Locating chimpanzee nests and identifying fruiting trees with an unmanned aerial vehicle.

ABSTRACT

Monitoring of animal populations is essential for conservation management. Various techniques are available to assess spatiotemporal patterns of species distribution and abundance. Nest surveys are often used for monitoring great apes. Quickly developing technologies, including unmanned aerial vehicles (UAVs) can be used to complement these ground-based surveys, especially for covering large areas rapidly. Aerial surveys have been used successfully to detect the nests of orang-utans. It is unknown if such an approach is practical for African apes, which usually build their nests at lower heights, where they might be obscured by forest canopy. In this 2-month study, UAV-derived aerial imagery was used for two distinct purposes: testing the detectability of chimpanzee nests and identifying fruiting trees used by chimpanzees in Loango National Park (Gabon). Chimpanzee nest data were collected through two approaches we located nests on the ground and then tried to detect them in UAV photos and vice versa. Ground surveys were conducted using line transects, reconnaissance trails, and opportunistic sampling during which we detected 116 individual nests in 28 nest groups. In complementary UAV images we detected 48% of the individual nests (68% of nest groups) in open coastal forests and 8% of individual nests (33% of nest groups) in closed canopy inland forests. The key factor for nest detectability in UAV imagery was canopy openness. Data on fruiting trees were collected from five line transects. In 122 UAV images 14 species of trees (N = 433) were identified, alongside 37 tree species (N = 205) in complementary ground surveys. Relative abundance of common tree species correlated between ground and UAV surveys. We conclude that UAVs have great potential as a rapid assessment tool for detecting chimpanzee presence in forest with open canopy and assessing fruit tree availability. UAVs may have limited applicability for nest detection in closed canopy forest.