南方三省黄麻品种丰产性稳定性测定及其分析方法的探讨

用Duncan、Finlay-Wikinson、Eberhart-Russell的方法和模式分析长江流域地区1980～1983年黄麻区域试验中9个品种的丰产性和稳定性。结果表明,参试品种间产量和稳定性参数都有相当大的差异和不同的适应性。为提高区域试验分析的准确性,应把Bartlett方差同质性检验列为联合方差分析的一个基本检验步骤。其联合方差分析的数学模型以选用混合模型即品种效应为固定模型,地点、年份效应为随机模型较为确切。用两种不同模式和方法估算品种稳定性,Eberhart较之Finlay的方法为佳,其定义和标准更符合生产实际。产量与稳定性参数回归系数b之间并无明显相关性。

AN INVESTIGATION ON THE STABLE HIGH YIELD CHARACTERISTICS AND ITS ANALYTICAL METHOD OF JUTE CULTIVARS IN THE THREE SOUTHERN CHINESE PROVINCES

By employing Duncan's, Finlay-Wik inson's, and Eberhart-Russell's, method and models. nine high yield and stable jute cultivars under experiment in the three southern Chinese provinces, 1980-1983 were analyzed.Result indicated that these cultivars had rather great differences and various degrees of adaptation regarding yield and stability parameters between cultivars. In order to increase the accuracy of regional experimental analysis. the Bartlett square deviation for the homogeneity probe must be made use of as a basic investigation step in the combined square deviation analysis. The mathematical model in the combined square deviation analysis should select the mixed model or known as cultivar effect of the fixed model. while location and year effects are more appropriate to be considered as random model.Comparing the two different methods or models in the evaluation of cultivar stability. Eberhart's is preferred to Finlay's. Its defination and standardization are closer to practical production. No significant correlation existed between yield and regression coefficient b stable parameter.