高密度种植可以提高大豆产量，但通过改良株高及叶柄性状以选育株型紧凑、抗倒伏性优异的高产品种是提高产量的重要途径。2017-2018年，我们比较了黄淮地区四个地点的短叶柄种质M657与三个对照品种产量相关性状、抗倒伏性和叶柄相关表型间的关系。结果表明，M657对高种植密度和倒伏性表现出极高且稳定的耐受性，尤其在最高密度8×105株ha^-1下表现依然优异。回归分析表明，较短的叶柄长度与抗倒伏性的增加显著相关。产量分析表明，M657在较高密度下获得了较高的产量，尤其在黄淮北片地区。在与地点、密度相关的倒伏性和产量方面，不同品种对株距、行间距的反应存在显著差异。植株的倒伏性与种植密度、株高、叶柄长度和有效分枝数显著正相关，与茎粗、单株粒数、单株粒重呈显著负相关。在当前大豆品种种植密度的基础上，适当增加种植密度有助于黄淮地区大豆的产量提高。本研究为在高密度种植系统中引入适宜高产的紧凑株型性状、建立黄淮地区大豆高产模式提供了极有价值的新种质资源。

本研究建立了叶柄长度检测方法，并对EMS诱变冀黄13获得的高光效新种质M657为材料，于2017-2018年度在北方、黄淮、南方共7个地点进行表型鉴定。与冀黄13相比，M657在北方春、黄淮海夏及南方夏种植时矮化、叶柄短表型稳定，M657株高与叶柄长度显著降低，有效分枝数增加，生育期延长2-7 d，单株粒重、百粒重下降；4个短叶柄新品系的选育为大豆株型改良提供了重要的亲本种质，同时证明了利用矮杆短叶柄新种质M657理想株型为耐密、高产大豆新品种的培育的可行性。

本研究在前期已发掘短片段InDels和SNPs基础上，基于全基因组重测序分析在一个重组自交系群体的两个亲本中品03-5373（ZP）和中黄13（ZH）之间检测到不均匀分布在大豆20条染色体上的13573个长片段InDels，其中，Chr11上最少，有321个，Chr18上最多，有1246个。长片段InDels在染色体两臂的平均密度显著高于着丝粒区，与大豆基因组注释基因的分布模式一致。位于基因区的长片段InDels有2704个，占总数目的19.9%，其中319个为可导致蛋白质序列截短或延长的大效应InDels。重点对前期鉴定的株高相关QTL（qPH16）进行分析，共鉴定到35个长片段InDels，并将其开发成InDel标记，其中26个InDel标记（74.3%）在ZP和ZH之间表现出明显的多态性。利用开发的标记结合已有的4个SNPs标记对由ZP和ZH衍生的242个重组自交系进行基因型鉴定和QTL定位，将qPH16的定位区间从原来的960 kb缩小到477.55 kb，包含65个注释基因。在SNPs和短片段InDels开发基础上，进一步开发长片段InDels，可为大豆重要农艺性状的遗传分析和分子辅助选择育种提供更加全面的遗传变异信息。

大豆株高是由主效或微效基因控制的重要农艺性状。在已报道的株高QTL中，绝大部分定位区间较大，限制了大豆株高分子调控机制的解析。增加遗传图谱的标记密度会显著地提高QTL定位的效率和准确性。本研究利用双亲中黄13和中品03-5373及其衍生的241个重组自交系（RILs）全基因组重测序数据，构建一个包含4011个重组bin标记、总遗传距离为3139.15 cM的高密度遗传图谱，相邻bin标记间的平均距离为0.78 cM。比较基因组分析表明，所构建的遗传图谱与大豆参考基因组具有较高的共线性。基于此图谱，在6个环境中共检测到9个株高QTL，包括3个新位点（qPH-b_11，qPH-b_17和qPH-b_18）。其中，两个环境稳定主效QTL qPH-b_13和qPH-b_19-1可解释10.56%～32.7%的表型变异。qPH-b_13和qPH-b_19-1被精细定位到440.12 kb和237.06 kb的基因组区间，分别包含54和28个注释基因。进一步的拟南芥同源基因功能和候选基因表达分析表明，基因Glyma.13G292600和Glyma.19G194100分别为qPH-b_13和qPH-b_19-1的候选功能基因。

叶片是植物的主要光合器官，对作物品种的产量起着重要作用。鉴定导致叶片表型变异的致病突变和候选基因是大豆籽粒增产的重要育种目标。以大豆品种中品661为背景，研究了EMS诱导的具有异常皱叶表型的大豆突变体DWARFCRINKLEDLEAF1 (DCL1)。为了研究与皱叶性状相关的基因组位点，我们从Zp661和DCL1的杂交中构建了F2分离群体。采用整体分离分析(bulk separation analysis, BSA)结合全基因组重测序方法，通过欧氏距离(Euclidean distance, ED)关联算法检测出12个总长度为20.32 Mb的候选基因组区域与目标性状连锁。测序结果显示，Glyma.19G207100基因第1外显子存在一个单核苷酸突变(C:G>T: A)。基于该SNP衍生的CAPS标记对候选基因进行了验证，结果表明亲本之间存在核苷酸多态性。因此，我们的研究结果表明Glyma.19G207100(命名为GLYCINE MAX DWARF CRINKLED LEAF 1, GmDCL1)是一个可能参与大豆突变体DCL1皱叶性状形态发生的候选基因。本研究为该基因的功能验证提供了基础，并为大豆增产育种提供了前景。

为筛选大豆香味种质，建立大豆叶片中香味特征化合物2-乙酰基-1-吡咯啉（2-acetyl-l-Pyrroline，2AP）的鉴定方法。本研究通过单因素及三因素四水平（L9 (34)的正交试验，以峰形、总峰面积及检测样品时间为考察指标，建立了利用气质联用仪（GC-MS）快速检测香味的方法，明确了仪器运行最佳参数包括：柱温70℃，进样口温度180℃，以及样品最优萃取时间条件（酒精含量1ml、NaCl含量0.1g，超声时间10min，萃取时间为1h）。该检测方法重复性好、简单快速、样本试剂消耗少，可精准快速测定2AP含量。利用该方法对不同地理来源的101个大豆基因型进行了分析筛选。结果表明， 2-AP平均含量为0.29ppm，变幅为0.094ppm到1.816ppm，遗传多样性指数为0.54。可被划分为3个等级，其中，1级香型大豆有7份，包括中龙608、黑农88、哈13-2958、红面豆、黑农82、黄毛豆、吉育21。本研究建立的方法及筛选的优异种质为大豆香味育种和基因发掘提供了技术和材料支撑。

The cultivated soybean (Glycine max (L.) Merr.) was distinguished from its wild progenitor Glycine soja Sieb. & Zucc. in growth period structure, by a shorter vegetative phase (V), a prolonged reproductive phase (R) and hence a larger R/V ratio. However, the genetic basis of the domestication of soybean from wild materials is unclear. Here, a panel of 123 cultivated and 97 wild accessions were genotyped using a set of 24 presence/absence variants (PAVs) while at the same time the materials were phenotyped with respect to flowering and maturity times at two trial sites located at very different latitudes. The major result of this study showed that variation at PAVs is informative for assessing patterns of genetic diversity in Glycine spp. The genotyping was largely consistent with the taxonomic status, although a few accessions were intermediate between the two major clades identified. Allelic diversity was much higher in the wild germplasm than in the cultivated materials. A significant domestication signal was detected at 11 of the PAVs at 0.01 level. In particular, this study has provided information for revealing the genetic basis of photoperiodism which was a prominent feature for the domestication of soybean. A significant marker-trait association with R/V ratio was detected at 14 of the PAVs, but stripping out population structure reduced this to three. These results will provide markers information for further finding of R/V related genes that can help to understand the domestication process and introgress novel genes in wild soybean to broaden the genetic base of modern soybean cultivars.

    The growth periods (GPs, from planting/emergence to reproductive stage 8 (R8) of soybean cultivars vary in different ecological regions, especially in China with a very complex soybean cropping system. In this study, a 3-yr experimental study was undertaken in three geographical locations of China from 2008 to 2010, including the Northeast (40.66–45.85°N), Huang-Huai (34.75–38.04°N) and southern (22.82–30.60°N) eco-regions with about 250 accessions in each region to clarify the classification of maturity group (MG) and identify the cultivars with stable GP to increase the knowledge about the GP distribution of soybean cultivars in China. GPs of soybean cultivars in different eco-regions were significant different with a gradual decrease from 115–125 d in the Northeast part to the 85–100 d in the southern part of China. The geographical location was the major factor for GP of cultivars from the Northeast, while the year of planting was the major factor affecting the stability of GPs in Huang-Huai summer and southern summer soybean. AMMI2 (additive main effects and multiplicative interaction)-Biplot analysis showed that the GPs of soybean cultivars from the Northeast eco-region have a comparatively satisfactory environmental stability. Moreover, soybean cultivars with moderate GP/MG and stable environment adaptability in different eco-regions were identified based on the linear regression and AMMI analysis, which was important for the accurate classification of soybean MGs in future. Taken together, our results reflected the genetic diversity, geographical distribution and environmental stability of the Chinese soybean GP trait. Soybean cultivars with stable GP for various Chinese eco-regions would be beneficial for Chinese soybean genetic improvement, varietal introduction, exchange, and soybean breeding program for wide adaptability.

Soybean is a widely planted genetically modified crop around the world. However, it is still one of the most recalcitrant crops for genetic transformation due to the difficulty of regeneration via organogenesis and some factors that affect the transformation efficiency. The percentages of resistant bud formation and transient expression efficiency are important indexes reflecting the regeneration and transformation efficiency of soybean. In this study, the percentages of resistant bud formation and transient expression of β-glucuronidase (GUS) were compared after treatment with sonication or surfactant and co-treatment with both. The results showed that treatment with either sonication or surfactant increased the percentage of resistant bud formation and transient expression efficiency. The highest percentages were acquired and significantly improved when cotyledon node explants were co-treated with sonication for 2 s and surfactant at 0.02% (v:v) using two different soybean genotypes, Jack and Zhonghuang 10. The improved transformation efficiency of this combination was also evaluated by development of herbicide-tolerant soybeans with transformation efficiency at 2.5–5.7% for different genotypes, which was significantly higher than traditional cotyledonary node method in this study. These results suggested that co-treatment with surfactant and sonication significantly improved the percentages of resistance bud formation, transient expression efficiency and stable transformation efficiency in soybean.

Soybean seed storage protein is one of the most important plant vegetable proteins, and β subunit is of great significance to enhance soybean protein quality and processing property. F2 segregated population and residual heterozygous lines (RHL) derived from the cross between Yangyandou (low level of β subunit) and Zhonghuang 13 (normal level of β subunit) were used for mapping of β subunit content. Our results showed that β subunit content was controlled by a single dominant locus, qBSC-1 (β subunit content), which was mapped to a region of 11.9 cM on chromosome 20 in F2 population of 85 individuals. This region was narrowed down to 2.5 cM between BARCSOYSSR_20_0997 and BARCSOYSSR_20_0910 in RHL with a larger population size of 246 individuals. There were 48 predicted genes within qBSC-1 region based on the reference genome (Glyma 1.0, Williams 82), including the two copies of β subunit coding gene CG4. An InDel marker developed from a thymine (TT) insertion in one copy of CG4 promoter region in Yangyandou cosegregrated with BARCSOYSSR_20_0975 within qBSC-1 region, suggesting that this InDel marker maybe useful for marker-assisted selection (MAS).