Fire research for conservation management in tropical savannas: Introducing the Kapalga fire experiment

Abstract Fire is a dominant feature of tropical savannas throughout the world, and provides a unique opportunity for habitat management at the landscape scale. We provide the background and methodology for a landscape-scale savanna fire experiment at Kapalga, located in Kakadu National Park in the seasonal tropics of northern Australia. The experiment addresses the limitations of previous savanna fire experiments, including inappropriately small sizes of experimental units, lack of replication, consideration of a narrow range of ecological responses and an absence of detailed measurement of fire behaviour. In contrast to those elsewhere in the world, Australia's savannas are sparsely populated and largely uncleared, with fires lit primarily in a conservation, rather than pastoral, context. Fire management has played an integral role in the traditional lifestyles of Aboriginal people, who have occupied the land for perhaps 50 000 years or more. Currently the dominant fire management paradigm is one of extensive prescribed burning early in the dry season (May-June), in order to limit the extent and severity of fires occurring later in the year. The ecological effects of different fire regimes are hotly debated, but we identify geo-chemical cycling, tree demography, faunal diversity and composition, phenology, and the relative importance of fire intensity, timing and frequency, as critical issues. Experimental units (‘compartments’) at Kapalga are 15–20km² catchments, centred on seasonal creeks that drain into major rivers. Each compartment has been burnt according to one of four treatments, each replicated at least three times: ‘Early’- fires lit early in the dry season, which is the predominant management regime in the region; ‘Late’- fires lit late in the dry season, as occurs extensively in the region as unmanaged ‘wildfires’; ‘Progressive’- fires lit progressively throughout the dry season, such that different parts of the landscape are burnt as they progressively dry out (believed to approximate traditional Aboriginal burning practices); and ‘Unburnt’- no fires lit, and wildfires excluded. All burning treatments have been applied annually for 5 years, from 1990 to 1994. Six core projects have been conducted within the experimental framework, focusing on nutrients and atmospheric chemistry, temporary streams, vegetation, insects, small mammals, and vertebrate predators. Detailed measurements of fire intensity have been taken to help interpret ecological responses. The Kapalga fire experiment is multidisciplinary, treatments have been applied at a landscape scale with replication, and ecological responses can be related directly to measurements of fire intensity. We are confident that this experiment will yield important insights into the fire ecology of tropical savannas, and will make a valuable contribution to their conservation management.     

Optimal Fire Regimes for Soil Carbon Storage in Tropical Savannas of Northern Australia

Modification of fire regimes in tropical savannas can have significant impacts on the global carbon (C) cycle, and therefore, on the climate system. In Australian tropical savannas, there has been recent, large-scale implementation of fire management that aims to decrease Kyoto-compliant non-CO₂ greenhouse gas emissions by reducing late dry season intense fires through strategic early dry season burning. However, there is no accounting for changes to soil C stocks resulting from changes to savanna fire management, although impacts on these pools may be considerable. We present a hypothesis that soil C storage is greatest under low intensity fires with an intermediate fire return interval. Simulations using the CENTURY Soil Organic Matter Model confirmed this hypothesis with greatest soil C storage under a fire regime of one low intensity fire every 5 years. Key areas of uncertainty for CENTURY model simulations include fine root dynamics, charcoal production and nitrogen (N) cycling, and better understanding of these processes could improve model predictions. Soil C stocks measured in the field after 5 years of annual, 3 year and unburned fire treatments were not significantly different (range 41–58 t ha⁻¹), but further CENTURY modelling suggests that changes in fire management will take up to 100 years to have a detectable impact (+4 t ha⁻¹) on soil C stocks. However, implementation of fire management that reduces fire frequency and intensity within the large area of intact savanna landscapes in northern Australia could result in emissions savings of 0.17 t CO₂-e ha⁻¹ y⁻¹, four times greater than reductions of non-CO₂ emissions.     

Landscape analysis of Aboriginal fire management in Central Arnhem Land, north Australia

Aim To describe the spatial and temporal pattern of landscape burning with increasing distance from Aboriginal settlements. Location Central Arnhem Land, a stronghold of traditional Aboriginal culture, in the Australian monsoon tropics. Methods Geographical information system and global positioning system technologies were used to measure spatial and temporal changes in fire patterns over a one decade period in a 100 × 80 km area that included a cluster of Aboriginal settlements and a large uninhabited area. The major vegetation types were mapped and fire activity was assessed by systematic visual interpretation of sequences of cloud‐free Landsat satellite images acquired in the first (May to July) and second (August to October) halves of the 7‐month dry season. Fire activity in the middle and end of one dry season near an Aboriginal settlement was mapped along a 90‐km field traverse. Canopy scorch height was determined by sampling burnt areas beside vehicle tracks. Results Satellite fire mapping was 90% accurate if the satellite pass followed shortly after a fire event, but the reliability decayed dramatically with increasing time since the fire. Thus the satellite mapping provided a conservative index of fire activity that was unable to provide reliable estimates of the spatial extent of individual fires. There was little landscape fire activity in the first half of the dry season, that was mostly restricted to areas immediately surrounding Aboriginal settlements, with burning of both inhabited and uninhabited landscapes concentrated in the second half of the dry season. The mean decadal fire indices for the three dominant vegetation types in the study area were three in the plateau savanna, two in the sandstone and five in the wet savanna. The spatial and temporal variability of Aboriginal burning apparent in the satellite analyses were verified by field traverse surrounding a single settlement. Fires set by Aborigines had low scorch height of tree crowns reflecting low intensity, despite generally occurring late in the dry season. Conclusions Our findings support the idea that Aboriginal burning created a fine‐scale mosaic of burnt and unburnt areas but do not support the widely held view that Aboriginal burning was focused primarily in the first half of the dry season (before July). The frequency and scale of burning by Aborigines appears to be lower compared with European fire regimes characterized by fires of annual or biennial frequencies that burn large areas. The European fire regime appears to have triggered a positive feedback cycle between fire frequency and flammable grass fuels. The widely advocated management objective of burning in the first half of the dry season burning provides one of the few options to control fires once heavy grass fuel loads have become established, however we suggest it is erroneous to characterize such a regime as reflecting traditional Aboriginal burning practices. The preservation of Aboriginal fire management regimes should be a high management priority given the difficulty in breaking the grass‐fire cycle once it has been initiated.     

Testing the grass-fire cycle: alien grass invasion in the tropical savannas of northern Australia

Abstract. Invasive alien grasses can increase fuel loads, leading to changes in fire regimes of invaded ecosystems by increasing the frequency, intensity and spatial extent of fires. Andropogon gayanus Kunth. (Gamba grass), a tall perennial grass from Africa, is invading ecosystems in the Top End of northern Australia. To determine whether A. gayanus alters savanna fire regimes, we compared fuel loads and fire intensities at invaded sites with those from native grass savannas. Savanna invaded by A. gayanus had fuel loads up to seven times higher than those dominated by native grasses. This higher fuel load supported a fire that was on average eight times more intense than those recorded in native grass savannas at the same time of year (means 15700 ± 6200 and 2100 ± 290 kW m⁻¹, respectively), and was the highest early dry season fire intensities ever recorded in the Northern Territory. These results suggest that A. gayanus is a serious threat to northern Australia's savannas, with the potential to alter vegetation structure and initiate a grass-fire cycle.     

Soil temperature and depth of legume germination during early and late dry season fires in a tropical eucalypt savanna of north‐eastern Australia

Abstract Temperatures that significantly increase seed germination of some tropical legumes (i.e. 80–100°C) were documented in the topsoil during the passage of early (May) and late (October) dry season fires in a tropical eucalypt savanna of north‐eastern Australia. Elevated temperatures penetrated at least 30 mm into the soil during the higher‐intensity, late dry season fires, but were only detected at 10 mm during the early dry season fires. The depth from which germination of two native legume forbs Galactia tenuiflora and Indigofera hirsuta occurred was positively related to the temperature elevation in the topsoil and was greater after late compared with early dry season fires. A broader range in germination depth, resulting in higher seedling densities, was recorded for I hirsuta after late dry season fires. These results suggest that seedling emergence of native leguminous forbs is likely to occur at a greater density after late rather than early dry season fires in tropical eucalypt savannas of north‐eastern Australia. Therefore, the season of burning, as a result of its relationship to fire intensity, can influence species composition through its effect on seed germination.     

Long‐term variability and rainfall control of savanna fire regimes in equatorial East Africa

Fires burning the vast grasslands and savannas of Africa significantly influence the global carbon cycle. Projecting the impacts of future climate change on fire‐mediated biogeochemical processes in these dry tropical ecosystems requires understanding of how various climate factors influence regional fire regimes. To examine climate–vegetation–fire linkages in dry savanna, we conducted macroscopic and microscopic charcoal analysis on the sediments of the past 25 000 years from Lake Challa, a deep crater lake in equatorial East Africa. The charcoal‐inferred shifts in local and regional fire regimes were compared with previously published reconstructions of temperature, rainfall, seasonal drought severity, and vegetation dynamics to evaluate millennial‐scale drivers of fire occurrence. Our charcoal data indicate that fire in the dry lowland savanna of southeastern Kenya was not fuel‐limited during the Last Glacial Maximum (LGM) and Late Glacial, in contrast to many other regions throughout the world. Fire activity remained high at Lake Challa probably because the relatively high mean‐annual temperature (~22 °C) allowed productive C₄ grasses with high water‐use efficiency to dominate the landscape. From the LGM through the middle Holocene, the relative importance of savanna burning in the region varied primarily in response to changes in rainfall and dry‐season length, which were controlled by orbital insolation forcing of tropical monsoon dynamics. The fuel limitation that characterizes the region's fire regime today appears to have begun around 5000–6000 years ago, when warmer interglacial conditions coincided with prolonged seasonal drought. Thus, insolation‐driven variation in the amount and seasonality of rainfall during the past 25 000 years altered the immediate controls on fire occurrence in the grass‐dominated savannas of eastern equatorial Africa. These results show that climatic impacts on dry‐savanna burning are heterogeneous through time, with important implications for efforts to anticipate future shifts in fire‐mediated ecosystem processes.     

Climate regime shift and forest loss amplify fire in Amazonian forests

Frequent Amazonian fires over the last decade have raised the alarm about the fate of the Earth's most biodiverse forest. The increased fire frequency has been attributed to altered hydrological cycles. However, observations over the past few decades have demonstrated hydrological changes that may have opposing impacts on fire, including higher basin‐wide precipitation and increased drought frequency and severity. Here, we use multiple satellite observations and climate reanalysis datasets to demonstrate compelling evidence of increased fire susceptibility in response to climate regime shifts across Amazonia. We show that accumulated forest loss since 2000 warmed and dried the lower atmosphere, which reduced moisture recycling and resulted in increased drought extent and severity, and subsequent fire. Extremely dry and wet events accompanied with hot days have been more frequent in Amazonia due to climate shift and forest loss. Simultaneously, intensified water vapor transport from the tropical Pacific and Atlantic increased high‐altitude atmospheric humidity and heavy rainfall events, but those events did not alleviate severe and long‐lasting droughts. Amazonia fire risk is most significant in the southeastern region where tropical savannas undergo long seasonally dry periods. We also find that fires have been expanding through the wet–dry transition season and northward to savanna–forest transition and tropical seasonal forest regions in response to increased forest loss at the “Arc of Deforestation.” Tropical forests, which have adapted to historically moist conditions, are less resilient and easily tip into an alternative state. Our results imply forest conservation and fire protection options to reduce the stress from positive feedback between forest loss, climate change, and fire.     

Fire regime shifts affect bird species distributions

Aim Fire is a major driver of ecosystem structure and process, and shifts in fire regimes are implicated in the decline of many species. Shifting fire regimes have been documented around the world, and fire frequency and extent is predicted to increase in many areas because of changes in both climate and land management. Here, we evaluate how predicted increases in fire frequency are likely to impact on species distributions. Location The tropical savannas of northern Australia. Methods We developed distribution models for 44 bird species using the modelling algorithm Maxent. Our models incorporated bird locality records and environmental variables including climate, total fire frequency and the subset of fire frequency occurring late in the dry season. We investigated the effect of increasing total fire frequency and increasing fire late in the dry season, on species distributions by projecting species model algorithms onto scenarios of incrementally increased total fire frequency. Results The probability of presence for most species was higher when fire frequency late in the dry season was low. Species showed a mixed response to an overall increase in total fire frequency, with one‐third predicted to increase in distribution. However, almost all species (98%) showed a decrease in predicted range with increased late‐dry season fire, and species distribution area was generally negatively correlated with an increase in late‐dry season fire. Main conclusions Our study highlighted the array of responses of species to increasing fire frequency and suggested that increased fire frequency late in the dry season is detrimental to most savanna‐restricted bird species. The understanding of individual species’ preferences for particular fire frequencies is important for informed conservation planning.     

When is a ‘forest’ a savanna,and why does it matter?

Savannas are defined based on vegetation structure, the central concept being a discontinuous tree cover in a continuous grass understorey. However, at the high‐rainfall end of the tropical savanna biome, where heavily wooded mesic savannas begin to structurally resemble forests, or where tropical forests are degraded such that they open out to structurally resemble savannas, vegetation structure alone may be inadequate to distinguish mesic savanna from forest. Additional knowledge of the functional differences between these ecosystems which contrast sharply in their evolutionary and ecological history is required. Specifically, we suggest that tropical mesic savannas are predominantly mixed tree–C₄ grass systems defined by fire tolerance and shade intolerance of their species, while forests, from which C₄ grasses are largely absent, have species that are mostly fire intolerant and shade tolerant. Using this framework, we identify a suite of morphological, physiological and life‐history traits that are likely to differ between tropical mesic savanna and forest species. We suggest that these traits can be used to distinguish between these ecosystems and thereby aid their appropriate management and conservation. We also suggest that many areas in South Asia classified as tropical dry forests, but characterized by fire‐resistant tree species in a C₄ grass‐dominated understorey, would be better classified as mesic savannas requiring fire and light to maintain the unique mix of species that characterize them.     

Savanna–forest hysteresis in the tropics

A simple dynamic model relating forest area in a region, its contribution to dry season precipitation and the effect on its own establishment was developed. The model equation shows hysteresis between forest and savannas as a function of imported dry season precipitation. Regions are either dominated by forests or savannas, with each ecosystem showing stability despite changes in imported dry season precipitation. Deforestation beyond a certain threshold value, however, could cause a collapse of forest ecosystems and replacement by savannas in marginal areas. The predictions of this model corroborate pollen core analysis in the Amazon basin, where historical stability of tropical forest cover has been shown despite global climate change.     

Short‐term effects of low intensity fire on soil macroinvertebrate assemblages in different vegetation patch types in an Australian tropical savanna

Abstract Early dry season fires are a common land management regime employed across the tropical savannas of northern Australia. The rationale is that this reduces fuel loads and so reduces fire risk in the latter part of the dry season. Despite the acceptance of fire as a major management tool the ecological effects of fire remain uncertain. Vegetation patches and their associated macroinvertebrates play a critical role in the capture and recycling of water and nutrients. The aim of this study was to examine the responses of soil macroinvertebrates, within different types of vegetation patches, to early dry season fires in tropical savanna woodland in northern Australia. The abundance of major macroinvertebrate taxa and functional groups, and taxon richness were quantified in three vegetation patch types 2 weeks before and 2 weeks after burning. Termites dominated the soil macroinvertebrate assemblage sampled. Fire led to significant decreases in ant and spider abundances and overall taxon richness. Functional group analyses showed significant overall declines in the abundances of macropredators and litter transformers. There were also interactions between fire and patch type; in tree patches, fire significantly reduced total macroinvertebrate abundance, as well as the abundance of termites and ecosystem engineers. These changes in soil macroinvertebrates will potentially influence patch functionality, with important implications for soil processes and landscape health.     

Pyric Herbivory and the Nexus Between Forage,Fire and Native and Introduced Large Grazing Herbivores in Australian Tropical Savannas

Earth’s tropical savannas typically support high biomass of diverse grazing herbivores that depend on a highly fluctuating resource: high-quality forage. An annual wet–dry cycle, fire and herbivory combine to influence forage quality and availability throughout the year. In the savannas of northern Australia, a depauperate suite of large native (marsupial) herbivores (wallaroos [Osphranter spp.] and the agile wallaby [Notamacropus agilis]) compete for resources with non-native large herbivores introduced in the late nineteenth century, particularly bovines (feral and managed cattle [Bos spp.] and feral water buffalo [Bubalus bubalis]) that now dominate the landscape. Anecdotal reports of recent population declines of large macropods and negative impacts of bovines highlight the need to better understand the complex relationship between forage, fire and abundance of native and introduced large herbivores. The pyric herbivory conceptual model, which posits complex feedbacks between fire and herbivory and was developed outside Australia, predicts that native and introduced large herbivores will both respond positively to post-fire forage production in Australian savannas where they co-occur. We used grazing exclosures, forage biomass and nutrient analyses and motion-sensor camera-trapping to evaluate the overall robustness of the pyric herbivory model in the Australian context, specifically whether forage quantity and quality are impacted by herbivory, season and fire activity, and which forage attributes most influence large grazing herbivore abundance. Forage quantity, as measured by live, dead and total herbaceous biomass and proportion of biomass alive, was higher inside herbivore exclosures, even at relatively low densities of herbivores. Forage quality, as measured by fibre content, was not affected by herbivory, however, crude protein content of live herbaceous biomass was greater outside herbivore exclosures. Recent fire was an important predictor of all measures of forage quantity and quality. Recent fire occurrence decreased overall quantity (biomass) but increased quality (decreased fibre content and increased crude protein content); late dry season fires resulted in forage with the highest crude protein content. The predictions of the pyric herbivory conceptual model are consistent with observations of the feeding behaviour of introduced bovines and some large macropods in northern Australian savannas, lending support to the global generality of pyric herbivory in fire-prone grassy biomes.

     

The role of seed banks in vegetation dynamics and restoration of dry tropical ecosystems

Abstract. This paper reviews studies on seed banks in tropical dry vegetation and, for comparison, some wet tropical and dry subtropical vegetation. A first general conclusion is that tropical seed banks are smaller than those in temperate ecosystems. Many studies are devoted to only one species, among which are several Acacia species, both tropical and subtropical, some of which reach densities in the seed bank of up to 10000 seeds/m². The lesser importance of the seed bank strategy in tropical species may be related to higher risks of seed loss through higher mortality (fire, predation, pathogens etc.), or to intermittent germination occasions in relation to ephemeral favourable conditions (e.g. rains during the dry season). Regarding vegetation dynamics, the importance of seed banks and regeneration from seeds is most pronounced in larger forest gaps. Examples are discussed of rapid succession to forest of savannas after protection from fire, which is partly due to regeneration from the seed bank. The significance of seed banks in vegetation restoration projects is outlined and the need for additional sowing of seeds of important species underlined.     

Phenology, Lignotubers, and Water Relations of Cochlospermum vitifolium, a Pioneer Tropical Dry Forest Tree in Costa Rica

We examined structural and physiological traits relevant to the phenology of the tropical dry forest (TDF) pioneer tree Cochlospermum vitifolium . Despite marked seasonality in rainfall, meristem activity occurred throughout the year. Leaves were produced almost continuously during the rainy season, while leaf shedding started early during drought, before changes in soil water content were observed. Phenological activity under drought included flowering and fruiting of leafless trees; bud break and shoot extension took place before the end of the dry season. Low wood density of C. vitifolium stems (0.17 g/cm³) and lignotubers (0.14 g/cm³) provided water and starch storage needed to support phenological events such as branch extension, leaf flushing, and reproduction during the dry season, and probably also contributed to survival following mechanical damage and fire, typical of early TDF successional stages. Lignotuber water and starch contents showed substantial seasonal variation, declining from the beginning of the dry season to their lowest levels at the time of reproduction and dry-season flushing. Stems progressively replaced lignotubers as main storage organs as tree size increased. Evidence for a role of water stores in buffering daily water deficits was weak. Leaf water potentials remained above −1.2 MPa and stomatal conductance below 350 mmol/m²/s, suggesting that gas exchange during the rainy season was limited to prevent xylem cavitation. Leaf shedding occurred when early-morning and mid-day Ψ_L converged at the rainy–dry season transition, without changes in lignotuber or soil water content, suggesting that leaves of C. vitifolium are closely tuned to atmospheric drought.
Abstract in Spanish is available at http://www.blackwell-synergy.com/loi/btp .     

Termite mound emissions of CH4 and CO2 are primarily determined by seasonal changes in termite biomass and behaviour

Termites are a highly uncertain component in the global source budgets of CH₄ and CO₂. Large seasonal variations in termite mound fluxes of CH₄ and CO₂ have been reported in tropical savannas but the reason for this is largely unknown. This paper investigated the processes that govern these seasonal variations in CH₄ and CO₂ fluxes from the mounds of Microcerotermes nervosus Hill (Termitidae), a common termite species in Australian tropical savannas. Fluxes of CH₄ and CO₂ of termite mounds were 3.5-fold greater in the wet season as compared to the dry season and were a direct function of termite biomass. Termite biomass in mound samples was tenfold greater in the wet season compared to the dry season. When expressed per unit termite biomass, termite fluxes were only 1.2 (CH₄) and 1.4 (CO₂)-fold greater in the wet season as compared to the dry season and could not explain the large seasonal variations in mound fluxes of CH₄ and CO₂. Seasonal variation in both gas diffusivity through mound walls and CH₄ oxidation by mound material was negligible. These results highlight for the first time that seasonal termite population dynamics are the main driver for the observed seasonal differences in mound fluxes of CH₄ and CO₂. These findings highlight the need to combine measurements of gas fluxes from termite mounds with detailed studies of termite population dynamics to reduce the uncertainty in quantifying seasonal variations in termite mound fluxes of CH₄ and CO₂.     

Contemporary landscape burning patterns in the far North Kimberley region of north-west Australia: human influences and environmental determinants

Aim This study of contemporary landscape burning patterns in the North Kimberley aims to determine the relative influences of environmental factors and compare the management regimes occurring on Aboriginal lands, pastoral leases, national park and crown land. Location The study area is defined at the largest scale by Landsat Scene 108–70 that covers a total land area of 23,134 km² in the North Kimberley Bioregion of north‐west Australia, including the settlement of Kalumburu, coastline between Vansittart Bay in the west and the mouth of the Berkeley River in the east, and stretching approximately 200 km inland. Methods Two approaches are applied. First, a 10‐year fire history (1990–1999) derived from previous study of satellite (Landsat‐MSS) remote sensing imagery is analysed for broad regional patterns. And secondly, a 2‐year ground‐based survey of burning along major access roads leading to an Aboriginal community is used to show fine‐scale burning patterns. anova and multiple regression analyses are used to determine the influence of year, season, geology, tenure, distance from road and distance from settlement on fire patterns. Results Satellite data indicated that an average of 30.8% (±4.4% SEM) of the study area was burnt each year with considerable variability between years. Approximately 56% of the study area was burnt on three or more occasions over the 10‐year period. A slightly higher proportion of burning occurred on average in the late dry season (17.2 ± 3.6%), compared with the early dry season (13.6 ± 3.3%). The highest fire frequency occurred on basalt substrates, on pastoral tenures, and at distances 5–25 km from roads. Three‐way anova demonstrated that geological substrate and land use were the most significant factors influencing fire history, however a range of smaller interactions were also significant. Analysis of road transects, originating from an Aboriginal settlement, showed that the timing of fire and geology type were the most significant factors affecting the pattern of area burnt. Of the total transect area, 28.3 ± 2.9% was burnt annually with peaks in burning occurring into the dry season months of June, August and September. Basalt uplands (81.2%) and lowlands (30.1%) had greater areas burnt than sandstone (12.3%) and sands (17.7%). Main conclusions Anthropogenic firing is constrained by two major environmental determinants; climate and substrate. Seasonal peaks in burning activity in both the early and late dry season relate to periods of optimal fire‐weather conditions. Substrate factors (geology, soils and physiognomy) influence vegetation‐fuel characteristics and the movement of fire in the landscape. Basalt hills overwhelmingly supported the most frequent wildfire regime in the study region because of their undulating topography and relatively fertile soils that support perennial grasslands. Within these spatial and temporal constraints people significantly influenced the frequency and extent of fire in the North Kimberley thus tenure type and associated land uses had a significant influence on fire patterning. Burning activity is high on pastoral lands and along roads and tracks on some tenure types. While the state government uses aerial control burning and legislation to try to restrict burning to the early dry season across all geology types, in practice burning is being conducted across the full duration of the dry season with early dry season burning focused on sandstone and sand substrates and late dry season burning focused on basalt substrates. There is greater seasonal and spatial variation in burning patterns on landscapes managed by Aboriginal people.     

Contrasting demographics of tropical savanna and temperate forest eucalypts provide insight into how savannas and forests function. A case study using Corymbia clarksoniana from north‐eastern Australia

Eucalypts (Eucalyptus spp. and Corymbia spp.) dominate many communities across Australia, including frequently burnt tropical savannas and temperate forests, which receive less frequent but more intense fires. Understanding the demographic characteristics that allow related trees to persist in tropical savannas and temperate forest ecosystems can provide insight into how savannas and forests function, including grass–tree coexistence. This study reviews differences in critical stages in the life cycle of savanna and temperate forest eucalypts, especially in relation to fire. It adds to the limited data on tropical eucalypts, by evaluating the effect of fire regimes on the population biology of Corymbia clarksoniana, a tree that dominates some tropical savannas of north‐eastern Australia. Corymbia clarksoniana displays similar demographic characteristics to other tropical savanna species, except that seedling emergence is enhanced when seed falls onto recently burnt ground during a high rainfall period. In contrast to many temperate forest eucalypts, tropical savanna eucalypts lack canopy‐stored seed banks; time annual seed fall to coincide with the onset of predictable wet season rain; have very rare seedling emergence events, including a lack of mass germination after each fire; possess an abundant sapling bank; and every tropical eucalypt species has the ability to maintain canopy structure by epicormically resprouting after all but the most intense fires. The combination of poor seedling recruitment strategies, coupled with characteristics allowing long‐term persistence of established plants, indicate tropical savanna eucalypts function through the persistence niche rather than the regeneration niche. The high rainfall‐promoted seedling emergence of C. clarksoniana and the reduction of seedling survival and sapling growth by fire, support the predictions that grass–tree coexistence in savannas is governed by rainfall limiting tree seedling recruitment and regular fires limiting the growth of juvenile trees to the canopy.     

Simulation of forest carbon dynamics based on a dry-matter production model

Effects of dry season upon a tropical rainforest ecosystem are analysed by employing the microcomputer model developed in a previous paper (Oikawa, 1985). Surplus production (P _s) illustrated in a three-dimensional figure linearly decreases with the length of dry season (λ). Simulation experiments demonstrate that all the three strata constituting the tropical rainforest ecosystem model are able to exist stably and perpetually unless λ exceeds four months. However, it is suggested that the dry season lasting five months brings about a transition from a tropical rainforest to a subtropical deciduous forest, because the upper story fails in a stable and perpetual occurrence at λ =5 months. Though the total living phytomass gradually decreases with prolonged dry season within four months, supplementary increase of living phytomass is observed in the middle story. A considerable amount of soil organic dead accumulated is also observed with increasing dry season, so that fairly constant ecosystem biomass including soil organic dead is predicted independent of the length of dry season so long as it is less than five months.     

Role of weather and fuel in stopping fire spread in tropical savannas

Analysis of wildfire extinguishment can help to identify the relative contribution of weather and management to the prevention of fire spread. Here we examine the role of weather, previous fire scars and other fuel interruptions at stopping the spread of nine large (mean 90 000 ha) late dry season fires in Arnhem Land, in the tropical savannas of northern Australia. Daily spread was mapped using Moderate‐resolution Imaging Spectroradiometer (MODIS) satellite imagery with a resolution of 250 m. We sampled points along the boundary of the fires and 1 km inside the boundary and compared conditions between the two sets. Using a combination of binomial regression and regression tree analysis, we found that recent burn scars (from the same year) were very effective at stopping fires. Where there was any recent burning within 500 m of a point, there was a 92% likelihood that it was a boundary. Interruptions such as roads, rivers and topography had small but significant effects. Vegetation type and vegetation greenness also had minor effects. Weather had a small effect via wind speed. This minor role of weather was reinforced by the fact that on most days the fires were both spreading and stopping at different parts of their perimeter. In these savannas, the weather in the late dry season is relatively invariant and is probably always conducive to some degree of fire spread. Here, interruptions to the fuel are critical to stopping fires. Nevertheless, for approximately half of boundary cases, the cause of stopping was not clear. This is probably due to the coarse scale of the analysis that does not reflect fine patterns of fuel arrangements.