Proceedings of the conference held 8-9 October 1994, Footscray, Melbourne
Biodiversity Series, Paper No. 8
Department of the Environment, Sport and Territories, 1996
23. Fuel dynamics, preplan and future research needs
NSW National Parks and Wildlife Service
The conservation and management problems of widespread prescribed burning are briefly outlined and the poor knowledge of fuel complexes, related to vegetation types, is emphasised; although generally fuel loads exceeding 8-10 tonnes/hectare are considered to pose a hazard. The limited data on fuel dynamics relate to specific communities whereas fuel management programs are regional; but most local studies indicate that fuels accumulate rapidly for seven years after a fire and then decline.
The flawed basis of most hazard and risk assessment to date is explained and the development of new computer-based models for the ACT and Victoria is briefly described. Results of recent detailed fuel studies, over a number of woodland sites in southern NSW, demonstrate the complexity of the fuel components (e.g. fine ground litter, mid- and upper strata) and the problems of interrelating rates of accumulation with intensity, residual loads or decomposition. Matrices relating soil type (erodibility) with gradient are being developed for steep catchments, but no fuel or fire behaviour prediction model has been developed for heathland.
Analyses of occurrence and distribution of recorded ignitions in southern NSW show that: the majority of unplanned fires originate from non-natural causes; arson will continue to be a major source of illegal ignitions, but is very difficult to assess or predict. It is suggested that this problem can only be addressed by further social research on community attitudes and levels of understanding of fire management programs – especially in urban bushland interface areas.
Several recommendations are made to establish effective linkages between research and fire management personnel. These include:
- increased input and co-ordination of fire research programs by a nationally recognised organisation;
- establishment of a national bushfire research unit;
- additional training opportunities for fire planning and management (e.g. through formal, nationally recognised tertiary courses);
- greater roles for distant committees in identification and support of research;
- further use of geographic information systems and predictive modelling fire management.
Key words: fuel dynamics, fire management, wildfire mitigation, fire-vegetation interdependence.
All land management agencies have an obligation to undertake soundly based fire management planning and the implementation of effective wildfire mitigation programs. Almost all fire management plans have been developed with the management of ground fuels as the core program, this predominately being achieved through the application of prescribed burning. The principles of prescribed burning are well-founded but unfortunately the application of such burning has and is often poorly planned and executed. This has led to many prescribed burns not meeting the desired results in terms of prescribed levels of fuel reduction, percentage of burned area, excessive scorch heights and unacceptable impacts on native flora and fauna.
As a result of the above problems increasingly widespread application of prescribed burning, particularly in nature conservation reserves, has been criticised and is of considerable concern to many professional and community interest groups. These concerns arise mainly from the lack of detailed data on the impacts of prescribed burning on native biota; even though many hundreds of papers have been written and presented at fire conferences over the past three decades on all aspects of fire management, prescribed burning, wildfire and the interdependence of fire and native vegetation.
Of greater concern is the implementation of prescribed burning for fuel reduction in areas and vegetation types where little or no assessment of fuel loads, accumulation rates and fuel structures has been undertaken. Similarly little or no quantified assessment of the hazardous nature of various fuel complexes is evident in many prescribed burning plans; the general assumption being that fuel loads exceeding 8 to 10 t.ha-1. pose a high fire hazard.
So, arguably the most significant natural area land management program continues to be implemented with little knowledge of the primary component (the fuel complex) which much fire management aims to manipulate, modify or reduce. Unfortunately, while this situation exists the polarised views on prescribed burning will continue to the detriment of nature conservation and wildfire mitigation programs alike.
Many studies on fuel dynamics, particularly that of fine litter fuel accumulation rates have been undertaken over the past 25 years with the results published in some 120 scientific papers. While much has been done the number of published papers on fuels and fuel dynamics is still relatively low, being only about five per cent of the published papers on fires in the Australian environment.
These few papers do present a good overview of the fine fuel component of the fuel complex, but with few exceptions the research has addressed fuel dynamics within a single vegetation type or vegetation association. This has provided a good basis for fuel management planning for the specified communities; however as fuel management programs are generally planned on a region-wide basis much extrapolation of data and many assumptions on fuel accumulation rates are made. Van Loon (1977), Walker (1979), Birk (1979), Fox et al. (1979), Raison et al. (1983), Tolhurst et al. (1992) have made considerable contributions to the understanding of fuel dynamics; all indicate that fuels accumulate rapidly for up to seven years after a fire and then decline towards a zero net accumulation rate where litter fall approximates litter accumulation (Fig. 23.1).
Kessell et al. (1982), Conroy (1987), O'Connell (1991), and others have modelled fuel dynamics providing techniques to extrapolate specific fuels data to regional fuel management programs. The capacity to do this is essential at present as there is renewed pressure to carry out more and more prescribed burning, as a response to the very extensive and damaging wildfires which occurred in urban bushland areas of Sydney in January 1994. The need for more extensive prescribed burning must be carefully considered; the issue is not one of 'doing more'; but one of firstly identifying where and when prescribed burning is appropriate and would mitigate potential impacts of high intensity fires on life and property. This hazard and risk analysis is dependent on detailed fuels data, which as noted above is limited.
To date most hazard and risk assessment has been carried out using a simple point score system to give a numerical 'hazard and risk value' to an area or site. This system again usually used general fuels data extrapolated from specific site data often inappropriate to the area being assessed. Such systems provide a guide to the hazards and risks but are one-off static values, not necessarily reflecting the real situation where hazard and risk values change as a response to changing fuel loads and other dynamic factors.
McRae (1993) has developed a computer based hazard and risk assessment model, covering the entire Australian Capital Territory (ACT). This system provides the opportunity to regularly update the fire hazard and risk status across the ACT, particularly at the urban bushland and grassland interface where most concern is focussed on the protection of life and property.
A similar approach using a geographical information system, has been taken in Victoria (Garvey 1995); both systems being of considerable benefit in planning land, fire and fuel management programs. Both systems still have a dependence on detailed fuels data acquired on a regular and continuous basis - which are unavailable for the majority of areas in all States. The required data include fuel loads, curing rates and flammabilities as influenced by the fuel structures and moisture content. Much scope therefore exists for the linking of fuels research, hazard and risk assessment and fire prediction through appropriate modelling and geographical information systems.
Table 1. Fuel Loads for forest and woodland vegetation associations in southern New South Wales (time since last fire greater than 25 years)
|Vegetation Association||Fuel Weight (t ha-1)||Altitudinal Range
|Eucalyptus delegatensis||28.5||43.5||69.8||1200 - 1400|
|E. delegatensis/E. viminalis||24.2||37.0||42.4||1100 - 1220|
|E. pauciflora||18.4||28.4||38.2||1000 - 1400|
|E. pauciflora/E. dalrympleana||16.5||24.8||35.1||950 - 1500|
|E. pauciflora/E. rubida||14.8||24.0||31.2||900 - 1420|
|E. fastigata||26.5||30.3||38.4||900 - 1050|
|E. viminalis/E. elata||23.4||27.9||34.7||100 - 300|
|E. mannifera/E. dives||13.7||18.6||22.3||200 - 250|
|E. sieberi/E. globoidea||20.6||24.1||28.2||300 - 900|
|E. sieberi||14.9||21.4||22.8||600 - 1100|
|E. consideniana/E. sieberi||17.4||20.1||21.5||300 - 650|
|E. sieberi/E. fraxinoides||16.9||19.8||22.6||900 - 1100|
|E. fraxinoides||15.5||21.8||24.4||1000 - 1200|
|E. fraxinoides/E. pauciflora/E. nitens||22.3||23.5||26.6||1100 - 1300|
|E. globoidea/Angophora floribunda||18.8||22.9||25.8||150 - 450|
|E. dalrympleana/E. radiata||22.0||26.0||30.4||1000 - 1100|
|E. dalrympleana/E. fastigata||23.7||27.4||31.0||1000 - 1100|
|E. radiata/E. bicostata/E. viminalis||18.8||28.5||37.3||550 - 900|
|E. stenostoma/E. sieberi/E. consideniana||16.5||18.1||18.9||400 - 900|
|E. tereticornis||19.0||20.8||21.5||150 - 300|
|E. macrorhyncha||6.4||10.2||16.1||500 - 900|
|E. dives/E. mannifera||8.5||12.6||18.2||450 - 950|
23.3.1 Data for some vegetation associations in southern New South Wales
As noted above much is now known of general trends in fuel accumulation rates in forest and woodland communities, but much remains to be learned of specific components of the fuel complex within individual vegetation associations. For example mean fuel accumulation rates and maximum fuel loads have been generated for a number of community types and depicted as mean fuel curves (Fig. 23.1); but the range and variability of fuel loads for any one vegetation association are seldom detailed. Tolhurst et al. (1992) indicated the seasonal variation and range of fuel loads for a dry sclerophyll forest in Victoria, while Conroy (1987) details the variability of fuel loads as a factor of sample numbers taken for each vegetation type over a geographical range. The following data for some forest and woodland associations in south-east NSW (Table 1) also indicate the wide variability of fuel loads even when a large number of samples are taken over many years. The range of aspects on which the individual vegetation associations occur is reflected in the range of fuel loads. Associations occurring on aspects from exposed (dry) to sheltered (moist) have the widest range and the greatest variability in fuel weights between sites. The ranges in fuel weights also overlap and hence there is a continuum in fuel weights across all sampled vegetation associations from the drier to the moister communities.
It is of interest that the more sites that are sampled for each vegetation association the wider becomes the range of fuel loads (Table 1), although the mean fuel loads may vary very little from the accumulation curve generated from data from a few sites of different ages since the last fire or by generating a fuel curve using the modified Olsen model (Fox et al. 1979; Walker 1979; Birk 1979; Birk & Simpson 1980; Kessell et al. 1982; Raison et al. 1983; O'Connell 1991).
This model expresses the relationship between litter production and decomposition and hence the accumulation of dead litter fuel (x) at time t (in years) as:
Xt = L/k(1 - e-kt) where L = KXmax.
where L is the litter production rate (t ha-1 yr-1), Xmax is the maximum litter standing crop at equilibrium, K is the decomposition constant ( the ratio of L to Xmax) and e is the base of the natural logarithm. This equation yields zero fuel at time 0 and allows the amount of fuel (Xt ) to approach the maximum equilibrium value (Xmax) as time (t) proceeds to infinity. The values of L and Xmax determine both the maximum fuel level reached and the rate at which this level is reached.
Caution must be exercised when using this equation to generate fuel accumulation curves upon which fuel management and fuel reduction/prescribed burning programs are planned. The equation makes three unrealistic assumptions which need to be accounted for:
- it assumes that fine dead litter fuel at time 0 immediately after a fire is zero and that all fires effect total fuel reduction
- it assumes that decomposition and litter fall are constant over time
- it assumes that fine dead fuel load is the only fuel input.
Point 1 above may be correct for high intensity wildfires where there may be a lag period of one to three years during which there is little fuel accumulation (Fig. 23.2). On the other hand, after a low intensity prescribed fire, fuel accumulations are generally rapid and reach maximum levels again within five to seven years of such a fire. More importantly the eight to 10 t ha-1 fuel load, identified in many management plans as the level below which fuels should be maintained, will be exceeded in two to three years after a prescribed burn in most eucalypt forests and woodlands (Kessell et al. 1978; Raison et al. 1993; Tolhurst et al. 1992). This can be modelled if the amount of fuel remaining after any fire is known; as it simply decays as a function of the decomposition constant, viz XRrt = XR e-kt where Xrt is the amount (T ha-1) of this residual fuel remaining at time t and XR is the amount of residual fuel immediately after the last fire. The fine fuel load is thus the sum of Xt and Xrt. Figure 23.3 depicts a family of fuel curves generated for different amounts of fuel reduction in a Eucalyptus delegatensis community all of which converge after 30 to 40 years.
While the total fuel load may be measured and/or modelled as an input to fire behaviour prediction, it is to be recognised that other characteristics of the fuel complex do contribute to fire behaviour. The flammability of the fuels is influenced by the depth and density (packing ratio and bulk density) of the fuel bed and the distribution / ground cover of the fuel. The packing ratio is the actual weight of fuel per unit volume expressed as: PR = bulk density / average particle size. The bulk density can be calculated from the total fuel loading and its average depth expressed as:
BD= Total fuel load(t ha-1) / Depth (m) x 10 (Kg/m³)
A further aspect in the research and use of fuels data is that measured or calculated fuels generally only represent the dead fine litter ground fuels; these may only be a proportion of the total fine fuel available to a fire. For example, in a high to extreme fire season such as may exist in a prolonged drought, much fine fuel in the form of twigs and leaves is held as standing crop on shrubs and small understorey trees (Good 1982; Kessell et al. 1982; Tolhurst et al. 1992). Similarly , fibrous barked trees can support considerable amounts of fine fuel on their trunks, both fuel types providing for the support of a wildfire and the carriage of a fire into upper canopies.
To account for these fuels, measurements of the fuel loads in each stratum need to be made (Figure 23.4). Rothermel (1972) developed a three-strata fuel model to incorporate these data into fire behaviour predictions, with the mid and upper strata fuel data being determined by the canopy separation and diameter method, viz.
% cover = 25 # / (1 + C)2
where C is the average gap or separation distance / average diameter or canopy diameter
The bulk density of the shrub fine fuels would be determined by :
BD = total dry shrub fuel weight (Kg) / 0.25m²
x depth of the foliage canopy
It is therefore obvious that if a detailed appreciation of fuels is to be gained, and effective use made of fuels data in fuel management programs, research on these components of the fuel complex needs to be undertaken. To date fuels research has been dominated by studies on fine ground fuels, but there is an obvious need to develop a multi-strata fuel model for eucalypt forests and woodlands.
Similarly, no fuel or fire behaviour prediction model has been developed to date for heathland communities. The structural variability and spatial distribution of fuels of heath communities poses a real challenge to research and management personnel. (Catchpole 1985, 1987a, 1987b; Catchpole & Catchpole 1993).
Many other research programs by Gill and others have related fuels to fire impacts on flora and fauna, but few have addressed the fuels, fire and catchment/soil stability relationships; catchment management being a major management program in many fire-prone areas (Brown 1972; Good 1973). This aspect of fire research needs to be further addressed as a conflicts often exists between fire mitigation programs (fuel management and prescribed burning) and catchment management programs. Compromises between the objectives for fire mitigation and those of catchment management need to be made but based on detailed vegetation, soils and fuels data.
Fire intervals of five to seven years are commonly advocated for fuel management and hazard reduction programs to maintain fuels below the threshold levels of eight to ten t ha-1; but such frequencies and fuel level objectives can be in conflict with those appropriate to the maintenance of soil stability, particularly in the forested mountain catchments of south-eastern Australia (Good 1981). A compromise between prescribed burning to reduce fuel loads (to mitigate against high intensity wildfires in catchment areas) and maintenance of an adequate cover of ground litter to mitigate against potential soil erosion must be made.
As an example the following matrix of soil type, erosion and slope steepness is a summary of data from studies of soils, erosion and fuel levels required to maintain catchment stability in the Snowy Mountains (Table 2).
|Podzolic red||Podzolic yellow||Solodics red||Solodics Yellow||Solodized solonetz.|
|Slope class||Litter loads (t ha-1)|
The above may be a part of hazard and risk analysis particularly in conservation areas and mountain catchments. But hazard and risk assessment also draws upon a knowledge of the occurrence and distribution of past fire ignitions (Table 3), the probability and distribution of future fire ignitions and the extent of fires. The latter is very difficult to determine or predict as the majority of unplanned fire ignitions across southern NSW for example, now originate from non-natural causes. Illegal ignitions (arson) now commonly cause wildfires and fire records indicate that arsonists are not ad hoc in their activities, nor random in where they carry out their activities. Generally their activities appear to be well planned and deliberate. The fire history data in Tables 3 and 4 indicate that this is a major fire mitigation problem.
For any prescribed burning aimed at mitigating potential fire ignitions to be successful, part of the hazard and risk assessment program must be the predictive modelling of potential ignition sources, particularly that of illegal ignitions. This would require a component of social research, both on the attitudes of the community to fire and the levels of appreciation of any fire management programs proposed for a region, particularly in urban bushland interface areas.
|Cause of ignition||Frequency (%)|
Broad-area prescribed burning may mitigate against potential fire ignition and high intensity wildfires, but the percentage of area burnt by unplanned fires tends to be commensurate with the percentage of arson ignitions (Table 3). While arson continues to be the major source of fire ignitions and arsonists continue to target areas distant from areas of recent wildfire mitigation programs (prescribed burning) the total area burnt by unplanned fires will continue to increase; this is of concern to many managers of natural areas. It should also be a concern that the very program implemented to mitigate against potential wildfires is itself a major contributor to unplanned fire occurrence (>20 per cent). The assessment of fire ignition potential and probability, together with fuel loadings is therefore an essential component of hazard and risk assessment; this needs to be undertaken if fire mitigation programs are to meet the prescribed objectives of fuel reduction, lower fire ignition potential and reduction in the area and distribution of wildfires.
The above suggests that while the principles of prescribed burning have been well researched and developed, the basis upon which many prescribed burning programs are planned and implemented is inadequate; detailed knowledge of the fuel complex and fire ignition probabilities are still lacking.
|Number unplanned ignitions||286|
|Number of prescribed burns||27|
|Average area of prescribed burns||3200 ha|
|No. of unplanned ignitions falling within areas prescribed burned <3 years prior to wildfire ignition||109|
|No. of unplanned ignitions that developed to >50 ha. in prescribed burn areas||16|
|Total area burned by unplanned in prescribed burn areas||4840 ha|
|Average area of unplanned fires in prescribed burned areas||302 ha|
|No. of unplanned ignitions outside areas prescribed burned||267|
|Average area burnt by unplanned fires outside prescribed burn areas||584 ha|
|Total area of wildfires||155 928 ha|
|Percentage of study area burnt by prescribed and wildfire||25%|
This is a major concern at the present time, with the expressed objective of a number of fire and land management agencies to do more burning such that a greatly increased area of recent prescribed burning (less than three years) results. This may or may not be appropriate, but much research on the very fuels to be burnt in these programs, must be undertaken before more extensive burning is embarked upon. It is not a matter of doing more prescribed burning but one of determining where and when it is appropriate. The latter will be very different for an urban bushland interface area as compared with an extensive tract of natural bushland such as a conservation reserve. The Sydney wildfires of January 1994 have given support to calls for more prescribed burning and this may be appropriate for the Sydney urban bushland areas, but this is not a rational basis for the implementation of even more extensive prescribed burning elsewhere, as currently proposed.
Birk and Simpson (1980) made note of the scarcity of long-term data on litter dynamics in Australian forests and this situation still exists to date. A priority in fuels research and fuels management is the establishment of a nation-wide system of fuel monitoring points for the assessment and appreciation of long-term trends in fuel accumulation and decomposition. Until this is done, defined objectives for prescribed burning, in terms of fuel management and fuel reduction, will not be possible and prescribed burning will remain a management activity of concern to many.
As the majority of fuels research programs have focussed on forest and woodland communities, little fuels data are available for heath communities to assist fire and vegetation management in these communities. This research is a high priority as, until a better knowledge of heathland fuels is gained, the development of an applicable and accurate fire behaviour prediction model will not be possible. The recent proposal for a collaborative inter-State program to address heathland fuels and fire behaviour is a recognition of the priority for this research on a national scale (J. Marsden-Smedley 1994 pers comm).
Other priorities for future fuels research have been considered earlier but, as so little fuels data are available, any future fuels research can make a major contribution to our understanding of the very component which fire managers expend so much effort on manipulating and reducing. Any future research on the other hand should be well co-ordinated and be integrated (as considered for the heathland program) to ensure maximum benefit to research and management alike.
This integration needs to be considered at two levels, one being the integration of research fields and the other, the integration of research findings into day-to-day management. The first involves collaboration between research groups and personnel to provide for integration of programs and projects, while the second and more challenging, requires a closer collaboration between fire research and land and fire management personnel. Professional fire management now requires the continuing input of research data, but the transfer of these data and findings in meaningful and useable formats to management remains a challenge – for both researchers and managers.
To achieve an effective link between research and fire management personnel several issues need to be addressed in the near future.
1. A high level of input to and co-ordination of fire research programs and/or projects by a nationally recognised authority or committee
A high level of co-ordination and integration of fire research will always be difficult to achieve due to the very different research directions and priorities of the various research institutes and land management agencies but a greater commitment to achieve closer working relationships must be made. The funding arrangements and priorities of the individual research organisations are themselves a major limitation to the achievement of acceptable levels of research collaboration and co-ordination. The funding base of the research organisation unfortunately is often the determinant of what fire research will be carried out, and what priority is actually placed on fire as as research field. This is an inappropriate basis for any research. The latter has contributed to the general lack of long-term fuels and ecological fire research programs as funding is seldom provided for long-term monitoring programs.
2. The establishment of a national bushfire research unit
The National Bushfire Research Unit was established following the 1983 Ash Wednesday bushfires in Victoria and South Australia, but unfortunately was disbanded after completion of the primary project which it was established to address.
A similar national research centre could be both a centre of excellence for fire research and also the co-ordinating authority for the integration of research carried out by other collaborating agencies. Such a centre should include and represent the full range of Commonwealth and State Government research agencies, and university and industry research personnel, to ensure the continuity and equity of funding.
3. The provision of additional training opportunities for fire planning and management personnel, through one or more formal and nationally recognised tertiary courses
Several short courses in fire management are currently offered by universities, eg Sydney University (OAC) and University of Southern Queensland but these are units within other land management courses. There is a demand for a nationally recognised degree course in fire science and management, to provide professional fire management personnel who can effectively interpret and use research findings in day-to-day management.
4. A greater role and input by district fire committees to the identification and support of research programs
These committees or similar exist in all States and as they have a wide representation of land management personnel, they should play a greater role in initiating and supporting management orientated fire research programs, particularly those long-term programs as noted above and which are seldom funded by research institutes. eg fuel monitoring.
These fire committees have a major role to play in bridging the gap that has long been identified, between fire research output, management demands and requirements for research and the integration of research data into management programs.
5. An increased commitment to fire management planning
Fire suppression plans have been very much a part of land management planning for many years, particularly in State forests, national parks and local government areas, many of these plans being part of a wider regional fire action/suppression plan.
On the other hand few well developed ecologically based fire management plans for these same areas are evident. Until such plans are developed the ecological issues and conflicts between fire mitigation and conservation objectives, will never be adequately resolved.
These plans should also identify priority research projects, which would benefit the further development and implementation of management activities such that no undue compromise of objectives for conservation or the mitigation of wildfire and the protection of life and property occurs.
These plans should incorporate hazard and risk assessment as well as detailed fuel analyses. The latter is the initial link between ecological considerations and that of fire mitigation and suppression programs.
6. A greater commitment to the use of geographical information systems and predictive modelling in fire management planning
Some use is made of these techniques by several government land management agencies but there is considerable scope for greater use to be made of computing techniques to improve fire management programs and to provide for the integration of research data into plans and programs.
Computer based predictive modelling programs of fuels, fire behaviour, suppression activities and fire impacts have been available since the early 1980s, but professional competitiveness between researchers and between managers in this area of applications research has been detrimental to the further development and use of these important techniques.
A large number of research papers have been published over the past three or four decades (Gill et al. 1994), but fuels research is not prominent, due possibly to the need for long-term monitoring in fuel studies. The geographical and seasonal variability of fuels even within a single vegetation community, makes it essential that research be carried out over many years to provide a full appreciation of the fuel complex such that fuel management programs including prescribed burning can be soundly planned and undertaken.
This fuels research, as with other areas of fire research, should be planned and carried out as a collaborative and national program; so fuel models for individual vegetation communities can be generated and used with confidence, as to accuracy of any fire behaviour predictions, particularly for communities which may occur over a range of geographical or edaphic environments.
Prescribed burning to manage fuels is now regularly carried out by aerial incendiary ignition methods, often covering very large areas of varying terrain within a single operation. This broadscale approach to hazardous fuel reduction is to be questioned particularly where it is implemented in hilly or mountainous terrain; as the range of vegetation and fuel types which exist over the many environmental gradients make the achievement of desired outcomes of burning all but impossible. Unless the full range of fuel types and loads are known, ineffective burning in terms of fuel reduction will occur; or at the other extreme, unacceptable fire intensities and damage to vegetation will continue to result and the practice of prescribed burning will continue to be of concern to ecologists and environmentalists.
Fuels research while attracting little attention in the past must be a high priority in the future if sound and effective fire management is to be achieved to the benefit of both the protection of life and property and conservation of native biota and other natural resources. This will only be achieved when a collaborative approach to fuels research is taken and the research data and output is effectively integrated into planning and management.
Birk, E.M. 1979, 'Overstorey and understorey litter fall in a eucalypt forest: spatial and temporal variability', Ausralian Journal of Botany, vol. 27, pp. 145-156.
Birk, E.M. & Simpson, R.W. 1980, 'Steady state and the continuous input model of litter accumulation and decomposition in Australian eucalypt forests', Ecology, vol. 61, pp. 481-485.
Brown, J.A.H. 1972, 'Hydrological effects of a bushfire in a catchment in southeastern NSW', Journal of Hydrology vol. 16, pp. 77-96
Catchpole, W. 1985, 'Predicting biomass for Casuarina nana for input into the Rothermel fire model', University of New South Wales, Mathematics Department Report, no 24/85
Catchpole 1987a, 'Fire models for heathland Part 1 – Casuarina nana', University of New South Wales, Mathematics Department Report, no. 5/87.
Catchpole 1987b, 'Fire models for heathland. Part 2 – Coastal heathland', University of New South Wales, Mathematics Department Report, no. 6/87.
Catchpole, W & Catchpole, E.A. 1993, 'Stratified double sampling of patchy vegetation to estimate biomass', Biometrics, vol. 49, pp. 295-303.
Conroy, R. 1987, 'Fuel data for fire management', in Bushfire Management in Natural Areas. Ed. R.Conroy, National Parks and Wildlife Service Sydney, pp. 43-59.
Fox, J., Fox, M.D. & McKay, G.M. 1979, 'Litter accumulation after fire in a eucalypt forest', Australian Journal of Botany vol. 27, pp. 157-165
Garvey, M. 1996, 'The use of geographic information systems to analyse wildfire threat', in Fire and Biodiversity, Biodiversity Series Paper No. 8, Biodiversity Unit, Department of the Environment, Sport and Territories, Canberra.
Gill, A.M., Moore, P.H.R. & Martin, W.K. 1994, Bibliography of Fire Ecology in Australia (Including Fire Science and Fire Management). Edition 4. NSW National Parks and Wildlife Service, Hurstville.
Good, R.B. 1973, 'Preliminary assessment of erosion following wildfires in Kosciuko National Park in 1973', Journal of the Soil Conservation Service vol. 29, pp. 191-199.
Good, R.B. 1981, 'The role of fires in conservation reserves', in Fire and the Australian Biota, Eds, A.M. Gill , R.H. Groves & I.R.Noble, Australian Academy Science, Canberra.
Good, R.B. 1982, 'The effects of prescribed burning in the sub-alpine area of Kosciusko National Park', M.Sc. thesis. Univ. of NSW.
Kessell, S.R., Potter, R.B., & Potter, M.W. 1982, Computer modelling in natural area management, Australian Natural Parks and Wildlife Service, Special Publication 9
O'Connell, A.M. 1991, 'Fuel quantity: the time dimension for forest litter', in Proceedings of the Conference Bushfire Modelling and Fire Danger Rating Systems. Eds, N.P. Cheney & A.M. Gill, pp. 109-118.
McRae, R 1993, 'Rural fire hazard assessment', in The Burning Question: Fire Management in N.S.W., Ed, J.Ross, University of New England Armidale.
Raison, R.J., Woods, P.V. & Khanna, P.K. 1993, 'Dynamics of fine fuels in recurrently burnt eucalypt forests', Australian Forestry, vol. 46, no. 4, pp. 294-302.
Rothermal 1972, 'A mathematical model predicting fire spread and intensity in wildland fuels', US. Department of Agriculture Forest Service Research Paper INT-115
Tolhurst, K.G., Flinn, D.W., Loyn, R.H., Wilson, A.A.G & Foletta, I. J. 1992, 'Ecological effects of fuel reduction burning in a dry sclerophyll forest: a summary of principal research findings and their management', Department of Conservation and Environment, Victoria, Research Report no. 349.
Walker, J. 1979, 'Aspects of fuel dynamics in Australia', CSIRO Div.Land Use Research, Tech. Report 79/7.
Van Loon 1977, 'Bushland fuel quantities in the Blue Mountains: litter and understorey', NSW. Forestry Commission Research Note no. 33