| Maize plays an important role in grain and forage production, so the high and stable yield has been a primary target for maize breeding. Kernel size, including kernel length, width, thickness, and yield components are complex traits under the control of multiple quantitative trait loci (QTL), and have influence on maize yield and quality. Therefore, kernel size and yield components become important indicators for improving grain yield. In this study, we constructed an F2segregation population with270individuals as mapping population, which was derived from two elite maize inbred lines, V671with large kernel size and cob, while MC with small kernel size and cob. F23family of each F2plant was used for phenotypic evaluation of20-kernel length (KL),20-kernel width (KL),20-kernel thickness (KT) and100kernel weight (HKW) as well as ear diameter (ED), ear length (EL), cob diameter (CD) and ear row number (ERN) in five environments. QTL mapping were carried out to investigate the genetic basis of eight traits mentioned above through single-environment analysis with composite interval mapping (CIM) and joint analysis with mixed linear model-based composite interval mapping (MCIM). The main results are summarized as follows.1. Analysis of variance based on phenotypic data showed significant difference (P<0.05) between families and environments, and that significant transgressive segregation has been observed in the range of traits variation. Broad-sense heritability of all traits ranged from0.821to0.957, out of which, the highest was ear row number, and the lowest was ear diameter; kernel-size traits generally had medium to high heritability. The abundant phenotype variations and genetic stability could ensure the feasibility of our study on QTL mapping of these traits.2. Correlation analysis results suggested that KL, KW and KT positively correlated with HKW. KT has a positive correlation with KW, while a negative correlation with KL was observed. ED, CD and ERN all positively correlated with each other. But, the correlation between HKW and ERN was negative. Additional analysis of regression showed that KL, KW, KT had a positive correlation with HKW, and the highest contribution to HKW was KT. HKW was negatively influenced by ERN.3. The genetic linkage map was constructed using256SSR markers with a length of1351.50cM across maize genome. The average distance between adjacent markers is5.28cM. Among the256markers in this study, frequencies of alleles in170sites were in accordance with the ratio of1:2:1. While a total of59markers showed a distorted segregation, mainly distributing on chromosome1,2,3,7,9and10. There was a large segregation hotspot at3.00-3.08bin on chromosome3, in which21markers were biased toward V671. Other distorted markers biased toward heterozygote. The relative position of markers on the linkage map was in good agreement with IBM2008Neighbors.4. Using CIM,115QTL for kernel size and yield components were found by single-environment analysis in270F2:3families, mainly distributed on chromosome1,2,4,5.6and9.6QTL, as the minimum, for KL accounted for1.18%-12.92%of phenotypic variation; and18QTL, as the maximum, for KT, accounted for0.84%-17.98%.51QTL were detected in at least two environments. Most QTL contributed less than10%phenotypic variation.21stable major QTL-explained>10%of phenotypic variation in at least one environment and detected in at least two environments based on single-environment analysis-were detected for KL, KW, KT, HKW, CD and ERN, except for ED and EL.5QTL, qKW1-2, qKT1-4, qCD9-2, qERN2-4and qERN9-1were expressed among all environments and also identified by joint analysis, four of which were as major QTL except forqCD9-2. A total of55QTL were detected via joint analysis with MCIM, of which44QTL were in agreement with in single-environment analysis results, including16major QTL; and the rest, as minor QTL, were only detected by joint analysis. Both single-environment and joint analysis indicated the kernel size and yield components were under the control of several stable major QTL and many minor QTL.5. Joint analysis for all environments showed that some minor QTL interacted with environment(s), at low degree of heritability. But the interaction between major QTL and environments was too small to be identified. Eight pairs of epistatic interaction loci (P<0.05) were detected with additive-by-additive, additive by dominance, dominance by additive or dominance by dominance interaction effects. The heritability of epistatic interactions were generally0.02%-2.98%and no significant interaction with environment was found. Epistatic interaction was always detected between significant QTL, but few were occurred in non-significant loci, indicating that epistatic effect was an important genetic basis of kernel size and yield components.6. The study identified several regions on bin2.05-2.07, bin4.08, bin5.01-5.03, bin9.03-9.04that were clustered with some overlapped QTL for kernel size and for yield components. QTL for kernel size were also concentrated on bin1.02-1.03, bin1.04-1.06, bin2.01-2.03and bin2.08; QTL in the region of1.09-1.10bin,2.03-2.04bin,6.04-6.05bin and9.05bin had a significant influence on yield components. Clustered QTL showed the phenotypic and genetic correlation of different traits, and indicated that the ’multi-factorial linkage’ of multiple QTL or pleiotropic regulator(s)in these QTL could be great valuable for improving grain yield via marker assisted selection (MAS). |
PDF Full Text Request