Application of microsatellite markers to the genetic improvement of Acacia in Vietnam

Acacia plantations and downstream wood processing industries make a major contribution to the Vietnamese economy. In 2014, Vietnam exported an estimated 5.4M ton of wood chip valued at about $300M and most of this from acacia plantations. These are mainly A. mangium from open-pollinated seed and cloned natural hybrid with A. auriculiformis. Breeding programs for A. auriculiformis, A. mangium and its hybrid were started in the 1990s in Vietnam. Recently, research on polyploid Acacia has been conducted under a collaboration between the University of Tasmania (UTAS) and the Vietnamese Academy of Forest Sciences (VAFS) that aims to produce sterile and outstanding triploid (3x) genotypes. Molecular markers could facilitate the development of improved breeding methods, lead to better deployment programs and expand our knowledge of acacia genetics. Microsatellite markers (SSR) are among the most informative molecular markers. However, there is a limited number of SSR markers currently available for acacia, especially ones that works well across taxa. Therefore, the development of additional microsatellite markers that will function in A. auriculiformis, A. mangium and their interspecific hybrid is required for future research. The overall aim of this study was to develop such markers and show that they can be very useful to acacia breeding and deployment programs in Vietnam. A highly informative (probability of identity= 4.1 × 10\(^{−13}\)) and easy to use set of 16 SSR markers was developed for acacia hybrid and their parental species. The markers were optimised for assay in four multiplex mixes and used to genotype range-wide samples of A. mangium; A. auriculiformis, and putative F1 hybrids. Ten of the markers were highly polymorphic in each species and proved useful for fingerprinting, studying diversity and parentage analysis. Six of the markers were much less polymorphic and because the species did not share any alleles these provided ‘species-diagnostic’ markers. The six species-diagnostic markers, in combination with the statistical package HIest, enabled accurate allocation of genotypes to the two pure species, F1 and F2 interspecific hybrids and backcrosses with good degree of accuracy (96%-100%). Many acacia hybrids identified using morphology in ongoing selection program in Vietnam (65 out of 160 putative hybrid clones) were found to be mistaken (most were pure A. auriculiformis) using the marker set. The set of SSR markers were then applied to verify the ‘purity’ status of putatively pure A. mangium and A. auriculiformis clones in paired clonal seed orchards. Approximately 4% of the genotypes in these seed orchards were found to be either F1 hybrids or backcrosses. This shows that inter-species contamination is an issue requiring management in both pure-species and hybrid breeding of these species in Vietnam. The acacia pollen dispersal pattern was also examined in these paired clonal seed orchards with a large number of open-pollinated progeny (5,400 seedlings) using a pooling strategy. The decay in the level of hybridisation with distance followed a power function with a negative exponent. There were no differences between species and no interaction between distance and species in pollen dispersal. The restriction of most F1 hybridisation to within 100 m of species separation presents clear opportunities to manage the genetic purity of pure species seed orchard as well as to improve hybrid seed yields when that is the goal. Breeding system and seed characteristics of induced allotetraploid acacia hybrid were compared to those of their diploid progenitors, as well as growth characteristics of their progenies. Despite the fact that peak flowering differed by two months between ploidy, there was overlap in flowering time between them, showing potential for production of triploids through open pollination. However, of the 1350 seedlings analysed none were triploid. Allotetraploid acacia hybrid produced bigger seeds than those in their diploid progenitors. The outcrossing rate of allotetraploid was very low (14%) in comparison with the diploid progenitors (87%). The inbreeding depression in growth at 12 months after planting in diploid seedlings produced by selfing (33%) was greater than that in allotetraploid (17%). By investigating the inheritance of molecular markers in outcrossed progenies, we found evidence that the allotetraploid acacia hybrid has intermediate or mix model inheritance (with both disomic and tetrasomic marker inheritance) and it thus appears to behave as a segmental allotetraploid. Thus, breeding programs aiming to develop advanced generation allotetraploid acacias may have to select for fertility. Molecular markers have many uses in support of acacia breeding. They can be used to detect hybrid individuals from open-pollinated seed sources and this is useful because identification using morphology is prone to misidentification. Secondly, DNA fingerprinting is a valuable tool for paternity analysis and monitoring clonal identity as well as genetic diversity and therefore has a role in monitoring controlled crossing programs, evaluating assumptions regarding pollination in seed orchards and verification of clonal material during propagation. The study also contributes new knowledge of the breeding system, reproductive and growth characteristics of allotetraploid acacia hybrid