Study On Mammal Monitoring And Roe Deer Habitat Selection Based On Camera Trap Technology

Using the camera trap monitoring and line transect methods, we investigated the mammal species diversity and relative abundance of mammal in Changbaishan National Nature Reserve from Oct.2013 to Dec.2014. The effectiveness of mammal diversity monitoring was assessed using different methods. Additionally, roe deer habitat selection was studied base on the camera trap data as one of mammal species.Two monitoring methods using camera trap (ie:kilometer grid and hectare grid) were used to monitor the mammal species in the study region for estimating speices diversity and density of animal populations.1) We used 584 camera traps, monitored for 26 200 days, and got 2 096 mammal individual pictures of independently and effectively. In addition,. Camera trap survey included 12 30-km2 and 9 50-hm2 monitoring area. Ihe camera trap survey habitat types include mixed broadleaf-conifer forest, korean broad-leaved forest, larch forest, and dark coniferous forest. In this survey,8 families and 18 mammal species were detected. The relative abundance of the wild boar (Sus scrofa) was the highest, followed by roe deer (Capreolus pygargus) and sable (Martes zibellina), whereas, the brown bear(Ursus arctos) was detected by camera trap only one time.2) The 335 mammal footprints were found, and 6 families and 9 species were also identified in winter sample plot survey, which contained 3 20-km2 sample plots. tRelative footprint encounter rate indicated that the relative density of the roe deer was the highest, followed by wild boar, sable, red deer(Cervus elaphus), sika deer(Cervus nippon). However, lynx (Felis lynx) and yellow-throated marten (Martes flavigula) were were low.3) Compared with the line transect survey method, the number of mammal species obtained from camera trap monitoring method were higher. However, for the assessment of the relative abundance in species population, the results of camera trap monitoring and winter quadrat survey were similar with that of line transect survey method. Whereas, the result of hectare grid method was higher than that of kilometer grid method, which may be resulted from the spatial autocorrelation effect of individuals. Therefore, the kilometer grid monitoring method not only can monitor the number of mammal species, but also nicely capable of estimate the species diversity.The occupancy model was used to analyze the habitat selection of the roe deer to reveal its habitat selection characterisiticsand benefit for habitat management based camera trap data. The results indicated that:1) slope played an important role in detection probability of camera trap, and with the increase of slope, detection probability increased.2) The factors influencing habitat occupancy probability of roe deer included aspect, slope, river, korean broad-leaved forest and so on. The distance from riverand korean broad-leaved forest showed positive effects, and aspect and slope showed negative in habitat occupancy probability of roe deer. 3)roe deer avoided the habitat types with slope steep, while, preferred the habitat types, such as mixed broadleaf-conifer forest and korean broad-leaved forest of the gentle slope, abundant.Changbaishan National Nature Reserve not only has high mammal species richness, but also the large natural species gene pool in northeast China, especially mammal resources. Once mammal resources change in Changbaishan National Nature Reserve, mammal diversity of surrounding areas will change. Therefore, the mammal resources of Changbaishan National Nature Reserve should be long-term monitored and the effective conservation methods were adopted to protect the species diversity and relative abundance of mammal according to what mammal species habitat selection characteristics. This sudy indicated that camera trap monitoring technology has a good application prospect in the both species diversity and relative abundance monitoring, as well as the application of species habitat selection research and future management.