【作者】 陈庆全；

【导师】 牟同敏； 余四斌；

【作者基本信息】 华中农业大学 ， 作物遗传育种， 2007， 博士

【摘要】 随着全球气候的快速变暖，极端高温出现的频率不断增加，高温热害已经成为世界稻作生产的制约因子之一。高温热害不仅导致水稻减产和稻米品质的降低，同时还将引起整个稻作制度的改变，因此水稻的耐热性研究变得日趋重要。水稻耐热性机制较为复杂，其耐热性遗传基础以及水稻耐热性育种进展缓慢。水稻许多重要农艺性状都是数量性状，水稻遗传改良的大部分工作就是改良这些数量性状，但由于其容易受环境的影响，遗传基础很复杂，所以水稻重要农艺性状的遗传基础研究及其遗传改良至今仍未取得突破性进展。因此，水稻重要农艺性状的遗传基础研究也一直倍受关注。本研究以2个优良的籼型恢复系材料T219和T226为亲本构建了202个株系的重组自交系(RILs)群体，利用人工气候室模拟自然条件下的高温环境(温度设置：30～37℃的日变温，33.5℃日平均温度)，进行抽穗开花时期的耐热处理，以耐热系数为水稻耐热性指标，对其抽穗开花期耐热性进行遗传分析。与此同时，在武汉和海南陵水两地不同的年份、不同的土壤类型、不同的气候条件、不同的田间管理条件以及在武汉的不同年份、不同的栽培季节和不同的栽培方式等多种不同环境下，分别考察了重组自交系的株高、抽穗期、单株产量、有效穗数、穗长、每穗颖花数、每穗实粒数、结实率、千粒重和穗着粒密度共10个重要农艺性状，剖析了在不同环境下各性状的遗传基础。主要研究结果如下：1．利用该RILs群体构建了遗传连锁图谱，该图谱包含189个SSR标记、14个连锁群，图谱总长1559.6cM，相邻标记平均遗传距离为8.3cM。标记顺序与已经发表的图谱有很好的一致性。2．利用人工气候室对RILs进行抽穗开花时期的耐热处理，以耐热系数为耐热性指标，利用混合线性模型复合区间作图法对水稻抽穗开花期耐热性进行了QTL定位分析，两年共定位7个耐热性QTL，分别位于第2、3、8、9和12染色体上，其LOD值为4.5～13.6，贡献率7.0-12.0％。其中位于第3染色体的RM570-RM148N段qHT3在两年同时检测到，且保持一致的效应方向。3．两年共发现有7对抽穗开花期耐热上位性QTL，分别位于第2、3、4、7、8和9染色体上，LOD值为5.9-7.4，贡献率较小，为0.4-1.4％。4．不同环境下的重要农艺性状相关性分析发现，除了千粒重与穗长、千粒重与有效穗数以及结实率与穗长之间的相关系数在5个环境下均未达到显著水平外，其余各性状间的相关系数均至少在1个环境中达到显著或极显著水平。其中有11对性状之间在所有5个环境下均表现一致的显著或极显著的正相关，5对性状之间在所有5个环境下均表现一致的显著或极显著的负相关。同时另有4对性状之间在不同的环境下出现正负完全相反的极显著相关性。5．AMMI模型进行的基因型和环境互作效应分析表明，RILs所有10个性状的基因型、环境以及基因型与环境互作效应均为极显著。其中每株有效穗数、结实率和单株产量存在较大的基因型与环境互作效应，而株高、抽穗期和千粒重的基因型与环境互作效应均较小。AMMI模型将10个性状的基因型与环境互作效应均分解为2个主成分，分别解释了它们在各性状的基因型与环境互作效应中的比例。其中以株高、抽穗期、穗长和每穗实粒数的AMMI1和AMMI2所能解释的基因型与环境互作效应为最高，分别为80.24％、80.09％、77.33％和70.72％。，而以每穗颖花数的为最小，为55.59％。6．采用区间作图法，在5个不同环境下对10个重要农艺性状(结实率为8个环境)进行了QTL定位分析，总共检测到129个QTL。不同性状在不同环境中所检测到的QTL数目和位置均有不同程度的差异。7．从定位的QTL数量来看，在不同的环境中所检测到的各性状的QTL总数不同。总数最多为45个，最少的为41个。在各个性状中，以穗长在所有环境中检测到的QTL总数目最多，共计为19个，最少的为千粒重，共7个QTL。8．从结果中发现，部分性状的某些QTL在不同环境下的表现较稳定，而且有的QTL的效应还较大。如结实率的qSR3-1、抽穗期的qHD12-1、千粒重的qKGW5-1等。同时某些性状如千粒重的QTL，在各不同环境间表现很小的差异。更多还原

【Abstract】 Study on heat tolerance in rice is getting more and more important, for high temperature stress has become one of the major factors exerting serious influence on rice production. And the genetic basis of heat tolerance in rice is very complex due to different mechanisms mixed together which resulted in slow progress on heat tolerance breeding in rice. Most of the important agronomic traits in rice are quantitatively inherited. Improvement of these quantitative traits is one of the major objectives in rice genetic breeding program. By now, progress on the reseach of the genetic basis and improvement of the major agronomic traits was not significant because of their complex mechanisms and environment effect. So researches on the genetic basis of the major agronomic traits in rice are always given much attention. In this study, we have developed a set of novel population with 202 recombinant inbred lines (RIL) derived from a cross between two indica varieties, T226 and T219, and then treated the RIL population under high temperature regimes in growth chamber during flowering stage, so as to assess the genetic effect of quantitative trait loci (QTLs) conferring heat tolerance at flowering stage in rice with coefficient of heat tolerance. And to dissect the genetic basis of the major agronomic traits in rice, genotype by environment (GxE) interaction and QTLs for ten agronomic traits under various environments including local, year, soil type and nurse manner were analyzed using the population with 202 RILs. The main results are summarized as following:1. A genetic linkage map consisting of 189 SSR markers, covering 1559.6 cM in whole genome, was constructed using the RIL population, with an average interval of 8.3 cM. 189 SSR markers were separated into fourteen groups. The linear order of markers in the linkage map was in good agreement with that published previously.2. Treated the RIL population under high temperature regimes in growth chamber during flowering stage and measured heat tolerance with the relative ratio of spikelet fertility of treated plants with high temperature in growth chamber and with optimal temperature in natural environment during flowering time, QTLs for heat tolerance at flowering stage were analysed by using a mixed linear-model. Seven main effect QTLs controlling heat tolerance during flowering time, located in chromosome 2, 3, 8, 9 and 12were detected in two-year experiments. A QTL, qHT3, located at RM570-RM148 on chromosome 3, was detected in two years, respectively.3. Epistatic interaction analysis showed seven pairs of epistatic QTL which involved 12 loci located in chromosome 2, 3,4, 7, 8 and 9 were detected.4. The correlation between traits was also analyzed. Most of the correlation coefficient between traits were significant at least at one environment. Eleven pairs of traits were with the same significant positive correlation and five pairs with the same significant negative correlation under all five environments. However there were four pairs with different direction of correlation at various environments. The direction of correlation between some traits was stable and some were changeable at different environment.5. GxE interaction was detected using AMMI statistical model. GxE interaction for all ten traits under five environments was significant. The AMMI model dissected the interaction component in two sects and explained the extent of interaction. Plant height (PH), heading date (HD) and 1000 grains weight (KGW) had the least GxE interaction, while Tillers per plant (TP), seed setting ratio(SR), spikelet density (SD) and yield per plant (YD) had higher GxE interaction. Of the total GxE interaction effect, the AMMI model explained 80.24% for PH, 80.09% for HD and 77.33% for panicle length (PL) and 70.72% for grains per panicle (GP). This model explained at least 55.59% of the total interaction effects for the traits observed.6. Interval mapping was used to detect QTL. 129 QTL for ten traits were detected under all various envirnments. The number and location of QTLs for defferent traits varied with various environments.7. Of QTL numbers, medium season at Wuhan in 2005 (E4) had detected forty five QTLs, the most in all various environments; Field environment at Lingshui in 2005 (E2) had detected forty one QTLs, the least in all various environments.. Of all traits, the most QTLs were for PL with nineteen, and the least QTLs was for KGW with seven. To the difference of detected location of QTL in all traits, PL and YD were larger and KGW was the least.8. The QTLs in some agronomic traits with biger LOD value and explanning higher parts of the variation, such as the QTLs for SR at qSR3-1, HD at qHD12-1 and KGW at qKGW5-1, were stable and detected under various environments. Some QTLs were only detected under special environments. The QTLs for KGW had little difference among environments.更多还原