摘要
在对华北和东北地区杓兰属(Cypripedium L.)植物种质资源调查基础上,应用数量分类学中的Q型聚类分析法对38份杓兰代表样本进行分类研究,并对20个形态性状进行R型聚类分析和主成分分析。Q型聚类分析将这38份样品在Lq1=0.25水平分为山西杓兰群和其他杓兰群两个类群,在Lq2=0.155水平将其他杓兰群分为3个亚群,结果显示大花杓兰(C.macranthos Sw.)与山西杓兰(C.shanxience S.C.Chen)亲缘关系最远,中间过度类型多为大花杓兰与黄囊杓兰(C.calceolus L.)之间的天然杂交后代。R型聚类分析树系图显示各性状分布较为分散,相互独立。主成分分析表明20个性状特征综合成的9个主成分中前5个主成分的累积贡献率高达80.81%,结合其与性状间的相关性,筛选出影响力大的13个重要性状。
Based on our Investigations for endangered Cypripedium wild resource in North and Northeast China, 38 representative samples were studied by Q model cluster analysis of numerical taxonomy, and 20 typical morphological characters were studied in R-model cluster analysis and the principal component analysis to establish the complete Cypripedium classification system. Q-model cluster analysis indicated that the 38 samples were separated into two groups： C. shanxience series group （group A） and the other Cypripedium group （group B） at Lql = 0.25： and the group B were separated into three subgroups at Lq2 = 0.155. It has showed there are nature hybrids between C. macranthos and C. calceolus. R-model cluster analysis indicated that most of selected morphological characters were independent from each other. The principal component analysis suggested that 20 characters can be combinated to 9 main components, and the cumulative contribution rate of the top five main components have arrived as much as 80.81%, 13 dominate characteristics among 20 were selected for their important influence.
出处
《园艺学报》
CAS
CSCD
北大核心
2013年第5期933-942,共10页
Acta Horticulturae Sinica
基金
北京市自然科学基金项目(6072014)
北京市园林局科技项目(YLHH2010002-02)
北京市公园管理中心科技项目(20100203)
关键词
兰科
Q型聚类分析
R型聚类分析
主成分分析
Orchidaceae
Q-model cluster analysis
R-model cluster analysis
principal componentanalysis
