QTL Analysis Of High Yield、Drought Tolerance And Salt Tolerance In Selected Population

Rice is the staple food for most Asian people. Rice production is thus highly relevant to the world food security. In recent years, with the deterioration of the environment, many abiotic stresses such as drought and salinity have become increasingly important in limiting rice production globally. Developing stress tolerant varieties is the most cost effective and promising approach to mitigate the problem.During the past 2 decades, DNA markers have been extensively used in genetic dissection of complex traits in rice and numerous quantitative trait loci(QTL) affecting a wide range of agronomic traits including abiotic stress tolerances have been identified. However, the identified QTL have rarely been used in marker-assisted selection in rice breeding because of their genetic background effects and their interactions with environments. Thus, it would be desirable to integrate QTL discovery with breeding. However, it is challenging for breeding populations since most materials in a breeding program consists of groups of advanced progeny each from high yield or specific target traits from many different bi-parental crosses by strong directional selection. It is not appropriate to map QTL affecting target traits in the breeding materials using the conventional QTL mapping methods because of the small population size of progeny from specific crosses and reduced phenotypic variations for target traits. To overcome the problems, this study was design to(1) develop the theoretical aspects and corresponding methodology to detect QTL affecting target traits in breeding populations;(2) to evaluate and validate the methodology using computer simulation and real data from breeding materials selected for grain yield, salt and drought tolerances from 4 different BC2 rice populations. The main results were summarized as follows:1. Development of the theory and methodology to detect QTL affect target traits using the selected breeding population Directional selection in selected population leads to the elevation of genotype frequency that associated to target traits and causes segregation distortion. Therefore, method developed for testing segregation distortion can be applied to detect the QTL related to target traits. However, there is a challenge for QTL mapping under selection that the reduced sample size and the variation in phenotypes obviously caused the reduction of statistical power in detecting QTL. To solve the problem, a generalized linear model method was developed to perform joint QTL mapping in multiple selected populations by detecting segregation distortion using a Bayesian statistics, Wald tests. This method involves adopting of the Markov chain to combine different breeding populations from the same environment and the generalized linear model to detect segregation distortion in the combined breeding populations. Three simulation results obtained:(1) QTL of large effect affecting target traits can be consistently detected even with small populations < 10 or with very high selection intensity;(2) under the same scenarios of selection and sample size, the detection power was higher for the dominance fitness model than the additive one;(3) multi-population segregation distortion analysis(MSDA) has greater statistical power in QTL detection than single populations;(4) seven QTL affecting grain yield were identified from 68 BC2F3 selected for high yield from 2 BC populations, 4 of which were verified by their close locations with cloned QTL genes and by mapping from corresponding random populations or progeny testing. Thus, this mapping by-selection(MBS) approach provides a new avenue for breeders to improve breeding progress, and can be applied to breeding programs not only in rice but also in other agricultural species including crops, trees and animals.2. Development of high yield, drought and salt tolerant introgression lines(ILs) Using the four backcross populations, 17 DT ILs were selected, among which 7 promising ILs that had improved DT and HY than JG88. Comparing to the JG88, 6 ILs had increasing ST. The promising ILs from the crosses between japonica lines has greater grain yield than that from the crosses between japonica and indica.3. Detection of QTL affecting drought and salt tolerances(DT and ST) using MSDAUsing 68 HY, 133 DT and 146 ST BC2 ILs from 4 Jigeng 88(JG88, the recurrent parent BC2F2 populations, plus four random BC2F2 populations of the same crosses and the MSDA method, 21 QTLs related to HY, DT and ST were identified and mapped to chromosomes 1, 2, 3, and 5, 6, 7, 11. One QTL at 78 c M on rice chromosome 1 was associated with both DT and ST, and so was a QTL at 26.2 c M of rice chromosome 6 with HY and ST.4. The transcriptome analysis of DT ILs CF16 and recurrent parent JG88The gene expression profiles of a DT BC2F4 IL CF16 carrying the target QTL and its RP, JG88 under drought stress and non-stress conditions were analyzed using RNA sequencing. Five differential expressed genes(LOC_Os07g01960, LOC_Os07g02260, LOC_Os07g05840, LOC_Os07g13490 and LOC_Os07g03040) were identified in the region that the q DTY7.2 located in. These genes provided useful candidates for the DT QTL to be further confirmed in future experiments.The developed high yielding, DT and ST JG88 ILs and the information of their carried QTL provided useful materials and genetic information to develop new varieties with significantly improved yield, DT and ST by QTL pyramiding in future.