Dissertations / Theses on the topic 'Fire ecology Australia'

The forests of south-eastern Australia, having evolved in one of the most fire-prone environments in the world, are characterized by many adaptations to recovery following burning. Thus forest ecosystems are characterized by rapid regenerative capacity, from either seed or re-sprouting, and mechanisms to recover nutrients volatilized, including an abundance of N2 fixing plants in natural assemblages. Soil physical, chemical and biological properties are directly altered during fire due to heating and oxidation of soil organic matter, and after fire due to changes in heat, light and moisture inputs. In natural ecosystems, carbon (C) and nitrogen (N) lost from soil due to fires are recovered through photosynthesis and biological N2 fixation (BNF) by regenerating vegetation and soil microbes.
This study investigated post-fire recovery of soil C and N in four structurally different sub-alpine plant communities (grassland, heathland, Snowgum and Alpine ash) of south-eastern Australia which were extensively burnt by landscape-scale fires in 2003. The amount and isotopic concentration of C and N in soils to a depth of 20 cm from Alpine ash forest were assessed five years after fire in 2008 and results were integrated with measurements taken immediately prior to burning (2002) and annually afterwards.
Because the historical data set, comprised of three soil samplings over the years 2002 to 2005, consisted of soil total C and N values which were determined as an adjunct to 13C and 15N isotopic studies, it was necessary to establish the accuracy of these IRMS-derived measurements prior to further analysis of the dataset. Two well-established and robust methods for determining soil C (total C by LECO and oxidizable C by the Walkley-Black method) were compared with the IRMS total C measurement in a one-off sampling to establish equivalence prior to assembling a time-course change in soil C from immediately pre-fire to five years post-fire. The LECO and IRMS dry combustion measurements were essentially the same (r2 >0.99), while soil oxidizable C recovery by the Walkley-Black method (wet digestion) was 68% compared to the LECO/IRMS measurements of total C. Thus the total C measurement derived from the much smaller sample size (approximately 15 mg) combusted during IRMS are equivalent to LECO measurement which require about 150 mg of sample.
Both total C and N in the soil of Alpine ash forests were significantly higher than soils from Snowgum, heathland and grassland communities. The ratio of soil NH4+ to NO3- concentration was greater for Alpine ash forest and Snow gum woodland but both N-fractions were similar for heathland and grassland soils. The abundance of soil 15N and 13C was significantly depleted in Alpine ash but both isotopes were enriched in the heathland compared to the other ecosystems. Abundance of both 15N and 13C increased with soil depth.
The natural abundance of 15N and 13C in the foliage of a subset of non-N2 fixing and N2 fixing plants was measured as a guide to estimate BNF inputs. Foliage N concentration was significantly greater in N2 fixers than non-N2 fixers while C content and 13C abundance were similar in both functional groups. Abundance of 15N was depleted in the N2 fixing species but was not significantly different from the non-N2 fixers to confidently calculate BNF inputs based on the 15N abundance in the leaves.
The total C pool in soil (to 20 cm depth) had not yet returned to the pre-fire levels in 2008 and it was estimated that such levels of C would be reached in another 6-7 years (about 12 years after the fire). The C and N of soil organic matter were significantly enriched in 15N and 13C isotopes after fire and had not returned to the pre-fire levels five years after the fire. It is concluded that the soil organic N pool can recover faster than the total C pool after the fire in the Alpine ash forests.

The honey possum Tarsipes rostratus is a tiny (7 - 12 g) highly specialised flower-feeding marsupial endemic to the south-western corner of Australia. The impact of fire on this small mammal was studied, over a 19-year period, in the Fitzgerald River National Park, a large (330,000 ha) area of relatively undisturbed heathland/shrubland, rich in the proteaceous and myrtaceous plants upon which the honey possum appears to rely for food. The honey possum is the most abundant and widespread mammal in this Park. Capture rates of honey possums were significantly related to the years since the vegetation was last burnt, annual rainfall in the preceding (but not the current) year, the season when trapping occurred, and the trapping grid operated. Capture rates declined markedly after fire and remained low (less than one third of those in long unburnt vegetation) for about 4 - 5 years following a fire. Rates of capture then increased steadily over the next 20 - 25 years, with maximal abundance recorded about 30 years after fire. Thereafter, there appeared to be a slight decline in capture rates, but even in the vegetation unburnt for longest (> 50 years since fire), honey possum abundance was substantial and relatively stable. In contrast to these changes in abundance, the structure of the honey possum population, with 79 % adults and 57 % males, appeared little influenced by fire history, annual rainfall, season or grid. The increase in the rates of capture of honey possums following fire paralleled the pattern of availability of cover in the vertical and, to a lesser extent, horizontal plane. Indeed, projective foliage cover took around 20 years after fire to reach levels similar to those available in areas unburnt for even longer. The trend in capture rates was also congruent with the maturation of the most frequently visited foodplants of honey possums, particularly Banksia nutans (summer flowering) and B. baueri (winter flowering). Areas long unburnt still contained shelter and foodplants adequate for honey possums even 50 years or more after fire, with only slight evidence of senescence. Pollen loads indicated that honey possums caught in burnt areas, where their preferred foodplants were absent, continued to feed on these favoured foodplants (Banksia and Dryandra spp.) at nearby unburnt areas. In addition, they also fed, in both burnt and long unburnt areas, upon a suite of other plant species that regenerated more rapidly from lignotubers and epicormic buds, as well as from seeds (e.g. Eucalyptus and Calothamnus spp.). Thus, honey possums appeared to persist with their preferences for feeding from a limited number of flowering plants despite some of these species not being available in recently burnt areas for many years. Nearby patches of unburnt vegetation can clearly be important refuges, feeding grounds and shelter for the few honey possums that visit recently burnt areas, and appear to be the source of honey possum colonists in the years following a fire. Capture rates were also greater following years when rainfall was higher than average. Indeed, rainfall had as great an influence upon capture rates as time since fire. Capture rates were also consistently higher over winter, and to a lesser extent over summer, than in either autumn or spring. Individual grids, even those close together in apparently similar vegetation with a similar fire history, still differed significantly overall in their capture rates of honey possums. This last finding has implications for the use of chronosequences in the study of post-fire changes in biota. Although not the primary focus of the study, data on the limited suite of other, far less abundant, small mammals present indicated that house mouse Mus musculus domesticus numbers peak soon after fire (about two years after fire), grey-bellied dunnart Sminthopsis griseoventer numbers somewhat later (about eight years after fire) and that southern bush rats Rattus fuscipes fiuscipes, like honey possums, are later successional species. Most species were present in vegetation over a range of post-fire ages, with data consistent with models based on sequential changes in relative abundance. Like many Australian mammals, the range of the honey possum has contracted substantially over the last 200 years and the coastal heathlands of the south-west are its last stronghold. In terms of its conservation, this study indicates that, if possible, management burns in these heathlands should be separated by intervals of at least 20 years between successive burns, and preferably even longer. If burns are required more frequently to meet other management priorities, it is highly preferable that they are small and patchy, rather than large scale. Such practices may help ensure the long-term survival of this unique, highly specialised and endemic marsupial.

The honey possum Tarsipes rostratus is a tiny (7 - 12 g) highly specialised flower-feeding marsupial endemic to the south-western corner of Australia. The impact of fire on this small mammal was studied, over a 19-year period, in the Fitzgerald River National Park, a large (330,000 ha) area of relatively undisturbed heathland/shrubland, rich in the proteaceous and myrtaceous plants upon which the honey possum appears to rely for food. The honey possum is the most abundant and widespread mammal in this Park. Capture rates of honey possums were significantly related to the years since the vegetation was last burnt, annual rainfall in the preceding (but not the current) year, the season when trapping occurred, and the trapping grid operated. Capture rates declined markedly after fire and remained low (less than one third of those in long unburnt vegetation) for about 4 - 5 years following a fire. Rates of capture then increased steadily over the next 20 - 25 years, with maximal abundance recorded about 30 years after fire. Thereafter, there appeared to be a slight decline in capture rates, but even in the vegetation unburnt for longest (> 50 years since fire), honey possum abundance was substantial and relatively stable. In contrast to these changes in abundance, the structure of the honey possum population, with 79 % adults and 57 % males, appeared little influenced by fire history, annual rainfall, season or grid. The increase in the rates of capture of honey possums following fire paralleled the pattern of availability of cover in the vertical and, to a lesser extent, horizontal plane. Indeed, projective foliage cover took around 20 years after fire to reach levels similar to those available in areas unburnt for even longer. The trend in capture rates was also congruent with the maturation of the most frequently visited foodplants of honey possums, particularly Banksia nutans (summer flowering) and B. baueri (winter flowering). Areas long unburnt still contained shelter and foodplants adequate for honey possums even 50 years or more after fire, with only slight evidence of senescence. Pollen loads indicated that honey possums caught in burnt areas, where their preferred foodplants were absent, continued to feed on these favoured foodplants (Banksia and Dryandra spp.) at nearby unburnt areas. In addition, they also fed, in both burnt and long unburnt areas, upon a suite of other plant species that regenerated more rapidly from lignotubers and epicormic buds, as well as from seeds (e.g. Eucalyptus and Calothamnus spp.). Thus, honey possums appeared to persist with their preferences for feeding from a limited number of flowering plants despite some of these species not being available in recently burnt areas for many years. Nearby patches of unburnt vegetation can clearly be important refuges, feeding grounds and shelter for the few honey possums that visit recently burnt areas, and appear to be the source of honey possum colonists in the years following a fire. Capture rates were also greater following years when rainfall was higher than average. Indeed, rainfall had as great an influence upon capture rates as time since fire. Capture rates were also consistently higher over winter, and to a lesser extent over summer, than in either autumn or spring. Individual grids, even those close together in apparently similar vegetation with a similar fire history, still differed significantly overall in their capture rates of honey possums. This last finding has implications for the use of chronosequences in the study of post-fire changes in biota. Although not the primary focus of the study, data on the limited suite of other, far less abundant, small mammals present indicated that house mouse Mus musculus domesticus numbers peak soon after fire (about two years after fire), grey-bellied dunnart Sminthopsis griseoventer numbers somewhat later (about eight years after fire) and that southern bush rats Rattus fuscipes fiuscipes, like honey possums, are later successional species. Most species were present in vegetation over a range of post-fire ages, with data consistent with models based on sequential changes in relative abundance. Like many Australian mammals, the range of the honey possum has contracted substantially over the last 200 years and the coastal heathlands of the south-west are its last stronghold. In terms of its conservation, this study indicates that, if possible, management burns in these heathlands should be separated by intervals of at least 20 years between successive burns, and preferably even longer. If burns are required more frequently to meet other management priorities, it is highly preferable that they are small and patchy, rather than large scale. Such practices may help ensure the long-term survival of this unique, highly specialised and endemic marsupial.

[Truncated abstract] Mediterranean southwest Australia, a global biodiversity hotspot, has nutrient deficient soils, exacting climatic conditions and is species rich with 7380 native vascular plant species, of which 49% are endemic. The region is expected to experience one of the world's highest degrees of biodiversity loss and change in the coming decades, with introduced species presenting a major threat. Limited knowledge is available on the mechanisms of ecosystem change associated with invasion and fire in this biodiversity hotspot region. Banksia woodland, an iconic complex species-rich natural ecosystem is one of the major vegetation types of the coastal sandplain, extending from 15 to 90 km inland and 400 kms along the west coast. The following hypothesis was tested to explore the ecological impacts of invasion: Is invasion of Banksia woodland by the introduced species Ehrharta calycina and Pelargonium capitatum accompanied by an alteration in ecosystem properties and processes, whereby the degree of change is related to fire frequency and abundance of introduced species? Different vegetation conditions, i.e. Good Condition (GC), Medium Condition (MC), Poor Condition invaded by Ehrharta calycina (PCe) and Poor Condition invaded by Pelargonium capitatum (PCp) were utilized for field assessments. ... In the soil seed bank, species numbers and germinant density decreased significantly for native and seeder (fire sensitive) species between GC sites and invaded sites. Surprisingly 52% of germinants at GC sites were from introduced species, with much of the introduced soil seed bank being persistent. Native species were dominated by perennial shrubs, herbs and sedges, while introduced species were dominated by perennial and annual grasses and herbs. Invasion by introduced species, associated with frequency of fire, altered the ecosystem, thus disadvantaging native species and improving conditions for even greater invasion within the Banksia woodland. Significantly higher soil phosphorus P (total) and P (HCO3) were found at PCe and PCp sites compared to GC sites. Leaf nutrient concentrations of phosphorus were significantly higher, and potassium and copper significantly lower in PCe and PCp sites, with introduced species having significantly greater concentrations than native species (except Manganese). This study demonstrated the key role of phosphorus in the Banksia woodland, in contrast to other research which identified nitrogen as the major nutrient affected by invasion. Higher levels of soil and leaf phosphorus, loss of species diversity and function, changes in fire ecology and canopy cover and a limited native soil seed bank make restoration of a structural and functional Banksia woodland from the soil seed bank alone unlikely. Without management intervention, continuing future fire is likely to result in a transition of vegetation states from GC to MC and MC to PC. The knowledge gained from this study provides a better ecological understanding of the invasive process. This enhanced understanding will enable the development of adaptive management strategies to improve conservation practices within a biodiversity hotspot and reduce the impact of the key threatening process of invasion.

Bushfire is, in terms of human lives lost, property destroyed, and damage to natural systems, by far the most urgent environmental problem in Australia. This thesis tries to answer a number of questions about bushfire behaviour, history, effects, and management, in the Wungong Catchment of Western Australia. It does so by an overtly cross-disciplinary approach, involving a mixture of the three main streams of human knowledge, namely the humanities, natural science, and social science.First, I offer a literature review of several hundred books and papers drawn from the three main streams of knowledge mentioned above. The review includes some discussion of ‘bushfire epistemology’, a currently vague and neglected matter.The concept of ‘place’ is important to humans, so I then give a straightforward geographical description of Wungong Catchment, with some mention of the history of bushfire. To describe the vegetation, I use inductive statistics, and a method developed by me from the ideas of Delaunay (1929) and Dirichlet (1850). Given that there are hundreds of plant species within the catchment, I use a landscape approach, and only sketch the main tree species, and two iconic plants, the balga and the djiridji, both of which are important to the original custodians of the catchment, the Nyoongar people. There is discussion of other people’s research into the effect of bushfire on seed banks, and the flowering intervals of some plants of the jarrah forest.To see if Western Australia is anomalous, or fits into the worldwide pattern of humans using fire as a landscape management tool, I then examine some records of bushfire in other lands, including Africa, Madagascar, India, and Europe. The thesis then looks at the history of fire in the jarrah forest of Western Australia, based on observations by early European explorers and settlers from 1826 onward, the views of various foresters, and some opinions of current Nyoongar Elders.Using a mixture of natural science, applied mathematics, and archaeology, I give the results of cleaning the stems of those ancient plants called grasstrees, or balga (Xanthorrhoea spp.). These carry the marks of former bushfires, stretching back to 1750. They confirm historical reports of frequent fire in the jarrah forest, at 2-4 year intervals, and a recent decline in fire frequency. This contradicts the view, held by some, that European arrival increased the frequency of fire.As support for the balga findings, I present a simple mathematical model of self-organization in bushfire mosaics. It shows how lengthy bushfire exclusion can lead to disastrous situations, in which large areas of landscape become flammable and unstable. It shows how frequent, patchy burning can maintain a stable bushfire mosaic, with mild, beneficial fires. In the next chapter, I offer mathematical suggestions on how current unstable mosaics can be restabilized, by careful reintroduction of such burning.In dry, south-western Australia, water supply is an important topic, and a better understanding of the hydrological effects of bushfire may help with both bushfire and water management. I draw upon the natural science of forest hydrology, and the effects of fire in catchments. The evidence comes not only from Australia, but also from the United States, and South Africa.Turning to social science, I introduce Professor Peter Checkland’s ‘Soft Systems Methodology’, and suggest how it could be applied in resolving complicated conflict about bushfire management. I finish in legal style, with a summing up, and a verdict on the use of bushfire as a land management tool in Wungong Catchment, and possibly in other flammable landscapes.

There is growing understanding of the importance of landscape mosaics and heterogeneity for biodiversity outcomes in Western Australia. However, there is limited information on the patchiness (spatial configuration of unburnt and burnt patches which occur at a range of spatial scales) within the perimeter of a single prescribed burn. Of particular concern is the idea that prescribed burning operations, carried out under very restricted weather and environmental conditions, can lead to structural and floristic homogenization of the area within a burn perimeter. This may be evident as reduced post-fire vegetation patchiness. Western Australian Jarrah (Eucalyptus marginata) forests are managed to reduce fuel loads to protect life and property. Additionally, specific biodiversity and fuel reduction programs are completed by means of low intensity controlled burns carried out by the Parks and Wildlife Service, part of the Department of Biodiversity, Conservation and Attractions. The northern Jarrah forest is thus an ideal study site to test whether prescribed burning homogenizes a within-burn area. The objective of this study was to determine the post-fire patchiness of vegetation following prescribed burning to assess the potential for homogenisation of the within-burn area of Jarrah forests. To achieve this, a set of post-burn surveys were created to determine the patchiness of the burnt areas within 18 prescribed burns. The quantitative post-burn assessment designed in this study was evaluated to determine if it is suitable to be used as a possible future management option. A survey was undertaken to obtain data on 29 environmental variables which capture the variability across an individual prescribed burn area to determine the within-burn fire characteristics. A Pearson correlation matrix table was constructed to determine the significant correlations between variables as well as the strength of the relationships (coefficient of determination, r2). The patchiness between seasons autumn and spring was compared. Generalised Linear Modelling (GLM), and the application of Akaike were used to identify which variables were essential and most significant in predicting patchiness, to produce a list of candidate models. The AICc indicated which models were the most parsimonious or plausible candidate models (AICc < 2). Structured Equation Modelling (SEM) was completed to determine which variables in the ‘most parsimonious’ model were the greatest contributors to the model. It was found that the number of vegetation patches decreased with an increase in fire intensity and percentage of area burnt. Autumn and spring burns were not significantly different in terms of overall patchiness and the majority of patches were found to be small, between 1 – 10 m in length. Surface Moisture Content (SMC) was the only pre-fire condition variable that negatively influenced prescribed burn fire intensity and subsequently patchiness, while time since last burn was found to increase the number of vegetation patches and SMC. The post-fire survey data obtained in this study on the 18 post fire sites was also compared to the post-burn assessments completed by Parks and Wildlife Service on the same sites. GLM and AICc showed that the variables Fire Danger Index, ash cover (%), area burnt (%), month of fire and the number of vegetation patches, are the most parsimonious and ‘best’ fit at predictors of patchiness within a prescribed burn area. While SEM showed that area burnt (%) was the most important predictor of patchiness. Within-burn patchiness appeared to be low in the study sites. Results from the prescribed burns in the northern jarrah forests showed that in most prescribed fires, 90 to 100% of the vegetation area was burnt with a limited number of unburned patches (up to 12 km-1) remaining. Although successful for fuel reduction burning, such large scale vegetation loss has been found to lead to structural homogenisation within a burn area, which in turn can result in long-term impacts on biodiversity. This study thus showed that prescribed burns in the northern Jarrah forest can homogenise the within-burn area and the implications of these findings are discussed in the broader context of landscape patchiness. This study also shows that a quantitative post-fire patchiness survey similar to that designed here should be developed as a future management option.

Invasive predators are major drivers of global biodiversity loss and their impacts may be worsened by other disturbances such as fire. I examined how the fire history of shrublands influences the ecology of feral cats Felis catus, dingoes Canis dingo and their prey species in Western Australia’s northern Wheatbelt region. A review of the literature revealed that feral cats inhabit a diverse range of ecosystems worldwide, but are generally recorded most often in habitat types characterised by a mixture of plant growth forms close to ground level. Cat habitat use is influenced by predation/competition, prey availability, shelter availability and anthropogenic resource subsidies. Relatively few studies were available for review and the strength of evidence contained within them was generally low, which highlighted the need for more rigorous field studies. I examined overlap in resource use between cats and dingoes using remote camera surveys and dietary analysis of scats. Both carnivores were recorded in all four major habitat types: recently burnt shrublands (10 to 14 years since last fire), long unburnt shrublands (34 to ~49 years), very long unburnt shrublands (> 50 years), and woodlands. Dingoes and cats preferred woodlands and very long unburnt shrublands respectively, but spatial overlap between the two species was still common. Mean diurnal activity time for feral cats was two and a half hours later than that of dingoes. The diet of feral cats was more diverse than that of dingoes and dietary overlap between the two carnivores was relatively low. Rabbit remains did occur relatively frequently in both cat and dingo scats, but small mammals, reptiles and birds were also common in cat scats, and macropods in dingo scats. Nine of the 15 prey species studied showed a preference for either recently burnt or long unburnt shrublands. Two small mammals and three reptiles were most abundant in recently burnt areas, while the abundance of one small mammal and three reptiles was highest in long unburnt areas. Using giving up density experiments, I showed that rodents exhibited differential foraging behaviour in the two vegetation fire ages. The rodents foraged for longer in sheltered compared to open microhabitats, but this pattern only occurred in recently burnt, not long unburnt shrublands, probably because the higher density of understorey vegetation in recently burnt areas provided the rodents with extra cover to hide and escape from predators. I also developed a new framework for conceptualising interactions between invasive predators and other ecological disturbances, such as fire, habitat fragmentation, and top -predator decline. The impacts of invasive predators can be classified as either functional (density -independent) or numerical (density -dependent), and they interact with other threats through both habitat -mediated (fire, grazing, land clearing) and community -mediated (top -predator decline, altered prey populations, anthropogenic resource subsidies) interaction pathways. The key findings of this thesis show that both old and young shrublands can be suitable habitat for feral cats; predator -prey dynamics are influenced by successional habitat stages; small mammals show behavioural, as well as population -level responses to fire; and that invasive predator management is likely to benefit from addressing multiple threats in unison.

The establishment of seedling regeneration is a key process in and indicator of ecologically sustainable forest management. The availability of seed and the creation of a suitable seedbed are recognised as important factors limiting seedling recruitment. A silvicultural method commonly used across northern and eastern jarrah forest blocks is shelterwood cutting. The primary objective of treating jarrah forest to shelterwood is to promote seedling regeneration in areas lacking sufficient advanced growth. Despite the widespread and progressive implementation of the shelterwood method, its application in jarrah forest has shown varying degrees of success. This thesis sought to investigate and better understand the roles of seed supply and seedbed condition in promoting successful seedling regeneration in shelterwood-treated jarrah forest. It addressed two questions from an ecological and management perspective. Firstly, could adequate seed supply and favourable seedbed conditions be effectively managed and produced in shelterwood-harvested coupes? Secondly, could adequate seed supply and suitable seedbed conditions be reliably produced to facilitate successful seedling regeneration following disturbance events, in this case post-harvest burning? A major effort was dedicated to developing a more accurate and practical method of assessing seed crops in individual trees. The final model produced a high degree of predictability (R² = 0.85), while still maintaining a high level of practicality for field application, with three easily measured variables being used (stem diameter combined with subjective assessments of capsule clump density and capsule clump distribution). The refined model dramatically improved estimates of crown capsule numbers from the previous model, with the R² value increasing from 0.29 to 0.85. The second major focus of the study was to assess the capacity of prescribed burns, under mild conditions, to produce seedbed conditions suitable for regeneration. Low intensity prescribed burns resulted in the production of suitable conditions for seedling regeneration; that is, leaf litter and understorey vegetation were reduced and ash beds were created. Ash bed production was heterogeneous within sites. This heterogeneity has been attributed to the capacity of low intensity prescribed burns to account for fine-scale variations in fuel quantity, continuity and condition. Patterns of pre-burn aerial seed crop size and seed fall following low intensity prescribed burning were also assessed. Canopy capsule crops showed a high degree of spatial and temporal variability, both in terms of seed quantity and maturation. Such variability has been attributed to individual trees or groups of trees responding differently to localised climatic events and/or interspecific site factors at each stage of the flowering cycle. The main source of this variability was shown to be the numbers and spatial distribution of super trees; that is, trees defined as having a stem diameter >60 cm and >20 000 capsules. The average rate of seed fall increased substantially following prescribed burning under mild conditions. Postharvest burning under the mild conditions of the current survey did not result in en masse seed fall. Rather, peaks in seed fall were observed in the first few weeks post-burn, followed by low level falls throughout the following year. Sites burnt in spring showed a higher and more consistent release of seed in the first few weeks following fire, whereas seed fall after autumn burning was more sporadic. The comparative and interactive roles that seed supply and seedbed conditions play in limiting recruitment of jarrah were also studied. Low seedling densities were recorded across all six burnt study sites. The fact that ample levels of post-burn seed fall produced such low seedling numbers suggested that adequate seed supply did not coincide with seedbed conditions suitable for mass seedling regeneration. Conditions favourable for seedling recruitment were highly variable within sites, since both seed supply and seedbed conditions were spatially heterogeneous. Fine-scale areas burnt to mineral soil showed an additive influence to the overwhelmingly dominant factor of seed supply on seedling recruitment. However, the capacity of low intensity burns to produce these seedbed conditions at a broad scale is limited. Results of this study suggest that successful stocking of shelterwood-treated jarrah forest is not always achievable following a disturbance event, such as post-harvest burning under mild conditions. The chances of a large seed supply coinciding with broad-scale seedbed conditions favourable for mass germination, emergence and establishment appear to be low. Successful stocking of shelterwood-treated jarrah forest is more likely to be a longer term outcome achieved through episodic recruitment, when favourable environmental conditions coincide with optimal seedbed conditions. Such episodic recruitment strategies may be common in resource-limited systems such as jarrah forest and other dry eucalypt forest systems, where conditions controlling the regeneration niche are often variable and unpredictable.

[Truncated abstract] Recovery of soil nutrients, microbial populations and carbon (C) and nitrogen (N) cycling processes are critical to the success of rehabilitation following major ecosystem disturbance. Bauxite mining represents a major ecosystem disturbance to the jarrah (Eucalyptus marginata) forest in the south-west of Western Australia. Mining has created a mosaic of mined areas in various stages of succession surrounded by non-mined forest areas. Initial site preparations within rehabilitation areas such as contour ripping alter soil structure (creation of mound and furrows) and over time also influence the distribution of vegetation and litter. Current performance criteria developed by industry, government and other stakeholders have determined that before post-bauxite mined areas of jarrah forest can be integrated back into normal forest management practises they should be functional and demonstrate resilience to normal forest disturbances such as fire. Furthermore, resilience should be of a manner comparable to non-mined analogue forest sites. Currently little is known of the resilience of microbial communities and C and N cycling in rehabilitation sites to normal forest disturbances such as prescription burning. As such, before rehabilitated jarrah forests can be successfully integrated into broad scale forest management regimes, a more thorough knowledge of the potential impacts of burning practises on the soil microbial community and C and N cycling processes in these systems is required. ... While there are similar rates of C and N cycling the underlying microbial community structure was distinctly different; implying a high degree of functional redundancy with respect to C and N cycling. Differences in the C and N cycling and structure of the microbial communities were likely to be due to differences in soil environmental conditions (i.e. soil alkalinity/acidity, soil moisture) and C substrate availability which influence the physiological status of the microbial community and in turn are related to successional age of the forests. Results also suggest that the measurement of CLPP can be a useful approach for assessment of changes in the functional ability of microbial communities. However, the interpretation of how well these rehabilitation forests have recovered heterotrophic abilities was greatly affected by the methodological approach used (e.g. MicroRespTM or Degens and Harris, 1997). Importantly, results from Chapter 4 and 5 suggested that the effects of a moderate prescription fire on C and N processes, CLPP and microbial community structure of 18 year old rehabilitation forests are likely to be short-lived (< 2 years). Furthermore, the effects of the moderate spring prescription fire were not large enough to decouple C and N cycling processes over the short-term (< 1 years) which suggests that by 18 years of age rehabilitation forests demonstrate comparable functional resilience to a moderate prescription burn.

The northern savannas are one of the largest biomes in Australia, extending across northern and north-eastern Australia, and its vegetation has been extensively altered as a result of pastoralism and inappropriate fire regimes following the loss of indigenous fire management (Fraser, 2001; Legge et al., 2019). Small-medium sized mammals and granivorous birds are the most at-risk taxonomic groups (Franklin, 1999; Franklin et al., 2005; Murphy et al., 2010; Woinarski et al., 2010; Woinarski et al., 2013), and this has largely been attributed to unsuitable fire management, predation by feral cats, grazing by introduced herbivores, cane toads and invasive grasses (Fraser, 2001; Legge et al., 2019; Woinarski et al., 2010). The Partridge Pigeon (Geophaps smithii) is a granivorous bird species endemic to the northern savannas of Australia and has faced major declines disappearing from half of its pre-European distribution (Davies et al., 2019; Franklin, 1999; Fraser et al., 2003), largely due to changes in burning regimes, grazing by exotic herbivores and predation by feral cats. As a result, the Partridge Pigeon is an excellent model species for examining savanna management practices (Davies et al., 2019; Fraser et al., 2003). The aims of this study were: 1) to quantify the changes in the distribution of G. s. blaauwi over time; 2) to assess these changes against IUCN criteria to re-evaluate the species’ current conservation status; 3) to determine the land tenure across the species’ current distribution to evaluate the importance of Indigenous owned and managed lands in the protection of this species; 4) to understand habitat selection of G. s. blaauwi and the influence of fire regimes on habitat selection at landscape scales; 5) and lastly to understand the influence of fine-scale variables on G. s. blaauwi habitat selection. Following my analysis of the changes of G. s. blaauwi range over time I determined that it should still be classed as Vulnerable based on IUCN criteria but recommended that more surveys be undertaken to better assess the poorly surveyed areas of their distribution. Analysis of land tenure indicated that this species mostly occurs on Native Title land (93%) and in areas under Indigenous Protected Areas management (49%). I highlight the importance of Indigenous owned and managed lands for protection of G. s. blaauwi and emphasises the critical role IPA and conservation areas may play in the protection of biodiversity and threatened species in Australia. Next, I was able to determine that the most important geological and vegetation structures for G. s. blaauwi are woodland and open woodland areas that occur on alluvium and colluvium. This allowed me to develop a clear conceptual model of what habitats and fire management practices are required to support G. s. blaauwi populations. Lastly, my investigations of fine-scale habitat variables found none of the assessed variables influenced G. s. blaauwi site occupancy.

Future climate projections suggest an increase in average temperature as well as a decrease in average winter rainfall across the south-west Australian floristic region (SWAFR). These adverse future climatic conditions will amplify the intensity and frequency of disturbance events such as drought and fire. Mediterranean forests within the SWAFR are prone to drought and fire disturbance and have acquired resilience through the selection of drought and fire tolerable species. However, shifts in the magnitude of these disturbance events could increase the recovery period required for recruitment, causing a shift in forest structure and decreasing the resilience of these ecosystems to future disturbances. In this study, we investigated above-ground biomass (AGB) accumulation of understorey plants at sites within the Northern Jarrah Forest (NJF) that have experienced different degrees of drought and fire intensity. We found that within a disturbance event, sites experiencing either more severe drought and fire intensities on average accumulated substantially more understorey AGB than sites subjected to both low drought and moderate fire intensities. This suggests that understorey species within the SWAFR gain a competitive advantage in high drought and fire severity conditions and are highly tolerant to drought. However, the increase in understorey AGB accumulation also suggests a shift in overall forest structure to more dense, compact, low-ground small stems, which is known to increase fire probability. An increase in fire probability shortens the time period between fire intervals and can detrimentally affect forest recovery, especially in drought conditions. Therefore, these changes may shift ecosystems within the SWAFR to a state of non-equilibrium and reduce resilience to future disturbance events.

La dynamique océanique, notamment dans les régions Australes, se caractérise par d'importantes fluctuations induites par les processus à mésoéchelle (tourbillons, 100-200km) et à sous-mésoéchelle (filaments, <50km). Ces processus sont connus pour stimuler fortement le développement du phytoplancton et avec lui l'ensemble de la chaine trophique. Toutefois les relations entre la distribution des proies et celle des prédateurs sont encore loin d'être bien connues pour la plupart des espèces marines. C'est en particulier le cas pour l'éléphant de mer austral (EDM) dont les proies souvent localisées à grandes profondeurs. Cette thèse s’est intéressée à étudier comment le comportement de plongée et de chasse des EDM est influencé par les variations à fine échelle de l'environnement. Une des originalités de cette thèse a été d’utiliser les mesures à très haute résolution récoltées par les éléphants de mer qui constituent un jeu de données 3-D in-situ unique à sous-mésoéchelle, puis de les combiner aux mesures satellites plus classiques de température et d'altimétrie. Ce travail a montré, pour la première fois à l’aide de données in-situ très haute résolution, l’importance des zones frontales à sous-mésoéchelle dans le comportement alimentaire d’un prédateur supérieur. De plus, elle a permis le développement d’une méthode originale d’identification des zones frontales à fine échelle et confirmé la pertinence de la méthode Quasi-Géostrophique de Surface (SQG) dans l’étude de la dynamique fine échelle, notamment dans les perspectives des futures missions spatiales à hautes résolutions (de type SWOT)
Ocean dynamics, especially in the southern ocean, are caracterized by strong fluctuation due to mesoscale (eddies, 100-200km) and submesoscale (filaments, <50km) processes. Theses processes are known to strongly stimulate primary production and with him the rest of the trophic chain. However, in marine ecosystems, relationship between prey and predator distribution remain challenging to understand. Such complexe link exist within the Southern Elephant Seal (SES) and their deep diving prey.This PhD worked to understand how the SES diving and foraging behaviour is impacted by submesoscale variation of the environment. Two original aspects of this work was first to use the very high resolution measurement from SES as an unique 3-D in-situ submesoscale dataset and then to combine it to, more classic, satellite temperature and altimetry measurement. This work showed, for the first time the in-situ impact of submesoscale frontal regions on one top predator foraging behaviour. Furthermore, during this study, an original method to identify fine scale frontal regions was developped. Finally, this PhD confirm the relevant use of the Surface Quasi-Geostrophic method in the study of fine scale dynamics, especially in the possibility of high resolution spatial missions (such as SWOT)

The disruption of natural fire regimes has threatened animal species in many ecosystems around the world. A combination of prescribed burning and fire suppression is often used to promote successional variation in vegetation (i.e. fire mosaics), with little knowledge of how this will affect animal persistence. Understanding the processes that govern species responses to fire regimes is essential to build a predictive capacity for ecological fire management. I examined life-history, demographic (survival, reproduction and mortality) and dispersal attributes of reptiles to investigate mechanisms of fire responses in reptiles. I studied reptiles in conservation reserves of semi-arid southern Australia dominated by mallee vegetation (multi-stemmed Eucalyptus spp. with a shrubby understory). An introduction describes the ecological and management context of my research (Chapter 1). A community-level framework was used to determine if a generalised model of fire responses could be developed based on traits shared by groups of species (Chapters 2-3). I found a number of fire responses in reptiles that were previously undetected in analyses of smaller, but substantial subsets of the same data (Chapter 2). Nocturnal burrowers tended to be early-successional, while diurnal leaf-litter dwellers tended to be late successional, but a trait-based model of succession had limited power to describe responses among the community. I also documented some observations that suggested non-burrowing reptiles were more vulnerable to mortality during wildfire than burrowers (Chapter 3). A species-level framework was then used to examine variation in demographic and dispersal attributes within species among different post-fire successional stages (Chapters 4-7). These studies focussed on three species with significant and contrasting responses to fire: Amphibolurus norrisi (Agamidae; mid/late successional species), Ctenotus atlas (Scincidae; late successional) and Nephrurus stellatus (Gekkonidae; early/mid successional). Using mark-recapture modelling (Chapter 4), I described changes in abundance of N. stellatus that incorporated detectability, and showed that variation in survival and fecundity are possible drivers of this species strong population response to fire. Microsatellite DNA data were then used to examine gene flow in the three target species and gain insights into the effects of fire on dispersal. Chapter 5 begins this section with a description of the markers I used to generate the genetic data. I then used spatial models of landscape resistance to assess the importance of post-fire succession and other landscape features (e.g. topography) on gene flow in the three species (Chapters 6 and 7). For N. stellatus these analyses were combined with direct observations of movement (Chapter 6). Results showed that long-unburnt vegetation restricts dispersal in N. stellatus, which may result from, or contribute to its decline in population density with increasing time since fire. In Chapter 7 I found that fire affected gene flow in A. norrisi, but not in C. atlas, while genetic diversity in both species was affected by post-fire succession. My thesis demonstrated how examining demographic and dispersal attributes of reptiles can give insights into the mechanisms underlying species responses to fire. I concluded by providing management recommendations and highlighting key points for future research on fire ecology (Chapter 8). -- provided by Candidate.

In fire-prone ecosystems, plant species exhibit a range of traits which allow them to persist under certain fire regimes. Two critical traits are resprouting and regeneration from seed. Resprouters and seeders can coexist within plant communities, but their relative proportion varies. This research investigated variation in patterns of community composition, in terms of fire response traits, across a mountainous region of south-eastern Australia. The relative importance of fire regimes, habitat, vegetation cover and resource gradients in influencing these patterns was tested. Associations between leaf traits and fire response types were also explored to investigate whether traits related to fire persistence can be considered as part of a broader spectrum of plant ecological strategies, and assist in explaining landscape patterns. In January 2003 unplanned fires burned through the region, creating the opportunity to investigate the responses of locally occurring plants. An initial study of observed field data demonstrated that some species varied in their response between sites. Furthermore, observed responses often differed from those reported in a fire response trait database. As such, the decision was taken to use observed site-specific data on fire response traits where possible to analyse community composition across the region. Data were subsequently collated from five post-fire surveys conducted across 284 sites in the study area. Resprouting was a dominant response of species in the region. However, woody obligate seeders, although infrequent, were often structurally important. Within-species variation was identified, with variability in post-fire seed regeneration particularly evident in trees. The results unexpectedly identified 'dataset' as the factor most strongly associated with the observed patterns, raising important questions about the use of data collated from different sources. Obligate resprouters were favoured at sites burned with no short intervals between fires, whereas several woody obligate seeding types were in higher proportions at sites burned with one or two short intervals. This suggests that some degree of fire is required for obligate seeder persistence. However, the possibility that woody obligate seeders had already been eliminated from the region prior to this study, due to short inter-fire intervals, could not be discounted. Significant differences between habitats were identified, with woody obligate seeders found in high proportions on rock outcrops and in wet grasslands. However, trends in the proportions of fire response types across habitats could not be reconciled with trends in habitat characteristics such as canopy height and understorey cover. Understorey cover was negatively associated with seeders and facultative resprouters, suggesting an important influence of competition post-fire. Leaf trait comparisons showed that facultative resprouters differed from obligate seeders and obligate resprouters, indicating that this strategy requires an investment into rapid acquisition of resources and fast growth to compete. Differences in leaf traits in species with a canopy seed bank were also identified, raising some important considerations regarding approaches to classifying species by fire response traits. This research illustrates the value of considering the relative importance of fire in relation to other factors for developing a thorough understanding of the processes governing community composition in fire-prone ecosystems. Importantly, the approaches taken in this thesis also revealed a range of issues regarding the application of trait data, which are discussed in relation to trait-based ecological research and management.

Recurrent wildfires strongly affect the distribution and composition of biotic communities, and can also be highly destructive to human populations and infrastructure. Consequently, a number of fire management approaches (such as fire suppression, prescribed burning and selective clearing) are routinely implemented in many countries, with the aim of reducing the risk of catastrophic fires. While some of these approaches are known to be incompatible with the persistence of some plant and animal species, the extent to which inappropriate fire management is a threatening process for frog populations is largely unknown. Any effects of fire on frog populations that do occur are likely to be complex, because fire both kills individual animals, and may also influence frog movement through terrestrial locations (i.e. migration and dispersal). Further research on this topic would also be useful from a scientific perspective, in that frog populations can serve as interesting test cases for the investigation of fire ecology theories, which have typically been constructed for other animal taxa. In this thesis, I address these research gaps on frog responses to fire, through a program of research undertaken in Booderee National Park, southeastern Australia. The resulting papers are arranged by increasing spatial scale, moving from responses of frog assemblages at breeding ponds (chapters I & 2), to frog behavior in the pond margin (chapter 3), to terrestrial movements at the landscape scale (chapter 4). In combination, these studies constitute a detailed investigation of the effects of recurrent fire on frog populations. This body of research represents an in-depth assessment of the viability of ecological theory for explaining variation in frog assemblages following fire.

Fungi are key components in forest ecosystems, being involved in decomposition of plant biomass and the cycling of nutrients in forest soils. Despite their importance little is understood about the influence forest management practices, such as long-term prescribed burning and forest conversion are having on soil fungal communities. Part of the work described in this thesis investigated the effects of long-term repeated prescribed burning on the total soil fungal community, the diversity of mycelial communities of ectomycorrhizal fungi and the influence of biennial prescribed burning on the cellulolytic soil fungal community using stable isotope probing techniques. The influence of long-term repeated prescribed burning on soil fungal communities was investigated through a series of studies conducted at Peachester State Forest, Queensland, Australia. This site has been the centre of a long-term repeated prescribed burning experiment, established since 1972, consisting of plots subjected to biennial, quadrennial or no burning. Denaturing gradient gel electrophoresis (DGGE) was used to show that long-term prescribed burning significantly altered the total fungal community structure in the top 10 cm of soil, when compared with unburned plots. Hyphal ingrowth bags, used to target ectomycorrhizal (ECM) mycelia in soil, along with DGGE analysis, indicated that profiles of the soil fungal community from 2 yr burn plots significantly differed from those of the 4 yr burn and unburned plots. Following analysis of clone assemblages from the different burn regimes, results indicated that this difference reflected an altered ECM fungal community composition. 13C stable isotope probing (SIP), following the incubation of soil with 13C labelled cellulose, and DGGE analysis was found to significantly alter the active fungal community in the upper 10cm of soil at Peachester State Forest. Fewer active fungi in the 2 yr burn plots were found to have incorporated 13C compared to the unburned plots, strongly suggesting that the activities of cellulolytic fungi were negatively affected by the 2 yr burning treatment. The thesis also incorporated work that assessed the effect of forest conversion from native eucalypt to Pinus elliottii plantation on the soil fungal community at Beerburrum State Forest, Queensland, Australia. ITS and 18S RNA and DNA were used, along with terminal restriction fragment length polymorphism (T-RFLP) and DGGE analysis, indicating that total and active fungal communities differed significantly between the native eucalypt forest and first rotation P. elliottii plantation. This suggested that the conversion from native eucalypt forest to P. elliotti plantation significantly altered the soil fungal community at the Beerburrum site.
Doctor of Philosophy (PhD)

Invasive plants cause ecosystem degradation throughout the world, including the reduction of native plant density and diversity, and changes in ecosystem structure and function. Woody weeds often grow faster than native species and in invaded habitats produce larger and/or more seed and outshade other mid- and under-storey species. Boneseed Chrysanthemoides monilifera ssp. monilifera (L.) T. Norl. has caused the degradation of many temperate woodlands in Australia and has not yet reached its full potential distribution in this country. The control of this weed is therefore a high priority in Australia. Biological control agents have not controlled boneseed populations to date and no detailed integrated control strategies exist for different densities of mature boneseed plants and soil seed banks in native vegetation of varying levels of degradation. Fire, herbicides and manual plant removal have previously been used to control boneseed; however, substantial landscape scale control has not yet been achieved. Boneseed population control experiments were undertaken in two temperate woodlands in Victoria, Australia. In highly degraded temperate grassy woodlands at the You Yangs Regional Park west of Melbourne in Victoria and in a highly diverse native closed woodland at Arthurs Seat State Park in south-eastern Victoria. Several combinations of the weed control techniques of fire, herbicide application, hand-pulling of seedlings and distribution of competitive native grasses were found to control both mature boneseed populations and the large reserves of viable boneseed seeds in the soil. The efficacy of controlled burning, and the combination and timing of control techniques were found to vary according to differing densities of boneseed plants, viable soil seed banks and post-fire emergent seedlings. Where sufficient fine fuel existed, a warm, even, autumn burn consumed above ground biomass, killed the majority of viable boneseed seed in the soil, and caused the remaining boneseed seed to germinate. Spraying with glyphosate herbicide was as effective as metsulfuron-methyl herbicide for killing boneseed seedlings along with the secondary climbing weed Billardiera heterophylla (Lindl.) L.W.Cayzer & Crisp after fire. However, the use of glyphosate also killed all native species, resulting in bare ground. After fire in species rich vegetation, boneseed was eliminated where seed of the native C3 grass Poa sieberiana Spreng. had been broadcast onto the post fire ash-bed, and seedlings had been sprayed five months after the burn or where seedlings had been sprayed 12 months after burning. Boneseed control occurred when seedlings were sprayed five months after the burn. In degraded vegetation few boneseed seedlings remained where seedlings were sprayed 17 months after fire. Where insufficient rainfall occurred, hand-pulling flowering boneseed seedlings prevented new seed fall for 6 to 12 months. Suggestions are made for the integration of these methods with the establishment and proliferation of biological control agents. A new protocol for utilising several integrated control strategies for boneseed and other woody weeds in a mosaic at both the site and landscape scale is described. A mosaic would allow for a variety of native species responses to fire and other control methods and thus lead to heterogeneous ages and structures within the native vegetation following weed control.
http://proxy.library.adelaide.edu.au/login?url= http://library.adelaide.edu.au/cgi-bin/Pwebrecon.cgi?BBID=1457770
Thesis (Ph.D.) - University of Adelaide, School of Agriculture, Food and Wine, 2009

Snow is an important factor in the lives of flora and fauna in those regions where it occurs. Despite this, there is a relative lack of information about the ecological role of snow. In addition, on a global scale the majority of the research on snow ecology has been based in the boreal regions of North America and Eurasia. Insights from these areas may not extrapolate well to Australia. The distribution and physical characteristics of snow are highly variable both temporally and spatially. Its occurrence is affected by a range of factors acting at multiple scales. Working in snow covered areas, however, presents considerable practical problems, particularly for researchers attempting to sample organisms in the subnivean space between the base of the snowpack and the ground surface. As a result, most research has focused on small-scale projects because of logistical and animal-welfare issues. A technique was developed for this study for sampling small mammals beneath the snow using hairtubes fitted with bait and a removable adhesive surface that could be inserted into the subnivean space through a vertical PVC pipe. The technique provided a 39% detection rate with only 0.2% of tubes visited but not collecting hair samples. Using. this technique, it was possible to expand systematic sampling of small mammals in the subnivean space to larger scales at which snow cover can vary spatially and temporally particularly at the landscape scale. The main part of this research was conducted over two winters (2002-2003) at sites established in a series of valleys close to the Summit Road in Kosciuszko National Park, south-eastern Australia. Selection of sites was based on factors considered important in influencing the distribution of snow in the landscape and representative of the key vegetation types occurring in the subalpine zone. The resulting design consisted of 72 sites stratified by elevation (1501-1600 m, 1601-1700 m, 1701-1800 m), aspect (accumulating, ablating) and vegetation type (woodland, wet heath, dry heath, grassland) with each combination replicated three times. Each site consisted of three hairtube plots approximately 10 metres apart, at which small mammals were sampled. In addition, a range of biotic and abiotic factors including snow cover characteristics were measured throughout the winter at these same sites. In January 2003, a major bushfire burned 70% of the subalpine area of Kosciuszko National Park and damaged 83% of the sites established in 2002. As a result sampling during winter 2003 was limited to high elevation sites, along with a fifth habitat type (boulderfields). The snow cover that occurs in the main alpine and subalpine region of the· Snowy Mountains is primarily maritime in areas where there is sufficient accumulation, and ephemeral at lower elevations and ort higher ablating aspects. Maritime snow is generally deep (> lOOcm), with a density >0.30gcm-3, as a result of destructive metamorphism throughout the winter. The formation of depth hoar, which is considered to be important in acilitating the development of the subnivean space, does not occur under these conditions. Ephemeral snow is characterised by warm shallow snow that often melts before new snow is deposited. When snow was present, detections of dusky antechinus, Antechinus swainsonii and the bush rat, Rattus fuscipes were negatively correlated with snow depth and duration, and positively correlated with the complexity of structures and microtopography. At high elevations, detections were largely confined to boulderfields, and at mid- and low elevations, small mammals were detected primarily in habitats where the subnivean space was most extensive. Antechinus swainsonii and R. fuscipes responded differently to snow cover with the latter seeming better able to overwinter where snow cover was shallow and patchy. In contrast, A. swainsonii occurrence was correlated with the size of the subnivean space. The development of the subnivean space in the Snowy Mountains is dependent on the presence of structures such as shrubs, boulders and microtopographic features that are capable of supporting a snow layer above ground level. The temperature in the subnivean space was virtually constant beneath the snowpack, ranging between 0 and + 1 °C. When snow was patchy or absent, temperatures at ground level were highly variable with a minimum as low s -13°C and maximum as high as +47.5°C. Antechinus swainsonii and R. fuscipes were detected more regularly at sites that were thermally variable. At sites with deep and persistent snow cover (maritime snow), subnivean temperatures were stable, but small mammals were detected at low frequencies. At high elevations, boulderfields were favoured by small mammals during the nival period but were no different thermally from other habitats. The limitations imposed by snow cover on small mammals were further verified by a radio tracking study conducted during 2003 at Perisher Creek. That study investigated the home range size and activity patterns of R. fuscipes and A. swainsonii in relation to snow cover. Once continuous snow cover became established, the home range of both species contracted dramatically and there was an increase in home range overlap. Neither species showed any change in diurnal activity patterns. Rattus fascipes showed signs of social interaction during both seasons in contrast to A. swainsonii, which appeared to remain solitary. In winter, R. fuscipes nested communally at a single location, while during autumn the species appeared to use a number of nest sites. There was no significant change in daily activity patterns between autumn and winter in either species. R. fuscipes remained primarily nocturnal during both pre-nival and nival periods while A. swainsonii continued to be active throughout the diel cycle, although there was a slight shift in its peak activity time. Human activities can, have significant effects on the subnivean space and its residents. The physical characteristics of a range of modified snow types were investigated in the vicinity of several ski resorts in Kosciuszko National Park. Human activities associated with snow-based recreation, such as the creation of ski pistes, surface ski lifts and over-snow routes, involve compression of the snowpack and resulted in small or absent subnivean spaces and high snow cover densities compared to unmodified snow cover. To test the effects of the loss of the subnivean space on small mammals, the snowpack was experimentally compressed in high quality subnivean habitats. Detections of R. fascipes and A. swainsonii declined by 75-80%. Burnt sites from the 2002 study were used in 2003 to investigate the effect of removing vegetation on the subnivean space, to simulate the loss of structure associated with ski slope preparation. There was a significant reduction (p<0.0001) in the size of the subnivean space compared to unburnt sites regardless of habitat type. The key conclusions of the work reported in this thesis are listed below: • Snow conditions in the Australian Alps are markedly different from those of higher northern latitudes and altitudes. As a result, conclusions about snow/fauna interactions based on research in regions with particular snow cover types need to be carefully considered before attempting to extrapolate generalisations to other parts of the world. • The subnivean space can be formed either by passive or active processes. The former occurs when there are sufficient competent structures to permit the support of the snow pack above the ground surface, while the latter refers to the ability of small mammals to actively tunnel through relatively low density snow (depth hoar) and thus create their own subnivean space. In Australia the passive process dominates. • The widely held assumption that small mammals are dependent on the thermally stable conditions in the subnivean space was not confirmed. Rattus fascipes and A. swainsonii survive in the Australian Alps because they are able to exploit thermally variable environments. • Management of human activities in nival areas should focus on avoiding disturbance in areas where a subnivean space forms, particularly in high quality winter habitats such as boulderfields. • Global warming resulting from climate change is likely to provide conditions in the Australian Alps that favour an expansion of the distribution and population of R. fuscipes and A. swainsonii, but nival endemics such as Burramys parvus and possibly Mastacomys fuscus may be at a disadvantage. • The extent of alpine and subalpine environments in Australia will decrease in future, imposing greater pressure on a shrinking resource and raising the possibility of conflict between user groups and conservation imperatives.

Fire occurrence influences the distribution of plant species, and dynamics of plant populations, either independently from other factors or in interaction with them. Numerous studies have identified the effects of components of fire regimes (frequency, intensity and season of occurrence) on the population dynamics of individual plant species and the floristic composition of plant communities, both in Australia and in other fire-prone countries. Nevertheless, there has been considerably less research into understanding the causes of spatial variation in fire regimes and this will likely result in a major obstacle for the development of vegetation theory. Research into spatial and temporal patterns of fire regime and determining the extent to which this type of variation results in variation in the occurrence of plant species and hence the composition of plant communities, can help overcome this obstacle. Two hypotheses are constructed and addressed. They are:- (i) that because of the influence of a sites' neighbourhood, variation in fire regimes will exist for any particular set of sites that occupy a particular part of environmental space and are therefore otherwise similar, and (ii) that this variation in fire regimes will result in patterns of plant species occurrence and hence demonstrate that landscape induced pattems of fire regime are a fundamentally important component in determining the realised niche of plant species in spatially complex landscapes. These hypotheses are examined for a spatially complex landscape in the Australian Capital Territory region, Australia. Several distinct phases of research led to the conclusion that both hypotheses should be accepted for the study region. Firstly, a review of the available literature found that empirical approaches, and related statistical models, for determining long-term fire regimes provided data that was neither of sufficient length and accuracy nor of appropriate spatial resolution for examining landscape dependent patterns in fire regimes as outlined in hypothesis one. On the other hand, theoretical approaches which synthesise landscape patterns from the well­ understood processes that affect fire occurrence and behaviour proved to be a useful methodology, provided that the inadequacies in existing models, including their reliance on the North American approach to modelling fire spread, were addressed. These provisions were addressed by the construction of a new landscape-level process­ based dynamic simulation model known as FIRESCAPE. The development of the model involved parameterising and testing the Richardson weather generator for fire danger modelling and a re-assessment of the McRae lightning ignition model. The approach used in FIRESCAPE combines these models with existing models of terrain, solar radiation budgets, fuel moisture, soil moisture, fuel accumulation and fire spread to model spatial variation in fire regimes by the accumulation of information from spatially and temporally distinct fire events. Various analyses indicated that around 500 years was required to stabilise the variation associated with fire regime parameters for some sites, while sensitivity analysis found that the spatial variation in fire regime was not over-sensitive to any of the important fire spread parameters that summarise variation in the majority of input parameters. Further, the spatial pattern in fire regime was generally consistent with many general observations made in the real world, but v under-predicted the fire frequency of five stands of Eucalyptus pauciflora wh,ere detailed pyrodendrochronological measurements have been made. Model output demonstrated that considerable variation in fire frequency existed for all sets of environmentally similar sites that were analysed. This variation was la1rgely related to the position of the site with respect to its neighbourhood, and to a lesser extent the average annual solar radiation budget of the neighbourhood. This variation resulted in strong patterns of eucalypt species occurrence, as determined by generalised linear modelling of species occurrence in the lower Cotter River catchment. Many of the patterns were demonstrated to be ecologically sensible compared with general observations and independent models of the effects of variable fire frequency on the composition of plant communities. Therefore, it has been demonstrated that variation in fire regimes that arises from topographically complex landscapes can have important effects on the distribution of plant species within suites of environmentally similar sites. This is a major step forward in resolving the landscape-level effects of fire occurrence on plant species distribution.