摘要
为了建立完善的藤本月季分类评价体系,本研究以65个藤本月季为材料,筛选了23个表型指标,计算并分析了表型性状分布频率(Pi)、变异系数(CV)和Shannon-Weaver指数(H’),进行了R型聚类分析、主成分分析和有序样本最优分割聚类分析。结果显示:供试月季品种描述性状H’为0.64~1.70,数量性状变异系数CV为12.31%~95.86%,数量性状H’为1.30~2.03,藤本月季性状多样性有差异,除花数、刺数量外,数量性状变异潜力有限;R型聚类中部分性状呈强相关性,23个指标简化为18个性状指标;主成分分析中前6个主成分特征值大于1,累计贡献率为64.65%;有序样本最优分割聚类中基于主成分综合得分,将藤本月季分为4组,组间差异显著,体现不同的观赏特点。以上结果为藤本月季品种资源评价、品种选育与绿化应用提供了理论依据。
To improve the cluster and comprehensive evaluation system of climbing roses,65 climbing rose varieties were as materials.twenty-three phenotype characters were collected.Distribution frequency(Pi),coefficient of variation(CV),and Shannon-Weaver index(H’)were analyzed.R cluster,principal component analysis and sequential clustering method were conducted.The results showed that H’of description characters was 0.64~1.70,CV and H’of numerical characters were 12.31%~95.86%and 1.30~2.03,respectively.Character diversity varied in climbing roses.Except the number of flowers per inflorescence and prickles,the variation potential of numerical characters was limited.R cluster analysis showed some characters closely related and 23 characters were simplified into 18 characters.The principal component analysis showed that 6 principal components were extracted for their eigenvalue was above 1.The cumulative contribution of the top 6 principal components was 64.65%.Based on sequential clustering method,the comprehensive scores of the principal component of climbing roses were divided into 4 clusters.The classification of each cluster reflected the ornamental performance of climbing roses.This study provided theoretical basis for evaluation,breeding and application in urban landscaping of climbing rose varieties.
作者
吉乃喆
华莹
赵世伟
崔荣峰
周燕
Ji Naizhe;Hua Ying;Zhao Shiwei;Cui Rongfeng;Zhou Yan(Beijing Key Lab of Greening Plants Breeding,Beijing Academy of Forestry and Landscape Architecture,Beijing,100102)
出处
《分子植物育种》
CAS
北大核心
2023年第4期1294-1305,共12页
Molecular Plant Breeding
基金
北京市公园管理中心(ZX2019022
ZX2020023)资助。
关键词
藤本月季
表型多样性
主成分分析
Climbing rose varieties
Phenotype diversity
Principal component analysis
