空间效应模型在青贮玉米新品系评比试验中的应用  

Application of Spatial Effect Model in Comparative Test of New Silage Maize Lines

作　　者：焦金龙 向海游 杨子和 赵丽 张艳茹 王伟 吴元奇[1] JIAO Jinlong;XIANG Haiyou;YANG Zihe;ZHAO Li;ZHANG Yanru;WANG Wei;WU Yuanqi(Maize Research Institude,Sichuan Agricultural University,Chengdu 611130,China;College of Agronomy,Sichuan Agricultural University,Chengdu 611130,China)

机构地区：[1]四川农业大学玉米研究所,成都611130 [2]四川农业大学农学院,成都611130

出　　处：《四川农业大学学报》2020年第2期121-126,共6页Journal of Sichuan Agricultural University

基　　金：国家重点研发计划(2016YFD0300309);四川省教育厅项目(17EA0317)。

摘　　要：【目的】为了提高青贮玉米田间品种比较试验的准确性,降低空间变异对品种试验评比结果的影响。【方法】在非标准随机区组的条件下,利用SAS9.4对23个青贮玉米品种的单株产量进行方差分析和空间效应模型分析。【结果】研究结果表明:试验的误差存在典型的空间相关性,利用剩余误的误差的空间协方差可降低供试品种效应估计的误差和提高品种效应差异F检验与T检验的效率。在不同玉米品系平均单株产量分析中,空间效应模型在拟合效果、效应估计、误差估计方面都优于传统方差分析。田间试验存在的一些空间差异通过空间效益模型得到控制。【结论】遇到非标准随机区组设计时,建议采用空间效应模型分析不同青贮玉米品种田间试验数据,从而提高数据分析的准确性。【Objective】In order to improve the accuracy of field variety comparison experiment,reduce the effect of spatial variation on the result of variety test evaluation.【Method】Under the condition of non-standard random block,SAS9.4 was used to analyze the yield per plant of 23 silage maize varieties by variance analysis and spatial effect model.【Result】The results show:the error of the test has typical spatial correlation,the spatial covariance of residual error can be used to reduce the error of variety effect estimation and improve the efficiency of F test and T test.In the analysis of the average yield per plant of different maize strains,The spatial effect model is better than the traditional analysis of variance in fitting effect,effect estimation and error estimation.Some spatial differences in field experiments were controlled by spatial benefit models.【Conclusion】In the case of non-standard random block group design,it is suggested to use spatial effect model to analyze field test data of different silage maize varieties,so as to improve the accuracy of data analysis.

关 键 词：青贮玉米 非标准田间试验设计 方差分析 空间效应模型 

分 类 号：S513[农业科学—作物学]