Mapping genomic regions for grain yield and its components in Triticum turgidum L. var. durum across different environments

AKKIRAJU P; GOMÉZ P; RONCALLO PF; CERVIGNI, G D L; CARRERA A; CONTI V; MIRANDA R; WEHRHAHNE L; JENSEN C; BARIFFI H; ECHEIQUE V

St. Petersburg, Rusia

Congreso; 8th International Congress of Wheat.; 2010

VIR, CIMMYT, ICARDA, BGRI

Developing crop cultivars with high grain yield has been the principal aim of durum wheat breeding programs worldwide. The present work was focused on the identification of Quantitative Trait Loci (QTL) for grain yield and its components using a population of 93 RILs obtained from the cross between the variety Kofa and the line UC1113 by SSD. Six field trails using a randomized complete block design with three replications were conducted during 2006/07 and 2007/08. As plant materials the RILs population, the parental lines and eight local varieties from Argentina were used. Phenotypic values were registered for each yield component from 5 to 10 individual plants. An overall mean dataset was generated by averaging each trait data from overall replicates. The genetic map used in this QTL mapping was developed by Zhang et al. (2008, TAG 117 (8). p. 1361). Composite-Interval Mapping (CIM) was used to identify QTL for each trait in each environment and the average values across all the environments. Grain yield (Kg/ha) had to be recalculated based on yield components (spikes/m2, grain number per ear, thousand-grain weight) in order to decrease the effect of an unexpected genetic problem affecting some of the RILs (sterile white spikes). The present work identified a total of 103 significant QTLs, including 80 new QTL, for grain yield (Yld), thousand-grain weight (Tgw), test weight (Tw), plant height (Ht), peduncle length (Pl), harvest index (Hi), spikelet number per ear (Sne), grain number per ear (Gne), grain weight per ear (Gwe), spike fertility (Ft), grain number per fertile spikelets (Gnfs) and grain number per total spikelets (Gnts). For all the mentioned traits the RILs showed transgressive inheritance. In each environment the distribution of phenotypic values for most of the traits (except grain yield) was normal. The heritability for Tgw, Hlw, Ht, Pl, Sne and Gnts was moderate to high (46%-92%), being lower for Gwe and Ft (22%). However, for grain yield lower heritabilities (27%-57%) were resulted due to high environmental influence. The Genotype x Environment interaction was highly significant for all the traits. The threshold for declaring the presence of a significant QTL for each trait-environment combination was determined by 1000 permutations. QTL detected in different environments were considered to be the same if the estimated map position of their peaks was positioned within 15cM. A QTL was considered to be the “stable” if it appears atleast in three environments and in the overall mean. Additionally, “suggested QTL” have been reported if they appear atleast in two environments and in the overall mean. The LOD scores for the QTLs ranged between 1.9 (Hi) and 3.4 (Yld). Out of 103 significant QTL identified, Tgw showed the maximum number (14), where as Tw, Gwe and Ft showed the minimum (6). Out of 14 stable QTL identified, Ht and Pl showed the maximum number (5), while Gnts showed the minimum (1). One QTL each of Ht and Pl on chromosome 3B were stable across all the environments and also in overall mean. Out of 10 suggested QTL identified, Gnfs showed the maximum number (3), where as Tw, Gwe and Gnts showed the minimum (1). The phenotypic variation (R2) explained by the individual QTLs of all the considered traits ranged from 6.6% (Gnfs) to 45% (Ht). Both parental lines contributed with favorable alleles for the majority of traits, but the contribution from UC1113 was higher. Pleiotropic regions were identified for various traits over nine chromosomes, mainly 2B, 3A, 3B, 4A, 4B and 5A. The co-location of QTL on chromosome 3B involved ten yield components (Yld, Tgw, Tw, Ht, Pl, Gne, Gwe, Ft, Gnfs, Gnts), making it as an important region for marker-assisted selection. Positive and highly significant correlations were found between grain yield and its components. Gne, Gwe, Gnfs and Gnts showed positive correlations with grain yield in all the environments. Gne and Gnts showed strong positive correlations with grain yield (r = 0.82). The strongest positive correlation was found between Ht and Pl (r = 0.91). Indirect selection for these components can increase the grain yield of durum wheat. The reported QTL can be used for: introgression of favorable genes and QTLs into interesting plant materials, marker-assisted selection and germplasm screening for the evaluated traits.