Science Collected Works - Theses

Permanent URI for this collection

http://hdl.handle.net/11343/304

Filters

Reset filters

Settings
Statistics
Citations
Start date: 2000 × End date: 2018 ×

Search Results

Now showing 1 - 10 of 230
  • Item
    No Preview Available
    Household vulnerabiltiy to climate change in a rural Chinese country : does the local state matter?
    Rogers, Sarah May (University of Melbourne, 2014)
    The aim of this thesis is to understand the vulnerability of rural households to climate change in China and the influence of local government on that vulnerability. To achieve this aim, two research questions are addressed: 1) in what ways are rural households vulnerable to climate change?; and 2) in what ways does local government influence rural household vulnerability to climate change? The thesis applies a sustainable livelihoods approach to examining household vulnerability and is further informed by theories of vulnerability to climate change and theories of central-local government relations in China. It posits that local governments influence the vulnerability of rural households by mediating their access to livelihood assets and activities, and by shaping their coping and adaptive strategies. The research questions are addressed through the analysis of data collected in four apple-producing villages � two wealthier and two poorer � in a case study county called Jixian. Jixian is a nationally designated poverty county that lies on the Loess Plateau in Shanxi Province. The methods used are primarily qualitative: semi-structured interviews and direct observation, supported by secondary data collection. The thesis finds that all households in all villages are vulnerable to climate change due to water insecurity, but that some are more vulnerable than others as a result of differential access to land. Households in wealthier villages have better access to land such that they are better able to cope with climatic stress than households in poorer villages. In relation to local government influence, it is the county government that has the most influence through its allocation of resources to villages. Resource allocation is done in such a way as to favour the two wealthier, more developed villages, thus amplifying existing differences in access to resources between villages. Yet the county government�s influence on vulnerability is far from absolute, as the major drivers of household vulnerability in Jixian, which are access to land and water, lie beyond its control.
  • Item
    No Preview Available
    Structural evolution and hydrocarbon play fairways of the Gippsland basin
    Power, Michael Roderic. (University of Melbourne, 2003)
  • Item
  • Item
  • Item
    Thumbnail Image
    Interactions between sea-ice and the Southern Annular Mode
    Meneghini, Belinda Clare (University of Melbourne, 2012)
  • Item
    Thumbnail Image
    Expert knowledge for conservation : tools for enhancing the quality of expert judgement
    McBride, Marissa Frances Curteis (University of Melbourne, 2013)
    Expert knowledge is a widely-used tool in conservation decision-making when empirical data is scarce and decisions are imminent. Research has demonstrated both the potential value and the potential dangers of expert knowledge, and established the benefit of using systematic, structured procedures to elicit judgments from experts which anticipate and correct for the most serious and predictable frailties of expert opinion. However, to date, only a small handful of the applications of expert knowledge in conservation draw on these established, scientifically-based procedures and guidelines. As a developing field in conservation science, this thesis addresses this gap by investigating the potential role for a new methodology for structured expert elicitation to act as a flexible, non-intensive method for conservation practitioners to use that capitalises on recent empirical findings in the literature. I explore the use of this methodology in a series of different conservation contexts, as well as broader questions related to the way in which we define expertise in expert elicitation settings, the role for validation in expert knowledge applications, and the potential to improve the way in which we select between different possible techniques for use in structured elicitation by incorporating information on their relative costs and benefits into the selection process. First, I critically review the literature relevant to the structured elicitation of expert knowledge about uncertain events and quantities. I summarise key conclusions from the literature as they pertain to the elicitation of expert knowledge for conservation, and identify the research gaps and methodological developments in elicitation techniques that the work in this thesis will aim to build on and contribute to. Second, I demonstrate the use of structured elicitation tools in an applied ecology setting, the assessment of lUCN Red List extinction status for a set of nine threatened Australian bird species. I provide a full account of the process involved in implementing a structured elicitation process, and use this example as a way to illustrate for ecologists and conservation practitioners the potential for common judgmental biases to enter into the expert judgment process. The structured elicitation process was found to be well received by experts, and the outcomes were shown to he more defensible against challenges from opposing parties than informally elicited opinions. However, the process was regarded by participants as potentially too time intensive to be worthwhile for less divisive species, and the documentation of the time and resource commitments involved highlighted the additional effort involved in conducting a formal, rather than informal, expert elicitation process. Third, I demonstrate how explicitly representing attitudes toward uncertainty can be used to improve the rigor of conservation decisions made under uncertainty. Using the expert judgments from the previous chapter I demonstrate the role for explicitly eliciting expert uncertainty by (i) highlighting the impact that expert risk preferences can have on uncertain listing decisions, and (ii) illustrating improvements in the objectivity of listing decisions that arise from the use of formal methods for linking decisions about uncertainty attitudes to the calculation of utility given the costs and benefits involved. Results show that under the levels of uncertainty typically observed in lUCN Red List expert assessments, variation in risk attitudes led to changes in listing decisions of up to two categories. This suggests that removing arbitrary attitudes to risk from-the listing decision process allows for improvements in how uncertainty in listing decisions is resolved, and illustrates the benefits of adopting a formal uncertainty framework for separating the scientific side of risk assessment from the value-based side of determining 'acceptable risk'. Fourth, I evaluate expert predictions against empirical data to investigate levels of expert accuracy and calibration in ecology. Using four case studies, I asked groups of ecological experts to predict the outcomes of previously unseen research studies in their field, and quantify their levels of uncertainty. I find that experts are overconfident in their predictions of uncertainty, while student assessors are well-calibrated. Experts possess more knowledge than students, but are poorer at providing adequately wide bounds to encompass their full uncertainty. I demonstrate the lack of relationship between performance and traditionally accepted metrics such as years of experience, self-rated expertise and number of publications. Fifth, I use expert performance evaluation to investigate the utility of a cost-benefit framework for improving the efficiency with which managers deploy funds for structured elicitation. I collected information at a threatened species assessment workshop about the costs and benefits associated with four possible areas of investment in elicitation effort: number of participating experts, degree of expert specialisation, level of expert interaction, and method for combining expert opinions. I find that targeting investment into alternative elicitation procedures provides differential returns in accuracy but that these differences are outweighed by variation in costs when evaluating relative cost-effectiveness. Techniques such as using large groups of experts and face-to-face discussion rather than only statistical aggregation led to improvements in accuracy but represented comparatively poor investments when experts are not located locally. Relative cost-efficiency also shifted considerably depending on factors such as whether in-house facilitation was available, and the level of cost involved in preparing for the elicitation Finally, I synthesise the literature on expert elicitation to provide a set of nine steps suitable for experts to take to improve the quality of their own judgments, either on their own or with the help of a facilitator. For each step, I provide an explanation of the problem and information about relevant structured elicitation procedures, and suggestions for experts on how better to construct and communicate their judgments. These guidelines do not circumvent the need for a structured elicitation process, but may assist experts striving to give well-reasoned, quality judgments to obtain the best return for the investment of their time in an elicitation exercise.
  • Item
    Thumbnail Image
    Self-incompatibility in Raphanus raphanistrum (Wild Radish) : Insights and Applications
    Oon, Joshua Koh Chee (University of Melbourne, 2013)
    Wild radish (Raphanus raphanistrum) is a self-incompatible (SI) winter annual and is a major crop weed in Australia, being present in all Australian states and territories except the Northern Territory. Control of wild radish costs an estimated A$210 million per annum and the species has recently overtaken ryegrass as the worst weed to control in Western Australia (WA), with 93% of the WA populations resistant to one or more herbicides. There is thus an urgent need for herbicide-independent control measures that can target wild radish reproduction. This thesis aims to evaluate the potential of the self-incompatibility (SI) system of R. raphanistrum as a reproductive target and uses two approaches to activate the SI response in a constitutive manner: the first uses SP11 (S-locus protein 11) molecules and the second uses stigmatic ROS (reactive oxygen species) inducers. Chapter 1 is a literature review and surveys current understanding of the SI response, with a particular focus on the Brassicaceae which are the emphasis of this thesis. The SI response is controlled by a multiallellic locus, the S-locus, and pollen rejection or acceptance is determined by the interaction of the products of two tightly-linked S-locus genes: SP11, encoding a pollen-expressed protein ligand and SRK, encoding a receptor kinase expressed by the stigma. Binding of SP11 to its cognate SRK causes an allele-specific activation of the SI response resulting in rejection of self-pollen. SP11 thus has the potential to inhibit reproduction in R. raphanistrum in a specific and non-toxic manner. However, to act as a reproductive blocker, the allele-specific action of SP11 requires that the number of S-alleles present in the Australian R. raphanistrum populations to be low - information that is currently lacking. The chapter concludes with a statement of the aims of this thesis. Chapter 2 describes the identification and characterisation of S-alleles in a wild R. raphanistrum population (ML8) from WA (Western Australia). A diallel was used to demonstrate the presence of a robust SI system and compatibility relationships in the population were determined. SRK S-domain and kinase domain sequences were amplified from genomic DNA: in particular, S-domains for three S-alleles were isolated and physically linked to their respective kinase domains. SP11 sequences were amplified from floral bud RNA and their relationship to SRKs determined. Based on SRK kinase domain sequences and a theoretical estimation, a total of 30 S-alleles was predicted to be present in the entire Australian R. raphanistrum population - a number that is considered low and renders the SP11 molecule a viable reproductive blocker candidate. Chapter 3 reports further work on characterizing the gene products of SP11 and SRK, with particular emphasis on the interaction between these two SI determinants. Recombinant SP11 produced in E. coli was able to prevent the germination of otherwise compatible pollen on R. raphanistrum stigmas in an allele-specific manner. A recombinant version of SRK that was missing the kinase domain (eSRK) was transiently expressed in tobacco (Nicotiana benthamiana) leaves and the hypervariable sub-domain (HVR) of eSRK was produced in E. coli. However, none of the recombinant eSRK or HVR interacted with SP11 in pull-down and dot-blot assays. Chapter 4 describes the effect of ROS on the SI response in R. raphanistrum. Chemicals known to induce or inhibit ROS production were used to manipulate ROS levels in the stigma. Treatment with sodium pyruvate, ascorbic acid and menandione bisulphite led to >8-fold increase in ROS stigmatic levels. Stigmas with elevated ROS levels were able to reject compatible pollen in a manner that appeared identical to the rejection of incompatible pollen. The reasons why increased stigmatic ROS levels should lead to pollen rejection are discussed. This thesis concludes with a summary of key experimental findings in light of current understanding of SI and suggests potential avenues for further research. A discussion of an integrated weed management approach incorporating reproductive blockers described in this thesis with existing control strategies is also presented.
  • Item
  • Item
    No Preview Available
    Investigating genetic and morphological variability in the holly grevillea Grevillea aquifolium (Proteaceae)
    Downing, Trisha Lee. (University of Melbourne, 2012)
    Currently composed of 20 taxa, the holly grevilleas are distributed mainly throughout south-eastern mainland Australia and are an informal group of species with holly-like leaves and toothbrush inflorescences. The holly grevilleas as a group are of high conservation value, with the majority of species being endemics with restricted distributions and the more widespread species showing considerable variation across the landscape. The most- widespread and variable of these taxa, Grevillea aquifolium Lindl., is found in western Victoria and south-eastern South Australia and was the major focus of this project. The objectives of this study were to investigate phylogenetic relationships of the holly grevilleas, document morphological and genetic variation in G. aquifolium, and to determine whether segregate taxa could be recognised. The project included phylogenetic analyses based on cpDNA sequences (trnL-trnF, rpoB-trnC, trnQ-5'rps16 intergenic spacers), and a detailed study of G. aquifolium based on nuclear DNA microsatellites and morphology. Fifty-eight accessions, including 25 accessions of G. aquifolium, were used in the multi-cpDNA phylogenetic analysis of the holly grevilleas, while 135 accessions (106 G. aquifolium) were included in the phylogeographic analysis of G. aquifolium and its allied species. Levels of genetic variation based on nuclear DNA microsatellites included 312 individuals collected from across the geographic range of G. aquifolium. One hundred and thirty herbarium specimens were examined and scored for the multivariate pattern analysis of morphological variation within G. aquifolium. It is concluded from the phylogenetic analysis that the southern holly grevilleas are a monophyletic group (Chapter 2). The southern holly grevilleas include all taxa in the group except for G. renwickiana and G. scortechinii, which are found in northern New South Wales and southern Queensland. Related to the southern holly group were two more-northerly taxa, G. renwickiana and G. scortechinii subsp. sarmentosa, along with G. willisii, a species from north-eastern Victoria not traditionally classified in the holly grevillea group. Some clear incongruence between current species limits and the combined cpDNA tree was also identified, with not all species resolved as monophyletic. Some accessions of G. aquifolium had identical chloroplast DNA sequences to those of a number of other southern holly grevillea species (e.g., G. dryophylla, G. floripendula, G. steiglitziana); this was concluded to be most likely the result of introgression and/or incomplete lineage sorting. The geographic distribution of the taxa within the southern holly grevilleas suggested a vicariant pattern for the three major sub-clades identified ('G. repens �, 'G. dryophylla' and �G. aquifolium'), although the historical changes responsible for the pattern remain a matter of speculation. The study of the variable G. aquifolium revealed that chloroplast DNA (Chapters 2-3) gives a more fine-scale rendering of the pattern of genetic variation, compared to either microsatellite data or morphology. There is geographic separation of G. aquifolium chloroplast haplotypes on a broad scale, with respect to the South Australian, Little Desert/Mt Arapiles and Grampians populations (Chapter 3). The South Australian and Little Desert/Mt Arapiles populations each formed distinct lineages, while the Grampians recorded the largest number of haplotypes. Most of the diversity within the Grampians was found in the eastern ranges (nine haplotypes), with only two chloroplast haplotypes present in the western range. The fine-scale variation found within the Grampians indicates low seed-mediated gene flow and that the seeds of G. aquifolium do not move very far from the maternal plant, a result that is congruent with seed morphology. This study also revealed the presence of a stem-loop structure within the trnQ-5'rps16 intergenic spacer, a result not previously reported for this cpDNA non-coding region (Chapter 3). The microsatellite and morphological analyses of G. aquifolium (Chapters 4-5) were in broad agreement. These data showed differentiation of coastal and inland populations (see below), but neither the microsatellite-based analyses nor the morphologically-based phenetic analyses found a discernable geographic pattern of variation or genetic structure that could be used as evidence for the recognition of new taxa. It is clear that the South Australian and south-west Victorian (i.e., Kentbruck Heath) populations of G. aquifolium are genetically distinct from those in inland western Victoria (i.e., Grampians, Little Desert/Mt Arapiles). The five-locus microsatellite Structure analysis (Chapter 4) also identified the Little Desert as a distinct genetic lineage with affiliations to the Grampians. It is clear from the molecular analyses that the Grampians are a centre of genetic diversity for this taxon. However, apart from the lower number of chloroplast haplotypes present in the western-most ranges (Victoria Range; one shared and one unique haplotype), there was no clear phylogeographic pattern within the Grampians. The level of genetic diversity within the Grampians, compared with the other major disjunct populations, supports the hypothesis that the Grampians acted as a refugium for species during the climatic fluctuations of the Miocene and/or Pleistocene. With regards to taxonomy, the results of both the molecular and morphological analysis do not support the formal recognition of any of the horticultural forms of G. aquifolium.
  • Item
    No Preview Available
    Proteomic and metabolomic responses of rice suspension culture cells to salinity
    Liu, Dawei. (University of Melbourne, 2011)
    Salinity is one of the major environmental factors affecting agriculture globally. Studying the cellular regulatory pathways which are activated by salt stress and investigating the salt-specific response in protein and metabolite levels represent feasible approaches to improving our understanding of plant salt tolerance. To investigate salt-response mechanisms in plant cells, we used rice (Oryza sativa L.) suspension cultured cells (SCCs) as a model. Stress induction was monitored by following transcription changes of known stress-responsive genes and morphological analysis of cells in order to establish optimal salt stress conditions (Chapter 2). Proteomic approaches were performed by combining difference in-gel electrophoresis (DICE) (Chapter 3) and isobaric peptide tagging (iTRAQ) (Chapter 4) for relative quantification. These methods have been used to monitor protein expression level changes of rice SCCs. A GC-MS based metabolomic approach was used to monitor metabolite changes of rice SCCs (Chapter 5). This thesis aims to create a greater understanding of plant responses to salinity and lead, ultimately, to the production of crop varieties with improved salt tolerance. Chapter 1 is a literature review that provides an overview of our current understanding of the mechanisms contributing to salt tolerance in plants. This chapter also contains a brief over review of proteomics and metabolomics platform technologies that are related to this study. The establishment of the biological system is described in Chapter 2. A combination of morphological analyses, growth performance and transcriptional analyses using selected salt-induced genes were applied to more accurately evaluate the responses of rice SCCs to salinity. Growth of rice SCCs in either 100mM or 200mM increased the expression levels of a broad range of known salt-induced genes involved in cellular protection (OssalT and OsALDH7B6), salt uptake/exclusion (OsHKT1;5 and OsNHX1) and signaling transduction (OsCDPK7). The growth performance data indicates that 100mM NaC1 could significantly affect the both fresh and dry weight growth rate (decrease) during the exponential growth phase of rice SCCs. Moreover, 200mM NaC1 completely inhibited the growth of rice SCCs suggesting that rice cells were severely stressed at this salt concentration and had activated other pathways, such as apoptosis that would interfere with the analysis of the data. It was therefore concluded that rice SCCs collected during the exponential growth phase grown in lOOmM NaCl were the desired conditions for further proteomics and metabolomics experiments. Chapter 3 describes an extensive proteome reference gel-map of rice (Oryza sativa L.) cv. Japonica SCCs using DICE. One hundred and three proteins, including 68 significantly changed in abundance between control and salt treatment were identified. The results suggest that the responsive network of rice SCCs to salinity is highly complex with a wide range of proteins/enzymes involve in the salt response process. The salt responsive network includes proteins involved in the ROS scavenging and detoxifying processes (isocitrate dehydrogenase and glutathione S-transferase), ubiquitin regulated protein degradation processes (SCF protein complex, E1, E2 etc), Aux/IAA signaling pathway, signal transduction (14-3-3 protein family, elongation factor G) primary metabolism pathway (phosphoglucomutase etc), amino acid generation (homocysteine S-methyltransferase). Our results also indicate that the Aux/IAA pathway responses to salt stress may occur via a SCF complex mediated ubiquitation. In addition, the data suggest that the changes in abundance of some proteins under salt stress may be the results of post translational modification(s). The proteomics study was complemented using iTRAQ approach (Chapter 4). Using the same experimental design and same samples as in Chapter 3 provides a rare chance to compare these two primary approaches in current quantitative proteomic research. As expected, there was a significant increase in the number of proteins identified by iTRAQ over DICE. A total of 599 proteins were identified from 1180 distinct peptides and more than 49% of the identified proteins had at least two peptide matches per protein. The protein coverage spanned a range of both low and high abundance proteins. A small proportion of proteins (18) changed significantly in response to NaC1 (100mM) at any one time point. Twelve enzymes involved in glycolytic pathway and citric acid cycles, including 3 changed in abundance, were identified. Among 18 proteins changing in abundance, 6 proteins have functional domains interacting with redox signaling (or related to oxidation protection). These results suggest that anti-oxidation processes in rice SCCs under salt stress may regulated by redox signaling. A brief comparison between DICE and iTRAQ suggests the significant variations in both identification and quantitation by each method. The variation may indicate that a combination of proteincentric (DICE) and peptide-centric (iTRAQ) approaches are essential to acquire adequate data. The results of the proteomics analysis prompted us to extend our investigation to induct a metabolomics (Chapter 5). The metabolomic analysis provided insights into both developmental changes and salt-induced changes of rice SCCs. Both short-term (within 24 hrs) and long-term (up to 7 d) metabolic changes of rice SCCs in response to salt were examined. Statistic tools including PCA and HCA were applied to analysis this complex dataset. In total, 134 compounds were identified, including 30 amines and amides, 40 organic acids (OAs), 40 sugars, sugar acids and sugar alcohols, 21 fatty acids and sterols and 3 miscellaneous compounds. Amongst the identified metabolites, 44 were observed to change by at least 2 fold after the imposition to salt. Increases in amine and amide level may provide osmoprotection, altered OA levels may indicate the reduced metabolic activity and increased anti-oxidative capacity and high levels of sugars and sugar alcohols may contribute to osmoprotection. Chapter 6 is a summary of my conclusions and research highlights extracted from this multidisciplinary approach. Future directions and improvements according to my current knowledge and experience at the end of his PhD are described in this Chapter. This thesis presents one of the most comprehensive functional genomics (proteomics and metabolomics) investigations in the plant salt tolerance mechanism using rice SCCs. This study not only confirmed the results obtained from previous in planta studies related to plant salt tolerance but also provided some potential new salt responsive biomarkers in plants.