SEARCH

SEARCH BY CITATION

Keywords:

  • Darumbal;
  • Shoalwater Bay islands;
  • marine subsistence specialisation;
  • late Holocene change

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Shoalwater Bay Islands
  5. Otterbourne Island
  6. High Peak Island
  7. Collins Island
  8. Discussion
  9. Conclusion
  10. Acknowledgements
  11. References
  12. Supporting Information

Island archipelagos of the tropical coast of central Queensland include the most distant offshore islands used by Aboriginal Australians. Excavations on Collins, Otterbourne and High Peak Islands, located up to 40 km from the mainland, reveal evidence of offshore voyaging and marine specialisation in the Shoalwater Bay region for at least 5200 years. A time lag of up to 3000 years between island formation and systematic island use may reflect delayed development of key marine resources. Expansion of island use commencing around 3000–3500 years ago is linked to population increases sustained by synchronous increases in marine resources. Occupational hiatuses variously between 1000 and 3000 years ago are associated with increased ENSO activity. Intensified island use within the past 1000 years is primarily a social phenomenon associated with continuing demographic pressures and the development of more coastally and marine-focused mainland groups, with settlement patterns increasingly encompassing adjacent islands. The viability of risky offshore canoe voyaging was underwritten by two key high-return subsistence pursuits – hunting green turtles and collecting turtle eggs. In addition to subsistence and quartz quarrying, a key motivation for island visitation may have been socially restricted (e.g. ceremonial) practices.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Shoalwater Bay Islands
  5. Otterbourne Island
  6. High Peak Island
  7. Collins Island
  8. Discussion
  9. Conclusion
  10. Acknowledgements
  11. References
  12. Supporting Information

The long-term history of Aboriginal use of Australia's diverse offshore islands remains an important archaeological question that continues to generate debate (Hiscock 2008: chapter 9; Mulvaney & Kamminga 1999: chapter 18). In an important overview, Bowdler (1995: 956) concluded that there exists a “lack of obvious patterning” in Aboriginal use of islands beyond most island use taking place within the past 3000 years, with intensified use occurring within the past 1000 years. Recently, Sim and Wallis (2008) pointed out that little or no direct evidence exists for Aboriginal use of small offshore islands during the “initial post-insulation (island) phase” between approximately 6500 and 4500 years ago. This view challenges Barker's (2004) hypothesis for 9000 years of continuous occupation of the Whitsunday Islands on the central Queensland coast. Barker's (2004) social model for increased use of the Whitsunday Islands over the past 3000 years is also challenged by Sim and Wallis's (2008: 104) suggestion that similar increases in island use across northern Australia during the late Holocene were “a direct human response to weather regimes becoming more conducive to coastal habitation and watercraft travel”. We test the broader applicability of these divergent views with excavation results from islands off Shoalwater Bay, central Queensland. Using recent evidence for expansion of coral reefs across the southern Great Barrier Reef region within the past 3500 years, our research also tests Rowland's (1996) longstanding claim that sea-level change and changing configurations of marine resources influenced patterns of island use along the central Queensland coast. The Shoalwater Bay archipelago was formed between 12000 and 8000 years ago, when a series of hilltops across the continental shelf were flooded by rising seas associated with global warming at the end of the last Ice Age (Pleistocene). As such, the region provides an excellent opportunity to test the hypotheses of Barker, Rowland, Sim and Wallis, and to help answer two fundamental questions asked by the Darumbal – contemporary Traditional Owners of the region, with whom we have worked closely: What is the history of Aboriginal use of these islands and did it change through time?

Shoalwater Bay Islands

  1. Top of page
  2. Abstract
  3. Introduction
  4. Shoalwater Bay Islands
  5. Otterbourne Island
  6. High Peak Island
  7. Collins Island
  8. Discussion
  9. Conclusion
  10. Acknowledgements
  11. References
  12. Supporting Information

The waters and associated islands of the Shoalwater Bay region are located within the southern section of the Great Barrier Reef Marine Park World Heritage Area (Figure 1). The Shoalwater Bay mainland and inshore waters and islands also form the Shoalwater Bay Military Training Area, which was established in 1965. The region has a subtropical climate, with a mean annual rainfall of 1700 mm (mostly falling in summer) and a mean monthly temperature range of 32°C (January) to 10.5°C (July) (ABARE et al. 1993: 19–20; Catling et al. 1994: 120; Trnski et al. 1994: 246). The Shoalwater Bay region is aptly named, with winds producing 3 m + waves in open waters with little effort. Tidal ranges in the region are the highest for the Australian east coast at up to 7 m and tidal currents are strong, at up to 6 knots (ABARE et al. 1993: 174).

figure

Figure 1. The Shoalwater Bay study area.

Download figure to PowerPoint

The islands of Shoalwater Bay comprise the southern sections of the Northumberland Group (along with the Marble and Percy Islands) and can be divided into two subgroups: first, an inner subgroup, located within 25 km of the mainland and taking in Townshend Island and islands of the Cannibal Group (including Collins Island); and, second, an outer subgroup located 25–40 km off the mainland and extending from Otterbourne Island in the west to High Peak Island in the east (Figure 1). A 20 km wide gap of open sea separates the inner and outer groups. As such, the decision to canoe beyond the inner group to the outer group involves a high-risk voyage.

A key consideration for modelling Aboriginal occupation and changing use of central Queensland islands is the timing of island formation and development of marine resource habitats in relation to changing sea levels associated with the latter stages of the marine transgression over the past 10000 years. Extensive coral reefs, mangrove forests, seagrass beds and rocky platforms combine to create a vast and diverse marine fauna and food resource base for the region. An understanding of the chronological development of these key resource zones is critical to understanding long-term changes in Aboriginal use of the region.

Sea-level change and island formation

Rowland's (1996) hypothesis that most continental islands of the southern Great Barrier Reef region formed around 6000 years ago or later is not applicable to the offshore islands of Shoalwater Bay. An estimation of when rising sea levels reached and surrounded High Peak and Otterbourne Islands can be determined using recently reconstructed sea-level curves for the region (e.g. Larcombe et al. 1995: figure 5; Lewis et al. 2013: figure 2; Woodroffe 2009: figure 2; Yokoyama et al. 2006: figure 3) and local bathymetry (Australian Hydrographic Service 2004, 2009; Royal Australian Survey Corps 1986). The effects of Holocene fluvial inputs and sedimentation of the seabed upon these calculations is considered negligible, given that such sedimentation tends to wedge out to less than 1 m in thickness at depths of 15–20 m in the Great Barrier Reef region (Larcombe & Carter 2004; Larcombe & Woolfe 1999). While broad agreement exists on sea-level curves for the terminal Pleistocene and early Holocene (e.g. Larcombe et al. 1995; Lewis et al. 2013; Woodroffe 2009), some debate exists on the nature of sea-level changes in eastern Australia over the past 8000 years.

Rising sea levels associated with global warming after the height of the Last Glacial Maximum (LGM), ∼21000 years ago, saw sea levels rise and commence flooding of the continental shelf. At the LGM, the Pleistocene coastline, represented by the 120 m bathymetric line, was located 180 km north-east of High Peak Island (Page & Dickens 2005; Yokoyama et al. 2006). Woodroffe (2009) has proposed that after rising seas reached current levels ∼8000 calBP, that they continued to rise to at least 1.5 m above modern levels (peaking at + c.2.8 m around 5000 calBP) through to c.2300 calBP, before falling to modern levels only within the past 1000 years (Woodroffe 2009). Sloss et al. (2007) suggest that the highstand of +1.0–1.5 m occurred earlier, between 7700 and 7400 calBP, with the fall to modern levels similarly taking place over the past 2000 years. Lewis et al. (2008) argue for sea levels rising to +1.0–1.5 m around 7000 calBP, with modern levels attained only within the past 2000–1000 years along eastern Australia. Yu and Zhao's (2010) recent study of Holocene sea levels on Magnetic Island (Great Barrier Reef region) also found that after sea levels first reached current levels around 7300–7100 calBP, they continued to rise to a peak of +1.6 m at 5800 calBP and remained high before falling to modern levels after 2200 years ago. Perry and Smithers (2011) similarly consider Woodroffe's (2009) highstand of up to +2.8 m at 5000 calBP too high and too recent. In contrast to recent sea-level studies, Perry and Smithers (2011: appendix 2) follow Chappell's (1983) original hypothesis that sea levels have been “smoothly falling” to modern levels since the mid-Holocene highstand. Despite debate, Perry and Smithers (2011: 78) conclude that for the Great Barrier Reef region, “there is general consensus that the highstand was reached by ∼7000 cal yBP or slightly earlier, was around 1–1.5 m above present, and then fell to be near present for the past ∼1000 years”.

High Peak Island is located adjacent to the 50 m bathymetric line at a distance of 40 km from the mainland coast (Figure 2). The sea reached −50 m at around 12000 calBP, at which point High Peak Island was High Peak Hill, forming a headland (see Lewis et al. 2013: figure 2). High Peak Island became separated from the mainland to form an island, albeit much larger than its current configuration, as the sea rose to around −40 m around 11500 calBP. The modern configuration of High Peak Island probably occurred when the sea reached around −20 m around 10000–9000 calBP.

figure

Figure 2. The Shoalwater Bay region at different periods of sea-level rise.

Download figure to PowerPoint

Otterbourne Island is slightly later in formation because it is located closer to the current mainland coast. When rising seas reached the 30 m bathymetric line around 11000–10000 calBP, Otterbourne Island was a hilltop, located approximately 10 km inland (Figure 2). As the sea level rose to the 18 m bathymetric line in approximately 10000–9000 calBP, Otterbourne Island formed part of a large island that also included Hexham, Alnwick and Shields Islands, located approximately 6 km off what was then the mainland coast. Otterbourne Island (and the modern mainland coast located 26 km away) first attained their modern configuration when sea levels rose to current levels around 8000 calBP.

Collins Island is located 14 km from the mainland coast but was still part of the mainland when the sea reached the 20 m bathymetric line approximately 10000–9000 years ago (Figure 2). Around 9000 years ago, Collins Island was connected with Lingham Island as a large island, with the two islands separating and attaining their modern configurations when sea levels rose to current levels around 8000 calBP.

These sea-level reconstructions for the formation of High Peak, Otterbourne and Collins Islands are also broadly in line with Barker's (2004: 49–54) and Border's (1999: 130–131) reconstructed chronologies for the formation of islands of the Cumberland and Northumberland Groups.

Rocky reefs

Rocky reefs fringe numerous sections of mainland coast and the surrounds of most islands across the region, particularly the outer islands. The area and extent of rocky reefs around islands decreases moving from the inner group to the outer group of islands (Ayling et al. 1998). According to Trnski et al. (1994: 250), the headlands around the Shoalwater Bay region are “the most extensive rocky reefs north of NSW”. These reefs form a substrate for many different marine invertebrates (e.g. gastropods) and fish. In addition, red algae that grows on rocky reefs is a “significant wet weight component” of the diet of green turtles, “particularly in those animals caught in rocky reef or basking on rocks at low tide” (Limpus et al. 2005: 21). A range of rocky-platform gastropods, bivalves and chitons occur within the intertidal zone of rocky reefs across the region (e.g. Endean et al. 1956).

Coral reefs

Hard coral on average represents less than 40% coverage of reefs in the region, with another 40% represented by seagrass and algae (Ayling et al. 1998: 9). Most reefs are not true coral reefs, but “corals growing on rocks or rubble banks” (Ayling et al. 1998: 17). Strong tidal currents, high turbidity and large tidal ranges produce a “harsh environment” for corals in the Shoalwater Bay region and Northumberland Group (Ayling et al. 1998: 17; Kleypas 1996; van Woesik 1992; van Woesik & Done 1997). As a result, fringing coral reefs in the Shoalwater Bay region and Northumberland Group are poorly developed compared with adjacent sections of the Great Barrier Reef to the north (Whitsunday Islands) and south (Keppel Islands).

Mangroves and seagrass

Mangrove forests cover 270 km2 of the region and fringe 50% of the mainland coastline (particularly around Shoalwater Bay and Port Clinton). Small areas of mangroves can be found on inner islands such as Townshend and Leicester Islands and the Skull and Cannibal Groups (ABARE et al. 1993: 68; O'Neill 2008a). Seagrass meadows cover 130 km2 and are similarly concentrated within the more protected waters of Shoalwater Bay and Port Clinton across intertidal mudflats (Lee Long et al. 1997). Both mangroves and seagrass are important habitats for fish and marine invertebrates, with seagrass a significant food source for dugongs (Dugong dugon) and seagrass and mangroves important food sources for green turtles (Chelonia mydas) (Lanyon et al. 1989; Limpus et al. 2005). As such, the bulk of green turtles and dugongs in the region are found in close proximity to the mainland coast. Estimates suggest that “about 500 turtles per kilometre may be present over at least 30 kilometres of mainland coastline [of Shoalwater Bay] during the winter months” (O'Neill 2008b: 252). The Shoalwater Bay region has “possibly the second-largest concentration of green turtles on the east coast after Torres Strait” (Commission of Inquiry 1994: 95). Green turtle nesting sites have been recorded on Collins, Lingham and High Peak Islands off Shoalwater Bay and immediately north on a number of offshore islands (Percy Islands, Curlew Island and Pine Peak Island) (Limpus 2008: 121; Limpus et al. n.d.). Rogers (1888: 7) also observed green turtles nesting on High Peak Island and stated that “this was evidently a favourite laying place”. Furthermore, “if the eggs all hatched the sea around these islands would be thick with turtles” (Rogers 1888: 7). The region also supports the largest dugong population in the southern Great Barrier Reef region (Marsh et al. 1996; Slater & Stokes 1997). The “highest densities of dugongs were associated with inshore seagrass beds” (Marsh et al. 1996: 8). Rogers (1888: 11) recorded “big herds” of dugong in “Shoalwater Bay”. Thus, the most productive areas for green turtles and dugongs in the study area are within the intertidal zone of the more protected waters within Shoalwater Bay itself.

Terrestrial resources

Neither faunal nor floral surveys have been undertaken for Shoalwater Bay islands. A controlled faunal survey of South Percy Island, in the north-west of the study area, revealed only one possible mammal species (possum, Trichosurus sp.), a few snakes and skinks and numerous birds (Rees 1980). Border (1985: 107) rightly suggests that the other islands in the Northumberland Group exhibit a similar “depauperate representation of mammal species”. In marked contrast, the terrestrial mainland fauna of Shoalwater Bay is diverse (ABARE et al. 1993: 65–67; Catling et al. 1994; Schodde et al. 1992). The plant food resource base of these islands is poorly understood. Fresh water is available on all islands, especially in small creeks or depressions after rain.

Ethnographic context

The most detailed and authoritative analysis of nineteenth century Aboriginal social groupings in the Shoalwater Bay region is Memmott (1993). Based largely on the detailed records of Roth (1898, 1910) and Howitt (1904), Memmott (1993) concluded that the region took in the northern parts of the Darumbal language group and two of its four dialect groups – the Kuinmurbura in the west and the Ningebul in the east – each composed of six or seven and five or six patriclans, respectively. These patriclans had estates focused across various parts of the mainland. In contrast, the estate of Rundubura patriclan of the Ningebul dialect was located on Townshend Island. No other groups are identified as specifically island peoples of the region. This evidence suggests strongly that the Northumberland Islands were used primarily by groups whose estates were based primarily on the mainland. Memmott (1993: 19; 1994: 38) suggested that the islands of the Northumberland Group (including the Percy Islands) may have been used by the northern Darumbal.

Nineteenth century observations of Aboriginal people in the Shoalwater Bay region are scant, and limited mostly to the inner islands and mainland coast (Border 1999: table 1). Few records were found of people on the outer islands. Available recordings emphasise marine subsistence practices, and in particular turtle and dugong hunting and use of canoes. Two key recordings were made by Matthew Finders' expedition in 1802. On 30 August, the expedition visited an islet in the mangrove-dominated upper sections of Shoalwater Bay itself:

This islet had been visited by Indians, and several trees upon it were notched, similar to what is done by the people of Port Jackson when they ascend in pursuit of opossums. Upon the main, to the west of the islet, where I walked a mile inland, fire places and other signs of inhabitants were numerous, and still more so were those of kangaroo; yet neither that animal nor an Indian was seen. Around the extinguished fires were scattered the bones of turtle, and the shells of crabs, periwinkles, and oysters of the small kind; and in the low grounds I observed many holes, made apparently by the natives in digging for fern roots. (Flinders 1814: 46)

On 1 September, along the mainland coast, possibly between Raspberry and Shoalwater Creeks, Flinders (1814: 47) met up with a party of 16 Aboriginal people:

They had bark canoes which, though not so well formed, were better secured at the ends than those of Port Jackson; and in them were spears neatly pointed with pieces of quartz, for striking turtle. The number of bones lying about their fire places bespoke turtle to be their principal food; and the addition of shell fish, and perhaps fern roots, it is probably their sole support.

On 20 July 1820, Captain Phillip P. King in the Mermaid sailed into Port Clinton and met two Aboriginal men, one of whom gave them “a fishing-line spun and twisted of strips of bark, to the end of which was attached a hook made from a turtle-shell” (King 1827, I: 352–360).

On 23 June 1843, The HMS Fly expedition sailed past the “Townshend's Group” at night and John Sweatman recorded that “we observed a line of fires extending at regular distances completely across one of these islands” (cited in Allen & Corris 1977: 54). Whilst anchored in Port Clinton during February 1843, the expedition had direct interactions with numerous Aboriginal people, including some with fishing nets and baskets (Jukes 1847, I: 24–31).

Sometime in 1886–1887, Edmund de Norbury Rogers (1886–1887) was sailing within strong tidal current waters in the vicinity of the Mumford Islands and made a rare recording of turtle hunting and the dangers of canoe travel in the area:

About here Barcoo [Aboriginal man] told us a canoe was lost some years ago and two black boys drowned. They were looking for turtle in calm weather and had left their gins (wives) on this small rock, when, as often happens on this coast, a thunder-storm came up suddenly from the W. The frail canoe was soon swamped and the gins left alone. They made a fire and our Barcoo came over from Leicester Island and rescued them.

Furthermore:

Barcoo knows all the islands about here pretty well, and what animals are to be found upon them. The distribution seems to be rather strange, some islands have no opossums, others plenty, some no native bears, some no brown snakes, some have death adders, so as we walked round the beach [Collins Island] we lit the grass everywhere.

On a more sinister note, Barcoo reported that “a police inspector and black trackers” arrived on Collins Island in the early 1870s “and dispersed the blacks, with the usual results” in retribution for the murder of a local Chinese bêche-de-mer operator (Rogers 1886–1887).

During a dugong hunting trip to Shoalwater Bay in 1888, Rogers (1888: 9–10) recorded that:

The blacks used to hunt dugong in their bark canoes using a wooden or a bone peg with line attached which came away from the spear when it struck. As the bulls or cows weigh up to 14 cwt, measure 10 feet in length and 7 feet in girth, they are unmanageable with a canoe and the blacks left them alone. Besides they are not nearly as good eating as calves. Sometimes the natives use an ordinary bit of fencing wire instead of the wooden or bone peg; this when it enters the hide nearly an inch thick in the old males and very tough, bends over and holds firmly without a barb. When the sea-calf is growing exhausted they give the coup de' grace with a wooden peg driven into the valve like nostril. The dugong, which cannot do without air – and comes up to the surface to breathe every three minutes, thus dies of suffocation, and is towed ashore and eaten amid high festivities. At one time only the old men were allowed to eat dugong meat.

Also in the 1880s, Carl Lumholtz (1889: 340) participated in a dugong hunt in Shoalwater Bay. The party sailed out to an island (probably in the Cannibal Group) and noted:

Early the next morning two natives, who were to assist us in hunting, came rowing in a canoe from the mainland. One of them paddled the canoe while the other one kept baling out the water with a large shell.

The canoe of the natives here is made of three pieces of bark, one forming the bottom and two the sides. The pieces are sewed together with wood fibres, and there is nothing, by way of ribs, to keep the pieces of bark together; simply a small cross-piece to support the sides, nor are there row-locks or rudder. There is only room for two, and as the water continually pours in, one man is occupied in baling, while the other paddles on the two sides alternately with a stick about two yards long.

Three-piece canoes, measuring up to “8 or 10 feet” (2.4–3.0 m), were also described for the Darumbal by Roth (1898: 47–48). They are typical of the Northumberland and Cumberland (including Whitsunday) Groups and were clearly designed for open sea voyaging to offshore islands (Barker 2004: 39–41).

Archaeological fieldwork

In 1998, an archaeological site survey was conducted on selected islands of the Shoalwater Bay region (McNiven 1999a). This Stage 2 work was commissioned by the Darumbal–Noolar Murree Aboriginal Corporation for Land and Culture and geographically extended and complemented a previous archaeological site survey of the Shoalwater Bay Military Training Area carried out in 1996–1997 (McNiven & Russell 1997). The Stage 1 work was commissioned jointly by the Darumbal–Noolar Murree Aboriginal Corporation for Land and Culture, Queensland Department of Environment and the Department of Defence in response to recommendations of the 1993 Commonwealth Commission of Inquiry into the Shoalwater Bay Training Area (Commission of Inquiry 1994). As a follow-up to the Stage 2 survey, the Darumbal–Noolar Murree Aboriginal Corporation for Land and Culture and IM agreed to gain greater insights into the history of Darumbal use of the Shoalwater Bay islands by excavating a sample of previously recorded sites on three islands located at increasing distance from the mainland coast – Collins Island, Otterbourne Island and High Peak Island.

Otterbourne Island

  1. Top of page
  2. Abstract
  3. Introduction
  4. Shoalwater Bay Islands
  5. Otterbourne Island
  6. High Peak Island
  7. Collins Island
  8. Discussion
  9. Conclusion
  10. Acknowledgements
  11. References
  12. Supporting Information

Otterbourne Island is located 26 km from the mainland (Stanage Point) and 5.8 km from Hexham Island to the east. By island-hopping through the Cannibal Group, the longest single sea crossing is reduced to 19.0 km (from Mumford Island to Hexham Island). The island has a maximum length of 1.6 km and width of 0.5 km (area = ∼65 ha). It is topographically rugged, with steep slopes rising from the sea to an undulating ridge and plateau, elevated between 50 m and 83 m above sea level (ASL). Most of the island comprises Lower Cretaceous (>100 mya) quartz arenite (i.e. sandstone) and pebble conglomerate sedimentary rocks, most likely representing ancient lake beds, which are exposed as a thin 6 m thick capping over what appears to be ancient Palaeozoic metamorphic rocks (Murray 1975: 7–8). Soils are represented mostly by shallow pockets of brown to grey–brown silty sands and gravels over weathered bedrock (Grant et al. 1979: 86). Vegetation is coastal shrublands with “dry rainforest and closed scrub elements” (Cenko et al. 1990). During archaeological fieldwork, it was observed that the vegetation included eucalypt shrublands with scattered pandanus trees on some slopes. It appears that the island has been little used by Europeans. Beach deposits observed on the north coast contained pebbles and cobbles of quartz (poor quality), quartz conglomerate and metamorphic rocks. Outcrops of quartz conglomerate and quartzite were recorded at the western end of the island. Two sandy beaches, located on the north and south coasts, provide safe landing locations for canoes. In contrast, the remaining coastline features rock outcrops, most of which extend out to form extensive rocky platforms that are exposed at low tide. A number of green turtles were observed swimming across the reef on the south side of the island during the excavation of Site 4. The presence of the remains of juvenile turtles associated with a recent sea eagle roost 30 m from Site 4 suggests strongly that green turtles nest on Otterbourne Island.

Site 4 excavations

This extensive and rich midden deposit is located amongst outcrops of coarse-grained, friable sandstone on the edge of the flat plateau at an elevation of 60 m ASL and adjacent to the steep erosion slope that descends down to the sea on the south-west coast of the island. It was recorded by IM with Darumbal representatives Bob Muir and Theresa Chelepy-Roberts on 23 April 1998. The view from the site is spectacular and the Peninsula Range (mainland) located 42 km to the south-east can be seen easily on a clear day. The site is exposed over an area of ∼75 m × ∼35 m (Figure 3). Vegetation is stunted eucalypt shrubland with prickly pear cactus. Sediments are brown sandy loam. Midden deposit is exposed across the ground surface and extends under low sandstone overhangs in places. These overhangs are small and often less than 50 cm high. Severe site disturbance from megapod fowls has left a number of burrows associated with raking of surface sediments to construct a large mound nest on the site (Figure 3). The mound is over 1 m high and has a diameter of 10 m. It is covered with oyster shell and numerous stone artefacts. The bulk (<95%) of the matrix of the mound is soil and leaf litter. No recent human disturbance to the site was observed.

figure

Figure 3. The site plan of Otterbourne Island Site 4.

Download figure to PowerPoint

Two locations (Squares A and B) were selected for excavation (Figure 3). Square A is a 50 cm × 50 cm pit, positioned within a concentration of midden deposit located under a small and low protective overhang within a sandstone outcrop at the northern end of the site. It was orientated at an angle of 348°, with the north-east corner located 32 cm from the back wall of the small shelter. Square B is a 1 m × 1 m pit located within a separate midden concentration (the highest-density midden deposit observed at the site) adjacent to a sandstone outcrop at the northern end of the site. It was orientated at an angle of 45°, with the east corner located 121 cm from the back wall of the adjacent low shelter. Both squares were excavated by IM with Darumbal representatives Lance Aaskov, Andrew (AJ) Batchelor and Willie Hayden during 20–23 September 2002.

Square A was excavated using ten arbitrary Excavation Units (XUs), with all attempts made to avoid mixing different stratigraphic layers within the same XU. The XUs averaged 3.5 cm in thickness (range 1.2–5.5 cm). Excavation ceased at a maximum depth of 38.7 cm (south-east corner) due to increasing encroachment of sandstone bedrock and a dearth of cultural materials. The weight (to the nearest 0.1 kg) and volume (to the nearest 0.5 L) of each XU was recorded, and depth elevations were taken at the corners and centre of the square at the beginning and end of each XU. All excavated deposit was dry sieved through 2.1 mm mesh on the island and then wet sieved through 2.1 mm mesh back in Melbourne. Samples of <2.1 mm dry sieved sediment were taken for each XU. A total of 60.5 kg of deposit with a volume of 47.5 L was excavated.

Square B was excavated in a similar fashion to Square A. The 17 XUs averaged 5.5 cm in thickness (range 2.6–8.4 cm). Excavation proceeded to a maximum depth of 95.3 cm (north corner) and ceased due to difficulty of access, encroachment of rocks and an apparent dearth of cultural materials. XUs 9–17, taking in the lower 54 cm of the pit, were excavated as a 50 cm × 50 cm pit in the north quadrant of Square B due to time constraints. A total of 670.5 kg of deposit with a volume of 499.5 L was excavated. Both squares were backfilled with sediments from the sieve spoil heap.

Stratigraphy and sedimentary environment

The stratigraphy of Squares A and B is basic. Square A featured a single Stratigraphic Unit (SU1), which consisted of very dark grey (10YR 3/1) to very dark greyish brown (10YR 3/2), dry, unconsolidated sandy loam with numerous small fragments of exfoliated sandstone and small roots scattered throughout (Figure 4). The pH was a uniform 5.5 (acidic) (Supporting Table S1). Midden deposit was restricted mostly to the upper 5 cm of deposit, with stone artefacts recovered throughout the sequence.

figure

Figure 4. Stratigraphy, Square A, Otterbourne Island Site 4.

Download figure to PowerPoint

Square B similarly comprised dry, unconsolidated sandy loam, albeit separated into two stratigraphic units (Figure 5). SU1 took in the bulk of midden deposit down to a depth of ∼70 cm below the surface and contained numerous scattered rootlets. Colour varied from black (10YR 2/1), to very dark grey (10YR 3/1) to very dark greyish brown (10YR 3/2). SU2 represented the lower ∼25 cm of deposit and contained stone artefacts (and few other cultural materials) and scattered rootlets. It was mostly very dark greyish brown (10YR 3/2), becoming dark greyish brown (10YR 4/2) at the base. The pH was 7.5–8.0 (neutral) (Supporting Table S2). Both SUs exhibited numerous small fragments of exfoliated sandstone. The following analyses focus on Square B, given that Square A contained a similar range of cultural materials but at significantly lower amounts.

figure

Figure 5. Stratigraphy, Square B, Otterbourne Island Site 4.

Download figure to PowerPoint

Land snails. A considerable amount (3.0 kg) of land snail shell was recovered from Square B, compared to only 4.8 g recovered from Square A (Supporting Tables S1 and S2). Snail shell was found in all XUs of Square B, with highest densities occurring in XUs 6–10 (Figure 6).

figure

Figure 6. The density of cultural materials, Square B, Otterbourne Island Site 4.

Download figure to PowerPoint

Rocks. A total of 3.4 kg and 19.0 kg of rock rubble was recovered from Squares A and B, respectively (Supporting Tables S1 and S2). Nearly all were small fragments of local coarse-grained sandstone. The density of rock rubble was relatively consistent throughout the Square B sequence, with slightly higher levels in the lower half of the pit in XUs 10–17 (Figure 6).

Taphonomy and preservation

Apart from roots, little evidence of disturbance (e.g. burrows and/or insects) was observed during excavation of Squares A and B. Lack of mounding of midden deposits at both excavation locations suggests strongly that nest mounding associated with megapods, which currently inhabit the site, has not impacted the excavated deposits. Recovery of faunal remains through the entire Square B sequence, including small fish bones, suggests that faunal preservation in the square is good and that vertical changes in shell and bone for most of the sequence are not a function of differential preservation. This inference is consistent with the recording of pH values of 7.5–8.5 through the Square B sequence. Yet the paucity of faunal remains in the lower levels of Square B most likely reflects poor preservation to some degree, given the presence of stone artefacts in basal levels (Figure 6). The impact of poor preservation on faunal preservation is more apparent in Square A, where a paucity of shell and bone in most XUs matches a consistently low pH of 5.5. Disturbance of deposits from storm surges is impossible, given the site's 60 m elevation above the shoreline.

Radiocarbon dating and deposition rates

A comprehensive suite of 22 radiocarbon dates are available for the site – four from Square A and 18 from Square B (Table 1 and Figure 7). Dating was undertaken by the University of Waikato Radiocarbon Dating Laboratory in New Zealand. Radiocarbon dates were calibrated into calendar years using the online calibration program Calib 6.0 (Stuiver & Reimer 1993) and the Southern Hemisphere calibration dataset (ShCal04) (McCormac et al. 2004) for charcoal dates, and the marine calibration dataset (Marine09) (Reimer et al. 2009) for the marine shell date, using a ΔR value of 10 ± 7 years determined for central Queensland (Ulm 2002). A single, central best-point estimate was calculated for the irregular probability distribution of each date using the median calibrated age (following Telford et al. 2004) and rounded to the nearest 50 years for heuristic and discussion purposes.

figure

Figure 7. Calibrated radiocarbon dates, Shoalwater Bay islands (probability: white, 95.4%; black, 68.3%).

Download figure to PowerPoint

Table 1. Radiocarbon dates for Squares A and B, Otterbourne Island Site 4
Lab. no.SquareXUDepth (cm)SampleMethodδ13C‰ (± 0.2)Conventional 14C age (years BP)Calibrated age BP 68.3% (with probabilities)Calibrated age BP 95.4% (with probabilities)Median probability calBP
Wk-32248A21.2–4.5

Wood

Charcoal

1 fragment

0.11 g

AMS−25.6982 ± 25

800–846 (0.565)

862–889 (0.323)

896–906 (0.112)

791–920 (1.000)852
Wk-32247A510.2–13.0

Wood

Charcoal

1 fragment

0.08 g

AMS−25.3918 ± 25743–792 (1.000)

729–805 (0.885)

871–900 (0.115)

773
Wk-32246A715.9–19.9

Wood

Charcoal

1 fragment

0.03 g

AMS−25.2904 ± 25

735–775 (0.911)

782–787 (0.089)

723–801 (0.959)

874–897 (0.041)

762
Wk-14637A924.8–30.3

Wood

Charcoal

2 fragments

0.11 g

AMS−24.53059 ± 40

3081–3092 (0.054)

3114–3121 (0.028)

3141–3264 (0.831)

3297–3303 (0.023)

3305–3319 (0.064)

3042–3046 (0.003)

3063–3347 (0.997)

3201
Wk-14636B10–3.9

Oyster

Saccostrea cucullata

Fragments

37.91 g

Standard2.3586 ± 35

146–165 (0.171)

185–268 (0.829)

102–294 (1.000)210
Wk-32249B23.9–9.4

Wood

Charcoal

1 fragment

0.32 g

AMS−26.0128 ± 25

0-0 (0.004)

25–71 (0.457)

84–89 (0.038)

92–104 (0.096)

113–140 (0.272)

229–244 (0.133)

0–143 (0.844)

225–254 (0.156)

90
Wk-26911B39.4–14.2

Wood

Charcoal

1 fragment

0.16 g

AMS−24.0470 ± 30

474–480 (0.092)

486–513 (0.908)

340–352 (0.030)

451–526 (0.970)

497
Wk-32250B414.2–16.8

Wood

Charcoal

1 fragment

0.48 g

AMS−23.8768 ± 25656–680 (1.000)

573–587 (0.052)

646–693 (0.842)

697–722 (0.106)

670
Wk-32251B516.8–20.9

Wood

Charcoal

1 fragment

0.13 g

AMS−24.5771 ± 25657–681 (1.000)

575–585 (0.032)

648–722 (0.968)

671
Wk-26912B620.9–26.4

Wood

Charcoal

2 fragments

0.32 g

AMS−24.8751 ± 30

575–585 (0.176)

648–677 (0.824)

566–598 (0.235)

632–686 (0.736)

704–718 (0.028)

659
Wk-32252B726.4–32.6

Wood

Charcoal

1 fragment

0.17 g

AMS−25.9972 ± 25

795–822 (0.332)

830–842 (0.124)

864–905 (0.543)

774–783 (0.029)

786–915 (0.971)

850
Wk-32253B832.6–38.6

Wood

Charcoal

1 fragment

0.15 g

AMS−25.11129 ± 25

935–943 (0.075)

955–988 (0.6310

1032–1051 (0.294)

931–1013 (0.745)

1026–1055 (0.255)

979
Wk-32254B938.6–45.3

Wood

Charcoal

1 fragment

0.09 g

AMS−23.91514 ± 251312–1366 (1.000)1302–1394 (1.000)1345
Wk-26913B1045.3–52.6

Wood

Charcoal

1 fragment

0.30 g

AMS−23.51229 ± 30

1057–1144 (0.896)

1159–1169 (0.104)

982–1036 (0.162)

1046–1177 (0.838)

1098
Wk-26914B1152.6–58.8

Wood

Charcoal

8 fragments

0.15 g

AMS−25.52785 ± 302781–2853 (1.000)

2757–2885 (0.973)

2908–2921 (0.027)

2822
Wk-32255B1258.8–65.1

Wood

Charcoal

1 fragment

0.07 g

AMS−25.03474 ± 25

3616–3624 (0.042)

3627–3705 (0.935)

3712–3715 (0.023)

3574–3728 (0.905)

3748–3764 (0.025)

3792–3823 (0.071)

3668
Wk-26915B1365.1–69.5

Wood

Charcoal

2 fragments

0.18 g

AMS−24.13834 ± 32

4089–4163 (0.628)

4167–4180 (0.102)

4198–4231 (0.270)

3990–4044 (0.091)

4069–4259 (0.890)

4263–4267 (0.003)

4271–4288 (0.016)

4152
Wk-12990B1469.5–76.6

Wood

Charcoal

Fragments

3.02 g

Standard−25.24068 ± 62

4412–4580 (0.967)

4769–4780 (0.033)

4259–4270 (0.005)

4288–4651 (0.906)

4670–4702 (0.026)

4758–4808 (0.063)

4493
Wk-32256B1576.6–85.0

Wood

Charcoal

1 fragment

0.11 g

AMS−24.64534 ± 25

5044–5073 (0.174)

5106–5131 (0.140)

5165–5279 (0.686)

4977–5013 (0.076)

5033–5290 (0.924)

5167
Wk-32257B1685.0–89.7

Wood

Charcoal

1 fragment

0.10 g

AMS−24.54456 ± 25

4878–4942 (0.518)

4951–4979 (0.236)

5008–5036 (0.246)

4860–5053 (0.967)

5190–5213 (0.031)

5253–5256 (0.001)

4963
Wk-32258B1789.7–93.0

Wood

Charcoal

1 fragment

0.01 g

AMS−23.34494 ± 25

4965–5060 (0.667)

5113–5118 (0.020)

5186–5215 (0.164)

5222–5238 (0.058)

5242–5265 (0.090)

4877–4943 (0.109)

4947–5075 (0.527)

5105–5134 (0.049)

5164–5279 (0.315)

5038
Wk-12991B1789.7–93.0

Wood

Charcoal

1 fragment

0.03 g

AMS−23.94580 ± 45

5057–5187 (0.716)

5214–5223 (0.040)

5236–5243 (0.029)

5262–5302 (0.215)

4979–5008 (0.029)

5036–5317 (0.971)

5164

The dates range from 128 ± 25 BP (charcoal) (c.100 calBP) near the surface in XU2 (Square B) to 4580 ± 45 BP (charcoal) (c.5150 calBP) at the base in XU17 (Square B). As XU17 at the base of Square B contained 17 stone artefacts and is not culturally sterile (Figure 6 and Supporting Table S2), the antiquity of basal cultural materials at the site is considered to be at least 5200 years ago. However, the rapid drop-off in the density of bone, shell and charcoal in basal XUs suggests strongly that earliest occupation of the site was of lower intensity prior to 5200 years ago.

Square A dates fall into two groups, with the upper three dates in XUs 2–7 tightly clustered to between c.750 and 850 calBP and overlapping at one sigma and a near-basal date in XU9 of c.3200 calBP. Square B dates also fall into two groups – an upper group between 0 and 53 cm below the surface (XUs 1–10) spanning a period of 1250 years between c.100 and 1350 calBP and a lower group between 53 and 93 cm below the surface (XUs 11–17) spanning a period of 2350 years between c.2800 and 5150 calBP. Using these figures, the deposition rate of cultural deposit between c.2800 and 5150 calBP is ∼1.7 cm per 100 years, while between c.100 and 1350 calBP the rate is two and half times higher, at ∼4.2 cm per 100 years.

The vertical chronostratigraphic integrity of dates is good, with dates mostly increasing in age with depth. Slight inversions with the four older dates in XUs 15–17 dating to between c.4950 and c.5150 calBP are considered insignificant, given the overlaps at one sigma. The only true dating inversion occurs in XUs 9 and 10, with dates of c.1350 and c.1100 calBP, respectively. This inversion most likely reflects small-scale treadage disturbance or/and post-depositional mixing of sediments.

Occupational hiatus c.1400–2800 years ago

Comprehensive dating of every XU of Square B reveals a major hiatus in midden deposition and hence occupation of 1400 years between c.1400 and c.2800 years ago (Figure 7). Dating of Square A revealed an occupational hiatus of greater duration (2350 years) between c.850 and c.3200 calBP. Given the proximity of Squares A and B, it is clear that the hiatus in occupation of Site 4 can only be argued for the period c.1400–2800 years ago.

Charcoal and plant food remains

A total of 237.1 g of charcoal was recovered from Square B and 9.2 g from Square A (Supporting Tables S1 and S2). Most (90%) charcoal in Square B was recovered from XUs 1–9, with the highest densities recorded in XUs 1–6 dating to the past 650 years. Charcoal represents burnt wood, but some fragments are from fruits and tubers that may represent plant foods consumed at the site over the past 4000, and possibly 5000, years. The bulk of the identifiable carbonised macrobotanical remains (excluding wood charcoal) appear to come from two taxa: Cyperus and Pandanus. Every XU contained minor quantities of other small plant fragments, potentially of fruit, that are yet to be identified.

The most common botanical remains throughout the deposit are small, ovoid to ellipsoid rhizome tubers. They have annular nodes and internodes along the tuber length and circular rhizome scars or truncated remains of narrow connecting rhizomes. These features combined are indicative of Cyperaceae, the sedges. All genera of Cyperaceae that produce woody, ovoid tubers were considered; the closest matches being shown by both Cyperus and Bolboschoenus (Sharpe 1986; Stanley & Ross 1989). Although generally tuber morphology of these two genera is similar (e.g. broadly ovoid or globular, presence of annular nodes and rhizome scars), Bolboschoenus tubers are much larger than those of Cyperus and accordingly are too large to match the Site 4 specimens. The ancient tubers are similar in size and morphology to those of the genus Cyperus, which has been recorded in the Shoalwater Bay region (Australian Department of Defence 2009). Based on the known modern distributions of Cyperus spp. and the limited tuber information available in the published literature, it is not possible to definitively place a binomial on the determination of the Otterbourne tubers. Modern reference material of the species under consideration is required for a determination to be confidently made.

Cyperus tubers have previously been encountered in archaeological records, often dominating charred assemblages (e.g. Cyperus rotundus in Egypt: see Hillman 1989), where they were used for food by hunter–gatherers. The tubers of various Cyperus spp. are known plant foods used by Aboriginal communities in Australia and carbonised remains have been recovered archaeologically (Atchison et al. 2005; Bird et al. 2005; Hayes & Hayes 2007; Isaacs 1996; Low 1988; McConnell & O'Connor 1997; Stephens & Sharp 2009).

Pandanus remains consisted of small fragments of phalanges, the segments of Pandanus fruits. There are two possible species from which the fragments may derive. P. tectorius and P. brookei are the only species of the genus present in the modern flora of Shoalwater Bay (Australian Department of Defence 2009; Bostock & Holland 2010) and currently grow near Otterbourne Island Site 4 (pers. obs. 2002). Carbonised remains of Pandanus fruits have been recovered from Hill Inlet rock shelter in the Whitsunday Islands, and a range of other archaeological sites across northern Australia and the western Pacific Islands, including Papua New Guinea (Barker 2004: 125; Clarke 1988; Hall 1980; Kirch 1989; Kirch et al. 1995; Matthews & Gosden 1997; Summerhayes et al. 2010; Swadling et al. 1991; Yamaguchi et al. 2005).

The use of Pandanus by people is varied and multiple plant parts are known to be used for both food and materials. The ripe phalanges can be roasted and the seeds removed from the hard fruit segments to eat, but seeds can equally be eaten raw (Isaacs 1996; Meehan et al. 1979; Stephens & Sharp 2009). After the fruit segments have been processed for food, their remains can be used as fuel for fires (Di Piazza 1998).

Remains of Pandanus phalanges were highly fragmented and, as such, no quantification of remains has been attempted. Plant remains are represented as presence or absence (Supporting Table S3). In general, remains were rare in the lower XUs and became more abundant towards the surface. From XUs 8 to 1 (Square B), we see the continued presence of Cyperus tubers and Pandanus phalange fragments. The abundance of Cyperus and Pandanus correlates well with the abundance of charcoal and other biological remains per XU (Figure 6). Earlier occurrences of Pandanus and Cyperus in lower XUs cannot be confirmed, as the material is too degraded.

Marine vertebrates

Of the 2672.6 g of bone recovered from Square B, most was identified as either “large vertebrate” (85%) or “small vertebrate” (7%) (Supporting Table S4). Remaining bone was identified as turtle (6%), fish (including shark) (2%), lizard/snake (<1%) and bird (<1%). The negligible Square A bone assemblage (4.9 g) was similarly dominated by “large vertebrate” (73%), with trace amounts of turtle, fish, snake, bird and “small vertebrate”. It is likely that most if not all large vertebrate bone represents turtle (given the absence of alternative identifiable large vertebrates) and most small vertebrate bone represents fish. As such, nearly all of the bone assemblage from Square B probably derives from turtle.

Turtle. Turtle bone was indentified from fragments of osteoderm (upper carapace or plastron) and flipper phalanges and most probably comes from green turtle (Chelonia mydas). As turtle bone was found in XUs 2–15 of Square B, it is clear that the bulk of animal meat consumed at the site over the past 5000 years derived from turtles.

Fish. Fish bone was found in all XUs of Square B, indicating that along with turtle hunting, fishing was a persistent activity carried out while visiting Otterbourne Island over the past 5000 years. However, some fish bones may have entered the site naturally, given that during site recording in 1998 an active sea eagle roost containing bones of juvenile turtles, fish and snakes was observed ∼30 m south of the site. Using a binocular microscope at ×7, a total of 233 fish bones (including five otoliths) was examined for identification to the nearest taxon. With rare exceptions, bones were <10 mm in length. Although the pH for Square B sediments was neutral as described previously – contributing to optimal preservation – only teeth, otoliths and very dense portions of mostly fragmentary grinding elements, dentaries and premaxillaries, primarily of wrasses (Labridae), were preserved. Identifications were made using an extensive reference collection held at the Pacific Lab, University of Queensland (Weisler 2001: appendix 3) and consisting of ∼350 individual specimens.

Supporting Table S5 lists the identifications of 173 bones (74% of total fish elements). The wrasses (Labridae) account for 92% of all identified elements, with only 13 bones from eel, emperors (Lethrinidae), parrotfish (Scaridae), groupers (Serranidae) and shark. The preferred habitats of these fish reveal they were most likely obtained from reefs fringing Otterbourne Island. The three genus-level identifications were made using otoliths, which – with better-preserved specimens – could be identified to species (see Weisler 1993). The single eel bone, a very small right dentary, is not Muraenidae or Congride, but may be a member of the Worm Eels (Moringuidae) or Snake Eels (Ophichthidae), which spend most of their time buried in sand (Randall et al. 1990: 33, 43) (a sandy beach is downslope from Site 4). Parrotfish was identified by two isolated teeth from upper pharyngeals. The shark teeth may be from the lower jaw of the Whaler or Requiem sharks.

All but a few bones were from fish less than ∼10 cm long. Three bones could be matched to sizes of fish in the reference collection: an upper pharyngeal and premaxillary of a wrasse, weighing ∼300 g and ∼500 g, respectively, and a premaxillary of a grouper, weighing ∼2200 g. The grouper is a real anomaly and probably weighs more than the reconstructed weights of all other fish in the Square B assemblage. Consequently, it is likely that <4 kg of fish are represented in Square B. The small overall size of fish forming the Square B assemblage is also reflected in the small mean size of vertebrae (Figure 8).

figure

Figure 8. The range and mean diameters of vertebrae centra, bony fish, Square B, Otterbourne Island Site 4.

Download figure to PowerPoint

Snake/lizard. Most bones from snakes and lizards (all vertebrae) were recovered from XUs 1–7 in Square B, dating to within the past 850 years. While some of the remains may represent natural on-site deaths or eagle roost introductions, at least some are considered to have resulted from consumption by Aboriginal people.

Bird. Most bird bones (all small fragments of hollow limb) were found in XUs 1–5 in Square B, dating to within the past 650 years, with a few fragments also recovered from XUs 10 and 11, dating to c.1100–2800 calBP. At least some of these remains are considered to represent dietary items, albeit minor.

Marine invertebrates

Remains of marine invertebrates in Square B were dominated by shellfish (79.8 kg), with minor representations of coral (504.9 g) and crustaceans (including barnacles) (76.6 g) (Supporting Table S2). A comparatively small amount of similar remains (1.7 kg) was recovered from Square A (Supporting Table S1). The nutritional value of coral and barnacles is negligible, and these items may have been collected as a form of material culture (e.g. coral abraders) or inadvertently attached to other marine animals such as turtles (e.g. barnacles). The recovery of fragments of crab exoskeleton from most XUs reveals that these creatures were collected from intertidal rock pools and represented a minor part of local diets. However, nearly all marine invertebrate foods were obtained in the form of shellfish.

Shellfish. Marine shellfish were recovered from every XU in Square B, indicating that shellfish have been consumed for over 5000 years on Otterbourne Island. However, most marine shells (83% by weight) were recovered from XUs 1–8 dating to within the past 1000 years. The Square A marine shell assemblage was restricted mostly (99%) to XUs 1 and 2, within 4.5 cm of the surface and dating to c.750–850 calBP. Only those Diagnostic Anatomical Elements (DAE) suitable for MNI calculations were separated out for analysis. DAEs had to be >50% complete to avoid double-counting errors. The number of DAEs used for each taxon varied from two to three. For bivalves, two DAEs were used – left and right hinges. For gastropods, up to three DAEs were selected from the following – apex, anterior canal, posterior canal, operculum and umbilicus. For chitons, two DAEs were used – anterior valve and posterior valve. In all cases, MNI was calculated for each taxon within each XU by taking the highest count of any individual DAE. Taxa identification was undertaken using the shell reference collection of the Programme for Australian Indigenous Archaeology (Monash University) and reference texts (Lamprell & Healy 1998; Lamprell & Whitehead 1992; Wilson et al. 1993, 1994).

A total MNI of 6683 shellfish, divided into bivalves (MNI = 536), gastropods (MNI = 4387) and chitons (MNI = 1760), was calculated for Square B (Table 2). The range of shellfish species is limited to only nine taxa, most of which are nerites (Nerita spp.) (31%), chitons (26%), turbo (Turbo spp.) (10%) and rock oysters (Saccostrea cucullata) (8%) (Table 2). Square A produced only three taxa, dominated by oyster (MNI = 52), followed by chiton (MNI = 8) and turbo (MNI = 5). Over 99% of all shellfish are rocky-platform species available from the intertidal zone surrounding Otterbourne Island. The relative proportion of major shell taxa changes through the Square B deposit such that when chiton, oyster and turbo dominate shell diets, top shells and nerites decrease in relative significance, and vice versa. Thus, chiton, oyster and turbo were of greater significance in XUs 1–4 and 10–14, dated c.0–650 and c.1100/1350–4500 calBP, whereas top shells and nerites dominated XUs 5–9, dated c.650–1100/1350 calBP and, concomitantly, oysters almost drop out of the diet in XUs 7–10 during the period immediately following the occupational hiatus, dated c.1400–2800 calBP (Figure 9). Why these changes in shellfishing strategies occurred is difficult to know. However, both top shells and nerites can be harvested easily from rocks at low tide and probably represent the most easily procured shellfish in the Otterbourne Island shellfish assemblage, whereas oysters are amongst the most difficult to recover in terms of dislodgement from rocks. In addition, top shells and nerites tend to contain smaller edible portions compared to the larger shellfish such as oysters, chitons and turbos. As such, a focus on procurement of top shells and nerites may indicate a period when shellfishing was a more marginal and trivial subsistence pursuit.

figure

Figure 9. Vertical changes in major shell taxa proportions, Square B, Otterbourne Island Site 4.

Download figure to PowerPoint

Table 2. Marine shell taxa MNI, Square B, Otterbourne Island Site 4
TaxonTidal zoneSubstrateXUTotal MNI
1234567891011121314151617
Bivalves                    

 Rock oyster

Saccostrea cucullata

IntertidalRocks + mangroves16712520223879631230191361  532

 Pearl shell

Isognomon sp.

IntertidalRocks + coral12               3

 Asaphis

Asaphis violascens

IntertidalSand  1              1
Gastropods                    

 Nerites

Nerita sp.

IntertidalRocks + mangroves317648321434946183341458610381231  2071

 Top shell

Monodonta labio

Upper intertidalRocks + mangroves351687628932694163621066765431  1639

 Turbo

Turbo spp.

Intertidal + subtidalRocks + coral reefs2913572255856677736401920144   652

 Limpet

Patelloida saccharina

Middle/lower intertidalRocks510   31          19

 Thaidinae

Morula sp.

IntertidalRocks + coral 1    12   1 1   6
Chiton                    
PolyplacopheraIntertidal + subtidalRocks1641321027081188250208133193103893881  1760
Total MNI:  4326493191774131089135998642138816817281254006683
Number of taxa:  78655676555656400 

Material culture

Stone artefacts. A considerable number of flaked stone artefacts was recovered from Square A (n = 187, 232.20 g) and Square B (n = 1176, 1640.3 g) (Supporting Tables S1 and S2). The density of stone artefacts is relatively similar through most of the Square B sequence, except for much higher concentrations in XUs 10 and 11, dating to c.1100–2800 calBP (Figure 6). In terms of fracture types recorded for Square B, most artefacts are flakes (n = 1089, 93%), followed by cores (n = 58, 5%), flaked pieces (n = 19, 2%) and retouched flakes (n = 10, <1%) (fracture-type definitions after Hiscock 1984: 129). Most (n = 39, 67%) cores revealed signs of bipolar reduction. Most artefacts were manufactured from either quartz (n = 848, 72%) or volcanic (n = 267, 23%), with the remaining raw materials represented by quartzite (n = 34, 3%), chert (n = 23, 2%) and sandstone (n = 4, <1%). The proportion of artefacts within each raw material group exhibiting water-rolled cortex varied: quartzite (n = 10, 43%), volcanic (n = 61, 23%), sandstone (n = 1, 25%), quartz (n = 105, 12%) and chert (0%). Water-rolled cortex suggests strongly that the source of most raw materials was locally available pebbles and cobbles of similar raw materials found as beach deposits along the shoreline of Otterbourne Island. In contrast, the most likely source of chert is the adjacent mainland, where chert outcrops are known and chert represents 9% of raw materials recorded within stone artefact scatters (McNiven & Russell 1997: 47–48).

A number of chronological changes in stone artefact manufacture occur through the Otterbourne Island sequence (Table 3 and Figure 10). First, the proportion of quartz stone artefacts increases from 24% in XU17 at c.5150 calBP through to 100% in XU1 at c.100–200 calBP. Whereas quartz represents on average 55% of artefacts in XUs 8–17, dating to between c.1000 and 5150 calBP, in XUs 1–7, dating to the past 850 years, the proportion increases to 88%. Second, as the proportion of quartz increases through time, the proportion of volcanic raw materials decreases. Third, most chert artefacts (78% by number) were recorded in XUs 10 and 11, located on either side of the occupational hiatus dating between c.1400 and c.2800 calBP.

figure

Figure 10. Vertical change in the proportion of quartz, volcanic and chert stone artefact raw materials, Square B, Otterbourne Island Site 4.

Download figure to PowerPoint

Table 3. Stone artefact raw materials, Square B, Otterbourne Island Site 4
XUQuartz (milky), numberQuartz (milky and crystal), numberQuartz (total), numberQuartz (%)Volcanic, numberQuartzite, numberChert, numberSandstone, numberTotal number
 110010100.0000010
 28278986.410202103
 398210089.312000112
 45225476.15100271
 57778484.81500099
 68438790.6900096
 7103310695.55000111
 87658178.619030103
 93203265.31331049
107337650.063580152
112553042.930010070
121912048.81911041
131682468.61100035
141912057.11140035
151411540.52020037
161331645.71630035
1731423.5940017
Total7965284872.1267342341176

Ochre. Fragments of rock rich in red iron oxide soft enough to be scratched with a fingernail were found in Square A (19.0 g) and Square B (114.3 g) (Supporting Tables S1 and S2). A considerably higher density of ochre was found in XUs 11–13 in Square B, dating to c.2800–4150 calBP (Figure 6).

Pumice. A small amount (4.2 g) of pumice was found scattered through Square B between XUs 1 and 12, but none was recovered from Square A (Supporting Table S2). Highest concentrations of pumice were found in XUs 10 and 11, dating to c.1100–2800 calBP (Figure 6).

Summary

Otterbourne Island Site 4 is a dense midden deposit spanning mostly the past 5200 years, with no evidence for occupation during the period of initial island formation, 6000–8000 years ago. Comprehensive radiocarbon dating indicates a horizontal expansion of midden deposits around 3200 years ago, with most occupation taking place within the past 850 years. A hiatus period of little or no occupation took place between 1400 and 2800 years ago. Subsistence activities focused on exploitation of nearby marine resources such as fish, shellfish, crustacean and especially turtle. Rare insights into plant food resources were provided by carbonised remains of pandanus fruits and sedge tubers. The proportion of quartz artefacts increases through time, with the use of high-quality mainland cherts restricted to either side of the hiatus period.

High Peak Island

  1. Top of page
  2. Abstract
  3. Introduction
  4. Shoalwater Bay Islands
  5. Otterbourne Island
  6. High Peak Island
  7. Collins Island
  8. Discussion
  9. Conclusion
  10. Acknowledgements
  11. References
  12. Supporting Information

High Peak Island is the most outer island in the Shoalwater Bay region (after the Percy Islands) and is located 40 km from the mainland (Pinetrees Point) and 32 km from Townshend Island. It is the most easterly continental island off the Australian east coast. The island can be reached by island-hopping, with the largest leg running 21 km across open seas between Townshend Island and Cheviot Island. It is possible to see High Peak Island on a clear day from elevated areas at the northern end of The Peninsula (mainland) and vice versa. The S-shaped island has a maximum length of 2.2 km and width of 1.5 km, with an area of 219 ha. The island is Crown land within the Great Barrier Reef Marine Park (Cenko et al. 1990). The baseline geology of the island is not known, as it is located beyond the boundary of the 1:250000 Australian geological map (Murray 1975). However, during the archaeological survey it was felt that the island most likely represented part of the Lower Cretaceous (> 100 mya) pyroclastic and flow (lava) rocks that make up neighbouring islands (Murray 1975: 7).

A large beach ridge consisting of sands with coral and shell fragments extends along the large bay on the west coast of the island. Mixed in with these marine deposits are scattered pebbles and cobbles of volcanic rock, and very rarely milky quartz pebbles. Beach rock, represented by cemented coralline sands, also fringes the bay in front of the sandy ridge. Fronting the bay is a large coral reef, which formed mostly between 7800 and 5800 calBP (Kleypas 1996). Rocky coastline, often with steep cliffs, flanks the remainder of the island. Vegetation includes extensive areas of hoop pine (Araucaria cunninghamii) and dry rainforest patches (Cenko et al. 1990), along with coastal shrubland and casuarinas, acacias, figs and prickly pear (Opuntia sp.). Feral goats roam wild over the island, causing denudation of shrub vegetation and ground covering, which has resulted in extensive erosion and exposure of bare ground. Rogers (1888: 7) observed nesting green turtles on the island and active turtle nesting was similarly observed at the northern end of the sandy beach backing the large western bay during fieldwork in 2002. It is clear from Rogers (1886–1887) that High Peak Island has been visited by Europeans for well over a century. A small metal shed (toilet) and timber frames from another structure located on the west coast beach ridge suggest sustained camping on the island in recent times.

Site 10 excavations

A scatter of shells and stone artefacts was recorded over a 20 m × 10 m area within hoop pine forest on a flat ridge top elevated >100 m ASL at the south-west corner of the island. The site was recorded by IM and Darumbal representative Bob Muir on 21 April 1998. Surface sediments included tuff rock rubble and vegetation ground cover was minimal. Megapod nesting activity (e.g. raking) had disturbed the surface of some sections of the site. Excavation focused on a relative concentration of shells and stone artefacts located across a 4 m × 4 m area away from bird disturbance. A 3 m × 1 m pit comprising three contiguous 1 m × 1 m squares (Squares H10, H11 and H12) orientated at 310° (long axis of pit) was excavated by IM and Tosi (TJ) Cora Jnr, Joseph Cora and Carl Porter on 14–15 September 2002.

The shallow cultural deposit at Site 10 was sampled differently within each of the three excavation squares (Supporting Tables S6–S8). Square H11 was excavated down to a maximum depth of 12.9 cm (north-west corner) using four XUs. Squares H10 and H12 were excavated to increase the sample of near-surface cultural materials (mainly shells and stone artefacts). Thus, Square H10 was excavated down to a maximum depth of 3.0 cm (centre) using a single XU, while Square H12 was excavated down to a maximum depth of 3.4 cm (north-west corner) using two XUs. Excavation, sieving and backfilling methodology was the same as that employed at Otterbourne Island Site 4. Numerous charcoal fragments and stone artefacts were plotted in 3D and bagged separately to increase excavation precision. A total of 189.0 kg of deposit with a volume of 174.0 L was excavated.

Stratigraphy and sedimentary environment

The stratigraphy of all three squares was basic and consisted of unconsolidated, colluvial sediments with numerous rock inclusions and scattered fibrous roots. Sediments became lighter in colour with depth: very dark greyish brown 10YR 3/2 (XU1), dark greyish brown 10YR 4/2 (XUs 2 and 3) and greyish brown 10YR 5/2 (XU4). The density of cultural materials (e.g. marine shells and stone artefacts) decreases dramatically downwards over a vertical distance of 5 cm. Apart from marine shells and stone artefacts, other materials recovered from the site included charcoal, bone, crustacean shell, coral and red ochre. Neither land snail nor pumice was observed. The pH of Square H11 was acidic and ranged from 6.0 to 5.0 with depth (Supporting Table S7), suggesting that preservation conditions for shell and bone are poor. No evidence of disturbance to sediments (e.g. burrows and/or insects) was observed during excavation, apart from minor fibrous root infiltration.

Charcoal. A total of 16.9 g of charcoal was recovered from Square H11, with the highest densities occurring in XUs 2 and 3, located at a depth of between 2.6 and 5.3 cm below the surface (Supporting Table S7 and Table 4).

Table 4. Density of cultural materials, Square H11, High Peak Island Site 10
XUMarine shellStone artefactsCharcoal
(g/10 L of deposit)(number/10 L of deposit)(g/10 L of deposit)
1222.144.830.59
220.171.292.58
33.132.352.25
40.620.181.34

Radiocarbon dating and chronology

Four radiocarbon dates were obtained for Site 10 (Table 5 and Figure 7). Dating was undertaken by the University of Waikato Radiocarbon Dating Laboratory and calibrated as for dates for Otterbourne Island Site 4. Dates fall into two broad chronological groups – a recent group of three dates between c.200 and 500 calBP and an older group with a single date of c.3250 calBP. Dating inversions amongst the dates is not unexpected, given that the dated samples were obtained within 6.6 cm of the ground surface. It appears that most of the cultural deposit, including all of the midden shell, dates to the past 500 years and rests directly upon earlier evidence of activity represented solely by stone artefacts and charcoal dating to around 3250 years ago. Treadage, post-depositional disturbance and slow sedimentation has resulted in minor mixing of the two periods of occupation.

Table 5. Radiocarbon dates for Square H11, High Peak Island Site 10
Lab. no.XUDepth (cm)SampleMethodδ13C‰ (± 0.2)Conventional 14C age (years BP)Calibrated age BP 68.3% (with probabilities)Calibrated age BP 95.4% (with probabilities)Median probability calBP
Wk-1463810–2.6

Oyster

Saccostrea

cucullata

Fragments

36.73 g

Standard2.4599 ± 30

148–162 (0.120)

192–213 (0.154)

222–280 (0.726)

131–294 (1.000)233
Wk-3224424.3

Wood

Charcoal

1 fragment

(3D Find 3)

0.07 g

AMS−25.3187 ± 25

0–23 (0.140)

72–83 (0.083)

104–112 (0.052)

140–154 (0.124)

170–181 (0.070)

186–195 (0.052)

205–228 (0.208)

250–279 (0.273)

0–29 (0.117)

59–120 (0.208)

135–158 (0.112)

163–234 (0.315)

236–283 (0.249)

178
Wk-1463935.1

Wood

Charcoal

1 fragment

(3D Find 1)

0.10 g

AMS−24.33100 ± 403212–3337 (1.000)

3081–3092 (0.016)

3111–3125 (0.015)

3140–3366 (0.969)

3262
Wk-3224536.6

Wood

charcoal

Multiple Fragments

(3D Find 4)

0.32 g

AMS−24.5490 ± 25496–515 (1.000)

475–479 (0.009)

486–532 (0.991)

506

Occupational hiatus c.500–3300 years ago

Radiocarbon dating of Site 10 reveals a major depositional and occupational hiatus of 2800 years between c.500 and c.3300 years ago (Figure 7). Indeed, the jump in dates was recorded in a single XU (XU3 of Square H11). While it is possible such a hiatus represents the result of erosion of intervening deposits, the lack of lag deposits would suggest the hiatus does indeed reflect a period of 2800 years when little or no sediments and cultural materials were deposited at the site.

Marine fauna

The only bone recovered from the excavations was a single fish incisor from XU1 of Square H12. Small amounts of crustacean exoskeleton (0.22 g), barnacle shell (0.34 g) and coral (6.15 g) were recovered from Squares H10 and H11. While the fish and crustacean exoskeleton remains probably represent food remains, the barnacle shell most likely entered the site inadvertently attached to molluscs, while coral fragments represent manuports of unknown use.

Shellfish. Nearly all (94%) of marine shell in Square H11 was recovered from XU1 within 2.6 cm of the ground surface, with small quantities continuing down to XU4 at a depth of 10.7 cm (Supporting Table S7). It is likely that most, if not all, of the shells in XUs 3 and 4 of Square H11 were displaced vertically from XUs 1 and 2, given that all of the diagnostic MNI shell fragments recovered from Square H11 came from XU1. The range of shellfish recovered from all three excavation squares was small, represented mostly (84%) by oyster (Saccostrea cullculata) (MNI = 49), followed by nerite (Nerita sp.) (MNI = 5), chiton (MNI = 3) and Morula sp. (MNI = 1). All of these species could have been obtained from rocky substrates within the intertidal zone of High Peak Island.

Material culture

The only material culture recovered from the site was flaked stone artefacts (n = 41, 155.7 g) in Squares H10, H11 and H12, and a 1.0 g fragment of ochre with a ground facet (used as a source of pigment) recovered from XU4 of Square H11. Most (n = 14, 67%) stone artefacts in Square H11 were found in XU1 and it is possible that some of the small stone artefacts in XUs 2–4 have been vertically displaced downwards. However, the two largest stone artefacts excavated from the site, both weighing 29.7 g and measuring at least 5 cm in maximum length, were found lying horizontal and at an angle of ∼45°, at depths of 6.4 and 7.7 cm, respectively, in XU3, implying little or no post-depositional movement. Most artefacts were manufactured from volcanic raw materials (n = 26, 63%), followed by quartz (n = 15, 37%). Fracture types included flakes (n = 35, 85%), retouched flakes (n = 4, 10%) and cores (n = 2, 5%). All retouched flakes were volcanic and both cores were quartz and exhibited bipolar reduction. The mean size of volcanic flakes (3.0 g) was larger than that of quartz flakes (1.8 g).

Summary

High Peak Island Site 10 is a sparse and shallow midden deposit, dating to the past 500 years, preceded by a long hiatus and an initial phase of occupation represented by stone artefacts dating to around 3250 years ago, occurring well after island formation at least 9000 years ago. Marine subsistence specialisation is indicated by the presence of fish, shellfish and crustacean remains.

Collins Island

  1. Top of page
  2. Abstract
  3. Introduction
  4. Shoalwater Bay Islands
  5. Otterbourne Island
  6. High Peak Island
  7. Collins Island
  8. Discussion
  9. Conclusion
  10. Acknowledgements
  11. References
  12. Supporting Information

Collins Island is located 14.l km from the mainland, but can be accessed by island-hopping through the Skull Group. However, the final leg from Osborne Island still involves a sea crossing of 8.3 km. The island is the largest in the Cannibal Group and has a maximum length of 4.8 km and width of 1.3 km. Like all of the Cannibal Group, the island is made up of Carboniferous ocean-floor sedimentary rocks such as quartz greywacke and mudstone (see ABARE et al. 1993: 22–23, 161; Grant et al. 1979: 34; Murray 1975). These rocks appear to have been metamorphosed, a suggestion consistent with numerous large quartz veins observed on the southern and eastern coasts during the 1998 archaeological survey. On the south coast, milky quartz veins up to l m wide were observed associated with blocks of quartz up to 50 cm across. At the northern end of the east coast, a huge milky quartz vein, around 2 m thick, is exposed across the ground and down onto the beach and out into the sea (Figure 11). Vegetation consists mostly of stunted eucalypt woodland and Acacia shrubland. Casuarina and pandanus trees, and Grass trees (Xanthorrhoea sp.), lantana and prickly pear bushes are found around the coastline. A bêche-de-mer station operated on the island in the early 1870s (Rogers 1886–1887). In recent times, “Cannibal Seafoods” established an oyster processing and aquaculture (barramundi) operation within a lease across the north-west section of the island. Over 90% of the island appears to have been little modified by Europeans.

figure

Figure 11. Collins Island Site 24, looking south from the large quartz vein, September 2002 (photo: Ian J. McNiven).

Download figure to PowerPoint

Site 24 excavations

This site comprises extensive surface scatters of shells and stone artefacts and in-situ midden deposit extending for some 800 m along the entire northern two-thirds of the east coast of Collins Island (Figure 11). It was recorded by IM and Darumbal representative Bob Muir on 29 April 1998. The observable width of the site decreases northwards, from 30 m in the south to 5–10 m in the north. This change in width simply reflects decreasing surface erosion northwards along the coast. In the south, wind and rain have exposed large areas of ground surface, almost producing a “badlands” terrain. In the north, shrubland vegetation has maintained the integrity of the topsoil, which has resulted in less erosion of cultural deposits. As a result, the largest surface scatters of oyster shell and stone artefacts can be seen across the southern sections of the site. Here, quartz knapping floors with up to 30 artefacts/m2 were recorded. Other artefacts include milky quartz cobble cores and cobble manuports. Similar cobbles can be found along the adjacent beach. Erosion associated with the shoreline cliff face has exposed in-situ deposit along the northern 350 m of the site. This deposit is dominated by oyster shells (with a few chitons) and contains a high density of milky quartz artefacts. The dramatic increase in artefact density in this northern area is attributed to proximity to a large, 2 m wide milky quartz vein, which cuts through the site in this area. The vein starts well inland and erosion of surrounding softer sediments has left the vein exposed as a spectacular quartz “wall” running through the surf and into the sea (Figure 11). It is likely that raw material to make many of the quartz artefacts was quarried either from this vein or, more likely, collected as cobbles from the beach.

A 1 m × 1 m pit (Square A) was excavated 25 cm in from the 2.4 m high shoreline erosion face at a point located 64 m south of the huge quartz vein (Figures 11 and 12). The excavation area exhibits a range of low trees (mostly casuarinas and pandanus), shrubs (e.g. acacias and lantana) and groundcover (grass and pigface). The surface of Square A was covered in a 2–3 cm thick layer of dead casuarina “needles”. The pit was located adjacent to where a concentration of in-situ cultural materials (shells and stone artefacts) was exposed in section. It was positioned parallel to the cliff-line and orientated at an angle of 120°. The aim of the excavation was to obtain a stratified sample of cultural materials to provide insights in the antiquity of quarrying activity and associated artefact reduction at the site. Excavation, sieving and backfilling methodology was the same as that employed at Otterbourne Island Site 4. Square A was excavated down to a maximum depth of 71.2 cm (centre of pit), using 22 XUs with a mean thickness of 3.2 cm. Excavation ceased due to an apparent lack of cultural materials. A number of charcoal fragments and stone artefacts was plotted in 3D and bagged separately. A total of 1037.6 kg of deposit with a volume of 833.0 L was excavated. Excavations were undertaken by IM with Darumbal representatives Billy Mann, Malcolm Mann and Andrew (AJ) Batchelor during 3–7 September 2002.

figure

Figure 12. A cross-section through Square A, Collins Island Site 24.

Download figure to PowerPoint

Stratigraphy and sedimentary environment

Three Stratigraphic Units (SUs) were identified in Square A (Figure 13). SU1 is very dark greyish brown (10YR 3/2), unconsolidated sandy loam that extends from the surface down to a depth of ∼25 cm. Most shells and stone artefacts within the pit came from this unit. SU1 also contains numerous roots, especially in the upper 10 cm. It is homogenous in colour and texture, and the transition to SU2 is a little difficult to differentiate in places. SU2 is dark greyish brown (10YR 4/2) sandy sediment that extends down to ∼55 cm below the surface. It is slightly more consolidated than SU1 and contains far fewer roots (most of which are scattered fibrous roots) and few cultural materials. SU2 has a more gritty texture than SU1 and the lower sections appear to hold more moisture, which may relate to an apparent increase in clay content. The interface with SU3 is undulating and easily differentiated. SU3 extends across the base of the pit and is dark greyish brown (10YR 4/2), consolidated gritty clayey sand with a few scattered fibrous roots. The clay content of sediments in SU3 appears to increase with depth and increasing proximity to clay sediments underlying the pit. Clay sediments were encountered in the eastern corner of the square at the base of the pit. The amount of cultural materials in SU3 grades from minor to nil with depth. The pH ranged from 6.5 to 8.0 (neutral) (Supporting Table S9).

figure

Figure 13. Stratigraphy, Square A, Collins Island Site 24.

Download figure to PowerPoint

Land snails. A small amount (4.6 g) of land snail shell was recovered from XUs 1–6, with most (98%) in XUs 1–3 within 10 cm of the surface (Supporting Table S9). Absence of land snail shell in lower levels of the pit probably reflects lack of preservation of these thinly walled and fragile shells.

Pumice. A total of 243.2 g of pumice was recovered from XUs 1–18 (Supporting Table S9). Most (85%) pumice was restricted to XUs 1–15 within 45 cm of the surface, with highest densities within XUs 8–15, located between 19 and 45 cm below the surface (Figure 14).

figure

Figure 14. The density of cultural materials, Square A, Collins Island Site 24.

Download figure to PowerPoint

Rocks. A considerable quantity (14.2 kg) of rocks was recovered from Square A (Supporting Table S9). All of the different types of rocks recovered from the pit are available either across the adjacent beach and intertidal zone or naturally within the terrestrial sediments of Collins Island (e.g. lateritic pebbles). While many of the numerous unmodified quartz pebbles appear to have washed into the site from local terrestrial sources, a number of unmodified quartz cobbles found with the upper 11 XUs appear to derive from adjacent beach deposits and are probably manuports. Density of rocks is highest in the upper six XUs within 18 cm of the surface, peaking in XU 5 (Figure 14).

Charcoal. Most (98%) of the 42.6 g of recovered charcoal came from XUs 1–11, located within 31 cm of the surface (Supporting Table S9). The highest densities were found within the upper 16 cm of deposit, especially in XUs 3 and 4, between 6 and 14 cm below the surface (Figure 14).

Taphonomy and preservation

Sediments in Square A appeared to demonstrate high stratigraphic integrity, with minimal evidence of post-depositional disturbance apart from scattered roots and a single burrow void (XU17). All large quartz artefacts were lying flat, indicating minimal movement since initial deposition. No obvious marine shell decay profile was observed with depth. This finding is consistent with the pH values, which were 7.0–7.5 (i.e. neutral to mildly alkaline) for most XUs (range 6.5–8.0).

Radiocarbon dating and chronology

Seven AMS radiocarbon dates were obtained for Site 24 (Table 6 and Figure 7). Dating was undertaken by the University of Waikato Radiocarbon Dating Laboratory and calibrated as for dates for Otterbourne Island Site 4. Six dates are restricted to the past 300 years, including two “modern” dates. The earliest date is c.550 calBP from XU9. Considerable vertical inconsistencies occur, with the dates suggesting movement of charcoal within the deposit and mixing of sediments. The “modern” dates indicate post-depositional intrusion of charcoal into the deposit during the twentieth century. While it is possible that slightly older charcoal simply entered the site through disturbance, the absence of dates older than c.550 calBP suggests that the deposit does indeed date mostly to the past 600 years or so.

Table 6. AMS radiocarbon dates for Square A, Collins Island Site 24
Lab. no.XUDepth (cm)Sampleδ13C‰Conventional 14C age (years BP)Calibrated age BP 68.3% (with probabilities)Calibrated age BP 95.4% (with probabilities)Median probability calBP
Wk-3225922.1–5.7

Wood

Charcoal

1 fragment

0.06 g

−25.0 ± 0.2116 ± 29

0–0 (0.015)

25–71 (0.497)

84–87 (0.021)

93–103 (0.068)

113–140 (0.285)

229–242 (0.114)

0–143 (0.855)

225–254 (0.145)

87
Wk-3226049.6–13.4

Wood

Charcoal

1 fragment

0.54 g

−25.7 ± 3.1209 ± 26

147–158 (0.117)

163–220 (0.700)

266–282 (0.183)

0–25 (0.040)

71–84 (0.023)

88–92 (0.003)

103–113 (0.015)

140–229 (0.667)

244–300 (0.251)

194
Wk-32261513.4–15.7

Wood

Charcoal

1 fragment

0.03 g

−19.9 ± 3.2162 ± 26

0–27 (0.226)

60–63 (0.020)

66–114 (0.389)

137–147 (0.071)

221–232 (0.086)

238–265 (0.208)

0–151 (0.704)

214–274 (0.296)

109
Wk-32262717.7–19.3

Wood

Charcoal

1 fragment

0.05 g

−25.9 ± 0.2100.2 ± 0.1%ModernModernModern
Wk-14578921.8–25.2

Wood

Charcoal

Multiple fragments

0.14 g

−25.1 ± 0.2613 ± 38

537–563 (0.539)

603–628 (0.461)

518–572 (0.532)

588–645 (0.468)

568
Wk-322631127.8–30.7

Wood

Charcoal

1 fragment

0.07 g

−27.1 ± 0.2100.4 ± 0.1%ModernModernModern
Wk-145791852.5–55.2

Wood

Charcoal

Multiple fragments

(3D Find 1)

0.10 g

−25.8 ± 0.2105 ± 40

0–0 (0.027)

25–72 (0.459)

84–104 (0.121)

112–140 (0.267)

229–244 (0.126)

0–145 (0.831)

222–261 (0.169)

91

Marine vertebrates

Only a small amount of bone (5.6 g) was recovered from Square A (Supporting Table S9). It was restricted to XUs 1–7, within 20 cm of the surface, and was represented by “large vertebrate” (4.9 g), “small vertebrate” (0.3 g), turtle (0.3 g) and fish (0.3 g). It is likely that most, if not all, of the “large vertebrate” bone represents turtle and that most, if not all, of the “small vertebrate” bone represents fish.

Marine invertebrates

Marine invertebrate remains include molluscs, crustaceans, coral and foraminifera (Supporting Table S9). All (0.06 g) of the foraminifera discs were recovered from XUs 1–3, within 10 cm of the surface, and most likely blew into the site from nearby beach deposits. Coral fragments (188.2 g) were restricted to XUs 1–7, within 20 cm of the surface, and most represented small water-rolled fragments that also probably blew into the site from nearby beach deposits. One large (157.1 g) fragment of coral in XU5 clearly was brought to the site by people. Barnacles (1.2 g) were the only crustacean remains recorded and all were found in XUs 5–9, between 13 and 26 cm below the surface.

Shellfish. The marine shell assemblage included numerous water-worn pieces of shell, both small fragments and small whole shells (< 1 cm long), that clearly derived from adjacent beach deposits and most likely entered the site naturally through wind action. As such, shell analysis was restricted to non-water-rolled shells (whole and fragments) at least 15 mm in length that represented shells considered to have been collected by people for food. Only shell fragments with Diagnostic Anatomical Elements (DAE) for MNI analysis (see above for methodology) were investigated. As such, no marine shell weights are provided and vertical changes in marine shells are limited to MNI data.

A total MNI of 118 shells representing six taxa was identified from Square A (Table 7). DAE fragments were found consistently in XUs 1–9, with a single DAE (chiton) outlier in XU14. Thus, nearly all DAEs were recovered within 26 cm of the surface. The highest MNI values were found in XUs 4–6, located at a depth of 13–18 cm below the surface. In terms of numerical significance, shellfishing focused on chitons (MNI = 70, 59%) and oysters (MNI = 40, 34%). All of the identified shellfish taxa could have been obtained easily from rocky substrates and associated coral reefs within the intertidal zone of Collins Island. Indeed, oysters and chitons were the two major shellfish observed across the intertidal zone adjacent to the site during excavation.

Table 7. Marine shell taxa MNI, Square A, Collins Island Site 24
TaxonTidal zoneSubstrateXUTotal MNI
12345678914
Bivalves             

 Rock oyster

Saccostrea cucullata

IntertidalRocks + mangroves3247434103 40

 Pearl shell

Pinctada sp.

IntertidalRocks + coral  121     4
Type A     11     2
Type B      1     1
Gastropods             

 Turbo

Turbo spp.

Intertidal + subtidalRocks + coral reefs     1    1
Chiton             
PolyplacopheraIntertidal + subtidalRocks1418391033 170
Total MNI  46618461471331118
Number of taxa  2234532211 

Material culture

Stone artefacts. A total of 1066 flaked stone artefacts weighing 2702.8 g was recovered from Square A (Supporting Table S9). Artefacts range widely in size in terms of maximum length (3.6–104.5 mm) and weight (< 0.01–579.8 g), with most artefacts generally small (mean maximum length = 10.5 mm, mean weight = 2.5 g). All artefacts are made from quartz, which clearly reflects the proximity of the nearby quartz vein and associated beach cobble deposits. Most artefacts are flakes (n = 992, 93.1%), followed by flaked pieces (n = 53, 5.0%), cores (n = 15, 1.4%) and retouched flakes (n = 6, 0.6%). The low representation of cores is unusual for a site associated with quarrying of tool stone and may simply reflect sampling (e.g. the location of core reduction centres elsewhere along the extensive site) and/or removal of cores to other sites in the region. While some of the artefacts within Square A may have been made from quartz quarried from the huge vein, the presence of water-rolled cortex on 150 artefacts (14.1% of the total assemblage) indicates that cobble deposits located across the adjacent beach were more likely targeted as a source of quartz. Indeed, 10 of the 15 cores exhibited water-rolled cortex indicating the importance of beach cobble deposits. The use of cobbles may relate to the fact that water-worn cobbles often exhibit fewer hidden flaws compared to bedrock outcrops (Dickson 1981: 108). Nearly all artefacts appear to have been created through freehand percussion flaking, with only 11 flakes and six cores exhibiting bipolar reduction. The density of stone artefacts increases through the sequence and is highest in XUs 4–6 (Figure 14).

Ochre. As considerable ochreous fragments occur naturally within extensive local gravels deposits, only ochre fragments with a ground facet were considered to be cultural. No such fragments were identified.

Summary

Collins Island Site 24 is a low-density midden deposit, containing numerous quartz stone artefacts dating to the past 600 years – and well after island formation 8000–9000 years ago. Turtle and fish bone, along with shellfish remains, reveal subsistence practices focused on adjacent marine habitats. The unusually high amounts of quartz artefacts reflect quarrying of nearby beach cobble deposits associated with a huge quartz vein.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Shoalwater Bay Islands
  5. Otterbourne Island
  6. High Peak Island
  7. Collins Island
  8. Discussion
  9. Conclusion
  10. Acknowledgements
  11. References
  12. Supporting Information

Island trip viability and seasonality: turtle hunting and egg collection

Ethnohistorical and archaeological evidence presented in this study supports the view that Collins, Otterbourne and High Peak Islands hosted short-term visits by marine specialists. Furthermore, the range of cultural materials recovered from excavations on all three islands is consistent with targeted, specialised visits. For example, most of the identified bone at all three islands was fish and turtle, while the number of different shellfish taxa was equally restricted: High Peak Island Site 10 (n = 4), Collins Islands Site 24 (n = 6) and Otterbourne Island Site 4 (n = 9). The limited range of shellfish taxa exploited by visitors to Shoalwater Bay islands is emphasised when compared to the range of shellfish exploited by visitors to other islands in the region: over 30 taxa on Marble Island (Border 1985: 85) and over 22 taxa on the Keppel Islands (Rowland 1996: 204). Furthermore, the range of shellfish taxa exploited in the Whitsunday Islands is also limited to some extent: at least 11 taxa at Nara Inlet 1 and Border Island 1, at least ten taxa at Hill Inlet Rock Shelter 1, and at least nine taxa at Nara Inlet Art Site (Barker 2004: 69, 94, 109–112, 120–123). Fish remains were similarly restricted in terms of overall quantities and number of taxa. Most fish remains were recovered from Otterbourne Island Site 4, where six taxa were identified and fishing focused on small individuals (< 10 cm in length) and a single taxon (wrasse). This situation contrasts with higher quantities of larger individuals from ten taxa identified from Mazie Bay (Keppel Islands) to the south (Hermes 1984) and the higher quantities of larger individuals from seven taxa indentified from Nara Inlet 1 (Whitsundays Islands) to the north (Barker 2004: 76), where more permanent island occupation has been inferred. For Shoalwater Bay islands, fish bone data points to more opportunistic fishing practices suggestive of more ephemeral visitation. Furthermore, it seems unlikely that people would take on the hazards and dangers associated with offshore canoe travel to Otterbourne and High Peak Islands simply to exploit a limited range and number of marine shellfish and fish. The calorific returns on these subsistence items would have been minimal.

While the minor amount of turtle bone recovered from Nara Inlet 1 and Hill Inlet Site and the 9.9 g of turtle bone found scattered through the Nara Inlet Art Site sequence all date to the past 3000 years, the considerable amount of turtle bone (105.6 g) excavated from Border Island 1 dates to the past 7000 years (Barker 1998: 92, 2004: 79, 99, 113, 124). Border (1999) also found turtle bone dating to the past 3000 years in middens on Marble and Curlew Islands, representing “about 90% of the weight of all bone” excavated from the Curlew Island rockshelter midden (Border 1999: 134–135). No turtle bone was excavated from St Bees and Middle Percy Islands, but it was observed on the surface of a midden on South Percy Island (Border 1991: 10, 1999: 135). In this context, the 2.4 kg of definite and probable turtle bone dating to between 100 and 5200 years ago recovered from Otterbourne Island Site 4 not only represents the best dated sequence of turtle bone known for the Queensland coast during the second half of the Holocene, but also the highest quantity of turtle bone excavated from a central Queensland island site.

In marked contrast to shellfish and fish, the calorific rewards from hunting marine turtles are considerable. For example, around 45% of an averaged weight green turtle of 110–160 kg is edible meat, fat and organs (Nietschmann & Nietschmann 1981: 61). In addition, a single green turtle nest can yield over 100 eggs (Limpus 2008: 21). In this sense, shellfish and fish food remains pale into insignificance compared to the calorific windfall from hunting turtles and gathering turtle eggs. It is in this context that the location of numerous sites on both Otterbourne and High Peak Islands in elevated contexts on ridge tops overlooking the sea may be explained as turtle lookouts, similar to those documented ethnographically in Torres Strait (e.g. McIntyre-Tamwoy & Harrison 2004; Moore 1979: 153, 272).

Turtle hunting would have helped ensure the subsistence viability of offshore voyaging, especially if it was scheduled to occur at times of the year when turtles are much easier to see and hunt, such as during calm seas or during the turtle mating/breeding season, which in the southern Great Barrier Reef region is between mid-September and mid-November (Border 1999; Limpus 2008: 19). Similarly, female turtles tend to lay eggs in the region between mid-October and early April, with an average of five clutches per season per female (Limpus 2008: 19, 21). The tendency of female turtles to reuse turtle nesting sites would increase the predictability of successful turtle egg procurement during offshore voyaging (see Limpus 2008: 16). Thus, visits to offshore islands may have been scheduled to take place when turtles and turtle eggs were most accessible between mid-September and early April, as they provided a high return and reliable food resource. Significantly, Border (1999: 133) found that early observations of Aboriginal people on offshore islands of the Northumberland Group were mostly between September and June, which overlaps considerably with the seasonal availability of turtle resources. Indeed, Border (1985: 145) suggests that the island visits were scheduled to take advantage of “a marked increase in the availability of the turtle and bird resources on or around the islands”. The scheduling of island visits during the summer “wet” season also ensured reliable supplies of drinking water.

Testing the hypothesis for scheduling of island visits to between September and April to take advantage of turtle resources is not aided using the temporal availability of plant foods such as Pandanus fruits and Cyperus tubers recovered from Otterbourne Island Site 4. For example, P. tectorius and P. brookei, the likely species from Site 4, can produce fruit for all or most of the year (Wilson 2011: 227, 229). Similarly, Thomson et al. (2006: 10) note that while “the main fruiting season [for P. tectorius] is April–August” in northern Australia, it “may fruit occasionally during the off-seasons throughout the year”. In contrast, Meehan et al. (1979: 81–82) observed that Pandanus spiralis is exploited seasonally by the Gidjingali of Arnhem Land generally between July and October when the fruits (drupes) are edible. This species does not occur on the central Queensland coast. No information is available on the potential seasonal availability of Cyperus spp. on the central Queensland coast.

Island trip motivations: quartz quarrying and ceremonies

The fact that the mainland coast of Shoalwater Bay was much more productive in terms of easily procurable turtles suggests strongly that turtles were not the only reason to voyage offshore. The risk involved in exploiting marine resources on islands, especially offshore islands such as Otterbourne and High Peak, was far greater than the risk involved in exploiting similar marine resources along the immediate fringes of the mainland coast. As Rowland (1984: 28) aptly remarked in relation to Aboriginal voyaging to the Percy Islands, located some 60 km off the mainland coast, “it is difficult to conceive of any simple explanation as to why these islands were exploited when getting to them must have been a relatively hazardous and lengthy event”. As such, we posit that while turtle hunting and gathering of turtle eggs were the principal economic pursuits underpinning the subsistence viability and scheduling of offshore voyaging in the Shoalwater Bay region, non-subsistence-related activities may have been of equal or greater importance as motivating factors behind island visits. We hypothesise that beyond subsistence, an important motivating factor may have been social and perhaps related to taking advantage of the remoteness of islands where certain activities, such as ceremonies, could have been undertaken. The privileging of social over subsistence factors as the key reason for island visitation has been documented in other Australian contexts (McNiven 2000: see also Bowdler 1995: 954; Mills 1992: 38). That a ritual dimension may been associated with offshore canoe voyaging in the Shoalwater Bay region has been hypothesised by McNiven (2003: 344) on the basis of stone arrangements located across the intertidal zone of the adjacent mainland coast.

Collins Island reveals that specialised resource exploitation was also a motivational factor behind visiting inshore islands. Site 24 on Collins Island is the largest quarry and stone artefact site recorded in the region (including the adjacent Shoalwater Bay mainland) (McNiven 1999a). However, the east coast of Collins Island does not appear to offer any special features in terms of food resources. The rock platforms exposed at low tide and the waters fronting the site are similar to those found across many parts of Shoalwater Bay. The only major difference is the existence of the single largest milky quartz vein recorded to date in the region. As milky quartz is the most widespread stone tool raw material used across the Shoalwater Bay region (McNiven 1999a; McNiven & Russell 1997), it can be expected that the Collins Island quartz vein and associated beach cobble deposits represent an extremely important regional source of tool stone. Indeed, the site represents one of the largest known quarries with an associated reduction site on an Australian island and one of the largest milky quartz quarries recorded for any part of the continent (see Hiscock & Mitchell 1993). It would appear that Aboriginal use of the east coast of Collins Island was motivated primarily by a desire to exploit milky quartz to make into tools that were subsequently taken to other parts of the region, both islands and mainland. In this sense, the Collins Island quarry site has similarities to South Molle Island in the Whitsundays, where a huge basaltic quarry site attests to large-scale quarrying for tool stone over at least 9000 years (Barker 2004: 131; Barker & Schon 1994; Bryan et al. 2000; Lamb 2011; Lamb & Barker 2001).

Tool stone quarry sites have also been recorded on other islands in the Northumberland Group. Quartz quarries associated with quartz veins have also been recorded on Clara and Mumford North Islands, off Shoalwater Bay (McNiven 1999a), and to the north on Marble Island (Border 1985: 71). The small size of all these quarries suggests strongly that they were associated with provisioning visitors to these islands. In marked contrast, huge quarry sites such as established on Collins Island and South Molle Island provided tool stone for peoples using other islands and even peoples on the nearby mainland (for discussion of the use of South Molle Island stone across the Cumberland Island Group and adjacent mainland coast, see Barker 2004: 136). However, the existence of a series of quartz quarries along the southern mainland coast of Shoalwater Bay (McNiven & Russell 1997: 51) indicates that the Collins Island quarry may have primarily serviced peoples while visiting islands.

A separate Islander people?

In contrast to models of seasonal and ephemeral island use by mainland-focused peoples for southern Australia, Rowland (1996: 205) argues that “it is possible to suggest that a number of island based marine economies were operating within the Reef Province”. For example, following ethnographic evidence for a separate Keppel Islander group, he argues that the “range of shellfish, fish and sea mammals exploited on North and South Keppel and the size and permanency of hut structures implies the presence of a broad-based, permanent economic base from at least 3000 years ago” (Rowland 1980; Rowland 1996: 198; Rowland 2008: 92). Similarly, in the northern part of the central Queensland coast in the Whitsunday Islands group, Barker (2004: 148) argues that within the past 3000 years, “the total evidence suggests a trajectory that was followed to the status of separate ‘tribal’ entity, the archaeology revealing increased economic productivity, site patterning and use, a greater degree of marine specialisation and unique art styles, while the ethnohistory indicates the eventual development of possible linguistic differences and the identification of the island people as a permanent population”. Largely following ethnographic information, Border (1999: 131) similarly posits a separate “island people” for the Northumberland Group, taking in the islands of Shoalwater Bay. Thus, all three researchers concur that the islands of the central Queensland coast were used by island specialists whose territories focused on islands much more than the mainland. Furthermore, evidence exists that some islands, such as Townshend Island in Shoalwater Bay and the Keppels, maintained a permanent presence by separate and distinct named communities.

We argue that ethnographic and archaeological evidence does not support the view that the islands of Shoalwater Bay supported a distinct and separate Islander community. In contrast, we hypothesise that the islands were used by peoples whose territory also incorporated the adjacent mainland of Shoalwater Bay. Indeed, Border (1999: 131, 136) pointed out that in 1855 Captain Chimmo of the HMS Torch recorded that a group of 19 Aboriginal men, women and children who had been living on Middle Percy Island for at least three and a half months included people from “near” Port Clinton in the Shoalwater Bay region. Border (1999: 131) rightly concluded that these people “clearly … either had close connections with or were from the nearby mainland”.

While visits to the offshore islands of Shoalwater Bay were likely to have been short-term and seasonal, most inshore islands were likely to have been visited throughout the year. Indeed, Townshend Island, located 1.4 km from the mainland, was permanently occupied by the residential Rundubura patriclan. In this sense, use of the islands may have been analogous to use of the mainland, with different locations the focus of activities at different times of the year. Whereas on the mainland movement between different locales across the landscape was by foot, across the adjacent sea movement between different island locales across the seascape was by canoe. However, unlike mainland occupation and pedestrian mobility, movement across the Shoalwater Bay seascape by canoe was always difficult and potentially dangerous. The decision to voyage out to these islands to exploit certain resources and undertake various other activities always involved extra costs and risks compared to similar activities along the mainland coast. These extra costs and risks have shaped island use over the past 5000 years.

Island colonisation phase: >5000–3000 years ago

On the basis of evidence for >8000 BP marine resource use at Nara Inlet 1, Barker (1991: 106) argued that “people of the Whitsunday area were marine ‘adapted’ before the site was occupied”. Furthermore, use of marine resources in the lower levels of Nara Inlet 1 between at least 8150 (c.9000 calBP) and 6500 BP (c.7400 calBP) indicates that Aboriginal people had “successfully endured nearly 2000 years of continuous marine transgression in which time the sea level rose some 17.5 m and cut the site off from the mainland” (Barker 1991: 106). Barker (1991: 107) concluded that at least in the Whitsunday region, “changing sea level apparently had no significant effect on availability of, or access to, marine resources”. More recently, Barker (2004: 45) commented that “people on the coast were always coastal people, and that they moved easily and quickly and comfortably with changing coastline, exploiting quickly established and readily available marine resources”. These views were similarly taken up by Rowland (1996: 205), who suggested that “coastal resources in general whether on mainlands or islands continued to be used throughout the Pleistocene and Holocene”. He noted that the Nara Inlet 1 sequence “demonstrates that coastal areas were being used prior to the phase of major island development” (Rowland 1996: 195).

Otterbourne Island 4 reveals minor evidence of occupation prior to 5200 calBP. The fact that the site was used systematically between 5200 and 100 calBP suggests that it was a prime settlement location. As such, it would be expected that if people were using the island in a sustained and systemic way prior to 5200 calBP, then evidence of such occupation would have been recovered from Otterbourne Island 4. As such, it is hypothesised that a temporal gap of up to 3000 years may exist between earliest known major occupation of the island and likely island formation around 8000 years ago. Clearly, further excavations are required at Site 4 to identify basal cultural materials and dates for the commencement of occupation.

Rowland (1996: 197) hypothesised that as a result of mid-Holocene higher sea levels, “evidence of coastal occupation of the islands prior to 4500 years might now be lost”. This hypothesis was used by Rowland (1996: 197) to account for the absence of materials older than 5000 years at Mazie Bay on North Keppel Island. The plausibility of this hypothesis is weakened by recent research, which indicates that sea levels have been falling to modern levels over the past c.7000 years. Indeed an argument can be presented that falling sea levels would enhance the potential for preservation of mid-Holocene coastal sites. Furthermore, Barker (2004: 15) rightly pointed out that as coastal rock shelters are far less susceptible to coastal erosion processes, they provide a controlled context to test for differential preservation of materials through time. Although largely an open site, Otterbourne Island 4 similarly provides a test for the differential preservation hypothesis, given that it is elevated steeply some 60 m above the adjacent shoreline and well beyond the potential impact of erosion from storm surges and the like. As such, the >5200-year-old sequence at the site is not considered to be an artefact of erosion and selective preservation.

Beaton (1985) advanced a highly influential hypothesis that the “time-lag” between arrival of the sea to modern levels and deposition of marine resources in many coastal sites was due to delayed development of local marine habitats and associated resources. His “time-lag” model was based on excavations at Princess Charlotte Bay in north Queensland, where a temporal gap of “1500 years” was found between the earliest midden deposit dated to 4760 ± 90 BP (marine shell date) and the suggested “end of the marine transgression” around “6000 years ago” (Beaton 1985: 5). As the earliest date was obtained from a rock shelter, Beaton (1985: 9, 12) rightly concluded that the absence of pre-4760 BP deposits was not due to coastal erosion or in-situ decay. Alternatively, Beaton (1985: 12) argued that the lack of occupational evidence “is probably only realistically explained by allowing that there were no people living in the vicinity” until around 4760 BP. The reason the region was “unattractive” for occupation was linked to “less productive” marine resources reflecting dynamic coastlines associated with the final stages of the marine transgression and delayed development of relevant marine habitats following sea-level stabilisation (Beaton 1985: 12–13).

Is Beaton's (1985) “time-lag” hypothesis relevant to the issue of delayed marine resource use of Otterbourne Island? In terms of central Queensland islands, both Barker (1991) and Rowland (1996) point out that Nara Inlet 1 reveals that the “time-lag” hypothesis is not applicable and that marine resources are much more resilient than proposed by Beaton. Yet Beaton (1985: figure 6) acknowledged that in some areas of Australia middens do date back to – and, indeed, date to immediately before – the time of so-called sea-level stabilisation around 6000 BP. While the notion of sea-level stabilisation at 6000 BP has been superseded by recent research showing that the sea level continued to rise after reaching modern levels ∼8000 calBP and only fell back to modern levels within the past 2000 years (see above), the hypothesis of delayed marine resource development remains relevant. That inter- and intra-regional differences exist in the applicability of “time-lag” issues is also revealed by Barker's (2004) Whitsunday Islands excavation data. While Nara Inlet 1 reveals no evidence of occupational “time-lag”, <15 km away at Border Island 1 rock shelter, located 40 m up a “near-sheer cliff” above the sea within Cateran Bay, a “near basal” date of 6440 ± 90 BP (c.7300 calBP) was obtained, which is nearly 2000 years after the island attained its modern configuration around 9000 calBP (Barker 2004: 54, 107; Lamb & Barker 2001). To account for this temporal discrepancy, Barker (2004: 115) suggested that “occupation may not have occurred until development of extensive fringing reefs in Cateran Bay”. This hypothesis is curious in light of Barker's broader claims for minimal impact of sea-level change and marine resource availability on occupation trends in the region. For example, the hypothesis is not relevant to the shellfish assemblage from the site, which is dominated by rocky platform shellfish (e.g. nerites, turbans, top shells, oysters and chitons), considered highly adaptable motile species that adjusted quickly to sea-level change (Barker 2004: 13).

Perhaps more relevant to explaining delayed use of Border Island is marine turtle (most likely green turtle), which not only occurs throughout the Border Island 1 sequence but was a key prey species, in contrast to shellfish. As the preferred habitat for green turtles is fringing reefs, it is likely that the importance of the development of reefs at Border Island in terms of human colonisation/visitation thresholds was linked to green turtle biomass. We suggest that a similar hypothesis is relevant to understanding early use of Otterbourne Island. That is, while certain food resources such as rocky platform shellfish were readily available prior to 5000 years ago, the key subsistence item – green turtles – was not readily available due to poorly developed fringing reefs. As such, we argue that the green turtle biomass of Otterbourne Island only developed to the extent that it was considered viable for systematic exploitation by local Aboriginal people around 5000 years ago.

The linking of the delayed systematic use of Otterbourne Island to environmental change and increased resource abundance should not be seen in a deterministic sense. The viability threshold attached to decisions to visit Otterbourne Island and to hunt local green turtle stocks would have been socially defined and the economic equations of costs versus benefits socially determined. At this juncture, no evidence is available to indicate that decisions to travel 26 km from the mainland coast to Otterbourne Island were related to developments in canoe technology. Evidence of voyaging to Border Island over 7000 years ago and to the South Molle Island quarry over 9000 years ago reveals that sea voyaging canoes have been part of the regional technological repertoire for thousands of years prior to visiting Otterbourne Island. It is also likely that the decision to start visiting Otterbourne Island did not reflect improved sea conditions for offshore voyaging using canoes. Indeed, the frequency and amplitude of ENSO climatic events and associated climatic variability began intensifying around 5000 years ago (Donders et al. 2008; Gagan et al. 2004; Moy et al. 2002; Shulmeister & Lees 1995; Tudhope et al. 2001; Turney & Hobbs 2006), which if anything, would have made sea voyaging potentially more hazardous and complex in terms of scheduling and risk assessments (cf. Sim & Wallis 2008: 103). As the frequency of cyclones has remained reasonably constant over the past 5000 years (Hayne & Chappell 2001), it is likely that canoeing conditions and associated risks recorded ethnographically were similar to those encountered by the first visitors to Otterbourne Island.

Reef expansion phase: after 3500 years ago

Otterbourne Island Site 4 reveals a horizontal spread of midden deposit dating to c.3200 calBP with the establishment of the Square A sequence, while evidence for voyaging 40 km off the coast to High Peak Island was similarly dated to c.3250 calBP. While the quantities of cultural materials around 3000 years ago are minimal, their presence does suggest a spatial expansion of activities across the islands. This spatial expansion of activity matches with the research of Barker (2004), Border (1999) and Rowland (1999b), who similarly documented major changes taking place in the use of islands of the southern Great Barrier Reef region commencing around 3000–3500 years ago. For example, Mazie Bay, with occupation dating back to 4274 ± 94 BP (c.4400 calBP), is the only site excavated by Rowland on North and South Keppel Islands with an antiquity of greater than 1000 calBP (Figure 15 and Supporting Table S10). Rowland (1999b: 141–143) pointed out that a series of quantitative and qualitative changes took place at Mazie Bay midden after 3500 BP – an increase in fishing, a relative increase in fishing for whiting and Sparids, a relative decrease in wrasse fishing and turtle hunting, a decrease in the size of fish, and a relative increase in use of oysters and chitons compared to other shellfish. These subsistence changes coincided with a change in dune deposition at the site, leading Rowland (1999b: 148) to hypothesise that such changes may reflect independent evidence for stabilisation of sea levels to modern levels since 3500 BP.

figure

Figure 15. The occupation chronology for island sites, central Queensland (black bars, 95.4% span of calibrated radiocarbon dates; white bars, absence of radiocarbon dates).

Download figure to PowerPoint

Border (1999) provides preliminary results of excavations and radiocarbon dating of midden deposits on a series of islands in the Cumberland and Northumberland Groups north of Shoalwater Bay (St Bees Island, Curlew Island, Middle Percy Island and Marble Island). The earliest evidence for island use was a basal date of 3180 ± 150 BP (c.3300 calBP) from a rock shelter on St Bees Island (Figure 15 and Supporting Table S10). Although the dates reveal that occupation was confined mostly to the past 3000 years, Border (1999: 136) concedes that “pre-3000 BP occupation of the Mackay/Shoalwater Bay coast, and perhaps some of the islands, is a likely consideration”. Border (1999: 136) hypothesises that the apparent expansion of island occupation within the past 3000 years may coincide with movement of coastal peoples into the region or reflect delayed use of the islands (possibly due to delayed stabilisation of sea levels and development of coastal resources) by coastal peoples who had been in the region for at least 6000 years and, indeed, who had been following the transgressing Pleistocene shoreline westwards across the continental shelf.

Barker's (2004) excavations across the Whitsunday Islands of the Cumberland Group similarly revealed major changes taking place within the past c.3000 years (Figure 15 and Supporting Table S10). For example, basal dates of 2410 ± 80 BP (c.2400 calBP) and 2720 ± 120 BP (c.2800 calBP) were obtained for Nara Inlet Art Site and Hill Inlet 1 rock shelters, respectively. Similarly, at Nara Inlet 1 rock shelter, the bulk of cultural materials and highest discard rates (except for stone artefacts) took places in levels dating after an inferred date of c.3000 BP (c.3100 calBP) (i.e. located midway between dates of 2090 ± 50 BP and 3990 ± 60 BP: Barker 1999: 69). At the Border Island 1 rock shelter, a more complex pattern is evident, with higher discard rates for most cultural materials in lower levels dating 6000–7000 years ago, with increases in activity after 3260 ± 110 BP (c.3100 calBP) taking place following minimal site activity between 3000 and 6000 years ago. Qualitative changes within the post-2000 calBP levels of Nara Inlet 1 include the technological addition of bone points, shell scrapers and turtle shell artefacts, and the subsistence addition of green turtle, sand crab (Portunus pelagicus) and the mud/sand-dwelling shellfish Asaphis deflorata, Polymesoda erosa (cf. Gelonia coaxans) and Pinctada fucata (Barker 2004: 86–87). In terms of the shellfish additions, Barker (2004: 142) noted that “this may indicate a minor increase in local sedimentation and an expansion of macrophytic communities in Nara Inlet or nearby after this time or reflect changes in procurement choice relating to shellfish”.

Changes after 3000 years ago in the Whitsunday Islands indicate “a demographic restructuring of populations”, which involved “markedly increased regional and site use, including the exploitation of previously unoccupied islands and an intensification of the marine resource base”, according to Barker (2004: 148). Such restructuring was “brought about” by “changes in inter-regional social relations” that included a more “bounded” social system “in which territorial demarcation became more clearly defined and access to resource areas controlled or restricted” (Barker 2004: 150). Significantly, Barker (2004: 149) argued that cultural changes that took place more recently than 3000 years ago did not correlate with environmental change in the region and certainly did not reflect increased availability of marine resources. Indeed, he suggested that while “the full range of marine resources, including turtle and dugong, were available to coastal populations from a least the mid-Holocene … the opportunity to exploit the greater productivity of these resources was not widely acted upon until several thousands of years later” (Barker 2004: 149). While on theoretical grounds we concur with Barker (2004) that people do not respond mechanistically and deterministically to natural increases in bioproduction, we posit that increased use of islands of the southern Great Barrier Reef region commencing around 3000 years ago can be associated in part with new and emerging evidence for increased production and availability of key marine resources associated with development of coral reefs and mangrove systems (see Hiscock 2008: 170).

A recent summary of 176 core records and 211 associated radiometric dates taken from the Great Barrier Reef reveals two major pulses of reef development during the Holocene (Perry & Smithers 2011). In the southern Great Barrier Reef region, the two periods of reef initiation and accretion were constrained to ∼8700–5500 calBP and ∼3500 calBP to present (Perry & Smithers 2011: 79). For example, in the Shoalwater Bay region, coring of the fringing reef of High Peak Island reveals that it formed almost entirely between 7800 and 5800 calBP (Kleypas 1996; Perry & Smithers 2011: appendix 2). In contrast, coring at Middle Percy Island in the Northumberland Group suggested that most reef growth occurred within the past 3600 years (Kleypas 1996). During the mid-Holocene growth spurt, coral reefs in the southern Great Barrier Reef region grew vertically on average 5–9 mm/year (Perry & Smithers 2011: 81). The rapid response of early to mid-Holocene coral reefs to sea-level rise is consistent with historically recorded coral growth in association with modern sea-level rise on Heron Island (Scopélitis et al. 2011). The hiatus in reef growth between ∼5500 and 3500 calBP most likely reflected falling sea levels, with reactivation of growth within the past 3500 years reflecting stabilisation of the sea to current levels (Perry & Smithers 2011: 83). Significantly, most known examples of reef growth within the past 2300 years documented by Perry and Smithers (2011: figure 4) represent newly established reefs. As such, the spatial extent of coral reefs, and by extension reef-based resources (e.g. shellfish and fish), would have begun to increase considerably after 3500 years. Van Woesik (1992: 80, 121) suggested that “adverse conditions” for coral growth across the Northumberland Group “may have caused a time-delay in reef initiation (e.g. Middle Percy Island)”.

Insights into the developmental history of mangrove forests within the study area during the Holocene are limited by a lack pollen cores taken from mangrove sediments along the central Queensland coast. Remains of mud crab (Scylla serrata) in the lower levels of Nara Inlet 1 site in the Whitsunday Islands indicate that mangroves have been in the broader region for at least 9000 years (Barker 2004: 87). Wood from buried mangrove mud within Broad Sound has been dated to 6010 ± 240 BP and shows establishment of mangroves in the Shoalwater Bay region by at least c.6800 calBP (Burgis 1974: 35). However, it is unlikely that mangrove resources remained stable throughout the Holocene. A series of detailed studies of mangrove evolution in northern Australia reveals that following an early Holocene “Big Swamp” mangrove phase, mangrove forest decreased dramatically in area to modern levels variously after ∼6000 calBP (South Alligator River, Northern Territory: Woodroffe et al. 1988), ∼3100 calBP (Torres Strait, Queensland: Rowe 2007a, 2007b), ∼2700 calBP (Mulgrave River estuary, Queensland: Crowley 1996; Crowley & Gagan 1995) and ∼2000 calBP (Lizard Island: Proske & Haberle 2012). More locally, coring of Broad Sound inlet to the immediate north-west of Shoalwater Bay reveals a constant progradation of shorelines and movement of the mangrove zone progressively seaward over the past 7000 years (Burgis 1974: 35–36; Cook & Mayo 1977: 184–185; Cook & Polach 1973). To the south of Shoalwater Bay, coring of the lower Fitzroy River floodplain on the mainland opposite the Keppel Islands revealed extensive mangrove forests during the mid-Holocene, after which “mangroves have become more restricted to the margins of the creeks and main channel as fluvial sediments built up the floodplain, and have shifted seawards as the mouth of the estuary prograded into Keppel Bay” (Bostock et al. 2006: 49).

To what extent the spatial configuration and extent of mangroves within Shoalwater Bay similarly changed over the past 7000 years is difficult to determine. Certainly the area of mangroves within Shoalwater Bay is considerable (Figure 1) and it seems likely that the spatial extent of mangroves has expanded with prograding mudflats without the concomitant late Holocene replacement of mangroves with expansive saltflats on the landward side of mangrove forests, as seen in areas to the north (Broad Sound) and south (Fitzroy River). Similarly, it is also likely that prograding mudflats progressively increased the area of seagrass meadows. Thus, we hypothesise that expansion of coral reefs, mangroves and seagrass would have resulted in a dramatic increase in turtle, dugong, shellfish and fish resources in the study region over the past 3000 years. Such increases in marine resources would have been augmented by increased potential rookery sites, given that most sandy cays (the preferred location for turtle nesting) along the Great Barrier Reef similarly formed during the past 3000–3500 years (e.g. Fuentes et al. 2010; Hopley et al. 2007: 349–350; McLean & Stoddart 1978: 108; Woodroffe et al. 2007). In this sense, we posit an alternative hypothesis to Barker's (2004: 149) conclusion that there is “no case to be made for a general rise in bioproductivity in the area in the past 3000 years”. On the contrary, we hypothesise that expansion of island use over the past 3000 years was underwritten by expansion of key food resources such as turtles, which were central to the viability and success of offshore voyaging. Critically, and in line with Barker (2004), we do not believe that the demographic expansion that most likely accompanied the changes of the past 3000–3500 years simply reflected a deterministic response to increased resource availability. Alternatively, we posit that the decision to increase use of offshore islands reflected demographic pressures on the adjacent mainland (see Bowdler 1995: 956; O'Connor 1992: 58–59). Whether these demographic pressures related to demands for more food resources as a result of population increase, a reduction in availability in terrestrial food resources associated with drier climate (see Genever et al. 2003) and/or demands for new and separate territorial domains due to social fissioning remains to be seen (see McNiven 1999b).

We concur with Rowland (1996: 198) that offshore island use along the central Queensland coast was not contingent upon introduction of more advanced canoe technology. Beaton's (1985: 18) hypothesis that use of islands of Princess Charlotte Bay “2500 years ago probably had to await the introduction of the outrigger canoe of Papua-Melanesian origin” is not relevant to understanding early Aboriginal use of the offshore islands of Shoalwater Bay. Aboriginal people were visiting Otterbourne Island, 26 km off the coast, over 5000 years ago, which predates the earliest evidence for New Guinea migration into Torres Strait 2500–3000 years ago (McNiven et al. 2006) and similarly predates Austronesian (Lapita) expansions across the western Pacific, which commenced after 3500 years ago (Kirch 1997). Aboriginal people of central Queensland developed adequate bark canoes and associated seafaring skills to accomplish long-distance offshore voyaging many thousands of years prior to people entering Remote Oceania.

The ENSO contraction phase: 1000–2000 years ago

A dearth of radiocarbon dates occurs for midden deposits on islands of the central Queensland coast between 2000 and 1000 years ago (Figure 15). While it is possible that this dearth simply reflects issues of sampling, the existence of numerous radiocarbon dates covering the periods 3000–2000 and 1000–0 calBP on either side of the 2000–1000 calBP period suggests strongly that the dearth is real. Furthermore, while poor dating of most sites could easily contribute to an apparent dearth between 2000 and 1000 calBP, comprehensive dating of all 17 XUs comprising the Otterbourne Island Site 4 Square B midden sequence demonstrates a chronological gap (and by extension a depositional and occupational gap) between c.1400 and c.2800 calBP (Figure 7). An occupational hiatus between c.850 and c.3200 years ago was recorded in Square A of Otterbourne Island Site 4, while High Peak Island Site 10 also produced a depositional and occupational hiatus between c.500 and c.3300 years ago. Similarly, at Mazie Bay midden on North Keppel Island, Rowland (1999b: figure 1) documented a chronological gap between c.1100 and 3300 years ago, albeit separated by 35 cm of undated sediments. As such, we hypothesise that for Otterbourne Island, High Peak Island and North Keppel Island, and possibly for other islands off the central Queensland coast, a major period of occupational contraction covering hundreds of years occurred variably between approximately 1000 and 3000 years ago. While this pattern most likely reflects decreased overall use of islands, it is also possible that it reflects increased mobility and more ephemeral use of islands associated with encampments of shorter duration and minimal cultural discard.

The issue of the relationship between ENSO and late Holocene occupational gaps on northern Australian islands and mainland midden sites has attracted attention in recent years (e.g. Ulm 2011; Williams et al. 2010). For example, Sim and Wallis (2008: 100–101) identified chronological gaps across a suite of sites on Vanderlin Island in the Sir Edward Pellew Island Group, Gulf of Carpentaria, between about 4200 and 6700 calBP and between about 1700 and 2500 calBP. While the mid-Holocene “occupational hiatus phase” was attributed to lack of suitable watercraft and poor sea voyaging conditions, the late Holocene “low intensity or abandonment phase” was associated with decreased marine resource availability and poor sea voyaging conditions linked to increased “frequency” of ENSO events (Sim & Wallis 2008: 102–103). However, while research on the development of the modern ENSO system over the past 4000–5000 years reveals that the amplitude of ENSO events peaked between 1700 and 2500 years ago, the frequency of ENSO events decreases considerably during this period (Gagan et al. 2004: 136, figure 7; Moy et al. 2002: 164). Recent research supports the view that across the tropical Pacific Ocean, the “period of greatest ENSO variance in the Holocene occurred between 2000 and 1500 cal years BP” (Conroy et al. 2008: 1178).

An increased amplitude of ENSO activity represents increased variation between warm (El Niño) and cold (La Niña) conditions and can impact marine resources. For example, elevated sea temperatures from more extreme El Niño events have been correlated with dieback of corals across the southern Great Barrier Reef (McPhaden et al. 2006: 1741; Redondo-Rodriguez et al. 2012) and reductions in Pacific coral fish stocks (Lo-Yat et al. 2011). Yet studies have found that fish stocks in the Great Barrier Reef can increase following El Niño events (Cheal et al. 2007), implying that fish stocks would be lower during extended periods of lower ENSO frequency. Similarly, the numbers of green turtles at nesting sites along the Great Barrier Reef can vary up to three orders of magnitude in successive years, reflecting ENSO cycles with high breeding rates occurring around 2 years after major El Niño events and “crashes in nesting numbers” occurring around 2 years after major La Niña events (Limpus & Nicholls 1988: 402, 404; Limpus & Nicholls 2000). As such, it is likely that both fish and turtle stocks may have been lower between c.2500 and c.1700 years ago, when ENSO amplitude was higher and ENSO frequency lower. We hypothesise that these resource impacts resulted in decreased island visits and increased formation of ephemeral sites with minimal archaeological visibility. It is likely that the impact of ENSO-induced decreases in marine resources at Shoalwater Bay, like other parts of central Queensland, were complex and patchy depending upon spatial differences in the resilience of marine habitats (see Rowland 1999a). Such differences may account for variances in the chronology of occupational hiatuses seen at sites across the region between 1000 and 3000 years ago.

Cosgrove et al. (2007: 170) point out that “With ENSO came resource unpredictability, which heightened risk and uncertainty”. It is in this connection that the restricted appearance, albeit minor, of high-quality mainland chert artefacts at Otterbourne Island Site 4 to either side of c.1400–2800 cal BP occupational hiatus takes on significance. We speculate that use of mainland chert points to predominately mainland people undertaking short-term voyages to Otterbourne Island to hunt turtle during the lead into and out of the period of intensified ENSO activity. Furthermore, the use of highly curatable raw materials such as chert was part of a risk-mitigation strategy employed by more mobile peoples who may have experienced less time flexibility to obtain lower quality local raw materials due to increased time and effort spent in foraging for ENSO-stressed (i.e. less abundant and less predictable) food resources (Clarkson 2006: 155; Hiscock 1994, 2006; Shott 1986; Torrence 2000). Such reduced focus on use of time-consuming local resources may also help explain in part the marked drop-off in the exploitation and removal of oysters from rocks seen at Otterbourne Island Site 10 between c.850 and 1350 calBP. That risk mitigation was a factor influencing stone artefact technology of Aboriginal peoples using islands of the central Queensland coast, particularly during periods of rapid environmental change and resource stress, has also been argued by Hiscock (1994: 279) and Lamb (2011).

The social intensification phase: the past 1000 years

The most intensive period of occupation in the >5000 year history of Aboriginal visitation to Otterbourne Island was within the past 1000 years. Nearly all of the Square A deposit and more than half of the cultural deposit within Square B at Otterbourne Island Site 4 dated to this period. This dramatic quantitative increase in activity was seen also at Collins Island Site 24, where all of the deposit dated to the past 600 years, and at High Peak Island Site 10, where all of the midden deposit dated to the past 300 years. Barker (2004) documented a similar trend in the Whitsunday Islands, with the highest discard rates for most cultural materials at Nara Inlet 1 occurring after c.550 calBP. He argues that the ethnographic “system recorded historically may date from about 500 BP” (Barker 2004: 148). While this conclusion is plausible and probable, unfortunately the upper deposits of Nara Inlet Art Site, Hill Inlet 1 and Border Island 1 remain undated. Border (1999) found that half of the sites that he excavated across the central Northumberland Group dated to the past 1000 years (Figure 15). He argued that “this final phase is interpreted as being representative of the fully-fledged offshore economy, inferred from historical data” (Border 1999: 136). Perhaps of most significance are the Keppel Islands, where all of the sites excavated on South Keppel Island date to the past 800 years (Rowland 1982) (Figure 15). Rowland (1996: 204) pointed out that “there is also a change in the nature of deposits at Mazie Bay from about this time so that the development of true island based economies may be seen to date from the last 1000 years or less”. We also hypothesise that intensified occupation of Shoalwater Bay islands during the past 1000 years aligns with cultural configurations recorded ethnographically for the nineteenth century. However, as discussed above, we believe that the Shoalwater Bay islands were not inhabited by a separate Islander people, but by coastal peoples and marine specialists whose territory and settlement-subsistence system incorporated areas of coastal mainland and adjacent islands.

Despite maintaining mainland connections, we posit that within the past 1000 years, the use of the Shoalwater Bay islands intensified such that visitors from the mainland spent more time using the islands during the year compared to use prior to 1000 years ago. This extended use resulted in intensified exploitation of marine resources for food and concomitant intensified demands on quartz resources for tool manufacture. As a result, stone artefact assemblages on the islands became dominated by quartz, with the use of mainland raw materials (e.g. volcanics and chert) dropping to negligible levels. While reasons for intensified use of islands within the past 1000 years remain poorly understood, it is hypothesised that lack of evidence for major environmental changes producing increases in food biomass to parallel the scale of increased evidence for occupation implicate internally generated social changes. The fact that such dramatic changes also took place within the Whitsunday Islands to the north and the Keppel Islands to the south points to the existence of a broad-scale social phenomenon. A general chronological concurrence of such shared, macro-regional changes would require a degree of social integration and connectivity between communities along the central Queensland coast. Whatever the situation, an understanding of the long-term history of island use can only occur in conjunction with an understanding of the long-term history of the use of the mainland coast. Importantly, changing patterns of use of islands and the mainland are considered to have been mutually transformative, such that changes in use of the mainland subsequently structured changing use of islands, which in turn fed back to structure changes in mainland occupation, and so on.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Shoalwater Bay Islands
  5. Otterbourne Island
  6. High Peak Island
  7. Collins Island
  8. Discussion
  9. Conclusion
  10. Acknowledgements
  11. References
  12. Supporting Information

Our Shoalwater Bay excavation results demonstrate the continuing capacity of the central Queensland islands to produce archaeological evidence that both contributes to and challenges broader models of Aboriginal use of Australia's offshore islands. Indeed, the tropical coast of central Queensland provides the most compelling evidence available for any part of Australia for continuous, or near continuous, occupation of island archipelagos over the past 7000 years. Equally of importance, our Shoalwater Bay results reinforce previous insights for the broader central Queensland coast that the key environmental variables effecting island use relate to marine resources. In this sense, Sim and Wallis's (2008) hypothesis of the impact of climate change on island use during the late Holocene is considered of secondary importance, unless such changes can be shown to impact the availability of marine resources. While Sim and Wallis (2008: 103) do posit a relationship between increased storms (linked to increased ENSO activity) and diminution of mangroves and associated resources such as shellfish, the broader applicability and impact of this relationship remains to be demonstrated. Yet we endorse explorations of the impact of changes in ENSO amplitude upon marine resources, and on how Aboriginal people responded to such changes in terms of the perceived viability and risks of island voyaging. Whether or not such climatic changes similarly impacted exploitation of marine resources along the adjacent mainland coast of Shoalwater Bay, resulting in a similar occupational hiatus at sites, particularly midden deposits, variously between 1000 and 3000 years ago, is a question for future archaeological research. The possibility that mainland sites register such a hiatus is consistent with Ulm's (2006: 250) documentation of a period of “ephemeral low intensity occupation” between 2000 and 1000 years ago within midden sites of the southern Curtis Coast, located 100 km south of the Keppel Islands.

Following Rowland's pioneering research into past Aboriginal use of offshore islands of the southern Great Barrier Reef, our Shoalwater Bay islands research has built upon the detailed work of Barker (2004) and Border (1999) to further demonstrate long-term island occupation and reinforce the case for major broad-scale cultural changes taking place around 3000–3500 years ago and within the past 1000 years. Yet unlike the Whitsunday Islands to the north and the Keppel Islands to the south, the Shoalwater Bay region does not provide compelling evidence for a separate Islander people with minimal mainland contacts. To shed further light on this question of separateness will require comprehensive excavation of sites along the mainland coast of the Shoalwater Bay region. Furthermore, the boundedness of the Shoalwater Bay islands also needs to be explored and tested further, by excavating a wider range of local inshore and offshore island sites and extending the research of Border (1999) to other islands of the Northumberland Group (including the Percy Islands). In terms of explanatory frameworks, our research has tapped into new palaeoenvironmental evidence for major changes in marine resource productivity within the past 3000–3500 years that need to be tested in a broader range of island and mainland coastal contexts (e.g. coring of mangroves and freshwater swamps). Coupled with dramatic increases in occupational intensity on islands within the past 1000 years that appear to reflect complex social changes in regional socio-economic systems, the Shoalwater Bay region presents a broad range of opportunities to explore the extent to which social and environmental issues shaped the long-term history of the Darumbal people. Based on previous experiences, the results of this future research will have important implications for the broader central Queensland coast and Australian islands more generally.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Shoalwater Bay Islands
  5. Otterbourne Island
  6. High Peak Island
  7. Collins Island
  8. Discussion
  9. Conclusion
  10. Acknowledgements
  11. References
  12. Supporting Information

We offer special thanks to the Darumbal–Noolar Murree Aboriginal Corporation and Darumbal Enterprises for providing the rare opportunity to undertake fieldwork at Shoalwater Bay. In particular, we thank Doug Hatfield, Trevor Hatfield, David Hatfield and Terry Hatfield for their generous support. Funding for the archaeological survey of the islands of Shoalwater Bay was provided by a grant from Coastcare, with contributions from GBRMPA and the Queensland Department of Environment. Excavations were kindly funded through a Coastcare grant. We thank the Queensland Environmental Protection Agency for issuing a permit to allow our excavations to take place. Fieldwork was made possible by the hard work and good humour of Lance Aaskov, Andrew (AJ) Batchelor, Joseph Cora, Tosi (TJ) Cora Jnr, Willie Hayden, Billy Mann, Malcolm Mann and Carl Porter. Beverly Brigham, Kara Rasmanis, Cassandra Rowe and Toby Wood (School of Geography and Environmental Science, Monash University) drafted the figures. For assistance with radiocarbon dates, we thank Alan Hogg and Fiona Petchey (University of Waikato Radiocarbon Dating Laboratory). Robyn Jones, Catherine Longford, Joel Mayes and Alexia Peniguel kindly assisted with the laborious task of laboratory sorting the Otterbourne Island 4 midden materials. Helpful comments on an earlier version of this paper were kindly made by two referees, Peter White and Annie Ross. Finally, we thank Luke Godwin for his continuing generous support and assistance over many years.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Shoalwater Bay Islands
  5. Otterbourne Island
  6. High Peak Island
  7. Collins Island
  8. Discussion
  9. Conclusion
  10. Acknowledgements
  11. References
  12. Supporting Information
  • ABARE, AGS, BRS. 1993. Shoalwater Bay Military Training Area Resource Assessment. Australian Bureau of Agricultural and Resource Economics, Australian Geological Survey, and Bureau of Resource Sciences, Canberra.
  • Allen, J. and Corris, P. 1977. The Journal of John Sweatman. University of Queensland, St Lucia.
  • Atchison, J., Head, L. and Fullagar, R. 2005. Archaeobotany of fruit seed processing in a monsoon savanna environment: evidence from the Keep River region, Northern Territory, Australia. Journal of Archaeological Science 32: 167181.
  • Australian Department of Defence. 2009. Appendix 3: Native terrestrial plant species recorded in Shoalwater Bay Training Area. In State of the Environment Report for Shoalwater Bay Training Area 2008. Department of Defence, Russell.
  • Australian Hydrographic Service. 2004. North Reef to Port Clinton, 1:50,000 Hydrographic Map. Aus 820. Royal Australian Navy, Wollongong.
  • Australian Hydrographic Service. 2009. Port Clinton to Percy Isles, 1:50,000 Hydrographic Map. Aus 822. Royal Australian Navy, Wollongong.
  • Ayling, A.M., Ayling, A.L. and Berkelmans, R. 1998. Shoalwater Bay Fringing Reef Resource Assessment. Research Publication No. 54. Great Barrier Reef Marine Park Authority, Townsville.
  • Barker, B. 1996. Maritime hunter–gatherers on the tropical coast: a social model for change. In S. Ulm , I. Lilley and A. Ross (eds), Australian Archaeology ′95: Proceedings of the 1995 Australian Archaeological Association Annual Conference, pp. 3143. Tempus 6. Anthropology Museum, University of Queensland, St Lucia.
  • Barker, B. 1998. Use and continuity in the customary marine tenure of the Whitsunday Islands. In N. Peterson and B. Rigsby (eds), Customary Marine Tenure in Australia, pp. 8995. Oceania Monograph, 48. Oceania Publications, Sydney.
  • Barker, B. and Schon, R. 1994. A preliminary assessment of the spatial distribution of stone artefacts from the South Molle Island Aboriginal quarry, Whitsunday Islands, central Queensland Coast. Memoirs of the Queensland Museum 37 (1): 512.
  • Barker, B.C. 1991. Nara Inlet 1: Coastal resource use and the Holocene marine transgression in the Whitsunday Islands, central Queensland. Archaeology in Oceania 26: 102109.
  • Barker, B.C. 1999. Coastal occupation in the Holocene: Environment, resource use and resource continuity. In J. Hall and I.J. McNiven (eds), Australian Coastal Archaeology, pp. 119127. Research Papers in Archaeology and Natural History No. 31. ANH Publications, Department of Archaeology and Natural History, Research School Pacific and Asian Studies, Australian National University, Canberra.
  • Barker, B.C. 2004. The Sea People: Late Holocene Maritime Specialisation in the Whitsunday Islands, Central Queensland. Terra Australia, 20. Research School of Pacific and Asian Studies, Australian National University Press, Canberra.
  • Beaton, J.M. 1985. Evidence for a coastal occupation time-lag at Princess Charlotte Bay (north Queensland) and implications for coastal colonisation and population growth theories for Aboriginal Australia. Archaeology in Oceania 20 (1): 120.
  • Bird, D.W., Bliege Bird, R. and Parker, C.H. 2005. Aboriginal burning regimes and hunting strategies in Australia's Western Desert. Human Ecology 33 (4): 443464.
  • Border, A. 1985. Unchartered Waters: Hunter–Gatherer Specialisation on the Central Queensland Coast. Unpublished BA (Hons) thesis, University of New England, Armidale.
  • Border, A. 1991. NIAS: Northumberland Island Archaeological Surveys (1991). Unpublished report, Department of Environment and Heritage, Brisbane.
  • Border, A. 1999. Aboriginal settlement of offshore islands in the Southern Great Barrier Reef Province, central Queensland. In J. Hall and I.J. McNiven (eds), Australian Coastal Archaeology, pp. 129139. Research Papers in Archaeology and Natural History No. 31. ANH Publications, Department of Archaeology and Natural History, Research School Pacific and Asian Studies, Australian National University, Canberra.
  • Bostock, H., Ryan, D., Brooke, B., Packett, B., Hancock, G., Pietsch, T., Revill, A., Leeming, R., Moss, P. and Harle, K. 2006. Sediment Accumulation and Holocene Evolution of the Fitzroy River Lower Floodplain, Central Queensland, Australia. Technical Report 48. Cooperative Research Centre for Coastal Zone, Estuary and Waterway Management (Coastal CRC), Indooroopilly.
  • Bostock, P.D. and Holland, A.E. 2010. Census of the Queensland Flora. Queensland Herbarium, Department of Environment and Resource Management, Brisbane.
  • Bowdler, S. 1995. Offshore islands and maritime explorations in Australian prehistory. Antiquity 69: 945958.
  • Bryan, S.E., Ewart, A., Stephens, C.J., Parianos, J. and Downes, P.J. 2000. The Whitsunday Volcanic Province, central Queensland, Australia: Lithological and stratigraphic investigations of a silicic-dominated large igneous province. Journal of Volcanology and Geothermal Research 99: 5578.
  • Burgis, W.A. 1974. Cainozoic History of the Torilla Peninsula, Broad Sound, Queensland. Australian Bureau of Mineral Resources, Geology and Geophysics. Report 172. Australian Government Publishing Service, Canberra.
  • Catling, P.C., Mason, I.J., Richards, G.C., Schodde, R. and Wombey, J.C. 1994. The land vertebrate fauna of the eastern dunefields and tidal zone. In Commonwealth Commission of Inquiry Shoalwater Bay, Capricornia Coast, Queensland. Research Reports. Report No. 5 Volume A, pp. 117172. Australian Government Publishing Service, Canberra.
  • Cenko, C.I., Hotchin, D. and Barry, S. 1990. Vacant Crown land assessments for central region. Report on file at Queensland Department of Environment and Heritage, Rockhampton Office.
  • Chappell, J. 1983. Evidence for smoothly falling sea level relative to north Queensland, Australia, during the past 6000 yr. Nature 302: 406408.
  • Cheal, A.J., Delean, S., Sweatman, H. and Thompson, A.A. 2007. Spatial synchronicity in coral reef fish populations and the influence of climate. Ecology 88 (1): 158169.
  • Clarke, A. 1988. Archaeological and ethnobotanical interpretation of plant remains from Kakadu National Park, Northern Territory. In B. Meehan and R. Jones (eds), Archaeology with Ethnography: An Australian Perspective, pp. 123136. Department of Prehistory, Research School of Pacific Studies, The Australian National University, Canberra.
  • Clarkson, C. 2006. Lithics in the Land of the Lightning Brothers: The Archaeology of Wardaman Country, Northern Territory. Terra Australis, 25. ANU E Press, Canberra.
  • Commission of Inquiry. 1994. Commonwealth Commission of Inquiry Shoalwater Bay, Capricornia Coast, Queensland. Interim Report. Report No. 2. Australian Government Publishing Service, Canberra.
  • Conroy, J.L., Overpeck, J.T., Cole, J.E., Shanahan, T.M. and Steinitz-Kannan, M. 2008. Holocene changes in eastern tropical Pacific climate inferred from a Galápagos lake sediment record. Quaternary Science Reviews 27: 11661180.
  • Cook, P.J. and Mayo, W. 1977. Sedimentology and Holocene History of a Tropical Estuary (Broad Sound, Queensland). Department of National Development, Bureau of Mineral Resources, Geology and Geophysics. Australian Government Publishing Service, Canberra.
  • Cook, P.J. and Polach, H.A. 1973. A Chenier sequence at Broad Sound, Queensland and evidence against a Holocene high sea level. Marine Geology 14: 253268.
  • Cosgrove, R., Field, J. and Ferrier, A. 2007. The archaeology of Australia's tropical rainforests. Palaeogeography, Palaeoclimatology, Palaeoecology 251: 150173.
  • Crowley, G.M. 1996. Late Quaternary mangrove distribution in northern Australia. Australian Systematic Botany 9: 219225.
  • Crowley, G.M. and Gagan, M.K. 1995. Holocene evolution of coastal wetlands in wet-tropical northeastern Australia. The Holocene 5: 385399.
  • Di Piazza, A. 1998. Archaeobotanical investigations of an earth oven in Kiribati, Gilbert Islands. Vegetation History and Archaeobotany 7: 149154.
  • Dickson, F.P. 1981. Australian Stone Hatchets: A Study in Design and Dynamics. Academic Press, Sydney.
  • Donders, T.H., Wagner-Cremer, F. and Visscher, H. 2008. Integration of proxy data and model scenarios for the mid-Holocene onset of ENSO variability. Quaternary Science Reviews 27: 571579.
  • Endean, R., Stephenson, W. and Kenny, R. 1956. The ecology and distribution of intertidal organisms on certain islands off the Queensland coast. Australian Journal of Marine and Freshwater Research 7 (3): 317342.
  • Flinders, M. 1814. A Voyage to Terra Australis, 1801, 1802, 1803 in His Majesty's Ship the Investigator. Nicoll, London.
  • Fuentes, M.M.P.B., Dawson, J.L., Smithers, S.G., Hamann, M. and Limpus, C.J. 2010. Sedimentological characteristics of key sea turtle rookeries: Potential implications under projected climate change. Marine and Freshwater Research 61: 464473.
  • Gagan, M.K., Hendy, E.J., Haberle, S.G. and Hantoro, W.S. 2004. Postglacial evolution of the Indo-Pacific Warm Pool and El Niño-Southern Oscillation. Quaternary International 118–119: 127143.
  • Genever, M., Grinrod, J. and Barker, B. 2003. Holocene palynology of Whitehaven Swamp, Whitsunday Island, Queensland, and implications for the regional archaeological record. Palaeogeography, Palaeoclimatology, Palaeoecology 201: 141156.
  • Grant, K., Finlayson, A.A., Spate, A.P. and Fergusson, T.G. 1979. Terrain Analysis and Classification for Engineering and Conservation Purposes of the Port Clinton Area, Qld, Including the Shoalwater Bay Military Training Area. Technical Paper 29. Division of Applied Geomechanics, CSIRO, Melbourne.
  • Hall, J. 1980. Minner Dint: A recent Aboriginal midden on Moreton Island, southern Queensland. University of Queensland, Anthropology Museum, Occasional Papers in Anthropology 10: 94112.
  • Hayes, A. and Hayes, S. 2007. Ngambunyjarri: Thalanyjibarndi Yininyjarri. Thalanyji Plant Names and Uses. Wangka Maya Pilbara Aboriginal Language Centre, Onslow.
  • Hayne, M. and Chappell, J. 2001. Cyclone frequency during the last 5000 yrs from Curacoa Island, Queensland. Palaeogeography, Palaeoclimatology, Palaeoecology 168: 201219.
  • Hermes, M. 1984. Mazie Bay: An Analysis and Interpretation of the Faunal Component. Unpublished BA (Hons) thesis, Department of Prehistory and Anthropology, Australian National University, Canberra.
  • Hillman, G.C. 1989. Late Palaeolithic plant foods from Wadi Kubbaniya in Upper Egypt: Dietary diversity, infant weaning, and seasonality in a riverine environment. In D.R. Harris and G.C. Hillman (eds), Foraging and Farming: The Evolution of Plant Exploitation, pp. 207239. Unwin Hyman, London.
  • Hiscock, P. 1984. A preliminary report on the stone artefacts from Colless Creek Cave, northwest Queensland. Queensland Archaeological Research 1: 120151.
  • Hiscock, P. 1994. Technological responses to risk in Holocene Australia. Journal of World Prehistory 8 (3): 267292.
  • Hiscock, P. 2006. Blunt and to the point: Changing technological strategies in Holocene Australia. In I. Lilley (ed.), Archaeology of Oceania: Australia and the Pacific Islands, pp. 6995. Blackwell, Malden, MA.
  • Hiscock, P. 2008. Archaeology of Ancient Australia. Routledge, New York.
  • Hiscock, P. and Mitchell, S. 1993. Stone Artefact Quarries and Reduction Sites in Australia: Towards A Type Profile. Technical Publication Series No. 4. Australian Heritage Commission, Canberra.
  • Hopley, D., Smithers, S.G. and Parnell, K.E. 2007. The Geomorphology of the Great Barrier Reef: Development, Diversity and Change. Cambridge University Press, Cambridge, UK.
  • Howitt, A.W. 1904. The Native Tribes of South-East Australia. Macmillan, London.
  • Isaacs, J. 1996. A Companion Guide to Bush Food. Lansdowne Publishing, Sydney.
  • Jukes, J.B. 1847. Narrative of the Surveying Voyage of H.M.S. Fly. 2 vols. T. & W. Boone, London.
  • King, P.P. 1827. Narrative of A Survey of the Intertropical and Western Coasts of Australia Performed Between the Years 1818 and 1822. 2 vols. John Murray, London.
  • Kirch, P.V. 1989. Second millennium B.C. arboriculture in Melanesia: Archaeological evidence from the Mussau Islands. Economic Botany 43 (2): 225240.
  • Kirch, P.V. 1997. The Lapita Peoples. Blackwell, Oxford.
  • Kirch, P.V., Steadman, D.W., Butler, V.L., Hather, J. and Weisler, M.I. 1995. Prehistory and human ecology in Eastern Polynesia: Excavations at Tangatatau Rockshelter, Mangaia, Cook Islands. Archaeology in Oceania 30 (2): 4765.
  • Kleypas, J.A. 1996. Coral reef development under naturally turbid conditions: Fringing reefs near Broad Sound, Australia. Coral Reefs 15: 153167.
  • Lamb, L. 2011. Rock of Ages: South Molle Island Quarry, Whitsunday Islands: Use and Distribution of Stone through Space and Time. BAR International Series, 2262. Archaeopress, Oxford.
  • Lamb, L. and Barker, B. 2001. Evidence for early Holocene change in the Whitsunday Islands: A new radiocarbon determination from Nara Inlet 1. Australian Archaeology 53: 4243.
  • Lamprell, K. and Healy, J. 1998. Bivalves of Australia, Volume 2. Backhuys Publishers, Leiden.
  • Lamprell, K. and Whitehead, T. 1992. Bivalves of Australia, Volume 1. Crawford House Press, Bathurst.
  • Lanyon, J., Limpus, C.J. and Marsh, H. 1989. Dugongs and turtles: Grazers in the seagrass system. In A.W.D. Larkum , A.J. McComb and S.A. Shepherd (eds), Biology of Seagrasses: A Treatise on the Biology of Seagrasses with Special Reference to the Australian Region, pp. 610634. Elsevier, New York.
  • Larcombe, P. and Carter, R.M. 2004. Cyclone pumping, sediment partitioning and the development of the Great Barrier Reef shelf system: A review. Quaternary Science Reviews 23: 107135.
  • Larcombe, P. and Woolfe, K.J. 1999. Terrigenous sediments as influences upon Holocene nearshore coral reefs, Central Great Barrier Reef, Australia. Australian Journal of Earth Sciences 46: 141154.
  • Larcombe, P., Carter, R.M., Dye, J., Gagan, M.K. and Johnson, D.P. 1995. New evidence for episodic postglacial sea-level rise, central Great-Barrier-Reef, Australia. Marine Geology 127: 144.
  • Lee Long, W.J., McKenzie, L.J. and Coles, R.G. 1997. Seagrass Communities in the Shoalwater Bay Region, Queensland – Spring (September) 1995 and Autumn (April) 1996. Queensland Department of Primary Industries Information Series QI96042. QDPI, Brisbane.
  • Lewis, S.E., Wüst, R.A., Webester, J.M. and Shields, G.A. 2008. Mid-late Holocene sea-level variability in eastern Australia. Terra Nova 20: 7481.
  • Lewis, S.E., Sloss, C.R., Murray-Wallace, C.V., Woodroffe, C.D. and Smithers, S.G. 2013. Post-glacial sea-level changes around the Australian margin: A review. Quaternary Science Reviews. doi.org/10.1016/j.quascirev.2012.09.006.
  • Limpus, C. and Nicholls, N. 1988. The Southern Oscillation regulates the annual numbers of green turtles (Chelonia mydas) breeding around northern Australia. Australian Journal of Wildlife Research 15: 57161.
  • Limpus, C. and Nicholls, N. 2000. ENSO regulation of Indo-Pacific green turtle populations. In G. Hammer , N. Nicholls and C. Mitchell (eds), Applications of Seasonal Climate Forecasting in Agricultural and Natural Ecosystems, pp. 399408. Kluwer Academic Publishers, Dordrecht.
  • Limpus, C.J. 2008. A Biological Review of Australian Marine Turtles. 2: Green Turtle Chelonia mydas (Linnaeus). Environmental Protection Agency, Brisbane.
  • Limpus, C.J., Limpus, D.J., Arthur, K.E. and Parmenter, C.J. 2005. Monitoring Green Turtle Population Dynamics in Shoalwater Bay: 2000–2004. Research Publication No. 83. Great Barrier Reef Marine Park Authority, Townsville.
  • Limpus, C.J., Clifton, D., Griffin, K., Kemp, L., Gallagher, L., Fisher, S. and Parmenter, J. n.d. Survey of Marine Turtle Nesting Distribution in Queensland, 2000 and 2001: Broad Sound to Repulse Bay, Central Queensland. Unpublished report.
  • Low, T. 1988. Wild Food Plants of Australia. Angus and Robertson, North Ryde, NSW.
  • Lo-Yat, A., Simpson, S.D., Meekan, M., Lecchini, D., Martinez, E. and Galzin, R. 2011. Extreme climatic events reduce ocean productivity and larval supply in a tropical reef ecosystem. Global Change Biology 17 (4): 16951702.
  • Lumholtz, C. 1889. Among Cannibals. An Account of Four Years' Travels in Australia and of Camp Life with the Aborigines of Queensland. John Murray, London.
  • Marsh, H., Corkeron, P., Lawler, I., Preen, A. and Lanyon, J. 1996. The Status of the Dugong in the Southern Great Barrier Reef Marine Park. Great Barrier Reef Marine Park Authority, Townsville.
  • Matthews, P.J. and Gosden, C. 1997. Plant remains from waterlogged sites in the Arawe Islands, West New Britain Province, Papua New Guinea: Implications for the history of plant use and domestication. Economic Botany 51 (2): 121133.
  • McConnell, K. and O'Connor, S. 1997. 40,000 year record of food plants in the southern Kimberley Ranges, Western Australia. Australian Archaeology 45: 2031.
  • McCormac, F.G., Hogg, A.G., Blackwell, P.G., Buck, C.E., Higham, T.F.G. and Reimer, P.J. 2004. SHCAL04 Southern Hemisphere calibration, 0–11.0 cal kyr BP. Radiocarbon 46: 10871092.
  • McIntyre-Tamwoy, S. and Harrison, R. 2004. Monuments to colonialism? Stone arrangements, tourist cairns, and turtle magic at Evans Bay, Cape York. Australian Archaeology 59: 3142.
  • McLean, R.F. and Stoddart, D.R. 1978. Reef island sediments of the northern Great Barrier Reef. Philosophical Transactions of the Royal Society of London. Series A 291: 101117.
  • McNiven, I.J. 1999a. Darumbal Cultural Heritage, Shoalwater Bay Military Training Area. Stage 2: Islands & Peninsula Archaeological Survey. 2 vols. Report prepared for Darumbal-Noolar Murree Aboriginal Corporation for Land and Culture, Rockhampton.
  • McNiven, I.J. 1999b. Fissioning and regionalisation: The social dimensions of changes in Aboriginal use of the Great Sandy Region, coastal S.E. Queensland. In J. Hall and I.J. McNiven (eds), Australian Coastal Archaeology, pp. 157168. Research Papers in Archaeology & Natural History, No. 31. ANH Publications, Department of Archaeology and Natural History, Research School Pacific and Asian Studies, Australian National University, Canberra.
  • McNiven, I.J. 2000. Treats or retreats: Aboriginal island use along the Gippsland coast, southeastern Australia. The Artefact 23: 2234.
  • McNiven, I.J. 2003. Saltwater people: Spiritscapes, maritime rituals and the archaeology of Australian Indigenous seascapes. World Archaeology 35 (3): 329349.
  • McNiven, I.J. and Russell, L. 1997. Darumbal Cultural Heritage, Shoalwater Bay Military Training Area: Preliminary Archaeological and Anthropological Survey. Report prepared for Darumbal-Noolar Murree Aboriginal Corporation for Land and Culture, Rockhampton.
  • McNiven, I.J., Dickinson, W.R., David, B., Weisler, M., von Gnielinski, F., Carter, M. and Zoppi, U. 2006. Mask Cave: Red-slipped pottery and the Australian–Papuan settlement of Zenadh Kes (Torres Strait). Archaeology in Oceania 41 (2): 4981.
  • McPhaden, M.J., Zebiak, S.E. and Glantz, M.H. 2006. ENSO as an integrating concept in Earth science. Science 314: 17401745.
  • Meehan, B., Gaffey, P. and Jones, R. 1979. Fire and steel: Aboriginal exploitation of pandanus and some wider implications. University of Queensland, Anthropology Museum, Occasional Papers in Anthropology 9: 7396.
  • Memmott, P. 1993. Aboriginal Social History and Land Affiliations in the Rockhampton – Shoalwater Bay Region. Unpublished report to the Commonwealth Commission of Inquiry Shoalwater Bay, Capricornia Coast, Queensland.
  • Memmott, P. 1994. Aboriginal social history and land affiliations in the Rockhampton region. In Commonwealth Commission of Inquiry Shoalwater Bay, Capricornia Coast, Queensland. Research Reports. Report No. 5 Volume C, pp. 1106. Australian Government Publishing Service, Canberra.
  • Mills, R.A. 1992. Aboriginal Occupation of Lizard Island. Unpublished Masters thesis, University of Sydney.
  • Moore, D.R. 1979. Islanders and Aborigines at Cape York. Australian Institute of Aboriginal Studies, Canberra.
  • Moy, C., Seltzer, G., Rodbell, D. and Anderson, D. 2002. Variability of El Niño/Southern Oscillation activity at millennial timescales during the Holocene epoch. Nature 420: 162165.
  • Mulvaney, J. and Kamminga, J. 1999. Prehistory of Australia. Allen and Unwin, Crows Nest, NSW.
  • Murray, C.G. 1975. Port Clinton, Queensland – 1:250,000 Geological Series. Explanatory Notes. Sheet SF/56-9. Bureau of Mineral Resources, Geology and Geophysics. Australian Government Publishing Service, Canberra.
  • Nietschmann, B. and Nietschmann, J. 1981. Good dugong, bad dugong; Bad turtle, good turtle. Natural History 90 (5): 5563.
  • O'Connor, S. 1992. The timing and nature of prehistoric island use in northern Australia. Archaeology in Oceania 27 (2): 4960.
  • O'Neill, P. 2008a. Sea grasses, mangroves and saltmarshes. In L. Sommer (ed.), State of the Environment Report for Shoalwater Bay Training Area 2008, pp. 179208. Department of Defence, Canberra.
  • O'Neill, P. 2008b. Marine fauna. In L. Sommer (ed.), State of the Environment Report for Shoalwater Bay Training Area 2008, pp. 239270. Department of Defence, Canberra.
  • Page, M.C. and Dickens, G.R. 2005. Sediment fluxes to Marion Plateau (southern Great Barrier Reef Province) over the last 130 ky: new constraints on “transgressive-shedding” off northeastern Australia. Marine Geology 219: 2745.
  • Perry, C.T. and Smithers, S.G. 2011. Cycles of coral reef “turn-on”, rapid growth and “turn-off” over the past 8500 years: A context for understanding modern ecological states and trajectories. Global Change Biology 17 (1): 7686.
  • Proske, U. and Haberle, S.G. 2012. Island ecosystem and biodiversity dynamics in northeastern Australia during the Holocene: Unravelling short-term impacts and long-term drivers. The Holocene 22 (10): 10971111.
  • Randall, J.E., Allen, G.R. and Steene, R.C. 1990. Fishes of the Great Barrier Reef and Coral Sea. University of Hawaii Press, Honolulu, HI.
  • Redondo-Rodriguez, A., Weeks, S.J., Berkelmans, R., Hoegh-Guldberg, O. and Lough, J.M. 2012. Climate variability of the Great Barrier Reef in relation to the tropical Pacific and El Niño-Southern Oscillation. Marine and Freshwater Research 63 (1): 3447.
  • Rees, G. 1980. A Wildlife Management Programme on South Island National Park, Percy Isles (N.P. 27 – Northumberland). Unpublished report. Wildlife Management Section, Management and Operations Branch, Brisbane.
  • Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, W.J., Blackwell, P.G., Bronk Ramsey, C., Buck, C.E., Burr, G.S., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P., Hajdas, I., Heaton, T.J., Hogg, A.G., Hughen, K.A., Kaiser, K.F., Kromer, B., McCormac, F.G., Manning, S.W., Reimer, R.W., Richards, D.A., Southon, J.R., Talamo, S., Turney, C.S.M., van der Plicht, J. and Weyhenmeyer, C.E. 2009. Intcal09 and Marine09 radiocarbon age calibration curves, 0–50,000 years cal BP. Radiocarbon 51 (4): 11111150.
  • Rogers, E.deN. 1886–1887. Adventures in the Islands. Ms. Rockhampton and District Historical Society, Rockhampton.
  • Rogers, E.deN. 1888. A New Year's Cruise on the Queensland Coast. Ms. Rockhampton and District Historical Society, Rockhampton.
  • Roth, W. 1910. Social and Individual Nomenclature. North Queensland Ethnography Bulletin No. 18. Records of the Australian Museum 8 (1): 79106.
  • Roth, W.E. 1898. The Aborigines of the Rockhampton and Surrounding Coast Districts. A report to the Commissioner of Police, Queensland. Mitchell Library, Sydney.
  • Rowe, C. 2007a. Vegetation change following mid-Holocene marine transgression of the Torres Strait shelf: A record from the island of Mua, northern Australia. The Holocene 17 (7): 927937.
  • Rowe, C. 2007b. A palynological investigation of Holocene vegetation change in Torres Strait, seasonal tropics of northern Australia. Palaeogeography, Palaeoclimatology and Palaeoecology 25: 83103.
  • Rowland, M.J. 1980. The Keppel Islanders – Preliminary investigations. Australian Archaeology 11: 117.
  • Rowland, M.J. 1981. Radiocarbon dates for a shell fishhook and disc from Mazie Bay, North Keppel Island. Australian Archaeology 12: 6369.
  • Rowland, M.J. 1982. Further radiocarbon dates from Keppel Islands. Australian Archaeology 15: 4348.
  • Rowland, M.J. 1983. A note on corrections to radiocarbon dates from the Keppel Islands. Australian Archaeology 17: 134135.
  • Rowland, M.J. 1984. A long way in a bark canoe: Aboriginal occupation of the Percy Isles. Australian Archaeology 18: 1731.
  • Rowland, M.J. 1985. Further radiocarbon dates from Mazie Bay, North Keppel Island. Australian Archaeology 21: 113118.
  • Rowland, M.J. 1996. Prehistoric archaeology of the Great Barrier Reef Province: Retrospect and prospect. In P.M. Veth and P. Hiscock (eds), Archaeology of Northern Australia. Regional Perspectives, pp. 191211. Tempus 4. Anthropology Museum, University of Queensland, St Lucia.
  • Rowland, M. 1999a. Holocene environmental variability: Have its impacts been underestimated in Australian pre-history? The Artefact 22: 1148.
  • Rowland, M.J. 1999b. The Keppel Islands – “A 3000 year” event revisited. In J. Hall and I.J. McNiven (eds), Australian Coastal Archaeology, pp. 141155. Research Papers in Archaeology & Natural History, No. 31. ANH Publications, Department of Archaeology and Natural History, Research School Pacific and Asian Studies, Australian National University, Canberra.
  • Rowland, M.J. 2008. Colonization, environment and insularity: Prehistoric island use in the Great Barrier Reef Province, Queensland, Australia. In J. Conolly and M. Campbell (eds), Comparative Island Archaeologies, pp. 85104. British Archaeological Reports, Oxford.
  • Royal Australian Survey Corps. 1986. Port Clinton 1:250,000 Topographic Map. Series 1501, Sheet SF56-9, Edition 2. Department of Defence, Canberra.
  • Schodde, R., Catling, P.C., Mason, I.J., Richards, G.C. and Wombey, J.C. 1992. The Land Vertebrate Fauna of the Shoalwater Bay Training Area, Queensland. A CSIRO Survey for the Department of Defence, Division of Wildlife and Ecology, Canberra.
  • Scopélitis, J., Andréfouët, S., Phinn, S., Done, T. and Chabanet, P. 2011. Coral colonisation of a shallow reef flat in response to rising sea-level: Quantification from 35 years of remote sensing at Heron Island, Australia. Coral Reefs 30: 951965.
  • Sharpe, P.R. 1986. Keys to Cyperaceae, Restionaceae & Juncaceae of Queensland. Queensland Department of Primary Industries, Brisbane.
  • Shott, M. 1986. Technological organization and settlement mobility: An ethnographic examination. Journal of Anthropological Research 42: 1551.
  • Shulmeister, J. and Lees, B.G. 1995. Pollen evidence from tropical Australia for the onset of an ENSO-dominated climate at c. 4000 BP. The Holocene 5: 1018.
  • Sim, R. and Wallis, L.A. 2008. Northern Australian offshore island use during the Holocene: The archaeology of Vanderlin Island, Sir Edward Pellew Group, Gulf of Carpentaria. Australian Archaeology 67: 95106.
  • Slater, J. and Stokes, T. 1997. Shoalwater Bay (Dugong) Plan of Management. Great Barrier Reef Marine Park Authority, Townsville.
  • Sloss, C.R., Murray-Wallace, C.V. and Jones, B.G. 2007. Holocene sea-level change on the southeast coast of Australia: A review. The Holocene 17 (7): 9991014.
  • Stanley, T.D. and Ross, E.M. 1989. Flora of South-Eastern Queensland, Volume 3. Queensland Department of Primary Industries, Brisbane.
  • Stephens, K. and Sharp, D. 2009. The Flora of North Stradbroke Island. Queensland Herbarium, Environmental Protection Agency, Brisbane.
  • Stuiver, M. and Reimer, P.J. 1993. Extended 14C database and revised CALIB radiocarbon calibration program. Radiocarbon 35: 215230.
  • Summerhayes, G.R., Leavesley, M., Fairbairn, A., Mandui, H., Field, J., Ford, A. and Fullagar, R. 2010. Human adaptation and plant use in highland New Guinea 49,000 to 44,000 years ago. Science 330: 7881.
  • Swadling, P., Araho, N. and Ivuyo, B. 1991. Settlements associated with the inland Sepik–Ramu Sea. Indo-Pacific Prehistory Association Bulletin 11: 92112.
  • Telford, R.J., Heegaard, E. and Birks, H.J.B. 2004. The intercept is a poor estimate of a calibrated radiocarbon age. The Holocene 14 (2): 296298.
  • Thomson, L.A.J., Englbergerm, L., Guarino, L., Thaman, R.R. and Elevitch, C.R. 2006. Pandanus tectorius (pandanus), ver. 1.1. In C.R. Elevitch (ed.), Species Profiles for Pacific Island Agroforestry. Permanent Agriculture Resources (PAR), Hõlualoa, HI. Available from http://www.traditionaltree.org
  • Torrence, R. 2000. Hunter–gatherer technology: Macro and microscale approaches. In C. Panter-Brick , R.H. Layton and P. Rowley-Conwy (eds), Hunter–Gatherers: An Interdisciplinary Perspective, pp. 99143. Cambridge University Press, Cambridge, UK.
  • Trnski, T., Bray, D., Leis, J., McGrouther, M. and Reader, S. 1994. Survey of fishes of Shoalwater Bay Training Area, Queensland. In Commonwealth Commission of Inquiry Shoalwater Bay, Capricornia Coast, Queensland. Research Reports. Report No. 5, Volume A, pp. 235281. Australian Government Publishing Service, Canberra.
  • Tudhope, A.W., Chilcott, C.P., McCulloch, M.T., Cook, E.R., Chappell, J., Ellam, R.M., Lea, D.W., Lough, J.M. and Shimmield, G.B. 2001. Variability in the El Niño–Southern Oscillation through a glacial–interglacial cycle. Science 291: 15111517.
  • Turney, C.S.M. and Hobbs, D. 2006. ENSO influence on Holocene Aboriginal populations in Queensland, Australia. Journal of Archaeological Science 33 (12): 17441748.
  • Ulm, S. 2002. Marine and estuarine reservoir effects in central Queensland, Australia: Determination of the modern marine calibration curve. Geoarchaeology 17: 319348.
  • Ulm, S. 2006. Coastal Themes: An Archaeology of the Southern Curtis Coast, Queensland. Terra Australis, 24. ANU E Press, Canberra.
  • Ulm, S. 2011. Coastal foragers on southern shores: Marine resource use in northeast Australia since the late Pleistocene. In N.F. Bicho , J.A. Haws and L.G. Davis (eds), Trekking the Shore: Changing Coastlines and the Antiquity of Coastal Settlement, pp. 441461. Springer, New York.
  • Ulm, S. and Reid, J. 2000. Index of dates from archaeological sites in Queensland. Queensland Archaeological Research 12: 1129.
  • Van Woesik, R. 1992. Ecology of Coral Assemblages on Continental Islands in the Southern Section of the Great Barrier Reef, Australia. Unpublished PhD thesis, James Cook University, Townsville.
  • Van Woesik, R. and Done, T.J. 1997. Coral communities and reef growth in the southern Great Barrier Reef. Coral Reefs 16: 103115.
  • Weisler, M.I. 1993. The importance of fish otoliths in Pacific island archaeofaunal analysis. New Zealand Journal of Archaeology 15: 131159.
  • Weisler, M.I. 2001. On the Margins of Sustainability: Prehistoric Settlement of Utrōk Atoll, Northern Marshall Islands. BAR International Series, 967. Archaeopress, Oxford.
  • Williams, A.N., Ulm, S., Goodwin, I.D. and Smith, M. 2010. Hunter–gatherer response to late Holocene climatic variability in northern and central Australia. Journal of Quaternary Science 25 (6): 831838.
  • Wilson, A.J.G. 2011. Pandanaceae. In A. Wilson (ed.), Flora of Australia, Volume 39: Alismatales to Arales, pp. 211235. ABRS/CSIRO, Melbourne.
  • Wilson, B., Wilson, C. and Baker, P. 1993. Australian Marine Shells 1: Prosobranch Gastropods. Part One. Odyssey, Kallaoo, WA.
  • Wilson, B., Wilson, C. and Baker, P. 1994. Australian Marine Shells 2: Prosobranch Gastropods. 2. (Neogastropods). Odyssey, Kallaoo, WA.
  • Woodroffe, C.D., Chappell, J. and Thom, B. 1988. Shell middens in the context of estuarine development, South Alligator River, Northern Territory. Archaeology in Oceania 23 (3): 95103.
  • Woodroffe, C.D., Samosorn, B., Hua, Q. and Hart, D.E. 2007. Incremental accretion of a sandy reef island over the past 3000 years indicated by component-specific radiocarbon dating. Geophysical Research Letters 34 (L03602): 15.
  • Woodroffe, S.A. 2009. Testing models of mid to late Holocene sea-level change, north Queensland, Australia. Quaternary Science Reviews 28: 24742488.
  • Yamaguchi, T., Kayanne, H., Yamano, H., Najima, Y., Chikamori, M. and Yokoki, H. 2005. Excavation of pit-agriculture landscape on Majuro Atoll, Marshall Islands, and its implications. Global Environment Research 9 (1): 2736.
  • Yokoyama, Y., Purcell, A., Marshall, J.F. and Lambeck, K. 2006. Sea-level during the early deglaciation period in the Great Barrier Reef, Australia. Global and Planetary Change 53: 147153.
  • Yu, K.F. and Zhao, J.X. 2010. U-series dates of Great Barrier Reef corals suggest at least +0.7 m sea level ∼7000 years ago. The Holocene 20 (2): 161168.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Shoalwater Bay Islands
  5. Otterbourne Island
  6. High Peak Island
  7. Collins Island
  8. Discussion
  9. Conclusion
  10. Acknowledgements
  11. References
  12. Supporting Information
FilenameFormatSizeDescription
arco5016-sup-0001-si.zip332K

Table S1. Square A data, Otterbourne Island Site 4.

Table S2. Square B data, Otterbourne Island Site 4.

Table S3. Plant remains, Square B, Otterbourne Island Site 4. X, presence; *, possible occurrence, highly degraded specimen.

Table S4. Bone data, Square B, Otterbourne Island Site 4.

Table S5. Fish bone identifications, Square B, Otterbourne Island Site 4.

Table S6. Square H10 data, High Peak Island Site 10.

Table S7. Square H11 data, High Peak Island Site 10.

Table S8. Square H12 data, High Peak Island Site 10.

Table S9. Square A data, Collins Island Site 24.

Table S10. Radiocarbon dates for archaeological sites, central Queensland islands. Radiocarbon dates were calibrated into calendar years using the online calibration program Calib 6.0 (Stuiver & Reimer 1993) and the Southern Hemisphere calibration dataset (ShCal04) (McCormac et al. 2004) for charcoal dates, and the marine calibration dataset (Marine09) (Reimer et al. 2009) for marine shell dates, using a ΔR value of 10 ± 7 years determined for central Queensland (Ulm 2002). A single, central best-point estimate was calculated for the irregular probability distribution of each date using the median calibrated age (following Telford et al. 2004).

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.