| In China, the planting area for maize is24,634,000hm2, and the total yield of maize is121,310,000t with the unit area yield of4924.5kg/hm2. The planting area worldwide is137,500,000hm2, and total yield is602,200,0001with the unit area yield of4380kg/hm2. In USA unit area yield is10000kg/hm2. China’s per unit area yield of maize is higher than world average level, but lag far behind that of the USA. Compared with USA there is great space for development of hybrid maize product ability. The increase of China’s maize per unit yield has fluctuated since1990s, the reason for which is that the genetic basis of inbred lines is becoming narrower and narrower. Recently, new germplasm has been created and new inbred lines have been bred by introducing tropical-subtropical resources in maize in Institute of Agricultural Sciences of the Area Along Yangtse of Jiangsu. In order to evaluate the utilization value and genetic potential for further improvement of these inbred lines newly bred, we, firstly in this study, analysed general combining ability (GCA) and special combining ability (SCA) of11traits in9inbred lines (named as S1～S9) selected from the lines newly bred through seventy-two F1hybrids made by method3of Griffing diallel designs, and planted in Nantong and Nanjing, Jiangsu Province, respectively. Then, genetic analysis for9trait were conducted by using mixed major gene plus polygene inheritance models and joint segregation analytic method of P1, P2, F1, B1, B2and F2generations in two crosses made from the3elite inbred lines, S1, S3and S7. Lastly, QTL mapping was carried out for13traits by using a RIL population (109lines) made from a single cross hybrid of GY220/1145, using composite interval mapping method in both WinQTLCartographer2.5and QTLnetwork2.0softwares. Main research results obtained were as follows.1. Among the F1hybrids, genetic variances of non-additive were larger than those of additive in the variations of grain yield per plot, kernel number per row, ear length and kernel weigth per plant, and genetic variances of additive were greater than those of non-additive in the variations of row number per ear,1000-kernel weight, ear diameter,100-kernel volume, stalk weight per plant and growth duration. Genetic variances of additive of plant height different between Nantong and Nanjing. Inbred lines S7and S3had excellent GCA for kernel yield per plot,1000-kernel weight,100-kernel volume and kernel weigth per plant. Inbred lines S6and S3showed elite GCA for rows per ear and ear diameter. Inbred lines S9and S2showed good GCA for kernel number per row and ear length. Inbred lines S7and S2showed good GCA for high plant height and Inbred lines S4and S1showed good GCA for low plant height. Summing up the evaluation from the11traits, the best line was S3, then S7, S2,S1and S5in order, and last is S4. GCA for kernel number per row was improved in S9into which tropic and subtropic germplasm was introduced. Reciprocal effects of the traits were not significant at5%probability level except kernel number per row in both sites and plant height in Nantong.2. Kernel weight per ear and total weight per ear was controlled by two pairs of major gene with additive-dominance-epistatic effects plus polygenes with additive-dominance-epistatic effects, and the trait was mainly governed by major genes in both crosses.100-grain weight was controlled by two pairs of major gene with additive-dominance-epistatic effects plus polygenes with additive-dominance-epistatic effects in S1×S3, and by one pair of major gene with additive-dominance effects plus polygenes with additive-dominance effects in S3×S7, and was mainly governed by major genes in both crosses. Kernel row number was controlled by one additive-dominance major-gene and additive-dominance-epistasis polygenes, and was mainly governed by polygenes in cross S1×S3; whereas in cross S3×S7the trait was controlled by two additive-dominance-epistasis major-genes and additive-dominance-epistasis polygenes, and was mainly governed by major genes. Kernel number per row was controlled by one additive major-gene and additive-dominance polygenes, and was mainly governed by major gene in both crosses. Total effect of dominance was larger than that of additive in polygene loci. Ear length was controlled by additive-dominance-epistasis polygenes in cross S1×S3; whereas in cross S3×S7the trait was controlled by one additive-dominance major-gene and additive-dominance-epistasis polygenes. Ear diameter was controlled by one additive-dominance major-gene and additive-dominance-epistasis polygenes in cross S1×S3; whereas in cross S3×S7the trait was controlled by one wholly dominance major-gene and additive-dominance polygenes. Ear length and Ear diameter either was mainly governed by polygenes. Cob diameter was controlled by two additive-dominance-epistasis major-gene and additive-dominance polygenes in cross S1×S3; whereas in cross S3×S7the trait was controlled by two additive-dominance-epistasis major-gene and additive-dominance-epistasis polygenes, and was mainly governed by polygenes.100-kernel weight, kernel weight per ear, total weight per ear and kernel number per row was mainly governed by major genes.3.(1) Sixty-three loci were detected for the13traits by the CIM method. For trait of yield per ear,4QTLs, explaining7.8%～25.8%of phenotypic variation, were detected and2of the4were detected in both environments. For trait of rows per ear,4QTLs, explaining6.4%～11.7%of phenotypic variation, were detected and1of the4was detected in both environments. For trait of kernel number per row,4QTLs, explaining8.4%～17.2%of phenotypic variation, were detected and1of the4was detected in both environments. For trait of100-kernel weight,6QTLs, explaining7.0%～13.8%of phenotypic variation, were detected. For trait of total weight per ear,3QTLs, explaining8.5%～10.3%of phenotypic variation, were detected. For trait of ear length,3QTLs, explaining11.9%～22.0%of phenotypic variation, were detected. For trait of ear diameter,6QTLs, explaining6.7%～26.4%of phenotypic variation, were detected. For trait of cob diameter,3QTLs, explaining6.9%～17.0%of phenotypic variation, were detected. For trait of tip barren length,4QTLs, explaining6.5%～25.7%of phenotypic variation, were detected. For trait of kernel number per ear,3QTLs, explaining9.9%～16.5%of phenotypic variation, were detected and2of the3were detected in both environments. For trait of plant height,5QTLs, explaining6.4%～8.6%of phenotypic variation, were detected. For trait of ear height,4QTLs, explaining6.7%～12.8%of phenotypic variation, were detected. For trait of rough dwarf disease,6QTLs, explaining6.9%～17.6%of phenotypic variation, were detected and3of the6were detected in both environments. Six intervals with multiple effects were found among the intervals harboring loci. Locus between g5M5708-n55on23linkage conditioned yield per ear, total weight per ear, kernel number per row, ear diameter and kernel number per ear simultaneously. Locus between g4M3707-g6M6811on12linkage conditioned yield per ear, total weight per ear, kernel number per row, kernel number per ear and plant height simultaneously. Locus between g5M5813-g6M6808on9linkage conditioned row per ear, kernel number per ear and ear height simultaneously. Locus between g7M778-g4M3801on13linkage conditioned total weight per ear and yield per ear simultaneously. Locus between g8M8801-g8M8811on16linkage conditioned100-kernel weight and row per ear simultaneously. Locus between g5M5804-g5M5803on2linkage conditioned100-kernel weight and tip barren length simultaneously.(2) Nine main effect QTLs and7pairs of interaction between non main effect QTLs were detected by the QTLNetwork method. The9main effect QTLs were as follows. One was YE5-12, controlling yield per ear, explained7.4%of phenotypic variation. One was RE9-15, controlling row per ear, explained11.6%of phenotypic variation. Two were TWE5-12and TWE13-1, controlling total weight per ear, explained7.3%and6.6%of phenotypic variation respectively. Two were CD13-1and CD18-2, controlling cob diameter, explained7.7%and8.4%of phenotypic variation respectively. One was KNE12-5, controlling kernel number per ear, explained7.2%of phenotypic variation. One was PH5-10, controlling plant height, explained11.2%of phenotypic variation. One was KNE12-5, controlling kernel number per ear, explained7.2%of phenotypic variation. One was RDD2-22, controlling rough dwarf disease, explained9.0%of phenotypic variation. Seven pairs of interaction between non main effect QTLs were occurred linkages between1and3,5and18,3and5,5and19,7and12,1and2, and5and13, controlling yield per ear, row per ear,100-kernel weight, ear length, ear diameter, tip barren length and rough dwarf disease respectively, and explained4.7%-11.9%of phenotypic variation.(3) Six QTLs were detected simultaneously by the two genetic models, i.e. multiple regression model and mixed linear model. The6QTLs were YE5-12, controlling yield per ear, KNE12-5, controlling kernel number per ear, RDD2-22, controlling rough dwarf disease, RE9-15, controlling row per ear, PH5-10, controlling plant height, and TWE5-12, controlling total weight per ear. It is worth to study the6QTLs further since they are reliable QTLs. |
PDF Full Text Request