Adapting camera-trap placement based on animal behavior for rapid detection: A focus on the Endangered, white-bellied pangolin ().

Pangolin species are notoriously difficult to detect and monitor in the wild and, as a result, commonly used survey techniques fall short in gathering sufficient data to draw confident conclusions on pangolin populations, conservation status, and natural history. The white-bellied pangolin is a semiarboreal species that may be poorly detected in general mammal surveys even with modern techniques such as camera-trapping. As a result, population status information is often derived from hunting, market, and trafficking data. There is therefore a crucial need to improve camera-trap survey methods to reliably detect this species in its natural environment. Here, we test the influence of camera-trap placement strategy on the detectability of the white-bellied pangolin by comparing estimates from targeted ground-viewing camera-trapping and a novel log-viewing placement strategy adapted from local hunters' knowledge. Our results suggest that (1) deploying camera-traps to detect animals walking along logs is an effective strategy for recording several forest species, including the white-bellied pangolin, and (2) that camera-traps targeting logs are more efficient at detecting white-bellied pangolins than camera-traps viewing the ground (>100% increase in detection probability). We also found moderate evidence that there is a relationship between the white-bellied pangolin occurrence at our locality and elevation and weak evidence of an association with distance to the nearest river. Our results suggest an effective new monitoring approach allowing consistent detection of the white-bellied pangolin with moderate survey effort. This highlights the importance of harnessing local knowledge to guide the design of monitoring protocols for cryptic species.