Genome mapping of Eucalyptus globulus

This thesis covers the construction of genetic linkage maps of two Eucalyptus globulus parental trees (chapter 2), the detection of quantitative trait loci (QTL) carried out using these linkage maps (chapter 3) and a genetic analysis of rooting ability of micropropagated cuttings from two E. globulus families (chapter 4). Genetic linkage maps of two parent trees were constructed using 326 RAPD and 21 microsatellite markers. At a LOD score threshold of 4.9 for grouping and 3.0 for marker ordering, the male parent had 13 linkage groups consisting of 101 framework markers and the female parent had 11 linkage groups with 97 framework markers. Even though both parents originated from the same provenance it was found that polymorphic RAPD markers were readily detectable. Linkages between microsatellites previously reported for E. grandis/E. urophylla were conserved in the E. globulus cross. Segregation distortion of markers was found to be more prevalent than expected by chance. The linkage maps were used to detect QTLs based on 155 progeny grown in field trials. Twelve traits were included in the QTL analysis. These included: wood density (Pilodyn penetration), extent of early flowering (bud abundance), and growth (height at years one and two, stem diameter at years two, three, four and six, and relative incremental growth between years one and two, two and four, and four and six). Using interval mapping a total of eight QTL with LOD score peaks over 2.0 were detected, corresponding to seven map intervals. QTL detected included: two for cumulative growth; two for wood density; one for early flowering and three for relative incremental growth. Since the 155 progeny trees were grown at seven trial sites, an analysis of marker by site interaction was carried out with more marker by site interactions being found than expected by chance. Markers with significant QTL effects were examined for interaction with site with both of the cumulative growth QTL found to have significant site interaction. QTL stability with age was also analysed for these QTL, and both growth QTL were found to have a detectable association with the first measurements of height and diameter. A study of the rooting ability of cuttings grown in tissue culture from two families of E. globulus was undertaken with the aim of detecting QTLs for this trait. The two families were found to have significantly different rooting abilities indicating that genetic variation for the trait was present. Estimates of variance components relating to genetic and environmental effects indicated that within both families the variance due to genetic effects was small compared to the environmental variance with clonal repeatabilities of 0.17 and 0.14 for the two families. These estimates were used in power calculations based on t-tests of single markers under some simple assumptions. However even under optimistic circumstances the power for QTL detection was found to be too low to warrant the time and expense of genotyping required for QTL analysis.