Analysis of Tree Species Suitability for Plantation Forests in Beijing (China) Using an Optimal Random Forest Algorithm

编号 010034702

推送时间 20220613

研究领域 森林生态 

年份 2022 

类型 期刊 

语种 英语

标题 Analysis of Tree Species Suitability for Plantation Forests in Beijing (China) Using an Optimal Random Forest Algorithm

来源期刊 forest

期 第347期

发表时间 20220525

关键词 site quality evaluation;  site suitability;  random forest;  forest management; 

摘要 For afforestation, it is necessary to consider habitat conditions and their impact on specific tree species, in order to enable the selection of appropriate species to improve forest productivity and stand stability. Based on the 2014 Beijing forest management inventory data, we evaluated site quality using theoretical growth equations and quantile regression; we analyzed the effects of climate, topography, and soil variables on the growth of six main tree species using random forest models optimized by a genetic algorithm; and we mapped the potential habitat of six main tree species in Beijing. The results showed that climatic factors were the most important factors affecting tree growth. The prediction models had good accuracy, with an AUC of 0.75–0.85. Among the six main tree species studied, Pinus tabulaeformis Carr. was suitable for all of Beijing’s forest land. Platycladus orientalis (Linn.) Franco, Robinia pseudoacacia Linn. and Salix matsudana Koidz. were suitable for the mountainous areas, while Sophora japonica Linn. and Populus tomentosa Carr. were suitable for planting in the plains area of southeast Beijing. The optimized random forest model applied in this study gives insight into the distribution suitability of the main tree species in Beijing, and could serve as a reference for afforestation design. View Full-Text

服务人员 王璐

服务院士 蒋有绪

PDF文件 浏览全文