文献类型：期刊文献

中文题名：应用最大熵模型预测的欧李潜在适生区分布及气候变化对其的影响

英文题名：Distribution of Potential Suitable Areas of Cerasus humilis by Using Maxent Model and Effect of Climate Change on It

作者：张雅茜[1] 王淋[2] 包福海[1] 张淑宁[1] 红梅[1] 刘一超[1] 陈俊兴[1] 蔺悦[1] 敖敦[1] 白玉娥[1] 包文泉[1]

第一作者：张雅茜

机构：[1]内蒙古农业大学,呼和浩特010018;[2]中国林业科学研究院经济林研究所

年份：2023

卷号：51

期号：11

起止页码：54-62

中文期刊名：东北林业大学学报

外文期刊名：Journal of Northeast Forestry University

收录：CSTPCD;;北大核心:【北大核心2020】；CSCD:【CSCD_E2023_2024】；

基金：内蒙古农业大学青年教师科研能力提升专项项目(BR220114)。

语种：中文

中文关键词：欧李;最大熵模型;适生区;环境因子

外文关键词：Cerasus humilis;Maxent;Suitable area;Environmental factor

分类号：S662.5

摘要：为预测我国欧李适生区分布及气候变化对其的影响。利用R语言和ArcGis筛选出欧李(Cerasus humilis(Bunge)Sok)134个分布数据和25个环境因子;调用ENMeval程序包,优化最大熵生态位模型参数;根据Pearson相关分析和方差膨胀因子(VIF)分析,筛选出建模所需的环境因子,利用刀切法评估影响欧李适生区的主导环境因子。利用优化后的模型,构建欧李适生区的地理分布,推测末次间冰期、末次冰盛期和全新世中期的潜在适生区,分析其历史时期适生区分布变化,预测不同气候情景欧李未来分布区的变化趋势。模型预测结果表明:在默认参数设置时,调控倍率为1,特征组合为LQHPT,delta.AICc值为297.65。将模型参数优化后,在特征组合为LQH,调控倍率为2.5时,delta.AICc值为0,最大熵模型的拟合度最佳。据刀切法评估结果,降水量变异系数、最冷月极端低温、海拔、最湿月降水量、气温变异系数是影响欧李分布的主导环境因子。当前我国欧李的适生区主要分布于天津、北京、山东、河北地区;气候变化对欧李的影响较为明显;未来气候变暖将会引起欧李适生区的向北迁移,尤其在温室气体高排放浓度下更敏感,迁移距离更大。
In order to predict the distribution of the suitable area of Cerasus humilis and the influence of climate change on it,134 distribution data and 25 environmental factors were screened by using R language and ArcGis.The ENMeval package was invoked to optimize the Maxent parameter of the maximum entropy niche model.By Pearson correlation analysis and VIF variance inflation factor analysis,the environmental factors required for modeling werescreened out,and the knife cutting method was used to evaluate the leading environmental factors affecting the suitable area of C.humilis.The optimized model was used to construct the geographical distribution of the suitable area of C.humilis,and to predict the potential suitable areas of the last interglacial,the last glacial maximum and the middle Holocene.The distribution changes of suitable areas in the historical period were analyzed,and the future distribution trends of C.humilis were predicted under different climate scenarios.The model prediction results show that the delta.AICc is 297.65 at the default parameter setting with a modulation multiplier of 1 and a feature combination of LQHPT.After optimizing the model parameters to a feature combination of LQH modulation multiplier of 2.5,the delta.AICc is 0 and the Maxent model has the best fit.According to the knife method,Bio15,Bio6,Bio13 and Bio4 are the main environmental factors affecting the distribution of C.humilis.At present,the suitable areas of C.humilis are mainly distributed in Tianjin,Beijing,Shandong and Hebei.It was found that C.humilis was more sensitive in response to climate change.The future climate warming will lead to the northward migration of the C.humilis distribution area,especially under the high concentration of greenhouse gas emissions,which is more sensitive and the migration distance is larger.