Genetic dissection of the dwarfing effect of the apple rootstock 'M.9' : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Plant Biology at Massey University, Palmerston North, New Zealand

The dwarfing phenomenon in apple is mainly characterized by the ability of a rootstock to reduce the vegetative growth and ultimate size of the scion grafted onto it. Many hypotheses have been proposed to explain the dwarfing effect of rootstocks, from the production and translocation of hormones to the altered anatomy of the graft union. However, despite the numerous studies undertaken, none of the studies support a single hypothesis. This study focussed on identifying genetic markers for genomic regions influencing the dwarfing effect of 'Malling 9' ('M.9'). Two rootstock populations derived from crosses between 'M.9' and the vigorous rootstock 'Robusta 5' ('R5') were propagated and phenotyped at the HortResearch Havelock North Research Centre (New Zealand). Bulked segregant analysis (BSA) and QTL analysis were performed using phenotypic data collected from scions grafted onto the rootstock populations. Several genomic regions in 'M.9' and 'R5' were found to have a significant influence on the dwarfing phenotype and other related traits. The percentage of variation explained by these QTLs ranged from 4.2% to 57.2%. A large number of traits had significant variation associated with the major dwarfing QTL DWJ (identified previously by BSA), confirming the influence of this gene on tree architecture. To identify the genes responsible for the dwarfing effect, a microarray analysis on RNA extracted from bark tissues was performed to detect genes differentially expressed among dwarfing and vigorous rootstocks derived from the 'M.9' x 'R5' family. Following the mapping of 16 markers developed from 12 candidate genes, their position was compared with those of the QTLs identified previously and co­ localisations among genes and QTLs were identified. Results to date indicate that none of these particular CGs co-segregate with DW1. In order to estimate the number of different genetic sources of dwarfing present in commercial rootstocks, two SSR markers mapping about 0.5 cM away from the dwarfing QTL DWJ were screened over 58 rootstock accessions. The majority of the dwarf and semi-dwarf accessions screened carried the locus DWl, indicating that there may be only a single genetic source of dwarfing in apple rootstocks. The identification of markers for dwarfing will have a major impact on apple rootstock breeding, which currently relies on laborious phenotyping of individuals in breeding populations that presently takes at least 5-7 years to adequately perform. The identification of the genetic function of DWJ would provide an opportunity to develop dwarfing rootstocks for other members of the Rosaceae family for which such rootstocks have not yet been developed