Search

Commercial production of apricot is severely affected by sensitivity to climatic conditions, an adaptive feature essential for cycling between vegetative or floral growth and dormancy. Yield losses are due to either late winter or early spring frosts or inhibited vegetative or floral growth caused by unfulfilled chilling requirement (CR). Studies in this dissertation developed the first high-density apricot linkage map; followed by a comparative mapping strategy to validate conservation of synteny, genome collinearity and stable quantitative trait loci (QTLs) controlling CR and bud break between apricot and peach; and ultimately attempt to identify key candidate genes following a linkage disequilibrium-based association mapping approach to fine map the major CR QTL genomic regions. Following a two-way pseudotestcross mapping strategy, two high-density apricot maps were constructed using a total of 43 SSR (Simple Sequence Repeats) and 994 AFLP (Amplified Fragment Length Polymorphism) markers that span an average of 502.6 cM with an average marker interval of 0.81 cM. Twelve putative CR QTLs were detected using composite interval mapping, a simultaneous multiple regression fit and an additive-by-additive epistatic interaction model without dominance. An average of 62.3% ± 6.3% of the total phenotypic variance was explained. We report QTLs corresponding to map positions of differentially expressed transcripts and suggest candidate genes controlling CR. A majority of the QTLs were shown to be stable between both Prunus species, as well as similar trends in their QTL effects, with the allele for increasing the trait value mostly originating from the high chill parents. The denser apricot maps, due to more AFLP marker polymorphisms, provide a higher resolution to delineate QTLs to smaller genomic intervals, as well as splitting each of some of the peach QTLs into two. The comparative QTL mapping strategy presented here reveals the transferability of genetic information between two Prunus species, the characterization of stable QTLs, the utility of the maps to consolidate each other and to further validate previously identified CR QTL loci as a major controlling factor driving floral bud break. The LD-based association mapping was limited to marker dense genomic regions within and around previously detected major QTLs on linkage group (LG) 1 and 7. LD decayed below the centimorgan scale, indicating insufficient marker density averaged at 0.44 and 1.58 cM on LG1 and 7, respectively. Denser marker regions averaged at 0.1 and 0.7 cM on LG1 and 7, respectively, revealed significant LD estimates above the baseline threshold. We report significant marker-trait associations and underlying genes the markers were derived from. Our results demonstrate that an LD-based association mapping can be used for validating QTLs, fine mapping and detecting CGs in Prunus