【摘要】 基于Maxent模型,利用世界气候数据库的末次盛冰期至2070年前后的19个生物气候变量资料,在对存在样本点数量评估和主要参数筛选的基础上,综合Jackknife检验、置换重要值和百分比贡献率,探讨影响我国东北地区寒温带针叶林分布的主要环境因子,分析不同时期寒温带针叶林在东北地区潜在适生区的变化特征,结果表明:(1)Maxent模型的特征参数对环境变量响应曲线影响最大,调控倍频其次,最大背景点数最小;(2)特征参数为Threshold、调控倍频为1时,受试者工作曲线下面积(AUC)最高;(3)影响寒温带针叶林分布的主要温度因子为最暖季平均气温、年平均气温和最冷月最低气温,主要水分因子为年降水量、降水的季节变量和最干季降水量;(4)东北地区寒温带针叶林的主要潜在适宜分布区域在大兴安岭,末次盛冰期分布范围最大,为20.58万km2,从全新世中期范围开始收缩,到当代面积最小,在未来面积会有小幅增加。更多还原

【Abstract】 With the Maxent model,we used the data of 19 bioclimatic variables during the Last Glacial Maximum,Mid-Holocene,Present,the year 2050 and the year 2070 from the world climate database to evaluate the number of sample points and to screen the main parameters. And we investigated the environmental factors affecting the distributions of cold-temperate coniferous forest in Northeast China and analyzed its variation characteristics of potentially suitable habitat in different periods by means of Jackknife test,permutation importance and percent contribution. The influence of feature combinations( FC) on response curves of environmental variables was the greatest,followed by regularization multipliers( RM) and max number of background points( BC). When FC was threshold and regularization multiplier was 1,the area under curve( AUC) of receiver operating characteristic( ROC) was the largest. The main temperature factors affecting the distributions were mean temperature of warmest quarter( BIO10),annual mean temperature( BIO1) and min temperature of coldest month( BIO6). The main moisture factors were annual precipitation( BIO12),precipitation seasonality( BIO15) and precipitation of driest quarter( BIO17). The main potential suitable distribution area was Daxing’an Mountains. In the Last Glacial Maximum,the distribution range is the largest with 205 800 km~2 and it began to shrink from the Mid-Holocene. Its area was the smallest at present and would increase slightly in the future.更多还原