论文部分内容阅读
在土壤分类中综合运用多种数值分类方法可以取得较好的分类效果。本文应用了主组元分析、聚类分析,多组判别分析和最小支撑树等方法,将横断山区有代表性的30个森林土壤剖面分为六组,各组土壤间不仅在理化性质上有鲜明特征,而且与最重要的成土条件和森林植被类型基本吻合。在西南高山林区,树种与下木对土壤形成有重要作用,主要成土过程为酸性淋溶与腐殖化过程。通过选取的六个理化性质指标,主组元分析结果反映出了上述主导成土过程,而多组判别分析则进一步给予每组土壤以明确的分类域。这两种方法的结合可作为土壤数值分类的基本方法。
Comprehensive classification of soil in the use of a variety of numerical classification methods can achieve better classification results. In this paper, the principal component analysis, cluster analysis, multi-group discriminant analysis and minimum support tree and other methods, the representative of the Hengduan Mountains 30 forest soil profiles are divided into six groups, each group of soil not only in physical and chemical properties Distinctive features, but also with the most important soil conditions and forest vegetation types basically consistent. In the southwest alpine forest area, tree species and understory wood play an important role in soil formation. The main soil-forming process is acid leaching and humification. Through the six indicators of physical and chemical properties selected, the main component analysis results reflect the above-mentioned leading soil-cutting process, while the multiple-group discriminant analysis further gives each group of soil a clear taxonomic domain. The combination of these two methods can be used as a basic method of soil classification.