Abstract: International Hybrid Rice Observational Nursery (IRHON) of international Network of Genetic Evaluation for Rice (INGER) in 1994 consisted of 89 cultivars, conducted at 17 locations in six different countries. The augmented design was used in the nursery. The traditional ANOVA and linear models are ado longer able to ex plore cultivar and location interactions effectively in such large scale multilocation trials. In this paper the Additive Main Effects and Multiplicative interaction (AMMI) model is used, where the interaction is approximated by sum of a few multiplicative terms. By means of biplots the cultivars with specific adaptation to certain locations can be iden tified. Moreover, if the environmental data are available, then AMMI model will provide insight into the interaction. The power and the efficiency of AMMI models is demonstrated by analyzing IRHON data.

Key words: Additive Main Effects and Multiplicative interaction, augmented design, biplot, special adaptability, cultivar and site interaction

摘要： 国际水稻遗传评价网 （INGER）的1994年杂交水稻圃（IRHON）采用增广设计,包括了6个国家的17个地点和89个品种。为充分了解这类多点试验中关于品种和地点的交互作用，用加性主效应和乘积交互作用(AMMI)模型，以若干项乘积之和估算品种和地点的交互作用。借助双标图，可鉴别对地点有特殊适应性的品种。如果有参试地点的环境因子的数据，AMMI 模型还可帮助我们用环境因子解释品种和地点的交互作用。通过本文对1994年杂交水稻圃数据的分析,我们看到AMMI模型是分析多点试验数据的非常有效的工具，能为育种项目和决策部门提供非常有价值的品种适应信息。

关键词: 增广设计, 品种和地点交互作用, 特殊适应性, 加性主效应和乘积交互作用模型, 双标图