[Prediction of potential distribution of Carpinus cordata in China under climate change.]

Climate change seriously affects the geographical distribution of plants. Regional diffe-rences in plant response to climate change will provide important guidance for species introduction and conservation. Based on ArcGIS and MaxEnt model, we used 176 geographic information of Carpinus cordata and 13 climatic variables to reconstruct its current and future niche. The results showed that the model had a high credibility in simulating contemporary potential distribution areas. The AUC values of the test set and the training set of the model were 0.973 and 0.957, respectively. The main core suitable areas were concentrated in Qinling, Changbai Mountain and their adjacent areas, with other sporadic "island" distribution. C. cordata is not distributed in Guizhou, Jiangxi, Yunnan and Fujian, but the model predicted some suitable distribution areas in those provinces. With climate warming in the future, ecologically suitable areas of C. cordata would increase significantly, mainly as "shrinking to high altitude areas", "expanding northward", and "expanding eastward". However, core suitable areas would be slightly reduced, which would be manifested as "shrinking southward", "moderate stability", and "expanding northward". The response of C. cordata distribution to climate warming was obviously regional. Eastern Jiangsu, Anhui, and other places would become ecologically suitable areas for C. cordata because of their unique geographical location and climatic environment. The lower latitudes of the south, the original low-altitude areas might no longer be suitable for survival. The central Qinling region was a transition region from north to south, with strong buffer capacity, and climate warming had little effect on its distribution area. The Changbai Mountain and its adjacent areas at higher latitudes were more suitable for C. cordata. 气候变化严重影响植物的地理分布格局,植物分布对气候变化响应的区域差异性也将为引种和保存提供重要的指导.本研究基于ArcGIS与MaxEnt模型,利用176个千金榆地理信息和13个气候变量,对当前和未来的生态位进行重建.结果表明:模型模拟当代潜在分布区可信度极高,测试集与训练集AUC值分别为0.973和0.957.主要核心适宜区集中在秦岭、长白山及其各自的邻近区域,其他为零星的“岛屿”式分布.贵州、江西、云南、福建等地未发现千金榆的分布,但存在一定的适宜分布区.随着未来气候变暖,千金榆生态适宜区明显增加,主要表现为“向高海拔地区收缩”、“北扩”和“东扩”.但核心适宜区略有减少,具体表现为“南缩”、“中稳”和“北扩”.千金榆的分布对气候变暖的响应存在明显的区域性,在东部江苏、安徽等地,因其独特的地理位置与气候环境,该地区开始成为千金榆生态适宜区;在较低纬度的南方,原有的低海拔地区可能不再适宜千金榆生存;中部秦岭地区气候为南北的渐变区,有较强的缓冲能力,气候变暖对其分布区影响不大;较高纬度的长白山地区及其邻近区域愈发适宜千金榆生存.