School of Agriculture, Food and Ecosystem Sciences - Theses
Permanent URI for this collection
Search Results
1 - 5 of 5
-
ItemImpregnation of wood with stains( 2007)
-
ItemA study of the creep performance of microwave-modified radiata pine in different external environments( 2007)A new innovative timber treatment developed at the CRC Wood Innovations, which involves high intensity microwave application and resins impregnation, is intended to provide products for a range of applications including structural engineering members. This study has been undertaken to obtain experimental data and provide better understanding of the creep performance and long-term behaviour of the new products. In the experiment, fifteen samples, five untreated, five treated with MUF resin and five treated with Isocyanate resin were loaded in four-point bending at 30 percent of the matched samples' failing stress, in a protected external environment in Brisbane, Australia for a period of nine months to date. The treated samples were found to produce lower relative creep deformations than the untreated sample. The samples treated with MUF resin showed better creep resistance than the samples treated with Isocyanate resin. Data from the first 90 days were used to obtain parameters for the two chosen models: the power law model and the 5- parameter model. While both models provided good fitting for the data, the 5-element model was found to possess better extrapolation capacity beyond the regression period. An increase in the period of regression data from 90 days to 150 days significantly lowered the errors in both of the models.
-
ItemTowards association studies in Pinus radiata D.Don - populations and wood property candidate-genes( 2006)In Australia and New Zealand Pinus radiata D. Don wood quality is receiving increasing interest from tree breeders. This is partly due to declining resource wood quality associated with more advanced generation breeds leading to increased rejection and product downgrading in processing. While log segregation and wood grading at mill-door yields immediate benefits to processors the underlying cause is not addressed. The only long-term solution is to include wood quality in breeding programs. Wood quality traits are costly and difficult to measure. Marker-assisted selection offers a potential solution and quantitative trait loci (QTL) mapping studies have been undertaken with the aim of facilitating this. It is becoming increasingly clear that in widely outcrossing species with long generation times and very large genomes these approaches will not work. The identification of the underlying genetic sites, or tightly linked marker sites, would rectify this and in model species linkage-map based cloning has been used extensively. This approach is also not practical in species such as P. radiata. Association testing combined with a candidate-gene approach is therefore widely believed to be one of the only methods remaining. This approach uses a priori information to select and then test the phenotypic effects of variants within candidate-gene loci. Implementation of these studies relies heavily on the results of other investigations, especially those that generate DNA sequence information. Also required is detailed knowledge of the genetic population structure, the patterns of nucleotide diversity and the patterns of linkage disequilibrium. On a more practical level suitable populations need to be identified while the current methods for the collection and handling of samples for molecular investigations are limiting. The selection of candidate genes is also a non-trivial process. For candidate-gene association studies to be successful in P. radiata all these factors need to be addressed. This formed the main aim of this thesis. A multi-pronged approach was used. Firstly, at the population level, the genetic resources available for association studies were identified and the underlying genetic population structure of these resources and the patterns of nucleotide diversity and linkage disequilibrium were investigated. Secondly, improved methods for the collection and isolation of genomic DNA were developed and thirdly, a small set of wood quality candidate-genes were selected and further characterised with the aim of identifying those with the most promise of harbouring causative variation for inclusion in future association studies. This was achieved by literature based review, linkage mapping onto wood property QTL maps and neutrality testing. Results include support for previous population genetic studies showing P. radiata to have a complex genetic structure compared to most pine species. This study also indicated significant levels of migration between the three mainland populations. Within the candidate-genes two, cinnamyl alcohol dehydrogenase and sucrose synthase, showed interesting patterns of population differentiation and/or nucleotide diversity while the results for one other gene, korrigan, did not agree with previous investigations.
-
Item
-
ItemGrowth and yield models for South Australian radiata pine plantations: incorporating fertilising and thinning( 2001)This thesis describes the development of models to predict the volume growth response of South Australian radiata pine plantations to the interaction of the silvicultural tools of thinning and fertiliser used in combination. Some years ago this issue was identified as the component of the ForestrySA yield regulation system most in need of addressing and as a result a large thinning and fertiliser experiment was established. This was designed to determine whether a thinning and fertiliser interaction existed and to enable this interaction to be modelled. At the time it was established it was believed to be the only experiment of its kind in the world and this still appears to be the case. The thinning and fertiliser interaction models described in this thesis were designed to integrate with the models already implemented in the ForestrySA yield regulation system so that more precise predictions of future log availability can be provided, and improved management decisions can be made. Three sets of component sub models are described which operate at a stand level to: • predict the total volume growth of the main crop between the time of fertilising and the next thinning, approximately seven years hence; • predict the total volume growth of the portion of the stand which will be thinned (known as the thinnings elect) at the next thinning, between the time of fertilising and the next thinning; • predict the annual volume growth response of the stand between the time of fertilising and the next thinning. Further research is described to identify the data sets that are likely to be required for future analysis and revision of the South Australian growth and yield models. Adopting the future research recommendations will ensure that the consideration of the financial and economic benefit of alternative silvicultural prescriptions is broadened to include a more diverse range of sites and include log and wood quality considerations.