Identification of quantitative trait loci for grain yield in a recombinant inbred B-line population in sorghum.

摘要

In any crop, yield is typically the most complex trait to analyze, dissect, and make improvements. In recent years, sorghum improvement programs have observed diminishing rates of genetic improvement in grain yield. While several factors are responsible for this trend, it is clear that breeders must increase the genetic variability among elite sorghum germplasm and make better use of the extensive genetic resources in sorghum. Before this can be achieved, intensive work in understanding the genetic factors controlling yield potential must be done. This research reports on the identification of DNA loci controlling the phenotypic expression of grain yield and yield components such as plant height and maturity in sorghum (Sorghum bicolor L. Moench). A linkage map was created using a recombinant inbred line population from the cross between BTx3197 and BTx623 sorghum lines. The map consists of 138 DNA markers distributed along 13 linkage groups. The total genome coverage is nearly 900 cM with an average distance between markers of 13 cM. Phenotypic data was collected on the recombinant inbred line population per se in four environments (two locations, two years). While, phenotypic data on the recombinant inbred line population in testcross combination was collected in 2 environments (two locations, one year). Quantitative trait loci (QTL) analyses of population per se revealed a total of 14 genomic regions controlling traits for panicle number, panicle weight, panicle length, yield, days to mid-anthesis, plant height, and exsertion. These candidate QTLs explained from 10 to 50% of the variation for each trait, indicating the degree of complexity on some of the analyzed phenotypic traits. Finally, the presence of these genetic associations was corroborated in the testcross population to report their performance in hybrid combination. In this population, a total of 12 genomic regions were detected for all the traits but panicle number and panicle weight. Most of these genomic regions were unique in the testcross population indicating the importance of QTL evaluation in hybrid combination.