Improving Efficiency of Prairie Dog Surveys by Using a Small Copter Drone

Authors

Aaron B. Shiels, USDA, APHIS, Wildlife ServicesFollow
J. W. Fischer, USDA/APHIS/WS National Wildlife Research CenterFollow
Danika Spock, USDA APHIS, Wildlife Services, National Wildlife Research Center, Fort Collins
Meagan Allira, USDA APHIS, Wildlife Services, National Wildlife Research Center, Fort Collins

Date of this Version

2022

Citation

Proceedings, 30th Vertebrate Pest Conference (D. M. Woods, Ed.) Paper No. 21. Published December 27, 2022. 7 pp.

Comments

U.S. government work

Abstract

Prairie dogs are an accessible and enjoyed wildlife species in Colorado that require occasional surveys because populations can change abruptly due to plague outbreaks or human-induced control. We evaluated the use of small copter drones at four prairie dog colonies on Open Space and Mountain Parks lands, City of Boulder, to determine if this methodology improves efficiency over ground-based survey methods. We counted prairie dogs and burrows using two types of drones (DJI Matric 210 and Autel Evo II) at altitudes 100', 150', and 400' (burrows only). We recorded video and merged still images into orthomosaics prior to having USDA staff analyze this imagery. We then compared the drone imagery counts to those of our simultaneous ground-based counts of prairie dogs. We determined that 100' altitude mosaics produced using DJI Matric 210 drone were most accurate (closest to true, ground-based counts) for burrow abundance. We were not able to identify the best drone and altitude combination for drone-based prairie dog abundance. Overall video and mosaics both had similar accuracy in most prairie dog counts, however 150' video was more accurate than 100' video. One staff member counted burrows more closely to true than did the other. Both staff members required about the same amount of time to count and analyze imagery; videos could be evaluated slightly faster than mosaics (average of 3.8 hours vs. 5.5 hours per imagery) when counting prairie dogs, and burrow counts (of mosaics) generally took 2-3 times longer to analyze (averaging 8.1 hours per imagery; range: 3-13 hours) than did prairie dog counts. The labor requirement of using drones for burrow and prairie dog counts is far more time consuming (3-4 times longer per hectare) than having field staff conduct the traditional on-the-ground counts that include repeated prairie dog counts in a day. Until drone technology improves to allow targeting larger colonies (>2 km2) and automated detection and counting of wildlife become more commonplace, drone surveys are unlikely to be a more efficient technique than ground-based surveys for evaluating prairie dog abundances.