Abstract
Faba bean is a key protein feed and food worldwide that still requires accurate genomic tools to facilitate molecular marker-assisted breeding. Efficient quantitative trait locus (QTL) mapping in faba bean is restricted by the low or medium density of most of the available genetic maps. In this study, a recombinant inbred line faba bean population including 124 lines from the cross Vf6 x Vf27, highly segregating for autofertility, flowering time, plant architecture, dehiscence, and yield-related traits, was genotyped using the 'Vfaba_v2' SNP array. Genotypic data were used to generate a high-density genetic map that, after quality control and filtering, included 2,296 SNP markers. The final map consisted of 1,674 bin markers distributed across the six faba bean chromosomes, covering 2,963.87 cM with an average marker distance of 1.77 cM. A comparison of the physical and genetic maps revealed a good correspondence between chromosomes and linkage groups. QTL analysis of 66 segregating traits, previously phenotyped in different environments and years, identified 99 significant QTLs corresponding to 35 of the traits. Most QTLs were stable over the years and QTLs for highly correlated traits were mapped to the same or adjacent genomic regions. Colocalization of QTLs occurred in 13 major regions, joining three or more overlapping QTLs. Some of the pleiotropic QTL regions, especially in chromosome VI, shared the same significant marker for different traits related to pollen quantity and size, number of ovules per ovary, seeds per pod, and pod set. Finally, several putative candidate genes for yield-related traits, recently identified using a genome-wide association study, fall inside the colocalizing groups described in this study, indicating that, apart from refining the position of the QTLs and the detection of candidates, the dense new map provides a valuable tool for validation of causative loci derived from association studies and will help advance breeding programs in this crop.