
• Five field citizen scientists, including Jackson Hole High School students and Nature Mapping volunteers helped maintain 27 cameras
throughout spring, summer and fall
CONTINUED WORK
We are currently partnering with TNC California and Microsoft’s AI for Earth initiative to use an Artificial
Intelligence (AI) machine learning algorithm to automatically sort images to categories including animal, person, and no subject. The AI sorts and flags all images that contain no subject, the most common in the dataset often triggered by a moving branch or snowfall. This will allow us to filter out the photos containing no animals or humans before we upload to Zooniverse, reducing the workload for citizen scientists and allowing us to get data from the photos much more quickly. We are still developing the process for working with AI-sorted photos, but preliminary results show that the AI produced highly accurate classifications on a sample of 7500 photos.
Throughout the winter and spring, we will continue to sort and tag the half-million+ images collected last season, plus ~200,000 from the 2019 pilot season, with the help of the AI and citizen scientist volunteers. In spring 2021, each camera will be reset with new batteries and SD cards and continue capturing images. We do not plan to move the cameras to new locations so that we can gather comparable data across multiple years.
As we amass data from tagged and classified images, we will begin to analyze patterns of wildlife habitat use and daily activity patterns in relation to the type, intensity, and temporal pattern of recreation activity. We are actively collaborating with BTNF managers to ensure that our results will help inform revision of the upcoming Forest Plan, in which considerations of the effect of recreation on wildlife will be included for the first time. We are actively seeking and applying for continued funding for this research.
Photo: Wildlife image of a moose taken by trail camera.
ACKNOWLEDGEMENTS
We gratefully acknowledge the support of the Meg and Bert Raynes Wildlife Fund. We also thank the US Forest Service’s Citizen Science Competitive Funding Program for financial support for the project. We thank the Bridger-Teton National Forest, especially Linda Merigliano and Tim Ferris, for their support of this project. Chris Owens from Friends of Pathways have been instrumental in the planning and implementation of this project. Kyle Kissock of Jackson Hole Wildlife Foundation made a terrific video tutorial for our Zooniverse project. Finally, we thank the many, many volunteers that have assisted on the project so far, either in the field or on Zooniverse.
ADDITIONAL INFORMATION
Mailing Address: The Nature Conservancy in Wyoming, 258 Main St. Suite 200, Lander, WY 82520 Website Address: Nature.org/wyoming
Tax ID number: 53-0242652
Contact information for the person responsible for directing the funding: Lori Mathews, [email protected] Mission Statement: To conserve the lands and waters on which all life depends
Title of Grant: Monitoring wildlife and recreation use on the Bridger-Teton National Forest Amount Received: $2,500
LITERATURE CITED
Borkowski, J. J., P. J. White, R. A. Garrott, T. Davis, A. R. Hardy, and D. J. Reinhart. 2006. Behavioral responses of bison and elk in Yellowstone to snowmobiles and snow coaches. Ecological Applications 16(5):1911–1925. Cooper, C., L. Larson, A. Dayer, R. Stedman, and D. Decker. 2015. Are wildlife recreationists conservationists?
Linking hunting, birdwatching, and pro-environmental behavior. The Journal of Wildlife Management 79(3):446– 457.
Courtemanch, A. B. 2014. Seasonal habitat selection and impacts of backcountry recreation on a formerly
migratory bighorn sheep population in Northwest Wyoming. M.S. Thesis, University of Wyoming, Laramie, WY. Larson, C. L., S. E. Reed, A. M. Merenlender, and K. R. Crooks. 2016. Effects of recreation on animals revealed as
widespread through a global systematic review. PLOS ONE 11(12):e0167259.
Larson, C. L., S. E. Reed, A. M. Merenlender, and K. R. Crooks. 2019. A meta-analysis of recreation effects on vertebrate species richness and abundance. Conservation Science and Practice 1(10):e93.
