| Drought is a key limited factor for rice production and yield stability. In this study,149advanced backcross selected introgression lines (ILs) in six random BC2F2:5populations weredeveloped by elite restorers Shuhui527and Minghui86as recurrent parents, ZDZ057, Teqingand BG90-2as donors.8target traits were used for phenotypic analysis, correlation analysisand path analysis, including heading date (HD), plant height (PH), panicle length (PL), filledgrain number per panicle (FGP), total spikelet number per panicle (SPP), sterility fertility(SF), thousand grain weight (TGW) and grain yield per plant (GY). To identify the genomicregions contributing to yield and its components under drought stress and normal irrigationconditions, we detected quantitative traits loci (QTLs) by using Chi-square test and one-wayANOVA, respectively. The main results were as follows:1. Phenotypic performance of advanced backcrossing introgression lines under normalirrigation and drought stress suggested that drought stress caused a reduction in the averageyield per plant ranging from10.9%to63%. Under drought stress condition, yieldperformance of3populations in Shuhui527background is better than3populations inMinghui86background. Compared with recurrent parents, several traits even in the samebackgrounds presented different performance under normal irrigation and drought stressconditions. Such as an FGP, SPP and SF increasing in Shuhui527background, which haveconsistent performance, but have no significant difference was observed. However,3populations had different performance in Minghui86backgrounds. Under drought stresscondition, ILs has significant difference with recurrent parent. GY, FGP, SPP and PN havegreater effect. Moreover,23elite lines which had higher yield than recurrent parents wereidentified.2. The phenotypic correlation analysis and path analysis among yield and its componentsshowed that, GY was significantly correlated with PN, FGP and SPP in all populations undertwo conditions. Consistent performance was presented in the same populations under differentconditions. The path analysis of GY with PN, FGP and TGW in different populations undertwo conditions showed that GY can be increased by big panicle and many panicles. Anovaresults indicated that environment components were highly significant for yield and relatedtraits. For all traits explained an average of40.2%of the total traits variation in3populationswith Shuhui527background, and explained an average of7.2%in3populations with Minghui86background.3. Total62,67,50,19,27and59QTLs were detected in Shuhui527/ZDZ057,Shuhui527/Teqing, Shuhui527/BG90-2, Minghui86/ZDZ057, Minghui86/Teqing andMinghui86/BG90-2advanced backcross introgression populations, respectively. Among them,10,5,2,2,1and13QTLs were identified under two conditions, respectively. Moreover,7,8,3,1,3and10QTLs were detected by Chi-square text and one-way ANOVA.20,16,14,4,4and16tightly linked or pleiotropic QTLs were also identified. In addition,41QTLs weredetected by using the differences of trait under stress and irrigation, which were related to thedrought tolerance. Among the41QTLs,31QTLs were identified under two conditions, whichwere stably expressed and derived from the same parents. |
