44 ground-cover dendranthema varieties and 2 wild relatives were studied with 31 morphologic characters including 17 quantitative multistate characters, 12 quantitative multistate characters and 2 two-state characters by SPSS18.0. Q cluster was used in classification, and 31 characters were analyzed by R cluster and principal component analysis. 46 operational taxonomic units were divided into two types by Q cluster; one was Ground-cover Dendranthema varieties and the other was wild germplasm, and the varieties could further be divided into four small groups. R cluster analysis showed that 31 characters are significantly clustered into four groups, and some of them have significant correlation, but most of them are independent. In principal component analysis, the accumulative contribution of the first five principal components is 59. 00% . According to the total loading of characters, seven characters are more important in classification.%对44个地被菊品种和2个菊属近缘野生种中的重要性状采用SPSS18.0进行分析,其中数量多态性状17个,定性多态性状12个,二元性状2个.应用数量分类学中的Q型聚类方法进行分类,并对31个性状进行R型聚类分析和主成分分析,通过Q型聚类将46个分类运算单位分为特征明显的2大类群,即野生种和地被菊品种,品种中又可以分为4个小类；R型聚类分析结果将31个性状明显聚为4组,部分性状相关性较强,但多数表现较为独立；主成分分析中,前5个主成分累积贡献率达59.00％；根据各性状的总载荷量,选出分类中比较重要的7个性状.
展开▼
