Population genetics of native and domesticated Eucalyptus globulus

This thesis consists of firstly, a study of the local genetic dynamics occurring in a native population of Eucalyptus globulus, and secondly, an examination of genetic diversity within the Australian E. globulus breeding population. Fine scale spatial genetic structure within a continuous E. globulus forest on the Tinderbox Hills, Tasmania, was examined by genotyping mature trees (n = 168) and juveniles (n = 110) in a 140 m diameter sample site, using microsatellite markers. Spatial genetic autocorrelation revealed a significant decline in genetic similarity with distance between individuals, with the mature cohort displaying greater spatial genetic structure than the juvenile cohort. High-resolution analysis, using a combination of Bayesian clustering, ordination and spatial interpolation, revealed a complex pattern of genetic groups superimposed upon each other. Comparison between cohorts, and parentage analysis, indicated a directional shift in the distribution of genetic variation over generations, despite no differences in overall genetic diversity being detected between cohorts. Paternity analysis of 549 open-pollinated seed from nine trees identified 374 mating pairs. Sixty-seven percent of the pollinations involved trees from within the sample population where pollen dispersal was leptokurtic but not directional. Threshold relatedness values for full-sibling, half-sibling and unrelated relationships were calculated by simulation to quantify bi-parental inbreeding. On average, 10% of progeny were derived from self-pollination, with a further one percent and 13% derived from crossing between likely full- and half-siblings respectively. A further 43% of progeny were the product of unrelated individuals within the study site. Significant differences between families in the level of self-pollination and bi-parental inbreeding were revealed. Seed, juvenile and mature cohorts of the population showed high levels of observed and expected heterozygosity and there was little evidence for selection against homozygotes across cohorts. Genetic diversity within the first generation of the Australian E. globulus breeding population was compared with that observed between mature trees sampled from the entire natural distribution of the species. Observed and expected heterozygosity in the breeding populations was at least as high as that displayed by the native samples, and there was no evidence of increased inbreeding in the breeding population. The breeding population captured a significant amount of genetic variation from the natural - distribution of the species with most selections fitting closely to their native race of origin. However, a number of pedigree errors were detected at the level of race, family and genotype. This study provides a benchmark for monitoring genetic diversity during the course of domestication of E. globulus. Genetic diversity within the first generation of the Australian E. globulus breeding population was compared with that observed between mature trees sampled from the entire natural distribution of the species. Observed and expected heterozygosity in the breeding populations was at least as high as that displayed by the native samples, and there was no evidence of increased inbreeding in the breeding population. The breeding population captured a significant amount of genetic variation from the natural - distribution of the species with most selections fitting closely to their native race of origin. However, a number of pedigree errors were detected at the level of race, family and genotype. This study provides a benchmark for monitoring genetic diversity during the course of domestication of E. globulus. Genetic diversity within the first generation of the Australian E. globulus breeding population was compared with that observed between mature trees sampled from the entire natural distribution of the species. Observed and expected heterozygosity in the breeding populations was at least as high as that displayed by the native samples, and there was no evidence of increased inbreeding in the breeding population. The breeding population captured a significant amount of genetic variation from the natural - distribution of the species with most selections fitting closely to their native race of origin. However, a number of pedigree errors were detected at the level of race, family and genotype. This study provides a benchmark for monitoring genetic diversity during the course of domestication of E. globulus