欢迎访问《生态学杂志》官方网站,今天是 分享到:

生态学杂志

• 方法与技术 • 上一篇    下一篇

基于MaxEnt模型的珙桐在中国潜在适生区预测

王雨生1,王召海1*,邢汉发1,厉静文2,孙硕1   

  1. 1山东师范大学地理与环境学院, 济南 250014;2中国林业科学研究院荒漠化研究所, 北京 100091)
  • 出版日期:2019-04-10 发布日期:2019-04-10

Prediction of potential suitable distribution of Davidia involucrata Baill in China based on MaxEnt.

WANG Yu-sheng1, WANG Zhao-hai1*, XING Han-fa1, LI Jing-wen2, SUN Shuo1   

  1. (1College of Geography and Environment, Shandong Normal University, Jinan 250014; 2Institute of Desertification Studies, Chinese Academy of Forestry, Beijing 100091, China).
  • Online:2019-04-10 Published:2019-04-10

摘要: 珙桐为我国特有珍稀树种,预测珙桐潜在适生区可为珙桐物种保护与研究提供重要参考。基于387条中国境内样本分布数据和27个环境因子,利用MaxEnt模型与地理信息系统(geographic information system, GIS),对珙桐在中国的潜在适生区进行预测,并通过受试者工作特征(receiver operating characteristic, ROC)曲线对模型精度进行验证。结果表明:ROC曲线下面积(AUC)平均值为0.951,模型预测结果准确性较高,适合珙桐的潜在适生区预测;珙桐潜在适生区主要位于四川、云南、贵州、湖南、湖北、重庆、陕西和西藏等地区,高适生区呈不规则环状分布,西起横断山区,北达秦岭武当山荆山一线,东抵张家界雪峰山一带,南至贵阳,在西藏东南部的雅鲁藏布江河谷和安徽西部的大别山地区也有分布;气温年较差、年降水量、最冷月最低温、昼夜温差、降水量变异系数等5个环境因子对珙桐潜在适生区预测影响最大,其他环境因子,如地形、土壤、水文等影响较小。

关键词: 消落带, 富营养化, 分解, 氮磷释放

Abstract: Davidia involucrataBaill is a rare and precious tree species in China. The prediction of its suitable habitats can provide important reference for its protection and research. Based on 387 sample distribution and 27 environmental factors in China, the MaxEnt model and geographic information system (GIS) were used to predict the potential suitable habitats. The accuracy of the model was verified by receiver operating characteristic curve (ROC). The results showed that the mean value of the area under ROC curve was 0.951, and the accuracy of the model was high, which was suitable for predicting potential suitable habitats of D. involucrata. The potential suitable habitats for D. involucrata were mainly located in Sichuan, Yunnan, Guizhou, Hunan, Hubei, Chongqing, Shanxi, and Tibet. The highly potential suitable habitats distributed in an irregular circular shape, with the Hengduan Mountains to the west, the Qinling WudangJinshan Mountains to the north, the ZhangjiajieXuefeng Mountains to the east and the south to Guiyang. In addition, they are also distributed in the Yarlung Zangbo River Valley in the southeast of Tibet and the Dabie Mountains in western Anhui. The major factors determining the potential suitable distribution were annual temperature range, annual precipitation, minimum temperature of coldest month, diurnal temperature difference, and annual precipitation variation coefficient. Other environmental factors, such as topography, soil, hydrology, had limited influence on the distribution of D. involucrata.

Key words: reservoir riparian region, eutrophication, decomposition, release of nitrogen and phosphorus.