Skip to content
Frontiers
Search

Original Research

Frontiers in Environmental Archaeology ·

Volume 5 - 2026 | https://doi.org/10.3389/fearc.2026.1867558

MicroCT analysis of archaeological parenchyma identifies underground storage organs from late Holocene contexts at Birriwilk rockshelter, Manilikarr Country, northern Australia

  • Jenifer Pritchard1*corresponding author
  • Denis Shine2
  • Aleese Barron1
  • Rachel Wood3
  • Alfred Nayinggul4
  • Tim Denham1
  1. 1School of Archaeology and Anthropology, Australian National University, Canberra, ACT, Australia
  2. 2Irish Heritage School, Offaly, Ireland
  3. 3Oxford Radiocarbon Accelerator Unit (ORAU), School of Archaeology, University of Oxford, Oxford, United Kingdom
  4. 4Gunbalanya, NT, Australia
Article metrics0 viewsView details
PDFCiteMetrics

Abstract

MicroCT analysis of archaeological parenchyma fragments has identified Aboriginal exploitation of underground storage organs (USOs) at Birriwilk rockshelter, Arnhem Land, northern Australia. A variety of USO types were identified from late Holocene contexts, including bulbs/corms, dicotyledon primary/tap roots and monocotyledon stem tubers. The range of USOs exploited is consistent with ethnographic accounts of plant exploitation practices by Aboriginal communities in this region. Although lacking taxonomic specificity, these results demonstrate the value of using microCT to investigate charred archaeological parenchyma in order to understand past human use of soft-tissue, plant storage organs.

Keywords

Introduction

Abundant archaeological parenchyma was recovered during an archaeological excavation at Birriwilk rockshelter in Manilikarr Country, Arnhem Land, northern Australia (Shine 2014; Shine et al. 2013). Archaeological parenchyma ordinarily refers to vegetative storage parenchyma cells, which include underground storage organs (USOs) such as tubers, bulbs and corms that store starch and other carbohydrates (Hather 2000:1). However, parenchymatous tissues are found throughout plants and form a ground tissue around other more specialised plant tissues, such as vascular bundles, fibres and secretory bodies. Parenchyma cells are generally undifferentiated, thin walled and isodiametric. Parenchymatous tissues from USOs are the focus here.

A selection of charred archaeobotanical fragments from Birriwilk was subject to non-destructive microCT scanning, with subsequent virtual histological visualisation of internal structures. Virtual histology enabled comparison of archaeological parenchyma to modern reference specimens to enable the identification of different USO types, namely organographic determinations, with a view to eventually enabling taxonomic identification. Although taxonomic determination to genus or species level has not been possible, numerous fragments have been assigned to a type of parenchymatous storage organ whether bulb or corm, dicotyledon (dicot) primary or tap root, and monocotyloden (monocot) stem tuber.

The archaeobotanical project was designed to shed light on USO exploitation during occupation of the rockshelter in the late Holocene, thereby providing time depth for the exploitation of USOs by local communities documented ethnographically. Further, the project was designed to evaluate microCT as a method for the investigation of archaeological parenchyma, including the creation of a modern reference collection of USO-bearing species known to be exploited in the region. Although numerous ethnographic accounts and rock art depictions illustrate the significance of USOs to Aboriginal communities in northern Australia (Hammond 2019; Veth et al. 2018), there is relatively sparse archaeobotanical evidence for the exploitation of these plant parts (Florin et al. 2020, 2022). The results from Birriwilk demonstrate the value of using microCT analysis to document archaeological parenchyma and to reconstruct continuities in Aboriginal plant exploitation practices over millennia. More broadly, the research shows the need to revitalise research on archaeological parenchyma to provide more balanced and complete reconstructions of how people used plants in the past.

Cultural significance of USOs in northern Australia

Ethnographic and historic accounts across Arnhem Land document many aspects of Aboriginal culture, including the cosmological, ritual and dietary significance of USOs. Here, there is only sufficient space to allude to the centrality of USOs to various aspects of social life, with specific reference to the East Alligator River region in which Birriwilk is located.

Foremost, Berndt and Berndt (1989: 57–59) document the story of the ancestors Biriwilg [Birriwilk] and Ngalmoban, who both carry USOs of various types. Biriwilg is associated with lily roots (male version) and long yams (D. transversa; female version), while Ngalmoban carried man-gindjeg, bitter yams (D. bulbifera), which she spread throughout the landscape: “She was throwing them about at different places so they would grow there, and naming the places as she did so” (Berndt and Berndt 1989: 58). The narratives go on to describe how Biriwilg became embodied into a rock art image of herself, thereby giving her name to the rockshelter.

Another creation story from the Coburg Penisular and Alligator Rivers region describes the first ancestor Imbrerombera who leaves yams and spirit children as she travelled through the country. Two spirit children were left north of the mouth of the East Alligator River with murarowa, an edible ‘false yam’ of the Cyperus genus (Spencer 1914:277). A rite of passage to manhood on Melville Island conducted over several days involved the ritual use of hairy yams (Dioscorea sp.), itjimma kolamma, that were soaked overnight and cooked the following day. Here the yam was considered ‘hot’ and not good for eating though it was more widely eaten on the mainland (Spencer 1914:92–107).

During his 1847 expedition to Arnhem Land, Ludwig Leichardt observed people fanned out over the plains utilising a range of USOs. The USO types recorded included: swamp sedge rhizomes (probably Eleocharis dulcis) called allamurr in the East Alligator River region or marnatt near the coast; a species of Convolvulus, possibly native Ipomoea graminea, called imberbi on the Coburg Peninsula; and, the consumption of the lower stalks of Nelumbium, possibly Nymphaea sp. (Schrire 1982:20). Late in the dry season of 1874, Lewis (1922:131) recorded “great stacks” of yams piled up near the East Alligator River to be stored for the winter. Spencer (1914:27) also described the collection of lily roots called parijiliji (possibly Nymphaea sp., Nymphoides sp. or Nelumbo nucifera) by women and children in the Alligator Rivers region.

Baldwin Spencer (1914) recorded several ways of preparing yams for consumption. In Gagudju Country, a common method for cooking involved an earth oven called peindi. A pit of 1–2 feet deep was dug with a digging stick and lined with rounded stones called wairbi. A fire was created in the pit to heat the stones. Grass or green twigs were laid over the heated stones and the food was placed on top. The pit was then covered with earth and left until the food was cooked (Spencer 1914:393). Various treatments were used to prepare yams that were considered ‘hot’, namely acrid or toxic. In Wardaman Country, nongura was sliced, placed on an open fire for 15–30 min, pounded on stone and made into cakes, tjalangi. The cakes were cooked in the fire, pounded and cooked twice more before they were eaten. Another yam, magolu, was skinned and sliced, put in a basket and soaked for some hours, then pounded, made into cake and cooked in an earth oven (Spencer 1914:394). Some USO species utilised by the Gidjingali of the Blyth River included Ipomoea, three taxa of water lily (Nymphaea spp.), yams (Dioscorea transversa, D. bulbifera) and the spike rush (Eleocharis dulcis) (Jones and Meehan 1989:120–122).

Increasingly, Indigenous groups in northern Australia are recording their own biocultural heritage, namely the traditional knowledge relating to plant exploitation, often in collaboration with anthropologists and ethnobotanists (Clarke 2007; Puruntatameri et al. 2001; Roberts Garalnganyjak et al. 2011; Smythe and Van Sturmer 1985; Widijburru et al. 2010; Wightman and Jackson 1991; Winydjorrotj and Jawoyn Association 2005; Yunupinu et al. 1995). Despite the clear significance of USOs to Aboriginal lifeways in Arnhem Land, there is limited archaeobotanical information on how significant these plants were to people's diet in the past.

Archaeological investigation of USOs in northern Australia

Following the pioneering work of Hather (1993, 1994, 2000), globally there has been a relative neglect in the systematic study of archaeological parenchyma in comparison to other types of archaeobotanical remains, such as seeds, nuts and wood charcoal (Pearsall 2000). Notable exceptions include the exploitation of USOs by early modern humans in Africa (Wadley et al. 2020) and on Borneo (Barton et al. 2016), by Mesolithic communities in northwestern Europe (Kubiak-Martens 2002, 2006, 2016), by pre-agrarian communities in Southwest Asia (Arranz-Otaegui et al. 2018), as well as the recent adoption of domesticates introduced to New Guinea (Barron et al. 2022a; Pritchard et al. 2019). The research documented here is part of a broader resurgence of interest in the investigation of archaeological parenchyma, USO exploitation in the past, and the domestication of vegetatively propagated crops that collectively will provide a more balanced view of plant exploitation in the past (Denham et al. 2020).

Ethnographic information concerning the exploitation of USOs in Australia is not well corroborated by archaeological evidence. Exceptions are numerous examples of tubers and yams in many guises depicted in rock art across northern Australia, including: naturalistic yams, yams associated with human figures, anthropomorphised yams and other yam beings (yam birds, snakes and crocodiles) (Chaloupka 1993:138–145; Jelinek 1989:110, 185, 198, 368–389; Lewis 1988:271–301; Welch 2003). These depictions and transmogrifications demonstrate the ceremonial, totemic and mythological importance of USOs across extended time frames. Various rock art depictions of USOs have been taxonomically identified as Nymphaea species, Dioscorea transversa, Dioscorea bulbifera and Cyperus bulbosus (Hammond 2019; Veth et al. 2018).

The archaeobotany of macrobotanical remains in Australia has primarily focussed on the exploitation of seeds and nuts (Asmussen and McInnes 2013; Atchison et al. 2005; Clarkson et al. 2017; Cosgrove et al. 2007; see Denham et al. 2009a for review). Limited archaeobotanical recovery in the field and poor preservation conditions for plant remains in tropical regions contribute to the paucity of macrobotanical studies in northern Australia. Among the exceptions are well preserved archaeobotanical material at Madjedbebe (Clarkson et al. 2015, 2017; Florin et al. 2020, 2022) and at Anbangbang I; the latter includes remnants of unidentified desiccated tuber skins and rootlets of Nymphaea and Triglochin (Clarke 1985, 1989).

Excavation and sample selection

The excavation of Birriwilk was conducted as part of a community-led archaeological project (Shine 2014; Shine et al. 2013). The project entailed the excavation of three rockshelters on Manilikarr Country that were identified as being culturally significant by a Senior Traditional Owner, NaGodjok (now deceased), a member of the Nayinggul family. The intention of the original investigation was to provide archaeological time-depth for practices that were remembered as occurring at each site.

A single 1 m2 test unit was excavated in arbitrary excavation units (XUs) averaging 2.2 cm thick in cultural deposits, which increased to 3.6 cm thick from XU 29 downwards through deposits without cultural materials (Figure 1). Excavated material from all XUs was dry-sieved through 2.1 mm mesh, with the finer fraction from XUs 1–23 subject to flotation using 250 µm mesh to maximise archaeobotanical recovery. In sum, ephemeral human activities at Birriwilk predate 4,500 cal BP, with more sustained use occurring after 750 cal BP and only infrequent site visitation over the last 200–250 years (Shine et al. 2013). The major period of use from 750 to 250/200 cal BP is interpreted to comprise hunting and foraging in local wetlands, a technological shift to bifacial quartzite points, and increased artefact discard rates, as well as ochre grinding (Shine et al. 2013). The investigation of archaeological parenchyma at Birriwilk that is reported here sheds light on how people used USOs as part of their foraging activities in the distant and recent pasts.

Archaeological investigations at Birriwilk: upper—photographs of **(A)** rockshelter prior to excavation, **(B)** rock art panel on backwall of rockshelter, and **(C)** test unit during excavation with base of XU27 exposed; lower—stratigraphic sections of test unit walls depicting XUs and previous radiocarbon samples (Shine et al. 2013: Figure 4).
Figure 1. Archaeological investigations at Birriwilk: upper—photographs of (A) rockshelter prior to excavation, (B) rock art panel on backwall of rockshelter, and (C) test unit during excavation with base of XU27 exposed; lower—stratigraphic sections of test unit walls depicting XUs and previous radiocarbon samples (Shine et al. 2013: Figure 4).

Macrobotanical materials (>2.1 mm) were sorted into categories of macrocharcoal, nuts, seeds, archaeological parenchyma and other organic remains. The resultant archaeological parenchyma portion is primarily associated with human use of the rockshelter, namely to be cultural, because of general correspondences to, and hence associations with, frequencies of faunal remains, stone artefacts and macrocharcoal throughout most of the record (Figure 2; Supplementary Table S1). Of note, though, these general correspondences differ for the last 200–250 years, when microcharcoal peaks and stays relatively high while other cultural indicators decline towards the present. The macrocharcoal signal for the last 200–250 years plausibly represents burning in the immediate landscape of the rockshelter rather than cultural use of the site itself.

Multiplot depicting distributions of archaeological parenchyma (×100 in grams), lithic artefacts (by number), macrocharcoal and bone (by weight in grams) against excavation units (XUs 1–38). Solid lines show general trends. Previous radiocarbon dates at Birriwilk from Shine et al. (2013: Table 1); see Supplementary Table S1 for datasets and Supplementary Figure S13 for multiplot of new and old radiocarbon dates.
Figure 2. Multiplot depicting distributions of archaeological parenchyma (×100 in grams), lithic artefacts (by number), macrocharcoal and bone (by weight in grams) against excavation units (XUs 1–38). Solid lines show general trends. Previous radiocarbon dates at Birriwilk from Shine et al. (2013: Table 1); see Supplementary Table S1 for datasets and Supplementary Figure S13 for multiplot of new and old radiocarbon dates.

The archaeological parenchyma portion of the archaeobotanical materials collected from Birriwilk comprised 397 fragments. Initially, the surface morphology of each archaeological parenchyma fragment was studied under reflected light microscopy (RLM) using a Leica M80 stereo microscope (visual magnification ranges of 7.5×–60× with IC80 camera) or Leica MZ FLIII fluorescence stereo microscope (visual magnification ranges of 8×–100× with AxioCam HR3 camera). The initial RLM assessment of archaeological parenchyma fragments was designed to visually identify samples with potentially diagnostic features that would aid organographic or taxonomic identification using microCT. Archaeological parenchyma samples were not deliberately fractured or cut using microtome or hand-held razor blade during RLM assessment.

Fifteen fragments with a variety of different cell types, configurations and other surficial features were subsequently selected for microCT scanning using guidelines proposed by Hather (1993:3–8, 1994:73–74, 2000) (Supplementary Table S2). While the presence of parenchyma cells of vegetative storage organs characterises archaeological parenchyma; additional features, such as unique cellular structures, cell anatomy and overall morphology of a fragment are often important for identification to organographic and taxonomic groupings.

MicroCT scanning and visualisation

In general, preparing specimens for imaging and archaeobotanical identification by various forms of light and electron microscopy techniques is usually destructive, often requiring preparation of a freshly cut surface using microtome or razor blade for imaging. X-ray microcomputed tomography (microCT) offers a potentially non-destructive technique providing greater sample coverage for imaging, internal and external visualisation, analysis and identification that is gaining wider usage in archaeobotanical research (Barron et al. 2017, 2020; Barron et al. 2022b; Barron and Denham 2018, 2022; Pritchard et al. 2019; Whitau et al. 2017).

The Birriwilk archaeological parenchyma was scanned at the National Laboratory for X-ray Micro Computed Tomography (CTLab) at the Australian National University using a HeliScan MicroCT system (large volume helical-CT) (Latham et al. 2008; Myers et al. 2011; Varslot et al. 2011). Due to size variation 1–4 fragments were mounted in PVC sample tubes of diameters 10–14 mm and stabilized with packing foam. These were then scanned at 60 kV and a current of 60 mA with a 3 mm aluminium filter for durations of between 10 and 15 h, resulting in voxel sizes of between 2.9 micrometre/micron (μm) and 4.9 μm. The resolution of microCT data is influenced by the diameter of a sample, where a reduction in resolution occurs with an increase in sample size. Any decrease in resolution can cause fine details to become blurred or lose sharpness.

Images of archaeological parenchyma were rendered from the microCT scans using Drishti (v2.6.3; Limaye 2012) to simulate images obtained under reflected light microscopy (RLM) and scanning electron microscopy (SEM). All analyses were non-destructive; namely, they were undertaken without physical dissection of the archaeobotanical parenchyma fragment.

The digital images rendered from the microCT scans of the archaeological parenchyma fragments were initially examined with a virtual slice just below the surface of the fragment to reveal the cellular orientation within the fragment. Virtual slices were then ‘cut’ along transverse and longitudinal cellular planes (which are at right angles to each other) throughout the fragment to determine if there were diagnostic structural, anatomical and overall morphological characteristics. Individual snap shots were taken of the virtual slices at representative points within the fragment to show potentially diagnostic internal features and, ideally, to provide virtual images that would enable comparison with dissected specimens under RLM and SEM. However, it is acknowledged that the resolution of microCT analyses here, namely 2.9–4.9 µm, is not equivalent to the resolution attainable using SEM, namely, c. 10 nanometres (nm). Consequently, microCT visualisations and SEM images are not comparable, yet they are complementary. The resolution of microCT used here can characterise the overall morphology and internal arrangement of anatomical features, such as the distribution of vascular tissue, biogenic crystals and cellular orientation; SEM provides greater definition in the investigation and characterisation of individual anatomical elements.

Modern reference collection

Candidate species of USOs known to have been utilized in northern Australia were evaluated from ethnobotanical literature (Table 1; Atchison and Head 2012; Clarke 2007; Jones and Meehan 1989; Maiden 1889; Puruntatameri et al. 2001; Roberts Garalnganyjak et al., 2011; Smythe and Van Sturmer 1985; Widijburru et al., 2010; Wightman and Jackson 1991; Winydjorrotj and Jawoyn Association 2005; Yunupinu et al., 1995). This information was cross-referenced against the species of USOs growing within a ten kilometre radius of the rockshelter recorded in the Atlas of Living Australia (http://www.ala.org.au, accessed March 3, 2018). Desiccated specimens of most candidate taxa were obtained from the Australian National Herbarium in Canberra, the Northern Territory Annex, and charred specimens from Jon Hather's Archaeological Parenchyma Collection at the Institute of Archaeology University College London (subsampled by Tim Denham and Aleese Barron in 2017). The modern reference collection comprised 20 species known to produce edible USOs in northern Australia, including monocot tubers, rhizomes, corms and bulbs, dicot tubers and tap roots (Supplementary Table S3). Reflected light microscopy (RLM), scanning electron microscopy (SEM) and microCT imagery of reference specimens demonstrating the different types of USOs growing in the region are presented in Supplementary Figures S1–S12.

Table 1. Kundjeyhmi plant names for modern reference specimens (source primarily Fox and Garde 2018).

Taxonomic classificationCommon name(s)Kundjeyhmi nameFox and Garde 2018: page #Figure #
Cyperus bulbosusBush onion––S1
Colocasia esculentaTarodjamandarr/milekena126S2
Amorphophallus galbraSnakeskin lilyCheeky yaman-didjkanku119S3
Vigna lanceolataPencil yamBush carrot––S4
Eriosema chinenseRed yambulubbi/kurrkeb131S5
Ipomoea gramineaGrass leaf yaman-wohmi163S6
Curcuma australasicaNative turmerican-binjdjarrang129S7
Cynogetum dubiumWater yaman-buled/burlhde130S8
Thysanotus banksiiFringed lily––S8
Curculigo ensifoliaGrass leaf yaman-mulbirrk/an-burda128S8
Typha domingensisBullrushkundalkb–S9
Dioscorea bulbiferaCheeky yaman-kinjdjek157S10
Dioscorea transversaLong yamkarrbarda/kayawal158S11
Dioscorea esculentaLesser yam––S12

Note names are not identical to Kunwinjku plant names used by the Njanjma, but are comparable: for instance, names are generally cognate except for ‘an-‘ prefix in Kundjeyhmi becomes ‘man-‘ prefix in Kunwinjku.

The Kundedjnjenghmi names for taro (Fox and Garde 2018: 126).

General name for grasses, rushes and sedges obtained from online Bininj Kunwok dictionary (https://www.njamed.com/) viewed 5 January 2023.

Archaeobotanical results

Due to charring conditions and other taphonomic effects, the degree of discernible cellular structure and anatomical detail within the fifteen fragments varied greatly (Supplementary Table S4). There was insufficient information discernible to provide an identification for seven fragments (2a, 4a, 6a, 8a, 8d, 16g, 16h) and Fragment 30a resembled bone.

Other fragments show recognisable cellular structures and morphological characteristics that are insufficient for taxonomic identification, but are diagnostic of the types of USO organ utilised, including bulb/corm, dicot primary root/tap root and monocot stem tuber.

Corms/bulbs

Fragments 4h, 5d and 18b show cellular arrangements suggestive of a bulb or corm. Fragments 4h and 5d contain circular regions consistent with a stem base (Figures 3a–d) and adjacent cellular orientations characteristic of divisions between leaf margins in bulbs. The latter feature is also present in Fragment 18b (Figure 3h). Corms differ from bulbs in being solid throughout, and the stem base is at the top of the USO rather than at the base. A possible bulb candidate growing in the region is Cyperus bulbosus (Supplementary Figure S1) and possible corm candidates are Eleocharis dulcis and members of the Araceae, such as Colocasia esculenta (Supplementary Figure S2) and Amorphophallus galbra (Supplementary Figure S3).

MicroCT virtual images of Birriwilk fragments of possible bulbs and dicot tap roots. Fragment 4h: **(a)** sliced across surface showing internal structure underneath—orientation unclear. A circular region below (white arrow) suggesting possible bulb stem base with linear cell pattern in longitudinal view above. Patterning suggestive of divisions between leaf margins (black arrows) around a central circular feature; and, **(b)** sliced perpendicular to previous view showing internal structures—the dense longitudinal cell structure is in transverse view (arrow indicates position of circular feature). Fragment 5d: **(c)** internal transverse view. Central ’stem’ structure indicated by arrow. Note the curvature across this surface. Dark lines could indicate divisions between leaf margins; and, **(d)** upper image-radial view, lower image- tangential view (arrow indicates central ’stem’). This structure suggests a bulb or possibly corm. Fragment 12a: **(e)** whole fragment outer surface—arrow indicates direction of virtual slice to reveal cellular detail; and, **(f)** internal transverse view. Cell size graduating from larger (likely outer) to smaller (likely inner) in concentric rings indicated by arrows. Remnants of vascular tissue indicated by larger circular holes within these layers—likely dicot. Fragment 15a: **(f)** upper—outer surface, lower- internal transverse view showing concentric arrangement consistent with a tap root. Fragment 18b: **(g)** internal oblique slice showing structural arrangements suggestive of leaf margins in a bulb.
Figure 3. MicroCT virtual images of Birriwilk fragments of possible bulbs and dicot tap roots. Fragment 4h: (a) sliced across surface showing internal structure underneath—orientation unclear. A circular region below (white arrow) suggesting possible bulb stem base with linear cell pattern in longitudinal view above. Patterning suggestive of divisions between leaf margins (black arrows) around a central circular feature; and, (b) sliced perpendicular to previous view showing internal structures—the dense longitudinal cell structure is in transverse view (arrow indicates position of circular feature). Fragment 5d: (c) internal transverse view. Central ’stem’ structure indicated by arrow. Note the curvature across this surface. Dark lines could indicate divisions between leaf margins; and, (d) upper image-radial view, lower image- tangential view (arrow indicates central ’stem’). This structure suggests a bulb or possibly corm. Fragment 12a: (e) whole fragment outer surface—arrow indicates direction of virtual slice to reveal cellular detail; and, (f) internal transverse view. Cell size graduating from larger (likely outer) to smaller (likely inner) in concentric rings indicated by arrows. Remnants of vascular tissue indicated by larger circular holes within these layers—likely dicot. Fragment 15a: (f) upper—outer surface, lower- internal transverse view showing concentric arrangement consistent with a tap root. Fragment 18b: (g) internal oblique slice showing structural arrangements suggestive of leaf margins in a bulb.

Fragment 5d was compared to desiccated and charred modern reference specimens of Cyperus bulbosus (Supplementary Figure S1). The cellular organisation evident in virtual radial and tangential views of fragment 5d (Figure 3d) is like the regular cellular arrangement of Cyperus bulbosus (Supplementary Figure S1f). The widest cross section of the stem base of the charred modern reference specimen is 753.55 µm. This is of the same order of magnitude as the ’stem’ base of Fragment 5d at 578.36 µm. In virtual transverse view (Figure 3c), fragment 5d has curved striations suggestive of leaf margins observed in Cyperus bulbosus (Supplementary Figures S1c,g,h).

Dicot primary roots/tap roots

Fragments 12a and 15a have an overall morphology suggestive of a dicot primary root or tap root tuber with cell layers arranged in concentric rings between cambial layers. Fragment 12a contained clear remnants of vascular tissue within the concentric rings (Figures 3e,f). While individual vascular elements were not detectable in Fragment 15a, the overall pattern (Figure 3g) resembles the microCT image of the central zone of xylem of Eriosema chinense (Supplementary Figure S5f).

Possible candidates in the region include Ipomoea graminea, and members of the Fabaceae family such as Vigna lanceolata, Vigna vexillata and Eriosema chinense. Examination of reference USO specimens of these species revealed distinct cell layers in concentric rings, although the number and thickness of these cell layers varied from species to species (Supplementary Figure S4–S6). Clearly defined vascular tissue was observed within the concentric cell layers of dried and charred USOs of Vigna lanceolata and dried Vigna vexillata. The arrangement of the vascular tissue differed in these two species. In Vigna lanceolata there was a well-defined wide band of evenly spaced rows of xylem in a continuous ring inside the outer periderm and cortical layers (Supplementary Figures S4b,c,e). The vascular tissue was observed but was less marked in Vigna vexillata (Supplementary Figure S4f). The vascular tissue was present but not so clearly defined within the densely packed cell layers of the dried and charred specimens of Eriosema chinense (Supplementary Figure S5e), Ipomoea graminea and Ipomoea abrupta (Supplementary Figure S6). The arrangement of vascular tissues and cell layers of dicot primary roots contrast with those of monocot root tubers such as the native turmeric Curcuma australasica (Supplementary Figure S7), the ‘water yam’ Cycnogetum dubium, the ‘grass leaf yam’ Thysanotus banksii, the elongated corm of Curculigo ensifolia (Supplementary Figure S8) and the tuberous root of Typha domingensis (Supplementary Figure S9).

Monocot stem tubers

Two fragments (27a and 4e) of archaeological parenchyma were derived from monocot stem tubers. Given the significance of yams (Dioscorea spp.) to the cultural and economic life of Aboriginal communities in this region, these fragments were analysed and reported in detail. After extensive investigation, neither fragment can be allocated to a botanical taxon in accordance with plant identification confidence levels (Barton and Paz 2007; Paz 2001).

Fragment 27a measures 6.55 mm long and 4.12 mm wide. The outer layer of Fragment 27a is a network of irregularly shaped parenchyma cells interspersed with five well-defined features referred to here as warts (Figures 4a,b). Hather (2000: 24) describes wart-like surface features of the Dioscoreaceae family and exemplified in Dioscorea bulbifera: “The surface of many vegetative organs have warts the function of which is little understood. They are mostly extensions of the epidermis only and may contain oil or latex ducts.” The width and length of these features on Fragment 27a was 431.9 µm ± 39.4 µm and 535.5 µm ± 43.4 µm, respectively. They are discrete structures confined to periderm and cortical layers that do not extend beyond the endodermis into the inner part of the fragment (Figure 4e). Surface structures are typical of tubers of the Dioscoreaceae family, many of which are root abscission scars (Supplementary Figures S10b–e, S11d,e, S12a; Hather 2000:13, 21). However, the features on the surface of Fragment 27a are not consistent with root abscission scars because there appears to be no evidence of vascular tissue associated with these structures.

SEM surface images of fragment 27a: **(a)** entire fragment showing distribution of circular bud/wart-like features, **(b)** inset of bud/wart at higher magnification, and **(c)** surface parenchyma cells adjacent to buds/warts. MicroCT images of specific anatomical features: **(d)** virtual internal slice showing vascular bundles in transverse and longitudinal orientations indicated with arrows, **(e)** virtual internal slice of buds/warts on surface of fragment in transverse, **(f)** transverse view of vascular bundle, and **(g)** longitudinal view of vascular bundles—ground parenchyma cells are irregular and rounded.
Figure 4. SEM surface images of fragment 27a: (a) entire fragment showing distribution of circular bud/wart-like features, (b) inset of bud/wart at higher magnification, and (c) surface parenchyma cells adjacent to buds/warts. MicroCT images of specific anatomical features: (d) virtual internal slice showing vascular bundles in transverse and longitudinal orientations indicated with arrows, (e) virtual internal slice of buds/warts on surface of fragment in transverse, (f) transverse view of vascular bundle, and (g) longitudinal view of vascular bundles—ground parenchyma cells are irregular and rounded.

The outer periderm/cortex layer is between 1046.5 µm and 397.6 µm thick. This was thicker than the periderm/cortex layers measured in the modern reference Dioscorea tubers which, except for the D. transversa specimen with a maximum of 635 µm, did not exceed 303 µm. However, the dimensions of this layer are consistent with that of the aerial bulbils of Dioscorea bulbifera and bulbils also have warts on the surface exemplified in images published by Subasini et al. (2013). In addition, Hather (1994:74–75) notes that the outer periderm and cortex of aerial bulbils of Dioscorea bulbifera are thicker than that of the underground tuber. While partially fused, the multicellular outer layer is separated from the inner ground tissue. This junction is occupied by the water resistant protective endodermal cell layer in USOs though here it cannot be differentiated from a well-defined region of very small cells (Figure 4e); possibly the sclerenchymatous secondary cortex noted for members of the Dioscoreaceae (Raman et al. 2014). The inner ground tissue consists of larger parenchyma cells with dimensions of 35.8 µm ± 5.8 µm (n = 20). The ground parenchyma cells are interspersed with vascular strands throughout the fragment in an arrangement typical of monocot stem morphology. In transverse view widths were 108.5 ± 25.3 µm (n = 15).

The cell arrangement of the ground tissue is a disorganised network of parenchyma cells in both transverse and longitudinal views (Figures 4d–g). The overall shape of the parenchyma cells is of rounded irregular polyhedra tending towards pentagonal in 2D. The ground tissue is interspersed with vascular bundles in both transverse and longitudinal orientations (Figure 4d). This distribution and variable orientation of vascular bundles is present in monocot stems at the node where buds form. The orientation of the ground parenchyma is unlike that of the stem itself where parenchyma cells occur in linear alignments adjacent to the vascular tissue in longitudinal view (such as noted for sugarcane, Saccharum officinarum; Supplementary Figure S12g,h).

The vascular tissue appears to be comprised of a rounded multilayered bundle sheath of relatively small thick-walled cells (possibly sclerenchyma). Sclerenchyma cells are fibrous with thickened cell walls, which strengthen and provide structural support to plants. There are 5–6 large xylem vessels to one side of the bundle sheaths (Figure 4f), but the phloem is not discernible in these images. Charring often fuses the ground or pith parenchyma and phloem within USOs leaving rhexigenous or vesicular cavities (Hather 2000:72) or fused carbonised residue adjacent to the xylem vessels that are preserved to varying degrees (Hather 1993:6). This process is quite variable depending on the severity of the charring. Based on the orientation of the xylem relative to the bundle sheath, the vascular bundles are likely to be compound and collateral (xylem and phloem side by side in the same radius). Due to the disorganised nature of the vascular bundle arrangement within the fragment, it is difficult to determine the orientation of the xylem and phloem in the vascular bundles relative to the outer surface.

Fragment 27a has several anatomical features indicate monocot stem tuber and are suggestive of an aerial bulbil rather than an underground storage organ: the significant thickness of the outer periderm/cortex layer; the presence of surface structures resembling emerging buds or warts; the underlying vascular arrangement typical of a monocot stem; and, the disorganised ground tissue. However, although several anatomical elements are consistent with aerial bulbils of Dioscorea bulbifera that form in the leaf axis of the vine (Hather 1994: 74–75; Hather 2000: 24; Subasini et al., 2013; Raman et al., 2014), other elements, including detailed virtual histological analysis of the buds/warts, are not consistent with this species and are similar to archaeological parenchyma identified at Madjebebe as a monocot stem-based storage organ derived from “an aquatic or semi-aquatic USO” (Florin et al. 2020: 3, Figure 3c, Type A).

Fragment 4e is one of the largest fragments analysed measuring 6.78 mm in length and 4.82 mm in width. The internal structure is comprised of rounded variable parenchyma cells interspersed with vascular strands in a disorganised arrangement (Figure 5). There are irregular elongated cavities throughout the parenchyma tissue. The average widths and lengths of the parenchyma cells are 37.8 ± 4.7 µm and 47.5 ± 5.5 µm, respectively (n = 20), when vascular bundles are in transverse view. The seven vascular strands present are circular to ovate in transverse view where the average widths and lengths are 218.7 ± 41.0 µm and 264.4 ± 37.0 µm respectively. In longitudinal view, the average widths and lengths are 230.5 ± 44.7 µm and 492.8 ± 66.9 µm respectively (n = 6). The average parenchyma cell wall thickness is 5.8 ± 0.5 µm (n = 20). There is insufficient resolution to determine orientation of xylem and phloem.

SEM surface images of fragment 4e: **(a)** whole fragment, and **(b)** inset at higher magnification—surface parenchyma and vascular strand. MicroCT virtual internal views of **(c)** vascular strands (arrows) in longitudinal and transverse orientations (note segmented appearance in longitudinal strand) and elongated cavities (circled), **(d)** parenchyma removed showing disorganised arrangement of vascular tissue (characteristic of a monocot tuber), **(e)** detailed transverse view of multi-stranded internal vascular bundle suggestive of sclerenchyma to provide structural support to the tuber, **(f)** vascular bundle detail in longitudinal view surrounded by irregular organisation of parenchyma. Parenchyma cells are irregular rounded shapes and of variable size in both longitudinal and transverse views.
Figure 5. SEM surface images of fragment 4e: (a) whole fragment, and (b) inset at higher magnification—surface parenchyma and vascular strand. MicroCT virtual internal views of (c) vascular strands (arrows) in longitudinal and transverse orientations (note segmented appearance in longitudinal strand) and elongated cavities (circled), (d) parenchyma removed showing disorganised arrangement of vascular tissue (characteristic of a monocot tuber), (e) detailed transverse view of multi-stranded internal vascular bundle suggestive of sclerenchyma to provide structural support to the tuber, (f) vascular bundle detail in longitudinal view surrounded by irregular organisation of parenchyma. Parenchyma cells are irregular rounded shapes and of variable size in both longitudinal and transverse views.

While Fragment 4e lacks a discernible periderm and cortical layer and surface buds/warts, it has similar features to Fragment 27a suggesting they may be of the same genus. The shape and orientation of the parenchyma and vascular bundles in the ground tissue are quite similar. Both fragments have dense features surrounding the bundle sheaths (Figures 4f, 5e,f, respectively).

The spread and distribution density of the vascular bundles differ between the two fragments. In Fragment 27a there are more than 20 in transverse view while Fragment 4e has less than 10. In Fragment 27a, the vascular bundles are more densely positioned closer to the periderm and cortical layer decreasing with increasing distance from the fragment outer surface, which is typical tuber morphology. The vascular bundles of Fragment 4e are more sparsely arranged. Given that both fragments are of similar dimensions, the difference in density of vascular bundles suggests that Fragment 4e is likely to be from an internal position deeper inside the original specimen compared to Fragment 27a. This difference may indicate that the two fragments originate from specimens of different dimensions. There were no cavities resembling secretory bodies such as idioblasts and no evidence of raphides or druses in either fragment.

The cellular structure and overall morphological characteristics of Fragment 4e allow identification as monocot stem tuber, again with several anatomical elements consistent with Dioscorea spp (Ayensu 1972.:54–57; Barron et al. 2022b). However, anatomical similarities between Fragments 4e and 27a, as well as with Florin et al.’s (2020) archaeological parenchyma Type A, suggest that this is not definitive and a taxonomic identification is not yet possible.

AMS radiocarbon dating

Given the potential significance of the monocot stem tubers, Fragments 4e and 27a were directly AMS dated to less than 264 cal BP (S-ANU 60812) and 4,409–4,097 cal BP (S-ANU 60813), respectively (Table 2). The date for Fragment 4e corresponds to a previous date on wood charcoal from XU8 (147 ± 26 BP Wk-32918; Shine et al. 2013), indicating these deposits were formed immediately prior to, or following European colonisation. Direct dating of Fragment 27a securely dates the fragment to 4,409–4,097 cal BP, which broadly corresponds to an activity ‘unit’ extending from XU19 to XU27 dated to c. 4,800–4,100 cal BP (Shine et al. 2013; see Supplementary Figure S13). The direct AMS dating of significant archaeobotanical samples is now standard practice to overcome chronostratigraphic and taphonomic uncertainties at archaeological sites (following Smith 2001).

Table 2. Radiocarbon dates for archaeological parenchyma from text unit 1 at Birriwilk.

XULab codeTaxonomic identificationRadiocarbon age (BP)Calibrated date (cal BP) 95.4%
4S-ANU 60812Monocot stem tuber149 ± 26264-…
27S-ANU 60813Monocot stem tuber3,875 ± 364,409–4,097

Samples were pretreated with an ABA protocol prior to AMS analysis (Fallon et al. 2010; Wood et al. 2016). Dates were calculated according to Stuiver and Polach (1977) using a δ13C measured by AMS. Calibration was undertaken using OxCal v 4.4 (Bronk Ramsey 2009) and SHCal20 calibration dataset (Hogg et al. 2020). ‘…’ refers to calibrated probability distributions which may extend beyond the limit of the calibration curve.

Discussion

MicroCT is a non-destructive technique that enables 3D visualisation of the internal anatomy of samples, thereby enabling organographic determination, namely determination of organ-type. Virtual histology facilitates characterisation and description of internal, as well as external, anatomical features (see Barron et al. 2022b). The resolution of microCT used here, being 2.9–4.9 µm, hinders detailed investigation and description of individual anatomical elements, such as individual vascular tissues like phloem and xylem within vascular bundles; consequently, taxonomic identification is elusive.

Based on the results presented here, microCT provides an intermediate level of anatomical analysis between RLM and SEM that aids in characterising the internal organisation of anatomical elements. In terms of workflow, microCT can provide indications of organ-type, with SEM providing more detailed investigation of anatomical elements to determine taxonomic identification of different organ types. The non-destructive character of microCT leaves samples intact for subsequent preparation for SEM analysis.

Archaeobotanical evidence presented here indicates Aboriginal exploitation of multiple types of USO during the late Holocene, including bulbs or corms, dicot primary/tap roots and monocot stem tubers. Monocot stem tubers, such as yams (Dioscorea spp,) were known to grow and to have been exploited by Aboriginal communities across northern Australia (Bordulk et al., 2012:106; Denham et al. 2009b; Maiden 1889:22; Winydjorrotj and Jawoyn Association 2005:55; Yunupinu et al., 1995:34). It should also be considered that although some species may not be used today or in the recent past, they could have been processed and utilised in the distant past, perhaps detoxified and eaten as famine foods or had some ritual significance (Hiscock 1999; O'Connell 1995).

The density of cultural material at Birriwilk indicates ephemeral site use during the mid-Holocene, with increased site use after c. 4,000 cal BP (Figure 2; Shine et al., 2013). The distribution of the subset of charred archaeological parenchyma fragments peaks between c. 300 and c. 150 cal BP, which is slightly later than the lithic artefacts and before the entire macrocharcoal peak. Variable preservation conditions have been noted at archaeological sites in northern Australia where the preservation of charcoal decreases with increasing age (Kamminga and Allen 1973; Schrire 1982; Whitau et al., 2017); this also occurs at Birriwilk (Shine et al., 2013). However, two comparable monocot stem tubers—the modern Fragment 4e and the mid Holocene Fragment 27a—at Birriwilk both contained discernible cellular details that were useful for the determination of USO type. This suggests that individual fragments do retain identifiable cellular structures over extended timeframes.

Overall, the archaeobotanical findings are consistent with ethnobotanical information for the exploitation of food plants in the region. Possible candidate species include the monocot tubers of Dioscorea, the monocot bulbs of Cyperaceae Cyperus bulbosus, the dicot tubers of Convolvulaceae Ipomoea graminea, and members of the Fabaceae Vigna lanceolata and Eriosema chinense. These and other related species—such as Ipomoea abrupta, Vigna vexillata and Cyperaceae Eleocharis dulcis—are widely distributed and utilised across northern Australia (Bordulk et al. 2012:104–109; Roberts Garalnganyjak et al. 2011:66–70; Yunupinu et al. 1995:34–37).

For the selection of species in the reference collection, this project focussed on USO species in northern Australia that were eaten. Medicinal, ritual, and other cultural usages have also been documented for various species across northern Australia (Bordulk et al. 2012:215–225; Roberts Garalnganyjak et al. 2011:151–157; Spencer 1914:92–107, 277; Wightman and Jackson 1991:27). Expanding the modern reference collection to include species with a range of uses may better inform future analyses of the archaeological parenchyma at Birriwilk. Given the great intraspecific diversity and plasticity of many vegetatively reproducing and propagated plants, especially USOs (Denham et al. 2020; Hodge 2014), a greater range of intraspecies variation in the modern reference collection may reveal additional morphological traits to facilitate more taxonomically precise taxonomic identifications of archaeological parenchyma at Birriwilk and further afield across northern Australia.

Conclusion

This research represents an initial foray into the application of microCT to the characterisation of archaeological parenchyma at Birriwilk rockshelter in Arnhem Land. It demonstrates the ability of the technique to determine USO-type, with future potential for species identification when used in conjunction with higher-resolution SEM. A more extensive modern reference collection is needed to better represent the species diversity in the Arnhem Land region and to establish intraspecies phenotypic ranges to facilitate taxonomic identification and to ultimately address archaeobotanical questions relating to past plant-human interactions in northern Australia.

Archaeobotanical evidence at Birriwilk rockshelter indicates the presence of a range of USOs including bulbs or corms, dicot primary roots/tap roots and monocot stem tubers representing diverse plant exploitation practices in the past. While the continuity of specific plant exploitation practices documented during the historical period cannot be assumed, this preliminary evidence suggests that some of the plant species exploited during recent times were also exploited in the deep past. The organographic identification of a larger sample of archaeological parenchyma at Birriwilk could shed light on wider aspects of Aboriginal plant exploitation during the Holocene.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fearc.2026.1867558/full#supplementary-material

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Footnotes

Edited by: Simone Riehl, University of Tübingen, Germany

Reviewed by: Lucy Kubiak-Martens, University of Amsterdam, Netherlands

Ceren Kabukcu, University of Liverpool, United Kingdom

Statements

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Funding

The author(s) declared that financial support was received for this work and/or its publication. The research was funded by an Australia-Pacific Science Foundation Grant to Tim Denham and Denis Shine, and a Monash Research Accelerator Program Grant and Australian Research Council Future Fellowship (FT150100420) to Tim Denham.

Data availability

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

Author contributions

JP: Methodology, Formal analysis, Writing – original draft, Writing – review & editing, Data curation, Conceptualization, Investigation, Visualization, Validation. DS: Resources, Writing – original draft, Project administration, Writing – review & editing, Conceptualization, Investigation, Data curation, Methodology. AB: Validation, Formal analysis, Methodology, Writing – review & editing, Investigation, Writing – original draft, Visualization. RW: Writing – original draft, Methodology, Writing – review & editing, Investigation. AN: Writing – review & editing, Investigation, Writing – original draft, Conceptualization. TD: Conceptualization, Funding acquisition, Investigation, Writing – review & editing, Resources, Writing – original draft, Supervision, Data curation, Project administration, Validation, Formal analysis, Methodology, Visualization.

Acknowledgements

We acknowledge the Nayinggul family, the Manilikarr community and the Njanjma Aboriginal Corporation who made it possible to undertake this research, who have been kept informed through key stages of this community-led archaeology project (community reports, laboratory visit and discussions at ANU, as well as provision of article draft). The archaeological excavations were conducted with permits from the Aboriginal Areas Protection Authority (NT); Department of Natural Resources, Environment, the Arts and Sport (NT); and, the Northern Land Council. We thank the following people; Kim Parker and Tara Lewis who together with Denis Shine, sorted the archaeological parenchyma from the Birriwilk macrocharcoal; Levi Beeching and Michael Turner at the Australian National University microCT lab for their expertise; Stewart Fallon and Rebecca Esmay at the Australian National University radiocarbon lab for their expertise; Dave Albrecht at the Australian National Herbarium, Ian Cowie and Glenn Wightman at the Northern Territory Herbarium for their expert botanical and ethnobotanical advice and for sampling the precious collection of tubers for the modern reference collection; Karlee Foster at the Alice Springs Desert Park who collected a fresh sample of Vigna lanceolata to add to the reference collection; Anna Florin for comments and discussion on a previous draft; and, Daryl Wesley and Tristen Jones for their assistance with Community liaison. We also thank reviewers for their patience and feedback.

References

  1. Arranz-Otaegui A., Gonzalez Carretero L., Ramsey M. N., Fuller D. Q., Richter T. (2018). Archaeobotanical evidence reveals the origins of bread 14,400 years ago in northeastern Jordan. Proceed. Nat. Acad. Sci. (USA) 115, 7925 – 7930. doi:10.1073/pnas.1801071115
  2. Asmussen B., McInnes P. (2013). Assessing the impact of mid-to-late holocene ENSO-driven climate change on toxic macrozamia seed use: a 5000 year record from eastern Australia. J. Archaeol. Sci. 40, 471 – 480. doi:10.1016/j.jas.2012.06.005
  3. Atchison J., Head L. (2012). Yam landscapes: the biogeography and social life of Australian Dioscorea. The Artefact 35, 59 – 74.
  4. Atchison J., Head L., Fullagar R. (2005). Archaeobotany of fruit seed processing in a monsoon savanna environment: evidence from the keep river region, Northern Territory, Australia. J. Archaeol. Sci. 32, 167 – 181. doi:10.1016/j.jas.2004.03.022
  5. Ayensu E. S. (1972). Anatomy of the Monocotyledons VI. Discoreales. Oxford: Clarendon Press.
  6. Barron A., Denham T. P. (2018). A microCT protocol for the visualisation and identification of domesticated plant remains within pottery sherds. J. Archaeol. Sci. Rep. 21, 350 – 358. doi:10.1016/j.jasrep.2018.07.024
  7. Barron A., Denham T. P. (2022). A protocol for identifying archaeological parenchyma using microCT: steps towards a virtual reference collection. J. Archaeol. Sci. Rep. 46, 103682. doi:10.1016/j.jasrep.2022.103682
  8. Barron A., Fuller D. Q., Stevens C., Champion L., Winchell F., Denham T. P. (2020). Snapshots in time: microCT scanning of pottery sherds determines early domestication of sorghum (Sorghum bicolor) in east Africa. J. Archaeol. Sci. 123, 105259. doi:10.1016/j.jas.2020.105259
  9. Barron A., Mountain M.-J., Denham T. P. (2022a). MicroCT scanning and direct AMS dating of charred sweet potato (Ipomoea batatas) fragments from Nombe rockshelter in the highlands of Papua New Guinea. Archaeol. Ocean. 57, 146 – 149. doi:10.1002/arco.5266
  10. Barron A., Pritchard J., Denham T. P. (2022b). Identifying archaeological parenchyma in three dimensions: diagnostic assessment of five important food plant species in the indo-Pacific region. Archaeol. Ocean. 57, 189 – 213. doi:10.1002/arco.5276
  11. Barron A., Turner M., Beeching L., Bellwood P., Piper P., Grono E. (2017). MicroCT reveals domesticated rice (Oryza sativa) within pottery sherds from early neolithic sites (4150–3265cal BP) in Southeast Asia. Sci. Rep. 7, 7410. doi:10.1038/s41598-017-04338-9
  12. Barton H., Paz V. (2007). “ Subterranian diets in the tropical rain forests of sarawak, Malaysia,” in Rethinking Agriculture: Archaeological and Ethnoarchaeological Perspectives, eds. Denham T.P., Iriarte J., Vrydaghs L. (London, New York: Routledge), 50 – 77.
  13. Barton H., Paz V., Carlos A. J. (2016). “ Plant food remains from the niah caves: macroscopic and microscopic approaches,” in Archaeological Investigations in the Niah Caves, Sarawak, Vol. 2, eds. Barker G., Farr L. (Cambridge: McDonald Institute for Archaeological Research.
  14. Berndt C. H., Berndt R. M. (1989). The Speaking Land: Myth and Story in Aboriginal Australia. Melbourne: Penguin Books.
  15. Bordulk D., Dalak N., Tukumba M., Bennet L., Bordro Tingey R., Katherine M. (2012). Dalabon Plants and Animals,: Aboriginal Biocultural Knowledge from Southern Arnhem Land North Australia. Darwin: Department of Land Resource Management.
  16. Bronk Ramsey C. (2009). Bayesian analysis of radiocarbon dates. Radiocarbon. 51 (1), 337 – 360. doi:10.1017/S0033822200033865
  17. Chaloupka G. (1993). Journey in Time. the Worlds Longest Continuing art Tradition. the 50,000 Year Story of Australian Aboriginal Rock art of Arnhem Land. Hong Kong: The Good Times Group Pty. Ltd.
  18. Clarke A. (1985). “ Preliminary archaeobotanical anaylsis of the Anbanbang I site,” in Archaeological Research in Kakadu National Park, ed. Jones R. (Australia: National Parks and Wildlife Service and The Australian National University Prehsitory Pacific Studies), 77 – 96.
  19. Clarke A. (1989). “ Macroscopic plant remains,” in Plants in Australian Archaeology, eds. Beck W., Clarke A., Head L. St. Lucia, (Queensland: Anthropology Museum, University of Queensland), 54 – 89.
  20. Clarke P. (2007). Aboriginal People and Their Plants. Australia: Rosenberg Publishing.
  21. Clarkson C., Jacobs Z., Marwick B., Fullagar R., Wallis L., Smith M. (2017). Human occupation of northern Australia by 65,000 years ago. Nature 547, 306. doi:10.1038/nature22968
  22. Clarkson C., Smith M., Marwick B., Fullagar R., Wallis L. A., Faulkner P. (2015). The archaeology, chronology and stratigraphy of Madjedbebe (Malakunanja II): a site in northern Australia with early occupation. J. Hum. Evol. 83, 46 – 64. doi:10.1016/j.jhevol.2015.03.014
  23. Cosgrove R., Field J., Ferrier A. (2007). The archaeology of Australia’s tropical rainforests. Palaeogeograp. Palaeoclimatolog. Palaeoecol. 251, 150 – 173. doi:10.1016/j.palaeo.2007.02.023
  24. Denham T., Atchison J., Austin J., Bestal S., Bowdery D., Crowther A. (2009a). Archaeobotany in Australia and new Guinea. Aust. Archaeol. 68, 1 – 10. doi:10.1080/03122417.2009.11681884
  25. Denham T., Barton H., Castillo C., Crowther A., Dotte-Sarout E., Florin S. A. (2020). The domestication syndrome in vegetatively propagated field crops. Ann. Bot. 125, 581 – 597. doi:10.1093/aob/mcz212
  26. Denham T., Donohue M., Booth S. (2009b). Revisiting an old hypothesis: horticultural experimentation in northern Australia. Antiquity. 83, 634 – 648. doi:10.1017/S0003598X00098884
  27. Fallon S. J., Fifield L. K., Chappell J. M. (2010). The next chapter in radiocarbon dating at the Australian national university: status report on the single stage AMS. Nucl. Instrum. Methods Phys. Res. Sect. B: Beam Interact. Mat. Atoms 268 (7-8), 898 – 901. doi:10.1016/j.nimb.2009.10.059
  28. Florin S. A., Fairbairn A. S., Nango M., Djandjomerr D., Hua Q., Marwick B. (2022). 65,000 years of changing plant food and landscape use at Madjedbebe, Mirarr country, northern Australia. Quat. Sci. Rev. 284, 107498. doi:10.1016/j.quascirev.2022.107498
  29. Florin S. A., Fairbairn A. S., Nango M., Djandjomerr D., Marwick B., Fullagar R. (2020). The first Australian plant foods at Madjedbebe, 65,000-53,000 years ago. Nat. Commun. 11 (2020), 924. doi:10.1038/s41467-020-14723-0
  30. Fox G., Garde M. (2018). An-me Arri-ngun: The Food we eat: Traditional Plant Foods of the Kundjeyhmi People of Kakadu National Park. Jabiru, NT: Gundjeihmi Aboriginal Corporation.
  31. Hammond J. W. (2019). The spatial distributon of unique motifs featuring Dioscorea bulbifera, the round yam, in western Arnhem Land rock art. Aust. Archaeol. 85 (2), 215 – 219. doi:10.1080/03122417.2019.1686237
  32. Hather J. G. (1993). An Archaeobotanical Guide to Root and Tuber Identification. Volume 1. Oxford: Oxbow books.
  33. Hather J. G. (1994). “ The identification of charred root and tuber crops from archaeological sites in the Pacific,” in Tropical Archaeobotany Applications and New Developments, ed. Hather J. G. (London: Routledge), 51 – 64.
  34. Hather J. G. (2000). Archaeological Parenchyma. London: Archetype Publications.
  35. Hiscock P. (1999). “ Holocene coastal occupation of western Arnhem Land,” in Australian Coastal Archaeology, eds. Hall J., Mcniven I. J. (Australia: Department of Archaeology and Natural History, Australian National University), 91 – 103.
  36. Hodge A. (2014). The plastic plant: root responses to heterogeneous supplies of nutrients. New Phytol. 162, 9 – 24. doi:10.1111/j.1469-8137.2004.01015.x
  37. Hogg A., Heaton T., Hua Q., Palmer J., Turney C., Southon J. (2020). SHCal20 southern hemisphere calibration, 0–55,000 years cal BP. Radiocarbon. 62, 759 – 778. doi:10.1017/RDC.2020.59
  38. Jelinek J. (1989). The Great Art of the Early Australians. the Study of the Evolution and Role of Rock Art in the Society of Australian Hunters and Getherres. BRNO: Moravian Museum-Anthropos Institute.
  39. Jones R., Meehan B. (1989). “ Plant foods of the gidjingali: ethnography and archaeological perspectives from northern Australia on tuber and seed exploitation,” in Foraging and Farming: The Evolution of Plant Exploitation, eds. Harris D.R, Hillman G. C. (London: Unwin Hyman), 120 – 135.
  40. Kamminga J., Allen H. (1973). Alligator Rivers Environmental Fact-Finding Study. Darwin: The Department of the Northern Territory.
  41. Kubiak Martens L. (2016). “ Scanning electron microscopy and starchy food in mesolithic Europe: the importance of roots and tubers in mesolithic diet,” in Wild Harvest: Plants in the Hominin and pre-Agrarian Human Worlds, eds. Hardy K., Kubiak Martens L. (Oxford, Philadelphia: Oxbow books), 113 – 133.
  42. Kubiak-Martens L. (2002). New evidence for the use of root foods in pre-agrarian subsistence recovered from the late mesolithic site at Halsskov, Denmark. Veg. Hist. Archaeobot. 11, 23 – 31. doi:10.1007/s003340200003
  43. Kubiak-Martens L. (2006). “ Roots, tubers and processed plant food in the local diet,” in Schipluiden, a Neolithic Settlement on the Dutch North Sea Coast c. 3500 cal BC., eds. Louwe Kooijmans L. P., Jongste P. F. B (Leiden: Analecta Praehistorica Leidensia 37/38), 339 – 352.
  44. Latham S., Varslot T., Sheppard A. (2008). “ Automated registration for augmenting micro-CT 3D images,” in Computational Techniques and Applications Conference (CTAC 2008). 532 – 548
  45. Lewis D. (1988). “ The rock paintings of Arnhem Land, Australia. social, ecological and material culture change in the post-glacial period,” in BAR Int. Series (Oxford: British Archaeological Reports), 415.
  46. Lewis J. (1922). Faught and won. Adelaide: W. K Thomas and Co. Available online at: (Accessed June 30, 2019).
  47. Limaye A. (2012). “ Drishti: a volume exploration and presentation tool ”, Proc. SPIE 8506, Developments in X-Ray Tomography VIII, 85060X (17 Oct 2012).. doi:10.1117/12.935640
  48. Maiden J. H. (1889). The Useful Native Plants of Australia: Including Tasmania. Syndey: Turner and Henderson.
  49. Myers G., Kingston A., Varslot T., Sheppard A. (2011). Extending reference scan drift correction to high-magnification high-cone-angle tomography. Opt. Lett. 36, 4809 – 4811. doi:10.1364/OL.36.004809
  50. O'Connell J. F. (1995). Ethnoarchaeology needs a general theory of behavior. J. Archaeol. Res. 3, 205 – 255. doi:10.1007/BF02231450
  51. Paz V. (2001) Archaeobotany and cultural transformation: pattens of early plant utilisation in northern wallacea. Unpublished PhD. University of Cambridge.
  52. Pearsall D. M. (2000). Paleoethnobotany: A Handbook of Procedures. 2nd edn. London: Academic Press.
  53. Pritchard J., Lewis T., Beeching L., Denham T. P. (2019). An assessment of microCT technology for the investigation of charred archaeological parenchyma from house sites at kuk swamp, Papua New Guinea. Archaeol. Anthropol. Sci. 11, 1927 – 1938. doi:10.1007/s12520-018-0648-0
  54. Puruntatameri J., Puruntatameri R., Pangiraminni A., Burak L., Tipuamantymirri C., Tipakalippa M. (2001). Tiwi Plants and Animals: Aboriginal flora and Fauna Knowledge from Bathurst and Melville Islands, Northern Australia. Northern Territory Botanical Bulletin No 24. Darwin: Parks and Wildlife Commission of the Northern Territory.
  55. Raman V., Galal A. M., Avula B., Sagi S., Smillie T. J., Khan I. A. (2014). Application of anatomy and HPTLC in characterizing species of Dioscorea (Dioscoreaceae). J. Nat. Med. 68, 686 – 698. doi:10.1007/s11418-014-0849-5
  56. Roberts Garalnganyjak J., Yanybarrak Conway S., Morgan R., Dirn.gayg A., Harris S., Burrbanja Farrar E. (2011). Mangarrayi and Yangman Plants and Animals: Aboriginal Biocultural Knowledge from Elsey and the Roper River, North Australia. Northern Territory Botanical Bulletin No 41. Palmerston, Northern Territory: Diwurruwurru-Jaru Aboriginal Corporation & Mimi Aboriginal Art and Craft & Northern Territory Department of Natural Resources, Environment, the Arts and Sport.
  57. Schrire C. (1982). The Alligator Rivers Prehistory and Ecology in Western Arnhem Land. Terra Australis 7, Canberra: Australian National University Press.
  58. Shine D. (2014) Changing places: an archaeological study of manilikarr country in western Arnhem Land. Unpublished PhD thesis. Australia: Monash University.
  59. Shine D., Wright D., Denham T. P., Aplin K., Hiscock P., Parker K. (2013). Birriwilk Rockshelter: a mid to late Holocene site in Manilikarr country, southwest Arnhem Land, northern territory. Aust. Archaeol. 76, 69 – 78. doi:10.1080/03122417.2013.11681967
  60. Smith B. D. (2001). Documenting plant domestication: the consilience of biological and archaeological approaches. Proceed. Nat. Acad. Sci. (USA) 98, 1324 – 1326. doi:10.1073/pnas.98.4.1324
  61. Smythe D., Van Sturmer J. (1985). The use of Plants by Aboriginal People in the Oenpelli Region of Western Arnhem Land. Canberra: Australian Insititute of Aboriginal Studies.
  62. Spencer W. B. (1914). Native Tribes of the Northern Territory of Australia. London: MacMillan and Co, Limited.
  63. Stuiver M., Polach H. A. (1977). Reporting of 14C data. Radiocarbon. 19 (3), 355 – 363. doi:10.1017/S0033822200003672
  64. Subasini U., Thenmozhi S., Sathyamurthy D., Vetriselvan S., Rajamanickam G. V., Dubey G. P. (2013). Pharmacognositic and phytochemical investigations of Dioscorea bulbifera L. Int. J. Pharm. Life Sci. 4, 2693 – 2700.
  65. Varslot T., Kingston A., Myers G., Sheppard A. (2011). High-resolution helical cone-beam micro-CT with theoretically-exact reconstruction from experimental data. Med. Phys. 38, 5459 – 5476. doi:10.1118/1.3633900
  66. Veth P., Myers C., Heaney P., Ouzman S. (2018). Plants before farming: the deep history of plant-use and representation in the rock art of Australia’s Kimberley region. Quat. Int. 489, 26 – 45. doi:10.1016/j.quaint.2016.08.036
  67. Wadley L., Backwell L., d’Errico F., Sievers C. (2020). Cooked starchy rhizomes in Africa 170 thousand years ago. Science 367, 87 – 91. doi:10.1126/science.aaz5926
  68. Welch D. (2003). Plant motifs in kimberley rock-art, Australia. Before Farm. 2003, 1 – 13. doi:10.3828/bfarm.2003.4.5
  69. Whitau R., Balme J., O'Connor S., Wood R. (2017). Wood charcoal analysis at Riwi cave, Gooniyandi country, western Australia. Quat. Int. 457, 140 – 154. doi:10.1016/j.quaint.2016.07.046
  70. Widijburru B., Northern Territory Deptartment of Natural Resources Environment the Arts and Sport & Diwurruwurru-Jaru Aboriginal Corporation. (2010). Ngarinyman Plants and Animals: Aboriginal Knowledge of flora and Fauna from Judbarra/Gregory National Park and the Victoria River Area, Northern Australia. Northern Territory Botanical Bulletin No 34. Palmerston, Northern Territory: Department of Natural Resources, Environment, the Arts and Sport.
  71. Wightman G., Jackson D. L. (1991). Alawa Ethnobotany: Aboriginal Plant use from Minyerri, Northern Australia. Northern Territory Botanical Bulletin. Darwin, Northern Territory, Australia: Conservation Commission of the Northern Territory.
  72. Winydjorrotj P., Association J. (2005). Jawoyn Plants and Animals: Aboriginal flora and Fauna Knowledge from Nitmiluk National Park and the Katherine Area, Northern Australia. Northern Territory Botanical Bulletin No 29. Darwin: Jawoyn Association & Northern Territory Government Department of Natural Resources Environment and The Arts.
  73. Wood R., Jacobs Z., Vannieuwenhuyse D., Balme J., O'Connor S., Whitau R. (2016). Towards an accurate and precise chronology for the colonization of Australia: the example of Riwi, Kimberley, western Australia. PLoS One 11 (9), e0160123. doi:10.1371/journal.pone.0160123
  74. Yunupinu B., Yunupinu-Marika L., Marika D., Marika B., Marika B., Marika R. (1995). Rirrat Jinu Ethnobotany: Aboriginal Plant use from Yirrkala Arnhem Land Australia. Northern Territory Botanical Bulletin No 21. Darwin: Parks and Wildlife Commission of the Northern Territory.

Citation

Pritchard, J., et al. (2026). MicroCT analysis of archaeological parenchyma identifies underground storage organs from late Holocene contexts at Birriwilk rockshelter, Manilikarr Country, northern Australia. Frontiers in Environmental Archaeology, Volume 5. https://doi.org/
Received
Accepted
Published
Volume
5 - 2026

Copyright

© 2026 Pritchard, Shine, Barron, Wood, Nayinggul and Denham.

License: https://creativecommons.org/licenses/by/4.0/

*Correspondence: Jenifer Pritchard

Article metrics

Views
0
Downloads
0
Citations
0
Shares
0

Related articles

Brief Research ReportMarine resource exploitation and maritime adaptation in the Southern Ryukyus during the Neolithic: a short report on the Shimotabaru shell mound, Okinawa, JapanR. Ono, C. Boulanger, Y. Najima, K. Oohori, C. KatagiriOriginal ResearchNew data from old bones: using bioarchaeology and forensics to reassess legacy data from ancient Abydos, EgyptRoselyn A. CampbellOriginal ResearchTropical arboriculture and root crop cultivation in the volcanically active landscape of ancient Arenal, Costa RicaVenicia Slotten, Rachel Egan, Payson D. Sheets