Evaluation of Habitat Preference of Wildlife Using Camera Traps and High Resolution RGB Data of Drones in Urban Forest Fringe Area

Fragmentation is destroying habitat, reducing food sources for wildlife and affecting their home-range. In the case of some wild animals, they advanced into urban area and damage farmland to look for food sources. In particular, Sus scrofa and Hydropotes inermis are representative species that threaten the agriculture area in Republic of Korea. Although capture, GSP tracking, and fence installation have been proceeded for their management, it requires a lot of manpower and time. To overcome this problem, drones and camera traps are being used, but it is rare that they are used in an integrated way. In this study, in order to evaluate habitat preference of S. scrofa and H. inermis introduced through the fridge area before the harvest season, 18 camera traps were evenly deployed in an area of 40 ha at intervals of about 100 m from March to June in 2021. In addition, the survey site was photographed every 2 weeks with a drone. Tree species, trails, cemetery, and mountain edges were extracted through drones, and the location of the mud bath was identified through field surveys. The frequency of detection of S. scrofa and H. inermis was counted through the camera traps. A total frequency of detection of S. scrofa was 354. a threefold increase in frequency of detection for S. scrofa in June compared to the previous month. A total frequency of detection of H. inermis was 173, and the frequency of detection doubled in April compared to the previous month. As a result of multiple regression analysis to analyze the relationship between environmental factors and frequency of detection, mud bath was selected as preferred habitats for S. scrofa, and cemetery were selected as habitats to avoid. For H. inermis, on the other hand, there was no specific preferred habitat, but they did not prefer the vicinity of the trail. S. scrofa prefered deciduous forest type (60%) and H. inermis appeared in Quecus forest type (70%). This study can suggest the possibility of analyzing habitat preference area for wildlife management by using high resolution RGB data of drones and camera traps.