Antler detection from the sky: deer sex ratio monitoring using drone-mounted thermal infrared sensors
Ito, Takehiko et al. (2022), Antler detection from the sky: deer sex ratio monitoring using drone-mounted thermal infrared sensors, Dryad, Dataset, https://doi.org/10.5061/dryad.jm63xsjcr
Sex differences in large mammals with sexual dimorphism are important ecological and evolutionary issues and key factors for wildlife management. To examine the potential use of drone (unmanned aerial vehicle; UAV) observation using thermal infrared images for sex ratio monitoring of deer, we conducted UAV surveys at night in a sparse forest located on the distribution periphery of sika deer (Cervus nippon) and wild boar (Sus scrofa) local populations during summer and winter. Of the 163 thermal infrared images of large mammals detected, 132 (81.0%) and 16 (9.8%) were identified for deer and wild boar, respectively. In addition, velvet antlers of deer were visually recognized during summer, and 92% of the detected deer were antlered. This biased sex ratio would be a characteristic in the distribution periphery of local deer populations. Therefore, monitoring abundance and sex ratio using thermal infrared sensors on UAVs can improve deer management especially in the distribution periphery of local populations.
Large mammals were detected using a thermal infrared camera (Zenmuse XT ZXTB19FP, DJI) on a UAV (Inspire 1 V2.0, DJI) in the study area of 1 km2 in Tottori, western Japan. Observation flights were conducted at nights without rain, snowfall, or strong wind during summer and winter in 2017. A pilot controlled the UAV and camera manually and searched the heat sources on the controller’s monitor, flying approximately 70–120 m above ground to search.
Eleven camera traps (seven Trophycam XLT HD MAX, Bushnell, and four Ltl-5210A, Little Acorn Outdoors) were set in the study area over the entire period of drone observation (June to December 2017). We recorded videos for 30 s after detection using the infrared sensor, and the interval times after each video recording were 3 min for Trophycam and 10 min for Ltl-5210A.