Multipoint-likelihood maximization mapping on 4 segregating populations to achieve an integrated framework map for QTL analysis in pot azalea (Rhododendron simsii hybrids)

Background: Azalea (Rhododendron simsii hybrids) is the most important flowering pot plant produced in Belgium, being exported world-wide. In the breeding program, flower color is the main feature for selection, only in later stages cultivation related plant quality traits are evaluated. As a result, plants with attractive flowering are kept too long in the breeding cycle. The inheritance of flower color has been well studied; information on the heritability of cultivation related quality traits is lacking. For this purpose, QTL mapping in diverse genetic backgrounds appeared to be a must and therefore 4 mapping populations were made and analyzed. Results: An integrated framework map on four individual linkage maps in Rhododendron simsii hybrids was constructed. For genotyping, mainly dominant scored AFLP (on average 364 per population) and MYB-based markers (15) were combined with co-dominant SSR (23) and EST markers (12). Linkage groups were estimated in JoinMap. A consensus grouping for the 4 mapping populations was made and applied in each individual mapping population. Finally, 16 stable linkage groups were set for the 4 populations; the azalea chromosome number being 13. A combination of regression mapping (JoinMap) and multipoint-likelihood maximization (Carthagene) enabled the construction of 4 maps and their alignment. A large portion of loci (43%) was common to at least two populations and could therefore serve as bridging markers. The different steps taken for map optimization and integration into a reference framework map for QTL mapping are discussed. Conclusions: This is the first map of azalea up to our knowledge. AFLP and SSR markers are used as a reference backbone and functional markers (EST and MYB) were added as candidate genes for QTL analysis. The alignment of the 4 maps on the basis of framework markers will facilitate in turn the alignment of QTL regions detected in each of the populations. The approach we took is thoroughly different than the recently published integrated maps and well-suited for mapping in a non-model crop.