摘要
采用组平均聚类法、模糊聚类法、主成分分析法对广东不同被地利用方式对土壤养分含量的影响进行聚类分析.综合3种分类方法,可分为5大类:第1类包括鹤山茶园和鹤山坡地花生;第2类包括鹤山茶园空地、林地有覆盖和无覆盖、芒果地、草地、荒坡地和丰顺坡地青榄;第3类包括鹤山坡地龙眼(新植)、番薯,丰顺坡地龙眼和和高明坡地芒果;第4类包括鹤山坡地龙眼,丰顺坡地茶园,曲江坡地花生和沙田柚;第5类包括高州坡地甘蔗,南雄坡地黄烟、银杏、三华李.结果表明3种分类方法均可用来进行分类,但主成分分析法损失了较多的信息。
Group mean cluster analysis(GCA), fuzzy cluster analysis(FCA) and principal component analysis (PCA) were used to elassny soil nutrient on hillsides in Guangdong Province. The soil nutrient on hillsides could be classified into five types by integration of the forve three methods: Type Ⅰ. soil with tea and groundnut in Heshan; Type Ⅱ. bare soil of the tea plantation, forest land with or without mulch,mango plantation, grass land, bare soil in Heshan and olive plantation in Fengshun; Type Ⅲ. soil planted to longan or sweet potato in Heshan, longan plantation in Fengshun and mango in Gaoming; Type Ⅳ.soil planted to longan in Heshan, tea in Fengshun, groundnut and pummelo in Qujiang; Type V. soil planted to sugarcane in Gaozhou, tobacco, gingko and plam in Nanxiong. The results showed that the three methods could be used, but the PCA method lost som information and the classification effect had limitations.
出处
《华南农业大学学报》
CAS
CSCD
北大核心
1999年第4期90-93,共4页
Journal of South China Agricultural University
基金
国家教委留学人员基金!125
广东省自然科学基金!950389