Search

Eucalyptus globulus Labill is a foundation tree species over its disjunct distribution in southeastern Australia. The quality of its pulp makes it the most important hardwood species in the world. The importance of E. globulus prompted the establishment of common gardens from seed collected across its geographic range. This enabled us to study the genetic structure of the species, its population boundaries, and gene flow using 444 trees from different open-pollinated families that were genotyped at 16 microsatellite loci. A Bayesian clustering method was used to resolve five genetically distinct groups across the geographical range. These groups were identified as regions, which varied in diameter from 38 to 294 km and contain 4 to 16 putative populations. For two of these regional groups, we used spatial autocorrelation analysis based on assignment of trees to their natural stands to examine gene flow within each region. Consistent significant local-scale spatial structure occurred in both regions. Pairs of individuals within a region showed significant genetic similarity that extended beyond 40 km, suggesting distant movement of pollen. This suggests that breeding populations in E. globulus are much bigger than traditionally accepted in eucalypts. Our results are important for the management of genetic diversity and breeding populations in E. globulus. Similar studies of a variety of eucalypts pollinated by insects and birds will determine whether the local-scale genetic structure of E. globulus is unusual.

Tree growth and vegetative propagation are complex but important traits under selection in many tree improvement programmes. To understand the genetic control of these traits, we conducted a quantitative trait locus (QTL) study in three full-sib families of Eucalyptus nitens growing at two different sites. One family growing at Ridgley, Tasmania had 300 progeny and two clonally replicated families growing at Mt. Gambier, South Australia had 327 and 210 progeny. Tree growth was measured over several years at both sites and percentages of roots produced by either stem cuttings or tissue culture were assessed in the two Mt. Gambier families. Linkage analysis of growth traits revealed several QTLs for later year traits but few for early year traits, reflecting temporal differences in the heritabilities of these traits. Two growth QTL positions, one on LG8 and another on LG11 were common between the Ridgley and Mt. Gambier families. Four QTLs were observed for each of the two vegetative propagation methods. Two QTLs for vegetative propagation on LG7 and LG11 were validated in the second family at Mt. Gambier. These results suggest that growth and vegetative propagation traits are controlled by several small effect loci. The QTLs identified in this study are useful starting points for identifying candidate genes using the Eucalyptus grandis genome sequence.

The neighbourhood model apportions offspring of individual mother plants to self-fertilization, outcrossing to males within a circumscribed area around the mother plant (the neighbourhood), and outcrossing to males outside the neighbourhood. Formerly the model was applied only to haploid pollen gametes in the offspring of conifers, but is extended so that it can be used with genotypic data from diploid offspring of both angiosperms and gymnosperms. In addition, it is shown that the mating parameters can be estimated without independent estimates of allele frequencies in the pollen pools outside the neighbourhood; thus the model might be applied effectively to natural populations exposed to unknown external pollen sources. Parameters of the neighbourhood mating model were estimated for a 10-year-old seed orchard population of the insect-pollinated tree, Eucalyptus regnans, in southeast Australia, which contained a mixture of two geographical provenances (Victoria and Tasmania). The mating patterns revealed were complex. Crosses between trees of the same provenance occurred three times more often than crosses between trees of different provenances. Levels of self-fertilization and patterns of mating within neighbourhoods were influenced by provenance origin, crop fecundity and orchard position (central vs. edge) of mother trees. Gene dispersal, however, was extensive, with approximately 50% of effective pollen gametes coming from males more than 40 m away from mother trees (average distance between neighbouring trees was 7.4 m). Thus, insect pollinators are efficient promoters of cross-fertilization in this orchard, with the result that the effective number of males mating with each female is large.

Cellulose microfibrils are the major structural component of plant secondary cell walls. Their arrangement in plant primary cell walls, and its consequent influence on cell expansion and cellular morphology, is directed by cortical microtubules; cylindrical protein filaments composed of heterodimers of alpha- and beta-tubulin. In secondary cell walls of woody plant stems the orientation of cellulose microfibrils influences the strength and flexibility of wood, providing the physical support that has been instrumental in vascular plant colonization of the troposphere. Here we show that a Eucalyptus grandisbeta-tubulin gene (EgrTUB1) is involved in determining the orientation of cellulose microfibrils in plant secondary fibre cell walls. This finding is based on RNA expression studies in mature trees, where we identified and isolated EgrTUB1 as a candidate for association with wood-fibre formation, and on the analysis of somatically derived transgenic wood sectors in Eucalyptus. We show that cellulose microfibril angle (MFA) is correlated with EgrTUB1 expression, and that MFA was significantly altered as a consequence of stable transformation with EgrTUB1. Our findings present an important step towards the production of fibres with altered tensile strength, stiffness and elastic properties, and shed light on one of the molecular mechanisms that has enabled trees to dominate terrestrial ecosystems.

Eucalyptus leucoxylon is a widespread woodland tree species found in southeastern Australia that has suffered from, and continues to be, threatened by the impacts of habitat clearance and degradation. Populations now consist predominantly of scattered individuals, and their conservation status is of increasing concern. We report the development and characterization of a set of eight highly polymorphic microsatellite loci for E. leucoxylon. The loci can be amplified in three PCR multiplexes and electrophoresed in a single lane, allowing rapid throughput of large numbers of samples. A total of 111 alleles were detected in 68 individuals with an average of 12.3 alleles per locus, a mean expected heterozygosity of 0.83, and a mean observed heterozygosity of 0.72. The combined probabilities of identity and probabilities of paternity exclusion allow an extremely precise level of individual identification, indicating that these microsatellite markers will be ideal for population genetic and parentage-type studies in E. leucoxylon. The markers also exhibited an average of 76% conservation within the subgenus Symphyomyrtus, to which E. leucoxylon belongs, and 53% conservation across other subgenera of Eucalyptus, demonstrating the potential of these markers in ecological and breeding studies in a wide range of Eucalyptus species.

Ten microsatellite loci were used to DNA profile 215 first and second generation breeding clones in the Australian radiata pine breeding program. The clones were essentially unique from one another and these DNA profiles were used to examine the sexual and asexual phases in the breeding program. Only 2.6% of ramets from 25 breeding clones were incorrect whereas 8.4% of progeny from 10 families were genetically incorrect. The lineages of second generation clones were checked for DNA profiles that were available for putative parents. Some parents appear to be incorrect on the basis of the DNA profiles and further sampling will be required to confirm these results across the breeding program. Microsatellite markers can now enable most questions to be resolved in breeding programs and also will allow novel breeding strategies to be used in the future.

Impacts of forest harvesting and regeneration practices on genetic diversity in the Australian native forest species Eucalyptus consideniana Maiden (yertchuk) were examined using 29 Mendelian DNA markers (18 RFLPs and 11 microsatellites). Two replicate logging coupes were studied from each of the two most commonly employed silvicultural treatments: clear felling with aerial re-sowing and the seed tree system. For each coupe, genetic diversity measures were compared between a sample of the sapling regeneration and a corresponding control sample from bordering unharvested trees. When calculations were performed over all 29 loci, significant reductions of allelic richness (AR), effective number of alleles (AE) and/or expected heterozygosity (HE) were detected on one or both of seed tree coupes, but on neither of the clear falls. When calculations were performed over the 11 microsatellites alone, all three measures, AR, AE and HE, were significantly reduced on both of the seed replicates but on neither of the two clear falls. In contrast, when the RFLPs were examined separately, there were no significant reductions of diversity on either of the two seed tree coupes or on the two clear falls. These results suggest that genetic erosion is more likely under the seed tree system than under clear-felling with aerial re-sowing and that there is greater statistical power to detect it with microsatellites than with RFLPs. A Monte Carlo simulation to test the statistical significance of the number of apparently lost or gained alleles showed that significant losses of alleles above specified threshold frequencies occurred only in the two seed tree replicates. Three of the four control and regeneration population pairs were significantly differentiated, as indicated by exact tests or by pairwise FST estimates. Comparisons of CONTML dendrograms, constructed for the regeneration populations only versus the control populations only, indicated that genetic drift was significantly promoted under forest management. No significant decreases in observed heterozygosity, or increases in the panmictic index (f), were observed in any of the comparisons suggesting that inbreeding was not promoted by a single rotation of forest management.

Potential impacts of regeneration practices ongenetic diversity in the Australian nativeforest species Eucalyptus sieberi L.A.S.Johnson. (silvertop ash) were assessed usingDNA markers. Three different silviculturaltreatments were examined: clear-felling withaerial re-sowing, and the seed tree system withsite preparation by either burning ormechanical disturbance. In addition, twounharvested stands were chosen as controls. Atotal sample of 825 trees were genotyped at 35Mendelian markers: 26 single-copy nuclear RFLPsand 9 microsatellites. No significantdifferences were found among the treatments inany of four population genetic statistics:allelic richness, effective number of alleles,expected heterozygosity and the panmictic index(f). Rare alleles were prevalent, and a MonteCarlo simulation showed that the apparent lossof four rare alleles from the saplingregenerants was highly statisticallysignificant. There was no evidence for recentbottlenecks from analyses of either the levelsof expected heterozygosity relative to thatexpected under mutation drift equilibrium, orthe allele frequency profiles. A dendrogram ofthe relationships between the sampledpopulations suggested that the seed tree systemmay result in the promotion of genetic drift(slight expansion of the dendrogram) whileaerial re-sowing of clear falls with the sameseedlot will lead to genetic homogenisation(contraction of the dendrogram). The apparentgenetic robustness of E. sieberi tonative forest regeneration practices isattributed to its local abundance combined withthe favourable properties of its reproductivebiology, as well as to the limitation that onlya single rotation was examined.

A genetic linkage map containing potential candidate loci for wood, fibre and floral traits has been constructed for Eucalyptus globulus (Labill.) based on the segregation of 249 codominant loci in an outbred F(1) population of 148 individuals. The map contains 204 RFLP loci, including 31 cambium-specific expressed sequence tags (ESTs) and 14 known function genes, and 40 microsatellite and five isozyme loci. Independent male and female maps were constructed, and the 98 loci (39%) that segregated in both parents were used to combine the parental maps into an integrated map. The 249 loci mapped to 11 major linkage groups ( n=11 in eucalypts) and a 12th small linkage group containing three loci that segregated in the male parent only. Total map distance is 1375 cM with an average interval of 6 cM. Forty one of the mapped loci identify known proteins (five isozymes) or sequences with known function (14 genes and 22 ESTs). The mapped genes include enzymes involved in lignin and cell-wall polysaccharide biosynthesis, and floral-development genes. This map will be used to locate quantitative trait loci for wood, fibre, and other traits in Eucalyptus.

Eight dinucleotide microsatellites were developed in Eucalyptus sieberi L. Johnson (silvertop ash), a member of the subgenus Eucalyptus. Transfer of six of these to the subgenus Symphyomyrtus and their Mendelian inheritance are demonstrated using a full-sib cross in Eucalyptus nitens. Genetic diversity parameters are presented for the eight loci based on a sample of 100 old-growth E. sieberi trees from a single natural stand. One locus, Es266, had an atypically high fixation index, and significantly deviated from Hardy-Weinberg equilibrium genotypic proportions, indicating the likely presence of null alleles. Two of the loci, Es076 and Es140, had many alleles that differed in size by only a single base pair, possibly because of short poly(A) or poly(T) stretches in their flanking regions. These two loci were by far the most polymorphic, but were difficult to score reliably on a capillary DNA sequencer. Reliability of scoring of these two one-base microsatellite loci was markedly improved by the incorporation of internal reference alleles into each sample analysed.Key words: SSRs, single base pair alleles, null alleles, internal reference alleles.

An integrated genetic linkage map, comprised of 219 RFLP and 33 microsatellite loci in 13 linkage groups, was constructed using two outbred pedigrees of Acacia mangium Willd. The linkage groups ranged in size from 23 to 103 cM and the total map length was 966 cM. Individual maps were made for each pedigree and the ordering of loci was consistent with the integrated map. The use of two independent pedigrees allowed a comparison of recombination rates between linked loci in male and female meioses as well as between parents. Differences were confined to specific regions and were not uniform across the male and female genomes or between genotypes. The heterogeneity in recombination frequencies did not result in major differences in the ordering of loci between pedigrees; hence, the integrated map provides a sound basis for QTL detection, leading to marker-assisted selection in A. mangium. It also provides a reference map for comparative genome analysis in acacias. The co-dominant markers used for mapping provide a useful resource in population studies and for quality control in acacia breeding programs. Detection of a relatively high proportion of selfs in pods derived from flowers which were not emasculated (30%), compared with emasculated flowers (0.01%), indicates that emasculation is desirable for efficient delivery of control-crossed seed in acacia breeding programs.

Random genomic probes were used to assess levels of restriction fragment length polymorphism (RFLP) in two 2-generation outbred pedigrees of Acacia mangium Willd. Probes were evaluated for their ability to detect polymorphic loci in each pedigree and to determine the relative efficiency of different restriction enzymes in revealing polymorphisms. Sixty two percent of the probes which detected single- or low-copy number sequences revealed polymorphisms with at least one restriction enyzme. HpaII was the most efficient in detecting polymorphism among first-generation individuals. The recognition sequence of HpaII contains a CpG dimer, suggesting that cytosines in the CpG sequence may be hotspots for mutation in plant genomes, as previously reported in bacterial and mammalian genomes. Mendelian inheritance of 230 loci was demonstrated based on single-locus segregation in second-generation individuals. Less than 5% of loci showed evidence of segregation distortion. The proportion of fully informative loci (15%) was lower than previously reported in eucalypts reflecting the lower level of genetic diversity in A. mangium. The RFLP probes are suitable for the construction of a high-density genetic linkage map in A. mangium. Cross-hybridisation of the A.mangium RFLPs to DNA from species representing the three subgenera of the genus Acacia indicates that these markers could be used in breeding programs of other diploid acacias, for comparative studies of genome organisation, and for phylogenetic studies.

Genetic diversity in the nuclear genome of Eucalyptus nitens was assessed using 40 anonymous RFLP loci. Information about the RFLP probes was available from the genetic linkage map of the species and probes that detected single loci were selected to cover the whole of the genome. Several loci that were monomorphic in the mapping pedigree were also used to avoid bias towards polymorphic loci. Variation was determined for 20 individuals from each of eight populations covering the natural range of the species. The proportion of rare alleles detected in populations ranged from 21% to 49%. The level of diversity within the nuclear genome was high (AS = 6.9, PS = 92.5%, HT = 0.445) compared with other species where RFLP surveys have been conducted. The level of variation detected was higher for RFLP loci than for isozyme loci, but the patterns of genetic diversity were similar for both marker types. Population similarities based on RFLP analysis of the nuclear genome were not concordant with those based on the chloroplast genome.

Genetic diversity in the nuclear genome of Acacia mangium was estimated using 57 anonymous RFLP loci for 10 individuals from each of 10 natural populations representing the geographical range of the species. The level of genetic diversity varied significantly among the populations, ranging from HE=0.01 on the island of Ceram to HE=0.21 in Muting, New Guinea. The small, geographically isolated populations of Daintree, Townsville, Ceram and Sidei had low levels of diversity (HE=0.01–0.09) whereas the large New Guinea and the Cape York Peninsula populations had higher levels of diversity (HE=0.16–0.21). There was evidence of genetic differentiation between populations (θ=0.35), 75% of which was attributable to differences between four geographical regions of Cape York and Daintree–Townsville in Australia, New Guinea and Ceram–Sidei. This may reflect restrictions to gene flow associated with past expansions and contractions of rainforests which occurred with climate and sea level changes. The level of variation detected with RFLPs was higher than previously detected with allozymes, supporting speculation that allozyme coding regions are more likely to be constrained by selection than noncoding regions. The distribution of variation within and among populations was consistent for the two marker types.

Regions of the genome influencing frost tolerance in an outbred family of Eucalyptus nitens have been identified. Two QTLs present on the same linkage group, but located 40 cM apart, were identified using single-factor analysis of variance. The QTLs explained between 7.7 and 10.8% of the phenotypic variation for frost tolerance in this family. Analysis of marker loci linked to the QTLs showed one of them to have a simple mode of action with the effect segregating from the male parent in the family. For the other QTL multiple alleles were identified. This QTL showed segregation from the female parent which gave a positive effect on frost tolerance; however, an allele segregating from the male parent was identified which showed a negative interaction with the allele for increased frost tolerance.

Regions of the genome influencing height and leaf area in seedlings of a three-generation outbred pedigree of Eucalyptus nitens have been identified. Three QTLs affecting height and two QTLs affecting leaf area were located using single-factor analysis of variance. The three QTLs affecting height each explained between 10.3 and 14.7% of the phenotypic variance, while the two QTLs for leaf area each explained between 9.8 and 11.6% of the phenotypic variation. Analysis of fully informative marker loci linked to the QTLs enabled the mode of action of the QTLs to be investigated. For three loci the QTL effect segregated from only one parent, while for two loci the QTL showed multiple alleles and the effect segregated from both parents in the pedigree. The two QTLs affecting leaf area were located in the same regions as two of the QTLs affecting height. Analysis of these regions with fully informative markers showed that both QTLs were linked to the same markers, but one had a similar size of effects and a similar mode of action for both height and leaf area, whilst the other showed a different mode of action for the two traits. These regions may contain two closely linked genes or may involve a single gene with a pleiotrophic effect on both height and leaf area. The QTL with the greatest effect showed multiple alleles and an intra-locus interaction that reduced the size of the effect. Assessment for two of the QTLs in a second related family did not show an effect associated with the marker loci; however, this was consistent with the mode of action of these QTLs and the pattern of inheritance in the second family.

A genetic linkage map for radiata pine (Pinus radiata D. Don) has been constructed using segregation data from a three-generation outbred pedigree. A total of 208 loci were analyzed including 165 restriction fragment length polymorphism (RFLP), 41 random amplified polymorphic DNA (RAPD) and 2 microsatellite markers. The markers were assembled into 22 linkage groups of 2 or more loci and covered a total distance of 1382 cM. Thirteen loci were unlinked to any other marker. Of the RFLP loci that were mapped, 93 were detected by loblolly pine (P. taeda L.) cDNA probes that had been previously mapped or evaluated in that species. The remaining 72 RFLP loci were detected by radiata pine probes from a PstI genomic DNA library. Two hundred and eighty RAPD primers were evaluated, and 41 loci which were segregating in a 1∶1 ratio were mapped. Two microsatellite markers were also placed on the map. This map and the markers derived from it will have wide applicability to genetic studies in P. radiata and other pine species.

The level of polymorphism using genomic and cDNA probes with a number of restriction enzymes and the inheritance of the RFLP loci was investigated in E. nitens. The polymorphism detected with 366 genomic and cDNA probes and three to six restriction enzymes was analysed in three-generation outbred pedigrees. No difference in the level of polymorphism detected with genomic versus cDNA probes was observed. There was a difference in the efficiency of detection of polymorphism with six different restriction enzymes, with three of the enzymes (BglII, DraI and EcoRI) showing substantially more polymorphism than the others. There was no significant correlation between the size of the DNA fragments generated by the enzymes and the detection of polymorphism. Several cases of restriction-site mutations resulting in a polymorphism were observed. The inheritance of 69 loci was analysed in two pedigrees resulting from interpopulational crosses. The majority of the loci segregated according to expected ratios with distortion observed in only 3% of loci. Probes from the cDNA library detected a greater proportion of loci with more than two alleles than did probes from the genomic library. The high polymorphism, large number of alleles, and ease of interpretation of RFLPs in E. nitens means that they will be useful in a range of applications such as genetic linkage maps and paternity analysis.

Variation in the chloroplast genome of Eucalyptus nitens was assessed for ten individuals from each of eight populations covering the natural range of the species. Twenty-five mutations were detected which were distributed over 13 haplotypes. The mutations were present both within and between populations. The level of nucleotide diversity within the species was high and the majority of the variation was distributed between populations and regions. The level and distribution of haplotype diversity were similar to those of the nucleotide diversity. The interpopulational variation may be related to the large range and disjunct populations that characterize the species, due to the effects of isolation and genetic drift. There were three mutations which were unique and fixed in one small population. Individuals in the species could be distinguished into two cpDNA groups. The distribution of these groups was not consistent with any geographical pattern. The distribution of the cpDNA groups suggests that they are of ancient origin and predate the isolation of the regions. The pattern of cpDNA variation is not consistent with patterns of variation in morphological traits.

Eucalyptus is characterized by high foliar concentrations of plant secondary metabolites with marked qualitative and quantitative variation within a single species. Secondary metabolites in eucalypts are important mediators of a diverse community of herbivores. We used a candidate gene approach to investigate genetic associations between 195 single nucleotide polymorphisms (SNPs) from 24 candidate genes and 33 traits related to secondary metabolites in the Tasmanian Blue Gum (Eucalyptus globulus). We discovered 37 significant associations (false discovery rate (FDR) Q < 0.05) across 11 candidate genes and 19 traits. The effects of SNPs on phenotypic variation were within the expected range (0.018 < r(2) < 0.061) for forest trees. Whereas most marker effects were nonadditive, two alleles from two consecutive genes in the methylerythritol phosphate pathway (MEP) showed additive effects. This study successfully links allelic variants to ecologically important phenotypes which can have a large impact on the entire community. It is one of very few studies to identify the genetic variants of a foundation tree that influences ecosystem function.

The 5S RNA genes inPinus radiata occur in two size classes of about 850 and 525 bp in length. Representatives from both the long and short size classes were cloned and sequenced. The primary difference in the two size classes was shown to be a 330 bp insertion in the spacer region of the long units. Within an individual breeding clone ofP. radiata there was some sequence heterogeneity between representatives of the short class. The 5S RNA genes in thirty pine species were characterised using either a clone of the short size class or a subclone of the 330 bp insertion characterizing the long size class. Eleven species of subg.Strobus were examined and all lacked the long type of unit of radiata pine. The New World species of subg.Pinus all had both short and long units whereas the Old World species had long units. The implications of these results for the evolution of the 5S DNA sequences and the phylogenetic relationships withinPinus are discussed.

Background There is little information about the DNA sequence variation within and between closely related plant species. The combination of re-sequencing technologies, large-scale DNA pools and availability of reference gene sequences allowed the extensive characterisation of single nucleotide polymorphisms (SNPs) in genes of four biosynthetic pathways leading to the formation of ecologically relevant secondary metabolites in Eucalyptus. With this approach the occurrence and patterns of SNP variation for a set of genes can be compared across different species from the same genus. Results In a single GS-FLX run, we sequenced over 103 Mbp and assembled them to approximately 50 kbp of reference sequences. An average sequencing depth of 315 reads per nucleotide site was achieved for all four eucalypt species, Eucalyptus globulus, E. nitens, E. camaldulensis and E. loxophleba. We sequenced 23 genes from 1,764 individuals and discovered 8,631 SNPs across the species, with about 1.5 times as many SNPs per kbp in the introns compared to exons. The exons of the two closely related species (E. globulus and E. nitens) had similar numbers of SNPs at synonymous and non-synonymous sites. These species also had similar levels of SNP diversity, whereas E. camaldulensis and E. loxophleba had much higher SNP diversity. Neither the pathway nor the position in the pathway influenced gene diversity. The four species share between 20 and 43% of the SNPs in these genes. Conclusion By using conservative statistical detection methods, we were confident about the validity of each SNP. With numerous individuals sampled over the geographical range of each species, we discovered one SNP in every 33 bp for E. nitens and one in every 31 bp in E. globulus. In contrast, the more distantly related species contained more SNPs: one in every 16 bp for E. camaldulensis and one in 17 bp for E. loxophleba, which is, to the best of our knowledge, the highest frequency of SNPs described in woody plant species.

Pollen of acacias is transported by insects as polyads, composite pollen grains. The polyad has enough pollen grains to fertilize all ovules within a flower and hence all seed within a pod may be full sibs. Isozyme markers were used to test this hypothesis in two populations of Acacia melanoxylon R.Br. The proportions of fruit pods with multiple paternity detected in two populations were 0.08 and 0.15. The proportions of fullsib pairs within pods estimated by the sibling pair method were 1 and 0.63 for the two populations. Comparison of the diploid paternal genotypes of pods of single paternity showed that the probability of a common pollen source for a pair of pods was high within globular clusters (0.35) or within inflorescences (0.46) but declined to 0.10 or 0.25 within the tree at random. Thus the reproductive system acted to reinforce a hierarchy of paternal correlation within each tree.

Summary • Leaves of eucalypt species contain a variety of plant secondary metabolites, including terpenoids and formylated phloroglucinol compounds (FPCs). Both terpene and FPC concentrations are quantitative traits that can show large variation within a population and have been shown to be heritable. The molecular genetic basis of this variation is currently unknown. • Progeny from a field trial of a three-generation mapping pedigree of Eucalyptus nitens were assayed for terpenes and FPCs. Quantitative trait loci (QTL) analyses were conducted using a map constructed from 296 markers to locate regions of the genome influencing foliar concentrations of these plant secondary compounds. • A large number of significant QTL for 14 traits were located across nine linkage groups, with significant clustering of QTL on linkage groups 7, 8 and 9. As expected, QTL for biosynthetically related compounds commonly colocated, but QTL for unrelated monterpenes and FPCs also mapped closely together. • Colocation of these QTL with mapped candidate genes from the various biosynthetic pathways, and subsequent use of these genes in association mapping, will assist in determining the causes of variation in plant secondary metabolites in eucalypts.

Linkage disequilibrium (LD) mapping using natural populations results in higher resolution of marker-trait associations compared to family-based quantitative trait locus (QTL) studies. Depending on the extent of LD, it is possible to identify alleles within candidate genes associated with a trait. Analysis of a natural mutant in Arabidopsis has shown that mutations in cinnamoyl CoA reductase (CCR), a key lignin gene, affect physical properties of the secondary cell wall such as stiffness and strength. Using this gene, we tested whether LD mapping could identify alleles associated with microfibril angle (MFA), a wood quality trait affecting stiffness and strength of wood. We identified 25 common single-nucleotide polymorphism (SNP) markers in the CCR gene in Eucalyptus nitens. Using single-marker and haplotype analyses in 290 trees from a E. nitens natural population, two haplotypes significantly associated with MFA were found. These results were confirmed in two full-sib families of E. nitens and Eucalyptus globulus. In an effort to understand the functional significance of the SNP markers, we sequenced the cDNA clones and identified an alternatively spliced variant from the significant haplotype region. This study demonstrates that LD mapping can be used to identify alleles associated with wood quality traits in natural populations of trees.

Regions of the genome influencing wood and fibre traits in Eucalyptus globulus Labill. have been identified in two full-sib pedigrees that share a common male parent. The first pedigree, cross A, contains 148 progeny, and the second pedigree, cross B, contains 135 progeny. Subsets of progeny of these two controlled crosses were planted at seven sites throughout Australia in 1990. Wood cores were taken at 0.9 m above ground in 1997, and wood and fibre traits were analysed for each individual. Three quantitative trait loci (QTL) affecting wood density, one QTL affecting pulp yield and one QTL affecting microfibril angle have been located in both pedigrees, using single-factor analysis of variance. Other QTLs affecting these traits, as well as fibre length and cellulose content, were located in cross A only.

Six related radiata pine ( Pinus radiata) full-sib families were used to detect and independently verify quantitative trait loci (QTLs) for resistance to Dothistroma needle blight, caused by Dothistroma septospora. The detection families had from 26 to 30 individuals each, and had either a common maternal (31053) or paternal (31032) parent; one family (cross 4) consisted of progeny from both parents, 31053 x 31032. Approximately 200 additional progeny from cross 4 were clonally replicated and planted at two sites, with at least five to seven ramets of each individual per site. Marker segregation data were collected from a total of 250 RFLP and microsatellite markers, and single factor ANOVAs were conducted separately for each family and marker. A number of putative associations were observed, some across more than one family. Permutation tests were used to confirm expected probabilities of multiple associations based on chance alone. Seven markers representing at least four QTLs for resistance to Dothistroma were identified as being significant in more than one family; one of these was significant at P0.05 in three families and highly significant at P0.01 in a fourth. Further confirmation was obtained by testing those markers that were significant in more than one of the detection families (or highly significant in cross 4) in the clonally replicated progeny from cross 4. Four QTL positions were verified in the clonal populations, with a total percent variation accounted for of 12.5.

Fifty microsatellite markers were developed and characterized in Pinus radiata, and from among these, a subset of 10 easily scored and highly polymorphic markers was selected for use in fingerprinting, quality control, and breeding applications. The markers were characterized based on reliable and reproducible amplification, observed and expected heterozygosities, number of alleles, a low frequency of null alleles, and a lack of close linkage with other selected markers. Allele numbers and frequencies were estimated using 24 first-generation breeding clones from Australia and New Zealand. Observed heterozygosities for the selected markers were all greater than 0.67, and there was an average of 10.5 alleles/locus. The occurrence of null alleles was checked with megagametophytes from mother trees for loci that appeared to be homozygous. The 10 markers are not closely linked (r < 0.20 and LOD > 3) to each other. The selected microsatellites fall into three discrete size classes, and with appropriate selection of fluorescent dyes for 5' end labeling, can be multiplexed with up to 6 markers/sample on an ABI PRISM 310 or similar instrument.Key words: pine microsatellites, multiplexing, null alleles.

International audience; A major focus of our research has been on using molecular technologies to guide breeding for high value wood and fibre traits in eucalypts. One approach has been to use genomic maps to locate and characterise QTL that control wood and fibre traits. A generic map for eucalypts has been developed that consists of codominant markers that can be assayed across the major commercial species. Also on the map are candidate genes for traits such as flowering and wood and fibre traits. QTL have been characterised for wood density, fibre length, pulp yield and microfibril angle. A number of these have been validated in a related pedigree. A subset of QTL colocates with candidate genes. Research is now focussed on DNA sequence variation in different parts of these candidate genes and determining if this variation is associated with variation in wood properties. Of particular interest are genes involved in fibre development and cell wall formation. A third approach is to profile the expression of genes in the tissue of interest and determine if this relates to variation in trait performance. Microarrays are being used to relate expression of several thousand genes in xylem tissue to variation in traits such as microfibril angle and density.; Génomique des critères de qualité du bois d'eucalyptus. Notre recherche a été concentrée sur l'utilisation de technologies moléculaires pour guider l'amélioration génétique de critères importants de la qualité du bois et des fibres. On a développé une carte générique pour les eucalyptus à partir de marqueurs co-dominants intéressant les principales espèces d'intérêt commercial. Figurent également sur cette carte des gènes candidats concernant des caractères tels que la floraison et la qualité du bois et des fibres. On a mis en évidence des QTL pour la densité du bois, la longueur des fibres, le rendement en pâte et l'angle des microfibrilles. Un certain nombre ont été validés sur des familles apparentées à généalogie connue. Un sous ensemble de QTL sont co-localisés avec des gènes candidats. On concentre maintenant les recherches sur la variabilité de séquences ADN de ces gènes candidats pour savoir si elle est associée à celle des critères de qualité du bois. Sont d'un intérêt particulier les gènes impliqués dans le développement des fibres et la formation des parois cellulaires. Une troisième approche consiste à profiler l'expression des gènes dans les tissus en cause et à rechercher s'il existe d'éventuelles relations avec la variabilité des caractères. On utilise des microréseaux pour relier l'expression de plusieurs milliers de gènes du xylème avec la variabilité de caractères tels que l'angle des microfibrilles et la densité du bois.

Eucalyptus camaldulensis Dehnh. is the most widely planted eucalypt in the tropics. Natural populations are riparian and sampling strategies for breeding programmes have assumed that gene flow among drainage basins is limited. RFLP variation, within and among 31 populations from river systems across northern Australia, was analysed to test this hypothesis. To allow comparisons within and between river systems, trees were sampled from up to three populations per river system. Allele frequencies were correlated with longitude for more than half the 33 RFLP loci surveyed. Genetic identity was greatest between populations in closest geographic proximity, irrespective of river system, suggesting that sampling strategies for breeding programmes should be based on geographic distance rather than river system. The level of genetic variation was similar throughout the geographic range examined (mean H(E) = 0.49). However, there was evidence of a barrier to gene flow between populations in the east and west of the species range. The RFLP data support morphological evidence of hybridisation between E. camaldulensis and E. tereticornis Sm. in several populations in northeast Queensland and the genetic divergence of E. camaldulensis subsp. simulata Brooker and Kleinig.

Mapping in forest trees generally relies on outbred pedigrees in which genetic segregation is the result of meiotic recombination from both parents. The currently available mapping packages are not optimal for outcrossed pedigrees as they either cannot order phase-ambiguous data or only use pairwise information when ordering loci within linkage groups. A new package, OUTMAP, has been developed for mapping codominant loci in outcrossed trees. A comparison of maps produced using linkage data from two pedigrees of Acacia mangium Willd demonstrated that the marker orders produced using OUTMAP were consistently of higher likelihood than those produced by JOINMAP. In addition, the maps were produced more efficiently, without the need for recoding data or the detailed investigation of pairwise recombination fractions which was necessary to select the optimal marker order using JOINMAP. Distances between markers often varied from those calculated by JOINMAP, resulting in an increase in the estimated genome length. OUTMAP can be used with all segregation types to determine phase and to calculate the likelihood of alternative marker orders, with a choice of three optimisation methods.

Microsatellite markers were developed in Acacia mangium Willd. to provide highly variable co-dominant markers for linkage mapping and studies of the breeding system. After an enrichment procedure 40% of colonies contained microsatellites in contrast with less than 1% from a non-enriched library. The majority of microsatellite sequences were AC repeats. Co-dominant segregation of alleles in two full-sib crosses of A. mangium was demonstrated at 33 microsatellite loci. The markers were highly variable relative to restriction fragment lengths polymorphisms (RFLPs). In the two pedigrees 53% of microsatellite loci were fully informative compared with 15% of RFLPs. Based on alleles detected among four parental genotypes, the microsatellites consisting of dinucleotide repeats were more polymorphic than those with tri- and tetra-nucleotide repeats. The microsatellite markers were not as transferable across species in the genus Acacia as RFLPs. Two thirds of the primers developed in A. mangium (subgenus Phyllodineae, section Juliflorae) amplified DNA from other species within the same section but failed to amplify in species from the subgenus Acacia. The availability of multiallelic, PCR-based, co-dominant microsatellite loci makes possible efficient studies of gene flow and breeding systems in A. mangium, a species with low allozyme variation.

An integrated genetic linkage map for E. nitens was constructed in an outbred three-generation pedigree. Analysis of 210 RFLP, 125 RAPD and 4 isozyme loci resulted in 330 markers linked in 12 linkage groups covering 1462 cM (n=11 in eucalypts). The 12th linkage group is comprised of only 5 markers and will probably coalesce with another linkage group when further linked loci are located. Co-dominant RFLP loci segregating in both parents were used to integrate linkages identified in the male and female parents. Differences in recombination frequencies in the two parents were observed for a number of pairs of loci, and duplication of sequences was identified both within and between linkage groups. The markers were distributed randomly across the genome except for the RFLPs in linkage group 10 and for some loci showing segregation distortion, which were clustered into three regions of the map. The use of a large number of co-dominant RFLP loci in this map enables it to be used in other pedigrees of E. nitens and forms a basis for the detection and location of QTL in E. nitens and other eucalypt species.

Polymorphisme de 30 loci etudie chez 11 populations d'Acacia mangium du Nord de l'Australie et de la Papouasie et comparaison avec la diversite genetique d'autres especes d'acacia

Variation in allele frequencies in the outcross pollen pool of Eucalyptus regnans F. Muell. throughout a flowering season

In a high-salt soluble fraction of the total protein from single seeds of Pinus radiata, up to 45 polypeptides were resolved on SDS-polyacrylamide gels. At least one-fifth of these polypeptides showed variation between seeds. In the 27,000--29,000 dalton region, two polypeptides were inherited as codominant alleles at a single locus and were shown to assort independently of another seed protein locus and three allozyme loci. A survey of 120 individuals from the five known native populations of P. radiata in California detected only the 27K and 29K alleles at the locus. In all populations, the 29K allele predominated, and the two island populations were monomorphic for the 29K allele. The 27 and 29 kdalton polypeptides were shown to have very similar amino acid sequences, and the allelic difference at this locus is most probably in the gene sequence for the polypeptide.

Despite the fact that forest trees are in early stages of domestication there has been little direct evaluation of either the origin of, or genetic diversity within the breeding material in tree improvement programs. Allozyme variation was used to compare the total genetic diversity in the breeding programs of P. radiata within Australia and the five wild populations in North America. The current breeding populations were very similar genetically and were essentially homogenous with only 1.8% of the variation among programs. The total genetic diversity in the species was 0.12, which is a low estimate compared to most conifers. Overall in the Australian material the genetic diversity was somewhat less. The comparison of allelic frequencies in the five native populations with the Australian material indicates that the Monterey and Año Nuevo populations were probably the major source of the original introductions and that a substantial portion of the genetic diversity in the two populations has been captured in current breeding programs. The three southern populations do not appear to be currently represented in the breeding programs. The implications for future breeding strategies are discussed.

Three seed crops of a Eucalyptus delegatensis population were assayed for their allozyme genotype at three loci to determine estimates of mating system parameters. In the pollen the allelic frequencies at each of the three loci were similar to those in the parents and the progeny. Overall there was a significant amount of inbreeding (23%) in the population. The levels of outcrossing in each crop were significantly different from each other indicating apparent temporal variation in outcrossing rates. The outcrossing rate was greatest in the oldest crop (85%) and lowest in the most recent crop (66%). Mean heterozygosity in the progeny of all three crops was less than the heterozygosity in the parents indicating that selection favours heterozygotes during the life cycle. The implications of a 'balanced' mixed mating system for a eucalypt breeding program are discussed.