The first Australian plant foods at Madjedbebe, 65,000-53,000 years ago

Abstract

There is little evidence for the role of plant foods in the dispersal of early modern humans into new habitats globally. Researchers have hypothesised that early movements of human populations through Island Southeast Asia and into Sahul were driven by the lure of high-calorie, low-handling-cost foods, and that the use of plant foods requiring processing was not common in Sahul until the Holocene. Here we present the analysis of charred plant food remains from Madjedbebe rockshelter in northern Australia, dated to between 65 kya and 53 kya. We demonstrate that Australia's earliest known human population exploited a range of plant foods, including those requiring processing. Our finds predate existing evidence for such subsistence practices in Sahul by at least 23ky. These results suggest that dietary breadth underpinned the success of early modern human populations in this region, with the expenditure of labour on the processing of plants guaranteeing reliable access to nutrients in new environments.

Full text

ARTICLE
The first Australian plant foods at Madjedbebe,
65,000–53,000 years ago
S. Anna Florin 1✉, Andrew S. Fairbairn1,2,3, May Nango4, Djaykuk Djandjomerr4, Ben Marwick5,
Richard Fullagar 6, Mike Smith 7,8, Lynley A. Wallis9,10 & Chris Clarkson 1,2,3 ✉
There is little evidence for the role of plant foods in the dispersal of early modern humans into
new habitats globally. Researchers have hypothesised that early movements of human
populations through Island Southeast Asia and into Sahul were driven by the lure of high-
calorie, low-handling-cost foods, and that the use of plant foods requiring processing was not
common in Sahul until the Holocene. Here we present the analysis of charred plant food
remains from Madjedbebe rockshelter in northern Australia, dated to between 65 kya and
53 kya. We demonstrate that Australia’s earliest known human population exploited a range
of plant foods, including those requiring processing. Our finds predate existing evidence for
such subsistence practices in Sahul by at least 23ky. These results suggest that dietary
breadth underpinned the success of early modern human populations in this region, with the
expenditure of labour on the processing of plants guaranteeing reliable access to nutrients in
new environments.
https://doi.org/10.1038/s41467-020-14723-0 OPEN
1School of Social Science, University of Queensland, Brisbane, QLD 4072, Australia. 2Australian Research Council Centre of Excellence for Australian
Biodiversity and Heritage, University of Wollongong, Wollongong, NSW 2522, Australia. 3Depatrment of Archaeology, Max Planck Institute for the Science
of Human History, Kahlaiche Strasse 10, 07745 Jena, Germany. 4Gundjeihmi Aboriginal Corporation, 5 Gregory Place, Jabiru, NT 0886, Australia.
5Department of Anthropology, University of Washington, Seattle, WA 98195, USA. 6Centre for Archaeological Science, School of Earth, Atmospheric and
Life Sciences, University of Wollongong, Wollongong, NSW 2522, Australia. 7College of Humanities, Arts and Social Sciences, Flinders University, Adelaide,
SA 5042, Australia. 8Centre for Historical Research, National Museum of Australia, Canberra, ACT 2601, Australia. 9Nulungu Research Institute, University
of Notre Dame Australia, Broome, WA 6725, Australia.
10
Present address: Griffith Centre for Social and Cultural Research, Griffith University, Brisbane, QLD
4111, Australia. ✉email: stephanie.florin@uqconnect.edu.au;c.clarkson@uq.edu.au
NATURE COMMUNICATIONS| (2020) 11:924 | https://doi.org/10.1038/s41467-020-14723-0 | www.nature.com/naturecommunications 1
1234567890():,;
Content courtesy of Springer Nature, terms of use apply. Rights reserved

The role of plant foods in the evolution and dispersal of
early modern humans (EMHs) has often been under-
estimated. A long-held focus on the notion of Paleolithic
populations as meat eaters and a lack of consistent archae-
obotanical recovery has frequently constrained analysis and
understandings of EMH diet to its animal components1. Exten-
sive use and processing of plant resources, and an associated
broadening of the diet, was therefore typically considered a late
Pleistocene/early Holocene phenomenon, linked to changing
foraging behaviours in the millennia prior to the emergence of
agriculture2,3. However, while plant foods may not make up the
dominant proportion of EMH diets globally, more recent research
into plant macro- and micro-fossils is breaking down this para-
digm: the use of plant foods, including those associated with later
agricultural transitions, such as grass seeds and underground
storage organs (USOs), is now evidenced in Middle Stone Age
sites in Africa and the Middle East4–7; the processing of toxic
plants (Dioscorea hispida and Pangium edule) is now dated to as
early as 46–34 kya in Niah Cave, Borneo8–10; the translocation of
yams (Dioscorea spp.) to high altitudes and management of
monodrupe pandanus stands, facilitated early use of highland
environments in New Guinea (~49 kya)11,12; and associated
plant-processing technologies, such as seed-grinding stones, are
linked to EMH dispersal into northern Australia13.
This shift in paradigm is particularly important when consider-
ing the southern dispersal of Homo sapiens out of Africa. Key
debates in this region have focused on both the ‘modernity’of EMH
populations involved in these migrations14,15 and the ‘pathways’
they may have followed16. Proponents of the single coastal dispersal
model suggest human populations expanded along coastal envir-
onments ~60 kya, moving quickly through Sunda and Wallacea,
and into Sahul17,18. This model emphasises the lure of high-ranked
coastal resources and suggests diet breadth was likely narrow in the
earliest phases of human expansion throughout this region. In
contrast, other models highlight early adaptations to non-coastal
environments by EMHs leaving Africa16,19,20, including to more
extreme ecosystems (e.g. rainforests8–10,21,22, high-altitudes11,12 and
deserts23,24). Current evidence for Pleistocene plant use in Sunda
and Sahul, while not necessarily related to the earliest phases of
human expansion in this region18, largely supports this latter
interpretation. This is because the intensive and multi-step pro-
cessing techniques required to make the identified plant foods
edible are indicative of both complex and flexible foraging, and the
kind of broad diet that underpins adaptations to more difficult
environments1,25.
Here we report on the charred plant macrofossils recovered
from the earliest layer of dense occupation at Madjedbebe.
Madjedbebe is a rockshelter in western Arnhem Land (northern
Australia) situated at the base of the Djuwamba Massif, an
escarpment outlier to the east of the Magela Creek floodplain
(Fig. 1). Its earliest, dense phase of occupation (Phase 2) con-
tains charcoal, abundant ground ochre, grinding stones,
including those used for seed-grinding, and a dense assemblage
of unique flaked stone artefact types and raw materials (>10,000
artefacts). This phase is dated to c.65–53 kya on the basis of an
extensive single-grain optically stimulated luminescence and
Madjedbebe
SUNDA
WALLACEA
SAHUL
Ivane Valley
Niah Cave Estuarine vegetation
Forest and woodland vegetation
Freshwater vegetation
Rainforest vegetation
Sandstone vegetation
Savannah vegetation
No available vegetation maps
Fig. 1 Site location. a Regional map showing the location of Madjedbebe, and other sites mentioned in-text, adapted from Norman et al. 201847;bcurrent
distribution of vegetation communities in proximity to Madjedbebe, developed from map data provided by Google, Landsat/Copernicus, TerraMetrics, and
vegetation maps of the Adelaide and Alligator Rivers area48,49.
ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-020-14723-0
2NATURE COMMUNIC ATIONS | (2020) 11:924 | https://d oi.org/10.1038/s41467-020-14723-0 | www.nature.com/naturecom munications
Content courtesy of Springer Nature, terms of use apply. Rights reserved

accelerator mass spectrometry radiocarbon dating regime
(Supplementary Fig. 1)13,26,27. The archaeobotanical assemblage
from Phase 2, therefore, provides the earliest known evidence for
an EMH diet in Sahul. We demonstrate that Australia’s earliest
known human population exploited a range of plant foods,
including some requiring processing. Our findings have impli-
cations for understanding EMH behavior, cognitive flexibility
and subsistence strategies at the eastern end of the modern
human dispersal arc.
Results
Phase 2 plant macrofossil assemblage. Plant macrofossil remains
were recovered from all phases of occupation at the site using
flotation (see Methods). The Phase 2 assemblage includes over
1000 non-wood plant macrofossils from a distinct hearth feature
(C1/43 A) and excavated sediment matrix (see Supplementary
Table 1). These macrofossils can be broadly categorised into four
distinct groups: (i) endocarp and mesocarp (‘nutshell’or ‘fruit-
stone’) from a variety of fruit and nut producing species; (ii)
vegetative parenchyma from USOs (‘roots and tubers’); (iii) stem
tissue from the Arecaceae (palm) family and (iv) various other
fragments of plant material. Within these groups, genus-, or
species-specific identifications have been made for five taxa, with
a further five types identified to broader botanical categories.
The paucity of charcoal prior to human occupation (Supple-
mentary Fig. 2a) and the presence of a diversity of edible species
preserved by burning suggests that the assemblage was largely
derived from human activities, specifically the cooking and
disposal of plant resources in hearths. This is corroborated by the
relatively high percentage (17%) of fragile parenchymatous tissue
from USOs preserved in the assemblage. USO-producing species
have evolved to survive and profit from bushfires, rapidly
regenerating new aerial shoots from their buried vegetative
organs28. These organs, themselves, are therefore unlikely to have
been charred within the site without human activity. However,
this does not preclude the inclusion of some of the Phase 2 plant
macrofossils via non-human agents.
Endocarp and mesocarp. Charred endocarps of five fruit and nut
taxa were identified: Buchanania sp.; Canarium australianum;
polydrupe Pandanus sp.; Persoonia falcata;andTerminalia sp.
(Fig. 2). All of these taxa are common in open forest and woodland,
and/or monsoon vine forest environments29,andthemajority
require little or no processing. Today, the abundant and easily
harvested ‘plums’of Buchanania spp. and Persoonia falcata are
highly sought after. The fruits can be eaten raw but are also often
ground into a paste, incorporating the endocarp and seed, prior to
consumption (MN, DjDj). Canarium australianum is a relative of
the widely used Melanesian tree-crop, galip (C. indicum), and has a
small oil-rich kernel (<1 cm) that is easily extracted with a single
blow from a hammerstone (DjDj). A number of Terminalia species
are edible, consumed either as fruits (T. carpentariae,T.ery-
throcarpa,T. ferdinandiana,T. microcarpa) or easily-extracted nuts
(T. grandiflora; MN, DjDj)30. However, there are also a number of
non-edible species of Terminalia in the Northern Territory.
Pandanus formed a substantial element of the early colonial diet
of Indigenous Australians in Arnhem Land and is also a valuable
material in weaving and fibrecraft31. There are two types of
polydrupe pandanus in Arnhem Land: P. spiralis and P. base-
dowii29. Only one fragment from Phase 2 at Madjedbebe, a portion
of mesocarp, can be securely identified as P. spiralis (Fig. 3f).
However, it is likely that the majority come from this species, since
P. basedowii only grow on the escarpment top and would have been
difficult to access from Madjedbebe (see Fig. 1b). Extracting kernels
from the fibrous and mechanically-resistant prismatic structure of
the P. spiralis drupe is a labour-intensive process when using stone
tools. The explorer, Ludwig Leichardt, recorded Indigenous groups
in the Gulf of Carpentaria in the 1840s using “large flat stones and
pebbles”to bash apart the drupes32. However, once open, the small
kernels are rich in fat (44–50%) and protein (20–34%)33.
Vegetative parenchyma. Three distinct types of vegetative par-
enchyma are present in the Phase 2 assemblage. These include
parenchymatous tissue from two types of monocotyledonous,
stem-based storage organs and a fragment from a secondary-root
storage organ (Fig. 3). Monocotyledonous stem-based storage
organ Type A is represented in Phase 2 only by charred fragments
of its skin tissue, which have distinct root abscission scars, or ‘eyes’,
and surface patterning (Fig. 3a, b). These fragments are comparable
to charred peelings generated by contemporary Indigenous Aus-
tralians when they remove the coarse external surface of cooked
USOs before consumption, often in proximity to the hearth in
which they were cooked (MN, DjDj; Fig. 3d). The presence of an
endodermis, in a larger fragment of this type recovered from Phase
3, allows for its further identification as an aquatic or semi-aquatic
USO (Fig. 3c). Endodermis is rarely present in stem tissue, except
in aquatic plants where they are significant in controlling water
balance within perennating stems34.
Arecaceae. There are two types of Arecaceae stem tissue present
in the Phase 2 assemblage (Fig. 4). Type A, characterised by the
presence of mostly two or more metaxylem elements per fibro-
vascular bundle, has a similar anatomy to that of Livistona spp.
palm tissue. Type B, represented largely by the peripheral section
of the palm stem, is characterised by the presence of only one
metaxylem element per fibrovascular bundle. Type B is only
found in the Phase 2 assemblage and may represent the periph-
eral xylem of Type A or another taxon of the Arecaceae. The
apex, or ‘heart’, and pith, or whole young stem, of several palms
present in open woodland and monsoon vine forest environments
in western Arnhem Land can be consumed (apex: Carpentaria
acuminata,Hydriastele ramsayi; apex and pith: L. benthamii, L.
humilis and L. inermis)30,35. While the apex of these palms may
be eaten raw or lightly roasted, the pith requires roasting for an
extended period (~12 h) prior to pounding. This process removes
the most fibrous elements from the otherwise starchy pith,
making the carbohydrates within the pith readily available for
consumption. While there are no distinguishing apical features on
the fragments of stem from the Phase 2 assemblage, their size
makes it impossible to securely identify them as basal (pith) stem.
Discussion
In summary, the plant macrofossil assemblage from Phase 2
provides evidence for the consumption of a range of plant foods
(~10) by the EMH population occupying Madjedbebe from
65–53 kya. These plant foods were foraged from open forest and
woodland, and, to a lesser extent, monsoon vine forest, and
aquatic environments, all of which were likely in proximity to the
site during this phase13. These plant foods provided a variety of
dietary macronutrients, comprising carbohydrates (USOs, palm
stems and fruits), fats and proteins (pandanus kernels and other
tree nuts). Some of these plant foods, such as the fruits and easily-
extracted nuts, represent readily available, high-ranked resources.
However, other plant foods present in the assemblage required
varying levels of processing prior to consumption. This included
the cooking (and peeling) of USOs and palm stems, likely the
further pounding of palm pith, and the laborious extraction of
Pandanus spiralis kernels. Furthermore, while there were no
edible seeds recovered in the plant macrofossil assemblage, resi-
due and usewear evidence from the Phase 2 grinding stone
NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-020-14723-0 ARTICLE
NATURE COMMUNIC ATIONS | (2020) 11:924 | https://doi.org /10.1038/s41467-020-14723-0 | www.nature.com/naturecommunications 3
Content courtesy of Springer Nature, terms of use apply. Rights reserved

ab
cd
ef
gh
ij
vb vb
fb
sl
sl
Fig. 2 Examples of endocarp from Phase 2. a,bBuchanania sp. endocarp from C2/46(HR), ascale bar is 1 mm, btransverse section, scale bar is 500 µm;
cCanarium australianum endocarp from C2/38(HR), scale bar is 500 µm; dC. australianum endocarp from C2/37(HR), close-up of internal surface, scale
bar is 500 µm; ePolydrupe Pandanus sp. endocarp from C2/42(HR), transverse section, scale bar is 500 µm; fP. spiralis mesocarp from C2/37, scale bar is
200 µm; g,hPersoonia falcata endocarp from C2/37(HR), gscale bar is 1 mm, hclose-up of internal surface, scale bar is 200 µm; i,jTerminalia sp. endocarp
from C2/37(HR), iscale bar is 1 mm, jtransverse section, scale bar is 200 µm. See the supporting online information and Supplementary Figs. 3-7 for
detailed identification proofs and the corresponding reference materials. sl seed locule, vb vascular bundle, fb fibrous bundle.
ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-020-14723-0
4NATURE COMMUNIC ATIONS | (2020) 11:924 | https://d oi.org/10.1038/s41467-020-14723-0 | www.nature.com/naturecom munications
Content courtesy of Springer Nature, terms of use apply. Rights reserved

assemblage has identified seed-grinding during this initial phase
of occupation13. While none of these plants are toxic, the pro-
cessing required to extract and make edible the nutritious com-
ponents from some of the taxa present is suggestive of multi-step
and labour-intensive processing techniques.
These findings, which predate existing evidence for the pro-
cessing of plant foods in Sahul by at least 23ky36,37, suggest that a
broader range of lower ranked plant foods was consumed during
early occupation of Sahul than envisaged by proponents of the
single coastal dispersal model17,18. This does not negate the
possibility that EMH populations first entered Sahul exploiting
high-ranked, coastal resources found on its now-submerged
coastline. However, the investment of labour and technology into
the extraction and processing of fruits, nuts, USOs and likely
seeds at Madjedbebe 65–53 kya, does suggest that a broad diet
was part of the toolkit employed by the EMH populations who
reached Sahul.
Indeed, as EMH populations crossed the Wallace Line, novel
fauna would likely have caused significant disruptions to their
hunting strategies. However, many of the families, and in some
ab
cd
ef
gh
ra
en
vb
ph
xy
xy
Fig. 3 Examples of vegetative parenchyma from Phase 2. a,bMonocotyledonous stem-based storage organ Type A from C2/41; adepicting skin-
patterning and root abscission scar, scale bar is 500 µm; bclose-up of root abscission scar, scale bar is 100 µm; ctransverse section of monocotyledonous
stem-based storage organ Type A from C2/32 A, depicting an endodermis and a series of closed collateral vascular bundles, arrows point to phytoliths on
edges of vascular bundles, scale bar is 500 µm; dMN peeling a ‘hairy’Dioscorea bulbifera tuber beside a hearth built to cook it, photo taken by SAF;
e,flongitudinal section of monocotyledonous stem-based storage organ Type B; escale bar is 300 µm; fclose-up of vascular bundle, scale bar is 20 µm;
g,htransverse section of secondary-root storage organ Type A from C2/39 A; gscale bar is 2 mm; hclose-up of central tract of xylem, scale bar is
500 µm. ra root abscission scar, en endodermis, vb vascular bundle, ph phloem, xy xylem.
NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-020-14723-0 ARTICLE
NATURE COMMUNIC ATIONS | (2020) 11:924 | https://doi.org /10.1038/s41467-020-14723-0 | www.nature.com/naturecommunications 5
Content courtesy of Springer Nature, terms of use apply. Rights reserved

cases genera (including Buchanania,Canarium,Livistona,Pan-
danus and Terminalia), of plants available across the southern
dispersal arc continue into Australia38,39. Therefore, the expen-
diture of labour in the preparation of a range of recognised plant
foods could have ensured reliable access to fats, proteins and
carbohydrates required to successfully move into the region.
The evidence for a broad plant food diet at Madjedbebe 65–53
kya is consistent with later Pleistocene archaeobotanical studies
conducted in Island Southeast Asia and Sahul8–12, and with
evidence for EMH diets in Africa and the Middle East4–6.As
such, it indicates that plant exploitation was a fundamental aspect
of EMH diets globally. Culturally transmitted botanical knowl-
edge, and the cognitive ability to perform multi-step and intensive
processing sequences likely contributed to the adaptability and
flexibility required by EMH populations to traverse continents
and colonise new environments around the world.
Methods
Archaeobotanical analysis. Two 1×1m
2columns (C3/1–27 and C2/28–57; C6/
1–15 and C5/16–72) of excavated sediment and all hearths and other features
identified during excavation were collected in their entirety for flotation (Supple-
mentary Fig. 2c). A hundred percent of the sediment from these contexts under-
went flotation, using a cascading ‘Ankara-style’flotation tank40.
All Phase 2 contexts in Square C2 (C2/46, C2/45, C2/44, C2/43, C2/42, C2/41,
C2/40, C2/39 A, C2/38 A, C2/38, C2/37 A and C2/37) and a preserved hearth in
Square C1 (C1/43 A) (Supplementary Fig. 2c) were analysed.
As flotation has been suggested to be less effective at recovering denser
macrofossils, especially endocarp fragments41, both the lighter fraction, or ‘flot’,
and the heavy residue from these contexts were analysed. All charred fragments >1
mm from both fractions were analysed. The relevant plant macrofossils were sorted
from the wood charcoal under low-powered light microscopy.
High-powered light microscopy and scanning electron microscope imaging was
used to compare the anatomical and morphological features of the archaeological
specimens to modern reference material from the region (see below). The
identification of this archaeological material was limited by two things: the size of
the modern reference collection; and the degree to which the botanical structures
vary by family, genus and species.
No attempt was made to quantify the proportion of the diet contributed by
different plants, analysis stopping simply at ubiquity. This is because many of the
food plants that constituted the diet of people inhabiting the site at 65–53ka would
not have been preserved archaeologically42. Indeed, many of the plant foods may
have been eaten raw and have, therefore, been less likely to come into contact with
fire, or have been processed and/or eaten away from Madjedbebe as a part of a
mobile foraging strategy43. This difference in preservation, determined by past
human behaviour, would have been compounded by differences in the preservation
rates and the level of identification possible for different taxa44.
Reference collection and ethnobotanical research. The modern reference
material was collected by SAF, MN, DjDj and research assistants over several
seasons in Kakadu National Park. This was carried out with the permission of the
Mirarr people, Gundjeihmi Aboriginal Corporation, Parks Australia and the
Australian Government (Permit to carry out Scientific Research in a Common-
wealth Reserve Permit No. RK870 and RK909; Access to Biological Resources in a
Commonwealth Area for Non-Commercial Purposes Permit No. AU-COM2015-
287, AU-COM2017-339, AU-COM2018-391). Alongside the production of a
modern reference collection, ethnobotanical research was also undertaken with
MN and DjDj, to define the material signature of plant exploitation practices in
western Arnhem Land. Plants were identified in the field by MN and DjDj, and
these identifications were then verified and furthered by the Northern Territory
Herbarium.
Plant samples are housed in the University of Queensland Archaeobotanical
Reference Collection, preserved dried, charred and in spirits. Following
Hather 200034, where underground storage organs, stems and roots were part of
the sample, stained thin-sections were produced (using a modified version of the
method outlined by Johansen 194045). This allowed for the anatomical structure of
these plant parts to be understood prior to their transformation through charring.
Reporting summary. Further information on research design is available in
the Nature Research Reporting Summary linked to this article.
Data availability
All elements necessary to allow interpretation and replication of results, including full
datasets and detailed archaeobotanical identification proofs are provided in
the Supplementary Information. Data and R code for Supplementary Fig. 1 are online at
https://doi.org/10.17605/OSF.IO/YDUZP46. Archaeobotanical material analysed in this
study will be kept in the Archaeology Laboratories of the University of Queensland until
2021. It will then be deposited in a Gundjeihmi Aboriginal Corporation keeping place.
The material will be publicly accessible upon request with permission from Gundjeihmi
Aboriginal Corporation (gundjeihmi@mirarr.net) and the corresponding authors. The
language, images and information contained in this publication includes reference to
Indigenous knowledge including traditional knowledge, traditional cultural expression
and references to biological resources (plants and animals) of the Mirarr people. The
source Indigenous knowledge is considered “Confidential Information”; traditional law
and custom applies to it and the Mirarr people assert ownership over it. Any Mirarr
related language, images and information are published with the consent of Gundjeihmi
Aboriginal Corporation as the representative of the Mirarr people for the purposes of
education and specifically for use only in the context of this published work. Please
ab
de
c
f
mx
ph
phy
mx
ph
phy
phy
fi
Fig. 4 Examples of Arecaeae family stem tissue from Phase 2. a–cTransverse section of Arecaceae stem Type A (cf. Livistona spp.) from C2/44, ascale
bar is 1 mm, bclose-up of fibrovascular bundle, scale bar is 100 µm, cclose-up of globular echinate phytoliths, scale bar is 30 µm; d,ftransverse section of
Arecaceae stem Type B from C2/44, dscale bar is 500 µm, eclose-up of fibrovascular bundle, scale bar is 200 µm; ftransverse section of Arecaceae stem
Type B from C2/45, close-up of globular echinate phytoliths, scale bar is 20 µm. See the supporting online information and Supplementary Fig. 8 for a
detailed identification proof of Arecaceae stem Type A (cf. Livistona spp.) and the corresponding reference material. ph phloem, mx metaxylem, phy
phytolith.
ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-020-14723-0
6NATURE COMMUNIC ATIONS | (2020) 11:924 | https://d oi.org/10.1038/s41467-020-14723-0 | www.nature.com/naturecom munications
Content courtesy of Springer Nature, terms of use apply. Rights reserved

contact Gundjeihmi Aboriginal Corporation to request permission to refer to any
Indigenous knowledge in this publication.
Received: 5 March 2019; Accepted: 29 January 2020;
References
1. Jones, M. in The Evolution of Hominin Diets: Integrating approaches to the
study of Palaeolithic subsistence (eds Hublin, J.-J. & Richards, M. P.) 171–180
(Springer, 2009).
2. Flannery, K. V. in The Domestication and Exploitation of Plants and Animals
(eds Ucko, P. J. & Dimbleby, G. W.) 73–100 (Duckworth, 1969).
3. Stiner, M. C. Thirty years on the “Broad Spectrum Revolution”and paleolithic
demography. Proc. Natl Acad. Sci. USA 98, 6993–6996 (2001).
4. Henry, A. G., Brooks, A. S. & Piperno, D. R. Plant foods and the dietary ecology
of Neanderthals and early modern humans. J. Hum. Evol. 69,44–54 (2014).
5. Larbey, C., Mentzer, S. M., Ligouis, B., Wurz, S. & Jones, M. K. Cooked starchy
food in hearths ca. 120 kya and 65 kya (MIS 5e and MIS 4) from Klasies River
Cave, South Africa. J. Hum. Evol. 131, 210–227 (2019).
6. Mercader, J. Mozambican grass seed consumption during the Middle Stone
Age. Science 326, 1680–1683 (2009).
7. Wadley, L., Backwell, L., d’Errico, F. & Sievers C. Cooked starchy rhizomes in
Africa 170 thousand years ago. Science 367 87–91 (2020).
8. Barker, G. et al. The ‘human revolution’in lowland tropical Southeast Asia:
the antiquity and behavior of anatomically modern humans at Niah Cave
(Sarawak, Borneo). J. Hum. Evol. 52, 243–261 (2007).
9. Barton, H. The case for rainforest foragers: the starch record at Niah Cave,
Sarawak. Asian Perspect. 44,56–72 (2005).
10. Barton, H. & Paz, V. in Rethinking Agriculture: Archaeological and
ethnoarchaeological perspectives (eds Denham, T. P., Iriarte, J. & Vrydaghs, L.)
50–77 (Left Coast Press, 2007).
11. Fairbairn, A. S., Hope, G. S. & Summerhayes, G. R. Pleistocene occupation of
New Guinea’s highland and subalpine environments. World Archaeol. 38,
371–386 (2006).
12. Summerhayes, G. R. et al. Human adaptation and plant use in highland New
Guinea 49,000 to 44,000 years ago. Science 330,78–81 (2010).
13. Clarkson, C. et al. Human occupation of northern Australia by 65,000 years
ago. Nature 547, 306–310 (2017).
14. Langley, M. C., Clarkson, C. & Ulm, S. From small holes to grand narratives:
the impact of taphonomy and sample size on the modernity debate in
Australia and New Guinea. J. Hum. Evol. 61, 197–208 (2011).
15. Mellars, P. The impossible coincidence. A single-species model for the origins
of modern human behavior in Europe. Evol. Anthropol. 14,12–27 (2005).
16. Groucutt, H. S. et al. Rethinking the dispersal of Homo sapiens out of Africa.
Evol. Anthropol. 24, 149–164 (2015).
17. Mellars, P., Gori, K. C., Carr, M., Soares, P. A. & Richards, M. B. Genetic and
archaeological perspectives on the initial modern human colonization of
southern Asia. Proc. Natl Acad. Sci. USA 110, 10699–10704 (2013).
18. O’Connell, J. F. & Allen, J. The restaurant at the end of the universe: modelling
the colonisation of Sahul. Aust. Archaeol. 74,5–17 (2012).
19. Boivin, N. L. et al. Ecological consequences of human niche construction:
examining long-term anthropogenic shaping of global species distributions.
Proc. Natl Acad. Sci. USA 113, 6388–6396 (2016).
20. Roberts, P. & Stewart, B. A. Defining the ‘generalist specialist’niche for
Pleistocene Homo sapiens. Nat. Hum. Behav. 2, 542–550 (2018).
21. Roberts, P. et al. Fruits of the forest: Human stable isotope ecology and
rainforest adaptations in Late Pleistocene and Holocene (approximately 36 to
3ka) Sri Lanka. J. Hum. Evol. 106, 102–118 (2017).
22. Wedage, O. et al. Specialized rainforest hunting by Homo sapiens ~45,000
years ago. Nat. Commun. 10, 739 (2019).
23. Blinkhorn, J., Achyuthan, H., Ditchfield, P. & Petraglia, M.
Palaeoenvironmental dynamics and Palaeolithic occupation at Katoati, Thar
Desert, India. Quat. Res. 87, 298–313 (2017).
24. Groucutt, H. S. et al. Homo sapiens in Arabia by 85,000 years ago. Nat. Ecol.
Evol. 2, 800–809 (2018).
25. Fullagar, R., Field, J., Denham, T. & Lentfer, C. Early and mid Holocene tool-
use and processing of taro (Colocasia esculenta), yam (Dioscorea sp.) and other
plants at Kuk Swamp in the highlands of Papua New Guinea. J. Archaeol. Sci.
33, 595–614 (2006).
26. Clarkson, C. et al. Reply to comments on Clarkson et al. (2017) ‘Human
occupation of northern Australia by 65,000 years ago’.Aust. Archaeol. 84,
84–89 (2018).
27. Wood, R. Comments on the chronology of Madjedbebe. Aust. Archaeol. 83,
172–174 (2017).
28. Specht, R. L. in Australian Vegetation (ed Groves, R. H.) 321–344 (Cambridge
University Press, 1994).
29. Northern Territory Government Flora NT: Northern Territory flora online,
http://eflora.nt.gov.au/home (2013).
30. Fox, G. & Garde, M. in An-me Arri-ngun, The Food We Eat: Traditional plant
foods of the Kundjeyhmi people of Kakadu National Park. (Gundjeihmi
Aboriginal Corporation, 2018).
31. Meehan, B., Gaffey, P. & Jones, R. in Readings in Material Culture Vol. 9
Occasional papers in anthropology (ed Lauer, P. K.) 73–96 (Anthropology
Museum, University of Queensland, 1978).
32. Leichhardt, L. in Journal of an Overland Expedition in Australia, from
Moreton Bay to Port Essington, a Distance of Upwards of 3000 Miles, During
the Years 1844–1845. (T. & W. Boone, 1847).
33. Low, T. in Wild Food Plants of Australia. (Angus & Robertson Publishing, 1991).
34. Hather, J. G. in Archaeological Parenchyma. (Archetype Publications, 2000).
35. Russell-Smith, J. et al. Aboriginal resource utilization and fire management
practice in Western Arnhem Land, monsoonal northern Australia: Notes for
prehistory, lessons for the future. Hum. Ecol. 25, 159–195 (1997).
36. Fullagar, R. & Field, J. Pleistocene seed-grinding implements from the
Australian arid zone. Antiq 71, 300–307 (1997).
37. Fullagar, R. et al. Evidence for Pleistocene seed grinding at Lake Mungo,
south-eastern Australia. Archaeol. Oc 50,3–19 (2015).
38. van Welzen, P. C., Parnell, J. A. N. & Ferry Silk, J. W. Wallace’s Line and plant
distributions: Two or three phytogeographical areas and where to group Java?
Biol. J. Linn. Soc. 103, 531–545 (2011).
39. Crisp, M. D., Isagi, Y., Kato, Y., Cook, L. G. & Bowman, D. M. Livistona palms
in Australia: ancient relics or opportunistic immigrants? Mol. Phylogenet. Evol.
54, 512–523 (2010).
40. Nesbitt, M. Plants and People in Ancient Anatolia. Biblic. Archaeol. 58,68–81
(1995).
41. Hageman, J. B. & Goldstein, D. J. An integrated assessment of
archaeobotanical recovery methods in the Neotropical rainforest of northern
Belize: flotation and dry screening. J. Archaeol. Sci. 36 2841–2852 (2009).
42. Hillman, G. C. in Foraging and Farming: The evolution of plant exploitation
(eds Harris, D. R. & Hillman, G. D.) Ch. 13, 207–239 (Unwin Hyman, 1989).
43. Mcarthur, M. in Anthropology and Nutrition (ed. Mountford, C. P.) 2,90–135
(Melbourne University Press, Melbourne, 1960).
44. Wright, P. Preservation or destruction of plant remains by carbonization?
J. Archaeol. Sci. 30, 577–583 (2003).
45. Johansen, D. A. Plant Microtechnique. (McGraw Hill, 1940).
46. Marwick, B. & Florin, S. A. The First Australian Plant Foods at Madjedbebe,
65,000–53,000 years ago. https://doi.org/10.17605/OSF.IO/YDUZP (2019).
47. Norman, K. et al. An early colonisation pathway into northwest Australia 70-
60,000 years ago. Quat. Sci. Rev. 180, 229–239 (2018).
48. Story, R. in Lands of the Alligator Rivers Area, Northern Territory 89-111
(Commonwealth Scientific and Industrial Research Organization, Australia,
1976).
49. Story, R. in Lands of the Adelaide-Alligator Area, Northern Territory Land
Research Series 114-130 (Commonwealth Scientific and Industrial Research
Organization, Australia, 1969).
Acknowledgements
The archaeobotanical and ethnobotanical research in this project was funded by a
Wenner Gren Dissertation Fieldwork Grant (9260), an Australian Institute of Nuclear
Science and Engineering Postgraduate Research Award (11877), a Dan David Scholar-
ship and an Australian Postgraduate Award awarded to S.A.F. The initial fieldwork and
excavation of Madjedbebe was funded by an Australian Research Council grant
(DP110102864) obtained by C.C., B.M., R.F., M.S. and L.W. The authors are grateful to
the custodians of Madjedbebe, the Mirarr Senior Traditional Owners (Yvonne Margarula
and MN) and our research partner, the Gundjeihmi Aboriginal Corporation, for per-
mission to carry out this research and publish this paper. We are also grateful to Justin
O’Brien and David Vadiveloo for assistance in the field. We thank Dr Xavier Carah for
implementing the archaeobotanical recovery program at Madjedbebe and for his assis-
tance alongside Elspeth Hayes, Kasih Norman, Ashleigh Rogers, Makayla Harding and
Kate Connell in the collection of plants and ethnobotanical data. This research was
completed in part in Kakadu National Park under Permit No. RK870, RK909; however,
the findings and views expressed are those of the authors and do not necessarily
represent the views of Parks Australia, the Director of National Parks or the Australian
Government. The authors acknowledge the assistance of the staff at Parks Australia, the
Northern Territory Herbarium, and the University of Queensland’s Archaeology, School
of Earth Sciences, and Centre for Microscopy and Microanalysis Laboratories (especially
Rachel Price, Linda Northdurft and Ian Cowie). We also thank the student volunteers
from the University of Queensland for helping sort the heavy residue. Finally, the authors
thank Quan Hua, Rachel Wood, Zenobia Jacobs, Elspeth Hayes and Alison Crowther for
their advice and expert opinion during the drafting process.
Author contributions
S.A.F. and A.F. conducted the archaeobotanical research. S.A.F., M.N. and Dj.Dj. con-
ducted the ethnobotanical research and modern reference collection. S.A.F., A.F. and
C.C. wrote the main text. M.N. and Dj.Dj. provided cultural knowledge and assisted in
NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-020-14723-0 ARTICLE
NATURE COMMUNIC ATIONS | (2020) 11:924 | https://doi.org /10.1038/s41467-020-14723-0 | www.nature.com/naturecommunications 7
Content courtesy of Springer Nature, terms of use apply. Rights reserved

the interpretation of the findings. S.A.F., B.M., C.C. and A.F. created the figures. C.C.,
R.F., B.M., M.S. and L.W. obtained the funding for and conducted the excavations.
Competing interests
The authors declare no competing interests.
Additional information
Supplementary information is available for this paper at https://doi.org/10.1038/s41467-
020-14723-0.
Correspondence and requests for materials should be addressed to S.A.F. or C.C.
Peer review information Nature Communications thanks Tim Denham, Anne Ford and
Juan José García-Granero for their contribution to the peer review of this work. Peer
reviewer reports are available.
Reprints and permission information is available at http://www.nature.com/reprints
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative
Commons license, and indicate if changes were made. The images or other third party
material in this article are included in the article’s Creative Commons license, unless
indicated otherwise in a credit line to the material. If material is not included in the
article’s Creative Commons license and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this license, visit http://creativecommons.org/
licenses/by/4.0/.
© The Author(s) 2020
ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-020-14723-0
8NATURE COMMUNIC ATIONS | (2020) 11:924 | https://d oi.org/10.1038/s41467-020-14723-0 | www.nature.com/naturecom munications
Content courtesy of Springer Nature, terms of use apply. Rights reserved

1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com