Search

The influences of wildfire through population dynamics and life history for two species of small mammals in a south-east Queensland heathland on Bribie Island are presented. Trapping results provided information on breeding, immigration and movement of Melomys burtoni (Grassland melomys) and Rattus lutreolus (Swamp rat). We first investigated and optimized the design of trapping methodology for producing mark-recapture population estimates to compare two adjacent populations, one of which was subjected to an extensive wildfire during the two year study. We consider how well rodents survive wildfire and whether the immediate impacts of fire or altered habitat have the greatest impact on each species. We found the R. lutreolus population was far more influenced by the fire than the M. burtoni population both immediately after the fire and over 18 months of vegetation recovery.

Summary Ecologists rely on models to explore tree compositional changes especially when occurring over decades or centuries. In this article, we construct a theoretical, mathematical model to investigate the long-term relationships between savanna stand structure, water resource availability and fire disturbance. We show how dry season length, rather than mean annual precipitation, leads to savanna stability, and how the soil properties and variation in annual rainfall distribution determines a climatic equilibrium for woody total basal area, with fire disturbance acting as a perturbation away from this state. This leads to our premise that rainfall and tree population dynamics drives the savanna state leading to grasses and fire, rather than grasses promoting fire to drive the savannas. The model predicts that savanna tree stands undergo cyclic variation in tree populations as a result of long-term population cycles and germination events. This outcome is true regardless of the presence of fire, however, fire does introduce a more heterogeneous stand size structure. Synthesis. Using a mathematically transparent model for water resource availability and stand structure in savannas, we demonstrate how seasonal rainfall distribution, specifically seasonal drought, acts as the primary determinant for stand structure through stand water dynamics, with frequent fire disturbance able to reduce the populations from the climatically induced state.

Invasive species pose a substantial risk to native biodiversity. As distributions of invasive species shift in response to changes in climate so will management priorities and investment. To develop cost-effective invasive species management strategies into the future it is necessary to understand how species distributions are likely to change over time and space. For most species however, few data are available on their current distributions, let alone projected future distributions. We demonstrate the benefits of Bayesian Networks (BNs) for projecting distributions of invasive species under various climate futures, when empirical data are lacking. Using the introduced pasture species, buffel grass (Cenchrus ciliaris) in Australia as an example, we employ a framework by which expert knowledge and available empirical data are used to build a BN. The framework models the susceptibility and suitability of the Australian continent to buffel grass colonization using three invasion requirements; the introduction of plant propagules to a site, the establishment of new plants at a site, and the persistence of established, reproducing populations. Our results highlight the potential for buffel grass management to become increasingly important in the southern part of the continent, whereas in the north conditions are projected to become less suitable. With respect to biodiversity impacts, our modelling suggests that the risk of buffel grass invasion within Australia’s National Reserve System is likely to increase with climate change as a result of the high number of reserves located in the central and southern portion of the continent. In situations where data are limited, we find BNs to be a flexible and inexpensive tool for incorporating existing process-understanding alongside bioclimatic and edaphic variables for projecting future distributions of species invasions.

In this study, systematic variation in tree morphology across a rainfall gradient in Australia's tropical savanna biome and its implications for carbon stocks and dynamics were quantified. The aim was to support efforts to manage fire regimes to increase vegetative carbon stocks as a greenhouse gas mitigation strategy. The height of trees for a given trunk diameter declines with decreasing rainfall from 2000 to 300 mm and increasing dry season length across the Australian savanna biome. It is likely that increasing dry season length is the main driver of this decline rather declining rainfall per se. By taking account of the response of total basal area to rainfall and soil type, stand structure, and tree height and diameter relationships, the carbon stocks in live trees were estimated to decline from about 34 t ha−1 in the wetter savannas to 6 t ha−1 in the drier savannas. These values are broadly consistent with field-based estimates. Because of the declining ratio of height to trunk diameter, trees of a given diameter in drier regions will be more likely to be killed by fires of a given intensity than trees in wetter regions. Thus single fires of given intensity are likely to have a greater proportionate impact on live tree carbon stock in drier savannas, but a much greater absolute impact in wetter savannas due to the greater total carbon stock. Projected decreases in early wet season rainfall under climate change scenarios, despite projections of little change in total precipitation in northern Australia, may lead to decreased carbon stock in live trees through two mechanisms: a reduction in total basal area and decreases in tree height for given trunk diameters.

We examined Pi2 pulsations observed simultaneously at low-latitude stations (L = 1.15 − 2.33) around the dawn terminator. Those Pi2 pulsations observed in the sunlit region were polarized in the azimuthal (D, positive eastward) direction. We found that the D component oscillations in the dark and sunlit regions were in antiphase, whereas the H component oscillated in phase. A statistical analysis indicates that these D component phase reversals occurred about 0.5 h sunward of the dawn terminator at 100 km in altitude, corresponding to the highly conducting E layer. The azimuthal polarization and D component phase reversals related to the dawn terminator cannot be explained by the existing models of low-latitude Pi2s (e.g., cavity resonance or substorm current wedge oscillations). Similar D component phase reversals were also found on the dusk side although the amplitude of the D component is smaller than that of the H component. We suggest that the meridional ionospheric current in the sunlit region adjacent to the dawn terminator drives the D component oscillations in antiphase with those D oscillations produced by the oscillatory field-aligned current (FAC) on the postmidnight side. The meridional current is expected to form a part of a current system that extends from the postmidnight FAC to the equatorial Cowling current. The D component oscillations in the Northern and Southern Hemispheres are also in antiphase, indicating that the current system is symmetric with respect to the equator.

Previous estimates of greenhouse gas emissions from Australian savanna fires have incorporated on-ground dead wood but ignored standing dead trees. However, research from eucalypt woodlands in southern Queensland has shown that the two pools of dead wood burn at similar rates. New field data from semiarid savannas across northern Australia confirmed that standing dead trees comprise about four times the mass of on-ground dead wood. Further, the proportion of total woody biomass comprising dead wood increases with decreasing fire frequency and a decreasing proportion of late dry season (August to December) fires. This gives scope for increasing the carbon stock in the dead wood pool with a reduced fire frequency. Following a previously published approach to quantify total dead wood loads in savannas, new and previously collected data on tree stand structures were used across the whole savanna zone to quantify dead wood loads in equilibrium with historic fire regimes. New parameters are presented for calculating dead wood dynamics including dead trees in Australia’s savannas.

We propose and demonstrate a new approach for the simulation of woody ecosystem stand dynamics, demography, and disturbance-mediated heterogeneity suitable for continental to global applications and designed for coupling to the terrestrial ecosystem component of any earth system model. The approach is encoded in a model called Populations-Order-Physiology (POP). We demonstrate the behavior and performance of POP coupled to the Community Atmosphere Biosphere Land Exchange model (CABLE) applied along the Northern Australian Tropical Transect, featuring gradients in rainfall and fire disturbance. The model is able to simultaneously reproduce observation-based estimates of key functional and structural variables along the transect, namely gross primary production, tree foliage projective cover, basal area, and maximum tree height. Prospects for the use of POP to address current vegetation dynamic deficiencies in earth system modeling are discussed.

Aim Many mangrove communities form bands parallel to the shoreline with each community dominated by a single species. However, the key determinants of mangrove species distribution across the intertidal zone are not well understood. We aimed to quantify the relationship between species' dominance and the hydroperiod (defined as the duration of inundation in a year), soil salinity and the salinity of inundating water for three dominant species, Sonneratia alba, Rhizophora stylosa and Ceriops tagal. Location An extensive (20,000 ha), largely intact mangrove forest in northern Australia, of some note as mangrove forests are threatened globally. Methods We related species dominance to the explanatory variables by applying two statistical modelling approaches: generalized linear models (GLMs), where a set of competing models were evaluated; and boosted regression tree models (BRTs), an approach that automatically captures interactions and nonlinear relationships between variables. Results Both GLM and BRT models achieved strong predictive performance for all species based on cross-validation, with receiver operating characteristics above 0.85 for all species, and 88% of deviance explained for S. alba, 42% for R. stylosa and 35% for C. tagal. All models indicated that the hydroperiod was the key variable influencing distribution, followed by soil salinity. The salinity of inundating water was the least informative variable in the models. Ecological space, determined by gradients in hydroperiod and soil salinity, was partitioned between the three species with little overlap. Main conclusions As anticipated changes in sea level will alter the hydroperiod, our findings are critical for global forecasting of future distributions of mangrove communities, and for the design of mitigation and adaptation measures.

Species distribution modelling (SDM) is a widely used tool and has many applications in ecology and conservation biology. Spatial autocorrelation (SAC), a pattern in which observations are related to one another by their geographic distance, is common in georeferenced ecological data. SAC in the residuals of SDMs violates the ‘independent errors’ assumption required to justify the use of statistical models in modelling species’ distributions. The autologistic modelling approach accounts for SAC by including an additional term (the autocovariate) representing the similarity between the value of the response variable at a location and neighbouring locations. However, autologistic models have been found to introduce bias in the estimation of parameters describing the influence of explanatory variables on habitat occupancy. To address this problem we developed an extension to the autologistic approach by calculating the autocovariate on SAC in residuals (the RAC approach). Performance of the new approach was tested on simulated data with a known spatial structure and on strongly autocorrelated mangrove species’ distribution data collected in northern Australia. The RAC approach was implemented as generalized linear models (GLMs) and boosted regression tree (BRT) models. We found that the BRT models with only environmental explanatory variables can account for some SAC, but applying the standard autologistic or RAC approaches further reduced SAC in model residuals and substantially improved model predictive performance. The RAC approach showed stronger inferential performance than the standard autologistic approach, as parameter estimates were more accurate and statistically significant variables were accurately identified. The new RAC approach presented here has the potential to account for spatial autocorrelation while maintaining strong predictive and inferential performance, and can be implemented across a range of modelling approaches.

My mother and my father taught me a lot, how to look after this land. Riding around this country on horseback we used to go to many places, burning along the way looking after the land. Times have changed, our country has changed a bit, but I still have good knowledge and it is my time to pass on what I have to my children and grandchildren. My mother and father would say if you look after this country, this country will look after you. Violet Lawson

In this study, key ecological modelling limitations of a process-based simulation model and a Bayesian network were reduced by combining the two approaches. We demonstrate the combined modelling approach with a case study investigating increases in woody vegetation density in northern Australia's tropical savannas. We found that by utilising the strengths of a simulation model and a Bayesian network we could both forecast future change in woody vegetation density and diagnose the reasons for current vegetation states. The local conditions of climate, soil characteristics and the starting population of trees were found to be more important in explaining the likelihood of change in woody vegetation density compared to management practices such as grazing pressure and fire regimes. We conclude that combining the strengths of a process and BN model allowed us to produce a simple model that utilised the ability of the process model to simulate ecosystem processes in detail and over long time periods, and the ability of the BN to capture uncertainty in ecosystem response and to conduct scenario, sensitivity and diagnostic analysis. The overall result was a model that has the potential to provide land managers with a better understanding of the behaviour of a complex ecosystem than simply utilising either modelling approach in isolation.

The health of arid and semiarid lands needs to be monitored, particularly if they are used to produce food and fiber, and are prone to loss of vegetation cover and soil. Indicators of landscape health based on remotely sensed data could cost-effectively integrate structural and functional attributes of land surfaces across a range of scales. In this paper, we describe a new index for remotely monitoring changes in the health of land. The new index takes important aspects of landscape structure and function into account by focusing on the potential for landscapes to lose or ‘leak’ (not retain) soil sediments. We combined remotely sensed vegetation patchiness data with digital elevation model (DEM) data to derive a quantitative metric, the landscape leakiness index, LI. This index is strongly linked to landscape function by algorithms that reflect the way in which spatial configuration of vegetation cover and terrain affect soil loss. Linking LI to landscape function is an improvement on existing indicators that are based on qualitatively assessing remotely sensed changes in vegetation cover. Using archived Landsat imagery and Shuttle Radar Topography Mission DEMs, we found for example that LI indicated improvements in the condition or health of a rangeland paddock that was monitored from 1980 to 2002. This paddock is located in central Australia and its improved health is documented by photographs and field data. Although the full applicability of LI remains to be explored, we have demonstrated that it has the potential to serve as a useful ecological indicator for monitoring the health of arid and semiarid landscapes.

The tropical savanna landscape of northern Australia is perhaps the most extensive and flammable ecosystem in the world. Like savannas worldwide, there still remains uncertainty as to the functioning of much of this system and the mechanisms maintaining a state between open grassland and woodland. We use the development of a process-based model to examine the processes operating on the eucalypt dominated tree component of the northern Australian savannas, focussing on the interaction of water availability and fire. A process-based, tree population dynamics model ( Flames) has been developed to understand the interactive effects of fire and rainfall variability on trees in Australian savanna ecosystems where there has been limited modelling to date. This paper shows that the model is capable of simulating the range of tree populations found in northern Australia by considering soil properties (texture and depth) and rainfall distribution. The model indicates that seasonal variability and inter-annual variability in rainfall is critical in defining the structure and density of tree stands in northern Australia. We also show that fire plays an important role in defining tree structure by removing the most sensitive (the very young and old) trees and that frequent fronting fires will lead to ongoing reductions in tree populations. While fire is an important factor in northern Australia, the model indicates its effects on tree populations at a sub-continental scale are secondary to the impact of rainfall variability.

A new continuous, cover-based, directional leakiness index, CDLI, is described that has a number of advantages over a binary-based, directional leakiness index, DLI, previously described in this journal. These indices are monitoring tools aimed to indicate the potential for gently sloping, arid and semi-arid landscapes to retain, not leak, resources, such as soils. To compute DLI, pixels in remotely sensed images had to be classified as being either vegetation patches or open interpatches. This simple binary classification procedure is unrealistic for images with relatively large pixels (e.g., 30-m pixel Landsat), where a single pixel is likely to be a mix of denser vegetation patches and more open interspaces. This mix of patch and interpatch within a pixel can be represented as proportional ground-cover. Because CDLI is based on these ground-cover data, which can vary continuously between 0–100%, it is an index of landscape leakiness that is applicable to larger scale, remotely sensed imagery. Using Landsat imagery, we illustrate how CDLI is useful for comparing the potential leakiness of four Australian arid and semi-arid sites differing in rangeland condition. When imagery is available for specific landscape sites through time, CDLI can be used to monitor changes in the potential leakiness of these sites. However, both CDLI and DLI assume that resource flows are one-directional, which restricts their application to images acquired from relatively gentle, uniform landscapes. A leakiness index applicable to landscapes with hilly terrain is being developed.

This paper highlights the relevant issues influencing the amount and arrangement of ground cover in savanna rangelands in Australia, and presents field measurements from hillslope scale flumes, which demonstrate how runoff and sediment loss vary with spatial patterns in ground cover. Hillslopes with relatively high mean cover, but with small patches bare of vegetation, are shown to have between 6 and 9 times more runoff, and up to 60 times more sediment loss than similar hillslopes that do not contain bare patches. The majority of sediment lost from the hillslopes is composed of fine (suspended) rather than coarse (bedload) material, although the absolute sediment loads are comparatively low. These low loads are considered to be the result of lower than average rainfall during the measurement period (2002–2005) and the high and prolonged rates of historical hillslope erosion that have exhausted the erodible material from the A-horizon. The collected data also demonstrate that a large proportion of soil is lost during the initial ‘flushing’ period of runoff events. The results presented have important implications for the management of savanna grazing systems by highlighting (i) the significance of bare patches in contributing to runoff and soil loss from hillslopes; (ii) the importance of having medium to high cover patches at the bottom of hillslopes for trapping and storing sediment and therefore reducing its entry into the stream network; and (iii) how maintenance of ground cover during the dry season reduces sediment concentrations in runoff occurring early in the wet season. Copyright  2006 John Wiley & Sons, Ltd.

In arid and semi-arid rangeland regions, landscapes that trap and retain resources, such as rain water, soil particles, and organic matter, provide more favorable habitats for vegetation and fauna, and are considered more functional than landscapes that lose, or leak, these essential resources. The cover and arrangement of perennial vegetation patches is an important indicator of whether landscapes retain or leak resources. Patchiness attributes, as descriptors of resource retention potential in landscapes, can be obtained from remotely-sensed imagery, such as aerial videography and high-resolution satellites where this imagery has been classified into perennial vegetation patch and open interpatch pixels. In this paper, we compare four landscape patchiness metrics on their ability to indicate how well landscapes potentially function to retain resources. Landscape patch attributes (e.g. patch cover and spacing) and on-ground inspection of soil and vegetation attributes were used to rate and rank four sites relative to their potential to retain resources. A directional leakiness index (DLI) that is highly sensitive to patch cover, shape, and configuration correctly and adequately ranked sites in the same order as our field ratings. The lacunarity index also correctly ranked sites, but showed little separation amongst sites with reduced potential to retain resources. The weighted mean patch size (WMPS) index and proximity index failed to correctly rank sites. The directional leakiness and lacunarity indices can be calculated for any remotely-sensed imagery that is of sufficient resolution to measure landscape patchiness at scales where processes of resource conservation are operating. For example, imagery of 0.2–1 m pixel sizes from arid and semi-arid rangelands can be classified into flow-obstructing patches and open non-obstructing interpatches. Such classified imagery and leakiness or lacunarity indicators can then be used to monitor changes in the resource retention potential of these landscapes. However, the applicability of these indicators for monitoring more humid vegetation types, and for assessing larger landscape areas (i.e. at coarser scales), needs to be evaluated.

The cover, number, size, shape, spatial arrangement and orientation of vegetation patches are attributes that have been used to indicate how well landscapes function to retain, not ‘leak’, vital system resources such as rainwater and soil. We derived and tested a directional leakiness index (DLI) for this resource retention function. We used simulated landscape maps where resource flows over map surfaces were directional and where landscape patch attributes were known. Although DLI was most strongly related to patch cover, it also logically related to patch number, size, shape, arrangement and orientation. If the direction of resource flow is multi-directional, a variant of DLI, the multi-directional leakiness index (MDLI) can be used. The utility of DLI and MDLI was demonstrated by applying these indices to three Australian savanna landscapes differing in their remotely sensed vegetation patch attributes. These leakiness indices clearly positioned these three landscapes along a function-dysfunction continuum, where dysfunctional landscapes are leaky (poorly retain resources).

Dominant species influence the composition and abundance of other species present in ecosystems. However, forecasts of distributional change under future climates have predominantly focused on changes in species distribution and ignored possible changes in spatial and temporal patterns of dominance. We develop forecasts of spatial changes for the distribution of species dominance, defined in terms of basal area, and for species occurrence, in response to sea level rise for three tree taxa within an extensive mangrove ecosystem in northern Australia. Three new metrics are provided, indicating the area expected to be suitable under future conditions (Eoccupied ), the instability of suitable area (Einstability ) and the overlap between the current and future spatial distribution (Eoverlap ). The current dominance and occurrence were modelled in relation to a set of environmental variables using boosted regression tree (BRT) models, under two scenarios of seedling establishment: unrestricted and highly restricted. While forecasts of spatial change were qualitatively similar for species occurrence and dominance, the models of species dominance exhibited higher metrics of model fit and predictive performance, and the spatial pattern of future dominance was less similar to the current pattern than was the case for the distributions of species occurrence. This highlights the possibility of greater changes in the spatial patterning of mangrove tree species dominance under future sea level rise. Under the restricted seedling establishment scenario, the area occupied by or dominated by a species declined between 42.1% and 93.8%, while for unrestricted seedling establishment, the area suitable for dominance or occurrence of each species varied from a decline of 68.4% to an expansion of 99.5%. As changes in the spatial patterning of dominance are likely to cause a cascade of effects throughout the ecosystem, forecasting spatial changes in dominance provides new and complementary information in addition to that provided by forecasts of species occurrence.

Savannas occur across all of northern Australia and are extensively used as rangelands. A recent surge in live cattle exports to Southeast Asia has caused excessive grazing impacts in some areas, especially near watering points. An important ecological and management question is “how resilient are savanna ecosystems to grazing disturbances?” Resilience refers to the ability of an ecosystem to remain in its current state (resist change) and return to this state (recover) if disturbed. Resilience responses can be measured using field data. These responses can then be modelled to predict the likely resistance and recovery of savannas to grazing impacts occurring under different climatic conditions. Two approaches were used to model resilience responses. First, a relatively simple mathematical model based on a sigmoid response function was used. This model proved useful for comparing the relative resilience of different savanna ecosystems, but was limited to ecosystems and conditions for which data were available. Second, a complex process model, SAVANNA, was parameterised to simulate the structure and function of Australian savannas. Simulations were run for 50 years at two levels of grazing to evaluate resistance and then for another 50 years with no grazing to evaluate recovery. These runs predicted that savanna grasslands were more resistant to grazing (changed less) than red-loam woodlands, which recovered relatively slowly from grazing impacts. The SAVANNA model also predicted that these woodlands would recover slightly slower under the climate change scenario projected for northern Australia. Keywords: Savannas, Modelling, Grazing, Recovery

As savannas are widespread across northern Australia and provide northern rangelands, the sustainable use of this landscape is crucial. Both fire and grazing are known to influence the tree–grass character of tropical savannas. Frequent fires open up the tree layer and change the ground layer from perennials to that dominated by annuals. Annual species in turn produce copious quantities of highly flammable fuel that perpetuates frequent, hot fires. Grazing reduces fuel loads because livestock consumes fuel–forage. This trade-off between fire and grazing was modelled using a spatially explicit, process-orientated model (SAVANNA) and field data from fire experiments performed in the Victoria River District of northern Australia. Results of simulating fire (over 40 years) with minimal or no grazing pressure revealed a reduction in the shrub and woody plants, a reduction in grasses, and no influence on the tree structure given mild fires. While mature trees were resistant to fire, immature trees, which are more likely associated with the shrub layer, were removed by fire. The overall tree density may be reduced with continual burning over longer time periods because of increasing susceptibility of old trees to fire and the lack of recruitment. Increases in stocking rates created additional forage demands until the majority of the fuel load was consumed, thus effectively suppressing fire and reverting to the grazing and suppressed fire scenario where trees and shrubs established. Keywords: Fire, Grazing, Tropical savannas, Modelling

Species distribution models (SDMs) are widely used to forecast changes in the spatial distributions of species and communities in response to climate change. However, spatial autocorrelation (SA) is rarely accounted for in these models, despite its ubiquity in broad-scale ecological data. While spatial autocorrelation in model residuals is known to result in biased parameter estimates and the inflation of type I errors, the influence of unmodeled SA on species' range forecasts is poorly understood. Here we quantify how accounting for SA in SDMs influences the magnitude of range shift forecasts produced by SDMs for multiple climate change scenarios. SDMs were fitted to simulated data with a known autocorrelation structure, and to field observations of three mangrove communities from northern Australia displaying strong spatial autocorrelation. Three modeling approaches were implemented: environment-only models (most frequently applied in species' range forecasts), and two approaches that incorporate SA; autologistic models and residuals autocovariate (RAC) models. Differences in forecasts among modeling approaches and climate scenarios were quantified. While all model predictions at the current time closely matched that of the actual current distribution of the mangrove communities, under the climate change scenarios environment-only models forecast substantially greater range shifts than models incorporating SA. Furthermore, the magnitude of these differences intensified with increasing increments of climate change across the scenarios. When models do not account for SA, forecasts of species' range shifts indicate more extreme impacts of climate change, compared to models that explicitly account for SA. Therefore, where biological or population processes induce substantial autocorrelation in the distribution of organisms, and this is not modeled, model predictions will be inaccurate. These results have global importance for conservation efforts as inaccurate forecasts lead to ineffective prioritization of conservation activities and potentially to avoidable species extinctions.

We developed a gain–loss approach to estimating the dynamics of the dead organic matter pool in northern Australian savannas consistent with the Intergovernmental Panel on Climate Change 2006 guidance. Previously, only emissions of methane and nitrous oxide were accounted in greenhouse gas budgets. The new approach is based on a modification of the Olson fuel accumulation equation. This modification includes the mean post-fire residue and accounts for losses from both fire and decomposition and for gains from grass production, tree litter fall and the death of trees. We apply the approach to a case study in western Arnhem Land, Northern Territory, Australia, in which the fire regime has been changed and the project has achieved recognised abatement of methane and nitrous oxide emissions. The carbon sequestration in the dead organic matter is ~85 times the annual emissions abatement in this project area and when expressed as an annual rate is ~3.5 times the abatement. We also show that the emissions abatement has previously been underestimated owing to an inability to fully account for dead organic matter dynamics. Future work could refine the models of dead organic matter to include seasonal dynamics in inputs.

Processes allowing coexistence of trees and grasses in tropical savannas have long intrigued ecologists. Early theories focused on climatic controls, but a conceptual model has emerged suggesting that savanna trees are subject to a fire-mediated recruitment bottleneck, with frequent fires preventing recruitment of saplings into the tree layer and maintaining biomass well below its climate-determined upper bound. We propose that this conceptual model has been overemphasised in northern Australia, where tree abundance is more strongly controlled by water availability. The dominant trees, eucalypts, have a remarkable capacity to grow through the ‘fire trap’ to reach fire-resistant sizes. This fire tolerance makes eucalypts relatively unresponsive to management-imposed reductions in fire frequency and intensity. Other trees in these savannas are typically more fire sensitive and respond positively to such management. There are suggestions that savanna fire management could lead to increases in woody biomass, but we contend that if tree biomass is strongly limited by water availability, then potential increases in tree biomass are relatively limited, at least in relation to the dominant eucalypt component. There is potential to increase the biomass of the more fire-sensitive non-eucalypts, but the upper bound of non-eucalypt tree biomass in these eucalypt-dominated systems remains poorly understood.

The stock, rates of sequestration and allocation of carbon were estimated for trees in 14 0.1-ha plots at Kapalga in Kakadu National Park, Northern Territory, using new allometric relationships of carbon stock to stem cross-sectional area and measured growth rates of trees. Carbon stocks of trees ranged from 12 to 58 t ha–1, with sequestration representing ~9% of the total stocks. More than half of the sequestered carbon is allocated to leaves and twigs and ~20% to wood. Only ~25% is retained in the live trees with leaf and twig fall accounting for 80%–84% of the total transfers to the environment. An alternative method of calculating sequestration rates from consideration of water use and carbon-isotope discrimination data had a close to 1 : 1 match with estimates from allometric relationships. We developed and applied algorithms to predict the impacts of fire on carbon stocks of live trees. This showed that the reduction in live carbon stocks caused by single fires increased with increasing intensity, but the impact was highly dependent on the tree stand structure.

We measured vegetation changes inside and outside two exclosures built in 1973 on red calcareous loam soils located in Conkerberry Paddock on Victoria River Research Station in northern Australia. These two exclosures were unburnt since their establishment in 1973 until exclosure 1 was unintentionally burnt late in the dry season (October) of 2001. Data from permanent transects and examples from photopoints illustrate that from mostly bare soils in 1973, total pasture biomass recovered relatively rapidly both inside and outside exclosures (in about five years). This initial recovery was primarily due to the establishment of annual grasses and forbs. After this five year period, there was a consistent increase in the biomass of perennial grasses, such as Heteropogon contortus and Dichanthium spp. Also in the first five years after exclosure, the exotic shrub, Calotropis procera,invaded the study area, but then largely disappeared in a period of lower wet-season rainfall in the late 1980s. The density of native tree species, particularly Hakea arborescens, Eucalyptus pruinosa and Lysiphyllum cunninghamii increased in general, but more so inside one or other of the exclosures. Although the late dry-season fire of 2001 reduced the density of larger H. arborescens and L. cunninghamii inside the exclosure at Site 1, this effect was not apparent for smaller trees and for trees outside this exclosure. Our findings show that savanna vegetation can change massively in the medium term (29 years) and that exclosure from cattle grazing can contribute to our understanding of the role of livestock in such change. However, exclosures by themselves do not provide adequate information about the processes leading to vegetation change replicated experimental studies are needed. That substantial increase in the biomass and proportion of perennial grasses occurred with light to moderate cattle grazing implies that these rangelands can be managed for production, although control of woody vegetation is an issue.

There is national and international concern that tree clearing and cattle grazing reduce habitat for native fauna. In this paper we quantify how the degree of clearing and the level of grazing change the patch structure and composition of vegetation in eucalypt woodlands, and how these habitat changes affect counts for 10 species including birds, reptiles and small mammals. These species were selected because they were abundant, hence providing the data needed for ordinations and regressions. We studied 37 sites occurring in two regions of central Queensland: Blackwater/Emerald and Alpha/Jericho. On each site, indices for the degree of tree clearing and the level of livestock grazing were assessed, the cover and size of tree groves, shrub thickets, log hummocks, termite mounds and perennial grass clumps were measured, and abundances of 10 common vertebrate species were estimated. As expected, the cover and size of tree groves declined and the cover of grass clumps increased as the degree of clearing increased. Native grass composition changed to introduced Buffel Grass as the level of grazing increased. Clearing affected fauna counts more than grazing did. The Grey Butcherbird Cracticus torquatus, Yellow-throated Miner Manorina flavigula, Striated Pardalote Pardalotus striatus and Pale-headed Rosella Platycercus eximius had significantly lower counts on cleared sites, but the Red-backed Fairy-wren Malurus melanocephalus had higher counts. The introduced House Mouse Mus musculus also had higher counts on cleared sites, but Carnaby's Skink Cryptoblephrus carnabyi had lower counts. Counts of the Weebill Smicrornis brevirostris, Bynoe's Gecko Heteronotia binoei and the Delicate Mouse Pseudomys delicatulus did not significantly change with clearing. Counts for the Pale-headed Rosella increased as the level of grazing increased, but counts for the other fauna species did not significantly change with our grazing index. Except for Bynoe's Gecko, fauna counts significantly changed with various vegetation patch attributes, particularly those strongly affected by clearing. From these results, we know which of the 10 fauna species are likely to decline or increase with clearing and grazing, and this knowledge can be used by land managers.

The propagation of compressional Pi 2 waves in the inner magnetosphere is investigated by analyzing the onset delay times between the ground and the geosynchronous altitude. We use the compressional component (northward) of magnetic data from low-latitude stations and the geosynchronous satellite ETS-VIII (GMLat. = -10.8°, GMLon. = 217.5°). The onset delays are determined by a cross-correlation analysis, and we analyzed the events with high waveform correlations (correlation coefficient greater than 0.75). Some of these high-correlation events have the properties of propagating waves; Pi 2 waveforms at the ground stations and the satellite were synchronized with each other when the data were shifted by onset delays. The results of the statistical analysis show that 87% of the Pi 2 onsets at a ground station (Kuju, GMLat. = 26.13°, GMLon. = 202.96°) were delayed from the Pi 2 onsets at ETS-VIII, and the average of the delay times was 29 sec. This clearly shows Pi 2 onsets (initial perturbations of Pi 2) propagated from the geosynchronous altitude to the low-latitude ground. The delay times tended to be larger around the midnight sector than around the dawn and dusk sectors. These results are consistent with two-dimensional propagation of fast waves estimated by the model of Uozumi et al. (J Geophys Res 114:A11207, 2009). The delay times are nearly identical to the travel time of fast waves from geosynchronous altitude to the low-latitude ground, and the local time variation of the delay shows the azimuthal propagation along the geosynchronous orbit. We conclude that the initial compressional perturbations of Pi 2 waves propagate radially and longitudinally as a fast wave in the inner magnetosphere.