【目的】 通过探究环境变化对山杨(Populus davidiana)分布的影响,为山杨资源的保护和开发提供理论支撑。【方法】 根据山杨的134条地理分布数据,结合18个气候、土壤及地形因子,基于Biomod2软件包构建组合模型,模拟我国山杨潜在分布区在未来3种气候条件模式下的空间分布格局变化,并确定影响山杨分布的主要环境变量。【结果】 我国山杨当前潜在适生分布区主要位于400 mm等降水线两侧较高纬度或较高海拔地区,总面积约为1 560 340.9 km²,约占我国陆地面积的16.2%,其中大兴安岭、长白山、太行山、秦岭、祁连山南麓、横断山、云贵高原等地区为山杨高度适生区;在未来气候条件下,山杨适生区整体呈向西南方向收缩趋势,生境适宜度总体呈下降趋势;影响山杨分布主要环境变量为最热月最高气温、年降水量和海拔;基于5个最优单一模型构建的组合模型比单一模型对山杨适生区预测结果更好,训练集平均受试者工作特征曲线下面积和真实技巧统计值分布为 0.91和0.73,预测准确度较高。【结论】 我国山杨空间分布格局主要受水热条件影响,海拔也是影响山杨分布的重要因素。在未来气候条件下,山杨分布区面积将随气候变暖的程度逐渐减少。以山杨作为用材林和生态公益林树种进行造林时,造林地点应选择未来生境适宜度变化不大的地区,以降低未来由于气候变化造成的损失。

【Objective】 This study aims to investigate the effects of changes in environmental factors on the distribution of Populus davidiana, and to provide theoretical support for the conservation and development of P. davidiana resources. 【Method】 This study applied Biomod2 to simulate changes in the spatial distribution pattern of P. davidiana in China's potential distribution areas under three future climatic conditions based on 134 geographical distribution data points of P. davidiana in China, combined with 18 climatic, soil and topographic factors. Then a combinatorial model based on the Biomod2 package was consturcted and identified the main environmental variables affecting the distribution of P. davidiana were identified. 【Result】 The current potential distribution areas of P. davidiana in China were mainly located at higher latitudes or higher altitudes on both sides of the 400 mm precipitation line, with a total area of about 1 560 340.9 km², of which the Greater Khingan Mountains, Changbai Mountains, Taihang Mountains, Qinling Mountains, southern foot of Qilian Mountains, Hengduan Mountains, Yunnan-Guizhou Plateau and other areas are the highest suitable areas for P. davidiana. Under future climatic conditions, the overall trend of suitable areas for P. davidiana will shrink to southwest China, and the overall trend of suitable areas was decreasing. The ensemble model constructed based on the five optimal single models had better prediction results for suitable areas for P. davidiana compared to the single model, and the area under the receiver operating characteristic curve and true skill statistics were distributed as 0.91 and 0.73, with higher prediction accuracy. 【Conclusion】 The spatial distribution pattern of P. davidiana in China was mainly influenced by water and heat conditions, while altitude was also an important factor affecting its distribution. Under future climatic conditions, the area of P. davidiana distribution will gradually decrease based on the degree of climate warming. When planting P. davidiana for timber forests and as an ecological public welfare forest species, planting sites should be selected in areas where habitat suitability will not change significantly in the future, to reduce future losses because of climate change.