关键词: 云南, 藏南, 印度野牛, 生态位模型, 橡胶林, 栖息地适宜性

Abstract

Gaur (Bos gaurus) are found in the Yunnan Province and Zangnan in southern Tibet in China. We conducted two field surveys in Xishuangbanna, Pu’er and Mt. Gaoligongshan in Yunnan Province, from February to March and November to December in 2016. We collected 47 valid occurrence locations of gaur by combining survey data and records from literature. Our analysis suggests that there are 180-210 gaurs in the Yunnan Province which face a serious survival crisis. No gaur signs were found in Mt. Gaoligongshan. Next, we used MaxEnt models to predict the potentially suitable habitats for gaur. We grouped 14 habitat predictor variables into five classes—terrain, land cover type, human footprint index, the distance to water and road, as well as climatic factors, and determined the contribution of each habitat factor to habitat suitability for gaur. The accuracy of our prediction models was accessed by the area under the curve (AUC) values for a receiver operating characteristic (ROC) curve. The validation showed that the results had high average AUC value of 0.994. The simulated potential habitat was divided into four classes—the most suitable habitat, moderately suitable habitat, low suitability habitat, and unsuitable habitat. The most suitable habitats for gaur are mainly located in southern Yunnan and Zangnan and spanned 4,987 km² and 13,995 km² respectively. Habitats with moderate suitability (total area = 32,778 km²) were located in the marginal areas of the most suitable habitats and in the Mt. Gongligongshan area. The most suitable habitats were mixed with habitat patches of moderate and low suitability for gaur in the southern parts of Xishuangbanna. In contrast, the central and northern parts of the study area were classified entirely as low suitable habitats and unsuitable habitats. The results of a Jackknife test indicated that temperature seasonality and isothermality had the strongest influence on habitat suitability for gaur, whereas terrain factor and precipitation had little effect. Temperature difference, land cover type and human footprint index were the main variables that explained patterns of gaur distribution. The results of land cover classification (using remote sensing) showed that rubber plantations have fragmented the suitable habitat and reduced landscape connectivity for gaur. We recommend that the relevant management authorities should protect natural forests, control the development of rubber plantations and other agricultural development in habitats suitable for gaur, and improve landscape connectivity to restore gaur populations in the landscape.

Key words: Yunnan, Zangnan, Bos gaurus, ecological niche model, rubber plantations, habitat suitability assessment