陆地棉耐盐性状与SSR分子标记的关联分析

本研究以134份陆地棉栽培种为试验材料,测定其在0.3%盐浓度(质量分数)下的出苗率,并使用74对SSR引物对这些材料进行基因组变异扫描。利用Structure2.3.4软件分析该自然群体的遗传结构,在此基础上采用Tassel2.1软件对耐盐性状与SSR分子标记进行关联分析,寻找与棉花耐盐性状相关的分子标记。研究结果表明:(1)134份陆地棉栽培种的出苗率呈极显著差异,并筛选出27个盐敏感材料和10个耐盐材料。(2)74个SSR分子标记共检测出148个多态性位点,涉及246个等位变异,变异范围为2~7,平均每个标记3.32个;基因多样性指数变异范围为0.0295~0.4959,平均值为0.2897;SSR分子标记多态性信息含量(PIC)变幅为0.0290~0.3729,平均值为0.2381。(3)通过群体结构分析,将该自然群体划分2个亚群体,分别包含89份和45份材料。(4)关联分析共发现8个与棉花耐盐性状相关的SSR分子标记位点,表型变异解释率变幅为2.91%~7.82%,平均值为4.32%。此研究结果可以为棉花耐盐性状分子标记辅助选择育种提供参考。     

陆地棉SSR标记遗传多样性及其与农艺性状的关联分析

分析陆地棉栽培种遗传多样性,通过关联分析寻找与棉花农艺性状相关联的分子标记,为分子标记辅助选择育种和提高棉花育种效率奠定基础。本文采用74个Simple sequence repeat(SSR)标记对172份陆地棉栽培种的基因组变异进行扫描,使用NTSYS-pc 2.20进行聚类,分析该群体遗传多样性;利用Structure 2.3.4软件分析群体结构,在此基础上结合田间表型数据,采用Tassel 2.1的一般线性模型(General linear model,GLM)进行关联分析,定位与农艺性状相关的QTLs。74个标记共检测到148个多态性位点,涉及246个等位变异,变异范围2~7个,平均等位变异数为3.32;引物的多态性信息含量(PIC)为0.0281~0.3733,平均值为0.2370;遗传相似系数变异在0.2816~1,平均值为0.5369,平均遗传相似系数为0.5369,表明我国陆地棉遗传基础狭窄,尽管国外及西北内陆棉区部分材料具有较丰富的遗传变异。聚类分析将该群体划分为12个亚群,不同棉区的材料交叉分布,且聚类结果基本与系谱吻合。群体结构分析却将172份供试材料划分为3个亚群;通过关联分析,发现30个位点与铃重、衣分、黄萎病抗性显著相关(P0.05),各位点对表型变异贡献率为2.24%~5.27%。     

SSR标记与陆地棉田间黄萎病抗性的关联分析

通过关联分析挖掘与陆地棉黄萎病抗性关联的分子标记,为陆地棉抗黄萎病性状的分子检测及标记辅助选择育种提供有益参考。选用在棉花基因组上均匀分布多态性较好的237个SSR标记对214份陆地棉材料的基因组变异进行了扫描,并使用Power Marker v3. 25分析群体的遗传多样性,Structure 2. 2分析群体结构,在此基础上结合3年黄萎病抗性鉴定数据,采用Tassel 2. 1中的GLM(Q)方法挖掘与黄萎病抗性相关的QTLs。结果表明,不同陆地棉材料黄萎病发病情况差异显著; 237个标记共检测到280个多态性位点,共计695个等位变异,变异范围2～6个,平均等位变异数2. 479 0;按照基因型数据可将该群体划分为2个亚群;通过关联分析,在不同年份中发掘与黄萎病抗性显著关联(P 0. 01)的SSR位点27个,其中有2个位点(BNL3442和BNL1064)在3年均被重复检测到,表型变异解释率最高分别为12. 10%和8. 02%。这些在不同年份中稳定存在的SSR标记位点有可能与抗病基因紧密连锁,有望用于棉花黄萎病抗性材料筛选和抗病基因挖掘。     

三个栽培种中天然彩色棉的遗传多样性及关联分析

本研究利用主要农艺性状和SSR标记分析了137份陆地棉、13份海岛棉以及27份亚洲棉等不同纤维色泽的种质资源。170对SSR引物共检测到1036个等位基因变异。其中有多态性的等位基因923个,平均每个SSR位点有5.43个等位变异,变化范围为1~10。位点多态性信息量(PIC)的变化范围为0.56~0.93,PIC在0.85以上的多态性SSR位点为122个。结果表明,实验材料存在着丰富的遗传多样性。在群体结构分析基础上,使用TASSEL软件中的混合线性模型(MLM),在陆地棉中得到与11个性状相关联的SSR标记23个;在海岛棉群体得到与5个性状关联的17个标记;在亚洲棉中检测到与8个性状关联的15个标记位点。在三个栽培种中与纤维色泽度相关的标记位点一共有6个,分别是CIR51-4、BNL1421-2、BNL2656-2、DPL570-2、GH268-6和TMB131-1,表型变异解释率从3.12%到18.97%。三个栽培种中天然彩色棉种质具有丰富的遗传多样性,为棉花的种间、种内杂交育种以及拓宽棉花新品种的遗传背景提供了物质基础和理论依据。     

陆地棉产量及农艺性状的SSR标记关联分析

【目的】挖掘与陆地棉产量和重要农艺性状密切关联的分子标记,为棉花分子标记辅助育种提供依据。【方法】本研究以147份陆地棉材料为供试群体,分别调查了7个环境下该群体的铃重和衣分数据,3个环境下的单株果枝数、单株铃数和株高数据。利用237对SSR标记对上述材料进行基因型检测,采用Tassel 2.1中的一般线性模型程序对上述5个性状进行关联分析。【结果】铃重、衣分、单株果枝数、单株铃数和株高的平均变异系数的变幅为7.28%～21.06%,平均值为13.47%。这表明供试群体具有较为丰富的表型多样性。237个标记在147份陆地棉中共检测到281个多态性位点,各位点上等位基因数的范围为2～6个,涉及690个等位基因,平均每个位点2.455 0个;多态性信息含量平均值为0.216 5;Structure2.2群体结构分析将147份供试材料划分为7个亚群;在3个及以上环境检测到的与铃重和衣分显著关联的标记位点分别有45个和1个,解释表型变异率在不同环境中最高分别达19.30%和11.58%(P0.01);在2个及以上环境检测到与单株果枝数、单株铃数和株高显著关联的标记位点分别有1、4和4个,表型变异解释率在不同环境中分别最高达9.96%、7.00%和5.75%,在BNL2448附近同时检测到与铃重、单株果枝数和株高显著关联的位点。【结论】通过关联分析挖掘了多个可重复检测到的与陆地棉产量及重要农艺性状关联的分子标记位点,此研究结果可为棉花高产育种分子辅助选择提供有益参考。     

不结球白菜形态性状及SSR遗传多样性关联分析

分析不结球白菜遗传多样性,筛选与形态性状相关联且具有多态性的标记位点,为不结球白菜的分子辅助育种选择、种质资源的利用和新品种选育提供依据。本研究以类型差异大、有南方特色的54份种质资源为材料,对其6个质量性状和5个数量性状进行3次田间调查,从40对引物中筛选出25对SSR引物进行遗传多样性分析,结合形态性状对54份不结球白菜种质资源进行统计分析。结果表明:6个质量性状间具有丰富的多样性,5个数量性状间有显著的差异性,其中SA变异系数最大,FC与L、W和SA呈显著正相关;25对SSR引物共检测到72个多态性位点,平均每对SSR引物产生2.88个多态性位点,PIC平均值为0.3831,说明不结球白菜SSR具有丰富的遗传多样性;54份不结球白菜的遗传相似系数变化范围为0.492~0.963,平均为0.742,在遗传距离0.6处可将供试材料划分为3个类群;通过关联分析发现5个与数量性状关联的标记位点,各标记的表型变异解释率在0.4657~0.6329,平均为0.5393。本试验获得的5个标记位点可为不结球白菜分子标记育种提供依据。     

陆地棉抗旱性与SSR分子标记的关联分析

【目的】了解陆地棉抗旱性和寻找与陆地棉抗旱性相关的分子标记,挖掘与抗旱性状相关的优异等位变异。【方法】以191份陆地棉栽培种组成的自然群体为材料,在萌发期采用质量分数为15%的PEG6000竖直滤纸法对191份材料进行胁迫处理,测定其发芽率。根据隶属函数值对萌发期抗旱性划分抗旱级别。利用关联分析的方法以74对SSR引物对该自然群体进行基因组变异扫描。在利用Structure 2.3.4分析该自然群体的遗传结构的基础上,使用Tassel 2.0的GLM模型将基因型数据与萌发期抗旱性状进行关联分析,寻找与抗旱性状相关的分子标记。【结果】筛选出强抗旱材料46份,中等抗旱材料59份,干旱敏感材料86份;群体结构分析将191份材料划分为2个亚群,分别包含123份和68份材料;通过关联分析获得与陆地棉萌发期抗旱性显著相关的分子标记15个,各位点对表型变异解释率为2.2%~7.49%。【结论】本研究筛选出强抗旱材料46份,获得与萌发期抗旱性相关的分子标记15个,可供棉花抗旱标记辅助选择育种利用。     

新疆陆地棉经济性状优异等位基因位点的遗传解析

【目的】基于SSR(Simple sequence repeat,简单序列)标记展开对陆地棉经济性状的关联分析,挖掘优异等位变异位点,解析新疆陆地棉经济性状的遗传基础,为新疆陆地棉的遗传机理研究和高效的陆地棉分子设计育种提供理论依据。【方法】利用筛选出覆盖棉花全基因组的73对SSR标记对新疆156份陆地棉品种进行多态性扫描;采用R语言绘制boxplot图,利用TASSEL软件进行关联分析,挖掘与产量、纤维品质性状相关联的优异等位变异位点。【结果】通过对6个环境下新疆陆地棉品种的产量、品质相关性状进行关联分析,获得与产量性状相关的等位变异位点10个,表型变异解释率为6.69%～9.88%,平均值为8.43%;与纤维品质性状相关的等位变异位点23个,表型变异解释率为3.73%～13.22%,平均值为7.52%。其中22个QTL(Quantitative trait locus)已被报道,有10个QTL的关联性状与前人研究一致。【结论】新疆陆地棉品种的群体遗传结构简单,连锁不平衡水平低,6个环境条件下表型性状变化趋势较稳定。基于SSR的关联分析,发掘了与产量和纤维品质相关的优异等位变异基因及聚合优异等位基因位点的典型材料。     

普通菜豆根系相关性状的关联分析

幼苗期根系发育对作物的生长发育具有重要作用。利用生长袋纸培系统对324份普通菜豆种质的主根长、根干重、根体积、根表面积等9个根系相关性状进行表型鉴定,并结合覆盖全基因组、有多态性的116对SSR标记,利用MLM（Q+K）模型进行表型和标记的关联分析。表型分析表明,324份材料的9个根系相关性状表型变异丰富,平均变异系数的变动范围是10.09%~37.03%;基因型分析表明,116个多态性SSR标记共检测到919个等位变异位点,每个标记的平均基因多样性指数为0.59,多态性信息含量（PIC）平均值为0.54,显示这些标记具有较高的基因多样性;群体结构分析表明,供试材料分为两个亚群,与普通菜豆起源于两个基因库对应;关联分析结果显示,以P<0.01作为显著条件,共检测到48个显著标记位点,其中有10个位点同时与2个以上性状相关联,有5个位点与前人研究结果一致。研究结果为进一步理解普通菜豆根系的遗传机理提供了理论参考,也为分子标记辅助选择改良普通菜豆根系奠定了基础。     

利用陆地棉MAGIC群体定位产量、生育期和株高性状的QTL

棉花的产量、生育期和株高是重要的农艺性状, 决定着棉花的经济效益和生产方式。为了研究这些性状的遗传基础, 利用8个亲本构建的包含960个株系的陆地棉MAGIC (multi-parent advanced generation inter-cross)群体和SSR标记对这些性状进行关联分析。对产量、生育期和株高共8个相关性状进行了3年3个地点共5个环境的田间试验。经过最佳线性无偏预测(BLUP)综合多环境的表型数据, 发现MAGIC群体比亲本有更丰富的表型变异。8个性状的广义遗传力(H ²)变化范围为0.17~0.71。结合284个SSR标记基因型数据, 利用混合线性模型对产量、生育期和株高相关性状进行关联分析, 分别检测到51、27和9个显著关联的位点, 这些位点都表现出微效性, 表明该陆地棉MAGIC群体在性状位点的挖掘方面具有高效性。检测到20个标记位点或区间控制多个性状, 还发现单株有效铃数、单铃皮棉重、第一果枝节位和株高存在染色体热点区域, 对多性状综合研究或单性状深入挖掘具有重要价值。本研究为后续深入利用MAGIC群体进行遗传研究提供参考, 一些表型优良的材料和关联到的位点为育种改良奠定了基础。     

Quantitative trait locus mapping of drought and salt tolerance in an introgressed recombinant inbred line population of Upland cotton under the greenhouse and field conditions

Drought and salt tolerances are complex traits and controlled by multiple genes, environmental factors and their interactions. Drought and salt stresses can result in more than 50% yield loss in Upland cotton (Gossypium hirsutum L.). G. barbadense L. (the source of Pima cotton) carries desirable traits such as tolerance to abiotic and biotic stress along with high fiber quality. However, few studies have been reported on mapping quantitative trait loci (QTL) for abiotic stress tolerance using a permanent bi-parental population in multiple tests. The transfer of drought and salt tolerance from Pima to Upland cotton has been a challenge due to interspecific hybrid breakdown. This issue may be overcome by using introgression lines with genes transferred from Pima to Upland cotton. In this study, four replicated tests were conducted in the greenhouse each for drought and salt tolerance along with another test conducted in a field for drought tolerance using an Upland recombinant inbred line population of TM-1/NM24016 that has a stable introgression from Pima cotton. The objectives of the study were to investigate the genetic basis of drought and salt tolerance and to identify genetic markers associated with the abiotic stress tolerance. A total of 1004 polymorphic DNA marker loci including RGA-AFLP, SSR and GBS-SNP markers were used to construct a genetic map spanning 2221.28 cM. This population together with its two parents was evaluated for morphological, physiological, yield and fiber quality traits. The results showed that drought under greenhouse and field conditions and salt stress in the greenhouse reduced cotton plant growth at the seedling stage, and decreased lint yield and fiber quality traits in the field. A total of 165 QTL for salt and drought tolerance were detected on most of the cotton chromosomes, each explaining 5.98–21.43% of the phenotypic variation. Among these, common QTL for salt and drought tolerance were detected under both the greenhouse and field conditions. This study represents the first study to report consistent abiotic stress tolerance QTL from multiple tests in the greenhouse and the field that will be useful to understand the genetic basis of drought and salt tolerance and to breeding for abiotic stress tolerance using molecular marker-assisted selection in cotton.     

Genetic Diversity and Association Analysis of Fiber Quality Traits with SSR Markers in Germplasm Resources of Early Maturity Upland Cotton in Xinjiang

[Objective] We explored the genetic diversity of early maturity upland cotton germplasm resources in Xinjiang, China. We discovered specific germplasm resources and elite allelic variation related to fiber quality. [Method] We used 219 early maturity upland cotton accessions in our experiments and investigated 15 agronomic traits in three environments. We used a total of 128 pairs of simple sequence repeat (SSR) primers to scan for polymorphism. We then used NTsys-pc2.1 software to analyze genetic diversity, and Structure2.3.1 and Tassle5.0 software to perform association analysis of fiber quality traits based on phenotypic effect values to identify elite allele variation and conventional materials. [Result] A total of 244 loci were amplified by 128 marker pairs in 219 samples with an average of 1.91 loci per marker. The polymorphic information content ranged from 0.13 to 0.86 with an average of 0.63. The distribution range of the genetic similarity coefficient between materials in this population was 0.42-0.99 with an average of 0.61. The genetic similarity coefficient was between 0.5 and 0.7 and accounted for 90.19%. Through association analysis, we detected 11 markers that were significantly (P<0.01) associated with fiber quality traits, and discovered seven typical materials with specific allele. [Conclusion] In total, 219 Xinjiang early-maturing upland cotton germplasm resources have low genetic diversity. Based on SSR association analysis, we discovered some specific allele variations and conventional materials related to fiber quality traits.     

Association Analysis of Agronomic and Yield-Related Traits in an Upland Cotton (Gossypium hirsutum L.) Germplasm Population Based on SSR Markers

[Objective] Molecular markers associated with agronomic and yield-related traits could provide useful references for improving the efficiency of Gossypium hirsutum breeding. [Method] In 147 upland cotton materials, 237 Simple Sequence Repeat(SSR) markers were assayed for polymorphisms. The population structure was analyzed using Structure 2.2. An association analysis between SSR markers and the agronomic and yield-related traits was performed using the Tassel 2.1 General Linear Model program. [Result] The average coefficients of variation for the five indicators varied from 7.28% to 21.06%, with an average of 13.47%, which indicated that the test population had a rich phenotypic diversity. The 237 SSR markers included 690 alleles belonging to 281 loci, with an average of 2.455 0 alleles per SSR marker (ranging from 2 to 6), and the average polymorphism information content value was 0.216 5. The test group could be divided into seven subgroups using Structure software. A genome-wide association analysis was conducted on five agronomic and yield-related traits from 2013 to 2015. There were 45 SSR loci and 1 locus significantly (P＜0.01) associated with boll weight and lint percentage, which were detected in 3 or more environments, and they explained 19.30% and 11.58%, respectively, of the phenotypic variation in the different environments (P＜0.01). Nine loci were significantly (P＜0.01) associated with agronomic traits at two or more bowing times, and one, four and four markers were associated with fruit branch number, boll number and plant height, explaining 9.96%, 7.00% and 5.57%, respectively, of the phenotypic variation in different environments. Sites significantly associated with boll weight, fruit branch number and plant height were simultaneously detected near BNL2448. [Conclusion] Many loci associated with the yield and other important agronomic traits of upland cotton were repeatedly detected in different environments using an association analysis. The results provide useful reference information for cotton molecular marker-assisted selection-based breeding.     

Mining Elite Sea-Island Cotton Germplasm Based on Phenotyping and SSR Markers

[Objective] An elite germplasm resource of sea-island cotton with outstanding traits was mined in order to accelerate the breeding process of new varieties. [Method] The core collections of sea-island cotton germplasm consisted of 178 accessions were used as experimental materials in this study. Analyses of variability and diversity were performed through detecting phenotypic data of six main breeding-targeted traits, including boll weight, boll number per plant, lint percentage, fiber length, fiber strength, and micronaire. The elite germplasm of sea-island cotton was selected according to 10% optimal sampling strategy based on the phenotypic value of each trait. The 120 pairs of polymorphic simple sequence repeat (SSR) primers were used to analyze the polymorphism of 178 accessions of the core collections. Then, we conducted the population structure and clustering analysis based on the genotyping results. According to the results of cluster analysis, the primary elite germplasm was further selected, and the final elite germplasm of sea-island cotton was identified. [Result] The results showed that there was a high variability and abundant genetic diversity in the 6 studied traits. In 178 accessions of sea-island cotton, 262 alleles were detected by 120 pairs of SSR primers, with an average of 2.18 loci. The average polymorphism information content (PIC) was 0.067 8-0.630 0, with an average of 0.296 0, showing moderate polymorphism. The cluster analysis showed that the core collection of sea-island cotton was divided into six groups. twenty-three elite germplasm resources of sea-island cotton were identified based on phenotypic value and cluster analysis of SSR markers. [Conclusion] The germplasm of sea-island cotton can be analyzed and evaluated based on the phenotyping and SSR markers, and then the elite germplasm of sea-island cotton can be identified. These results provided the material basis for the genetic breeding of sea-island cotton, as well as the important reference and basis for the mining and identification of crop elite germplasm.     

基于重测序的陆地棉InDel标记开发与评价

碱基插入/缺失(InDel)是基因组中丰富的遗传变异形式。InDel以其密度高、易于基因型分型等优点成为分子标记开发的理想来源。本研究利用262份陆地棉品系重测序数据鉴定的InDel位点,在全基因组范围内设计了3206个InDel标记并挑选均匀分布的320个标记进行验证。320个标记筛选出87个多态性标记,多态性率为26.88%。利用多态性标记对不同地理来源的262份陆地棉种质资源进行基因分型,共检测到160个等位位点;多态性信息含量(PIC)为0.0836～0.3750,平均值为0.3073;基因多样性指数变异范围为0.0874～0.5000,平均值为0.3876,表明我国陆地棉遗传基础相对狭窄。群体结构分析将262份陆地棉品系大致划分为2个亚群,聚类分析和主成分分析的结果与之基本一致。采用混合线性模型(Mixed linear model)对6个纤维品质性状的关联分析检测到65个关联位点(P 0.01),各位点对表型变异贡献率为2.57%～8.12%。本研究旨在利用重测序数据开发全基因组范围的可用于凝胶检测的InDel标记,为棉花种质资源研究和分子标记辅助选择育种提供便捷工具。     

Genetic Dissection of Allelic Loci Associated with Economic Traits of Upland Cottons in Xinjiang

[Objective] We completed anassociation analysis of economic traits for upland cotton using simple sequence repeat (SSR) markers. We then explored the allelic variation sites to analyze the genetic basis of economically important traits, studied the genetic mechanism of Xinjiang upland cotton, and aimed to accelerate efficient breeding of upland cotton. [Method] We carried out polymorphic scanning on 156 upland cotton varieties in Xinjiang by screening 73 pairs of SSR markers encompassing the whole cotton genome. We constructed boxplot maps using R statistical computing software and graphics language and used TASSEL software to correlate yield or fiber quality traits with significant allelic variation loci. [Results] We obtained 10 allelic variation loci related to yield traits using the correlation analysis of Xinjiang upland cotton varieties from six different environments. The interpretation rate of phenotypic variation ranged from 6.69% to 9.88% with an average of 8.43%. Twenty-three allelic variation loci associated with fiber quality traits and phenotypic variation interpretation rates ranged from 3.73% to 13.22% with an average of 7.52%. The 22 detected quantitative trait loci were reported in previous studies and 10 showed the same associated traits as previously reported. [Conclusion] The population genetic structure of Xinjiang upland cotton varieties is simple, the linkage disequilibrium level is low, and the phenotypic traits show a stable trend under six environments. Using association analysis, we discovered unique allelic variation genes related to yield and fiber quality and diverse allele loci.     

Association analysis of fibre traits in Gossypium arboreum accessions

Advances in the use of diploid Asiatic species in cotton breeding require an understanding of the relatedness and ancestry of diploid cotton accessions, and identification of simple sequence repeat (SSR) markers associated with agronomically important phenotypic traits, for example, fibre quality. Fifty‐six Gossypium arboreum germplasm accessions introduced from nine regions of Africa, Asia and Europe were evaluated for eight fibre characters (lint percentage, lint colour, elongation, micronaire, strength, 50% span length, 2.5% span length and maturity%) and genotyped with 98 SSR markers. When viewed across all accessions most of the SSR markers were polymorphic. Population structure analysis identified six main clusters for the accessions which corresponded to different geographic regions, indicating agreement between genetic and predefined populations. The general linear model method was used to disclose marker–trait associations. Marker–trait associations were investigated by fitting single marker regression models for phenotypic traits on marker band intensities with correction for population structure. This paper illustrates the potential of association mapping in diploid cotton, because existing phenotypic data, a modest number of SSR markers, and a pioneering statistical analysis, identified interesting associations.     

Mapping the QTLs underlying drought stress at developmental stage of sorghum (Sorghum bicolor (L.) Moench) by association analysis

This study was conducted to identify quantitative trait loci (QTLs) for drought tolerance in sorghum (Sorghum bicolor (L.) Moench) by association mapping using a simple sequence repeat (SSR)-marker-based diversity research set. Genotypic data for 98 SSR marker loci on ten chromosomes were used for the association analysis. The experiment was conducted under control (well-watered) and drought stress conditions, and the phenotypic values of 23 morphological traits were recorded. Drought tolerance was assessed by using a leaf drying score as a parameter of the tolerance/susceptibility: scores were assigned on a scale from 1 (most tolerant) to 9 (most susceptible). Under the control conditions, 17 QTLs associated with 12 traits were identified on chromosomes 1, 2, 4, 8, 9, and 10, with ?Log₁₀ (P) ranging from 2.5 to 7.6 and explaining 9.5–57.5 % of the total phenotypic variance for the traits. Under the drought stress conditions, nine QTLs associated with 8 traits were identified on chromosomes 1, 2, 3, and 10 that explained 9–61.2 % of the total phenotypic variance for the traits, with ?Log₁₀ (P) ranging from 2.5 to 3.5. QTLs for some traits were detected only under the drought stress condition, suggesting that these traits are important in drought tolerance. These QTLs could be used to further dissect the genetic and physiological basis of drought tolerance in sorghum.