Abstract: It is important to identify the geographical distribution of rare or endangered species in biodiversity conservation. Blacknecked crane (Grus nigricollis) is an alpine wetland’s flagship species, which belongs to a national Ⅰ level key protected wild species in China. Based on distribution and ecological environment data of blacknecked crane, we adopted the classification and regression tree (CART), treenet gradient boosting (TG), random forest (RF) and maximum entropy (ME) model to predict its suitable distribution area. The results showed that: (1) The main environmental factors affecting its geographical distribution are altitude, precipitation and temperature in autumn migration period, with a relative contribution rate of 9.7%, 8.9% and 7.3% respectively. Its suitable habitats are the areas with altitude of 3000-3950 m, mean monthly temperature of -1.2 to 0.2 ℃, and monthly precipitation of 28-32 mm in autumn migration period. (2) Combining with the AUC (area under curve), TSS (true skill statistics), and Kappa coefficients, the prediction results of random forest model showed the highest matching degree with the known distribution locations of black-necked crane (AUC=0.889, TSS=0.897, Kappa=0.613). (3) Its suitable distribution area is mainly concentrated in the southwest of Zoige County (Nenwa, Dazasi, Tangke, Xiaman) and the east of Maqu County (Awancang, Manrima). High probability distribution area is located in Tawa Village, Xiaman Town, Zoige County. The landscape fragmentation degree is small, with the aggregation index of 99.6 and the separation index of 4.6. The highsuitable area for black-necked crane accounts for 10.8%, the low-suitable area accounts for 20.7%, and 60.5% of the area are not suitable.

Key words: black-necked crane (Grus nigricollis), species distribution model (SDM), Zoige wetland, suitable distribution area.