品种区试的多对照设置及适应性参数估计法的改进

求聚类分析的终止判据,将所有参试品种的b_1值聚归为类间有显著差异的几个组,从而可以分析具体品种间b_1的差异程度。也可应用此法探讨多个对照种的选设问题。采用成对GE互作均方差异显著性测验,可以间接地提供参试品种适于哪些地区推广的信息。以品种区城试验多年的资料计算出各试点的平均产量作环境指数,用回归分析法估计出的适应性参数b_1值,可能会更好地反映各品种的实际适应情况。

STUDIES ON MULTI-CHECK DESIGN IN CULTIVAR REGIONAL TRIAL AND IMPROVEMENT FOR ADAPTABILITY PARAMETER ESTIMATIONS

A set of wheat regional trial data was used as an example to dcmonstrat the analysis approach. (1) Genotypes are grouped according to similarity of dissimilarity index, d(v_? V_1). When every group contains one check, and one chock only, the choice of chocks may be, regarded as ideal. (2) An analysis method is based on pairwise GE interaction mean squares (MS(GE)) between the checks and the test cultivars. (3)To improve the adaptability parameter estimations, the authors used the mean performance of all genotypes grown in an environment for multi-years as an environmental index, and also used a linear regression approach. The performance of each gcnotype was plotted against this index for each environment, and simple linear regression fitted by least squares to summarise the genotype's response.