基于R语言的AMMI模型和GGE双标图在大豆区试中的应用评价

Application Evaluation of AMMI Model and GGE Biplot Based on R Language in Soybean Regional Test

为提高甘肃省大豆品种的选育和应用效率,利用大豆区域试验数据,从基因型与环境的互作分析出发,对甘肃省大豆新品种的稳定性、适应性以及各试点的鉴别力进行全面评估。本研究采用AMMI模型与GGE双标图相结合的方法对甘肃省9个大豆品种在5个试验点的产量进行分析,结果表明,AMMI模型中主成分值(IPCA1、IPCA2)占总变异平方和的95%;其中‘中黄318’属于高产稳产性品种,而‘陇黄3号’和‘铁丰31’虽然产量较高,但其稳定性中等,适合在特定区域栽培。在5个试验点中,凉州分辨力最强,镇原分辨力较弱。综合运用AMMI模型和GGE双标图法,能够更准确直观地反映各品种生产力、稳定性和适应能力,以及在不同试验区域的分辨能力和代表性。

To improve the selection and application efficiency of soybean varieties in Gansu Province,data from soybean regional trials were used to comprehensively assess the stability and adaptability of new soybean varieties in Gansu Province and the discrimination power of each pilot site in terms of the interaction analysis between genotype and environment.This paper used the AMMI model combined with GGE double-labeled plots to analyze the yields of nine soybean varieties in Gansu Province at five trial sites.The results showed that the principal component values(IPCA1 and IPCA2)in the AMMI model accounted for 95%of the sum of squares of the total variation;among them,‘Zhonghuang 318’was a high-yield and stable variety,while‘Longhuang 3’and‘Tiefeng 31’had higher yields but moderate stability,making them only suitable for cultivation in specific areas.Among the five trial sites,Liangzhou had the strongest discrimination power and Zhenyuan was weaker.The combined use of the AMMI model and the GGE double-labeled map method can more accurately and intuitively reflect the productivity,stability and adaptability of each variety as well as the discrimination power and representativeness of each trial site.