2020-01-142019-11-12RAMOS, M. R. F. Análise comparativa dos métodos de avanço por Bulk e SSD na identificação de QTLs para produtividade de grão de arroz no cruzamento Epagri 108 X Irat 122. 2019. 145 f. Tese (Doutorado em Genética e Melhoramento de Plantas) - Universidade Federal de Goiás, Goiânia, 2019.http://repositorio.bc.ufg.br/tede/handle/tede/10297A relevant aspect of all rice breeding programs is the extensive genetic variability available and stored in germplasm banks. A major challenge is precisely how to select the most appropriate genotypes to meet the objectives of these programs. An interesting alternative is the assembly of core collections. Besides the characterization per se, the accessions that stood out for their genetic variability or productive performance were crossed in a diallel scheme. The resulting hybrids were self-fertilized to obtain generation F2, which was advanced by Bulk and SSD until F7. Among the most productive crosses, one in particular was interesting due to the genetic distance between the parents (RW = 0.91), and the high value of specific combining ability - Epagri 108 (Oryza sativa spp. Indica) x Irat 122 (Oryza sativa spp. Japonica). This study aimed to perform QTL analysis for plant yield and height using two populations of Epagri 108 x Irat 122 cross, advanced by SSD (generation F8) and Bulk (generation F7:8) methods. The 158 recombinant inbred lines of each method (SSD and Bulk) were evaluated for two years (2016/2017 and 2017/2018 seasons), in a 18x18 double lattice design with two replications, consisting of four-line plots of three meters in Palmital Farm (Goianira, GO). The RILs were genotyped by the DArTseq® methodology, which generated about 6,000 SNPs. The statistical model adopted for the grain yield data analysis was mixed linear model (MLM) through the R program. For the first and second year evaluations (2016/2017 and 2017/2018 seasons) and joint analysis (two years/seasons), the RILs-Bulk group presented higher grain yield averages when compared to the RILs-SSD and testers group. However, regarding the genetic variance component, the SSD group presented the highest estimate followed by Bulk and testers. Bulk-RIL yields ranged from 4,010.75 kg ha-1 to 5,815.42 kg ha-1, while SSD-RILs ranged from 3,321.76 kg ha-1 to 8,096.27 kg ha-1, both exceeding the testers group, which ranged from 2,754.30 kg ha-1 to 3,643.73 kg ha-1.For the plant height trait (ALT), in the first year, the plants ranged from 116 cm to 165 cm for RILs-Bulk. On the other hand, RILsSSD ranged from 91 cm to 177 cm, while the testers ranged from 100 cm to 104 cm. In the second year, RILs-Bulk ranged from 101 cm to 130 cm, while RILs-SSD ranged from 81 cm to 132 cm, while the testers presented heights from 96 cm to 117 cm. In the joint analysis, the testers presented the lowest heights. For QTL analysis, multiple interval mapping was used, with a total of 2,115 SNPs, and 3 QTLs were identified in the SSIL-RILs for the grain yield (PG) traitr, of which 2 QTLs were located on chromosome 6 (qGYLD6.1 and qGYLD6.2), one for the second year of experiment, with a phenotypic variation of 23.56%, and the other for the joint analysis, explaining 9.45% of the phenotypic variation. The other QTL was identified on chromosome 9 (qGYLD9) for the second year, with a phenotypic variation of 7.45%. For the trait height (ALT) a QTL on chromosome 1 (qPTHT1) was identified, with a phenotypic variation of 14.01%. For RILs-Bulk, with a total of 2,354 markers, 3 QTLs were identified for the PG character, two QTLs mapped on chromosomes 6 and 9 (qGYLD6 and qGYLD9), referring to the second year of evaluation, presenting a phenotypic variation of 21.65. % and 3.71%, respectively. In the joint analysis a QTL was mapped on chromosome 7 (qGYLD7), with phenotypic variation of 12.9%. For ALT no QTL was found in the RILs-Bulk. From the identification of these QTLs in haplotypic blocks, the next step will be the validation of markers in Embrapa germplasm bank accesses before being incorporated into the assisted selection routine, in order to identify materials with higher grain yield potential. For the Epagri 108 x Irat 122 cross, the SSD method was the most efficient in the generation of superior rice lines for grain yield, but at a higher operating cost than the Bulk method. RILs derived from both Bulk and SSD identified QTLs for the PG character; however, SSD identified a higher number of QTLs with greater effect on trait variation.application/pdfAcesso AbertoDialeloRILsMarcadores SNPsMapeamento por IntervaloProdutividade de grãosDiallelRILsSNP MarkersInterval MappingGrain yieldGENETICA::GENETICA VEGETALTese