摘要
以23个棉花品种接种18个黄萎病菌株的抗感表现(病情指数)为分析对象,介绍了广义线性模型分析无重复试验互作效应的方法,并对其结果与AMMI模型分析、常规的方差分析及数据转换后的分析结果进行了比较。结果表明:常规方差分析不能对无重复试验资料分析其互作效应,广义线性模型和AMMI模型能分析无重复试验的互作效应,但广义线性模型方法是更为简便而有效方法。
In studying the interaction between the strains and cotton varieties, the disease indices of 23 different resistant or susceptible varieties inoculated with 18 isolates of Verticillium dahliae were analyzed by ANOVA, generalized linear model (GLM), additive main effects and multiplicative interaction (AMMI) model, and data transformation. The results showed that, for unreplicated data, neither ANOVA nor data transformation has the capability to treat the interaction, GLM and AMMI can analyze it. Compared with AMMI, GLM is a more efficient and sampling method to analyze the interaction of unreplicated data.
出处
《浙江大学学报（农业与生命科学版）》
CAS
CSCD
北大核心
2003年第1期71-75,共5页
Journal of Zhejiang University:Agriculture and Life Sciences
基金
国家自然科学基金资助项目(39970122)
关键词
黄萎病菌
品种
棉花
黄萎病
互作
广义线性模型
病情指数
cotton
Verticillium wilt
Verticillium dahliae
interaction
generalized linear model
