北方春大豆品种农艺和品质性状分析与综合评价  

Analysis and comprehensive evaluation of agronomic and quality traits of spring soybean varieties in northern China

作　　者：聂波涛 刘德泉 陈健[1] 崔正果 侯云龙[1] 陈亮[1] 邱红梅[1] 王跃强[1] NIE Bo-Tao;LIU De-Quan;CHEN Jian;CUI Zheng-Guo;HOU Yun-Long;CHEN Liang;QIU Hong-Mei;WANG Yue-Qiang(Soybean Research Institute,Jilin Academy of Agricultural Sciences,Changchun 130033,Jilin,China;College of Agriculture,Yanbian University,Yanji 133002,Jilin,China)

机构地区：[1]吉林省农业科学院大豆研究所,吉林长春130033 [2]延边大学农学院,吉林延吉133002

出　　处：《作物学报》2024年第9期2248-2266,共19页Acta Agronomica Sinica

基　　金：吉林省科技发展计划项目(20220202012NC);吉林省农业科技创新工程(CXGC2022RCG005)和吉林省农业科技创新工程(CXGC2022RCY008)资助。

摘　　要：大豆种质资源是品种改良的关键。本研究对334份我国北方春大豆种质资源的主要农艺性状和品质性状进行了变异分析、相关性分析、聚类分析、主成分分析、分段线性回归分析和逐步多元回归分析。聚类分析结果表明,可将334份大豆种质划分为4个类群,其中类群Ⅲ主要为吉林地区杂交育成大豆品种,在本地区平均单产高,抗倒伏能力强,综合表现较好。主成分分析表明, 5个主成分分别代表产量因素、品质因素、株高因素、分枝与开花因素和开花至成熟时间因素,累计贡献率达到79.76%。使用DTOPSIS法对大豆品种进行评分,综合表现最好的是编号36的大豆品种。多元线性逐步回归分析表明,单株荚数、百粒重、油分、主茎节数、单株粒数对单产影响显著;油分、单株荚数、分枝数、株高和开花时间对蛋白质影响显著;蛋白质、百粒重、开花时间、开花至成熟时间、单株荚数和单产对油分影响显著。依据分段线性回归分析,并综合相关性分析和多元回归方程,构建高产大豆理想株型为:主茎节数尽量增加;五节高8.00~14.70 cm;单株荚数大于132.83个;分枝数在6个左右;开花时间57.16 d左右;开花至成熟时间大于61.29d,整个生育日期控制在霜降期以内;不倒伏或轻倒;单株粒数250.43粒左右;蛋白质38.87%左右;油分大于19.27%;百粒重19.45~22.55g。高产高油大豆理想株型为:株高113.50~114.45cm左右;五节高12.15~14.70 cm;主茎节数不超过25.15节;单株荚数132.83~142.65个;分枝数0~3.00个;不倒伏或轻倒;开花时间57.16 d左右;开花至成熟时间61.29~69.00 d;单株粒数249.63粒左右;百粒重19.25~22.55 g。高产高蛋白大豆理想株型为:株高113.50~114.45cm左右;五节高8.00~10.55cm左右;主茎节数19.42节左右;单株荚数132.83~138.96个;分枝数4.72个左右;倒伏性可为轻倒;开花时间63.00d左右;开花至成熟时间68.52d左右;单株粒数227.54~250.43粒;百�Soybean germplasm resources are the key to the improvement of soybean variety.Variation analysis,correlation analy-sis,cluster analysis,principal components analysis,piecewise linear regression analysis,and stepwise multiple regression analysis were conducted on the main agronomic and quality traits of 334 spring soybean germplasm resources in northern China.The re-sults of cluster analysis showed that 334 soybean germplasm resources could be divided into 4 groups,among which group III mainly include cross breeding soybean varieties in Jilin region,with the high average yield per unit,strong lodging resistance,and better comprehensive performance in this region.Principal components analysis showed that the five principal components repre-sented yield factor,quality factor,plant height factor,branching and flowering factor,and flowering to maturity time factor,with the cumulative contribution rate of 79.76%.The soybean variety with the best overall performance was No.36 using DTOPSIS method.Multiple linear stepwise regression analysis showed that the number of pods per plant,100-seed weight,oil content,the number of nodes on main stem,and the number of seeds per plant had significant effect on yield;oil content,the number of pods per plant,the effective branch number,plant height,and flowering time had significant effect on protein content.Protein content,100-seed weight,flowering time,flowering to maturity time,the number of pods per plant,and yield had significant effect on oil content.Based on piecewise linear regression analysis,combined with correlation analysis,and multiple regression equations,the optimum plant types for high-yield soybeans were established as follows:the number of nodes on main stem was increasing as much as possible.Five nodes height on the bottom was 8.00-14.70 cm.The number of pods per plant was more than 132.83.The effective branch number was about 6.The flowering time was about 57.16 days.The time from flowering to maturity was greater than 61.29 days,and the whole growth period was c

关 键 词：大豆 农艺性状 品质性状 主成分分析 分段线性回归 综合评价 

分 类 号：S565.1[农业科学—作物学]