关键词: 雪岭云杉（Picea schrenkiana）, 适生区分布, 环境因子, MaxEnt模型, 新疆

Abstract: Taking Picea schrenkiana as an object, based on 210 distribution sample data and 23 environmental variables, the MaxEnt model was applied to predict the potential distribution and important impact factors of Picea schrenkiana under the current climate environment. The accuracy of the model was tested using the Receiver Operating Characteristic curve （ROC curve）, and the contribution rate, replacement importance, and knife cut method of environmental factors were used to screen the dominant ecological factors. ArcGIS was combined to achieve visualization of suitable growth areas for spruce in Xueling. The results showed that the average AUC obtained by running the MaxEnt model repeatedly for 10 times was 0.953, indicating that the model had high prediction accuracy, and the predicted geographical distribution results of Xueling spruce are highly consistent with the actual distribution data points, the operation results were reliable; under the current climate environment, were high suitability areas of Picea schrenkiana were mainly distributed in the Tianshan Mountains, including the Ili River Valley, the mountains on both sides of the Tekes River Valley, the Borokonu Mountains, the Bogda Mountains, the western mountains of Jungar, the Alatao Mountains, and the mountains on both sides of the Fergana River Valley. The areas of high, medium, and low suitability areas respectively accounted for 5.44%, 10.75%, and 12.60% of the total suitability area; in the current climate environment, the dominant environmental factors affecting the geographical distribution of Picea schrenkiana were elevation, slope, driest monthly precipitation （bio14）, annual average temperature （bio1）, and warmest season precipitation （bio18）, among which altitude was the most important factor affecting the geographical distribution of Picea schrenkiana.

Key words: Picea schrenkiana, distribution of suitable habitat, environmental factors, MaxEnt model, Xinjiang