摘要
生物产量是青贮玉米的重要考评指标。采用相关分析、主成分分析、聚类分析、判别分析及岭回归分析方法,对我国西南地区37个青贮玉米品种的10个与青贮玉米生物鲜重相关的农艺性状进行了分析。结果表明:各农艺性状间相关性不同,利用单个性状与生物鲜重的相关性进行分析误差较大;主成分分析的4个主成分因子累计贡献率为78.106%;聚类分析可将参试品种分为四大类;判别分析表明判对概率为89.19%;通过岭回归分析建立了一套青贮玉米产量评价模型。
Biological yield is an important evaluation index for silage corn.This paper used methods such as correlation analysis,principal component analysis,cluster analysis,discriminant analysis and ridge regression analysis to analyze 10 agronomic traits related to the biological fresh weight of 37 silage corn varieties in southwestern China.The results showed that the correlation between various agronomic traits was different,and there was a significant error in analyzing the correlation between individual traits and biological fresh weight.The cumulative contribution rate of the four principal component factors in principal component analysis was 78.106%.Cluster analysis could classify the tested varieties into four categories.Discriminant analysis showed that the probability of judgment was 89.19%.A yield evaluation model for silage corn was established through ridge regression analysis.
作者
杨云
金容
夏清清
李仕伟
杨凯
蒲全波
YANG Yun;JIN Rong;XIA Qingqing;LI Shiwei;YANG Kai;PU Quanbo(Nanchong Academy of Agricultural Sciences,Nanchong Sichuan 637000;Shehong City Vocational School,Shehong Sichuan 629200)
出处
《现代农业科技》
2024年第16期16-19,共4页
Modern Agricultural Science and Technology
基金
国家玉米产业技术体系项目(CARS-02-88)
南充市科技局项目(22JCYJPT0031)。
关键词
青贮玉米
生物产量
农艺性状
相关分析
主成分分析
聚类分析
判别分析
岭回归分析
silage corn
biological yield
agronomic trait
correlation analysis
principal component analysis
cluster analysis
discriminant analysis
ridge regression analysis
