大豆生育期组鉴定分组方法的比较研究

Comparative study on identification methods of soybean maturity group

对以往研究中的大豆生育期组鉴定方法进行比较,为选择适宜的鉴定方法提供参考。分别应用区间法、聚类分析法、回归分析法对90份大豆品种进行生育期组鉴定。结果表明,三种方法均能较好地对供试品种进行分组。聚类分析法适用于待分组材料熟期类型丰富、数量均衡,但标准品种的数量较少,生育期组值不精确的情况,其分组的临界值需要根据生育期组的组间距离进行校正。区间法适用于各组内标准品种生育期严格以生育期组的中位数对称分布,对分组临界值界定比较准确。回归分析法适用于标准品种生育期组值精确,标准品种数量较多的情况,可以精准地鉴定供试品种的生育期组。任何鉴定方法的结果都应以确定标准鉴定地点为前提,其鉴定结果应该符合生育期组定义的组间距离等要求。通过三种鉴定方法的对比,分析了各种方法的适用情况和结果,为各地大豆生育期组鉴定提供了理论依据。

The identification methods of soybean maturity group in previous studies were compared to provide reference for selecting appropriate identification methods.Interval method,clustering methodology and regression analysis method were used to identify 90 soybean varieties in their growth period groups.The results showed that the three methods could group the tested varieties well.Clustering analysis method is applicable to the situation that the maturation period of materials to be grouped is rich and balanced,its number of standard varieties is small,and the value of the maturity group is not accurate.The critical value of the grouping needs to be corrected according to the distance between the maturity group.Interval method is suitable for the growth period of standard varieties in each group strictly according to the median distribution of the maturity group,and the definition of the grouping critical value is more accurate.Regression analysis method is applicable to the situation that the value of standard variety maturity group is accurate and the number of standard variety is large,so it can accurately identify the growth period group of tested variety.The results of any identification method should be based on the premise of determining the standard identification site,and the identification results should meet the requirements such as the distance of the maturity group defined by the growth period group.Through the comparison of three identification methods,the applicability and results of each method were analyzed,which provided theoretical basis for the identification of soy‐bean maturity groups in different region.