论文部分内容阅读
用水稻、小麦和大豆的 8个三重复品系试验的产量结果 ,进行了作物育种品系试验的三种分析方法与随机完全区组分析 (RB)的比较。此三种分析方法分别是 ,行列区组分析 (RC)、行列区组及近邻小区分析 (RCN)和行列区组及以小区丛数为协变量的协方差近邻分析 (COV)。结果表明 ,以误差均方作为相对效率指标 ,RC、RCN、COV相对于RB的平均相对效率分别为 136 .38± 31.2 2 %、 14 7.38± 38.96 %和 2 19.6 6± 2 6 .0 8%。由此建议在作物育种品系比较试验中 ,在产量鉴定时要重视对小区株数等性状的调查以供协方差分析 ,采用行列区组及近邻小区分析是一种值得提倡的简易可行方法。
A comparison of three analysis methods for crop breeding lines with randomized complete-region analysis (RB) was conducted using the yield results of eight triplicate lines of rice, wheat and soybean. The three methods of analysis are Row Group Block Analysis (RC), Row Group and Neighboring Community Analysis (RCN) and Row Group, and Covariance Nearest Neighbor Analysis (COV) with cohort of clusters as covariates. The results showed that the average relative efficiencies of RC, RCN and COV relative to RB with mean error square as the relative efficiency index were 136.38 ± 31.2 2%, 14 7.38 ± 38.96% and 2 19.6 6 ± 26.080% . Therefore, it is suggested that in the comparative experiments of crop breeding lines, attention should be paid to the investigation of the traits such as the number of trees and so on for the analysis of covariance during the identification of yield. It is a feasible and feasible method to adopt the analysis of the ranks and neighborhoods.