THE EFFECTS OF FIRE ON SOLUBLE ROCK LANDSCAPES
Abstract
Fire and its effects have long been scrutinised in relation to fauna, flora and erosion. Except for some causal observations in conjunction with other studies, research has not been specific to the effects on rock or landform types. The vulnerability of soluble rock landscapes to geomorphic disturbance by fire is demonstrated by the mechanical process of spalling, the chemical process of calcination and the erosional process due to sediment movement and loss of soil support when vegetation is removed. The combined action of this allows for faster weathering action and interruption to the karst geomorphic process. Fire is therefore an important component of karst management.
Introduction
The management of ecosystems is often coloured by the requirements of others to protect their interests such as commercial ventures or to protect life and property. Abundant evidence has been produced as to the degradation of Australian ecosystems by the activities of the Europeans in the last 200 years. An issue in management actions has been the use of fire as a tool and the lack of knowledge about fire in the development of our present ecosystems. This is compounded by speculation on the effects of past Aboriginal burning.
Soluble rocks, because of their very nature, often develop complex and dynamic physical systems known as karst landforms. These landforms then support a broad ecosystem below and above the surface of the earth. Karst landforms are the product of the geomorphic process and while they have morphological differences due to the climatic regime, the main landform features tend to be identical.
This work is to establish what effects fire can have on the rocks that are the geological base to produce these landforms and the interference this may cause to the geomorphic process. As these types of landforms have an underground component (caves) there is the tendency for information about past process to be stored in this environment. This information can be used to look at these past processes and answer some of the questions that are often asked.
Study Area
Location
The Jenolan Caves is one of the best known limestone caves in Australia. The area is centred at ~33047S, 150005E, and has an altitude range from 1200m at the northern end of the reserve down to 635m where the river flows across the eastern boundary of the Reserve.
The climate at Jenolan is cool and moist with an average rainfall of 800mm. Many days are foggy and misty. This allows wet sclerophyll forest to develop in many of the small side valleys while on the more exposed ridges dry sclerophyll forests predominate. The soils tend to be skeletal on the limestone with low sparse shrubby vegetation.
Fire History.
Fire history is not well documented at Jenolan but it appears that the most serious fires have reached the area from the direction of the Five Mile Hill road. The Jenolan River Valley has not had a serious fire for decades, although some minor fires have been ignited by lighting strike. Several areas of the reserve have been subject to controlled burns, the most frequent being around the Five Mile houses.
According to the 1980 report by the Bush Fire Council, in the past forty years, fires have started during the dry spring months in the lower Kowmung and the Jenolan River. These have burnt upwards along these water courses in a westerly direction for long periods, emerging along a line from Mt Werong north to Jenolan State Forest where the first effective suppression could be undertaken. The times of reaching this point have been usually in early summer. During the 1956-57 fire season several cottages were lost within the Jenolan Caves Reserve area during a major fire.
Geology
Most work to date has been carried out on the Jenolan Caves Limestone which is on the margin of the Tasman Geosyncline. The Jenolan Caves Limestone is Upper Silurian in age and unconformably overlies laminated cherts and andesites to the west and silicic volcaniclastics to the east. The Silurian sequence is unconformably overlain to the east by upper Devonian Lambie Group sediments and the Silurian rocks are intruded by granite plutons NE and S of the caves. Permo-Triassic rocks occur at ~1150m within 3km of the caves and Lower Carboniferous rocks are overlain by Permian and Triassic sediments 15km to the east (Kiernan 1986).
The limestone is a detrital carbonate sequence and consists primarily of algal-stromatoporal biomicrite and biomicrudite, recrystallised in places to sparry calcite, and is massive, compact and consists of 96-99% CaCO3 (Chalker, 1971).
Mineralogy
An analysis of limestone from the Grand Arch, Jenolan Caves, from Carne and Jones (1919).
| CaCO3 | 97.62% |
| MgCO3 | 0.91% |
| Fe2O3 + Al2O3 | 0.18% |
| Gangue | 1.04% |
| Sum | 99.75% |
Consequence Of Fire
Spalling
The first noted event of breakdown of limestone (spalling) by fire is made by Kranjc 1987 in reference to a fire in Postojna Cave, Yugoslavia "made a lot of damage: first parts of galleries were covered with dense soot and the whole blocks of rock collapsed because of heat."
Spalling has been recorded on rocks other than limestone (Emery 1944; McNickle 1985) as a result of fires. It is a similar process to exfoliation. Spalling on many rock types, and in common with larger-scale exfoliation, is due to a variety of processes. Spalling on non-soluble rocks need not follow lithological boundaries.
E. Hamilton-Smith, 1989 notes at Cutta Cutta, the first concern is the direct effect of repeated fire on minor solution features (karren). Effects, chiefly blunting of sharp edges and minor spalling, appear to be confined to areas of high fuel load especially where fallen trees lie across the karren. Frequent fires probably enhance this effect.
In the Junee-Florentine area, Tasmania, some very large pieces of spalled karren were seen after regeneration fires. Evidence of spalling on limestone at Jenolan comes from efforts of residents to landscape the gardens around the cottages and to clear limestone outcrops. The practice of lighting fires over them results in the outcrop breaking up into small pieces that can be removed.
Spalling is known to occur after fires in areas which are not usually weathered in this way. This has been shown to have destroyed surface solution features in areas not normally subjected to fire, if fuel reduction burning is carried out. The Bendethera limestone, being already prone to this type of mechanical weathering, is particularly affected by fires which appear to have contributed to the production of colluvium. The fact that the limestone has an almost continuous and low cover of Acacia covenyi means that comparatively more heat will be generated during the course of fires, at a close range to the limestone (Ian Houshold 1989).
The underground sediments at Jenolan have laminae of limestone rock fragments that may relate to single fire spalling events. While fire spalling may be the source of the fragments, other processes could be due to cold climate conditions. Marques, 1990 found that environmental modification derived from fire devegetation induces more severe soil temperature conditions and gives rise to periglacial type processes normally absent in this zone. Accumulation of all this material in one single stratum underground may relate to cave entry. Stratified surface deposits with fire-generated layers may be reworked and enter a cave system as a mixed deposit of varying age. In that case, correlation with surface deposits is necessary.
Calcination
The burning of limestone (known as calcination) for the production of quicklime or unslaked lime (calcium oxide) was a major industry associated with limestone outcrops from the early 1800s up to recent years (the dissociation temperature for calcite is 898°C). Dissociation always proceeds gradually from the outside surface inward (Boynton 1980). Azbe 1939 has detected traces of surface dissociation as low as 742°C with high calcium stone.
Acid insoluble residues averaged 0.68% (n=6) and consisted of silica(quartz). Reduction of the limestone by acid to insoluble residues, revealed the presence of unidentified hydrocarbons. At high temperatures these would become volatile and ignite, therefore their presence might reduce the threshold temperature at which limestone disaggregation occurs.
Materials lost during burning were not identified but moisture may account for a large proportion. When water is added to the calcinated material, the carbon monoxide combines with it giving off heat and forming calcium hydroxide (commercially known as slaked lime). If temperatures from a bush fire are sufficient to cause the calcining to occur, any following rain would then create slaked lime; this is a powder that would erode at a fast rate. This needs to be investigated under controlled field conditions and it may account for pale calcium rich laminae in some cave sediments. Although such laminae may be due to phosphate minerals.
Microscopic examination of the slaked lime showed a proportion of silica consistent with the acid insoluble residues. Fires burning hot enough to cause calcinination would release the silica micro-clasts but such release would also be expected from normal weathering processes.
Thin sections showed a mineralogical dissimilarity between the fossils and the micrite cement, disaggretion would then be uneven allowing the fossils to stand out. Fire burnt limestone does tend to show this. Weathering rates between fossils and micrite would also be different.
Samples of limestone heated to 800°C in the laboratory exhibited disaggregation from the surface inwards to a depth of 1cm. Loss of weight was proportional to size with 23.03% loss to sample of 39.74gms with a 15.63% loss to a sample of 77.87gms. There is a clear relationship between spalling and surface area.
Erosion
"In the Holocene the region underwent, like most of the Dinaric Karst, the first settling intensification in the first millenuim BC in the time of Illyrians who were known in history as pastoral livestock breeders. To obtain pastures they burnt forests. In the relatively humid sub-Mediterranean climate the annual fall of leaves is substantial in such forests, and since leaves can not be blown away because of the great thickness of woods, they are permanently spread also over less-pointed outcropping stones. For this reason the stony karstic appearance was blurred in forests. It became visible after the first mass burnings of forests, when the humus cover on the higher stones was also burnt and wind could blow soil away from the sporadic depressions. B. Gusic (1957) could personally follow the changes of landscape character on the karst after the burning of forest at the end of the 19th century in the southeastern Dinaric karst" (Gams 1991).
Kiernan (1988(d)) considers the effects of regular control burning and that of a severe naturally-occurring fire that may result without hazard control-burning. Kiernan comes to the conclusion that while a natural fire may cause erosion and sedimentation to occur on a catastrophic scale, the fact that underground drainage channels can be affected by minor sedimentation means that the presence or absence of sediment is of greater importance than its intensity.
Repeated regular firing, such as is used to protect rainforest remnants, can contribute to soil instability both at the surface and at depth — there is evidence of this on the Park with subsidence dolines being noted near the Guy Cave and Sculpture Cave outcrops. Subsidence dolines are not unnatural features but their restriction in this environment to those areas demonstrating frequent repetitive burning gives rise to concern.
The watertable caves common in the Tindall Limestone are particularly prone to sedimentation problems as the incoming waters have far higher sediment-carrying capacities than the ground-work outflows. The role of fire in enhancing subsidence doline development is twofold. Firstly, fires often reduce infiltration capacity and expose large areas of the soil surface leading to runoff and ponding around the perimeter of rock-outcrops encouraging transport of fires into cave systems below. This can be through joints and solution tubes at sizes as small as three millimetres or perhaps even less. The second effect of fire in encouraging soil breakdown is in the reduction of strength created by destruction of the root mass either by straight decomposition and the loss, by smouldering, of larger roots. In either case soil transport into cave systems is readily enhanced with possible deleterious effects on the karst systems (Hamilton 1989).
Alternate Evidence
Charcoal deposits
Charcoal deposits or laminae are frequently used to develop fire histories and are often seen in underground sediments. As they often gather in the ponded areas of the sediments and because they float, it is suspected that accumulation may not be related to individual fire events. Some reworking is common at Jenolan.
Sequences of charcoal in the alluvium of McKeown's Valley have been identified by the author. A study of a doline situated in the playing fields revealed a significant amount of carbon. "Every interval had a significant charcoal component, the charcoal having formed as a result of bush fires. Thus bush fires have been an active feature of the area for a long period of time" (Kelly, 1988).
Recorded fires in the Jenolan Caves area in 1957 and 1923 imply that the fires passed along the ridges well above the karst. These fires would have been major contributors of charcoal to the karst sediment yet have no relationship to fires on the limestone outcrops. Charcoal on its own cannot be regarded as a signature of fire on limestone itself but rather indicates fire in the regional context.
Management
Kiernan (1988(b)) considers the style of fire management and hazard reduction burning as being of considerable concern to cave managers. Kiernan's concerns are principally associated with the effect of fire on soil erosion rates and the resulting effects of sediment upon the karst. The NSW NP&WS (1987) also notes that fire can influence cave microclimates, affect minor solution features and alter hydrologic regimes, cultural values and vegetation.
Open cave systems have always been subject to the effects of fire. Caves throughout Australia show evidence of this, whether it be ash layers seen within sediments or discolouration of speleothems (e.g. Yarrangobilly Caves, NSW). The question that arises is to what degree should controlled burning be used in the process of hazard reduction.
The NSW NP&WS(1987) considers that fire regimes on limestone should not differ from those occurring naturally in a particular environment (referring to the Cooleman Plain area of NSW).
Kiernan (1988(d)) considers the effects of regular control burning and that of a severe naturally-occurring fire that may result without hazard control-burning. Kiernan comes to the conclusion that while a natural fire may cause erosion and sedimentation to occur on a catastrophic scale, the fact that underground drainage channels can be affected by minor sedimentation means that the presence or absence of sediment is of greater importance than its intensity.
Cameron McNamara (1988) considers that fuel reduction burning is required for the management of the Jenolan Caves (NSW). They suggest that where possible hand clearing be given priority over localised hazard reduction burning which should only be carried out where there is no alternative in the aim of protecting life and property. They also consider that any hazard reduction burning that is required, should take place in line with the natural fire regime of the vegetation and that naturally-occurring fire should only be controlled if it threatens life or property.
Where possible the fire history of an area should be established. No broad scale hazard reduction burning should take place unless it conforms to the natural fire regime of the area. If the natural fire regime of the area is not known, no broad scale hazard-reduction burning should take place. Localised hazard reduction burning for the purposes of the protection of human life (i.e. small areas around houses and access routes) can be carried out if there is no alternative, such as viable hand-clearing.
In any area where hazard-reduction burning is to be carried out there is a need to be fully aware of the environment to be burned. This would mean the preparation of a detailed fire management plan. Such a plan should include thorough documentation of vegetation associations, flora, fauna, fuels and fire control advantages. It should provide for proper planning for future prescription burns and allow for a co-operative effect with neighbouring land managers (Gellie, 1987).
Where possible fire management practices for karst areas should be on a catchment wide basis.
There is little or no use in adopting karst protective fire management practices in the areas around cave areas when sediments from areas of the catchment with different fire management practices are dumping sediments into the cave area (Elems 1990).
Fire has long been a factor in the environment. However little realistic information, from a management perspective, on historical and prehistoric patterns of fire is available. For example in the Sassafras Creek Valley, sediment deposits in both Baldocks and Quarry caves are clearly associated with fire in the landscape albeit 1800 and 12000 years BP respectively (Kiernan 1989). How these two events relate to the fire frequency and intensity regimes at those times, or indeed today, is far from clear (Spate 1990).
Fire frequency, intensity and seasonality are major factors influencing the distribution of faunal and vegetation communities. The latter are, of course, prime determinants of karst processes.
While there is very little documentation on the effects of fire on karst landforms there is much evidence on effects of fire on flora, fauna, soils and catchment process. In the Junee-Florentine area fire has caused limestone to spall and calcination destroying karren features. Hamilton-Smith et al (1989) note that in the Cutta Cutta karst, Northern Territory, fire has contributed to soil instability resulting in accelerated development of subsidence dolines.
Wildfires are events over which mankind has little control. While procedures can be put in place to safeguard life and property, fire-vulnerable natural resources also require protective measures. The protective value of control or prescribed burning may be arguable. However, prescribed fires do increase the frequency of burning which may have long term effects on soils, biotic communities and catchment stability.
The development and use of fire trail networks may aggravate many other problems such as the spread of weeds, soil erosion and the accessibility of sensitive sites to increased recreational and other pressures. There is a need to develop prevention and suppression strategies for karst areas specifically as distinct from general fire planning.
The Kosciusko National Park Plan of Management has, amongst its specific objectives (Section 8.3.1), the following: "to protect important natural features, especially alpine areas, restricted, rare or endemic plant or animal communities and limestone areas". Fire management is important to the protection of these features.
The Management Unit, occupying as it does much of the headwaters of the Goodradigbee River, has important catchment values contributing very high quality water to the Burrinjuck Dam on the Murrumbidgee. Because of the nature of the karst hydrologic system, Cave Creek supplies the bulk of the base flow in the Goodradigbee during dry periods. Flow durations from Blue Water-holes system are very long and considerably buffer seasonal and longer rainfall variability. As well as catchment values the area has other important values as discussed elsewhere in this plan.
Fire can, of course, influence many of these values; for example, cave microclimates can be changed or minor solution features destroyed in addition to changes to vegetation, soil stability, hydrologic regime and cultural values. Fires are also part of the environment but at unknown frequencies and intensities.
Secondly, the open woodland and cave land surfaces require a fire management strategy to coincide with the conservation, landscape and regeneration work that is to be carried out. In general terms fire is considered incompatible with limestone environments. as such, mechanical fire hazard reduction is required within the Reserve together with protective measures to ensure the reserve is not liable to fire from adjacent freehold lands and from the Mitchell Highway (NSW Department of Lands, 1989).
Conclusions
That fire breaks down the rock is evident from the examined areas and is activated by a mechanical and chemical process resulting from the heat of the fire. There is sufficient evidence that erosion is also an environmental component as a result of fires. Taken in the total context, fire alters, accelerates and synthesises the geomorphic process on soluble rock landforms. The question of fire frequency as a result of natural igniters and does controlled/prescribed burns have the same impact is then raised. It is noted that on the calcarenites of southwest Western Australia that damage to the vegetation was not as severe on the prescribed burnt area of Margaret River as the bushfire area of Yanchep. But the limestone in both areas had spalling of equal concentration and severity. There is also evidence while mostly anecdotal, but supported by some fire records, that naturally occurring fires may be rare on limestone outcrops in the Eastern states. "Observations at Bendethera indicate that most major bushfires tend to burn around the limestone outcrops, not over them" (Davies 1986).
References
AZBE, V.J. (1939) ASTM Symposium on Lime, 15pp
BOYNTON,ROBERT S.1980. Chemistry and Technology of Lime and Limestone. 2nd Edition. A Wiley-Interscience Publication. JOHN WILEY & SONS, INC
CAMERON MCNAMARA (1988). Jenolan Caves Reserve Draft Plan of Management, Volume 1, Draft Plan, Cameron McNamara
COOLEMAN PLAIN KARST AREA MANAGEMENT PLAN (1985)- a Supplementary Plan to the Kosciusko National Park Plan of Management. NSW National Parks & Wildlife Service. Pages 53 & 54
CHALKER, L. 1971 Limestone in the Jenolan Caves Reserve. BSc Hons. thesis, University of Sydney
DAVIES, S.J. AND GILLIESON, D.S. 1986. The Limestone flora at Bendethera, Unpublished Report to NPWS South-Eastern Regional Office. 8pp
PADOV. pp.1-8
GELLIE, N. and DE GOVRIK, K. Fire Management Plan for Winburndale Nature Reserve. New South Wales National Parks and Wildlife Service
MARQUES,M.A. SOLE,A. MORA,E. LLASAT,C. (1990) Frost Action Features as a Consquence Of Fire Devegetation In A Montane Mediterranean Environment. Catena by Catena Verlag, D-3302 Cremlingen-Destedt, W. Germany vol.17,pp 163-174
Guide Book, Edit I. Gams, P. Habic. Ljubljana. pp.29-38
McNICKLE H P 1985 'An Introduction to the Spear Hill Rock Art Complex, Northwestern Australia' Rock Art Research 2 [1]; 48-64
EMERY K.O. 1944 'Brush Fires and Rock Exfloration.' American Journal of Science 242: 506-508
HAMILTON-SMITH. E, HOLLAND. E, MOTT. K, SPATE. A, (1989), Cutta Cutta Caves Nature Park. Draft Plan of Management, Conservation Commission of the Northern Territory
HOUSHOLD, IAN. (1989), Bendethera Fieldtrip report 7 - 9/2/89. NPWS South Eastern Region Report
KELLY, B. 1988. The Hydrogeology of the Northern Limestone Jenolan Caves and related Land Management Issues, BSc Hons Thesis, University of N.S.W.