Recreational trail traffic counts and trail proximity as a driver of ungulate landscape utilization
Data files
May 05, 2025 version files 117.43 MB
-
Daily_20250502.rda
118.31 KB
-
FlatDetectionFile_20250502.csv
117.30 MB
-
README.md
5.10 KB
Abstract
With continual growth in recreational trail use, it is becoming increasingly complicated to balance demands for outdoor recreation opportunities with wildlife conservation. To better understand how mule deer (Odocoileus hemionus) and Rocky Mountain elk (Cervus canadensis) respond to trail-based recreation, we deployed remote cameras in a paired study design to assess ungulate encounter rates relative to recreation traffic counts and distance from trails. Our methodology allowed us to estimate the magnitude of human activity on public land trails which can be challenging, but is a key step in understanding recreational effects on animal utilization, especially if any interaction in trail proximity and human traffic rates exist. Paired cameras provided a comparison of animal encounter rates on and off-trail at varying trail proximities, and the on-trail camera also yielded daily recreation traffic counts to assess how animals respond to varying traffic for each camera pair. Elk avoided busier trails and showed a strong positive response to distance from trails, while mule deer exhibited a positive response to higher traffic counts and no measurable effect of trail proximity. We identified a distance to trail threshold of 600-655 m, where elk utilization shifted from lower-than-expected utilization to higher-than-expected, indicating this elk population was redistributed to some degree within 57.5% of the study area’s extent. Quantifying recreation traffic counts was key to understanding species’ responses to recreation, highlighting the need for managers to consider both trail placement and human traffic regulation to mitigate recreation impacts.
https://doi.org/10.5061/dryad.80gb5mkz9
We have submitted our raw detection data (after images were processed using the CPW Photo Warehouse, FlatDetectionFile_20210329.csv) and the prepared dataset (Daily_20250502.rda) at the daily resolution, and used for analysis.
Descriptions
FlatDetectionFile_20210329.csv
Raw detection data after camera images were processed using the CPW Photo Warehouse (UTM identifiers removed).
- GroupID: Integer, ID representing a group of paired camera
- LocationName: Text (String), Concatenated location name containing GroupID and camera distance to trail where zero is the on-trail camera (e.g., "151_0")
- FileName: Text (String), Image filename
- ImageDate: DateTime, Date and time the image was captured (MM/DD/YYYY H:MM:SS, timezone = "MST")
- DetectionID: Text (String), Unique identifier for a detection event (in
{}format) - Individuals: Integer, Number of individuals detected
- SpeciesID: Integer, ID for 'species' classification
- CommonName: Text (String), Camera trap classification labels. 'Species' include wildlife species, recreators (vehicle types and human activities), domestic animals, and unclassified or unknown categories.
- DetailID (Integer or "NA") and DetailText: Because recreators often triggered a chronological succession of images representing a single individual or same party as they passed by the on-trail camera, the first image of any recreation type was assigned a “new party” attribute. If DetailID = 4, DetailText designates the "New party".
- Comments: Text (String), General comments from observer classification
- ObsID: Integer, Observer ID
- LastName: Text (String), Observer's last name
- ActiveStart: DateTime, Start of the cameras active period (MM/DD/YYYY H:MM:SS, timezone = "MST")
- ActiveEnd: DateTime, End of the cameras active period (MM/DD/YYYY H:MM:SS, timezone = "MST")
Daily_20250502.rda
Data prepared for analysis (daily recreation traffic rates incorporating deer and elk counts by site and environmental characteristics at each camera).
Identification & Time
- PAIRID: Identifier for matched camera pairs or sampling units.
- day_of_year: Numeric day of the year (1–366).
- LOCNAME: Location name or ID, usually showing pair and treatment (whether on [0] or off trail [>0], e.g., 1_0, 1_5).
Wildlife Encounters
Density-based metric of deer utilization intensity at each camera, calculated using a time-in-front-of-camera approach that accounts for detection zone size, monitoring duration, and group size
- DEER_Encounter: Daily deer encounter rate (numeric 0 - 4.95).
- ELK_Encounter: Daily elk encounter rate (numeric 0 - 16.13).
Trail & Traffic
- DIST: Distance to route (trail, road) in meters (0 - 1810).
- TRAFFIC: Daily number of recreator detections (0 - 412).
- ROUTETYPE: Route classification (e.g., 0 = non-motorized, 1 = motorized).
Environmental/Spatial Variables
- ELEV: Elevation in meters (source 10m DEM - USGS).
- slope: Slope in degrees (source 10m DEM - USGS).
- ASP90 / ASP180: aspect on x and y aspect respectively (source 10m DEM - USGS).
- TRI500: Terrain ruggedness index at 500 m search window (source 10m DEM - USGS).
- TPI500: Topographic position index at 500 m search window (source 10m DEM - USGS).
- PERNENN_D: Distance to perennial stream in meters (source 1:24K DRGs - USGS).
- DINT: Distance to intermittent stream in meters (source 1:24K DRGs - USGS)
- paved_euc: Distance to nearest highway (source Colorado Dept. of Transportation).
- conifer: Binary presence of conifer (source LANDFIRE - USFS).
- DFE: Distance from forest edge in meters (source LANDFIRE - USFS).
- HD500m: Housing density at 500m seach winter (source Gunnison County Parcel Data).
- On_off_trail: Categorical whether the location is "On-trail" or "Off-trail".
- StructureKD: Distance to nearest structure (source Building footprints - Microsoft Bing).
- CC: Canopy cover percent (source LANDFIRE - USFS).
Temporal
- WEEKEND: 1 = weekend, 0 = weekday.
- hunt_new: designates whether day is within a hunting season (categorical: "None" = no hunting, "Primitive" = archery, muzzleloader, "Rifle").
Human Activity
- MOTO: Count of motorized recreation detections.
- NONMOTO: Count of non-motorized recreation detections.
- The following are counts of detections from specific recreator types or modes of travel: "Car.Truck.SUV", "Dog.Off.Leash", "Dog.On.Leash", "Horseback", "Motorcycle", "Mountain.Bike", "On.foot.non.mechanized", "Pack.goat", "Pack.horse", "Pack.mule", "Quad.bike", "Side.by.side.OHV", "Skier", "Snowmobile", "Stroller", "Unknown.nonmotorized.recreator", "Unknown.OHV", "Walking.with.bike"
Code/Software
R is required to open Daily_20250502.rda; the file was created using version 4.2.3.
Detection data from camera trap photographs. Cameras were deployed in paired study design on- and off-trails to simulatenously collect animal and recreation data. This repository includes:
- the raw detection data (after images were processed using the CPW Photo Warehouse) (FlatDetectionFile_20210329.csv)
- R object named 'Daily' prepared for analysis containing daily recreation traffic, daily ungulate encounters, and environmental covariates for each camera (location data removed) (Daily_20250502.rda)