Candidate Genes and Stable QTL for Grain Yield and Seed Size in Durum Wheat
Abstract Background: In wheat grain yield is expressed as the product of different components. Among these, thousand kernels weight (TKW) reflects the combination of several grain related traits including grain length (GL), grain width (GW) and area. Grain weight is also affected by phenological traits, such as heading time (HT) and plant height (PH). To detect stable QTL and candidate genes involved in phenotypic control of grain yield, a recombinant inbred line (RIL) population derived from two elite durum wheat cultivars (Liberdur and Anco Marzio) was evaluated for yield components and grain related traits for three growing seasons in southern Italy. The mapping population was genotyped with a 90K SNP array and a high-density genetic linkage map with 5134 markers was obtained.Results: A total of 30 QTL were detected on the durum RIL population including 9 stable QTL for TKW (2 QTL), GL, GW (2 QTL), AREA, HT and PH (2 QTL) distributed on 1B, 2A, 3A and 6B chromosomes. Interestingly, a QTL cluster mapped on 2A included a major QTL for HT explaining at least 70% of phenotypic variance and co-located with a QTL for YLD, TKW, GL and GW and AREA, respectively. In the physical position of this QTL cluster a photoperiod sensitivity gene (Ppd-A1) was found. Serine carboxypeptidase, Big Grain 1 and β-fructofuranosidase candidate genes were mapped in clusters containing stable QTL. Candidate genes involved in auxin metabolism were also found in QTL clusters in which a QTL for AREA was declared. Conclusions: This study showed that yield components and phenological traits had higher inheritances than grain yield, allowing an accurate stable QTL cluster detection. This was a powerful requisite to physically map QTL on the reference durum wheat genome and to identify candidate genes strongly affecting the genetic grain yield network.