Australian Oryza are an understudied and underexploited genetic resource for rice improvement. Four species are indigenous: Oryza rufipogon, Oryza meridionalis, Oryza australiensis are widespread across northern Australia, whereas Oryza officinalis is known from two localities only. Molecular analysis of these wild populations is required to better define the distinctness of the taxa and the extent of any gene flow between them and rice. Limited collections of these wild populations are held in seed and DNA banks. These species have potential for domestication in some cases but also have many traits of potential value in the improvement of domesticated rice. Stress tolerance (biotic and abiotic) and grain quality characteristics in these populations may be useful.

Spotted gums (genus Corymbia, section Politaria) occur as a species replacement series along the eastern seaboard of Australia, their distributions marked by regions of disjunction and sympatry. Their taxonomy remains controversial, with species assignment often challenging and reliant on knowledge of geographic origin as well as subtle morphological or leaf-oil variation. In the present paper, we explore a classification for spotted gums, without assuming predefined geographic or taxonomic groups but instead using genetic structure at microsatellite marker loci (n = 9) and a Bayesian model-based clustering approach implemented in Structure software. The C. torelliana outgroup (n = 21; section Cadagaria) formed a well resolved cluster (minimum pairwise Fst = 0.19). Four populations were evident within the spotted gums (n = 93) but structure was weak (pairwise Fst range 0.13–0.05). Geographic distance, topography and distribution disjunction were major determinants of structure, with migration among populations approximating a linear stepping-stone model. Corymbia maculata was resolved as a taxon and had the greatest genetic distance from any other population (minimum pairwise Fst 0.08). Three clusters were evident within the northern taxa but alignment with taxonomic groupings was poor. C. citriodora material from north of a major disjunction in central Queensland formed a Northern population. C. citriodora, C. variegata and C. henryi material south of this disjunction but north of the Border Range, formed a Central population, whereas a Southern population comprised C. variegata and C. henryi from predominately south of the Border Range.

The genus Corymbia is closely related to the genus Eucalyptus, and like Eucalyptus contains tree species that are important for sub-tropical forestry. Corymbia’s close relationship with Eucalyptus suggests genetic studies in Corymbia should benefit from transfer of genetic information from its more intensively studied relatives. Here we report a genetic map for Corymbia spp. based on microsatellite markers identified de novo in Corymbia sp or transferred from Eucalyptus. A framework consensus map was generated from an outbred F2 population (n = 90) created by crossing two unrelated Corymbia torelliana × C. citriodora subsp. variegata F1 trees. The map had a total length of 367 cM (Kosambi) and was composed of 46 microsatellite markers distributed across 13 linkage groups (LOD 3). A high proportion of Eucalyptus microsatellites (90%) transferred to Corymbia. Comparative analysis between the Corymbia map and a published Eucalyptus map identified eight homeologous linkage groups in Corymbia with 13 markers mapping on one or both maps. Further comparative analysis was limited by low power to detect linkage due to low genome coverage in Corymbia, however, there was no convincing evidence for chromosomal structural differences because instances of non-synteny were associated with large distances on the Eucalyptus map. Segregation distortion was primarily restricted to a single linkage group and due to a deficit of hybrid genotypes, suggesting that hybrid inviability was one factor shaping the genetic composition of the F2 population in this inter-subgeneric hybrid. The conservation of microsatellite loci and synteny between Corymbia and Eucalyptus suggests there will be substantial value in exchanging information between the two groups.

The grain morphology of 17 wild rice relatives were studied by light and scanning electron microscopy and compared to two cultivated rice varieties (Oryza sativa cv. Nipponbare and O. sativa cv. Teqing). Observations were made of the grain colour, size and shape. Grains from wild rice species exhibited a variety of colours that have potential aesthetic and nutritional value. The grains of these species exhibited a wide array of sizes and shapes, but still fell within the standard classification scale that rice breeders use for routine breeding evaluation. These results highlight the potential of these species as whole grain foods or as sources of novel alleles in conventional rice breeding programmes.

This chapter discusses the approaches that have been used in the isolation of DNA from plant tissues. A summary of approaches is presented in the appendix.

Quantitative trait loci (QTL) detection was carried out for adventitious rooting and associated propagation traits in a second-generation outbred Corymbia torelliana × Corymbia citriodora subspecies variegata hybrid family (n=186). The parental species of this cross are divergent in their capacity to develop roots adventitiously on stem cuttings and their propensity to form lignotubers. For the ten traits studied, there was one or two QTL detected, with some QTL explaining large amounts of phenotypic variation (e.g. 66% for one QTL for percentage rooting), suggesting that major effects influence rooting in this cross. Collocation of QTL for many strongly genetically correlated rooting traits to a single region on linkage group 12 suggested pleiotropy. A three locus model was most parsimonious for linkage group 12, however, as differences in QTL position and lower genetic correlations suggested separate loci for each of the traits of shoot production and root initiation. Species differences were thought to be the major source of phenotypic variation for some rooting rate and root quality traits because of the major QTL effects and up to 59-fold larger homospecific deviations (attributed to species differences) relative to heterospecific deviations (attributed to standing variation within species) evident at some QTL for these traits. A large homospecific/heterospecific ratio at major QTL suggested that the gene action evident in one cross may be indicative of gene action more broadly in hybrids between these species for some traits.

While cultivated rice, Oryza sativa, is arguably the world's most important cereal crop, there is little comparative morphological information available for the grain of rice wild relatives. In this study, the endosperm of 16 rice wild relatives were compared to O. sativa subspecies indica and O. sativa subspecies japonica using scanning electron microscopy. Although the aleurone, starch granules, protein bodies and endosperm cell shapes of the cultivated and non-cultivated species were similar, several differences were observed. The starch granules of some wild species had internal channels that have not been reported in cultivated rice. Oryza longiglumis, Microlaena stipoides and Potamophila parviflora, had an aleurone that was only one-cell thick in contrast to the multiple cell layers observed in the aleurone of the remaining Oryza species. The similarity of the endosperm morphology of undomesticated species with cultivated rice suggests that some wild species may have similar functional properties. Obtaining a better understanding of the wild rice species grain ultrastructure will assist in identifying potential opportunities for development of these wild species as new cultivated crops or for their inclusion in plant improvement programmes.

No abstract