【作者】 黄中文；

【导师】 盖钧镒；

【作者基本信息】 南京农业大学 ， 作物遗传育种， 2008， 博士

【摘要】 产量是大豆育种最重要的性状。大豆产量是一个综合性状,其遗传基础复杂。研究产量密切相关的性状或产量因素,是揭示产量性状遗传基础的有效途径。大豆产量主要影响因素可归为3类:（1）光合生理有关性状:生物量累积与分配、叶部性状等;（2）产量因子:收获指数、百粒重、荚数、节数以及分枝性状等;（3）倒伏性。前2者是产量决定因素,后者是高产限制因素。目前对这3类产量主要影响因素的遗传研究还比较薄弱。因此,本文以科丰1号和南农1138-2组建的大豆重组自交家系NJRIKY为材料,在遗传背景相对控制下,通过2年田间试验的表型评价,研究高产家系生物量累积与分配动态特征;产量与这3类性状之间的关系,并定位相关性状的QTL。与美国Florida大学合作发展出2套新的遗传统计模型,分别用于大豆株高生长QTL与环境互作、茎杆和全株生物量异速生长性状QTL的检测。本研究结果为分子标记改良大豆产量性状,较全面地揭示NJRIKY群体产量遗传基础,提供分子标记信息。1.通过测定出苗后25 d至鼓粒期生物量累积和分配数据,获得大豆高、中、低产家系生物量积累与分配的动态特征:（1）地下部、地上部生物量与产量具极显著的相关动态,随生长进程,相关系数逐渐增加,至鼓粒期（R5～R6期）相关系数达到最大;（2）大田条件2500～2800kg hm^-2以上的高产家系,在约19万株/ha的情况下,地下部、地上部生物量累积显著高于中、低产家系,其地下部累积量为300-330kghm^-2（R1）、500 kg hm^-2（R3）和650 kg hm^-2（R5）,最大值660-700kg hm^-2,地上部累积量为1500～1600 kg hm^-2（R1）、3100～3400 kg hm^-2（R3）和5500～6500 kg hm^-2（R5）,最高值7200～7800 kg hm^-2,最大累积值均出现在鼓粒期;（3）高产家系各器官生物量分配动态特征是,茎秆、叶柄占同时期全生物量的比例两年平均分别为29.4%（R1）,32.0%（R3）,30.8%（R5）、27.1%（R6）和10.5%（R1）、11.7%（R3）、10.6%（R5）8.2%（R6）,显著高于中、低产家系;叶、根比例分别为46.0（R1）、41.2%（R3）、34.1%（R5）、25.4%（R6）和7.1%（R1）、12.6%（R3）、9.7%（R5）、7.9%（R6）,显著低于中、低产家系。2.在NJRIKY群体目标性状较宽的变异范围内,进一步研究了R5期生物量、收获期生物量、收获指数、R1和R5期叶面积指数与产量,以及收获指数与生物量的回归关系。结果表明,R5期生物量与产量呈指数回归关系,R5期生物量累积在5500～6500kg hm^-2时,产量不再随生物量增加而增加,进一步验证了上述高产家系R5期生物量累积标准具有普适性;收获期生物量与产量具直线关系,试验范围内未发现生物量上限;R1、R5期叶面积指数与产量具直线关系,但同一产量水平对应的叶面积指数变幅较宽;表观收获指数与产量呈指数曲线相关,小于0.42时与产量具正变关系,大于0.42时与产量具负变关系。收获期生物量与表观收获指数呈负向的指数曲线相关,说明当生物量累积不再是高产品种的限制因素时,产量的进一步突破途径是提高收获指数。产量因子与产量的相关研究结果是,除有效分枝和分枝荚数外,产量与根重、冠层宽和冠层高、荚数、荚粒数、百粒重、主茎荚数,分枝荚数、主茎节数等均呈极显著相关;其中,根重、冠层宽和冠层高与产量的相关系数相对较高。3.NJRIKY群体产量及产量相关性状QTL的检测结果:（1）检测到9个产量QTL,贡献率6%～17%。其中qYDB1-1和qYDD1a-2是两年均能稳定表达的产量QTL,累积解释产量变异的29%;另外2个产量QTL,qYDD2-1和qYDD2-2,贡献率分别是13%和16%,也归为主效产量QTL;其余的5个产量QTL贡献率均在10%以下。说明产量性状存在主效QTL控制的同时,也受效应较小的QTL位点的影响,表现出主效基因和微效基因共同决定的遗传体系。（2）检测到R1、R3和R5期生物量累积有关的QTL分别有6、9和6个,其中C2连锁群的2个标记区间,GMKF059a～satt319和O连锁群GMKF082b～satt3313,同时检测到这3个时期的生物量QTL,且相应QTL贡献率均较大,是控制生物量累积的主效QTL。检测到10个收获期生物量QTL,其中,qBMB1-2、qBMC2-1和qBMO-1,是主效QTL。（3）两年共检测到10个表观收获指数QTL,贡献率6%～22%。其中qHIB1-2、qHIO-1和qHIO-2贡献率较大且具有年份稳定性,是控制表观收获指数的主效QTL。（4）发现9个根重QTL,贡献率在5.1%～21.1%之间。B1连锁群的qRTB1-1两年能稳定表达,贡献率在10%以上,是根重有关的主效QTL。（5）检测到5个R1期叶面积指数QTL,贡献率在6.4%～17.2%之间,2年未检测到稳定表达的QTL。R5期检测到6个叶面积指数QTL,贡献率7.3%～26.2%。位于B1连锁群的qLAIR3B1-1具有年份稳定性,且贡献率较大,是叶面积指数主效QTL;两年共检测到4个冠层宽有关的QTL,贡献率在6.3%～13.1%之间。冠层高有关的QTL两年共检测到12个,贡献率为5.2%～9.2%。其中的O连锁群标记区间satt262-satt173,是冠层高和宽共同的QTL位点。（6）百粒重、荚粒数和单株荚数分别被检出6、2和1个QTL,解释相应的表型变异在6.9～15.7%之间。主效QTL是qSWB1-1、qSNPPO-2和qPPB1。发现分枝荚数、主茎荚数、主茎节数和有效分枝QTL分别为5、3、8和3个,贡献率6.3%～13.7%。（7）比较产量及产量相关的性状QTL定位结果发现,（1）在9个产量QTL中,有5个QTL位点,同时发现有多个产量相关性状QTL,以产量主效QTL之一的qYDB1-1所在的B1连锁群上分布的各性状QTL最多。（2）产量及产量相关性状,多数QTL成簇分布在B1、C2和O连锁群。这一方面解释了这些性状间相关的遗传原因,也说明产量与其相关的性状既存在共同的遗传基础,也有各自自身遗传控制体系。将产量主效QTL和产量密切相关性状QTL位点结合起来,共同作为产量的遗传控制体系,对深入揭示产量的遗传控制基础将是有意义的。4.对倒伏性评价指标的遴选研究结果是:在设计的4种倒伏性评价指标中,根据指标与倒伏程度的相关程度、物理意义、测度简便性和环境稳定性等4个判断标准,遴选出鲜重力矩（PF）的综合表现最优,并对PF和倒伏程度进行QTL分析。在NJRIKY群体共检测到7个倒伏程度相关的QTL,2年未检测到相同的QTL;检测出7个倒伏势有关的QTL,分别解释表型变异的5%～12%,其中qPFC2-2,2年均能稳定表达,贡献值较大,是倒伏势主效QTL。5.在原有功能作图的框架下,扩展建立了2套遗传模型,分别用于检测株高生长QTL与环境互作效应和控制茎杆和全株生物量异速生长QTL的定位。结果,在C2和O连锁群发现2个控制株高生长的QTL。通过互作检验,发现这些QTL具有显著的基因型-环境互作效应,并估计了株高QTL在整个生长和发育阶段的时间序列效应,测验了不同环境下特定QTL基因型生长曲线的区别。利用异速生长模型,成功检测到4个控制茎杆和全株生物量异速生长的QTL。这2套模型,为破译任何生物性状的表达以适应生境和非生境中的遗传控制结构,研究复杂表型的遗传网络结构,以获得与性状的发育模式和进程的调节机制有关的重要信息,奠定了初步的遗产学研究思路和方法。更多还原

【Abstract】 Yield is the most important trait for soybean breeding.It is a valuable strategy to study yield related traits or yield components for revealing genetic basis of yield since yield is believed a composite trait by many researchers.There are many traits relating yield including morphological,photosynthesis,physiological traits and yield components et al, all which can be summed uo to three kinds of trats,i.e.,biomass accumulation and partition, leaf-related traits and lodging in terms of the reports published.The two former determine size of yield and the latter can restrict a high yield.It is clear only for relation of yield with maturity time and growth habit in QTL locations.There is no report for QTL mapping for biomass,harest index and leaf area index.Although study on QTL mapping for lodging of soybean has been conducted for many years,there is absent of an affective method for precisely evaluating lodging due to demand to corresponging field environment.Crop breeders need a method for detecting lodging pential of plant.This paper employed a RIL population from a cross of Kefeng 1 and Nannong 1138-2 and tested phenotypic values of traits across two years to study dynamic characteristics of biomass accumulation and partition and relationships and QTL mapping for yield,biomass accumulation,harvest index,leaf area index canopy related traits and yield components.We work together with departent of statistics,university of Florida,to develop two genetic models for detecting genotype environment interactions for growth curves of plant height and QTL regulating ontogenetic allometry of stem with whole biomass.The aim of this paper is to supply molecular marker information and to study genetic basis of yield for yield enhancement of soybean.Through testing data of biomass accumulation and partition from 25 d after emergence to seed filling time,dynamic characteristics showed that（1） yield was positively and significantly correlated with under and above ground biomass accumulation and their correlation increased in the process of growth with the peak correlation coefficients during R5（start of seed filling） to R6（medium of seed filling）.（2）The lines with yield at above 2500～2800 kg ha^-1 and stand plants at 190 000 ha^-1,had a significant higher biomass accumulation than medium yield and low yield lines.Biomass accumulation of the high yield lines was 300～330 kg ha^-1（R1）,500 kg ha^-1（R3）,650 kg ha^-1（R5） and 700 kg ha^-1 （the peak） for under ground part,and 1500～1600 kg ha^-1（R1）,3100-3400 kg ha^-1（R3）, 5500～6500 kg ha^-1（R5） and 7200～7800 kg ha^-1（the peak） for above ground part.Both of under and above ground biomass accumulation reached to peak at R5～R6.（3）Comprising with the medium and low yield lines,the high yield ones had significantly higher mean proportions of petiole and stem biomass to the whole biomass across the two years with 10.5%（at R1）、11.7%（at R3）、10.6%（at R5） and 8.2%（at R6） for petiole and 29.4%（R1）, 32.0%（R3）,30.8%（R5）和27.1%（R6）for stem,and significant lower proportions of leaf and root with 46.0%（R1）、41.2%（R3）、34.1%（R5） and 25.4%（R6） for leaf and 17.1%（R1）、12.6%（R3）、9.7%（R5）and 7.9%（R6） for root.Biomass in harvest time had the biggest correlation coefficient with yield followed by leaf area index,root,canopy width and canopy height and the lowest ones were yield components including pod number,seed weight,seed number per pod.The regression for traits obtained was as following.There appeared some negative exponential correlation between yield and biomass at R5,with the biomass of 5 500～6500 kg ha^-1 as the highest turning point.A linear correlation of yield with biomass at harvest stage was detected,but without upper limit of the biomass found in the present experiment.Leaf area index had a linear relation with yield and there did not occur the upper limit for size of biomass.There was an exponential correlation between yield and apparent harvest index,with 0.42 as the turning point,positive relationship as less than 0.42 and negative relationship as larger than 0.42.Biomass in harvest negatively related with harvet index which shows that harvest index will become a restricting factor as biomass accumulation to some high size.Analizing results using method of Pearson simple correlation showed that yield was closely and significantly related with root weight,canopy width,canopy height,seed weight,pod number,seed number per pod,pod number per node,pod number in branch and node number in main stem excluding effective branch and pod number in branch.Result of QTL mapping in the NJRIKY population showed that,（1）Nine yield QTLs were detected in NJRIKY population distributing with explaining ratios from 6%to 17%. Of those yield QTLs there are two major ones,qYDB1-1 and qYDD1a-2,which could been detected across the two years and other two QTL,qYDD2-1 and qYDD2-2 were also major yield QTLs for their biger axplaning retios（13%and 16%）.The other five yield QTL appeared effects less than 10%.All the information showed the genetic basis of yield were composited of major genes and less efects’ genes.（2） 6,9 and 6 QTLs for biomass at R1, R3 and R5 were identified in the population,respectively,with 2 of them being detected across R1 through R5 in both years.9 QTLs for root weight were found in five lingage groups with R² from 5.1%to 21.1%.The major QTL for root weight was qRTB1-1.Ten biomass QTLs were found with three QTLs,qBMB1-2、qBMC2-1 and qBMO-1,as major QTLs for biomass.（3）Ten apparent harcest indexes QTL have been detected with explain ration from 2%to 22%.qHIB1-2、qHIO-1 and qHIO-2 were major QTLs for this trait.（4） 9 QTLs for root weight were dected and the major QTL was qRTB1-1 with a explaining effect more than 10%.（5） 5 and 6 QTLs for LAI at R1 and R5 were detected in the population with R² from 6.4 to 26.2%.qLAIR3B1-1 was the major QTL for LM.4 and 12 QTLs for canopy width and canopy height.The marker interval,satt262-satt173,was the same for the two traits.QTLs for seed weight,seed number per pod and pod number were 6, 2 and 1 respectively with R² from 6.9%to 15.7%and corresponding major QTL was qSWB1-1,qSNPPO-2 and qPPB1.5,3,8 and 3 QTL were found for pod number at branch, pod number at main stem,node number of main stem and effective branch.Most of QTLs for yield and yield related traits were distributed in B1,C2 and O linkage groups.Among the marker intervals of yield QTLs,were found to have QTLs conferring biomass and apparent harvest index,leaf area index which implied a partial same genetic basis among the three traits.We can integrate linkage groups of QTLs of yield with yield closely related traits as gentic component of yield,which will benefit for revealing genetic basis of yield.Results of studying on lodging resistance indices and related QTLs showed that fresh matter moment（PF） had the best representation among the four indices studied which denoted lodging potential.QTL analysis was performed also for lodging score and lodging potential.Seven QTLs for lodging score were found but no common one between the two years.There were seven QTLs for lodging potential with explaining 5%～12%phenotypic variation,with the same QTL qPFC2-2 in two years as major QTL.We have mapped dynamic traits with university of Florida together.Two genetic models were developed by extending composite functional mapping for estimating the effects of QTL-environment interactions on growth curves of plant height and allometry of stem with whole biomass respectively.Three QTL for plant height wer detected locating B1, C2 and O linkage groups,we have characterized the dynamic patterns of the genetic effects of the QTLs governing growth curves of plant height and estimated the global effects of the underlying QTLs during the course of growth and development,and test the differentiation in the shapes of QTL genotype-specific growth curves between different environments.Employed allometry gentic model,four QTLs controlling allometry of stem with whole biomass were found.The two models have successfully detected several QTLs that cause significant genotype-environment interactions for plant height growth processes. The model provides a basis for deciphering the genetic architecture of trait expression adjusted to different biotic and abiotic environments for any organism and will help to study the genetic architecture of complex phenotypes and,therefore,gain better insights into the mechanistic regulation for developmental pattems and processes in organisms.更多还原