There are several issues of contention regarding the recent history of fire in Australia and the Sydney region in particular. This debate has recently been re-ignited by Flannery’s (1994) book The Future Eaters, which popularised the debate concerning Aboriginal use of fire (eg. Cleland, 1957; Tindale, 1959; Jones 1969; Hallam 1975; Horton, 1982). Flannery (1994) claimed that regular burning of the Australian landscape meant that the vegetation encountered by European settlers was a cultural artifact resulting from thousands of years of manipulation by Aborigines. This interpretation relies heavily on evidence gleaned from writings of early settlers, especially in the Sydney area, which may be misinterpreted and/or exaggerated (Williams and Gill, 1995; Benson and Redpath, 1997). Ryan et al. (1995) and Langton (1998) have since proposed the application of Aboriginal burning practices in the modern setting for conservation reasons. They suggest that the introduction of low intensity, high frequency fire will return the landscape to its 'natural' state as was maintained by the Aborigines.

The way in which fire regimes changed with the coming of European settlers is another issue surrounded in contention (Head, 1989). It is generally reported that when Aboriginal influences were removed there was a build up of fuel and the clearing of land for agricultural purposes saw an increase in the frequency and intensity of fires (Recher and Christensen, 1981; Pyne, 1992; Flannery, 1994; Kirkpatrick, 1994). A policy of fire suppression followed (Recher and Christensen, 1981; Pyne, 1992; Flannery, 1994), which has in recent times given way to hazard reduction burning (Pyne, 1992). This generally constitutes low intensity fires at pre-determined intervals often out of the normal fire season (McLoughlin, 1998).

This study aims to apply to the analysis of charcoal to a ²¹⁰Pb dated sediment core from Jibbon Lagoon in Royal National Park, to the south of Sydney. Notably, this differs from previous palaeoecological studies in the Sydney region in several ways:

1. the recent sediment chronology is better constrained through the application of ²¹⁰Pb dating;
2. it utilises larger charcoal size fractions; and
3. it attempts to compare the sedimentary charcoal record to the recent history of fire at the study site to test the methodology.

Jibbon Lagoon (located 20km south of the Sydney CBD in Royal National Park, NSW at 34° 5’S, 151° 9’E) is a fresh-water shallow lagoon occupying a deflation hollow below sea level. There are no streams flowing into the lagoon: precipitation and runoff from the surrounding catchment are the only surficial flows of water. The bedrock underlying the lagoon consists of Hawkesbury Sandstone, and the main feature of the catchment is a moderately steep hill to the south-west of the lagoon that leads to the township of Bundeena. The soils of the area are highly erodible, skeletal, sandy-loams with low humus content, poor water-retaining properties and low fertility.

The Jibbon area experiences a temperate coastal climate and the fire season is during spring and summer. Royal National Park supports a complex mosaic of eucalypt woodland, heaths and open scrub communities (NPWS, 1994). The exact composition of these assemblages is highly dependant on the time since last fire as the dominant species (Banksia ericifolia, Casuarina distyla and Hakea teretifolia) are fire sensitive and are unable to persist if the fire frequency is too high or too low (NPWS, 1994; Wright, 1996).

It is thought that Aboriginal people from the Dharawal language group occupied Royal National Park for at least 7500 years prior to European settlement (Megaw, 1969; Goldstein, 1976; Flood, 1990; Wright, 1996). The abandonment of traditional lifestyles probably occurred in the early Nineteenth Century (Carter, 1969). The European history of the study area began as a large property known as Yarmouth Estate. This was subdivided in 1898 to form the township of Bundeena (Pettigrew and Lyons, 1979).

The recent fire history (back to 1942) of the study site was derived from the National Parks and Wildlife Royal National Park GIS (post 1968) and from Keith (1995) who analysed aerial photographs.

Sediment cores were removed from a site adjacent to the south-western shore, below a steep slope, where it was thought that the delivery of charcoal would be best. The charcoal content of the sediment was assessed in three size classes. Charcoal was quantified using the point count method (Clark, 1982) and wet sieving (Long et al., 1998), which resulted in size fractions of 125-250m m and >250m m. These larger size fractions should travel shorter distances and reflect local fire (Clark, 1988 1990; Morrison, 1994; Tinner et al., 1998).

Pollen slides were produced following standard palynological techniques (Faegri and Iverson, 1975; Moore et al., 1991). For the point count method at least 1650 random points were applied. The concentration of Alnus, Pinus and Banksia pollen grains present was also determined. Charcoal was expressed both as a concentration and influx, with the latter calculated using sedimentation rates derived from the ²¹⁰Pb analysis.

The amount of sand present in each of the (macro-) charcoal sub-samples was estimated. The bulk density, water content, dry weight (at 105° C for ³ 12 hours) and loss-on-ignition (2hrs at 550° C) of the samples were determined. Lead-210 dating was applied to 0-4, 4-6, 6-9 and 9-11cm intervals and contiguous 1cm depths below that to 21cm. The sediment accumulation rates were calculated based on the constant rate of supply model as described by Appleby and Oldfield (1978).

One of the aims of this research was to derive a sensitive index of fire activity by comparing documented historic fires with their sedimentary expression. It was hoped that the fire events of the recent European period at Jibbon Lagoon would be reflected as distinct peaks in the charcoal record. Investigations into the documented fire history of the Jibbon area resulted in a large number of fires, however the fires of 1942, 1946, 1964, 1977, 1988 and 1994 were considered 'signature' fires to be examined against the sedimentary charcoal record.

The fires of January 1994 and October 1988 were devastating, and hence it was expected that the upper section of the sediment core would contain abundant charcoal. In this section, however, both the concentration and influx of charcoal was relatively low. It was thought that this result may have reflected the high organic content of this upper section of the sediment column, so the (macro-) charcoal concentrations were re-calculated as number of particles per gram of material not ignited in the LOI procedure. Even with this recalculation, the most recent fires (1994 and 1988) are not unambiguously reflected in the charcoal record.

Notably several peaks in the charcoal record did occur at a depth that approximated the timing of earlier historic fires. This includes the 1976/77 fire (the charcoal peak at 4-4.5cm depth), the 1964 fire (6-7cm) and the peak in charcoal (centred on 10 cm depth) seems to well reflect the large fires of 1942 and 1946. It is uncertain why these fires show up as peaks in the record when the two most recent fires do not.

The charcoal values between the European and pre-European periods at Jibbon Lagoon are very different. The pre-European period is characterised by consistently low concentrations of macroscopic charcoal with only one apparent large local fire event (at 47-49cm). Even at it's lowest, the charcoal of the post-European period exceeds the pre-European levels, suggesting that fire has been more prevalent during this time, and that the fire regime does not mimic the pre-European pattern.

The early European period (from 16.5-19cm viz. 1850-1916) in the sediment column contained the lowest influx of charcoal of the European period. This is likely to reflect the absence of fire, an interpretation supported by the relatively high concentration of Banksia pollen as it has been well established that species of Banksia are negatively affected by high fire frequency (Keith, 1995; Keith and Tozer, 1997). This may reflect a period of little human influence in the catchment, representing the period between the decline of Aborigines in the Sydney region and the expansion of Bundeena.

The influx of charcoal is high in the section of the core corresponding to ca. 1939 to 1954 (9.5cm to 14cm). This period covers WWII, during which time there was a large influx of people moving to the growing township of Bundeena (Pettigrew and Lyons, 1979). This also corresponds to a period of increased sedimentation, suggesting that the increase in fire activity in the catchment is largely responsible for the post-European increased rate of sedimentation in Jibbon Lagoon.

The generally low levels of charcoal in the pre-European period is likely to reflect either the absence or rarity of fire, or that fires were small, and as such the vegetation surrounding any burnt patch impeded the transport of charcoal to the lagoon. The rationale for the broadcast use of fire by the Aboriginal people must be considered. In the vegetation surrounding Jibbon Lagoon there may have been little advantage in the use of fire to manipulate resources, as is implied by Jones' (1969) "Fire-Stick Farming" hypothesis.

The finding that there has been a noticeable increase in fire activity since the arrival of Europeans at Jibbon contradicts the findings of other studies carried out in the Sydney Basin (Kodela and Dodson, 1989; Johnson, 1994; Martin, 1994; Dodson et al., 1995). Nonetheless, a post-European increase in charcoal and presumably fire has been recorded before: by Boon and Dodson (1992) at Lake Curlip in east Gippsland, Victoria and by Gell et al. (1993) at a site on the Delegate River, also in southeastern Victoria.

There are a number of reasons for the differences between the current study and previous Sydney palaeoenvironmental studies. Perhaps of greatest concern is the temporal control of previous studies. No study to date in the Sydney basin has utilized ²¹⁰Pb dating, but rather they rely on the presence of Pinus pollen in the sediment as a proxy boundary between the pre- and post-European periods. In this study Pinus pollen is not present in the sediment until a depth of 12.5cm (ca. 1945). Aside from methodological differences it is possible that site specific features are of importance. This may include different groups of Aborigines, different vegetation communities or notably, the proximity of urban areas.

As concluded by Bowman (1998) the management of our conservation estate must have the clear aim of conserving extant biodiversity, meaning that the lessons of the past may not be entirely relevant given that conditions are now different. Nonetheless, knowledge of pre-European fire history may help to ensure that management decisions are made in a well informed manner. This study suggests, for example, that Aboriginal burning in the heath and Angophora forest that dominates the Jibbon Lagoon catchment was relatively rare. In contrast, fire in the European period, particularly from the 1930s onwards, was frequent. This study thus supports the view held by Clark (1983, pp. 32) that "the frequency and areal extent of Aboriginal burning may well have been overestimated and European burning underestimated".

(References are available on request: s.mooney@unsw.edu.au)