Silcrete quarries and artefact distribution in the Central Queensland Highlands, Eastern Australia.
A geoarchaeological study of silcrete in the Central Queensland Highlands highlights the importance of geological context in understanding the procurement, use and transport of artefacts. This study shows, by focusing on a newly discovered silcrete quarry, that silcrete formed beneath basalt by silicification of stream sediments and underlying sandstone. Silcrete outcrops are largely restricted to cliffs where the base of the basalt is exposed, narrowing the location of potential quarries. Silcrete artefacts are abundant throughout the region, and there are clear patterns in the use and distribution of different qualities of silcrete. Fine-grained grain-supported silcrete is the dominant artefact lithology because of its excellent flaking quality: it is sufficiently versatile to make flake scrapers and retouched microlithic tools. At one occupation site, the mid-Holocene appearance of backed artefacts and points was accompanied by an increase in the use of fine-grained grain-supported silcrete and a decrease in the exploitation of coarse-grained matrix-supported silcrete; the latter was less suited to the manufacture of more curated tools. With increasing distance from silcrete outcrops, there is a decrease in the percentage of silcrete artefacts and an increase in the expedient use of local lithologies. However, silcrete's flaking quality justified its dispersal over distances of >80 kilometres from the nearest outcrop.
Keywords: silcrete, quarry, central Queensland, dispersal.
Silcrete is probably the most widespread artefact lithology in Australia (e.g. Mulvaney & Kamminga 1999; White and O'Connell 1982), due to its abundance and often high-quality flaking properties (Webb & Domanski 2008). In the Central Queensland Highlands, stone tools from archaeological excavations in caves and rock shelters include significant amounts of silcrete, but because it was not identified as such in some previous archaeological studies--for example, Mulvaney and Joyce (1965) -its importance as a lithic resource in this region has thus far been underestimated.
In this paper, we use analysis of samples from a newly discovered silcrete quarry in the Central Queensland Highlands at Ogres Thumb, together with data from previously excavated assemblages at other sites, to assess the availability of silcrete, variation in its physical characteristics and the likely distribution of quarries in the region. We also consider the factors that may have influenced the selection of silcrete by Aboriginal toolmakers at different stages in the cultural sequence. However, because studies of raw material should incorporate a solid understanding of its geological distribution and physical properties (Tykot 2003: 63), we first discuss the physical characteristics and formation of silcrete.
WHAT IS SILCRETE?
Silcrete is a term first used by Lamplugh (1902) to describe a strongly indurated material resulting from surface or near-surface low-temperature silicification of weathered bedrock, regolith and/or unconsolidated sediments (Summerfield 1983; Webb & Golding 1998). Silcretes have been reported from all continents except Antarctica, but are most widespread in Australia, southern Africa and Europe (Nash & Ullyott 2007). Silcretes are brittle, often with a lustrous conchoidal fracture, and very resistant to weathering. Silcrete layers generally have limited vertical thickness (<2 metres) and variable lateral extent, and frequently outcrop as low clifflines and cap-rocks on mesas.
Silcretes can be subdivided into pedogenic, groundwater and pan/lacustrine types (Nash & Ullyott 2007). Pedogenic silcretes have irregular bases that merge into the underlying sediments, and normally contain structures indicative of soil-forming processes--for example, nodules, cutans and columns--that overwrite the original textures of the parent material. By contrast, groundwater silcretes are massive, with sharp upper and lower boundaries, exhibit good preservation of host structures and lack pedogenic features (Ullyott et al. 1998). They probably reflect silicification at the water table (Milnes & Thiry 1992). Pan/lacustrine silcretes form thin sheets or nodules within, or adjacent to, ephemeral lakes, pans or playas.
Silcretes are arbitrarily defined as containing >85% Si[O.sub.2] (by weight), but many have >95% Si[O.sub.2] (Summerfield 1983). The minor amounts of very fine-grained titanium oxide (anatase) and iron oxides present influence the silcrete colour, imparting cream/pale yellow streaks and overall red/orange colours, respectively.
During silcrete formation, there is an input of silica into the pre-existing material; this infills the original porosity and/or replaces the non-quartz component, particularly the clays (Webb & Golding 1998). As a result, silcretes contain both a detrital component inherited from the host material (mostly quartz grains) and secondary quartz precipitated during silcrete formation (Figure 1).
Silcretes result from low-temperature surface or near-surface silicification, and are not produced by metamorphic, igneous or moderate to deep burial processes (Nash & Ullyott 2007). Therefore, flint and chert (of marine or replacement origin), metamorphic quartzite and silicified sandstone not formed near the ground surface are different rock types. Silcrete may be distinguished from other siliceous rocks by its lustrous conchoidal fracture; it is so strongly cemented that it fractures evenly through grains and matrix, whereas many quartzites fracture around the grains. Microcrystalline silcrete can often be distinguished from chert by the presence of cream-coloured streaks of very fine-grained anatase (titanium dioxide) that form during the silicification process (Webb & Golding 1998). Furthermore, in situ silcrete invariably occurs as near-surface layers or nodules.
In Australian archaeology, the identification of silcrete has been fraught with problems. It was not until Hughes et al. (1973) examined artefacts from sites in southern New South Wales (NSW) that it was acknowledged that silcrete dominates many archaeological assemblages throughout Australia. Prior to this, silcrete artefacts were often identified as silicified sediment, orthoquartzite or quartzite (see discussions in Hughes et al. 1973; Sullivan & Simmons 1979).
More recently, Australian archaeologists have shown much greater recognition of the importance of silcrete as a stone resource, focusing on the quarrying, heat treatment, fracture mechanics, reduction and dispersal of silcrete (e.g. Byrne 1980; Doelman 2008; Domanski et al. 1994; Hiscock 1993; Robins 1997; Rowney & White 1997; Webb 1993; Webb & Domanski 2008). Yet the similarity of silcrete to other rock types continues to cause difficulties for archaeologists identifying lithic materials in the field, and it is important to emphasise that its two main distinguishing characteristics (smooth fracture surfaces through the grains, creamy streaks in microcrystalline varieties) will allow the positive identification of most silcrete artefacts.
Silcretes are classified according to their micromorphology; that is, the relative proportions of detrital quartz grains inherited from the host material, quartz cement and matrix. Cement forms by precipitation of silica in the original pore spaces between the grains, and has two common morphologies: syntaxial quartz overgrowths (rims in crystallographic continuity with the quartz crystals of the grains (Figure 1 A) and fibrous quartz (chalcedony) rims around detrital grains (Figure 1B). Matrix was originally composed of varying proportions of clay and silt-sized quartz grains; during silcrete formation, the clay material is replaced by microcrystalline quartz.
Summerfield (1983) defined four type of silcrete on the basis of micromorphology: grain-supported, matrix-supported (or floating), microcrystalline and conglomeratic. Grain-supported silcrete consists of sand-sized detrital quartz grains touching each other (Figures 1A-C); between the grains is either cement (syntaxial quartz overgrowths (Figure 1A) or, less commonly, fibrous quartz (Figure IB) or matrix (a mixture of quartz silt and microcrystalline silica replacing clay; Figure 1C). Grain-supported silcretes with syntaxial cement have a distinctive sugary texture in hand specimen. Floating (or matrix-supported) silcrete consists of >5% detrital quartz grains floating in a matrix (Figures 1D and I E). Microcrystalline (or matrix) silcrete consists of clay replaced by microcrystalline quartz, with varying proportions of scattered silt-sized detrital quartz grains (Figure IF). Conglomeratic silcrete contains pebbles >4 millimetres across. The grain-supported and matrix-supported categories have been further classified into fine-grained (detrital quartz grains dominantly <0.25 millimetres; Figures 1A, 1C and 1D) and medium-grained (detrital quartz grains dominantly 0.25-0.5 millimetres; Figures IB and 1E) (Doelman et al. 2001; Holdaway et al. 2004). Silcrete may contain coarse-sand-sized (>0.5 millimetres) and/or pebbly grains, but these varieties were less commonly used as artefact materials.
The microfabric/grain-size classification is often an accurate predictor of the flaking properties of the silcrete (Doelman et al. 2001; Webb & Domanski 2008) and is therefore very useful when determining the relationship between lithology and tool type.
SILCRETE IN AUSTRALIA
Silcrete is abundant in Australia and has often been referred to in previous literature as "grey billy"; it occurs across almost the entire continent (Langford-Smith 1978; Young 1985). Two distinct geographical associations have been differentiated: pedogenic, groundwater and (occasionally) pan/lacustrine silcretes throughout the inland arid regions (e.g. Doelman 2008; Wopfner 1978); and massive groundwater silcretes along the eastern coast, most commonly in association with basalts (Webb & Golding 1998).
The silcretes of the Central Queensland Highlands form part of the latter association, which occurs throughout the more humid parts of eastern Australia, from Tasmania through Victoria and NSW to north Queensland (e.g. Young 1985; note that Young's localities in Queensland west of the Great Dividing Range fall into the inland association). The eastern Australian silcretes generally show a close spatial relationship to basalts, outcropping directly beneath or at most a few hundred metres away, and are commonly known as "sub-basaltic" silcretes (Langford-Smith 1978). Many studies have suggested a genetic relationship between silcretes and basalts in eastern Australia (e.g. Gunn & Galloway 1978; Oilier 1991; Taylor & Smith 1975; Webb & Golding 1998), on the basis of the close spatial connection between the two and the substantial amount of introduced silica (greater than in the inland silcretes; Webb & Golding 1998).
The silica input that leads to the formation of the sub-basaltic silcretes was most likely supplied by basalt weathering; the alteration of the mafic minerals, volcanic glass and feldspars in basalt releases large amounts of dissolved silica (Eggleton et al. 1987). The source of the silica that formed the inland silcretes in Australia is less certain, but it may have been released by weathering of the volcaniclastic Early Cretaceous sediments in which many of these silcretes have formed (Tait 1998).
LITHIC RAW MATERIALS IN AUSTRALIAN ABORIGINAL ARCHAEOLOGY
Stone artefacts are the most durable and widespread form of evidence of the Aboriginal occupation of Australia, and stone artefact research in this country (e.g. Fullagar 1994; Hiscock 1998; Hiscock & Clarkson 2000; Holdaway & Stern 2004; Mulvaney 1977) has increasingly focused on the spatial relationship between stone artefacts and their sources, and the apparent preferential selection of some lithologies for particular types of artefacts.
Sourcing studies usually assess the distance that particular rock types were transported or traded (e.g. Corkill 2005; David et al. 1992; McBryde 1984; Mitchell 1994; Sheppard 1997; Smith 2006; Ulm et al. 2005). Distance from source may have influenced technological strategies, including more intensive core reduction of exotic materials (Attenbrow et al. 2008; Hiscock 1984), extending the use-life of stone implements through resharpening or reshaping (Byrne 1980; Clarkson 2002; McNiven 1993), manufacturing smaller implements (Attenbrow et al. 2008), establishing portable quarries (Webb 1993) and employing greater levels of residential mobility and tool curation (McNiven 1994). However, the simple expectation that raw material conservation strategies will be more apparent as distance from source increases (the "distance decay" theory) is not always supported by the evidence (e.g. Faulkner & Clarke 2009; White & McDonald 2010: 34-35), and so it is important that the specific nature of such relationships is demonstrated rather than assumed (Hiscock & Clarkson 2000: 102; Shiner 2008: 89).
Other technological and typological studies have sought to understand the apparent preferential selection of particular rock types. For example, the high selectivity of microcrystalline siliceous material (including silcrete) for tula adzes (e.g. Doelman 2008: 133; Gould 1978: 827-830; Smith 2006: 393-395) reflects the suitability of this lithology for adzing hard wood (Gould 1978: 827-829; Kamminga 1985: 17).
THE CENTRAL QUEENSLAND HIGHLANDS
The Consuelo and Buckland Tablelands form the most elevated central part of the Central Queensland Highlands (Figure 2), situated just south of the Tropic of Capricorn and 400-500 kilometres from the coast, and comprise adjacent low-relief plateaus at an elevation of around 1100 metres. They have been deeply dissected into cliffed gorges and steep-sided valleys by streams such as Carnarvon Creek, which has incised up to 600 metres below the tableland surface.
The geology of the area consists of horizontal, Lower Jurassic, interbedded sandstones and shales: the three conformable formations present are, in stratigraphic order, the Precipice Sandstone, the Evergreen Formation and the Hutton Sandstone (Exon 1968). These strata were uplifted and eroded in the Late Mesozoic--Early Tertiary to form the Central Queensland Highlands: the greater part of the erosional development of the landscape occurred at this time (Young & Wray 2000). In the mid-Tertiary, basalt lavas of the Buckland Volcanic Province erupted: on the Consuelo and Buckland Tablelands these form a large eroded shield volcano of middle to late Oligocene age (24-28 Ma: Sutherland 1985; Webb & McDougall 1967). Silcrete outcrops discontinuously through the Central Queensland Highlands (Gunn & Galloway 1978), often as layers up to 3 metres thick immediately beneath the basalt; these shed large blocks of silcrete downslope and into nearby creeks.
Within the gorges deeply incised into the tablelands, the lower slopes are cliffed in the aptly named Precipice Sandstone. Above the cliff top is a prominent bench cut into the more recessive Evergreen Formation and Hutton Sandstone. A second, higher cliff represents the overlying basalt lavas, above which lies the low-relief surface of the tablelands.
THE ARCHAEOLOGY OF THE CENTRAL QUEENSLAND HIGHLANDS
The archaeology of the Central Queensland Highlands has played a very important role in the development of our understanding of Australian prehistory. The most famous site of the region, Kenniff Cave, provided the first radiocarbon-dated evidence that human occupation of the continent extended into the Pleistocene (Mulvaney 1964). The three-phase cultural sequence observed by Mulvaney and Joyce (1965) at Kenniff Cave and The Tombs was subsequently verified by Beaton's (1977, 1991a,b) excavations of Rainbow Cave, Wanderer's Cave and Cathedral Cave, and Morwood's (1981) excavations of Native Well 1, Native Well 2, Turtle Rock and Ken's Cave, using not only stone artefacts, but also rock-art styles. While the evidence of vertical movement of stone artefacts through the sandy sediments of some of the sites suggests that we should be cautious about assigning precise dates to these cultural phases (Richardson 1992, 1996; Stern 1980), the existence of two relatively abrupt cultural transitions in the region seems to be beyond doubt.
The first two phases were seen as continent-wide technologies; they became known as the Australian Core Tool and Scraper Tradition (Bowler et al. 1970: 52) and the Australian Small Tool Tradition (Gould 1969: 234), respectively. Transition from one to the other was thought to occur around 4000 years ago (Mulvaney 1969). These cultural phases have provided a foundation for broad models of demographic, social and economic change in the middle to late Holocene (e.g. Bowdler 1981; Lourandos & Ross 1994; Lourandos 1997: 142-143, 300-321; Morwood 1987), although more recent studies (e.g. Hiscock 2003; Hiscock & Alien 2000; Hiscock & Attenbrow 1998; McNiven 2000; Shawcross 1998; Slack et al. 2004; Smith 2006) have raised questions about their significance as pan-continental concepts.
Morwood's (1984) description of the stone tool industries of the Central Queensland Highlands shows that the Core Tool and Scraper Industry (~19000-4300 BP) comprised a restricted range of flake scrapers and core tools of variable size, and only very occasional use of the grinding technique. Stone artefact densities were low during the early Holocene. The Small Tool Industry (~4300-2000 BP) saw an increase in occupational intensity, which was accompanied by blade technology and many new small implement types, including geometric microliths, Bondi points, backed points, pirri points, eloueras, burren and tula adze slugs and thumbnail scrapers. Edge-ground axes made their first appearance at this time and grindstones became much more common. Core tools and flake scrapers also continued to be made. The Recent Industry (<2000 BP) was characterised by an apparent reduction in deposition rates, the disappearance of backed blades, eloueras, pirri points and thumbnail scrapers, and the appearance of Juan knives and large scrapers of unconventional type.
SILCRETE QUARRIES IN THE CENTRAL QUEENSLAND HIGHLANDS
Silcrete quarries, using either silcrete outcrop or displaced blocks and scree, lie near or immediately adjacent to several excavated occupation sites (Kenniff Cave, Turtle Rock and Native Well: Morwood 1981; Mulvaney & Joyce 1965; Figure 2). A previously unrecorded quarry at Ogres Thumb was discovered during the present study; it represents an excellent example of the type of quarry found in the region.
The Ogres Thumb quarry
The quarry lies at the south-eastern edge of the Consuelo Tableland (Figures 1 and 2), on the boundary between the Tertiary basalt and the Boxvale Sandstone Member of the Evergreen Formation (Exon 1968). The 120 metre thick Hutton Sandstone, which overlies the Evergreen Formation, is absent at the quarry site, though it occurs to the north-west, along the side of the gorge (Figure 3). The quarry site most likely represents a drainage line in the pre-basalt topography, where there has been ~120 metres of erosion to remove the Hutton Sandstone; pre-basaltic erosion was extensive throughout the area (Young & Wray 2000). The silcrete layer is underlain by a weathering profile developed in the Boxvale Sandstone Member, consisting of a mottled red layer over 2 metres thick that grades down into soft sandstone.
The quarry is located on the crest of the ridge that forms the drainage divide between streams flowing south into Carnarvon Creek and north into Arch Creek (Figure 3). The steeply descending ridge crest here flattens out for 250 metres and forms a bench approximately 50 metres wide, covered in open eucalypt forest. The central area of this bench is covered with a thin (2-8 metres) remnant of basalt. Along the north and south sides, the underlying silcrete forms a gently sloping (~5[degrees]) platform about 15 metres wide, terminating in a small breakaway (cliff) approximately 1 metre high. This is the total thickness of the silcrete. Below the silcrete cliff, the sides of the ridge slope away at approximately 25[degrees] and are littered with large blocks of silcrete.
There is evidence of extensive quarrying at this site. The silcrete bench on both north and south sides of the ridge is covered with flakes at densities of up to 500 per square metre. Fractured conchoidal surfaces are common on the silcrete outcrop and on displaced silcrete blocks downslope from the outcrop. Flakes and corestones of silcrete can also be found lying on the basalt upslope of the outcrop. On the eastern tip of the ridge, there has been a lot of digging into the slope, although it is not clear why this would have been necessary, given the good outcrop all round the ridge.
The silcrete is pale buff with a glassy conchoidal fracture. It is a fine-grained grain-supported silcrete, composed of moderately well-sorted fine-quartz grains (average 0.2 millimetres; Figure 4A) with very minor zircon, rutile and tourmaline. The quartz grains are cemented by syntaxial quartz overgrowths (rims in crystallographic continuity with the quartz crystals of the grains; Figure 4A). In hand specimen, it has the distinctive sugary appearance typical of this type of silcrete.
The silcrete has formed by silicification of a clean, well-washed sand deposit; this is quite different to the weathering profile developed on the underlying Boxvale Sandstone Member, which contains an extensively ferruginised opal-clay matrix with cutans and rootlets. Thus the silcrete has not formed by silicification of the Boxvale Sandstone Member or the soil developed on it. Instead, it represents a silicified clean quartz sand, probably deposited by the Tertiary stream on the pre-basalt surface that eroded away the Hutton Sandstone in this area.
Other silcrete quarries
The silcrete at the other known quarry sites in the Central Queensland Highlands (Kenniff Cave, Turtle Rock and Native Well; Figure 2) is all located immediately beneath the basalt capping the tableland, and was quarried either in situ or from displaced blocks and scree (Morwood 1981; Mulvaney & Joyce 1965). This reinforces the strong spatial link between silcretes and basalts that is evident elsewhere in eastern Australia.
The silcrete at Kenniff Cave (identified as quartzite by Mulvaney & Joyce 1965) is very similar to that at Ogres Thumb; it has the distinctive sugary appearance in hand specimen, and is fine-grained and grain-supported, consisting of well-sorted quartz clasts, cemented by syntaxial overgrowths (Figure 4B). This silcrete, like that at Ogres Thumb, may have formed, in well-washed quartz sand in stream channels beneath the basalt, which perhaps explains why it is found only in discrete sites (former Tertiary stream courses).
At Turtle Rock (Table 1), examination of the artefacts excavated by Morwood (1981) showed that they are predominantly (>75%) made of fine-grained grain-supported silcrete almost identical to that at Ogres Thumb and Kenniff Cave (this material was identified as fine silcrete by Morwood 1981). In addition, ~15% of artefacts are composed of coarse-grained matrix-supported silcrete (quartzite of Morwood 1981), in which coarse-sand-sized clear quartz grains are separated by a grey microcrystalline matrix. There is also a small proportion (<1%) of microcrystalline (matrix) silcrete (the duricrust silcrete of Morwood 1981).
The material excavated by Morwood (1981) at the Native Well site was not available for re-examination during this study, but using the reinterpretation of Morwood's lithological identifications at Turtle Rock (see above; Table 1), fine-grained grain-supported silcrete makes up a much smaller proportion of the assemblage (~25%), with ~15% each of coarse-grained matrix-supported and microcrystalline silcrete, and a much larger proportion of non-silcrete material.
The geological description of the Native Well quarry site by Morwood (1981) indicates that the silcrete is underlain by coarse-grained friable sandstone. It is likely that the different types of silcrete here have formed by silicification of different parent materials. Silicification of a well-washed Tertiary stream sand and the underlying Lower Jurassic sandstone probably formed the fine-grained grain-supported and coarse-grained matrix-supported silcrete, respectively. A mudstone was silicified to form the microcrystalline silcrete: this was probably an interbed within the Lower Jurassic sandstone.
The geological distribution of basalt and its close association with the formation of silcrete allows us to identify the placement of Aboriginal stone artefact quarries in the Central Queensland Highlands. It is predicted that silcrete quarries will be restricted to cliffs where the base of the basalt is exposed, limiting the potential area of quarrying to within and around the Consuelo and Buckland Tablelands (Figure 2). The low relief of the large area of basalt to the north of the tablelands (Figure 2) means that the base of the basalt in this area is generally not well exposed, so even if silcrete has developed beneath the basalt here, it will not be readily accessible. Transported silcrete was used as an artefact lithology outside the Consuelo and Buckland Tablelands; silcrete boulders and cobbles with a waterworn cortex were utilised in the Arcadia Valley to the east (Cochrane et al. 2012).
SILCRETE ARTEFACTS AT OCCUPATION SITES IN THE CENTRAL QUEENSLAND HIGHLANDS
Excavated occupation sites that are not associated with quarries are scattered through the Central Queensland Highlands (Figure 2), and information from these can be used to identify the distribution and use of the different types of silcrete.
Stone artefacts from an excavation on Arch Creek, <5 kilometres north of the Ogres Thumb quarry, are predominantly fine-grained grain-supported silcrete identical to that at the quarry (based on examination of material held in the Queensland Museum).
Examination of the material excavated from Rainbow Cave by Beaton (1991a) showed that it is dominated by fine-grained grain-supported silcrete (~50%; Table 1), but there is also a proportion of matrix-supported silcrete with clear quartz grains in a dull grey microcrystalline matrix. It is notable that the fine-grained silcrete artefacts have a smaller mean long axis (1.7 centimetres) than artefacts of matrix-supported silcrete (2.7 centimetres). Rainbow Cave is located ~12 kilometres east of the limit of the basalt outcrop (Figure 2), so there is no immediately adjacent source of silcrete. The component of matrix-supported silcrete indicates that the silcrete source was probably not the Ogres Thumb quarry, which lies about 45 kilometres to the north-west (Figure 2); instead, there is probably a source along the basalt cliffs closer to the cave.
The Tombs (Figure 2) is an occupation site adjacent to a cliffed residual of Precipice Sandstone: artefacts are dominated by fine-grained grain-supported silcrete (identified as quartzite by Mulvaney & Joyce 1965; this material was not re-examined during the present study). There are also substantial numbers of quartz, jasper and chert artefacts, derived from pebble bands in the sandstone cliffs at the site (Mulvaney & Joyce 1965). The nearest basalt outcrop to The Tombs is ~5 kilometres to the north-west (Figure 2), so any silcrete outcrop would have been at least this far away, and the nearby creek has a sand bed, so there is no local source of silcrete cobbles. At The Tombs, silcrete was clearly preferred for artefact manufacture over the more locally available chert, quartz and jasper pebbles.
Ken's Cave is located over 170 kilometres west-north-west of the other sites (Figure 2), and examination of the artefacts excavated by Morwood (1981) showed that they are composed mostly of quartz and chert (Table 1), with minor amounts of fine-grained grain-supported silcrete (4%) and matrix-supported silcrete (6%). Some of the tools are made of silcrete (e.g. the blade illustrated as figure 8E in Morwood 1981). Ken's Cave is ~80 kilometres north-west of the nearest outcrop of basalt and a similar distance from the nearest creek with basalt in its catchment (Figure 2), so the silcrete at the site must have been transported at least this distance.
The artefacts from Wanderer's Cave and Cathedral Cave, which were excavated by Beaton (1991a,b), were not available for examination, and Beaton (1977: 39) eschewed raw material classification, so there is no published data on the rock types present and the proportion of silcrete is not known. However, the stone materials at Wanderers Cave were described as "reasonably fine-grained siliceous rocks like those of Rainbow Cave" (Beaton 1991a: 24), so it is likely that these two caves, which are only 6 kilometres apart, have the same stone assemblages.
Not surprisingly, the number of artefacts recovered from sites that were associated with quarries tended to be higher than those that were not (Table 2). Nevertheless, there is still a rigorous pattern, which shows that across the Central Queensland Highlands, the occupation sites with silcrete as the main raw material are at or close to silcrete outcrops, where cliffs expose the base of the basalt. There was some expedient use of local non-silcrete lithologies. At sites that were some distance from silcrete outcrops (e.g. Ken's Cave, at least 80 kilometres), materials such as quartz and chert are predominant, but silcrete was nevertheless preferred for some tool types (see below), indicating that it was sufficiently highly regarded as an artefact lithology to justify wide dispersal.
THE RELATIONSHIP BETWEEN SILCRETE TYPE AND ARTEFACT MANUFACTURE
There is a strong correlation between silcrete microstructure, its mechanical strength and the type of tools made from it (Webb & Domanski 2008). Microcrystalline silcrete with few impurities has a high compressive strength, comparable to that of obsidian and flint (Domanski et al. 1994), and is therefore suitable for systematic production of microblades and other fine-profiled tool forms. For example, tools at Tibooburra in western NSW are dominated by microcrystalline silcrete (called amorphous silcrete by Doelman 2008), and this was almost the only lithology used for the manufacture of tula adzes.
The fine-grained grain-supported silcrete at Kenniff Cave and Ogres Thumb quarry is a high-quality material: its low fracture toughness means that it is relatively easy to flake by percussion knapping techniques (Webb & Domanski 2008). However, it has only moderate compressive strength, so it is slightly lower-quality than microcrystalline silcrete. Testing of a hafted knife made from fine-grained grain-supported silcrete from Arcadia Valley (to the east of the Central Queensland Highlands) showed that it is very tough, with no sign of failure and only some edge wear after carving hard wood for more than 16 hours.
The physical properties of the fine-grained grain-supported silcrete in the Central Queensland Highlands have dictated how it was worked, used and transported over time. At Kenniff Cave, it was the predominant material used throughout both the early Australian Core Tool and Scraper Tradition and the following Australian Small Tool Tradition (Morwood 1981, 1984; Mulvaney & Joyce 1965). This type of silcrete had sufficient versatility for the manufacture of the whole range of artefacts of both assemblages, including the fine pressure and delicate percussion retouching required for the mid-Holocene backed artefacts and points, so the technological changes did not require the use of a different raw material type (Webb & Domanski 2008).
At Native Well, two different types of silcrete were utilised: fine-grained grain-supported silcrete and coarse-grained matrix-supported silcrete. The mid-Holocene appearance of the small tool industry at this site was accompanied by an increase in the use of the former and a decrease in the use of the latter (Morwood 1981), indicating that the fine-grained silcrete is a higher-quality material. The coarse grains in the matrix-supported silcrete represent microstructural defects and are likely to lead to a high frequency of step fracture terminations, offsetting the probable increase in compressive strength due to the presence of a microcrystalline matrix (Webb & Domanski 2008). Manufacture of the mid-Holocene backed artefacts and points, associated with fine pressure and delicate percussion retouching (Flenniken & White 1985; Mulvaney & Kamminga 1999), was facilitated by the use of fine-grained grain-supported silcrete. However, the mid-Holocene tula and burren adzes at Native Well were made exclusively of chert rather than fine-grained silcrete (Morwood 1981), probably because the silcrete was more prone to edge fracturing, due to its relatively low fracture toughness (Webb & Domanski 2008).
A newly discovered silcrete quarry at Ogres Thumb on the Consuelo Tableland in the Central Queensland Highlands is in fine-grained grain-supported silcrete that formed beneath the mid-Tertiary basalt capping the tableland. Other silcrete outcrops in these highlands also occur beneath the basalt, verifying the strong link between silcretes and basalts evident throughout eastern Australia. From the geological distribution of basalt and its close association with silcrete, it is possible to predict that silcrete quarries are largely restricted to cliffs where the base of the basalt is exposed, limiting the potential area of quarrying to within and around the Consuelo and Buckland Tablelands. In addition, transported silcrete boulders and cobbles have been utilised as a silcrete source where outcrops are not available. There is a relatively low likelihood of silcrete quarries within the large, low-relief area of basalt to the north of the highlands, because the base of the basalt here is generally not well exposed.
Silcrete artefacts are abundant throughout the region, demonstrating the importance of silcrete as a raw material source. Fine-grained grain-supported silcrete is the dominant artefact lithology around the Consuelo and Buckland Tablelands because of its excellent flaking quality, making it amenable to the manufacture of flake scrapers as well as retouched microlithic tools. At Kenniff Cave, this type of silcrete was the predominant material used throughout the Holocene, because it had sufficient versatility to make the whole range of artefacts present. In contrast, at Native Well, where both fine-grained grain-supported and coarse-grained matrix-supported silcrete were utilised, the mid-Holocene appearance of backed artefacts and points was accompanied by an increase in the use of grain-supported silcrete and a decrease in the use of matrix-supported silcrete, because the latter was less suited to the manufacture of the more curated tools. This was a simple response to the technological requirements of artefact manufacture and the functional requirements of the new tool types.
With increasing distance from the silcrete quarries in the Central Queensland Highlands, there is a decrease in the percentage of silcrete artefacts and an accompanying increase in the expedient use of local lithologies such as quartz and chert. However, even at locations at least 80 kilometres from the nearest silcrete outcrop, silcrete is still present in the artefact assemblages and was preferred for some tool types, indicating that the flaking quality of the silcrete justified its dispersal over considerable distances.
This study highlights the importance of understanding the geological context of a region to gain further insight into the procurement, use, transport and distribution of artefacts; and, in particular, how people made decisions, in an unchanging geological context, to select a particular silcrete type for particular tools.
This study was carried out under Cultural Records Permit No SW/96/01/RES(A), issued by the Queensland Minister for the Environment. We wish to thank Richard Robins of the Queensland Museum for arranging access to stone artefact collections from the excavations at Turtle Rock, Rainbow Cave and Ken's Cave, and Chandra Jayasuriya who drew the maps in Figure 3.
Attenbrow, V., Doelman, T. and Corkill, T. 2008. Organizing the manufacture of Bondi points at Balmoral Beach, Middle Harbour, Sydney, NSW, Australia. Archaeology in Oceania 43: 104-119.
Beaton, J.M. 1977. Dangerous Harvest: Investigations in the Late-prehistoric Occupation of Upland South-East Central Queensland. Unpublished PhD thesis, Australian National University, Canberra.
Beaton, J.M. 1991a. Excavations at Rainbow Cave and Wanderer's Cave: Two rockshelters in the Carnarvon Range, Queensland. Queensland Archaeological Research 8: 3-32.
Beaton, J.M. 1991b. Cathedral Cave: A rockshelter in Carnarvon Gorge, Queensland. Queensland Archaeological Research 8: 33-84.
Bowdler, S. 1981. Hunters in the highlands: Aboriginal adaptations in the Eastern Australian Uplands. Archaeology in Oceania 16:99-111.
Bowler, J.M., Jones, R., Allen, H. and Thome, A.G. 1970. Pleistocene human remains from Australia: A living site and human cremation from Lake Mungo, western New South Wales. World Archaeology 2: 39-60.
Byrne, D. 1980. Dynamics of dispersion: The place of silcrete in archaeological assemblages from the Lower Murchison, Western Australia. Archaeology and Physical Anthropology in Oceania 15: 110-119.
Clarkson, C. 2002. Holocene scraper reduction, technological organization and landuse at Ingaladdi Rockshelter, Northern Australia. Archaeology in Oceania 37: 79-86.
Cochrane, G.W.G., Habgood, P.J., Doelman, T., Herries, A.I.R. and Webb, J. 2012. A progress report on research into the stone artefacts of the southern Arcadia Valley, central Queensland. Australian Archaeology 75: 98-103.
Corkill, T. 2005. Sourcing stone from the Sydney region: A hatchet job. Australian Archaeology 60:41-50.
David, B., Bird, R., Fullagar, R. and Little, L. 1992. Glassy obsidian artefacts from north Queensland: The Nolan's Creek source and some archaeological occurrences. The Artefact 15: 25-30.
Doelman, T. 2008. Time to Quarry: The Arehaeology of Stone Procurement in Northwestern New South Wales, Australia. BAR International Series, 1801. Archaeopress, Oxford.
Doelman, T., Webb, J. and Domanski, M. 2001. Source to discard: Patterns of lithic raw material procurement and use in Sturt National park, northwestern New South Wales. Archaeology in Oceania 36: 15-33.
Domanski, M., Webb, J.A. and Boland, J. 1994. Mechanical properties of stone artefact materials and the effect of heat treatment. Archaeometry 36: 177-208.
Eggleton, R.A., Foudoulis, C. and Varkevisser, D. 1987. Weathering of basalt: changes in rock chemistry and mineralogy. Clays and Clay Minerals 35: 161-169.
Exon, N.E 1968. Eddystone, Queensland, 1:250,000 Geological Series--Explanatory Notes. Bureau of Mineral Resources, Geology and Geophysics, Canberra.
Faulkner, E and Clarke, A. 2009. Artefact assemblage characteristics and distribution on the Point Blane Peninsula, Blue Mud Bay, Arnhem Land. Australian Archaeology 69: 21-28.
Flenniken, J.J. and White, J.E 1985. Australian flaked stone tools: A technological perspective. Records of the Australian Museum 36: 131-151.
Fullagar, R. 1994. Traces of times past: Stone artefacts into prehistory. Australian Archaeology 39: 63-72.
Gould, R.A. 1969. Puntutjarpa rockshelter. Archaeology and Physical Anthropology in Oceania 4: 229-237.
Gould, R.A. 1978. The anthropology of human residues. American Anthropologist 80:815-835.
Gunn, R.H. and Galloway, R.W. 1978. Silcretes in south-central Queensland. In T. Langford-Smith (ed.), Silcrete in Australia, pp. 51-71. Department of Geography, University of New England, Armidale.
Hiscock, E 1984. Raw material rationing as an explanation of assemblage differences: A case study of Lawn Hill, northwestern Queensland. In G. Ward (ed.), Archaeology at ANZAAS, pp. 178-190. Canberra Archaeological Society, Canberra.
Hiscock, E 1993. Bondaian technology in the Hunter Valley, New South Wales. Archaeology in Oceania 28: 65-76.
Hiscock, P. 1998. Revitalising artefact analysis. In T. Murray (ed.), Archaeology of Aboriginal Australia, pp. 257-265. Allen and Unwin, Sydney.
Hiscock, P. 2003. Quantitative exploration of size variation and the extent of reduction in Sydney Basin assemblages: A tale from the Henry Lawson Drive Rockshelter. Australian Archaeology 57: 64-74.
Hiscock, E and Allen, H. 2000. Assemblage variability in the Willandra Lakes. Archaeology in Oceania 35: 97-103.
Hiscock, E and Attenbrow, V. 1998. Early Holocene backed artefacts from Australia. Archaeology in Oceania 33: 49-62.
Hiscock, E and Clarkson, C. 2000. Analysing Australian stone artefacts: An agenda for the twenty first century. Australian Archaeology 50: 98-108.
Holdaway, S., Shiner, J. and Fanning, E 2004. Hunter-gatherers and the archaeology of discard behaviour: An analysis of surface stone artefacts from Sturt National Park, western New South Wales, Australia. Asian Perspectives 43: 34-72.
Holdaway, S.D. and Stern, N. 2004. A Record in Stone. The Study of Australia's Flaked Stone Artefacts. Aboriginal Studies Press, Canberra.
Hughes, P.J., Sullivan, M.E. and Lampert, R.J. 1973. The use of silcrete in southern coastal N.S.W. Archaeology and Physical Archaeology in Oceania 8: 220-225.
Kamminga, J. 1985. The pirri graver. Australian Aboriginal Studies 198512: 2-25.
Lamplugh, G.W. 1902. Calcrete. Geological Magazine 9: 75.
Langford-Smith, T. 1978. A select review of silcrete research in Australia. In T. Langford-Smith (ed.), Silcrete in Australia, pp. 1-11. Department of Geography, University of New England, Armidale.
Lourandos, H. 1997. Continent of Hunter-Gatherers: New Perspectives in Australian Prehistory. Cambridge University Press, Cambridge, UK.
Lourandos, H. and Ross, A. 1994. The great "intensification debate": Its history and place in Australian archaeology. Australian Archaeology 39: 54-63.
McBryde, I. 1984. Kulin greenstone quarries: The social contexts of production and distribution for the Mt William site. World Archaeology 16: 267-285.
McNiven, I. 1993. Raw material proximity and bevel-edged tool use, Teewah Beech, Southeast Queensland. Archaeology in Oceania 28: 138-143.
McNiven, I.J. 1994. Technological organization and settlement in southwest Tasmania after the glacial maximum. Antiquity 68: 75-82.
McNiven, I.J. 2000. Backed to the Pleistocene. Archaeology in Oceania 35: 48-52.
Milnes, A.R. and Thiry, M. 1992. Silcretes. In I.E Martini and W. Chesworth (eds), Developments in Earth Surface Processes 2, Weathering, Soils and Palaeosols, pp. 349-377. Elsevier, Amsterdam.
Mitchell, S. 1994. Stone exchange networks in northwestern Arnhem Land: Evidence for recent chronological change. In M. Sullivan, S. Brockwell and A. Webb (eds), Archaeology in the North, pp. 188-200. Australian National University, Darwin.
Morwood, M.J. 1981. Archaeology of the central highlands: The stone component. Archaeology in Oceania 16: 1-52.
Morwood, M.J. 1984. The prehistory of the Central Queensland Highlands. In E Wendorf and A. Close (eds), Advances in World Archaeology, pp. 325-380. Academic Press, New York.
Morwood, M.J. 1987. The archaeology of social complexity in south-east Queensland. Proceedings of the Prehistoric Society 53: 337-350.
Mulvaney, D.J. 1964. The Pleistocene colonization of Australia. Antiquity 38: 263-267.
Mulvaney, D.J. 1969. The Prehistoo, of Australia. Thames & Hudson, London.
Mulvaney, D.J. 1977. Classification and typology in Australia: The first 340 years. In R.V.S. Wright (ed.), Stone Tools as Cultural Markers: Change, Evolution, Complexity, pp. 263-268. Australian Institute of Aboriginal Studies, Canberra.
Mulvaney, D.J. and Joyce, E.B. 1965. Archaeological and geomorphological investigations on Mt Moffatt Station, Queensland, Australia. Proceedings of the Prehistory Society New Series 31: 147-212.
Mulvaney, D.J. and Kamminga, J. 1999. Prehistory of Australia. Allen and Unwin, Sydney.
Nash, D.J. and Ullyott, J.S. 2007. Silcrete. In D.J. Nash and S.J. McLaren (eds), Geochemical Sediments and Landscapes, pp. 95-143. Blackwell, Oxford.
Ollier, C. 1991. Aspects of silcrete formation in Australia. Zeitschrift fur Geomorphologie 35:151-163.
Richardson, N. 1992. Conjoin sets and stratigraphic integrity in a sandstone shelter: Kenniff Cave (Queensland, Australia). Antiquity 66:408-418.
Richardson, N. 1996. Seeing is believing: A graphical illustration of the vertical and horizontal distribution of conjoined artefacts using DesignCAD 3D. Australian Archaeology 95: 81-95.
Robins, R.P. 1997. Patterns in the landscape: A case study in nonsite archaeology from southwest Queensland. Memoirs of the Queensland Museum Cultural Heritage Series 1 : 23-56.
Rowney, M. and White, J.P. 1997. Detecting heat treatment on silcretes: Some experiments with method. Journal of Archaeological Science 24: 649-657.
Shawcross, W. 1998. Archaeological excavations at Mungo. Archaeology in Oceania 33: 183-200.
Sheppard, P.J. 1997. Characterisation of cherts from sites in southwest Tasmania. Archaeology in Oceania 32: 47-53.
Shiner, J. 2008. The intensity of raw material utilisation as an indication of occupational history in surface stone artefact assemblages from the Strathbogie Ranges, central Victoria. Australian Aboriginal Studies 2008/2: 80-92.
Slack, M., Fullagar, R., Field, J. and Border, A. 2004. New Pleistocene ages for backed artefact technology in Australia. Archaeology in Oceania 39: 131-137.
Smith, M.A. 2006. Characterizing Late Pleistocene and Holocene stone artefact assemblages from Puritjarra rock shelter: A long sequence from the Australian desert. Records of the Australian Museum 58: 371-410.
Stern, N. 1980. Taphonomy: Some Observations about its Place in Archaeology. Unpublished BA (Hons) thesis, University of Sydney, Sydney.
Sullivan, M.E. and Simmons, S. 1979. Silcrete: A classification for flaked stone artifact assemblages. The Artifact 4:51-60.
Summerfield, M.A. 1983. Petrography and diagenesis of silcrete from the Kalahari Basin and Cape coastal zone, southern Africa. Journal of Sedimentary. Petrology 53: 895-909.
Sutherland, P.L. 1985. Regional controls in eastern Australian volcanism. In P.L. Sutherland, B.J. Franklin and A.E. Waitho (eds), Volcanism in Eastern Australia, pp. 13-32. Special Publication 1. Geological Society of Australia, Sydney, NSW.
Tait, M. 1998. Geology and Landscape Evolution of the Mt Wood Hills Area, near Tibooburra, Northwestern New South Wales. Unpublished honours thesis, La Trobe University, Melbourne.
Taylor, G. and Smith, I.E. 1975. The genesis of sub-basaltic silcretes from the Monaro, New South Wales. Journal of the Geological Society of Australia 22: 377-385.
Tykot, R.H. 2003. Determining the source of lithic artifacts and reconstructing trade in the ancient world. In EN. Kardulias and R.W. Yerkes (eds), Written in Stone: The Multiple Dimensions of Lithic Analysis, pp. 59-85. Lexington Books, Lanham, MD.
Ullyott, J.S., Nash, D.J. and Shaw, P.A. 1998. Recent advances in silcrete research and their implications for the origin and palaeoenvironmental significance of sarsens. Proceedings of the Geologists' Association 109: 255-270.
Ulm, S., Cotter, S., Cotter, M., Lilley, I., Clarkson, C. and Reid, J. 2005. Edge-ground hatchets on the southern Curtis Coast, central Queensland. In I. Macfarlane (ed.), Many Exchanges: Archaeology, History, Community and the Work of Isabel McBryde, pp. 323-342. Aboriginal History Monograph 11. Aboriginal History Inc., Canberra.
Webb, A.W. and McDougall, I. 1967. A comparison of mineral and whole rock potassium-argon ages of Tertiary volcanics from central Queensland, Australia. Earth and Planetary Science Letters 3: 41-47.
Webb, C. 1993. The lithification of a sandy environment. Archaeology in Oceania 28:105-111.
Webb, J.A. and Domanski, M. 2008. The relationship between lithology, flaking properties and artefact manufacture for Australian silcretes. Archaeometry 50: 555-575.
Webb, J.A. and Golding, S.D. 1998. Geochemical mass-balance and oxygen-isotope constraints on silcrete formation and its paleoclimatic implications in southern Australia. Journal of Sedimentary Research 68:981-993.
White, B. and McDonald, J. 2010. Lithic artefact distribution in the Rouse Hill Development Area, Cumberland Plain, New South Wales. Australian Archaeology 70: 29-38.
White, J.P. and O'Connell, J.E 1982. A Prehistory of Australia, New Guinea and Sahul. Academic Press, Sydney.
Wopfner, H. 1978. Silcrete of northern South Australia and adjacent regions. In T. Langford-Smith (ed.), Silcrete in Australia, pp. 93-141. Department of Geography, University of New England, Armidale.
Young, R.W. 1985. Silcrete distribution in eastern Australia. Zeitschrift fur Geomorphologie Neue Folge 29: 21-36.
Young, R.W. and Wray, R.A.L. 2000. Contribution to the theory of scarpland development from observations in central Queensland, Australia. Journal of Geology 108:705-719.
JOHN WEBB, BRIAN FINLAYSON, GRANT COCHRANE, TRUDY DOELMAN and MARIAN DOMANSKI
JW: Environmental Geoscience, Department of Agricultural Sciences, La Trobe University; BF: Department of Resource Management and Geography, The University of Melbourne; GC, TD: Department of Archaeology, University of Sydney; MD: Environmental Geoscience, Department of Agricultural Sciences, La Trobe University.
Correspondence: John Webb, Environmental Geoscience, Department of Agricultural Sciences, La Trobe University, Victoria 3086, Australia. Email: firstname.lastname@example.org
Table 1. Percentages of stone artefact lithologies from excavation sites in the Central Queensland Highlands; identification of silcrete types at Native Well based on comparisons with Turtle Rock identifications (see text for details). Grain- Matrix- supported supported silerete silerete Rainbow Cave ([dagger]) No. 250 72 % 43.6 12.6 Kens Cavet ([dagger]) No. 11 16 % 3.9 5.6 Turtle Rock ([dagger]) No. 3873 770 % 77.3 15.4 Native Well ([double dagger]) No. 197 100 26.5 13.4 Matrix Quartz Chert silerete Rainbow Cave ([dagger]) No. 0 71 71 % 0 12.4 12.4 Kens Cavet ([dagger]) No. 40 97 88 % 14 34 30.9 Turtle Rock ([dagger]) No. 45 14 271 % 0.9 0.3 5.4 Native Well ([double dagger]) No. 127 1 196 17.1 0.1 26.3 Sandstone Basalt Shale Rainbow Cave ([dagger]) No. 27 21 49 % 4.7 3.7 8.6 Kens Cavet ([dagger]) No. 1 0 0 % 0.4 0 0 Turtle Rock ([dagger]) No. 2 5 0 % 0 0.1 0 Native Well ([double dagger]) No. 115 6 0 15.5 0.8 0 Petrified wood Rainbow Cave ([dagger]) No. 3 % 0.5 Kens Cavet ([dagger]) No. 7 % 2.5 Turtle Rock ([dagger]) No. 0 % 0 Native Well ([double dagger]) No. 0 0 ([dagger]) Counts from boxed collections in the Queensland Museum. ([double dagger]) Counts from Morwood (1981). Table 2. The number of artefacts recovered from the Central Queensland Highlands sites discussed in the text. Note that the artefact numbers for Native Well are much greater than in Table 1, which refers only to the retouched artefacts from this site; Morwood (1981) did not provide raw material identifications for a artefacts. No. of artefacts Sites associated with quarries Kenniff Cave >23000 Turtle Rock 5013 Native Well 1 9948 Native Well 2 3849 Sites not associated with quarries Arch Creek Unknown Rainbow Cave 573 The Tombs 2324 Ken's Cave 285 Cathedral Cave 5848 Wanderer's Cave 297
|Printer friendly Cite/link Email Feedback|
|Author:||Webb, John; Finlayson, Brian; Cochrane, Grant; Doelman, Trudy; Domanski, Marian|
|Publication:||Archaeology in Oceania|
|Date:||Dec 1, 2013|
|Previous Article:||Prehistory in a nutshell: a Lapita-age nut-cracking stone from the Arawe Islands, Papua New Guinea.|
|Next Article:||Distribution and extirpation of pigs in Pacific Islands: a case study from Palau.|