【作者】 张晗；

【导师】 田纪春；

【作者基本信息】 山东农业大学 ， 作物遗传育种， 2014， 博士

【摘要】 小麦（Triticum aestivum L.）是具有分蘖成穗特性的禾本科植物，其单位面积产量由单位面积穗数、每穗粒数和粒重（一般用千粒重表示），即产量“三因素”构成。在一定范围内每一个因素的增加，都可提高产量。本研究利用3个遗传群体对小麦的产量及其三要素进行非条件和条件QTL分析，从单个QTL/基因水平上阐明了小麦产量与其构成“三因素”呈正相关，而产量构成“三因素”之间一般呈负相关的遗传基础。鉴定到增加单位面积穗数而不显著减少穗粒数和千粒重、或增加穗粒数而不显著降低千粒重和单位面积穗数的有利QTL/基因，为结合分子标记辅助育种方法，聚合单位面积穗数、每穗粒数和粒重的有利等位基因，培育大穗、大粒和多粒的产量大幅度提高的小麦新品种具有十分重要的意义。获得如下主要研究结果：1．整合了含有182个家系的RIL群体（简称C群体）和含有256个家系的RIL群体（简称D群体）的遗传图谱，获得了一张包括802个位点的整合分子遗传图谱，其中DArT标记位点734个、SSR标记位点62个、TaGW2-CAPS（粒重基因）标记位点1个、Glu-A1和Glu-D1标记位点各1个和Wx-A1、Wx-B1和Wx-D1蛋白亚基标记位点各1个。该整合图谱含有31个连锁群，覆盖小麦基因组的21条染色体，共覆盖小麦基因组长度6034.1cM。单个染色体长度为3.5-425.3cM，标记间平均距离为7.52cM。2．利用QTL IciMapping v3.3软件，对C群体、D群体和含有168个家系的DH群体（简称DH群体）进行条件和非条件QTL分析，共检测到控制产量及产量三要素的92个加性非条件的QTL，其中有11个QTL在2个环境中均被重复检测到，5个QTL在3个环境中均被重复检测到，1个QTL在5个环境中均被检测到，这些QTL为环境间稳定表达的QTL。32个QTL的贡献率大于10%，其中QTKW-D-2D-3.2、QTKW-D-2D-2.1和QKNPS-DH-7B-2.1贡献率分别高达64.40%、31.91%和34.07%，这些为控制产量或产量构成因素的主效QTL。34个QTL位于1A、2B、2D、3A、3B、4A、4B、6A和6D染色体的同一区段，表现为“一因多效”。3．在3个群体中，当单位面积产量分别给定单位面积穗数、穗粒数和千粒重及单位面积穗数、穗粒数和千粒重相互给定条件下，共检测到181个条件QTL，3个群体中分别可解释4.59%-20.29%、5.14%-35.63%和3.62%-69.59%的表型变异，除了染色体3D和7D外，共覆盖了其他19条染色体。4．利用非条件QTL和条件QTL对比分析方法，首次在QTL水平上解析了产量与三要素及三要素之间的遗传关系。检测到17个条件QTL，其中QSN-DH-2B、QSN-DH-3A、QSN-DH-6D、QSN-D-1A-1.1和QSN-D-3B-2.1等5个QTL在提高单位面积穗数的同时，并不导致穗粒数的降低； QKNPS-DH-2B-2.1和QKNPS-D-5B提高穗粒数的同时，并不导致单位面积穗数的降低；QKNPS-DH-1A、QKNPS-DH-2D-1.1和QKNPS-DH-6A等3个QTL在提高穗粒数的同时，并不导致千粒重和单位面积穗数的降低；QTKW-DH-5B、QTKW-DH-7B和QTKW-C-4B-2.1等3个QTL在提高千粒重的同时，并不导致穗粒数的降低；QTKW-DH-4B和QTKW-D-2B-1.1等2个QTL在提高千粒重的同时，并不导致单位面积穗数和穗粒数的降低；QY-DH-2D-1.1和QY-DH-3A是提高产量的QTL。这些QTL是克服小麦单位面积穗数、每穗粒数和粒重之间负相关矛盾的重要的基因位点。5．在C群体和D群体中，共检测154个QTL，其中52个为非条件QTL，102个为条件QTL。利用BioMercator V3.0软件的QTL映射功能，对154个QTL进行了映射分析，其中有149个QTL映射到C和D群体的公共图谱上，但有33个QTL位于gap上，最后映射到公共图谱的QTL有116个。6．共检测到控制产量、单位面积穗数、穗粒数和千粒重的41对非条件的上位性QTL和63对条件的上位性QTL，其中有10对上位QTL解释的表型变异大于40%，尤其是位于wPt-8492-wPt-1454区间的QSN-C-2B与位于Xgwm190-wPt-6429区间的QSN-C-5D、位于wPt-5704-wPt-667891区间的QSN-C-3B与位于wPt-666615-wPt-666008区间的QSN-C-6D，位于wPt-5313-Xgpw7646区间的QSN-C-3D和位于wPt-666615-wPt-666008区间的QSN-C-6D分别解释高达59.07%、51.65%和50.77%的表型变异。加性QTKW-DH-3A-2.1与上位性QTKW-DH-3A位于Xcfa2170-Xbarc51同一个区间内。更多还原

【Abstract】 Common wheat (Triticum aestivum L.) is a Poaceae plant characteristic of spike-bearingtillers.Spike number per m2, kernel number per spike and kernel weight (generally representedby thousand-kernel weight) are the “three main components” determining the yield per m2.Increase in any one component can improve wheat yield. In this research, conditional andunconditional quantitative trait locus (QTL) analysis of yield and its three main componentswas conducted using3populations. The aim was to explore the genetic mechanismsunderlying the significant positive correlations between yield and its three main componentsand the negative correlations among the three main components at single QTL/gene level.Identification of the QTLs or genes that have positive effects on increasing the spike numberper m2without reducing significantly kernel number per spike and kernel weight, orincreasing the kernel number per spikewithout reducing significantly the spike number per m2and kernel weight will contribute to combining those beneficial QTL or genes for spikenumber per m2, kernel number per spike and kernel weight using molecular marker-assistedtechniques and developing high yielding wheat cultivars with larger spike and more andheavier grains.The main results are as follows:1. An integrative genetic map consisting of802loci was obtained by integrating twogenetic maps that had been constructed based on182families (“C” population) and256families (“D” population). The new map consisted of31linkage groups contained734DArTmarkers,62SSR markers, one TaGW2-CAPS marker, two HMW-GS markers and three Wxprotein subunit markers, and covered21chromosomes of the wheat genome and a total lengthof6034.1cM. Single chromosome length varied from3.5to425.3cM, with an averagebetween marker distance of7.52cM.2. A total of92unconditional additive QTL controlling yield and the three yieldcomponents were detected using IciMapping v3.3. Of these QTLs,11were detected in eachof two environments, five in each of three environments, one in each of five environments. Therefore, these QTL were expressed stably across environments.32QTLs had contributionrates of more than10%. Among which, QTKW-D-2D-3.2, QTKW-D-2D-2.1andQKNPS-DH-7B-2.1were main effect QTLs and had contribution ratge of64.40%,31.91%and34.07%, respectively.34QTLs were located on the same segments on1A,2B,2D,3A,3B,4A,4B,6A and6D chromosomes and showed pleiotropic effects.3.181conditional QTLs were detected in the three populations when yield per m2wasconditioned on spike number per m2, kernel number per spike and thousand-kernel weight andwhen spike number per m2, kernel number per spike and thousand-kernel weight wereconditioned on each other. These QTLs accounted for4.59%-20.29%,5.14%-35.63%and3.62%-69.59%of the phenotypic variations for the DH, C and D population, respectively andwere located on19chromosomes except3D and7D.4. The genetic relationships between yield and the three main yield components andamong the three main yield components were dissected using comparative analysis ofconditional and unconditional QTL for the first time.17QTLs were detected. Of these, fiveQTLs, i.e. QSN-DH-2B, QSN-DH-3A, QSN-DH-6D, QSN-D-1A-1.1and QSN-D-3B-2.1,could increase spike number per m2without reducing kernel number per spike. Two QTLs, i.e.QKNPS-DH-2B-2.1and QKNPS-D-a5B, could increase kernel number per spike withoutreducing spike number per m2. Three QTLs, i.e. QKNPS-D H-1A、QKNPS-DH-2D-1.1、QKNPS-DH-6A, could increase kernel number per spike without reducing spike number perm2and kernel weight. Three QTLs, i.e. QTKW-DH-5B、QTKW-DH-7B and QTKW-C-4B-2.1,could increase kernel weight without reducing kernel number per spike. Two QTLs, i.e.QTKW-DH-4B and QTKW-D-2B-1.1, could increase kernel weight without reducing spikenumber per m2and kernel number per spike. Two QTLs, i.e. QY-DH-2D-1.1and QY-DH-3A,could increase yield but had no effect on spike number per m2. The afore-mentioned QTLs aresome of the important loci that contribute to overcoming the contradictions among spikenumber per m2, kernel number per spike and kernel weight5. A total of154QTLs, including52unconditional and102conditional ones, weredetected in C and D population. Map analysis on the154QTLs was conducted using the QTLprojection function of BioMercateor V3.0. As a result,149of the154QTLs were projectedonto the common genetic map of the two populations. However,33of them were located on the gaps. Finally,116QTLs were projected onto the common map.6.41pairs of unconditional epistatic and63pairs of conditional epistatic QTLs for yieldand the three main yield component traits were detected in this study. Of these,10pairs ofepistatic QTLs accounted for more than40%of the phenotypic variation. Especially, forQSN-C-2B and QSN-C-5D (located in interval wPt-8492-wPt-1454and intervalXgwm190-wPt-6429, respectively), QSN-C-3B and QSN-C-6D (located in intervalwPt-5704-wPt-667891and interval wPt-666615-wPt-666008, respectively), QSN-C-3D andQSN-C-6D (located in interval wPt-5313-Xgpw7646and interval wPt-666615-wPt-666008,respectively), the explainable phenotypic variation was59.07%,51.65%and50.77%,respectively. The additive QTL QTKW-DH-3A-2.1and the epistatic QTL QTKW-DH-3A werelocated in same interval, i.e. Xcfa2170-Xbarc51.更多还原