This chapter contains the archaeological site reports, including excavation and sampling methods, chronology, and the results of the analyses, for:

• Agajang-guwa
• Anamanba
• Aningarra’s Camp 1974
• Gulukula Mounds 1974, 1978
• Guna-jengga 1974
• Gupanga wangarr an-dakal a-yurra
• Ji-bena 1978, 1979
• Jilangga a-jirra
• Jinawunya 1978
• Jurnaka 1978
• Lorrkon a-jirrapa East 1974
• Lorrkon a-jirrapa West 1974, 1978
• Mu-garnbal 1973
• Muyu a-jirrapa Coastal Shell Midden 1974, 1978
• Muyu a-jirrapa Shell Mound 1974, 1978
• Ngarli ji-bama 1974
• Yuluk a-jirrapa Mounds 1974 (Map 3.1).

Site descriptions come from Shell Bed to Shell Midden (Meehan 1982a), the field notes of Betty Meehan (BM) and Rhys Jones (RJ), and the radiocarbon submission sheets. Gupanga, Lalarr gu-jirrapa and Ngarli ji-bama were home bases for the An-barra when Betty and Rhys did their fieldwork in 1972–73 (Figure 2.1, Map 2.2; Meehan 1982a, 26–9).

Chronology

Not all sites recorded by Betty and Rhys were selected for dating and analysis. The original dating of the excavation and sampling program undertaken in the 1970s was done by the ANU Radiocarbon Laboratory. In the 2000s, the Centre for Archaeological Research at ANU (CAR) provided funding for radiocarbon dates. Samples (shell and charcoal) were sent to the ANU Radiocarbon Laboratory and to the Waikato Radiocarbon Dating Laboratory in New Zealand. The results are listed in Table 3.1. Further dating and analysis of more sites is possible in the future, as there are suitable samples of charcoal and shell available, stored at MAGNT in Darwin.

Agajang-guwa

Site description

Agajang-guwa is an area just south of Jinawunya where there is a shell midden on a transverse dune chenier running east–west.

Sampling

Rhys took a Dosinia juvenilis sample from the lower level of the midden in 1978.

Chronology

The sample returned a date of 1,060–683 cal BP (ANU-11209) (Table 3.1).

Anamanba

Site description

Anamanba lies on the eastern side of Balpilja Swamp (Maps 2.1 and 3.1). There is a large earth mound composed of termite nest and some shell lenses. This mound, like many others in the area including Ji-bena (see below), is on the bank of a freshwater billabong.

Sampling

Anamanba mound was augured by Jones and Chappell in 1999. Shell samples were collected from 2.3 m below the surface of the mound.

Chronology

A sample of Geloina coaxans (ANU-11208) returned a date of 796–423 cal BP (Table 3.1).

Aningarra’s Camp 1974

Site description

Michael Aningarra and his family camped on this shell midden, which sits on a coastal sand dune at Gupanga – a contemporary home base on the west bank near the mouth of An-gartcha Wana (Figure 2.1, Map 3.1). The upper levels of the midden were still occupied in 1974.

Excavation

In 1974, after clearing the surface, Betty and Rhys excavated a 50 × 50 cm square down to sterile sand and took two solid samples: Upper Level (0–6 cm), Lower Level (20–25 cm). The midden itself was very dense and contained much fine charcoal and ash with many shells that were burnt and crushed. The total depth of the excavation was 25 cm. The samples were later sieved through 1.5 mm mesh (RJ 1974, FN Bk 1, 63–5).

Chronology

Lower Level was dated from a Dosinia/charcoal pair (ANU-2013). A date of 515–130 cal BP was obtained on shell and 505–0 cal BP on charcoal (Table 3.1).

Results

There is a range of both marine and mangrove shellfish present in the upper and lower levels (Appendix 2). The dominant shellfish species by MNI are Dosinia and Marcia (Table 3.2), by NISP Dosinia and Modiolus and, by weight, Dosinia and Mactra. All are marine species, which dominated in both the upper and lower levels, with slightly more emphasis on marine in the lower levels. Mangrove species remain consistently less in both upper and lower levels (Figures 3.1–3.3).

Gulukula Mounds 1974, 1978

Site description

The Gulukula chenier is the base for several large discrete shell mounds up to 5 m in height and 30 m in diameter (Map 2.1, Figure 3.4). Known as the Gulukula or Dog Mounds, they are the most famous on An-barra land. They lie inland, about 1 km from Guna-jengga on the present-day coast, and 2 km west of An-gartcha Wana (Map 3.1). “According to the An-barra, at some time in the past, the first dog in the country piled up these extensive mounds of shells with his paws, as dogs do with earth when they are digging a hole” (Meehan 1982a, 167).

Frank Gurrmanamana described the mounds as:

Excavation

Gulukula was sampled twice, in 1974 and 1978. In 1974, Betty excavated a 50 × 50 cm test pit, 10 m from the top of one of the mounds that she describes as being the largest of the Gulukula shell mounds, 5 m high and 50 m wide. The pit was 25 cm in depth. Two solid samples were taken, Bags 1 and 2, as well as a dating sample from the bottom of the pit. There was shell only, no charcoal being obvious (BM 1974, FN Bk 2, 61). The deposit was sieved through 4 mm mesh. In 1978, a solid sample was taken 10 cm below the surface inside a goanna burrow (BM 1978, FN Bk 5, 29). Jones and Chappell took another sample from 25 cm depth in 1999.

Chronology

From Gulukula 1974, the shell (Dosinia) sample taken from 25 cm gave a date of 875–465 cal BP (ANU-2021). From Gulukula 1978, a shell sample (Dosinia) taken from 80 cm depth was dated to 666–392 cal BP (Wk-24232) (Table 3.1; Meehan 1983). In 1999, Jones and Chappell took another sample that was dated to 641–312 cal BP (Blyth 299) (Table 3.1).

Results

Gulukula 1974

Both marine and mangrove species (Appendix 2) were in the sample taken from the top of a Gulukula Mound in 1974. The dominant species by MNI, NISP and weight were marine, including Dosinia, Modiolus and Mactra (Table 3.3). In the sample taken from the side of a Gulukula Mound, Dosinia was the dominant species, and the other species appeared in negligible quantities.

Gulukula 1978

Like the 1974 samples, the dominant species by MNI, NISP and weight were marine, dominated by Dosinia. The other species appeared in negligible quantities (Tables 3.4 and 3.5).

 

Guna-jengga 1974

Guna-jengga, 3 km west of Lalarr gu-jirrapa, is associated with an important fish trap, built across a creek flowing into the sea, which was used in the mid-dry season (Map 3.1; Meehan 1982a, 36, 113, 166). Frank Gurrmanamana said it is rrawa, a remembered camp site occupied during the “cold weather time” (i.e. not a Dreaming site) (RJ 1973, FN Bk 21, 26).

Rhys described Guna-jengga and the fish trap back in 1972:

Betty described the use of the fish trap in April 1973:

She went on to say, however, that it was not always reliable:

Site description

The archaeological site of Guna-jengga is a coastal midden consisting of a modern “dinner time” camp overlying an older midden deposit (Figure 3.5). The site is about 200 m from the beach where there is a well behind the beach berm.

Rhys described the site in 1973:

In 1974, Rhys reflected on storm damage at Guna-jengga:

Sampling

In 1974, Rhys sampled the midden:

Chronology

Rhys speculated on the age of the site when he took the sample: “My guess is that the date will prove to be significantly different from ‘modern’ but will not have a large antiquity – i.e. of the order of 200–400 yrs BP” (RJ 1973, FN Bk 23, 4). He proved to be correct. A Dosinia/charcoal pair (ANU-2020) returned an age of 495–0 and 417–0 cal BP respectively (Table 3.1).

Results

Again, the dominant shellfish is marine, Dosinia (Table 3.6, Figure 3.7).

Gupanga wangarr an-dakal a-yurra (where the Dreaming white ochre representing diyama lies)

This site is a fossilised shell bed cemented by coarse calcareous sand, containing a mix of shells including Dosinia and Marcia, exposed intertidally 400 m north of Gupanga (Map 3.1; Meehan 1982a, 56).

Rhys described it as follows:

Betty notes:

Chronology

A Dosinia sample (ANU-2024) returned an age of 1,695–1,214 cal BP (Table 3.1).

Ji-bena 1979

The earth mound Ji-bena is in open woodland adjacent to the freshwater swamp, Balpilja, and close to the mangroves lining An-gartcha Wana. It is now 8 km from the sea (Maps 2.1, 2.2 and 3.1, Figure 3.8). It is a large circular earth mound approximately 80 m in diameter and 2 m high (Figure 3.9). According to Frank Gurrmanamana, it was occupied in the late wet season (RJ 1979, FN Bk 1, 79).

Excavation

In 1979, Betty and Rhys excavated a metre square test pit in 15 spits that reached 1.65 m in depth (Figures 3.10 and 3.11) (BM 1979, FN Bk 1, 51–89; RJ 1979, FN Bk 1, 57–9, 82–9).

Chronology

Ji-bena was dated to 1,284–790 cal BP (ANU-2817), 10 cm above base. Four other sequential dates were obtained, the youngest being 710–333 cal BP (ANU-2815), 50 cm below the surface (Table 3.1; Brockwell et al. 2005; Meehan 1988a; White et al. 1990).

Results

Shell, bone and stone artefacts were recovered from the Ji-bena excavation. Relative quantities are recorded in Table 3.7. Two pieces of glass were also found in Spit 1, maximum depth 8 cm.

Shell

Twenty taxa were identified: the marine taxa Dosinia, Modiolus, Mactra, Nassarius, Tellina, Volema, Tegillarca, Marcia, Saccostrea, Volegalea, Neverita, Melo amphora, Cymbiola and Syrinx; the mangrove taxa Cerithidea, Telescopium, Terebralia, Geloina, Cassidula; and the terrestrial taxon Xanthomelon. The relative proportions of taxa are shown in Table 3.8. They are listed with the numerous species first. Dosinia juvenilis dominated the assemblage by MNI and weight. Marine shellfish dominated until c. 700 years BP when they disappeared from the assemblage (Figure 3.12).

Other faunal material

A total of 206.1 g of bone was recovered from the Ji-bena excavation. Most bone was found in the upper part of the excavation post c. 500 years BP. The distribution is recorded in Table 3.9 and Figure 3.13.

Turtle was the dominant taxon, followed by mammal, fish, reptile and bird (Table 3.10). Identified mammal bone includes the teeth and jaws of the northern brushtail possum (Trichosurus arnhemensis), northern brown bandicoot (Isoodon macrourus), an unspecified rodent and macropod. Fish bones include the vertebrae, spines and cranium of two common north Australian fish, the fork-tailed catfish (Arius sp.) and barramundi (Lates calcarifer). Both species can live in either fresh or saltwater. Neither bird (long bones) nor reptile remains (vertebrae) could be identified to family or species level. Freshwater taxa (turtle) became dominant in the assemblage c. 700 years ago until the recent past (Figure 3.14).

Stone assemblage

Altogether there were 71 stone artefacts weighing 1,496.9 g recovered from the excavation at Ji-bena, with an average density of 46.5 artefacts per cubic metre. The artefacts were concentrated in the upper part of the site post 500 years BP (Table 3.11, Figure 3.15). The most common item was flaked quartzite, including one quartzite bipolar core. There were also some ground ochre, haematite and sandstone artefacts (Table 3.12).

Jilangga a-jirra

Jilangga a-jirra is located 3 km from the eastern bank of An-gartcha Wana, and about 1 km from the coast (Map 3.1). There is a thin shell midden scatter, located on top of an inland vegetated dune running parallel to the coast. It lies adjacent to a large shell mound, which is a wet season camp belonging to the Gu-lalay who are related to the An-barra. Stone tools were observed on the surface of the midden, along with mammal and fish bones (Meehan 1982a, 38, 163).

Sampling

A dating sample was taken 15–20 cm from the top of the midden, which had been dug out to make a cooking pit for six wallabies.

Chronology

A Dosinia/charcoal pair (ANU-2022) returned dates of 1,235–703 cal BP and 452 to modern cal BP (Table 3.1). The discrepancy in dates could be due to sampling from disturbed deposit.

Jinawunya 1978

Jinawunya is a coastal area on the western side of An-gartcha Wana, on the western margin of An-barra land near the Gulukula shell mounds (Map 3.1, Figure 3.16; Meehan 1982a, 12). It is a camping place, popular for shell-fishing and fishing, and for collecting Pandanus nuts, vegetables, fruit and wood from trees on the chenier dune behind the beach (Map 2.1). Rhys and Betty made an archaeological transect here (Meehan 1982a, 107, 110–11, 165; BM 1978 FN, Bk 5, 39–44). No excavation was undertaken at Jinawunya but dating samples were taken from the large shell midden on a dune adjacent to the beach.

In 1983, Rhys collected a shell sample (Dosinia juvenilis) from 20–30 cm below the surface of the midden. We visited Jinawunya in 2015 and collected a shell sample (Anadara antiquata) from the surface of the midden, which had been half eroded away by wave action (Figure 3.16). This was the same place where doctoral student Bethune Carmichael (2015) interviewed Betty Ngurrpangurrpa for his film Places in Peril regarding her perceptions of climate change. We also collected recently dead shells to be compared with archaeological samples in an isotope analysis study to investigate late Holocene changes in sea surface temperatures.

Chronology

The 1983 shell sample (ANU-11206) was dated to 740–430 cal BP. The 2015 sample (Wk-42262) returned a date of 40–0 cal BP (Table 3.1). The discrepancy in dates is likely due to sampling, as the 1983 sample was taken from below the surface and the 2015 sample from the surface.

Jurnaka 1978

Jurnaka (meaning corner) is located on the coast, west of the mouth of An-gartcha Wana in a mangrove area, near where Jurnaka Creek flows into it (Map 3.1). A shell mound, measuring 24 × 25 m, was deposited on a small natural rise on the coast. A nearby modern midden contained diyama (Marcia) and mussels on the surface. The maximum depth of the mound was 0.5 m. There was clean sand underneath, then dense midden to the surface (BM 1978, FN Bk 5, 21–6).

Although oysters (Saccostrea) were not important to the An-barra diet in terms of quantity, they were highly prized and sought after. Called an-guljaraba, they come from mangrove areas around the mouth of An-gartcha Wana, one of the most popular areas being Jurnaka. Betty explains:

Sampling and excavation

The following description of the excavation of the shell mound is taken from Meehan (BM 1978, FN Bk 5, 21–6). 

Chronology

A Dosinia sample was taken from the base of the shell mound and dated to 373–14 cal BP (Wk-24234) (Table 3.1).

Results

The dominant shellfish was Dosinia with other species appearing in low quantities (Table 3.13, Figure 3.17).

Lorrkon a-jirrapa East and West 1974

The Lorrkon a-jirrapa coastal shell middens are located on a chenier ridge 3 km from the western bank of An-gartcha Wana and 2 km from the coast (Maps 2.1 and 3.1). The northern edge is bordered by a salt clay pan, and the southern side is black soil plain. A jungle vine thicket is situated at its base. The midden area is extensive, over 400 m long and 3 m high.

Betty and Rhys describe the site:

Sampling

At Lorrkon A-Jirrapa West, Betty and Rhys excavated a 50 cm square test pit to a depth of 30 cm in the middle of the midden. A bulk sample was taken from 20–25 cm.

At Lorrkon a-jirrapa East, a midden sample was taken 10–20 cm below the surface (RJ 1974, FN Bk 2, 52).

Chronology

A Dosinia/charcoal pair (ANU-2012) was taken from the bulk sample at Lorrkon a-jirrapa West 1974 and returned dates of 1,179–750 and 1,345–740 cal BP respectively (Table 3.1).

Results

In both the East and West shell midden samples, Dosinia was the dominant shellfish. However, Lorrkon a-jirrapa East also had a high proportion of Modiolus. Both samples showed mangrove exploitation (Table 3.14).

Lorrkon a-jirrapa West 1978

Lorrkon a-jirrapa West was surveyed and excavated in 1978 (RJ 1978, FN Bk 1, 54; BM 1978, FN Bk 5, 36–9).

Excavation and sampling

The following description of the excavation comes from Betty’s field notes (BM 1978, FN Bk 5, 2–3).

Chronology

A shell sample (Dosinia juvenilis) (Wk-24233) was taken from Spit 5 (80–90 cm) at the base of the excavation and dated to 1,166–821 cal BP. Jones and Chappell took two additional shell samples from the midden cap in 1999. They were dated to 1,433–1,046 (Blyth 499) and 1,228–772 cal BP (Blyth 599) (Table 3.1).

Results

Only the shellfish from Lorrkon a-jirrapa were analysed, not the fish and mammal bone, stone artefacts or worked shell mentioned by Rhys in his field notes. They are available at the MAGNT in Darwin.

There are a wide variety of marine and estuarine shellfish represented in the Lorrkon a-jirrapa West 1978 excavation. However, Dosinia and other marine species dominate all levels by MNI, NISP and weight (Table 3.15, Figure 3.18).

 

Mu-garnbal 1973

The site of Mu-garnbal lies on the east side of An-gartcha Wana, about 5 km from the coast at the juncture of the mangroves and the black soil plain (Map 3.1).

Mu-garnbal is owned by the Martay. Betty describes the site and its environment:

The site is named after a canoe as described in Rhys’ field notes:

Sampling

Betty and Rhys recorded Mu-garnbal and took a sample from its low dense midden on 27–29 June 1973 (Meehan 1975, Appendix 4, p. A33).

Chronology

A Tegillarca granosa sample (Wk-17746) was dated to 3,724–3,345 cal BP (Table 3.1).

Results

The sample from Mu-garnbal is dominated by Tegillarca, a marine species, with some mangrove species (Table 3.16).

Muyu a-jirrapa Coastal Shell Midden 1974, 1978

Site description

Muyu a-jirrapa lies on the west bank of An-gartcha Wana, 1 km from its mouth and c. 1 km north of Gupanga. There is a linear midden 500 m long (Figure 3.19), with a shell mound behind (Figure 3.20). The midden was stratified on top of a sand dune and covered by jungle thicket in some parts. The edge of the midden was being eroded by the river. It was recorded by Betty and Rhys in 1974 and excavated in 1978 (Figures 3.21 and 3.22). It was still used occasionally as a camping and cooking site in 1973–74.

The site was described by both Betty and Rhys:

By 1979, the midden was mostly destroyed: “The sea had eroded it [with] two large scallops into the beach/midden edge – last Barra [wet season] – no doubt held up by uprooted trees – will go next Barra” (RJ 1979, FN Bk 1, 26).

Sampling 1974

The shell midden was sampled for dating in 1974:

Rhys took three solid samples and one small extra carbon sample from a depth of 20–25 cm (RJ 1974, FN Bk 1, 84).

Survey and excavation 1978

The site was surveyed and excavated in 1978 (BM 1978, FN Bk 5, 45–6) (Figure 3.22). The excavation is described by Betty in her field notes (BM 1978, FN Bk 5, 55–73):

Chronology

A date of 1,392–987 cal BP (ANU-2014) was obtained from the 1974 Dosinia sample (Table 3.1).

Results

Dosinia was the dominant shellfish, both by MNI and weight, with some mangrove species (Tables 3.17 and 3.18, Figure 3.23).

Muyu-a-jirrapa Shell Mound 1974, 1978

Site description

Muyu a-jirrapa Shell Mound is small with a dense deposit of shell, located on the western bank of An-gartcha Wana about 2 km from its mouth. Rhys described the site:

Sampling 1974

Excavation 1978

Muyu a-jirrapa Shell Mound was excavated by Betty in 1978 in a metre square (BM 1978, FN Bk 5, 81–2, 88–99). The excavation reached a depth of 80 cm in eight spits onto clean sand.

Chronology

Rhys speculated on the date: “It is quite possible that this mound is older than the other Kopanga middens tested” (RJ 1974, FN Bk 1, 90). And, indeed, he was correct. A Dosinia juvenilis sample taken from the lowest level of the 1978 excavation yielded a date of 1,356–867 cal BP (ANU-2816) (Table 3.1).

Results

Unfortunately, this assemblage was lost in the Canberra bushfires of 2003 when ANU’s Weston store was burnt to the ground and was never analysed. However, we do have Betty’s notes from the excavation that give some idea of the contents of the shell mound. Shells included:

• Ana-mula an-ika [Terebralia]
• An-bambula [Geloina]
• An-demburela [Tellina]
• An-dirrbula [Dosinia]
• An-guljuraba [Saccostrea]
• An-jalabaykarda [Melongenidae]
• Diyama [Marcia]
• Ngandipurdurda [Tegillarca]
• Nornda [Telescopium]
• Ana-jirralanggula [Cymbiola]

Some shells were very small, especially an-dirrbula. Rhys’ observation from his 1974 sample is that the mound contained “mostly small Anderabula” (RJ 1974, FN Bk 1, 90).

Ngarli ji-bama 1974

Site description

Ngarli ji-bama is a home base on an inland dune associated with fossil shell beds. It is located 4 km east of An-gartcha Wana, and about 1 km from the coast, and is occupied in the dry season (Figure 2.1; Maps 2.2 and 3.1). It has a good supply of water from wells surrounding the site. Although it lies close to the sea, the main shell beds (Modiolus micropterus [now modulaides]) lie 3 km away (Meehan 1982a, 31, 66–7). It is a large open site with a series of shallow midden deposits strung around the circumference. They contain burnt clay, hearth material and shells.

Sampling

“Rhys went off to have a look at some old sites (and, hopefully to take some shell samples)” (BM 1974, FN Bk 3, 2). In 1974, a solid sample was taken from a midden by a banyan tree, one of a series of shallow midden deposits from the large open site. It was sieved through 1.5 mm mesh.

Chronology

A Mactra abbreviata sample was dated to 286–0 cal BP (Wk-17747) (Table 3.1).

Results

The sample was dominated by Mactra abbreviata (Mytilidae, a marine species) (Table 3.19).

Yuluk a-jirrapa Mounds 1974

Yuluk a-jirrapa contains an impressive series of shell mounds, 1 km inland, adjacent to the home base of Ngarli ji-bama, on the eastern side of An-gartcha Wana (Map 3.1). The An-barra say they were formed by the large stingray Yuluk, who made these mounds as he flapped his way along the dune (Meehan 1982a, 167–8). The mound that was sampled was 24 m in diameter and 1.5 m high.

Sampling

Rhys took a sample at a depth of 10–15 cm.

Chronology

A Dosinia juvenilis/charcoal pair (ANU-2023) was dated to 636–165 and 240–23 cal BP respectively (Table 3.1).

Results

The dominant shellfish at Yuluk a-jirrapa was Coecella horsfieldii, a marine bivalve species found in intertidal coarse sand (Table 3.20; Meehan 1982a, 168).