School of Agriculture, Food and Ecosystem Sciences - Theses
Permanent URI for this collection
Search Results
1 - 10 of 402
-
ItemThe mechanisms through which fire shapes plant life cycles in heathlands( 2023-11)Fire is a key driver of plant diversity, and many plants have adaptations that help them thrive in fire–prone ecosystems. However, changes to fire activity threaten thousands of plants worldwide. To understand the future of plant populations under fire regime changes, empirical research on fire’s influence on demographic processes is required. This thesis explored how patterns of fire influence plant species across their life cycle, from seeds to mature life stages, and how this relates to plant functional traits. I examined a Mediterranean-type heathland ecosystem as a case study, to examine the mechanisms through which fires impact plants at different life-stages, including those that take place above and below ground. I established 57 study sites in Gariwerd, southeastern Australia, which has experienced substantial variation in fire history. First, I investigated whether knowledge of plant traits can be used to make robust predictions for how fire influences plant relative abundance. I deductively assigned species to plant functional types, based on their persistence traits, establishment capacity, and the timing of key life stages, and made a priori predictions on how relative abundance changes as a product of time since fire. Using empirical data I collected on species relative abundance, I then built nonlinear models to test species’ model conformity to a priori predictions for plant functional types. Predictions of the direction of changes in relative abundance (increase or decrease from 0-81 years since fire) were correct for 18 of 24 species modelled. Predictions of the shape of changes in relative abundance were not as accurate, but still useful: 13 out of 24 species showed ‘excellent’ conformity with shape predictions, 7 ‘good’ conformity, and 4 ‘poor’. This suggests plant functional types can be used to generalise fire responses across species that share similar traits, and thus inform fire management and biodiversity conservation. Second, I examined how fire severity and time since fire interact to influence plant maturity. I collected data on the proportion of plants that had reached reproductive maturity at a site. I used this field data, alongside satellite-based fire severity mapping, to build non-linear models of plant-fire relationships. The results indicated that the proportion of mature plants was influenced by time since fire, regardless of fire severity. For example, for Banksia marginata, the proportion of mature plants increased from 13% (1-year post-fire) to 58% (15 years post-fire), and maturity of this species showed minimal variation between low and high severity fire. Interestingly, no relationships were observed between time since fire and the relative abundance of plants. That is, only when plant life stages were considered, did I detect an effect of fire on plants. Ecological studies that distinguish between plant life stages will help to predict the impacts of fire on populations and enhance decision-making. Third, I investigated how time since fire and mean fire interval influence canopy seedbank production, based on a suite of plant traits. I surveyed all individual plants with canopy cones present at each of the 57 study sites. On each mature individual, I measured plant height and width, and counted the number of cones. I sampled a subset of these cones across individual plants, and then germinated them in a laboratory trial. I used regression models to explore the relationship between fire frequency and variables relating to different aspects of canopy seedbank production. The interval between fires influenced canopy seedbank production and viability. For example, no canopy cones were observed on plants at short mean fire intervals: such as fire intervals more frequently than every 18 years for the obligate seeder tree Callitris rhomboideia. Quantifying the fire intervals which supports canopy seedbanks provides a new understanding of an important above ground process and helps to determine how frequently to burn ecosystems containing serotinous species. Last, I examined how time since fire and fire frequency influence the occurrence of different species in the soil seedbank and, again, examined ecological relationships through the lens of plant traits. I sampled the soil seedbank at 57 sites, treated soil samples with heat and smoke product to promote germination, and grew seedlings in a germination trial lasting 14 months. I used non-linear modelling to explore relationships between fire and species occurrence. Fire frequency influences the occurrence of species in the soil seedbank, and the nature of these relationships depends on plant traits such as plant and seed longevity. For example, frequent fires (every <15 years) will reduce the occurrence of herbaceous species with long-lived seed. However, for other types of plants, such as perennials with short-lived seed, I observed no relationship between fire and soil seedbank occurrence, demonstrating many species have soil seedbanks resilient to frequent fires. Overall, my research advances understanding of how fire impacts different species and groups of plants across their life cycle. Notably, a mix of field research, laboratory studies and empirical models provide evidence that the traits of plants can be used to identify how fire affects species in the soil and canopy seedbanks, and as juvenile and mature plants. By examining plant life stages above and below ground, this work also helps to define the fire regimes that support plants in the heathy woodlands of Gariwerd. Because it is based on mechanisms, I anticipate that the trait-based approaches I have developed and tested could be used to understand and predict fire-related changes in plant populations in a wide range of ecosystems.
-
ItemPredicting redistribution of species and communities under environmental change: Improving the reliability of predictions across time( 2023-04)Ecological models used to forecast range change (range change models; RCM) have recently diversified to account for a greater number of ecological and observational processes in pursuit of more accurate and realistic predictions. Theory suggests that process-explicit RCMs should generate more robust forecasts, particularly under novel environmental conditions. RCMs accounting for processes are generally more complex and data-hungry, and so, require extra effort to build. Thus, it is necessary to understand when the effort of building a more realistic model is likely to generate more reliable forecasts. During my thesis, I investigated how explicitly accounting for processes improves the temporal predictive performance and transferability of RCMs. I first identified key knowledge gaps, and the challenges of evaluating temporal predictive performance and transferability. One of the main challenges is the lack of robust metrics to assess predictive performance and transferability. To address this I implemented and tested the use of new emerging tools to enable fair comparisons of predictive performance across samples with varying degrees of imbalance (e.g. species with low and high observed prevalence). I then tested a couple of hypotheses related to whether modelling observational processes explicitly results in better forecasts. In particular, I evaluated under what circumstances the benefits of explicitly accounting for imperfect detection and allowing information sharing across multiple species are retained when the models are extrapolated to generate predictions beyond the training temporal window. The findings should shed light on how to address remaining knowledge gaps, and how to generate more reliable forecasts on species’ responses to global change scenarios.
-
ItemTrees Need Closure Too: Unveiling The Molecular Control Of Wound-Induced Secondary Vascular Tissue Regeneration In Trees( 2023-09)Trees play a pivotal role in terrestrial ecosystems and are an important natural resource. These attributes are primarily associated with the capacity of trees to continuously produce woody tissue from the vascular cambium, a ring of meristem cells located just beneath the bark between phloem and xylem tissue layers. Long-lived trees are exposed to a myriad of biological and environmental stresses that may result in wounding, leading to a loss of bark and the underlying vascular cambium. This affects both wood formation and the quality of timber arising from the tree. In addition, the exposed wound site is a potential entry point for pathogens that cause disease and may even lead to the death of the whole plant. In response to wounding, trees have the capacity to regenerate lost or damaged tissues at a wound site. Investigating gene expression changes associated with different stages of wound healing reveals complex and dynamic changes in the activity of transcription factors, signalling pathways and hormone responses. This thesis investigated molecular regulators of wound-induced secondary vascular tissue (SVT) regeneration. It summarises current literature on primary and secondary vascular tissues and bark wounds and related revascularisation processes, specifically on genes and hormones. Using this information, eight genes from Eucalyptus, including WUSCHEL RELATED HOMEOBOX 4 (EgrWOX4), Arabidopsis thaliana HOMEOBOX GENE 8 (EgrATHB8), CORONA (EgrCNA), PHABULOSA/PHAVOLUTA (EgrPHX), REVOLUTA (EgrREV), AUXIN RESPONSE FACTOR 5 (EgrARF5), PIN-FORMED 1 and 3 (EgrPIN1 and EgrPIN3) were chosen for subsequent experiments on wound-induced SVT regeneration. During these in-planta experiments, Induced Somatic Sector Analysis (ISSA) was used as a molecular tool to assess promoter activity and gene function of these candidate genes in wild-type stems and those where auxin transport was chemically inhibited. Endogenous auxin (IAA) concentrations were quantified using LC-MS to understand how varying auxin concentrations might be required for proper vascular tissue patterning during various stages of regeneration. Results show that the remaining xylem tissues on the wound surface regenerate all lost tissues in a four-step process. EgrPIN1/3 are expressed in all tissue types, EgrWOX4, EgrARF5 and EgrREV predominantly in cambium tissues and EgrATHB8, EgrCNA and EgrPHX in cambium and xylem tissues. Overexpressing micro-RNA-resistant REV leads to faster regeneration rates, while over-expressing miR166 and chemical inhibition of polar auxin transport leads to slower regeneration rates. Samples from overexpression experiments and auxin inhibition also lead to defects in cell anatomies, arrangement, and organisation. Quantification of IAA levels suggests alternating high and low auxin signalling during different stages of regeneration. Together, this thesis provides novel insights into spatial-temporal expression patterns of the selected molecular regulators and discusses how they relate to our current understanding of vascular cambium formation and xylem differentiation during secondary growth. Based on the findings, I propose a model for wound healing that provides the conceptual foundations for future studies aiming at understanding this intriguing process.
-
ItemMycorrhizal associations of Prasophyllum R.Br. (Orchidaceae) and the conservation of its threatened species(University of Melbourne, 2010)
-
ItemGenetic variation in Eucalyptus regnans F. Muell(University of Melbourne, 2010)Past climate changes and associated range contraction, expansion and introgression events are regarded as major forces shaping the evolution of the genus Eucalyptus. Previous investigations in south-eastern Australia have identified some putative refugia and recolonised areas for tall eucalypt forests during the last glacial cycle but the response of individual species are very little understood and in some regions, the evidence for migration or persistence is completely absent. Here I uncover the response of Eucalyptus regnans and two species (E. delegatensis and E. obliqua) which often occur with it and among all of which there is likely to be introgression. These results have important implications for understanding forest responses to climate change and for current gene pool management. Chloroplast variation in Eucalyptus regnans suggests that in spite of the narrow ecological tolerances of the species and the harsh environmental conditions of the Last Glacial Maximum (LGM), E. regnans was able to persist locally, in or near its current range, often in what would have been unfavourable conditions. Therefore, rather than the waves of recolonisation or long distance dispersal experienced by tree species in much of Europe or North America, E. regnans, with the exception of the Central Highlands region of Victoria in mainland Australia and the central region of Tasmania, does not appear to have migrated great distances. A comparative phylogeographic study of three south-east Australian Eucalyptus species (E. delegatensis, E. obliqua, and E. regnans) with different tolerances, in terms of cold, drought and soils revealed that, despite ecological differences, the three species have responded to past climatic changes in a similar way, by persisting in multiple refugia in similar locations. Inter-specific sharing of haplotypes was extensive, and fixation of shared, regional haplotypes, was more likely in areas postulated as treeless at the LGM. It seems likely that the process of recolonisation, in particular, results in loss of chloroplast diversity within species and interspecific haplotype sharing. A nuclear microsatellite study of range wide variation in E. regnans found a high level of diversity, and, in contrast to the findings from the study on chloroplast variation, a generally low level of geographic differentiation. The findings do not support the hypothesis of a high level of genetic differentiation in the species that was suggested by the chloroplast data and its disjunct distribution. No loss of nuclear diversity was found at locations fixed or nearly fixed for the hypothesised recolonising chloroplast haplotypes. I argue that a loss of diversity at recolonised locations has been offset by hybridisation with, and genetic displacement of a pre existing (species was already present at site), compatible species. Extensive pollen mediated gene flow both during and after recolonisation from E. regnans in refugia may have led to the displacement of a pre-existing species and the homogenisation of nuclear microsatellite allele frequencies but retained the pattern of chloroplast variation from the displaced species. This scenario would explain the maintenance of high nuclear and low chloroplast variation at recolonised locations and the low level of genetic differentiation between recolonised locations and adjacent refugia. These results suggest that E. regnans is a species with a strong phylogeographic structure but a low level of nuclear differentiation. Across the species range there is a complex mosaic of populations with different histories of colonisation, persistence and introgression. Phylogeographic patterns in this study indicate that the major biotic responses of tall eucalypt forest to climatic changes were persistence and resilience. The implications of these findings are discussed in the context of defining gene pool management strategies for E. regnans. They generally support the conservative management objectives of the current seed transfer guidelines for commercial forestry operations in both Victoria and Tasmania: seed should be sourced from the harvested coupe if possible or from nearby stands matched for environment. In the central highlands of Victoria and central Tasmania where there is low between stand differentiation longer distance transfers may be more acceptable provided the environment is matched. In other regions such as East Gippsland where long-term, in situ persistence has resulted in strong differentiation between stands, then transfer of seed over relatively small geographic distances will disrupt the established patterns of evolutionary lineages. Given the evidence of the importance of introgression in E. regnans and its relatives, the current policy of maintaining the pre-existing species mix on regeneration should also assist in preserving the collective adaptive potential of the species at a site.
-
ItemCommunity dynamics in natural resource governance : building adaptive management capacity towards ecological sustainability(University of Melbourne, 2010)The aim of the research was to analyze community dynamics and collective action for sustainable natural resource governance in decentralized Indonesia. The research was an ethno-methodology research in which in-depth interview and participant observation were used for data collection. Data analysis was carried out by examining the distribution of narratives provided by the respondents, and by carrying out a thematic analysis in which emerging themes were used to produce a complex and coherent narrative of the discourse found within the case study site. The research aims to explore the various practices of natural resource governance and the complex social relations which influence collective action for the sustainable governance of natural resources. Natural resource governance in modem Indonesia is marked by the tension between the centralized policy strategy of the Suharto period and the reactive strategy of Post-Suharto decentralization. To some extent, decentralization led to devolution of power and opportunities for local resource users to make consequential decisions over the natural resources upon which they depend. However, this approach rested upon the capacity of communities to reach a consensus untainted by local politics, commercial imperatives and traditional power stmctures. Moreover, decentralization had not given the majority access to strategic and structural decision making power. Empirical findings from Tongke Tongke�s mangroves in Sinjai, South Sulawesi, suggest that social institutions and local mles came into play and the people felt honored to protect the resource on behalf of the community. These social institutions took the form of neighborly ties, collective identity, reciprocity and social and ecological responsibilities. Access to Tongke Tongke�s mangroves was not free but governed by local and informal mles to maintain its benefits for the good of the community. The community, through the elders, was determining access and making decisions about management on behalf of them all. Community members acted in a way that benefited the overall good even when they were avowing individual rights. The thesis argued that individuals evolved behavior which was commensurate with their responsibilities, leading to innovative power stmctures which were more locally sensitive and environmentally appropriate. The case study in the village of Tongke Tongke within the Regency of Sinjai suggested a rebuttal of Hardin�s Tragedy of the Commons. In line with Ostrom�s theory, the commons is governed by local and often informal mles which induce behavior that are in line with a collaborative mentality to maintain its benefits for the good of the community. Nonetheless, as suggested by Bookchin and argued in the thesis, collective natural resource governance is also about individuals who resist and accept the shaping of civic collaboration and ecological sustainability. In addition, barriers and enablers for sustainable natural resource governance need to emerge from local contexts; they could not emerge as a consequence of top down devolution alone. Moreover, no preparation of local communities could be made to assume the unintentional consequences of complex power relations. In line with Etzioni�s theory, empirical findings suggest that real power relationships in real resource management contexts can undermine the possibility of democratic and equitable consensus making. The thesis argued that social reciprocity, identity validation and symbolic capital can motivate resource users to behave in line with a collaborative mentality for mangrove protection.
-
ItemMaking sense of social research in forestry : an exploration of interactions in an interdisciplinary research process(University of Melbourne, 2010)Many social scientists believe their disciplines bring a critically important perspective to environmental problems (Ewert, Baker, & Bissix, 2004; Lockie, Higgins, & Lawrence, 2001; Reser & Bentrupperbaumer, 2001; Roughley & Salt, 2005). Problems involving natural resource management, particularly competing land uses, are often thought to inherently be social problems. Despite this, social researchers have faced challenges in contributing to solving these problems. A critical challenge is lack of understanding of social science by natural scientists, policymakers and non-academic research partners, such as industry. The complex nature of these problems requires coordination between these stakeholders. Gibbons et al. (1994) have described the trend towards collaborations which involve a broader range of academic and societal stakeholders as a new �mode� of knowledge production. Inroads have been made to understanding how knowledge production occurs in the new �mode�, especially in the area of study focused on interdisciplinary collaboration. Some of this work has specifically focused on collaboration across the �great divide� of natural-social sciences in natural resource management and environmental problem settings (e.g. L�l� & Norgaard, 2005), yet it is still unclear how social scientists can, and do, contribute alongside others. The aim of this study is to understand and conceptualise the processes in which social scientists and others engage to make social research usable. This study uses a narrative ethnographic approach (B. Tedlock, 1991), taking advantage of my position as a collaborator studying collaboration between social researchers and foresters in the Cooperative Research Centre Forestry (CRC) Communities Project based in Australia. Principal methods involved participant-observation over 21 months, detailed field note-taking, interviewing members of the Communities Project, and analysis by renarrating field experiences. Studies of natural-social scientific interdisciplinarity provided the expectation that there would likely be miscommunication, misunderstandings and missed opportunities. Instead, social researchers and industry partners managed to make connections. Social researchers were able to communicate the messages they saw as important (even when these had negative implications for industry partners) while industry partners increased their understanding of the methods, findings and implications of social research. I found that this was facilitated by two interactive processes. The first, I labeled �reframing�. This involved participants engaging in questioning and clarification as well as mutually learning each others� vocabularies and concepts. I labeled a second process �stretchy narratives�, referring to participants� stretching, or expansion of their original narrative to incorporate critical features of others� narratives. These interactive processes constituted integrative practice in this interdisciplinary collaboration, adding to previously abstract conceptualization of integration (e.g. Klein, 2001b; Newell, 2001). Working with industry partners, I redefined the nature of our differences, finding that they were focused on practical measures such as clear project explanations and valuable research outcomes. This, and the lack of obvious �disciplinary� struggle, suggests that in some �interdisciplinary� collaborations, particularly those with non-academic stakeholders, philosophical differences may be less relevant than social scientists have argued (e.g. Eigenbrode et al., 2007). Given the notion that disciplinarians have somewhat fixed epistemological positions, project partners� perspectives were more fluid than anticipated. My research suggests that those working in natural-social science collaborations in environment and natural resource management problem settings can make social research usable by adopting a gradual and incremental approach to educating and learning from others, engaging opportunistically in teaching moments.
-
ItemPlacing the native pest in the postcolonial city : the relocation of Melbourne's flying foxes(University of Melbourne, 2010)As long as wildlife inhabits the urban environment, humans will experience interspecies conflicts. And as long as these conflicts occur, questions will be raised about who and what belongs in city places, and what it means to put native pests �in their place�. The ways we deal with animals judged to be �out of place� have effects: not only for the species and spaces involved, but also for our very ideas about these species and spaces, and for our ideas about the place of nature in our postcolonial urban lives. Human interventions with animals matter because they are not only reflective of the socio-spatial order, but productive of it. This thesis, grounded in cultural geography, investigates the way native animals figure in the socio-spatial orderings of the contemporary Australian city. It aims to make an original contribution to animal geography by paying sustained attention to the ways wild animals are placed discursively and practically by humans, and in turn, how those wild animals place themselves. I pursued this research by following the 2003 relocation of Melbourne�s colony of native greyheaded flying-foxes from the Royal Botanic Gardens (RBG) to Yarra Bend Park. I used this case of animal transgression and human intervention to make wider claims about the paradoxical place of the native pest in the postcolonial city. I took an in-depth and interpretive approach to the research, with the aim of providing qualitative, contextualised claims. My methods included: targeted interviews with key stakeholders; participant observation activities; field site visits; news media reviews; and content analysis of key documents and websites. The thesis structure follows the chronological journey of the relocation: from the flying-fox colony�s discursive rejection from the RBG via the Save Our Gardens Campaign; to the determination of the new roosting site by the Grey-headed Flying-fox Alternative Campsite Task Force; the embodied relocation of the flying- foxes across Melbourne; and, finally, the discursive and material work undertaken to accommodate the colony in Yarra Bend Park. The thesis concludes that, while the relocation has set a precedent for dealing with native pests, its supposed success sits in tension with the actual nature of the relocation. The relocation achievement was not as tidy, successful, nor as human as many post-relocation claims suggest. The relocation was constituted by a messy, hybrid process largely shaped by the bats, and resulted in hybrid and unstable outcomes.
-
ItemConflicting agendas : illicit drugs, development and security in the Wa special region of Myanmar(University of Melbourne, 2010)This thesis examines the reasons behind the persistent conflict and underachievement in global efforts to control illicit drug production in the South and argues for a new approach. This underachievement is explored through a detailed case study of a remote illicit drug producing area in Myanmar, the Wa Special Region (WSR), and through a comparative examination of the illicit drug problem in Afghanistan. Using extensive field interviews, results from existing quantitative studies and drawing on literature that examines the relationship between security and development, a comprehensive picture is painted that describes the perspectives, motivations and actions of both Northern and Southern actors over time with respect to three key inter-related themes: drug control, security and development. The conceptual framework that is used for the thesis uses security and development research and illicit drug studies to map and analyse conflict in actor policy around these three themes. The thesis makes a number of findings and recommendations. Firstly, new information on an under-researched region, its people and leaders is presented. The WSR is a poorly understood region that has been dependent on illicit opium cultivation and trade for over a hundred years and a more detailed understanding of its history, security concerns and recent actions may assist in future policy development towards the region. Secondly, the thesis argues that, at a relatively localized level, each drug growing region is context specific and the details of local economies and political complexes at this level are often not well understood. This lack of understanding can contribute to project failure or irrelevance of drug control efforts. More consistent progress on the illicit drug control issue requires that the complexity of individual production areas needs to be analysed, particularly local governance structures, the underlying economy and perceptions and actions of local leaders. The thesis also shows that drug control is characterised by conflicting perspectives and actions by local and external actors. Policies are often based on self-interest and may be single-minded in that they ignore the interdependency of drug control, security and development. A general inattention to gathering evidence through research and analysis of the underlying social, economic, political and cultural factors, compounds this conflict and actors are informed by a partial assessment of the drug control problem. Limited interaction between different actors leads to lack of coordination, coherence and opportunities for consensus-building on drug control action and policy, which may adversely impact local populations or the level of drug production. In some cases, local actors may be criminalised by external ones foreclosing opportunities for continuing cooperation and progress. This thesis argues that local actors and governance structures are paramount and must be considered in areas where illicit drug production is to be addressed. Local leaders must be respected and local achievement should be recognised and reinforced. Further, that the security concerns of all actors including local leaders need to be considered and compared as a starting point in any drug reduction programme. The thesis recommends that drug control should be underpinned by evidence-based complexity analysis of drug producing regions that incorporates both local and external actors and be overseen by a single UN body to build consensus. Finally, there must be a concomitant willingness on the part of powerful international actors to share intelligence and to pursue a more cooperative approach to drug control and less unilateralism.
-
ItemQuantification and mitigation of gaseous nitrogen emissions from pasture and cropping systems(University of Melbourne, 2010)Due to the application of nitrogen (N) fertilizers and animal excreta agricultural soils emit large quantities of ammonia (NH3), nitrous oxide (N2O) and nitric oxide (NO) to the atmosphere. As well as negative environmental effects, such as contribution to acid precipitation and enhancement of the greenhouse effect, the processes that produce these gases are the main cause of the low efficiency of N fertilizers. A series of field studies were carried out in Australia and China to quantify emissions of NH3, N2O and NO to investigate how farming practices impact on the soil N dynamics responsible for their loss and to recommend strategies to reduce them. A micrometeorological method using a backward Lagrangian stochastic model (bLs) was used to determine NH3 volatilization resulting from grazing a pasture by a herd of dairy cows and from subsequent urea application. Emissions were 4.9 and 1.9 g N cow"1 d"1 in summer and autumn, respectively, with the larger losses in summer due to higher temperatures and evaporation rates. Loss of fertilizer N in summer was 1.2% of applied- N when irrigation followed fertilizer application immediately and 5.3% in autumn when irrigation took place 3 days after application. Measurements of N2O emissions were made with 100 static chambers from an approximately 1 ha area in the same pasture a few days after the sequence of grazing, fertilizer application and irrigation. The key soil physical and chemical properties from under each chamber were analyzed. The N2O emissions were extremely variable, ranging from 45-765 (average 165) ng N m"2 s"1 in the summer and 20-953 (average 138) ng N m"2 s"1 in autumn. Emissions showed spatial dependence up to 73 and 51 m in autumn and summer, respectively, and were significantly correlated with soil nitrate (r2 = 0.79) and ammonium (r2 = 0.41) through the effects of animal traffic in the field, indicating both nitrification and denitrification contributed to N2O emissions. Ammonia loss from fertilizer application to winter cereals in western Victoria was measured following application of urea, ammonium sulfate and urea-ammonium nitrate. Ammonia loss varied with location and fertilizer type and ranged from 1.3 to 23% of N applied. The variation with location seemed to be due to a combination of factors including soil texture, soil moisture when fertilizer was applied and rainfall (how soon and how much) after fertilizer application. Addition of a urease inhibitor [N-(n-butyl) thiophosphorictriamide] to a urea treatment reduced losses by more than 80%. Emissions of NOx and NH3 from a maize crop fertilized by deep-placed urea in northern China were 1.2% and 6.5% of the applied N, respectively. The loss of NOx was far greater than that reported by others, possibly because the measurements were made continuously in this study rather than as spot measurements with static chambers. Soil analyses to depth indicate that little nitrate was leached down the profile and a study using soil cores and acetylene inhibition to measure denitrifying activity suggested that the rate of denitrification was very slow in this soil. The results suggest that in this soil with a slow denitrification rate and little leaching, deep placement of the urea to limit NH3 volatilization is an effective method for decreasing N loss to the atmosphere and increasing fertilizer use efficiency. The soil N dynamics of an irrigated maize system in which stubble retention and stubble burned treatments were superimposed over treatments of varying N fertilizer rate were studied at a site near Whitton, NSW. Field measurements using 15N-labelled urea recovery from microplots with and without the nitrification inhibitor ammonium thiosulfate (ATS), showed significantly more fertilizer N was recovered in the grain from a stubble incorporated treatment than a stubble burned treatment (P < 0.05), and more fertilizer N remained in the soil in the stubble burned treatment. This indicated lower denitrification losses of fertilizer N applied to the stubble burned system. The high N recovery in grain in the stubble incorporated treatment was thought to be due to improved soil structure in the heavy texture soil. Inclusion of 5% ATS in the fertilizer formulation did not appear to have an impact on N recovery of fertilizer N. Nitrous oxide emission was strongly dependent upon fertilizer rate and there was no statistically significant effect of stubble management on N20 emission rate found using static chambers. The effects of soil fertility on emission of NO and N20 from mixed pasture converted to high clover pasture, and subsequently to a cereal crop were studied for 21 months. More NO and N20 was emitted from the high fertility than from the low fertility paddocks likely due to restriction of legume growth by phosphorus deficiency. This effect was amplified by increased N cycling due to higher stocking densities on the high fertility pasture and stubble-grazed wheat treatments. Overall, changing the land use from a mixed clover grass pasture to a wheat crop resulted in increased emissions of NO and N2O (2.03 kg as NO and 2.09 kg as N2O). Most of the N2O was emitted when the water filled pore space (WFPS) was high and most of the NO was emitted when the WFPS was low suggesting that the N2O was formed during denitrification and the NO was formed by nitrification.