基于BLUP和GGE双标图的林木多地点试验分析

Forestry multi-environment trial analysis based on BLUP and GGE biplot

【目的】建立基于无偏预测值(BLUP)与基因型主效加基因型-环境互作效应(GGE)双标图的分析模型,以提高林木多点试验数据分析的准确性。【方法】以火炬松36个基因型在6个试验地(S1~S6)的种子产量为基础数据,利用ASReml软件对实测数据进行空间变异结合因子分析法的模型拟合,以获取各地点下每个基因型的BLUP值;从试验地划分、试验地评估和林木基因型评估3个方面,对原始数据、BLUP数据进行GGE双标图分析与比较。【结果】BLUP数据具有明显的空间变异,比原始数据具有更高的产量变异解释能力;原始数据和BLUP数据的试验地分组结果一致,均分为2组,但BLUP数据的试验地点间的相关关系变弱;原始数据的理想试验地为地点S5,而BLUP数据为地点S1;原始数据和BLUP数据的最理想基因型均为21,但2种数据高产和稳产基因型的一致性比较低。【结论】基于BLUP与GGE双标图相结合的模型,可用于林木多点试验分析,其比原始数据的GGE双标图分析结果更为可靠。

【Objective】This study established an analysis model based on best linear unbiased prediction（BLUP）and genotype main effect plus genotype-by-environment interaction（GGE）biplot to improve the accuracy of forestry multi-environment trial analysis.【Method】The BLUP data of all genotypes in each site was obtained by ASReml software from model fitted by spatial effects with factor analytic method using seed yield data of 36 genotypes of Pinus teada at six trail sites（S1-S6）.Then the original data and BLUP data were analyzed and compared by GGE biplot from mega-environment analysis,test-environment evaluation and genotype evaluation.【Result】BLUP data had obvious spatial variability pattern,and had higher yield variation ability than the original data.The experimental sites in original data and BLUP data were both divided into two same groups,but the correlations between sites were weaker in the BLUP data.The ideal site of the original data was S5,while that for BLUP data was S1.The ideal genotype of both the original data and BLUP data was 21,but the consistency of genotypes with high yield and stability between the two data was low.【Conclusion】The BLUP and GGE biplot based model could be used for forest multienvironment trial analysis and it was more reliable than GGE biplot on original data.