Application of molecular marker technologies for genetic improvement of forest plantation species in Indonesia and Australia

FST/2000/122: Application of molecular marker technologies for genetic improvement of forest plantation species in Indonesia and Australia

Indonesia

CSIRO Forestry and Forest Products, Australia

Dr Gavin Moran
Phone: 02 6281 8208
Fax: 02 6281 8233
Email: gavin.moran@csiro.au

• Gene Technics, Australia
• Gadjah Mada University, Indonesia
• Centre for Forest Biotechnology and Tree Improvement, Indonesia

$393,733

01/01/2001 - 31/12/2002

01/01/2003 - 30/06/2003

Dr John Fryer

Plantations using species of the Australian genera Eucalyptus and Acacia have been established in many parts of the tropics and subtropics because the species used are fast-growing and produce wood that can be used for fuel, pulp, roundwood and sawlogs - and are also a source of non-wood forest products such as oils and tannins. The plantations are now important components of the rural economies of many developing countries and also in Australia, where many farmers are planting eucalypts in the southern parts as an alternative to farming livestock.
Although plantation acacias in Indonesia have proved successful, their growth rates have been well below the known optimums for the species. Poor provenance selections were used in some plantations and, in others, early seed orchards were found to have high levels of inbreeding that depressed growth rates.
In Australia, trees from both genera have been improved by selective breeding, but the application of newer molecular techniques will enable further improvements in productivity to take place. Indonesia, where both tree genera occur naturally, is one of the few countries in the region with sufficient skills and facilities in the forestry sector to take advantage of the new technology, if given some assistance.

The project aimed to improve breeding programs of Acacia and Eucalyptus through collaboration to marry molecular marker technology with existing research on controlled breeding.

The project comprised six parts - three related to A. mangium in Indonesia and three to Eucalyptus in Australia. In Indonesia the team identified the parents of selected lines in open-pollinated selections, and obtained DNA fingerprints to create a database of the mother plants from which the seed collections were made. Further study enabled the scientists to assess the extent of inbreeding and gene flow in seed orchards. The team aimed to establish mini-breeding arboreta based on improved orchard design and using better quality genetic stock.
In Australia the research concentrated on characterising quantitative trait loci (QTL) for certain important traits in Eucalyptus globulus and E. grandis. (QTLs are the regions on the chromosome that code for particular features, such as rooting ability or frost tolerance.) The QTLs were verified by laboratory, glasshouse and field trials, as a basis for selection of germplasm to improve the quality of outputs from breeding programs, prior to selecting germplasm for incorporation into these into a crossing program to create superior breeding lines.

CSIRO had developed DNA marker technologies in acacias prior to this project. The primers and all available information for the 33 microsatellite loci developed by CSIRO FFP were transferred to the molecular laboratory at CFBTI. Optimisation of microsatellite assays was carried out at CFBTI with assistance of CSIRO personnel and 16 primers are now optimised to work in the CFBTI laboratory. A key output was training of CFBTI staff in CSIRO laboratories to efficiently carry out microsatellite screening and to competently perform genetic analysis of molecular marker data.

A first generation open pollinated progeny/provenance trial was assessed for growth and form and plus trees selected. Open pollinated seed from the plus trees in subline D were planted in a 2nd generation progeny trial assessed for growth at 2 years of age. Two hundred and fif-seven trees, including the plus tree selections, were DNA profiled by determining their multilocus genotypes at 15 microsatellite loci. All trees could be uniquely distinguished from one another. Microsatellite markers are sufficiently powerful to enable efficient monitoring and quality control in Indonesian breeding programs of A. mangium.

Progeny from each of 10 plus trees were assayed at the same 15 loci, and parentage analysis carried out to determine the unknown male parent. About half of the progeny could be assigned both parents with some confidence. There was no significant relationship between degree of genetic relatedness of parents and the performance of the progeny. Progeny could be classified into groups due to interfamily, inter provenance and self crossing. There were no differences between these groups in diameter growth but some significant differences were present in form. It also highlights that these methods can allow classification of breeding material into structures previously unavailable in acacias and most tree breeding programs. This is a very exciting result. However, there appeared to be very significant errors in progenies sampled which were largely not due to laboratory mis-scoring. This result if found to be more broadly applicable to genetic trials has serious implications for the breeding programs of A. mangium in Indonesia and suggests that CFBTI can play a central role in monitoring breeding programs.

A provenance field trial in South Sumatra was used to determine genetic diversity levels in 9 provenances covering the main geographic range of A. mangium, and the land race from Subanjeriji. For each provenance 30 randomly selected trees were assayed for their microsatellite genotypes at 12 loci and population genetic parameters estimated. Heterozygosities were very high with an average He of 0.66 across all trees but with marked differences in levels of diversity between geographic regions. The land race had a very low level of observed heterozygosity with evidence of significant inbreeding. A major objective was to test whether genetic diversity parameters at the individual tree level and the provenance level were associated with growth. For 15-month growth data only means at the provenance level were made available whereas 5 year data were available at the individual tree level. At the provenance level regressions of heterozygosities with mean heights and mean diameters of 15 month data were significant. However the provenance means for 5 year diameter data were not significantly related to heterozygosities. Similarly, at the individual tree level 5 year diameter growth was not significantly associated with observed heterozygosity. The fifteen month results are suggestive that early selection in the nursery based on levels of observed heterozygosities may be possible. Nevertheless, selection has to be at the individual tree level so further research is needed to firmly establish this relation for individual trees. Analysis of growth data suggests some uncertainty as to whether the sampling was sufficiently random in selecting trees for this study.

In acacias it has proved extremely difficult to generate large numbers of seed from control pollinations. An alternative approach was to be tested in which minibreeding arboreta were used to generate open-pollinated progeny involving a limited number of parents but using multiple clonal copies of each parent. The seed could be assigned full parentage using microsatellite analysis. In the second half of 2001, two pollarding experiments were carried out on 2-year-old operational plants to generate shoot cuttings. In both experiments viable cuttings were not propagated. The need to clonally propagate plus tree selections in breeding programs for a variety of reasons makes it essential that the clonal propagation issue be resolved as soon as possible. The development of minibreeding arboreta in this project did not proceed because of lack of success at macropropagation .

Leaf rust, Atelocauda digitata, infected young 2nd generation progeny trials in South Sumatra in early 2001. The progeny tests of subline D and F were assessed for rust susceptibility to investigate the genetic control of rust resistance. Differences between families in leaf rust susceptibility were highly significant mainly due to one very susceptible family from the Subanjeriji land race. In subline F the differences between families was not significant and heritability was very low at 0.02. With such low heritabilities QTL detection will be very difficult and not recommended and phenotypic selection, preferably based on clonal trials, would be a better approach in the breeding program.

A genetic linkage map for the hybrid eucalypt pedigree was constructed with 117 progeny and 125 markers. The map consisted of 11 linkage groups and a total size of 1100cM with an average of 11 markers per linkage group. The map shows close agreement with our previously published maps of E. globulus and E. nitens. There was evidence of significant segregation distortion on parts of two linkage groups perhaps due to mating incompatibility between the species. In this pedigree the overall percentage rooting for stem cuttings was low at 22.7% probably reflecting the globulus background in the pedigree. QTL analysis consisted of single factor analysis of variance for each locus and was carried out for mean rooting across all three settings. Frost tolerance was assessed using an electrical conductivity measure and ANOVA analyses were on both the first assessment alone and means across all three assessments. Six QTL regions for rooting of stem cuttings were located on the genetic map and two of these appeared to be the same QTL as previously reported in E. nitens. For frost tolerance five QTL regions on the map were found. The strongest QTL was on linkage group one and had not previously been characterised. Two QTL for frost tolerance appeared to be the same as reported in E.nitens. Hopefully further genomics research will collocate genes to these QTL and allow development of methods to more efficiently select for these traits in breeding programs.

Cloned progeny from the two pedigrees were planted in a field trial near Deniliquin in NSW in 2002. When old enough they will be assessed for commercial traits and a combined selection strategy developed for rooting, frost and the commercial traits.