Utilization Of Genetic Mapping Results In Designing Wheat Breeding Methods

The major objective in wheat breeding is to develop new variety with high yield, good quality and robust biotic and abiotic resistances. Studies on breeding methods can provide useful information which faciliates the practical wheat breeding. Great achivements have been made in wheat conventional breeing, but there are still some questions to be answered. Questions in wheat breeding such as distribution of donor segment length from different mating designs and how to use genetic results to have more targeted breeding can be investigated though simulation.In this study, the Qu Line breeding simulation software was used to simulate different mating systems in deriving recombinant inbred lines(RILs), in order to quantify length of the donor chromsomal segment. One UK reference population, i.e. Avalon×Cadenza DH lines, was used to indentify gene information on yield and yield-related traits. The identified information was then used to design the target genotype based on the ojective to improve grain yield in UK. The other UK genetic population, i.e. Malacca×Hereward RILs, was used to indentify gene information on baking quality traits. The identified information was then used to design the target genotype based on the major requirements from consumers in UK. Design-led wheat breeding approaches were demonstrated by the identified gene information in the two populations. Parenatal genetic contribution and length of chromosome segment of parents in two populations were analysed as well. Majors results were given below.1. Donor segment length from different mating systemsWhen length of the target donor segment was less than 10 cM, the minimum population size decreased by the times of backcrossing and random mating. As length of chromosome increased from 100 c M to 400 c M, after foreground and background selection, length of donor segment for one time, two times, and three times of backcrossing ranged at 44.72 c M-67.29 c M, 37.09 c M-47.46 c M, 29.26 c M-37.77 c M, respectively. Compared with foreground selection only, foreground plus background selection can reduce length of backcrossing by 5%-20%, and reduce length of intermating by 26%-46%.2. Breeding design to improve the grain yield using the UK Avalon×Cadenza DH reference populationA total of 163 QTL for grain yield(GY), yield components, plant height(PH), ear emergence(EM), solid stem(SS) and yellow rust resistance(Yr) were identified in the UK Avalon×Cadenza DH reference population. The requirements for new wheat cultivars in UK in recent years are: similar PH and EM to Avalon, having Rht-D1 b and Vrn-A1 b alleles, high TGW and GN, long and wide grains, large root system, resistance to diseases, and maximum GY. Based on these requirements, the breeding target genotype was designed from the identified QTL in the reference population. Three single crosses, i.e. DH109×DH160, DH61×DH182 and DH27×DH61, and three-way cross or modified double cross among three parents DH182, DH109, DH98 can generate target genotype. Among the three single crosses, DH61×DH182 resulted in the largest number of target genotypes. Moreover, simulation results revealed that the F3-DH method was superior to selected bulk and modified SSD in achieving more target genotype and higher genetic gain. When DH182 was used as the third parent or the common parent, the three-way cross or modified double cross resulted in more target genotypes. Up to 4.69 more target genotype was obtained from modified double cross compared with three-way cross.3. Breeding design to improve bread baking quality using the UK Malacca×Hereward RIL populationA total of 19 QTL were identified for eight baking quality traits from 84 RILs derived from Malacca and Herewerd grown in two years. Expected baking quality consists of high loaf volume, high luminance, fine bread structure with a large number of small size, thin-walled cells and soft texture, based on which the breeding target genotype was designed from the identified QTL in the reference population. Target genotype was achieved from single cross RIL104×RIL17, topcrosses or double crosses among RIL107, DH62 and DH11. For cross RIL104×RIL17, multi-stage selection methods such as two-stage selection and three-stage selection were effective to reduce population size, work load, and cost in screening molecular markers. Using RIL107 as the third parent or common parent, the respective three-way cross or modified double cross resulted in the highest number of target genotype because the unfavorable linkage of QTL for L* and bread firmness was considered. Moreover, double crosses retained 1-2 more target genotype than three-way crosses. Therefore, when favorable genes were dispersed in three parents, double cross was more efficient to pyramid genes no matter whether linkage was included or not.