QTL Analysis For Stover Yield And Quality Traits And Their Genetic Relationship Using Two Connected F_2:3 Populations In Maize

Maize stover with rich crude fiber and high energy, is the main coarse forage for ruminant animal in northern china. In china, 150 billion kilograms of maize stover was produced annually. Improving maize stover yield and quality traits has important role in increasing economic benefits and social effects in maize production. High-oil corn has high commercial value and broad developmental prospects owing to its advantages in both kernel nutritional quality and stover quality. Previous researches about QTL mapping for several stover yield and quality traits have been conducted with normal maize germplasms, but such researches using high-oil maize were only limited on grain quality and ear-kernel traits. In this study, an elite high-oil maize inbred line GY220 developed from ASK germplasms was crossed with two elite dent maize inbred lines 8984 and 8622 to generate two connected F2:3 populations (P1F2:3 and P2F2:3). High-density genetic maps were constructed using SSR markers. The field experiments were conducted under two different environments, at Luoyang in spring and at Xuchang in summer. QTL analyses for 13 stover yield and quality traits were conducted by composite interval mapping (CIM) method. The interactions between detected QTL were analysis by multiple interval mapping (MIM) method. Joint QTL analysis were done using CIM method of multiple traits analysis among stover traits, and stover traits with their significantly related grain quality traits, ear-kernel traits, and plant traits, respectively. An intergrated map was constructed and consensus QTLs were identified using meta-analysis method in Biomercator 2.1. Our first objective was to reveal the molecular genetic mechanism of stover yield and quality traits, and the genetic relationship among stover traits and between them with other traits. The second objective was to identify major QTL for stover traits and their effective linked markers. These results will do great help in breeding for stover yield and quality traits, and in MAS, fine mapping, and map-based cloning for major QTL. The main results were as follows:1. Transgressive segregation was observed for all 13 stover traits in the two F2:3 populations. Normal distribution was observed for most traits except for 3 traits in P1F2:3 population and 6 traits in P2F2:3 population. Joint analysis of variance showed that F2:3 family lines, environments and the interactions between genotype and environment for most traits were all significant or highly significant. Heritabilities for fresh stover yield (FSY), crude protein content (CPC), in vitro dry matter digestibility (IVDMD) and neutral detergent fiber (NDF) were relatively high, ranging from 0.781 to 0.917, while those for hemicellulose concentration (HCC) and in vitro cell wall digestibility (INNDFD) were low, from 0.046 to 0.427.2. Totally, one hundred and forty four QTL were detected for 13 stover traits in the two F_2:3 populations under two environments and in combined analysis. 70 and 74 QTL were detected in P1F_2:3 population and P2F_2:3 populations, respectively. They were located on all ten chromosomes. Phenotypic variation explained by a single QTL varied from 4.8% to 20.1% in P1F_2:3 population and from 5.1% to 20.4% in P2F_2:3 population. Limited digenic interactions among detected QTL were found, and their effects were low. The number of QTL expressing additive, partially dominant, dominant and over-dominant effects were 10, 41, 20 and 73, respectively, which reflected that partially dominance and over dominance played the most part role in the genetics of stover traits. The positive alleles of 54.2% QTL for IVDMD and IVNDFD, and the negative alleles of 64.0% QTL for ADF and NDF were contributed by the high-oil parent line GY220, and the positive alleles of 77.8% QTL for FSY, and those of 57.1% QTL for NDF were contributed by the two dent corn parent lines.3. The stability in genetic backgrounds and environments for QTL of stover yield and quality traits were low. No common marker intervals associated with the same trait was found for the two populations. Only three QTL for CPC, and each one QTL for FSY and IVNDFD were detected in adjacent marker intervals, which were located at bin 2.04, 7.04, 7.04-7.05, 3.06-3.07 and 4.08-4.09, respectively. QTL for EEC and CPY were all detected at the same bin 9.03 and 5.03. Eighty-six QTL (accounting for 75.4%) were only detected under one environment. Twenty six QTL (accounting for 22.8%) were detected under both environments or under one environment and in combined analysis. Only 2 QTL (accounting for 1.8%) were commonly detected under two environments and in combined analysis. Among the 49 QTL with phenotypic variations higher than 10%, qlFSY1-5-1/qFSY1-5-1, qlDMC2-3-1/qDMC2-3-1, qlCPC1-7-1/qCPC1-7-2, qlCPC1-7-2/qCPC1-7-3, qlCPC2-5-1/qxCPC2-5-1/qCPC2-5-1, qxCPC2-5-2/qCPC2-5-2 and qlIVNDFD2-1-2/qxIVNDFD2-1-1/qIVNDFD2-1-1 showed high stability across different environments, and they could be used as the main objective QTL in further studies and in MAS.4. Significant correlations were almost consistently observed among three pairs of stover yield traits, thirteen pairs of stover quality traits, six pairs of stover yield with quality traits, between CPC with grain protein content, PH and EH, and between FSY and DMY with PH and EH in both populations and under different environments. Significant positive correlations were observed between ADF with NDF, and between IVDMD with IVNDFD. IVDMD, IVNDFD, FSY and DMC were all negatively correlated with ADF and NDF.5. Multiple trait joint analysis could improve QTL detection for significantly correlated traits. One thousand and four multiple-trait QTL were detected for 18 pairs of positively correlated traits and 12 pairs of negatively correlated traits, including three stover yield traits, seven stover quality traits, one grain quality traits and three plant traits. Six hundreds and seventy four new QTL were detected. Those QTL for correlated traits on several chromosomes might be QTL with pleiotropy or tight linked QTL.6. The linkage length of the integrated map for the two populations was 2 388.67 cM, which included 274 SSR markers, with an average interval of 8.72 cM between adjacent markers. Thirteen and 42"consensus"QTL were detected for 144 QTL associated with 13 stover traits, and 465 QTL associated with all 32 traits (including stover traits, grain quality traits, ear-kernel traits and plant traits), respectively. Most QTL for correlated traits were distributed on chromosomes in cluster. QTL for plant traits and stover yield traits, and for stover quality traits, grain quality traits and ear-kernel traits, were often located on chromosome regions for the same"consensus"QTL. QTL for stover traits were located on chromosomes 7, 3 and 5 in hotspot. True major QTL for stover traits might exist on three chromosome regions, 98.85 cM on chromosome 5, 197.03 cM on chromosome 3, and 114.33 cM on chromosome 7.7. CPC was the most important stover quality traits, since it was significantly correlated with several stover yield and quality traits, grain quality traits, and plant traits. CPC was positively correlated with FSY, IVDMD, IVNDFDD and grain protein content, and negatively correlated with DMC, ADF, NDF, PH and EH. Multiple trait joint QTL analysis among them reflected that their related QTL were pleiotropic or tight linked. Therefore, it was not easy to improve stover yield and quality traits simultaneously. Considering the breeding objective for stover yield and quality traits and their genetic correlations, it was very difficult to improve stove quality traits while increasing stove yield. It was possible to improve stover quality through conducting positive MAS on grain protein content or CPC, and to increase stover yield by planting large population.