Objectives and methods The impact of sambar deer on alpine ecosystem in the Yushan National Park has increased in recent years. This project aims to 1) reveal the composition and distribution of mid- to large-sized mammals in Shinkang area by monitoring tracks and signs along the hiking trails and camera traps, 2) establish long-term monitoring area and transects to evaluate the impact of sambar deer on the recruitment and structure of forests, 3) estimate population size of sambar deer via camera traps and capture-recapture model. Important findings Sixteen species of mammals in 9 family and 4 orders were found. Among them, Formosan lesser weasel and Formosan gem-faced civet were first recorded in Shinkang area. This area is the southern-most distribution for the Formosan lesser weasel and the highest distribution for the Formosan gem-faced civet. Sambar deer has the highest relative abundance among all mammals. Track densities of the four ungulate species were higher than 2009, which may be related to the closure of the National Park this year. The occurrence index of the Formosan rock-monkey was significantly lower in winter than in other seasons, but it is unclear whether this phenomenon is due to seasonal dispersal or changes in the frequency of ground activity. Three hundred and ninety-two trees were debarked along the hiking tracks this year. Among them, 70% were Taiwan hemlock. Debarking happened most frequently in June and August. Tree species and the frequency of debarking were highly correlated (χ2 = 257.7，p < 0.0001). Taiwan hemlock, Taiwan spruce, and Taiwan white fir were the three most frequently debarked species, while Taiwan pieris, Eurya loquaiana, and Taiwan red pine were the three least frequently debarked species. In the long-term monitoring areas, 70% of the trees were Taiwan hemlock and Taiwan white fir. Sambar deer debark individuals of DBH lower than 70 cm and kill Taiwan hemlock of DBH < 25 cm and Taiwan white fir of DBH < 10 cm. The number of sambar pellets was not related to number of saplings of the two species. However, the magnitude debarking was related to the occurrence index of sambar deer. Therefore, the OI of sambar deer is a potential index for the damage on the two tree species. The occurrence index of sambar deer is only positively correlated with the density of Yushania niitakayamensis but is independent with coverage, height, and density of any other plant species. In Yushania niitakayamensis area, OI of sambar deer is positively correlated with density and average height of conifer saplings. In conifer forests, OI of sambar deer is positively correlated with density of Yushania niitakayamensis. This facts indicate the activity of sambar deer is correlated with the density of food. Population density of sambar deer was experimentally estimated at two areas. The results showed that the model to estimate population size should incorporate variability of capture probability among individuals and capture occasions. The population size of sambar deer was estimated to be 54 individuals, equivalent to 0.95 individuals per hectare in Bushin camping area. The estimation at the upstream of Lakuin river was not successful due to bad weather. The most appropriate density of camera and duration of capture are to be evaluated in future research. Suggestions The National Park should use the monitoring procedure to reveal the population trend of large ungulates such as sambar deer and their impacts on habitats and gradually establish quantitative information for wildlife management. The sambar deer is in need of conservation while also have impacts on the environment. Therefore, any management practice should base on solid and quantitative information. The Yushan National Park should consider both ecosystem function and needs of the society for the management of sambar deer and its habitat to establish a paradigm for other National Parks. The activity patterns of three species of ruminants recorded this year could be considered as natural patterns under no or low human disturbance. This information can be used as a base for the evaluation of the impacts of tourists, constructions, or hunting activities on wildlife. In order to establish scientific and persuasive management polices and practices, the National Park should request statistical background in methodology and data analysis in all monitoring projects.