Proceedings of the conference held 8-9 October 1994, Footscray, Melbourne
Biodiversity Series, Paper No. 8
Department of the Environment, Sport and Territories, 1996
16. The effects of fuel reduction burning on forest soils
Victorian College of Agriculture and Horticulture
Throughout Australia approximately a million hectares of forest are burnt annually with low-intensity, prescribed fires in order to reduce the likelihood of wildfires or to minimise their impact when they occur. Fuel reduction, or prescribed burning, has been a forest management strategy in most States for approximately thirty years. While it is assumed that wildfires have a greater effect on forest soils than fuel reduction burning, little work has been carried out to assess the impact of low-intensity fires on the erosion hazard, nutrient status and physical properties of forest soils.
Studies in the Gembrook State Forest indicated that the erosion hazard was substantially increased (at least ten fold), that significant nutrient movement occurred and that some important physical properties were affected by a fuel reduction fire. Runoff plots were constructed prior to a fuel reduction fire, the litter layer sampled and some soil physical properties measured. Following the fire, runoff from the plots was collected weekly for approximately four months and sediment, calcium, magnesium, potassium, sodium and phosphorus levels in the runoff were determined.
The soil of a forest ecosystem is one of its principal components and it may be that the short-term protection afforded by fuel reduction burning could cause a long-term loss of productivity. This would be the likely consequence of a fire regime which increases soil and nutrient loss. Approximately 100 times more calcium and magnesium, at least 30 times more phosphorus and 20 times more potassium were lost in runoff from burnt plots. However some major concerns regarding sustainable production are the volatilisation of nutrients during a fire and the regular short circuiting of the biogeochemical cycling of nutrients. A substantial proportion of nutrients are cycled through the litter on a forest floor. It is precisely this layer of litter that a fuel reduction fire aims to destroy. The litter layer must also ultimately supply much of the soil organic matter. The long term implications of a high frequency of burning on soil organic matter levels has yet to be determined. The forest litter also provides a protective layer which reduces the erosive power of rain.
Forest managers face a real dilemma over the use of fuel reduction burning. On the one hand it is an easily applied technique which provides some short-term protection from devastating wildfires. On the other hand its long-term impact on forest productivity, resulting from soil and nutrient losses and changed soil properties, certainly need to be more carefully assessed.
Key words: fuel reductions, burning effects, forest soils, erosion hazard, nutrient status, physical properties.
Fuel reduction burning is a routine management practice in Victorian forests and throughout Australia. Nearly three percent, or more than 200,000 hectares, of the States forests are burnt annually in an attempt to minimise the incidence and intensity of wild fires (Johnston et al. 1982). Throughout Australia approximately a million hectares of forest is burnt with low-intensity fires annually. It is possible then to estimate that some twenty million tonnes of forest litter may be burnt annually and this figure ignores the mass of understorey vegetation also burnt. The rationale behind the use of fuel reduction burning is that damage to forests depends largely on fire intensity and that excessive damage to forests can be avoided by preventing a large build-up of fuel. The practice of fuel reduction burning aims to reduce the levels of combustible fuels below a threshold level of ten to fifteen tonnes per hectare (Good 1981).
The immediate effects of fire on forest vegetation depends largely on the intensity but longer-term effects are influenced by the frequency and season of burning. These three aspects are the components of a fire regime. While fire is a natural environmental variable in the Australian landscape – to attempt to exclude it is naive and foolish – it is worth emphasising that specific native forest ecosystems have evolved under the influence of specific fire regimes rather than fire per se. There is a large and growing body of information regarding the effects of a changed fire regime on vegetation structure and floristic diversity in forest ecosystems. However while the lower and middle strata of vegetation are likely to be most affected by a changed fire regime the forest product we are trying to protect may also suffer.
While the soil of a forest ecosystem is one of its principal components much less work has been carried out to assess the effects of a changed fire regime on forest soils. If we are to manipulate, or simplify, forest ecosystems then we need to be very sure of the effects of our management practices on forest soils. The impact of repeated low-intensity fires on nutrient and soil physical properties must be assessed. The cumulative effects of high frequency prescribed fires on forest soils may be as significant as less regular, high-intensity wildfires. The selection of the most appropriate fire regime should depend on: a desire to reduce the likelihood of wildfires, and reduce their intensity if they do occur; maintenance of the long-term stability of the forests and their long-term productivity.
It may be that the short-term protection afforded by fuel reduction burning causes a long-term loss of productivity. This would be the likely consequence of a fire regime which increases soil and nutrient loss, impacts on nutrient pathways or adversely changes soil physical properties.
Regardless of its intensity any fire will lead to some increase in the exposure of the surface soil to the agents of erosion. The potential erosion hazard will be influenced by the soil erodibility, the erosivity of subsequent rainfall, exposure of the soil to wind and site characteristics such as the degree and length of slope. The speed of litter recovery will also be important as will the time taken to re-establish a vegetation cover on the soil. This can occur quickly following a fire as the fire mobilises nutrients and warms the soil as well as scarifying hard seed coats.
A high frequency of burning can eliminate some species if they are burnt before they set seed. The ash bed effect reflects a short-term improved availability of nutrients. Balanced against this effect is the possibility of a long-term depletion of nutrients by leaching, in run-off or, and possibly most significantly, through volatilisation. Raison (1980) estimated that these losses may be in the order of 60 per cent of the nitrogen and sulphur and 50 per cent of the phosphorus in the fuel. The impact of repeated low intensity burns on the nutrient status of the forest soil has still to be assessed.
In order to measure soil and nutrient loss following a prescribed low intensity fire, runoff plots (Riley et al. 1981) were established at two sites prior to fuel reduction fires in autumn 1985. Two plot lengths were used in an attempt to establish the most appropriate size of this type of runoff plot. These sites were in the Wombat and Gembrook State Forests. Eight plots were constructed at the Wombat site approximately six kilometres south-west of Daylesford and six plots were constructed in the catchment of the Black Snake Creek in the Gembrook State Forest. Fuel was sampled (Tolhurst, K. 1985 pers comm; Bull, H. 1985 pers comm) and soil samples collected from the surface 10cm prior to the fires taking place. Fire data were also provided by Department of Conservation, Forests and Land staff.
Despite the action of raking litter from around the control plots, covering them with aluminium sheeting and supervising the burn, the fuel on the control plots at the Wombat site caught fire when the fire was carried out on 15 March, 1985. Fortunately the control plots were protected at the Gembrook site. Following the fires, runoff was collected weekly from both sites, sub-sampled (Muir-Smith 1986) and the sub-samples analysed for sediment, calcium, magnesium, potassium, sodium and phosphorus. Sampling continued for over three months.
Undisturbed cores were also taken from burnt and unburnt sites for bulk density determinations. The dispersive nature of the soils prior to, and following, the fires were assessed (Rengasamy et al. 1984). Electrical conductivity, pH and texture were also assessed both prior to, and following, the fire at Wombat and from burnt and unburnt sites at Gembrook.
Due to the destruction of the control plots at the Wombat site during the fuel reduction fire, no basis for comparison exists between burnt and unburnt plots at that site. The loss of the controls was unfortunate since runoff and sediment yields were higher at the Wombat site despite lower rainfall and lower slope angles of plots (Hall 1986).
The higher sediment yields were probably due to the finer textured soils at Wombat and the lower sorptivity of that soil. However since comparison between burnt and unburnt plots can only be made at the Gembrook site only those results are presented.
Variation in the micro-topography of the plots greatly affects runoff characteristics. One burnt plot consistently provided substantially lower yields of runoff. This was the consequence of a small gully which effectively shortened that plot (Y) to two metres in length. This gully only became apparent following removal of the litter during the fire.
With the exception of the aberrant plot Y the burnt plots generated approximately 40 times the runoff of the unburnt control plots (Table 1).
Plot W 4m x 2m control
Plot X 10m x 2m control
Plots U and Y 10m x 2m burnt
Plots V and Z 4m x 2m burnt
Clearly, removal of litter and ground cover increased nutrient and sediment movement off the burnt plots.
In Table 1 it can be seen that sediment was moved off burnt plots at a rate of approximately ten times that from the unburnt control plots.
(iii) Monovalent cations
The losses of the monovalent cations Sodium (Na+) and Potassium (K+) were at least twenty times greater than for the control plots (Table 1).
(iv) Bivalent cations
There was approximately one hundred times the Calcium (Ca 2+) and Magnesium (Mg 2+) lost from burnt plots (U, V and Z) than from the control plots (Table 1).
Only very small amounts of soluble phosphorus were lost from the burnt plots in surface runoff (1.82-0.08mg/square metre). These could still represent significant losses of phosphorus from the nutrient cycle (Muir-Smith 1986). Of more concern is the possible volatilisation of phosphorus which occurs at relatively low temperatures (approx. 550°C) and the interruption of the biogeochemical cycling of nutrients through litter. Attiwill (1980) estimated that 32 per cent of the annual demand for phosphorus in a Eucalyptus obliqua (L' Herit.) forest was cycled through the litter. It is precisely this litter that is the target of a fuel reduction fire.
(vi) Soil texture
The results of soil particle size analysis indicated a decrease in the percentage clay content in the surface 10cm following the fire at Gembrook (Table 2)
|Fraction (%)||Unburnt 21-March||Burnt 21-July||Unburnt 31-May||Burnt 31-May|
* fine sand 8%; coarse sand 54%
+ fine sand 10%; coarse sand 53%
(vii) Bulk density
Bulk density increased at Wombat and decreased at Gembrook following fire. Such apparently contradictory results were referred to in a literature review by Clinnick (1984). The decrease at Gembrook is possibly related to the indicated change in texture.
|Site - date||Bulk Density (g/cm³)||% Porosity|
|Wombat pre-fire - 21 March||1.363||48.57|
|Wombat post-fire - 12 July||1.459||44.94|
|Gembrook unburnt - 31 May||1.296||51.09|
|Gembrook burnt - 31 May||1.134||57.21|
(viii) Electrical conductivity and pH
No substantial change occurred as a result of the fire in pH or electrical conductivity.
(ix) Soil dispersion
No spontaneous dispersion occurred but all samples dispersed upon mechanical agitation; this indicated an increased erosion hazard following fire when the protective litter layer is removed (Rengasamy 1994 pers comm).
A major concern regarding the practice of fuel reduction burning is that pre-fire fuel levels may be regained in four years or less (Raison et al. 1983). This places in doubt the validity of short-term benefits of protection from wildfires when weighed against the long-term implications of a changed
fire regime. The risks involved include increased erosion rates, nutrient losses in smoke and ash (Constadine 1984), nutrient loss or relocation in runoff (Hall 1986) and possible changes to soil physical properties.
Little information is known about the long-term effects of an altered fire regime on levels of soil organic matter; although a reduced level would seem the likely consequence of the regular burning of forest litter. Soil organic matter should not be confused with measurements which include charcoal; this distinction is necessary as the amount of charcoal present is massive but relatively inert.
Depending upon the texture of the soil a reduction in organic matter will have implications for nutrient and water holding or soil structure. In sandy soils much of the cation exchange capacity and water holding capacity is associated with the soil organic matter fraction. In finer textured soils the soil organic matter can be a significant contributor to improved soil structure.
Similarly, little is known of the likely effects on forest productivity of the interference in the biogeochemical cycling of nutrients through the forest litter. Attiwill (1980) showed that the proportion of gross annual demand of nutrients in a Eucalyptus obligua forest supplied through the litter were: in the order calcium (60 per cent), > magnesium (40 per cent), > phosphorus (32 per cent), > potassium (17 per cent). It is precisely this litter which is the target of a fuel reduction fire. The short circuiting of these nutrient pathways, particularly in respect to phosphorus with its low volatilisation temperature and general low availability in Australian soils, is a major concern. The volatilisation of nitrogen can also occur at the relatively low temperatures experienced in fuel reduction burns, and could also be significant for the long-term productivity of forests. The loss of sulphur is another cause of concern; its significance as a nutrient has only recently been fully appreciated.
Soil characteristics such as the degree of dispersion (Durgin 1985) can also be altered by fire. Ash leachate can change the dispersion characteristics of a forest soil making them more prone to erosion. Bulk density and surface texture can also be altered by fire. These changes are likely to vary depending on both the above and below ground components of the particular forest ecosystem as well as the nature of the fire.
Long-term studies are required to evaluate the effects of a changed fire regime on soil properties and forest productivity; when ecosystems are manipulated in such a dramatic way we need to be confident of the long-term consequences. With the current level of knowledge we can do little more than guess, but the short-term loss of nutrients and soil indicate that the long-term productivity of forests is at risk.
It should be emphasised that this study is reporting results obtained from a single fire event and it is dangerous to extrapolate the results too far. However the results of the study in the Gembrook Forest indicated that nutrients are mobilised following a fire and that some of the mobilised nutrients could be lost in runoff or at least relocated downslope. Significant soil properties can also be altered, including dispersion characteristics which could lead to an increase in the erosion hazard. Of much greater concern are the consequences of: a long term reduction in soil organic matter levels; the loss of substantial amounts of nutrients in smoke; the short circuiting of biogeochemical pathways; the possibility of wetter soils as a consequence of a reduction in understorey vegetation; the effect that wetter soils may have on the spread of Phytophthora cinnamomi; the effect that such a dramatic change to the fire regime will have on the carbon cycle and hence on other nutrient cycles.
High frequency, low-intensity fires can effectively simplify the forest ecosystem. The risk we take is that simple systems are less stable. When we manipulate a forest ecosystem we should be confident of the long term consequences, but with our current level of understanding we can do little better than guess about the possible implications.
Results of phosphorus analysis were provided by J. Muir-Smith (1986). His contribution is gratefully acknowledged. I also thank staff of the V.C.A.H.-Burnley for their assistance in the preparation of this paper and staff of the Department of Conservation, Forests and Land for data and assistance during this research.
Attiwill, P.M. 1980, 'Nutrient cycling in a Eucalyptus obliqua (L'Herit). Forest IV nutrient uptake and nutrient return', Australian Journal of Botany, vol. 28, pp. 199-222.
Clinnick, P.F. 1984, A Summary – Review of the Effects of Fire on the Soil Environment. Soil Conservation Authority, Victoria.
Constadine, M.L. 1984, 'Prescribed burning and forest nutrition', Ecos, vol. 42, pp. 9-12.
Durgin, P.B. 1985, 'Burning changes the erodibility of forest soils', Journal of Soil and Water Conservation, vol. 40, no. 3, pp. 299-301
Good, R.B.1981, 'The role of fire in conservation reserves', in Fire and the Australian Biota, (Eds) Gill, A.M., Groves, R.H. & Noble, I.R. Australian Academy of Science.
Hall, R.G. 1986, The effects of fuel reduction burning on the physical and chemical properties of forest soils and the erosion hazard of these fires. M.Env.Sci. thesis, Monash University, Melbourne.
Johnston, J.B., McKittrick, J., Flynne, D.W. & Brown, H.G. 1982, Fire Protection and Fuel Reduction Burning in Victoria. Report to Minister of Forests, September 1982.
Muir-Smith, J. 1986, The effects of fuel reduction burning on nutrient cycling, surface runoff and sediment loss in a eucalypt forest, M.Env.Sci. thesis, Monash University, Melbourne.
Raison, R.J. 1980, 'A review of the role of fire in nutrient cycling in Australian native forests, and of methodology for studying the fire-nutrient interaction', Australian Journal of Ecology, vol. 5, pp. 15-21.
Raison, R.J., Woods, P.V., & Khanna, P.K. 1983, 'Dynamics of fine fuel in recurrently burnt eucalypt forests', Australian Forestry, vol. 16, no. 4, pp. 294-302.
Rengasamy, P., Greene, R.S.B., Ford, G.W. & Mehanni, A.H. 1984, 'Identification of dispersive behaviour and the management of red-brown earths', Australian Journal of Soil Research, vol. 22, pp. 413-431
Riley, S.J., Crozier, P. & Blong, R.J. 1981, An inexpensive and easily installed runoff plot. Journal of the Soil Conservation Service NSW, vol. 37, no. 3, pp. 144-147.