Citizen science project on urban canids provides different results from camera traps but generates interest and revenue
Data files
Feb 04, 2025 version files 302.98 KB
-
AdDataByDay.csv
5.69 KB
-
cameraCovariates.csv
3.90 KB
-
coyoteDetectionHistory.csv
3.70 KB
-
coyoteSeparatedDetections.csv
12.40 KB
-
Days_Cameras_Working_by_Week.csv
3.42 KB
-
foxSeparatedDetections.csv
9.66 KB
-
PathFencedOpen.csv
999 B
-
peopleSeparatedDetections.csv
112.28 KB
-
README.md
9.35 KB
-
redfoxDetectionHistory.csv
3.70 KB
-
sightingsCorrected.csv
137.87 KB
Abstract
As urbanization increases, wildlife increasingly encounters people. Coyotes (Canis latrans) and red foxes (Vulpes vulpes) are two canid species that have readily adapted to urban environments. Citizen science has emerged as a low-cost method of collecting data on urban-adapted species that can benefit management agencies but may provide different results than traditional methods. We analyzed data collected by citizen scientists and via motion-triggered camera traps to see how each related to the anthropogenic features of distance to roads, building density, and median household income and the natural feature of distance to water. We also investigated the potential benefits of advertising the citizen science project on social media. We used occupancy models to analyze data from a grid of 67 cameras across Wichita, Kansas, USA, from March 2023 to February 2024. We used generalized linear models to evaluate data collected simultaneously from a website we created and advertised on social media where members of the public could report sightings of urban canids. The camera-trap occupancy models suggested that red fox occurrence was only related positively to building density and coyote occurrence was related negatively to building density and positively to income. The citizen science models suggested that sighting reports of both species were more likely closer to roads, at intermediate building densities, and in high income neighborhoods. Coyotes and red foxes were both most likely to be detected by people during crepuscular periods but most likely to be detected by cameras at night. We also found that advertisements increased sighting reports and generated six times as much revenue for the wildlife agency than was spent. Our study suggests that citizen science data differs from camera-trap data by tracking human activity patterns and distribution, but citizen science projects can provide other benefits such as generating interest in and revenue for management agencies.
README: Citizen science project on urban canids provides different results from camera traps but generates interest and revenue
https://doi.org/10.5061/dryad.d51c5b0ck
Description of the data and file structure
Here are CSV files with the data used in “Citizen science project on urban canids provides different results from camera traps but generates interest and revenue.” Most of the files are data collected by camera traps with two exceptions: “AdDataByDay.csv” and “sightingsCorrected.csv”.
Files and variables
File: AdDataByDay.csv
Description: This is data on how well the advertisements performed in terms of reach and return on investment. Each row represents one day.
Variables
- Date: Date of Ad data
- Amount spent (USD): the amount of money in USD spent by the general public on the KDWP website after clicking on our advertisements each day.
- Purchases conversion value: the amount of money in USD the KDWP spent on advertisements each day.
File: cameraCovariates.csv
Description: This is data on each camera location along with its covariate values.
Variables
- CamLoc: The camera ID (A letter plus number code) combined with the location (either A or B). All cameras started at A (except A8), and if they needed to be moved, they were moved to B.
- RoadDistM: A measure of how close the sighting was to a road in meters.
- DistWaterM: A measure of how close the sighting was to a body of water in meters.
- BuildingDensity: A measure of the building density in a 1km radius around the sighting. Building density is measured as a percent of land area.
- Income: A measure of the median household income in USD in a 1km radius around the sighting.
File: DaysCamerasWorkingByWeek.csv
Description: This shows how many days each week the camera traps were active for.
Variables
- CamLoc: The camera ID (A letter plus number code) combined with the location (either A or B). All cameras started at A (except A8), and if they needed to be moved, they were moved to B.
- SP1: How many days the camera was active for spring week 1.
- SP2: How many days the camera was active for spring week 2.
- SP3: How many days the camera was active for spring week 3.
- SP4: How many days the camera was active for spring week 4.
- SP5: How many days the camera was active for spring week 5.
- SU1: How many days the camera was active for summer week 1.
- SU2: How many days the camera was active for summer week 2.
- SU3: How many days the camera was active for summer week 3.
- SU4: How many days the camera was active for summer week 4.
- SU5: How many days the camera was active for summer week 5.
- UT1: How many days the camera was active for fall week 1.
- UT2: How many days the camera was active for fall week 2.
- UT3: How many days the camera was active for fall week 3.
- UT4: How many days the camera was active for fall week 4.
- UT5: How many days the camera was active for fall week 5.
- WI1: How many days the camera was active for winter week 1.
- WI2: How many days the camera was active for winter week 2.
- WI3: How many days the camera was active for winter week 3.
- WI4: How many days the camera was active for winter week 4.
- WI5: How many days the camera was active for winter week 5.
File: redfoxDetectionHistory.csv
Description: These data show whether a red fox was detected on each camera each week.
Variables
- CamLoc: The camera ID (A letter plus number code) combined with the location (either A or B). All cameras started at A (except A8), and if they needed to be moved, they were moved to B.
- SP1: Red fox detection (1 for yes, 0 for no) for spring week 1.
- SP2: Red fox detection (1 for yes, 0 for no) for spring week 2.
- SP3: Red fox detection (1 for yes, 0 for no) for spring week 3.
- SP4: Red fox detection (1 for yes, 0 for no) for spring week 4.
- SP5: Red fox detection (1 for yes, 0 for no) for spring week 5.
- SU1: Red fox detection (1 for yes, 0 for no) for summer week 1.
- SU2: Red fox detection (1 for yes, 0 for no) for summer week 2.
- SU3: Red fox detection (1 for yes, 0 for no) for summer week 3.
- SU4: Red fox detection (1 for yes, 0 for no) for summer week 4.
- SU5: Red fox detection (1 for yes, 0 for no) for summer week 5.
- UT1: Red fox detection (1 for yes, 0 for no) for fall week 1.
- UT2: Red fox detection (1 for yes, 0 for no) for fall week 2.
- UT3: Red fox detection (1 for yes, 0 for no) for fall week 3.
- UT4: Red fox detection (1 for yes, 0 for no) for fall week 4.
- UT5: Red fox detection (1 for yes, 0 for no) for fall week 5.
- WI1: Red fox detection (1 for yes, 0 for no) for winter week 1.
- WI2: Red fox detection (1 for yes, 0 for no) for winter week 2.
- WI3: Red fox detection (1 for yes, 0 for no) for winter week 3.
- WI4: Red fox detection (1 for yes, 0 for no) for winter week 4.
- WI5: Red fox detection (1 for yes, 0 for no) for winter week 5.
File: coyoteDetectionHistory.csv
Description: These data show whether a coyote was detected on each camera each week.
Variables
- CamLoc: The camera ID (A letter plus number code) combined with the location (either A or B). All cameras started at A (except A8), and if they needed to be moved, they were moved to B.
- SP1: Coyote detection (1 for yes, 0 for no) for spring week 1.
- SP2: Coyote detection (1 for yes, 0 for no) for spring week 2.
- SP3: Coyote detection (1 for yes, 0 for no) for spring week 3.
- SP4: Coyote detection (1 for yes, 0 for no) for spring week 4.
- SP5: Coyote detection (1 for yes, 0 for no) for spring week 5.
- SU1: Coyote detection (1 for yes, 0 for no) for summer week 1.
- SU2: Coyote detection (1 for yes, 0 for no) for summer week 2.
- SU3: Coyote detection (1 for yes, 0 for no) for summer week 3.
- SU4: Coyote detection (1 for yes, 0 for no) for summer week 4.
- SU5: Coyote detection (1 for yes, 0 for no) for summer week 5.
- UT1: Coyote detection (1 for yes, 0 for no) for fall week 1.
- UT2: Coyote detection (1 for yes, 0 for no) for fall week 2.
- UT3: Coyote detection (1 for yes, 0 for no) for fall week 3.
- UT4: Coyote detection (1 for yes, 0 for no) for fall week 4.
- UT5: Coyote detection (1 for yes, 0 for no) for fall week 5.
- WI1: Coyote detection (1 for yes, 0 for no) for winter week 1.
- WI2: Coyote detection (1 for yes, 0 for no) for winter week 2.
- WI3: Coyote detection (1 for yes, 0 for no) for winter week 3.
- WI4: Coyote detection (1 for yes, 0 for no) for winter week 4.
- WI5: Coyote detection (1 for yes, 0 for no) for winter week 5.
File: PathFencedOpen.csv
Description: This shows whether the cameras were placed on a Road (R), Trail/Path (P), Open area (O), or in an enclosed area (F).
Variables
- CamLoc: The camera ID (A letter plus number code) combined with the location (either A or B). All cameras started at A (except A8), and if they needed to be moved, they were moved to B.
- Spring: Spring location (Road (R), Trail/Path (P), Open area (O), or in an enclosed area (F))
- Summer: Summer location (Road (R), Trail/Path (P), Open area (O), or in an enclosed area (F))
- Fall: Fall location (Road (R), Trail/Path (P), Open area (O), or in an enclosed area (F))
- Winter: Winter location (Road (R), Trail/Path (P), Open area (O), or in an enclosed area (F))
File: sightingsCorrected.csv
Description: This is our citizen science data which consists of urban canid sightings by members of the public from March 2023 to February 2024. Each row is a canid sighting.
Variables
- num: Number of report
- Species: species that was sighted
- Date: Date of sighting
- Time: Time of sighting
- Remote_cam: Whether the sighting was from a remote camera and/or in person
- RoadDistM: A measure of how close the sighting was to a road in meters.
- DistWaterM: A measure of how close the sighting was to a body of water in meters.
- BuildingDensity: A measure of the building density in a 1km radius around the sighting. Building density is measured as a percent of land area.
- Income: A measure of the median household income in USD in a 1km radius around the sighting.
File: coyoteSeparatedDetections.csv
Description: These data display the individual detections of coyotes on the camera traps.
Variables
- Number: Detection number
- Camera: camera ID
- DateTime: Date and Time
- Species: species of animal detected
- RadianTime: Time of day in radians from 0 (midnight) to 3.14 (noon) to 6.28 (midnight).
File: foxSeparatedDetections.csv
Description: These data display the individual detections of red foxes on the camera traps.
Variables
- Number: Detection number
- Camera: The camera ID (A letter plus number code).
- DateTime: Date and Time
- Species: species of animal detected
- RadianTime: Time of day in radians from 0 (midnight) to 3.14 (noon) to 6.28 (midnight).
File: peopleSeparatedDetections.csv
Description: These data display the individual detections of people on the camera traps.
Variables
- Number: Detection number
- Camera: The camera ID (A letter plus number code).
- DateTime: Date and Time
- Species: species of animal detected
- RadianTime: Time of day in radians from 0 (midnight) to 3.14 (noon) to 6.28 (midnight).
Methods
Data on red foxes, coyotes, and human activity patterns were collected from a public website where people could report sightings and from camera traps.