The Improvement of Rice Through the Use of Marker-Assisted Backcrossing and Genome-Wide Association Study

Abstract

The first study of this M.S. thesis is on the incorporation of a blast resistance gene into rice cultivar Trinity. Rice blast disease, caused by the fungus Magnaporthe oryzae, is a significant biotic stress in modern-day rice agriculture. The Pi-b gene provides resistance to multiple races of M. oryzae and can be incorporated into rice cultivars. Rice cultivar Trinity has qualities favorable to the Texas rice industry, but it lacks major blast-resistant genes. The objective of this study was to incorporate the Pi-b gene into Trinity through marker-assisted backcrossing (MABC) and determine the genetic similarities between the backcrossed lines and Trinity. Trinity and Saber served as the recurrent parent and donor parent, respectively. Using MABC, individual plants that possessed the Pi-b gene ‘Pibdom’ DNA marker were identified, allowing backcrosses to be made within that same generation. Two backcross generations produced 38 BC2F1 lines with genetic similarities over 87.4% compared to Trinity, of which 25 have the Pi-b gene. These BC2F1 lines are recommended for further backcrossing to produce BC3 lines, followed by yield trial and disease rating test evaluations. The second section of this thesis is on the genome-wide association study (GWAS) of tiller density. Increasing tiller density has the potential to improve rice grain yield, and it is necessary to identify loci that are associated with this trait. The objective of this study was to perform GWAS on a variable population of 208 rice accessions to find single nucleotide polymorphisms (SNPs) associated with tiller density. The genotyped accessions were planted at Texas A&M AgriLife Research Center in Beaumont, Texas, and phenotyped for tiller density in 2020 and 2021. Tiller density ranged from 170 to 1,092 tillers/m2 in 2020 and from 170 to 870 tillers/m2 in 2021. The GWAS models used in the analyses were the mixed linear model (MLM), multiple loci mixed model (MLMM), and fixed and random model circulating probability unification (FarmCPU). The models identified 31 SNPs linked to 15 genes, including Os04g0503600 and Os04g0504000, which are significantly associated with tiller density and can be analyzed in future studies to determine the mechanism by which they impact tillering.