摘要
在玉米育种过程中,为了选育具有优良性状的玉米杂交种,要对玉米杂交种进行品种比较试验,筛选较优品种,在筛选过程中不仅要分析各参试品种的产量表现,还要综合考虑生育期、株高、穗位高、穗部性状等,才能对各参试品种做较全面的综合评价。聚类分析是一种多元统计方法,对试验品种进行综合分析,得出全面的品种评判。本文采用间比法顺序排列试验,设2次重复的试验方法,应用聚类分析法对云南省农科院配制的28个玉米杂交品种,选用莫A和京杂6号作为对照品种进行聚类分析。研究发现此方法可以弥补常用的方差统计分析方法针对产量等单一性状的局限,为品种试验结果的综合利用提供了科学的依据。
To breed and screen maize hybrid varieties with good quality,trals of variety comparison should be carried out.Besides yield potential,the characteristics of growth period,plant height and height of upper ear and etc.should be taken into consideration in the process of variety screening,and only in this way comprehensive evaluation can be established.Cluster analysis is a pluralistic statistic analysis methods frequently used for comprehensive evaluation on varieties.This study applied parttion comparison method and designed two repeats of field trials.Cluster analysis was conducted on 26 maize hybrid varieties produced by Yunnan Academy of Agriculture Sciences,and 2 varieties of Mo A and Jingza 6 were set as controls.This study found that this method can make up the limitation where only the yield characteristic was analyzed by the traditional way of analysis of variance,and it can also provide basis for scientific utilization of trial results.
出处
《西南农业学报》
CSCD
北大核心
2014年第2期480-484,共5页
Southwest China Journal of Agricultural Sciences
基金
云南省现代农业玉米产业技术体系专项资金
关键词
聚类分析
玉米
杂交种
筛选
Cluster analysis
Maize
Hybrid varieties
Screening
