Abstract 232 – A combination of aerial damage detection and in situ DNA sampling in crop fields for genetic monitoring of problem brown bears in Hokkaido, Japan.

Yuri Shirane, Mami Kondo, Hino Takafumi, Kazuki Miura, Tsutomu Mano, Hifumi
Tsuruga

Human-wildlife conflict in agricultural lands is a serious global issue that affects the survival of
wildlife populations and human safety. Previously, wildlife damage to crops has been
investigated by radio telemetry to track animal movements and measure habitat selection, and
by diet analysis using stable isotope analysis of biological samples. However, these methods are
difficult to obtain a profile of problem individuals involved in crop damage within an area and
to continuously survey damaged fields on a detailed spatial and temporal scale. Here, we
attempted to combine unmanned aerial survey and ground-based genetic sampling to
determine if a large number of bears in the vicinity of the agricultural lands were causing
damage or if the same individuals were causing damage repeatedly in brown bears of Hokkaido,
Japan. In Yakumo Town, located in southwestern Hokkaido, where dent corn is widely
cultivated every year, we conducted drone flights to take aerial photographs of cornfields. By
photographing for two to five consecutive days from late June to September in 2015–2019 and
2021–2022, fresh damaged patches of each field were identified based on the unique textural
characteristics of healthy and damaged patches in the field. Then, we visited each damaged
spatches to collect fresh genetic samples (e.g., hair, feces, and residual saliva on partially-
consumed corn) of brown bears, and examined 9 loci of microsatellites for individual
identification. Results revealed that most of the individuals causing damage are males, and that
same individuals repeatedly caused damage to the one corn field over multiple years. In
addition, some individuals were causing damage in a different corn field than the previous year.
Our individual-based methods for analyzing crop damage can be used to better understand the
relationship between the number of problem individuals, the area damaged, and the
landscape, leading to more efficient management to reduce human-bear conflict.