Spatial Structure and Population Genetic Variation in a Eucalypt Species Complex

Abstract

In this study, the relative influences of selection, gene flow, and other evolutionary forces on the spatial structure of genetic variation within a eucalypt species complex (the spotted gums: genus Corymbia, section Politaria) were assessed. The study investigated the spatial genetic structure among four putative species of spotted gum (broad-scale), as well as within a single population (fine-scale)of one species, using both molecular and quantitative markers. The spotted gum complex occurs naturally across a range of 2500 km in eastern Australia. Spatial genetic variation within and between the four putative spotted gum species was examined using both chloroplast and nuclear markers. No significant differentiation was found between the three northern species of the complex, C. citriodora, C. variegata and C. henryi. The southern species, C. maculata, shared no haplotypes with any of the three northern species. These results disagree in part with those reported in a previous allozyme based study in which C. henryi was found to be significantly divergent from C. variegata (with which it is sympatric) and more closely aligned with C. maculata. Re-analysis of the allozyme data provided evidence of selection acting at the PGM2 locus within populations of C. variegata and C. henryi. The exclusion of this locus from the data set led to concordance between the cpDNA and nDNA analyses. Restricted gene flow and evidence of isolation by distance were identified as the dominant processes influencing the contemporary distribution of the cpDNA haplotypes. No geographic structure of haplotypes was found and complex genealogical relationships between haplotypes indicated the combined effects of past fragmentation, range expansion and possible long distance dispersal events. The variation and spatial structure in both neutral molecular markers and quantitative genetic traits were compared to explore the relative influences of dispersal and selection within a single eucalypt population. Both mature trees (n=130) from a natural population of C. variegata and their progeny (n=127) were sampled. A very high outcrossing rate (98%) was estimated for the population based on data from seven microsatellite loci. This suggested regular pollen–mediated gene flow into the population, further supported by the observed high levels of genetic diversity and polymorphism. Significant positive spatial structure was found between parent trees occurring up to 150 m apart in the natural forest, although genetic distance between these individuals suggested limited relatedness (i.e. less than half-sib relatedness). The effect of pollen-mediated gene flow appears, therefore, to swamp any effect of nearest neighbour inbreeding which has been reported in other studies of eucalypt populations and has been attributed to limited seed dispersal. Resistance to the fungal disease Sporothrix pitereka (Ramularia Shoot Blight) was measured on progeny from each of the population study trees. Substantial resistance variability was found, along with a high estimate in heritability of resistance (0.44 ± 0.06), indicating significant additive genetic variation within the population. Spatial analysis showed no significant spatial structure with resistant and susceptible genotypes apparently distributed randomly throughout the population. The lack of concordance between the molecular and quantitative markers suggests that there may be a cost to resistance. Temporal variation in the severity of disease outbreaks may have then led to differential selection of seedlings across many generations, maintaining variability in disease resistance and facilitating the apparent random distribution of disease resistant and susceptible genotypes throughout the population. C. variegata is an important commercial forestry species. The identification of strong genetic control in the disease resistance trait, as well as significant adverse genetic and phenotypic correlations between susceptibility and growth traits, will aid future breeding programs. Controlled crosses between resistant genotypes from this population should result in strong genetic gains in both resistance and growth, with little costs associated with inbreeding depression due to the highly outcrossed nature of the population.