【摘要】 AMMI模型是一种有效的加性主效应和乘积交互作用 (Additivemaineffectsandmulti plicativeinteraction ,简记为AMMI)模型 ,它发展了分析基因型与环境互作的统计方法。本文以1996和 1997年 (秋播 )的云南省油菜区域试验的小区产量为资料 ,应用AMMI模型分析研究油菜区域试验。通过分析得出 :使用AMMI模型能提高估计的准确性、提高选择的概率。例如 ,此组区域试验用线性回归模型分析时 ,剩余残差占交互作用的 74 60 % ;而在AMMI模型中 ,残差仅占交互作用的 16 63%。在AMMI分析中 ,显著的主成分轴 (PCA1～ 3)用尽量小的自由度捕捉尽量大的SS (其中 ,PCA1用 2 4 %的互作自由度捕捉了 4 8 72 %的互作SS) ,而剩余大自由度对应小的SS ,作为试验误差处理 ,提高试验精确度。特别是AMMI模型利用对应分析图可以直观评价品种的稳定性和较适应地区 ,为其推广某一品种和在某一区域推广这一品种等科学决策提供依据。当显著的PCA轴多于 3个以上时 ,可通过D值的大小来确定品种的稳定性。由此得出结论 ,此模型可提高试验精确度 ,减少试验重复数 ,评价品种更精确 ,选择出的品种更可靠 ,且能够直观地得出各参试品种的稳定性和较适应地区。它是一种花费小、简便易行、经济实用的统计分析方法更多还原

【Abstract】 The genotype-by-environment interaction effect analyzed using the additive main effects and multiplicative interaction (AMMI) statistical model was a useful method, particularly developed a large genotype-by-environment interaction.In Yunnan province rape varieties regional trials in two years result in balanced data with 10 genotypes grown in 10 environments were analyzed by AMMI model.AMMI model firstly used for additive analysis of variance, and then for the multiplicative principal components analysis (PCA) model to the residual from the additive model. It can be used to analyze interaction by PCA,So its uses can improve precision of yield trials and reduce replicates of trials. Meanwhile, correspondence analysis can be used in the AMMI to evaluate variety stability and adaptation regions, So the AMMI model is more accurate to estimate yield, and it is more reliable in screening superior genotypes.更多还原