Abstract
We investigated mammal fauna of 12 public pastures in Tochigi prefecture, Japan, with two survey methods, camera trapping and field sign survey, and compared the results between these methods. We also examined the features of camera trapping data for future technical refinements. A total of 16 wild mammal species including 9 out of 13 medium-size and large-size mammal species in Tochigi were detected by camera trapping (12 species) and sign survey (14 species). Although the detectability of mammal signs was seasonally affected, such effects were not observed in camera trapping, suggesting that camera trapping is available throughout the year. The contingency of detected species and the correlation of abundance index (frequency of photos and signs) between these two methods were weak (kappa statistics: 0.43, 95% CI: 0.29-0.56) and mild (r=0.58, p<0.05), respectively. These methods were considered to be complementary. It took an average of 218.4 camera-days (range: 24.4-780.9) to reach the maximum species number detected in each pasture while a single remote camera per pasture was not enough to reach the maximum. There may be greater effectiveness of a fixed intermission between photographing in the remote cameras as suggested, comparing among different types of remote cameras. Practically, camera trapping in mammalian faunal surveys in pastures should be conducted with several remote cameras per pasture for at least 3 months with supplemental field sign surveys.
