ArticlePDF Available

Abstract and Figures

Aboriginal Australians represent one of the longest continuous cultural complexes known. Archaeological evidence indicates that Australia and New Guinea were initially settled approximately 50 thousand years ago (ka); however, little is known about the processes underlying the enormous linguistic and phenotypic diversity within Australia. Here we report 111 mitochondrial genomes (mitogenomes) from historical Aboriginal Australian hair samples, whose origins enable us to reconstruct Australian phylogeographic history before European settlement. Marked geographic patterns and deep splits across the major mitochondrial haplogroups imply that the settlement of Australia comprised a single, rapid migration along the east and west coasts that reached southern Australia by 49–45 ka. After continent-wide colonization, strong regional patterns developed and these have survived despite substantial climatic and cultural change during the late Pleistocene and Holocene epochs. Remarkably, we find evidence for the continuous presence of populations in discrete geographic areas dating back to around 50 ka, in agreement with the notable Aboriginal Australian cultural attachment to their country.
Content may be subject to copyright.
180 | NATURE | VOL 544 | 13 APRIL 2017
ARTICLE doi:10.1038/nature21416
Aboriginal mitogenomes reveal 50,000
years of regionalism in Australia
Ray Tobler1*, Adam Rohrlach2,3*, Julien Soubrier1,4, Pere Bover1, Bastien Llamas1, Jonathan Tuke2,3, Nigel Bean2,3,
Ali Abdullah-Highfold5, Shane Agius5, Amy O’Donoghue5, Isabel O’Loughlin5, Peter Sutton5,6, Fran Zilio5, Keryn Walshe5,
Alan N. Williams7, Chris S.M. Turney7, Matthew Williams1,8, Stephen M. Richards1, Robert J. Mitchell9, Emma Kowal10,
John R. Stephen11, Lesley Williams12, Wolfgang Haak1,13§ & Alan Cooper1,14§
At the time of initial human colonization (around 50 ka)
1,2
, Australia
and New Guinea were connected as a single landmass (termed Sahul)
that remained contiguous until separated by rising sea levels around
9 ka (ref. 3). Despite this, the initial Sahul colonists appear to have
rapidly diverged into distinct New Guinean and Australian popu-
lations, with limited signs of subsequent gene flow4–12—although
genetic data remains sparse. Little is known about the post-coloni-
zation diversification of Australian lineages or the effects of major
environmental and cultural changes over the last 50 thousand years
(kyr). Palaeoclimatically, these include continental-scale aridifi-
cation and cooling of Australia during the Last Glacial Maximum
(21 ± 3 ka), warming in the early Holocene (9–6 ka), and intensification
of the El Niño/Southern Oscillation during the mid-to-late Holocene
(4–2 ka)13,14. Substantial changes in the cultural record are not observed
until the terminal Pleistocene and Holocene, and include the formation
of the Panaramittee art style, the spread of the Pama–Nyungan group
of languages across most of the continent, and the increase in diversity
and complexity of technology and resource exploitation15,16. Aboriginal
history is inextricably interwoven with the Australian landscape and
is culturally expressed through the central importance of kin group
attachment to ‘country’, and further reinforced through Songlines and
Dreaming narratives
17
. Close relationships to the landscape are likely to
have played an important role in surviving the extreme environmental
changes of late Pleistocene Australia.
Reconstructing the genetic history of Aboriginal Australia is greatly
complicated by past government policies of enforced population
relocation and child removal that have eroded much of the physical
connection between groups and geography in modern Australia.
However, a unique opportunity is provided by a remarkable set of hair
samples and detailed ethnographic metadata collected with permission
from more than 5,000 Aboriginal Australians during expeditions run by
the Board for Anthropological Research (BAR) from the University of
Adelaide between the 1920s and 1970s (Supplementary Information).
The extensive genealogical and geographical information collected with
the samples allows detailed reconstruction of the genetic and historical
relationships between Aboriginal Australian groups before the effects
of European colonization.
Dataset
We obtained informed consent from hair donors or their families
(Supplementary Information) to perform genetic analyses and
sequenced complete mitogenomes from hair samples of 111 indi-
viduals across three different Aboriginal communities (Point Pearce,
South Australia; Cherbourg, Queensland; Koonibba, South Australia;
Supplementary Information). Using the genealogical and cultural
metadata, we traced the geographic origin of each individual (referred
to as BAR samples) as far back as possible along the ancestral maternal
lineage. The resulting broad geographic range is shown in Extended
Data Fig. 1. We identified 54 unique mtDNA haplotypes, which fell
into the five major mitochondrial haplogroups S, O, M, P and R
that have been described previously for Aboriginal Australia9,10,12
(Supplementary Information). Phylogenetic relationships were
Aboriginal Australians represent one of the longest continuous cultural complexes known. Archaeological evidence
indicates that Australia and New Guinea were initially settled approximately 50 thousand years ago (ka); however, little is
known about the processes underlying the enormous linguistic and phenotypic diversity within Australia. Here we report
111 mitochondrial genomes (mitogenomes) from historical Aboriginal Australian hair samples, whose origins enable us to
reconstruct Australian phylogeographic history before European settlement. Marked geographic patterns and deep splits
across the major mitochondrial haplogroups imply that the settlement of Australia comprised a single, rapid migration
along the east and west coasts that reached southern Australia by 49–45 ka. After continent-wide colonization, strong
regional patterns developed and these have survived despite substantial climatic and cultural change during the late
Pleistocene and Holocene epochs. Remarkably, we find evidence for the continuous presence of populations in discrete
geographic areas dating back to around 50 ka, in agreement with the notable Aboriginal Australian cultural attachment
to their country.
1Australian Centre for Ancient DNA, School of Biological Sciences, The University of Adelaide, Adelaide, South Australia 5005, Australia. 2School of Mathematical Sciences, The University of
Adelaide, Adelaide, South Australia 5005, Australia. 3ARC Centre of Excellence for Mathematical and Statistical Frontiers, The University of Adelaide, Adelaide, South Australia 5005, Australia.
4Genetics andMolecular Pathology, SA Pathology, Adelaide, South Australia 5000, Australia. 5South Australian Museum, Adelaide, South Australia 5005, Australia. 6School of Biological Sciences,
The University of Adelaide, Adelaide, South Australia 5005, Australia. 7Palaeontology, Geobiology and Earth Archives Research Centre, and Climate Change Research Centre, School of Biological,
Earth and Environmental Sciences, University of New South Wales, Sydney, New South Wales 2052, Australia. 8School of Archaeology and Anthropology, College of Arts and Social Sciences,
Australian National University, Canberra, Australian Capital Territory 0200, Australia. 9Department of Biochemistry and Genetics, La Trobe University, Melbourne, Victoria 3086, Australia.
10Alfred Deakin Institute, Deakin University, Melbourne, Victoria 3125, Australia. 11Australian Genome Research Facility, The Waite Research Precinct, Adelaide, South Australia 5064, Australia.
12Community Elder and Cultural Advisor, Cherbourg, Queensland, Australia. 13Department of Archeogenetics, Max Planck Institute for the Science of Human History, 07745 Jena, Germany.
14Environment Institute, The University of Adelaide, Adelaide, South Australia 5005, Australia.
*These authors contributed equally to this work.
§These authors jointly supervised this work.
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Article reSeArcH
13 APRIL 2017 | VOL 544 | NATURE | 181
analysed with other full mtDNA haplotypes from Aboriginal
Australians and Melanesians (44 and 25 samples, respectively, 123
unique mtDNA lineages in total).
Dating the colonization of Sahul
The timing of human arrival in Australia was estimated using the
age of the most recent common ancestor (TMRCA) for the different
Australian-only haplogroups, calculated using a molecular clock
with substitution rates calibrated with ancient European and Asian
mitogenomes18. Although these TMRCA values are likely to be minimal
estimates given the limited sampling, they group in a narrow window
of time from approximately 43–47 ka (Fig. 1 and Extended Data
Figs 2, 3), consistent with previous studies (Supplementary
Information). To examine the accuracy of this molecular age estimate
we re-analysed a comprehensive suite of radiocarbon and optically
stimulated luminescence ages from early archaeological sites across
Sahul using currently available calibration datasets19 and the phase
function in OxCal 4.2.4. The resulting independent estimate for initial
colonization of Sahul, 48.8 ± 1.3 ka, is a close match to the genetic age
estimates (Fig. 1 and Supplementary Table 4). Indeed, the basal splits
between haplogroups O, S and N13, P and R, M16 and M42 (Fig. 1)
might reflect the initial within-Australia events, around 50 ka. However,
we have taken a conservative approach and assumed these reflect
lineages present in the initial population colonizing Sahul, as suggested
by the presence of basal sister clades of Melanesian and Aboriginal
Australian lineages within haplogroups M and P (Fig. 1).
Aboriginal Australian phylogeography
Phylogenetic analysis of all Aboriginal Australian samples with reliable
geographical information (74 BAR samples and two from previous
mtDNA studies8,14, 76 lineages in total; see Methods), revealed large-
scale phylogeographic patterns for each major haplogroup (Fig. 2). For
example, none of the haplogroup O lineages were found in eastern
Australia, which was dominated by haplogroups P, S and M42a. Within
the two main Australian P-clades (based around P5 and P4b1) there
was a clear split between northeastern and Riverine/South Australia
(Fig. 2). Similar patterns are observed in the other major haplogroups,
indicating that Aboriginal Australian mitochondrial lineages have
undergone limited amounts of dispersal over time, and related line-
ages are grouped geographically. Furthermore, the basal lineages within
each major haplogroup were mostly in northern Australia, presumably
reflecting early divergences as members of the founding populations
remained while others moved south where more derived lineages were
observed. Together with the deep divergences among the mtDNA
lineages, these results suggest that populations were structured by the
initial major population movements following colonization around
50 ka (Fig. 1).
To verify that the small sample sizes are not biasing the phylogeo-
graphic patterns, we used a novel correlation test based on the results
of a multiple correspondence analysis to examine the 76 mtDNA
lineages with reliable provenance. This method is a generalization, for
individual haplotypes, of the principal component analysis used for
population genetic analyses of diploid genotypes. The major axes of
variation among the pooled haplotype data are determined and then
used to test for significant correlations with supplementary variables
of interest. The test showed strong phylogeographic clustering among
Aboriginal Australian mtDNA lineages, and a significant correlation
between the phylogenetic structure between and within each hap-
logroup and both the latitudinal and longitudinal origin of the samples
(Table 1 and Extended Data Table 1). As a second test for relative geo-
graphic structure, we applied a Mantel test to find correlations between
pairwise distances for individuals calculated from geographic and
genetic coordinates (from the multiple correspondence analysis). We
also found significant correlations between these distances, both within
and between haplogroups, indicating (geographically) neighbouring
individuals were closely related genetically (Table 1 and Extended
Data Table 1). These findings confirm that there was strong phylogeo-
graphic clustering among Aboriginal Australian mtDNA lineages before
European colonization, differentiated along latitudinal and longitudinal
gradients, indicating that there were very limited amounts of geographic
*
50 010203040
Time (ka)
M
P
O
S
45.7
(35.4–56.6)
45.7
(36.1–56.1)
43.8
(33.6–55.9)
42.5
(31.6–61.3)
47.2
(36.9–58.6)
M42
P5
P4b1
S2
S1
R
N
O1
O2
M16
LGM
Figure 1 | Australian mtDNA phylogeny. Phylogenetic analysis of
Aboriginal Australian and Melanesian (dashed grey lines) mitogenomes
using BEAST31, showing the four major haplogroups detected in Australia
(in colour), along with other Aboriginal Australian lineages not used in
dating analyses (solid black lines). The age of the most recent common
ancestor (TMRCA) and 95% highest posterior density intervals were
calculated for each Aboriginal-Australian-only clade (red dots) using
human mitochondrial evolutionary rates calibrated with Palaeolithic
European and Asian mitogenomes18,32 to minimize the effects of rate
temporal dependency33,34 (see Methods). The posterior distributions for
each TMRCA are shown behind the phylogeny, in matching colours.
The dark grey box represents the initial colonization of Australia indicated
by archaeological evidence at 48.8 ± 1.3 ka (see Methods). The light grey
box indicates the period when mitochondrial lineages were still sorting
into Australia or New Guinea/Melanesia, which occurred during the
initial colonization of Sahul. Genetic divergences during this time (for
example, between M16 and M42, or O and N) might have occurred outside
Australia, and were excluded from TMRCA calculations. The short branch
length of an ancient S2 sequence14 reflects the radiocarbon-dated age
of the specimen. The early Holocene diversification of lineages within
haplogroup O2 is indicated with an asterisk. LGM, Last Glacial Maximum.
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Article
reSeArcH
182 | NATURE | VOL 544 | 13 APRIL 2017
dispersal given the long time periods involved. Similarly, an additional
set of Aboriginal Australian mtDNA genomes recently generated as
part of a genomic study
12
show a concordant phylogeographic distribu-
tion to the patterns in our data (Extended Data Fig. 4). However, these
sequences are not available and the samples lack information about
pre-European distributions, complicating historical analysis.
Migratory patterns and regionalism within Australia
The phylogeographic distribution of the major Aboriginal haplogroups
are consistent with coastal colonization models of Australia
20,21
where
the initial Sahul colonizers spread across northern Australia, and then
south along the east (haplogroups P, S, M42a) and west (haplogroups
O, R) coasts in parallel clockwise and counter-clockwise movements
(Fig. 3). The disjunction between haplogroups O and S in central
southern Australia (Fig. 2) potentially reflects a meeting of the two
movements. Limited genetic surveys in Tasmania are consistent
with this model, because haplogroups P, S and M were detected, but
not haplogroup O or R (ref. 22). A major migration corridor is also
apparent between northeastern and southern Australia, potentially
along the Murray–Darling River23.
The 49–45 ka age range recently reported from Warratyi rock
shelter24, Flinders Ranges, South Australia is close in age to the
earliest sites reported from northern Australia1. To similarly con-
strain the timing of human arrival in the far southwest of Australia, we
re-examined the multi-dated sequence of Devil’s Lair, southwestern
Australia (Extended Data Fig. 5) along with continental-wide earliest
occupation ages (Supplementary Table 4). The resulting age estimate
(47.8 ± 1.5 ka), together with multiple early occupation sites across
southern Australia (Fig. 3 and Extended Data Fig. 6) suggest the initial
expansion around Australia was very rapid, perhaps taking only a few
thousand years. The initial human colonization considerably preceded
the extinction of the last megafauna25, as indicated by the presence
of the Diprotodont Zygomaturus at 42 ka just south of the Flinders
Ranges
26
, and this temporal overlap is similar to the pattern recently
reported for South America27.
The marked population structure of deeply diverged Aboriginal
Australian mitogenomes appears to date back to the original arrival of
people on the Australian part of Sahul. These patterns are surprising
given the pronounced environmental changes that have occurred since
P
OS
M
P4b1
P5
S1
M42a
S2
M16
Sample size
136
O2
O1
Figure 2 | Australian mtDNA phylogeography. Phylogeographic
distributions of Aboriginal Australian mitogenome haplotypes, grouped into
the four major haplogroups O, S, P and M with timescales calculated using
an ancient-DNA-calibrated molecular clock (see Methods). Lineages from
samples in the current study (circles) are shown at the location of the oldest
known maternal ancestor recorded in genealogical and geographic data,
generally before the effects of European colonization. Triangles represent
data from modern samples reported in previous studies. The size of the
symbols reflects the number of identical haplotypes as indicated in the
figure. Identical sequences from the same location were pruned, whereas
those from multiple locations were only used where they could not be
explained through genealogical records. Coloured circles and lines represent
haplotypes with known geographical provenance, with colours matching the
cluster assignments of the multiple correspondence analysis (Supplementary
Table 3), whereas grey (empty) circles represent the geographic distribution
of samples not falling within each specific haplogroup. Previously published
haplotypes that lack detailed geographic data histories are shown with
yellow triangles (and black lines) for each haplogroup, whereas those with
no associated locations are shown on the tree as black branches alone.
Map data was sourced from the Oak Ridge National Laboratory Distributed
Active Archive Center (https://webmap.ornl.gov/wcsdown/wcsdown.
jsp?dg_id= 10003_1).
Table 1 | Australian phylogeography test results
Haplogroup O S
M (without
M16) P
Longitude 0.6395
(0.0629)*
0.3351
(0.0016)* * *
0.642
(0.0929)*
0.7796
(0.0002)* * *
Latitude 0.5010
(0.0083)* * *
0.5977
(0.0006)* * *
0.8560
(0.0055)* * *
0.8690
(4 × 106)* * *
Mantel test 0.3352
(0.0176)* *
0.2695
(0.0374)* *
0.3273
(0.0953)*
0.4488
(3 × 106)* * *
Tests based on multiple correspondence analysis of phylogeographic structure within the major
Aboriginal Australian haplogroups reveal signicant correlations with latitude and longitude,
implying lineages are likely to be found in certain geographic locations. Mantel tests conrm the
lineages are grouped geographically on the landscape, implying that neighbouring individuals
are expected to share common ancestry (see Methods). For each haplogroup, the correlation
coecient is given for the dimension with the most signicant correlation in the case of longitude
and latitude, along with the P value in brackets (* P < 0.1; * * P < 0.05; * * * P < 0.01). Although not
every principal dimension is signicantly correlated with geography, we would not expect that this
is the only driver for lineage distribution.
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Article reSeArcH
13 APRIL 2017 | VOL 544 | NATURE | 183
initial colonization. The most extreme example of this is the widespread
aridification and cooling of the Last Glacial Maximum, during which
archaeological models suggest pronounced geographic contraction of
populations and abandonment of large parts of the continent28. The
diversity and grouping of Aboriginal Australian mitogenome data
indicate that Aboriginal Australian populations survived these changes
without large-scale movements, although there is potential evidence
for a late-glacial (approximately 15 ka) re-expansion into the Western
central desert (Extended Data Fig. 4 and Supplementary Information).
Notably, both the diversity of mitochondrial lineages and population
size estimates during this time period do not suggest severe population
bottlenecks (Fig. 1 and Extended Data Fig. 7), indicating that many
populations survived in local refugia that may have been cryptic to the
archaeological record29.
Holocene intensification
The rapid diversification of derived haplotypes within haplogroup O2 is
indicative of a population expansion around 7 ka in southern Australia
(Fig. 1), but this is the only obvious genetic signal that coincides with
the mid-Holocene climatic optimum (9–6 ka) and the increasing
accessibility of the arid interior to hunter–gatherer groups13,15. The
above suggests that the extensive cultural changes evident during
the Holocene, including the establishment of Panaramittee rock
art, spread of the Pama–Nyungan languages, adoption of complex
and diversified technologies (for example, seed grinding, wooden
toolkits), advanced food-processing techniques (of, for example,
Macrozamia plants), and greater reliance on marine resources,
may have been the result of demographic change and/or cultural
transmission, rather than population movement or replacement
15
. In
this regard, recent archaeological models propose that rapid demo-
graphic growth during the Holocene led to reduced mobility and a
consequent greater investment in technology15. It is also possible that
some cultural changes were entirely male-mediated, and therefore
not apparent in mtDNA data. Recent genomic data from modern
Aboriginal Australians has been used to tentatively link the spread
of the Pama–Nyungan languages to an early Holocene population
expansion in northeast Australia, and limited gene flow to the rest
of Australia
12
. However, the strength of the genetic signal for both
the population expansion and movement remains ambiguous at best
(Supplementary Information).
Discussion
The long-standing and diverse phylogeographic patterns documented
here are remarkable given the timescale involved, and raise the pos-
sibility that the central cultural attachment of Aboriginal Australians
to ‘country’ may reflect the continuous presence of populations in
discrete geographic areas for up to 50 kyr. The very limited geographical
movement of populations over time is consistent with observations
of nomadic sedentism in recent Aboriginal Australian societies,
where ranging was anchored in localized, collective and stable land/
language ownership units, and occurred within a broad environmental
region17 (Supplementary Information). This form of subsistence (and
territoriality) might also explain the notable lack of exchange
between New Guinea and Australian mitochondrial lineages,
despite a land bridge between the two until about 9 ka. Overall,
these patterns are similar to recent reports of marked mitochondrial
phylogeography in early South American populations30, and
raise the possibility that hunter–gatherer groups were capable of
exhibiting pronounced regionalism, or at least female philopatry,
over prolonged time periods.
The mitochondrial dates reported here for Aboriginal Australian
arrival and dispersal appear considerably older than recent estimates
from nuclear-genomic data
12
that suggest a single ancestral population
started to differentiate as recently as 10–32 ka, following an admixture
event with Denisovans around 43 ka. The latter event, at least, is incon-
sistent with the Australian archaeological record that does not support
the presence of Denisovans, indicating that any admixture must have
occurred before the colonization of Sahul around 50 ka. This raises the
possibility that the molecular-dating analyses of the nuclear-genomic
data have been confounded by complex population histories, including
multiple hominin introgressions12 and/or patterns of selection
(Supplementary Information). By contrast, when combined with
detailed phylogeographical data, mitogenome dating may provide a
less complex alternative to reconstructing human colonization patterns
in situations such as Australia.
Online Content Methods, along with any additional Extended Data display items and
Source Data, are available in the online version of the paper; references unique to
these sections appear only in the online paper.
Received 9 August 2016; accepted 24 January 2017.
Published online 8 March 2017.
1. Roberts, R. G., Jones, R. & Smith, M. A. Thermoluminescence dating of a
50,000-year-old human occupation site in northern Australia. Nature 345,
153–156 (1990).
2. O’Connell, J. F. & Allen, J. The process, biotic impact, and global implications of
the human colonization of Sahul about 47,000 years ago. J. Archaeol. Sci. 56,
73–84 (2015).
Tropical rainforest
Temperate forest
Palaeovegetation reconstruction
at sea level –120 m
Grassland
Savana/dry forest
Steppe
Xerophytic wood-shrubland
Desert
O2
P5
P
M42
a
(
P, S, M
)
43–39 ka
New
Guinea
Sahul
Sunda
47 ka
46 ka
42–40 ka
46–45 ka
45–44 ka
44–40 ka
40–39 ka
44 ka
48 ka
41 ka
44 ka
42 ka
40 ka
45 ka
46 ka
40 ka
Warratyi rock
shelter
Devil’s
Lair
R12
49–45 ka
O
1
M
42
O
1
M
42
R
S
Lake Carpentaria
Figure 3 | The peopling of Australia. Model of the peopling of Australia
combining genetic and archaeological data, showing approximate, and
stylised, coastal movements of haplogroups O and R (west) and P, S, and
M (east). The inferred movement of S into the interior is influenced by
the path of a recent study on water sources and human movement21.
Data from other studies where pre-European distributions are unclear
are indicated with a dagger (†), and include a potential late-glacial
movement into the western central desert region (blue dashed arrows;
see Methods). Early archaeological sites in Australia and New Guinea
(black dots) are given with mean ages for earliest occupation of sites in
each region (Supplementary Table 4). Insufficient data were available for
sites with white dots, which were not used in the age model for the initial
Sahul colonization date but provide independent age controls. Ages in
southwestern and south central Australia, at Devil’s Lair (49–46 ka) and
Warratyi rock shelter (49–45 ka), suggest that the overall population
movements were rapid and that the coastal regions of Australia were
colonized within a few thousand years. Approximate late Pleistocene
vegetation reconstructions are shown (from ref. 35). The map was adapted
from the figure in ref. 36, originally constructed by J.S.
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Article
reSeArcH
184 | NATURE | VOL 544 | 13 APRIL 2017
3. Lewis, S. E., Sloss, C. R., Murray-Wallace, C. V., Woodroe, C. D. & Smithers, S. G.
Post-glacial sea-level changes around the Australian margin: a review.
Quat. Sci. Rev. 74, 115–138 (2013).
4. Redd, A. J. & Stoneking, M. Peopling of Sahul: mtDNA variation in aboriginal
Australian and Papua New Guinean populations. Am. J. Hum. Genet. 65,
808–828 (1999).
5. Rasmussen, M. et al. An Aboriginal Australian genome reveals separate human
dispersals into Asia. Science 334, 94–98 (2011).
6. Fehren-Schmitz L. et al. A re-appraisal of the early Andean human remains
from Lauricocha in Peru. PloS One. 10, e0127141 (2015).
7. Bergström, A. et al. Deep roots for Aboriginal Australian Y chromosomes.
Curr. Biol. 26, 809–813 (2016).
8. Nagle, N. et al. Antiquity and diversity of aboriginal Australian Y-chromosomes.
Am. J. Phys. Anthropol. 159, 367–381 (2016).
9. Hudjashov, G. et al. Revealing the prehistoric settlement of Australia by
Y chromosome and mtDNA analysis. Proc. Natl Acad. Sci. USA 104,
8726–8730 (2007).
10. van Holst Pellekaan, S. Genetic evidence for the colonization of Australia.
Quat. Int. 285, 44–56 (2013).
11. Heupink, T. H. et al. Ancient mtDNA sequences from the First Australians
revisited. Proc. Natl Acad. Sci. USA 113, 6892–6897 (2016).
12. Malaspinas, A.-S. et al. A genomic history of Aboriginal Australia. Nature 538,
207–214 (2016).
13. Reeves, J. M. et al. Palaeoenvironmental change in tropical Australasia over the
last 30,000 years—a synthesis by the OZ-INTIMATE group. Quat. Sci. Rev. 74,
97–114 (2013).
14. Fitzsimmons, K. E. et al. Late Quaternary palaeoenvironmental change in the
Australian drylands. Quat. Sci. Rev. 74, 78–96 (2013).
15. Williams, A. N., Ulm, S., Turney, C. S. M., Rohde, D. & White, G. Holocene
demographic changes and the emergence of complex societies in prehistoric
Australia. PLoS One 10, e0128661 (2015).
16. Ulm, S. ‘Complexity’ and the Australian continental narrative: themes in the
archaeology of Holocene Australia. Quat. Int. 285, 182–192 (2013).
17. Sutton, P. Native title in Australia: an ethnographic perspective. (Cambridge Univ.
Press, 2003).
18. Fu, Q. et al. A revised timescale for human evolution based on ancient
mitochondrial genomes. Curr. Biol. 23, 553–559 (2013).
19. Reimer, P. J. et al. IntCal13 and marine13 radiocarbon age calibration curves
0–50,000 years cal BP. Radiocarbon 55, 1869–1887 (2013).
20.
Bowdler, S. in Sunda and Sahul: Prehistoric Studies in Southeast Asia, Melanesia and
Australia (eds Allen, J., Golson, J. & Jones, R.) 205–246 (Academic Press, 1977).
21. Bird, M. I., O’Grady, D. & Ulm, S. Humans, water, and the colonization of
Australia. Proc. Natl Acad. Sci. USA 113, 11477–11482 (2016).
22. Mcallister, P., Nagle, N. & Mitchell, R. J. Brief communication: the Australian
Barrineans and their relationship to Southeast Asian negritos: an investigation
using mitochondrial genomics. Hum. Biol. 85, 485–502 (2013).
23. White, J. P. & O’Connell, J. F. A Prehistory of Australia, New Guinea, and Sahul.
(Academic Press, 1982).
24. Hamm, G. et al. Cultural innovation and megafauna interaction in the early
settlement of arid Australia. Nature 539, 280–283 (2016).
25. Saltré, F. et al. Climate change not to blame for late Quaternary megafauna
extinctions in Australia. Nat. Commun. 7, 10511 (2016).
26. Roberts, R. G. et al. New ages for the last Australian megafauna: continent-wide
extinction about 46,000 years ago. Science 292, 1888–1892 (2001).
27. Metcalf, J. L. et al. Synergistic roles of climate warming and human occupation
in Patagonian megafaunal extinctions during the last deglaciation. Sci. Adv. 2,
e1501682 (2016).
Supplementary Information is available in the online version of the paper.
Acknowledgements We acknowledge the support and involvement of the Point
Pearce, Cherbourg and Koonibba communities and the individual families.
We also acknowledge the work of N. Tindale, J. Birdsell and members of
the original Board for Archaeological Research expeditions collecting the
specimens. We thank the South Australian Museum, Australian Research
Council, University of Adelaide Environment Institute, the Genographic
Project and Bioplatforms Australia for support, and S. Ulm, G. Gower, I.
Mathieson, L. O’Brien, S. Easteal, M. Vilar, C. Stringer and ACAD colleagues for
helpful comments and advice. The Aboriginal Heritage Project webpage is
https://www.adelaide.edu.au/acad/ahp/, and this work was carried out under
the auspices of the University of Adelaide Human Research Ethics Committee,
project approval H-2014-252.
Author Contributions The project was conceived by A.C., W.H. and P.S. and
directed by A.C. and W.H. Archival research and community outreach was
led by I.O., A.A-H., S.A., A.O., F.Z. and L.W. with A.C., W.H., R.T. and R.J.M. The
genetic sequencing was performed and coordinated by W.H., P.B., M.W., S.R.
and J.R.S., and the genetic analysis by W.H., R.T., A.R., J.S., J.T., N.B., B.L. and A.C.
Archaeological and anthropological interpretations were provided by P.S., C.T.,
A.N.W. and K.W. The manuscript was written by A.C. and R.T., with critical input
from P.S., C.T., A.N.W., A.R., J.S., W.H. and all other co-authors. R.T., J.S., A.N.W.
and A.R. compiled the Supplementary Information.
Author Information Reprints and permissions information is available at
www.nature.com/reprints. The authors declare no competing financial
interests. Readers are welcome to comment on the online version of the paper.
Correspondence and requests for materials should be addressed to
A.C. (alan.cooper@adelaide.edu.au).
Reviewer Information Nature thanks P. Bellwood, C. Lalueza-Fox and the other
anonymous reviewer(s) for their contribution to the peer review of this work.
28. Williams, A. N., Ulm, S., Cook, A. R., Langley, M. C. & Collard, M. Human refugia
in Australia during the Last Glacial Maximum and terminal Pleistocene:
a geospatial analysis of the 25–12ka Australian archaeological record.
J. Archaeol. Sci. 40, 4612–4625 (2013).
29. Smith, M. The Archaeology of Australia’s Deserts (Cambridge Univ. Press,
2013).
30. Llamas, B. et al. Ancient mitochondrial DNA provides high-resolution time
scale of the peopling of the Americas. Sci. Adv. 2, e1501385 (2016).
31. Drummond, A. J. & Rambaut, A. BEAST: Bayesian evolutionary analysis by
sampling trees. BMC Evol. Biol. 7, 214 (2007).
32. Posth, C. et al. Pleistocene mitochondrial genomes suggest a single major
dispersal of non-Africans and a late glacial population turnover in Europe.
Curr. Biol. 26, 827–833 (2016).
33. Ho, S. Y. W. et al. Time-dependent rates of molecular evolution. Mol. Ecol. 20,
3087–3101 (2011).
34. Rieux, A. & Balloux, F. Inferences from tip-calibrated phylogenies: a review and
a practical guide. Mol. Ecol. 25, 1911–1924 (2016).
35. Balme, J., Davidson, I., McDonald, J., Stern, N. & Veth, P. Symbolic behaviour
and the peopling of the southern arc route to Australia. Quat. Int. 202, 59–68
(2009).
36. Cooper, A. & Stringer, C. B. Paleontology. Did the Denisovans cross Wallace’s
Line? Science 342, 321–323 (2013).
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Article reSeArcH
METHODS
Samples. The 111 hair samples used in the present study were originally collected
during anthropological expeditions to one of the following communities:
Cherbourg, Queensland (23 samples), Point Pearce, South Australia
(41 samples) and Koonibba, South Australia (47 samples) (Extended Data Fig. 1
and Supplementary Table 1). Consent was obtained from the original donors,
or their descendants, according to protocols detailed in the Supplementary
Information. Six of the Koonibba samples were collected during an expedition
to the area between 13 and 25 August 1928, all remaining samples were obtained
from the extensive Harvard and Adelaide Universities Anthropological Expeditions
lead by N. B. Tindale and J. B. Birdsell that took place from 13 May 1938 to 30 June
1939. Hair was collected from different parts of the body, but all samples used in
the current study consist of small locks of hair that were cut with permission from
the head of participants. Since the initial collection date, the hair samples have
been stored in sealed paper envelopes. The envelopes are currently secured in a
restricted-access storage room maintained by the South Australian Museum. For
each sample, a portion of the hair (between 20–190 mg) was removed from each
envelope for use in the present study.
Ancient DNA analysis. The hair samples from Cherbourg and Point Pearce were
soaked in 3.5 ml of 1% bleach, rinsed in 7 ml of water, and subsequently 3.5 ml of
100% ethanol and before being air-dried. For the Koonibba samples, we applied
2 washes in 3 ml of water, a subsequent wash in 3 ml of 100% ethanol, followed
by air-drying. Each sample was digested for 1 h under constant rotation at 55 °C
in 4 ml of a digestion buffer containing 75 mM Tris pH 8.0, 50 mM NaCl (Sigma-
Aldrich), 0.5 mgml1 Proteinase K (Life Technologies), 50 mM DTT (Promega)
and 0.75% SDS (Life Technologies). After lysis, samples were centrifuged at
4,600 r.p.m. for 1 min and the supernatant was pipetted into 100 μ l silica suspen-
sion and 16 ml modified binding buffer (90% QG Buffer (Qiagen), 1.3% Triton
X-100 (Sigma-Aldrich), 25 mM NaCl (Sigma-Aldrich) and 0.2 M sodium acetate
(Sigma-Aldrich)), and left for 1 h at room temperature under constant rotation.
Silica suspensions subsequently pelleted using a centrifuge at 4,600 r.p.m. for 5 min,
and the supernatant was discarded. The silica pellet was washed three times in
80% ethanol and centrifugation at 13,000 r.p.m. for 1 min. After the last wash, the
pellet was air dried for 30 min and resuspended twice in 120 μ l of a pre-warmed
(at 50 °C) mix of EB buffer (Qiagen) and 0.05% Tween 20, and incubated for
10 min. After centrifugation at 13,000 r.p.m. for 1 min, a final 240 μ l extract was
obtained. Subsequently, 60 μ l extract was purified using a MinElute Reaction
Cleanup Kit (Qiagen) following the manufacturer’s protocol.
Double-stranded libraries were prepared following standard protocols
30,37,38
,
using short Illumina adapters with dual 5-mer (non-Koonibba samples) or
7-mer (Koonibba samples) internal barcodes. For the Koonibba samples partial
uracil-DNA-glycosylase (UDG) treatment
39
was performed for DNA repair in the
first step of library construction. Libraries for the Koonibba sample extracts were
amplified using Platinum Taq HiFi (Invitrogen), whereas the Cherbourg and Point
Pearce samples were amplified using isothermal amplification (TwistAmp Basic
kit, TwistDx Ltd). The latter were enriched by hybridization using mitochondrial
RNA baits prepared in-house and finally amplified using full-length 7-mer indexed
Illumina adapters (see ref. 6 for a full explanation of the protocol). Libraries
were pooled and sequenced in a HiSeq 2× 100 PE run. The Koonibba libraries
were amplified using full-length 7-mer indexed Illumina adapters and shotgun
sequenced in MiSeq (2× 150 PE) and NextSeq (2× 150 PE) Mid Output runs at
the Australian Genome Research Facility.
Mapping and consensus calling. Raw Illumina reads were processed using
the PaleoMix v1.0.140 pipeline. AdapterRemoval v2 (ref. 41) was used to trim
adaptor sequences, merge the paired reads, and eliminate all reads shorter than
25 bp. Filtered reads were then mapped to the Reconstructed Sapiens Reference
Sequence (RSRS) mitochondrial reference genome
42
with BWA v0.6.2 (ref. 43). The
minimum mapping quality was set to 25, seeding was disabled and the maximum
number or fraction of open gaps was set to 2. MapDamage v2 (ref. 44) was used
to check that the expected mapping and damage patterns were observed for each
library and re-scale base qualities for the non-repaired libraries (see Supplementary
Table 2 for library statistics).
All mtDNA genome consensus sequences were called using Geneious v9.1.3
(ref. 45). For each sample, reads were remapped to the RSRS reference using the
Geneious mapper (default settings, serial mapping iterated five times). To call a
base, each region required a coverage 3, with a majority allele frequency 0.75.
The resulting consensus sequences were then inspected by eye, with particular
attention being paid to the hypervariable regions and nucleotide positions
previously identified as being problematic on the phylotree website (http://www.
phylotree.org/)46. All ambiguous sites were called as ‘N’.
Identical haplotypes were collaps ed into a single haplotype sequence. Individuals
with genealogical information that indicated a shared common maternal ancestor
were checked for sequence similarity, and were identical in all but two cases where
they differed by a single nucleotide. These cases were subsequently maintained
as separate mtDNA haplotypes. For all individuals where identity by maternal
descent was unknown, two sequences were deemed as identical if their sequences
shared all diagnostic variants for a given haplogroup. After combining all common
haplotypes, a total of 54 non-redundant consensus sequences were determined
(from 111 original samples; Supplementary Table 1). The resulting consensus
haplotypes cover all the major mtDNA haplogroups previously described for
Australia (Supplementar y Information).
Phylogenetics. To help determine the timing of the split between Melanesian
and Australian populations, and the colonization history of Australia, the
phylogenetic software BEAST (v1.8.3)
31,47
was used on 123 complete (or mostly
complete) mtDNA genomes (54 unique Aboriginal Heritage Project (AHP) con-
sensus samples combined with 44 Australian and 25 Melanesian publicly available
sequences; see Supplementary Table 1). The non-AHP sequences were obtained
from the mitochondrial database mtDB48 and two recently published papers5,11.
Before analysis, all 123 mtDNA genomes were aligned to the RSRS with BLAT
49
and then analysed with a custom R script, so that indels were removed and only
point mutations relative to the RSRS were used in the subsequent analyses.
The TN93+ G6 model of nucleotide substitution was selected through
comparison of BIC scores using ModelGenerator v0.85 (ref. 50), a GMRF skyride
model
51
was used to allow for a complex population histor y, with a relaxed uncor-
related log-normal clock52 to account for rate heterogeneity between lineages
(a strict clock was empirically rejected as ucld.stdev posterior distribution did
not include zero). Monophyly was constrained for all major haplogroups and
the ancient sequence hap97 was given a tip date log-normal prior distribution
with a mean of 1,250years and a standard deviation of 0.7 (95% of the dates fall
between 500 and 3,000years; based on estimates from ref. 11). Two mutation
rates with normally distributed priors were applied, using the values from ref. 18
(mean = 2.67 × 10
8
substitutions per site per year, s.d. = 2.6 × 10
9
) and from
ref. 32 (mean = 2.74 × 108 substitutions per site per year, s.d. = 2 × 109).
These two rate estimates were chosen as they both use state of the art tip-dating
calibration methods to infer mutation rates, thereby providing inferences that
minimise the effects of rate temporal dependency on late Palaeolithic events33,34.
In particular, the mutat ion-rate estimates reported in refs 18,32 are based on 10 and
66 radiocarbon-dated ancient sequences, respectively. Notably, the calibration dates
for these ancient sequences are distributed across 46,000–4,000 ka and cover both
haplogroups M and N, a scenario that is well–suited for comparison with Australia,
both in terms of temporal coverage and mtDNA diversity. Separate BEAST
phylogenies were inferred for the combined set of Melanesian and Australian
lineages using the mutation rate from ref. 18 (Fig. 1 and Extended Data Fig. 2)
and ref. 32 (Extended Data Fig. 3). A phylogeny based on Australian lineages only
was also inferred using the mutation rate from ref. 18 and used to determine the
palaeodemography of Australia (Extended Data Fig. 7).
All parameters showed sufficient sampling (indicated by effective sample
sizes above 200) after 20,000,000 steps, with the first 10% of samples discarded as
burn-in. Notably, the two different mutations rates produced TMRCA estimates
for the major haplogroups within 1.5kyr of each other (Extended Data Figs 2, 3),
with posterior mutation-rate estimates that were also highly similar (mean
rate = 2.70 × 10
8
(ref. 18), mean rate = 2.74 × 10
8
(ref. 32)), indicating that the
choice of prior distribution for the mutation rate had little effect on our dating.
Multiple correspondence analyses. A useful tool for detecting and analysing
demographic structure in genetic data is principal components analysis (PCA)53.
When working with non-autosomal data, PCA cannot be applied to any
(satisfactory) recoding of sequence data (unless it is manually, that is, subjectively,
sorted into haplogroups). Multiple correspondence analysis (MCA) is an analysis
technique for data exploration and dimension reduction for categorical data. MCA
is a generalization of PCA to categorical variables and can therefore be applied to
raw sequence data. MCA has been independently rediscovere d many times since its
original development, and as such can also be found under titles including ‘optimal
scaling’, ‘dual scaling’ and ‘homogeneity analysis’
54
. MCA was originally developed
for the analysis of survey data, so that responses that were commonly (or rarely)
reported together could be efficiently identified. We apply the same notion but
treat single nucleotide polymorphisms (SNPs) as survey questions, and observed
SNP markers as responses.
We restricted t he MCA to AHP samples and two Australian mtDNA haplotypes
derived from ancient samples whose origin was assumed to be the area in which the
specimen was collected
5,11
(Supplementary Table 3). Unfortunately, we have been
unable to obtain the mtDNA data from a recent Aboriginal genomics study
12
to
use in the MCA analyses, although these samples may have had limited utility for
phylogeographic analysis given the large-scale relocation of Aboriginal Australians
after European arrival. However, we have included the reported sample locations
and mtDNA lineages in geographic plots to examine the consistency with our
results (Extended Data Fig. 4). For the AHP samples, geographic locations were
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Article
reSeArcH
determined for each individual using the relevant genealogies to trace maternal
ancestry as far back as the archival information allowed. Importantly, the broad
distribution of the female ancestors for the AHP samples collected from each of
the three sampling locations (Extended Data Fig. 1) reflects the forced relocation
of Aboriginal Australians from their traditional territories, and highlights the
difficulties associated with obtaining valid phylogeographic information using
only modern samples.
Identical samples were treated separately if they came from different
geographical locations, as these most likely represented more distant family
relationships not captured in the genealogical information. This resulted in 76
unique sequences (Supplementary Table 3). Restricting the analyses to these
samples ensured that the underlying phylogeographic signal was not diluted by
the addition of sequences from modern individuals that are likely to have been
affected by forced- displacement or child-removal policies and typically lack
genealogical information. Independent MCA analyses were run for all samples
combined and for each haplogroup separately. We excluded the M16 lineage from
the M haplogroup tests, because this was a deeply divergent Australian lineage that
clusters among Melanesian samples and thereby most likely represents a pre-Sahul
split (Fig. 1).
We cleaned the aligned sequence data by removing any homogeneous (unin-
formative) sites, and any sites containing missing data. Unlike PCA analyses, we
are not forced to filter out triallelic SNPs and thereby can retain the information
contained within these sites
53
. For M sequences in an alignment, the MCA analysis
will return M 1 principal dimensions of length J Q, where Q is the number of
cleaned SNPs of interest, and J is given by,
=
=
JJ
i
Q
i
1
where J
i
is the number of alleles observed at SNP i, for i = 1, …, Q. These principal
dimensions are analogous to the principal components returned from PCA
analyses, and the dimensions are ordered by the amount of inertia (analogous to
variability in PCA) that they explain. Dimensions with associated eigenvalues less
than 1/Q are discarded as they explain less variation than expected (analogous to
the threshold of 1 for the eigenvalues in PCA)55. The retained coordinates are then
used for the visualization of the relationships between individuals, investigation
of correlation between the dimensions and geographic variables, and clustering
for genetic similarity. We carried out our MCA analysis using the FactoMineR
package56.
Clustering via k medoids. Identifying points in n-dimensional space based on
similarity inferred through Euclidean distance is not a new problem. By far the most
popular clustering algorithm is the k-means clustering algorithm
57
. We used the
closely related k-medoids algorithm instead, which, instead of using a centroid for
each cluster, forces one of the observed data points to be the centre of the cluster.
In doing so, the inter- and intra-cluster distances are more robust to noise and
outliers
58
. We consider an exhaustive range of values for k, and a ‘best’ number of
clusters must be chosen. Unlike the p ossibly subjective ‘elbow method’, used in PCA
through scree plots, we instead calculate
sk
, called the ‘average silhouette’59, for each
value of k. The value of k that maximises
sk
is chosen. However to avoid ‘over-fitting’
the number of clusters, we apply a leave-one-out jack-knife approach to both
identify if influential individuals exist in the data and to obtain some measure of
variability for the values of
sk
. We c ar ried out our clustering methodology using the
cluster package60, in the R statistical programming language61.
Testing for correlation. We tested for geographic correlation through two methods
that seem similar, but are subtly different in their interpretation. First we applied
the Mantel test, which is a test for correlation between two distance matrices62. One
distance matrix contains the pairwise Euclidean distances between individua ls wit h
respect to their geographic location, and the second distance matrix contains the
genetic distances, calculated from the coordinates of the MCA. The null hypothesis
is that there is a perfectly mixed population (that is, pan-mixia), so that rejection
of the null hypothesis indicates some genetic clustering on the landscape. At the
cost of statistical power, we use the Spearman correlation coefficient, as it is unrea-
sonable to assume strictly linear relationships. We us ed 10
5
permutations for each
test. Second, we calculated the correlation between the longitude and latitude of
individuals, and the retained principal dimensions. We perform a standard test
for correlation under Spearman’s ρ (for the same reasons indicated in the Mantel
test). All tests of correlation use the AS 89 algorithm for calculation of P values63.
Although these tests may appear similar, the Mantel test is a relative test of
geographic correlation that simply tests for some clustering with respect to local
geographic location. In essence, the Mantel test investigates if certain combina-
tions of SNP markers are often found within close proximity. With the test of
significant correlations between the principal dimensions and either longitude or
latitude, not only distance, but also direction is important. Hence, the correlation
tests are absolute tests for identifying if combinations of certain SNP markers can
be linked to certain geographical locations on the landscape, with respect to the
entire sampling region. We performed the Mantel test using the vegan package64,
and the standard correlation tests in the R statistical programming language61. The
full list of P values from Spearman’s correlation and the Mantel tests are shown in
Extended Data Table 1.
The Australian archaeological record. Devil’s Lair, southwest Australia. To mo re
precisely constrain the time of arrival of modern humans in southwest Australia,
we analysed a comprehensive multi-dating suite of ages for Devil’s Lair, one of
the earliest archaeological sites in southwestern Australia
65
. The dates comprise
radiocarbon dating (pretreated using acid–base–acid or ABA, and acid–base–
acid stepped combustion, or ABOX–SC, pretreatment), optically stimulated
luminescence, electron-spin resonance (derived using an early uptake model) and
U-series dating. Devil’s Lair (34°9 S, 115°4 E) is a single-chamber cave (floor
area 200 m
2
) formed in the Quaternary dune limestone of the Leeuwin–Naturaliste
Ridge, 5 km from the modern coastline and approximately 250 km south of Perth
(Western Australia). Archaeological investigation over the past four decades has
identified a stratigraphic sequence in the cave floor deposit that consists of 660 cm
of sandy sediments, with > 100 distinct layers, intercalated with flowstone and
other indurated deposits65–67. Archaeological evidence for intermittent human
occupation extends down to layer 30 (around 350 cm depth), with hearths, bone
and stone artefacts found throughout. The lower part of layer 30 represents a fan
of redeposited topsoil that accumulated rapidly after widening of the cave mouth,
and contains the earliest evidence for occupation of the cave. Below layer 30,
six stone artefacts have been identified, including a single specimen each from
layers 32–35, 37 and 38. No artefacts have been found below layer 38.
The age model was created with OxCal v.4.2.4 using a Poisson process
deposition model (P_sequence)
68
with the ‘general outlier’ analysis option
69
of
all ages as reported in ref. 65. The outlier option was used to detect ages that fall
outside the calibration model for the sequence and, if necessary, down-weight
their contribution to the final age estimates. Radiocarbon ages were calibrated
using the SHCal13 calibration dataset70. Taking into account the deposition model
and the actual age measurements, the posterior probability densities quantify the
most probable age distributions. Notably, the lowest artefact in the sequence
is constrained by age estimates obtained using all four dating techniques (but
excluding the ABA radiocarbon (14C) ages, which reached background levels
around 40 ka)65, providing confidence in the calculated age for this level. Using this
approach we derive an age for layer 30 (lower) for cave occupation of 47.1 ± 0.8 ka
and the lowest artefact (layer 38) of 49.5 ± 1.1 ka (Extended Data Fig. 5).
Early colonization of Australia. We extended this approach across Australia, and
examined radiocarbon and optically stimulated luminescence ages associated with
the lowest cultural horizons in early Australian archaeological sites (Extended Data
Fig. 6 and Supplementary Table 4) to estimate the timing of colonization across
the continent. Here we used the Phase model option in OxCal v.4.2.4 (ref. 68) with
general outlier analysis detection (probability = 0.05)69. Notably, the Phase opt ion
is a grouping model which assumes no geographic relationship between samples
(in contrast to the P_sequence used above, which assumes a stratigraphic relation-
ship between dated levels). The model simply assumes that the ages represent a
uniform distribution between a start and end boundary68. Terrestrial samples were
calibrated using the SHCal13 dataset70; marine ages were converted to calendar ages
using the Marine13 calibration dataset19 and corrected for regional Δ R (marine
reservoir age) with reported values for Papua New Guinea (372 ± 64years)71 and
the east Indian Ocean (43 ± 81years)72,73. Using this approach, and incorpo-
rating the age calculated above from Devil’s Lair, we derive an age estimate for
human arrival in Australia (the start of continental occupation) as 48.8 ± 1.3 ka
(Extended Data Fig. 3). Notably, this age estimate includes the luminescence-dated
Northern Territory sites of Malakunanja II and Nauwalabila I (ref. 74–76), which
are statistically indistinguishable from the timing of occupation continent-wide.
Our estimated timing of human arrival is consistent with the minimum age
obtained from the Huon Peninsula
77,78
and the recently reported ages obtained
from Warratyi Rockshelter in the Flinders Ranges24.
Data availability. The datasets generated and analysed during the current study
are available in the European Nucleotide Archive repository, and are accessible
through accession number PRJEB15344. Additional d ata related to this paper may
be requested from the authors.
37. Meyer, M. & Kircher, M. Illumina sequencing library preparation for highly
multiplexed target capture and sequencing. Cold Spring Harb. Protoc. 2010,
pdb.prot5448 (2010).
38. Knapp, M., Stiller, M. & Meyer, M. Generating barcoded libraries for multiplex
high-throughput sequencing. Methods Mol. Biol. 840, 155–170 (2012).
39. Rohland, N., Harney, E., Mallick, S., Nordenfelt, S. & Reich, D. Partial uracil-
DNA-glycosylase treatment for screening of ancient DNA. Phil. Trans. R. Soc.
Lond. B 370, 20130624 (2015).
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Article reSeArcH
59. Rousseeuw, P. J. Silhouettes: a graphical aid to the interpretation and
validation of cluster analysis. J. Comput. Appl. Math. 20, 53–65 (1987).
60. Maechler, M., Rousseeuw, P., Struyf, A., Hubert, M. & Hornik, K. Cluster Analysis
Basics and Extensions. R package version 2.0.4. CRAN (2016).
61. R Development Core Team. R: a language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna, Austria http://www.R-
project.org (2013).
62. Legendre, P. & Fortin, M. J. Comparison of the Mantel test and alternative
approaches for detecting complex multivariate relationships in the spatial
analysis of genetic data. Mol. Ecol. Resour. 10, 831–844 (2010).
63. Best, D. J. & Roberts, D. E. Algorithm AS 89: The upper tail probabilities of
Spearman’s Rho. Appl. Stat. 24, 377–379 (1975).
64. Dixon, P. VEGAN, a package of R functions for community ecology. J. Veg. Sci.
14, 927–930 (2003).
65. Turney, C. S. M. et al. Early human occupation at Devil’s Lair, southwestern
Australia 50,000 years ago. Quat. Res. 55, 3–13 (2001).
66. Dortch, C. E. & Dortch, J. Review of Devil’s Lair artefact classication and
radiocarbon chronology. Aust. Archaeol. 43, 28–32 (1996).
67. Dortch, C. Devil’s Lair, an example of prolonged cave use in South-Western
Australia. World Archaeol. 10, 258–279 (1979).
68. Bronk Ramsey, C. & Lee, S. Recent and planned developments of the program
OxCal. Radiocarbon 55, 720–730 (2013).
69. Bronk Ramsey, C. Dealing with outliers and osets in radiocarbon dating.
Radiocarbon 57, 1023–1045 (2009).
70. Hogg, A. et al. SHCal13 Southern Hemisphere calibration, 0–50,000 years cal
BP. Radiocarbon 55, 1889–1903 (2013).
71. Petchey, F., Phelan, M. & White, J. P. New Δ R values for the southwest Pacic
Ocean. Radiocarbon 46, 1005–1014 (2004).
72. O’Connor, S., Ulm, S., Fallon, S. J., Barham, A. & Loch, I. Pre-bomb marine
reservoir variability in the Kimberley region, Western Australia. Radiocarbon 52,
1158–1165 (2010).
73. Bowman, G. M. Oceanic reservoir correction for marine radiocarbon dates
from northwestern Australia. Aust. Archaeol. 20, 58–67 (1985).
74. Roberts, R. G. et al. The human colonisation of Australia: optical dates of
53,000 and 60,000 years bracket human arrival at Deaf Adder Gorge,
Northern Territory. Quat. Sci. Rev. 13, 575–583 (1994).
75. Roberts, R. G. et al. Single-aliquot and single-grain optical dating conrm
thermoluminescence age estimates at Malakunanja II rock shelter in northern
Australia. Anc. TL 16, 19–24 (1998).
76. Bird, M. I. et al. Radiocarbon dating of organic- and carbonate-carbon in
Genyornis and Dromaius eggshell using stepped combustion and stepped
acidication. Quat. Sci. Rev. 22, 1805–1812 (2003).
77. Groube, L., Chappell, J., Muke, J. & Price, D. A 40,000 year-old human
occupation site at Huon Peninsula, Papua New Guinea. Nature 324,
453–455 (1986).
78. Roberts, R. G. Luminescence dating in archaeology: from origins to optical.
Radiat. Meas. 27, 819–892 (1997).
40. Schubert, M. et al. Characterization of ancient and modern genomes by SNP
detection and phylogenomic and metagenomic analysis using PALEOMIX.
Nat. Protocols 9, 1056–1082 (2014).
41. Lindgreen, S. AdapterRemoval: easy cleaning of next-generation sequencing
reads. BMC Res. Notes 5, 337 (2012).
42. Behar, D. M. et al. A “Copernican” reassessment of the human mitochondrial
DNA tree from its root. Am. J. Hum. Genet. 90, 675–684 (2012).
43. Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows–
Wheeler transform. Bioinformatics 25, 1754–1760 (2009).
44. Jónsson, H., Ginolhac, A., Schubert, M., Johnson, P. L. F. & Orlando, L.
mapDamage2.0: fast approximate Bayesian estimates of ancient DNA damage
parameters. Bioinformatics 29, 1682–1684 (2013).
45. Kearse, M. et al. Geneious basic: an integrated and extendable desktop
software platform for the organization and analysis of sequence data.
Bioinformatics 28, 1647–1649 (2012).
46. van Oven, M. & Kayser, M. Updated comprehensive phylogenetic tree
of global human mitochondrial DNA variation. Hum. Mutat. 30, E386–E394
(2009).
47. Drummond, A. J., Suchard, M. A., Xie, D. & Rambaut, A. Bayesian
phylogenetics with BEAUti and the BEAST 1.7. Mol. Biol. Evol. 29, 1969–1973
(2012).
48. Ingman, M. & Gyllensten, U. mtDB: Human mitochondrial genome database, a
resource for population genetics and medical sciences. Nucleic Acids Res. 34,
D749–D751 (2006).
49. Kent, W. J. BLAT—the BLAST-like alignment tool. Genome Res. 12, 656–664
(2002).
50. Keane, T. M., Creevey, C. J., Pentony, M. M., Naughton, T. J. & Mclnerney, J. O.
Assessment of methods for amino acid matrix selection and their use on
empirical data shows that ad hoc assumptions for choice of matrix are not
justied. BMC Evol. Biol. 6, 29 (2006).
51. Minin, V. N., Bloomquist, E. W. & Suchard, M. A. Smooth skyride through a
rough skyline: Bayesian coalescent-based inference of population dynamics.
Mol. Biol. Evol. 25, 1459–1471 (2008).
52. Drummond, A. J., Ho, S. Y. W., Phillips, M. J. & Rambaut, A. Relaxed
phylogenetics and dating with condence. PLoS Biol. 4, e88 (2006).
53. Patterson, N., Price, A. L. & Reich, D. Population structure and eigenanalysis.
PLoS Genet. 2, e190 (2006).
54. Abdi, H. & Valentin, D. in Encyclopedia of Measurement and Statistics
(ed. Salkind, N.) 651–657 (Thousand Oaks, 2007).
55. Greenacre, M. Correspondence Analysis in Practice. (CRC Press, 2007).
56. Lê, S., Josse, J. & Husson, F. FactoMineR : an R package for multivariate
analysis. J. Stat. Softw. 25, 1–18 (2008).
57. Lloyd, S. P. Least squares quantization in PCM. IEEE Trans. Inf. Theory 28,
129–137 (1982).
58. Kaufman, L. & Rousseeuw, P. J. Clustering by means of medoids. Statistical
Data Analysis Based on the L 1-Norm and Related Methods. First International
Conference 405–416416 (1987).
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Article
reSeArcH
Extended Data Figure 1 | The geographical distribution of the oldest
recorded maternal ancestors for the hair sample donors. Despite
being collected from three different historical locations—Cherbourg
(Queensland), Point Pearce and Koonibba (both South Australia)—the
broad distribution of the maternal ancestors of the hair sample donors
demonstrates the massive displacement experienced by Aboriginal
Australians after European colonization. This pattern illustrates why the
accurate reconstruction of Aboriginal Australian genetic history ultimately
relies upon samples or genealogical records that capture patterns prior to
this displacement. Map data was sourced from the Oak Ridge National
Laboratory Distributed Active Archive Center (https://webmap.ornl.gov/
wcsdown/wcsdown.jsp?dg_id= 10003_1).
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Article reSeArcH
Extended Data Figure 2 | Sahul phylogenetic tree calibrated using
the mitogenome rate from ref. 18. BEAST31 phylogenetic tree of
123 Australian and Melanesian mtDNA lineages, which was calibrated
using the ancient mitogenome rate in ref. 18 to minimize the impacts of
temporal dependency33,34 and improve estimation of the timing of the
founding migrations. The major mitogenome haplogroups are shown at
the base of each clade, and posterior support values are provided for all
nodes.
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Article
reSeArcH
Extended Data Figure 3 | Sahul phylogenetic tree calibrated using mitogenome rate from ref. 32. As for Extended Data Fig. 2, except that rate
calibration used the mitogenome rate from ref. 32.
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Article reSeArcH
Extended Data Figure 4 | Australian phylogeography incorporating
mtDNA lineage information from modern samples reported in ref. 12.
The additional samples from ref. 12 are shown as stars and are distributed
according to their reported locations of collection, all other sample
information is presented in an identical manner to Fig. 2. The mtDNA
haplogroups from ref. 12 are coloured according to the system used in
Fig. 2, with haplogroups not previously shown (that is, R, R12, M42,
P3b and S5) indicated with new colours that are described beneath the
relevant haplogroup map (we have added the two R haplogroups on the
P haplogroup map, as this is the closest sister clade). As in Fig. 2, mtDNA
samples from other studies are shown in yellow, with the samples from
ref. 12 having a yellow dot to indicate this status. Map data was sourced
from the Oak Ridge National Laboratory Distributed Active Archive
Center (https://webmap.ornl.gov/wcsdown/wcsdown.jsp?dg_id=
10003_1).
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Article
reSeArcH
Extended Data Figure 5 | Age-depth model for Devil’s Lair, south-
western Australia. The age-depth model was generated with OxCal
v.4.2.4 (ref. 68) using the Poisson process (outlier) deposition model.
Original ages with 68% uncertainty (prior to modeling) with laboratory
codes shown on left hand side. Prior (light grey) and posterior (dark
grey) probability distributions are plotted. The blue and green envelopes
describe the 68% confidence interval for the sedimentary units below and
above layer 30 (lower) respectively.
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Article reSeArcH
Extended Data Figure 6 | Locations of the early occupation sites used
to estimate the timing of the colonization of Sahul. Sites used for
colonization time estimation are shown as black dots, with white dots
indicating sites that were used to provide independent age controls. Sites
names: 1, Buang Merabak; 2, Matenkupkum; 3, Huon Peninsula; 4, Ivane;
5, Kupona na Dari; 6, Yombon; 7, Nawarla Gabarnmang; 8, Malakunanja
II; 9, Nauwalabila I; 10, Carpenter’s Gap; 11, Riwi; 12, Djadjiling;
13, Ganga Mara; 14, Jansz; 15, Mandu Mandu; 16, Upper Swan; 17,
Devil’s Lair; 18, Allen’s Cave; 19, GRE8; 20, Ngarrabullgan; 21, Menindee;
22, Cooper’s Dune (PACD H1); 23, Lake Mungo; and 24, Warreen Cave.
Additional information for these sites including phase calibrated age
ranges for initial occupation is provided in Supplementary Table 4. Phase
calibrations were performed using OxCal v.4.2.4 (ref. 68) and resulted in
an estimate of the initial colonization of Sahul at 48.8 ± 1.3 ka. The map
was adapted from the figure in ref. 36, originally constructed by J.S.
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Article
reSeArcH
Extended Data Figure 7 | Palaeodemography of Australian
mitogenomes. GMRF Skyride51 analysis of the 98 Australian-only mtDNA
lineages showing the estimated effective maternal population size since
the initial colonization of Sahul around 50 ka (see Methods). Owing to the
lack of available calibration points, the palaeodemographic curve should
be considered relatively approximate. Nonetheless, there is no obvious
indication of a major population bottleneck during the Last Glacial
Maximum (around 21–18 ka). Line, median and grey shading, 95% highest
posterior densities.
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Article reSeArcH
Extended Data Table 1 | Complete Australian phylogeography test results
Spearman’s ρ for correlation with longitude, latitude (with associated P value in parentheses; * P < 0.1; * * P < 0.05; * * * P < 0.01) and the cumulative percentage of inertia (CI%, condence interval)
captured for each principal dimension (rst three rows for each haplogroup), along with Spearman’s ρ for the Mantel test (with associated P value), for haplogroups M (without M16), O, P (including
P4b1 and P5 separately) and S, and the pooled samples (All). Analyses were performed on the 76 samples with reliable provenance (see Methods and Supplementary Table 3).
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
... Notably, the cooling and aridification in Australia during the LGM (cf. De Deckker et al., 2021) led to pronounced geographic contractions of human populations and the abandonment of large parts of the continent (Williams et al., 2013), followed by a deglacial re-expansion of populations (Tobler et al., 2017). ...
... A new chronology constrains the early dispersal of modern humans out of Africa across southern Asia into "Sahul" (northern Australia and New Guinea connected by a land bridge at times of glacially lowered sea level; see Saltré et al., 2016) to ∼ 65-50 ka BP (Clarkson et al., 2017;Tobler et al., 2017). The further settlement comprised a single, rapid (within a few thousand years; Tobler et al., 2017) migration along the eastern and western coasts, with Aboriginal Australians reaching the south of Australia by ∼ 49-45 ka BP. ...
... A new chronology constrains the early dispersal of modern humans out of Africa across southern Asia into "Sahul" (northern Australia and New Guinea connected by a land bridge at times of glacially lowered sea level; see Saltré et al., 2016) to ∼ 65-50 ka BP (Clarkson et al., 2017;Tobler et al., 2017). The further settlement comprised a single, rapid (within a few thousand years; Tobler et al., 2017) migration along the eastern and western coasts, with Aboriginal Australians reaching the south of Australia by ∼ 49-45 ka BP. It is also clear that humans were present in Tasmania by ∼ 39 ka BP (Allen and O'Connell, 2014) and in the arid center of Australia by ∼ 35 ka BP (Smith, 2013). ...
Article
Full-text available
The Leeuwin Current, flowing southward along the western coast of Australia, is an important conduit for the poleward heat transport and inter-ocean water exchange between the tropical and the subantarctic ocean areas. Its past development and its relationship to Southern Ocean change and Australian ecosystem response is, however, largely unknown. Here we reconstruct sea surface and thermocline temperatures and salinities from foraminiferal-based Mg/Ca and stable oxygen isotopes from areas offshore of southwestern and southeastern Australia, reflecting the Leeuwin Current dynamics over the last 60 kyr. Their variability resembles the biomass burning development in Australasia from ∼60–20 ka BP, implying that climate-modulated changes related to the Leeuwin Current most likely affected Australian vegetational and fire regimes. Particularly during ∼60–43 ka BP, the warmest thermocline temperatures point to a strongly developed Leeuwin Current during Antarctic cool periods when the Antarctic Circumpolar Current (ACC) weakened. The pronounced centennial-scale variations in Leeuwin Current strength appear to be in line with the migrations of the Southern Hemisphere frontal system and are captured by prominent changes in the Australian megafauna biomass. We argue that the concerted action of a rapidly changing Leeuwin Current, the ecosystem response in Australia, and human interference since ∼50 BP enhanced the ecological stress on the Australian megafauna until its extinction at ∼43 ka BP. While being weakest during the Last Glacial Maximum (LGM), the deglacial Leeuwin Current intensified at times of poleward migrations of the Subtropical Front (STF). During the Holocene, the thermocline off southern Australia was considerably shallower compared to the short-term glacial and deglacial periods of Leeuwin Current intensification.
... Archaeological sites in the Kimberley are slightly younger and have modelled start ages for onset of occupation that overlap extensively with one another Norman et al. 2022;Veth et al. 2019;Wood et al. 2016) (Table 1). This pattern is replicated in southwest and southeastern Australia, with occupation appearing virtually synchronously across the southern half of the mainland ( Figure 1; Table 1) (Hamm et al. 2016;Tobler et al. 2017;Turney et al. 2001). ...
... Six stone artefacts thought to have washed into the cave were found in sediments between Layers 32 and 38 with a Bayesian modelled age of 51.7-47.3 ka (Tobler et al. 2017), and they could indicate a human presence in the region (Dortch 1979). Four of these six artefacts are made of calcrete, a material available within the cave, while two very small artefacts are made from quartz (Dortch 2004: 96). ...
... The earliest clear evidence for human occupation of the cave is contained in a thick organic layer (Layer 30 lower) that has been interpreted by Balme et al. (1978) as representing the creation of an opening su ciently large to allow people to enter the cave (Balme et al. 1978;Turney et al. 2001). Using OSL and ABOX-SC radiocarbon ages, the Bayesian modelled date for the earliest evidence of occupation of the cave is 48.7-45.5 ka (Tobler et al. 2017;Turney et al. 2001). ...
Article
Mainland Australia was connected to New Guinea and Tasmania at various times throughout the Pleistocene and formed the supercontinent of Sahul. Sahul contains some of the earliest known archaeological evidence for Homo sapiens outside of Africa, with a growing record of early complex social, technological, and artistic life. Here we present an overview of the oldest known sites in Australia along with key evidence pertaining to the dynamic cultures of early Aboriginal peoples. We review debates surrounding the age of first settlement and present evidence for the earliest technology, economy, and symbolism in Australia, emphasizing maritime skills, a large founding population size, novel technology, and adaptation to a wide range of environments.
... When applying quantitative values to these birth/death rates, we find that >1000 and <3000 people would have been needed at 50 ka to result in values of between ~380,000 and 1.15 million Aboriginal people at the time of European invasion. While these initial seeding values may appear high, they conform closely with genomic studies (Tobler et al., 2017) Between the initial peopling of Sahul and European invasion, Williams (2013) provided a range of quantitative population values. At a general level, these suggest that for much of the Pleistocene, populations were likely <50,000 (~1 person/385 km ), values that al lowed exploration of much of the continent, but unlikely to support establishment of per manent regionalism. ...
... It was not until the Early Holocene that values exceed this, with a steady increase from ~100,000 at 6 ka (1 person/70-80 km ) to ~250,000 (1 person/30 km ) at 2.5 ka and culminating at ~1.15 million (1 person/7 km ) at 0.5 ka. In contrast, genomic research suggests long-term continuity in distinct regional populations from 42 ka (Tobler et al., 2017). Some modelling also indicate potentially early populations at sat uration levels could have been up to 6.5 million across Sahul shortly after arrival (Brad shaw et al., 2021). ...
Chapter
The use of radiocarbon data as a proxy for past human demography has become common in many parts of the world with increasingly sophisticated techniques developed in the last decade. Australian archaeologists have been at the forefront of this research. Using this technique, the authors show that at a continent scale Aboriginal population remained low—in the tens of thousands—throughout the Pleistocene, followed by a stepwise growth in the last 10,000 years, and culminating at ~1.15 million in the Late Holocene. The Last Glacial Maximum resulted in significant disruption to populations, with recent evidence suggesting a prolonged recovery from the event, hindered by sea-level change through the terminal Pleistocene. This chapter hypothesises that increasing population during the Early Holocene, along with environmental packing from a reducing landmass, established the conditions contributing to the complex societies observed later in the Holocene and up to the ethnographic period. While radiocarbon approaches to exploring demography have been subject to frequent criticism, virtually all are explicitly addressed in sophisticated applications of the approach, and the authors’ findings continue to be proven robust as more archaeological data becomes available. For instance, the authors’ demonstration that initial seeding population at ~50,000 years ago was ~1000–3000 people and likely involved a deliberate act of exploration has been validated by a plethora of recent studies. The authors suggest a number of temporal and spatial areas that should form the focus of further archaeological research to fill in current knowledge gaps.
... The highly diverse Aboriginal and Torres Strait Islander Peoples of Australia have survived and adapted to climate changes such as sea level rise (SLR) and extreme rainfall variability during the late Pleistocene era, through intimate place-based Indigenous knowledge in practice and while losing traditional land and sea country ownership (Liedloff et al., 2013) (Cross-Chapter Box INDIG in Chapter 18) including during the Late Pleistocene era (Golding and Campbell, 2009;Nunn and Reid, 2016). They belong to the world's oldest living cultures, continually resident in their own ancestral lands, or 'country', for over 65,000 years (Kingsley et al., 2013;Marmion et al., 2014;Nagle et al., 2017;Tobler et al., 2017;Nursey-Bray and Palmer, 2018). The majority of the Australian Indigenous Peoples live in urban areas in southern and eastern Australia, but are the predominant population in remote areas. ...
... Geographic variation in oral microbiota was previously reported among huntergatherers and traditional farmers living in the Philippines [30]. Indigenous Australians have close connections to Country: in many cases, the ancestors of living Indigenous Australians lived in a particular location for at least 45 000 years, even if recent colonial disruptions mean that not all Indigenous Australians live on Country today [49]. During this time, Indigenous Australians' microbiota could have adapted to specific environments or cultural practices such as diets or traditional medicines. ...
Article
Full-text available
Background and objectives Aboriginal Australians and Torres Strait Islanders (hereafter respectfully referred to as Indigenous Australians) experience a high burden of chronic non-communicable diseases (NCDs). Increased NCD risk is linked to oral diseases mediated by the oral microbiota, a microbial community influenced by both vertical transmission and lifestyle factors. As an initial step towards understanding the oral microbiota as a factor in Indigenous health, we present the first investigation of oral microbiota in Indigenous Australian adults. Methodology Dental calculus samples from Indigenous Australians with periodontal disease (PD; n = 13) and non-Indigenous individuals both with (n = 19) and without PD (n = 20) were characterized using 16S ribosomal RNA gene amplicon sequencing. Alpha and beta diversity, differentially abundant microbial taxa and taxa unique to different participant groups were analysed using QIIME2. Results Samples from Indigenous Australians were more phylogenetically diverse (Kruskal–Wallis H = 19.86, P = 8.3 × 10−6), differed significantly in composition from non-Indigenous samples (PERMANOVA pseudo-F = 10.42, P = 0.001) and contained a relatively high proportion of unique taxa not previously reported in the human oral microbiota (e.g. Endomicrobia). These patterns were robust to stratification by PD status. Oral microbiota diversity and composition also differed between Indigenous individuals living in different geographic regions. Conclusions and implications Indigenous Australians may harbour unique oral microbiota shaped by their long relationships with Country (ancestral homelands). Our findings have implications for understanding the origins of oral and systemic NCDs and for the inclusion of Indigenous peoples in microbiota research, highlighting the microbiota as a novel field of enquiry to improve Indigenous health.
... Similar to the Greater Cape Region, this likely reflects the high diversity and prevalence of USO-bearing taxa within the SWAFR. In addition, we suggest that the finding of Botha et al. (2019) that contemporary Khoe-San plant use is representative of continual human use of up to 160,000 years in the Greater Cape is likely also in southwestern Australia for the length of Noongar occupation (i.e. more than 50,000 years (Tobler et al. 2017)), which Hallam (1989) also suggested based on archaeological evidence from the Swan Coastal Plain. Archaeological studies focused on detection of plant residues on Noongar grinding implements may help to shed further light on temporal scale. ...
Article
Full-text available
Aims and background Underground storage organs (USOs) have long featured prominently in human diets. They are reliable year-round resources, especially valuable in seasonal climates. We review a significant but scattered literature and oral recounts of USOs utilised by Noongar people of the Southwest Australian Floristic Region (SWAFR). USOs are important to First Nations cultures in other geophyte-rich regions with Mediterranean climate, with specialist knowledge employed, and productive parts of the landscape targeted for harvest, with likely ecological interactions and consequences. Methods We have gathered Noongar knowledge of USOs in the SWAFR to better understand the ecological role of Noongar-USO relationships that have existed for millennia. Results We estimate that 418 USO taxa across 25 families have Noongar names and/or uses. Additionally, three USO taxa in the SWAFR weed flora are consumed by Noongar people. We found parallels in employment of specific knowledge and targeted ecological disturbance with First Nations’ practice in other geophyte-rich floristic regions. We found that only in 20% of cases could we identify the original source of recorded USO knowledge to an acknowledged Noongar person. Conclusion This review identified that traditional Noongar access to USOs is taxonomically and geographically extensive, employing specific knowledge and technology to target and maintain resource rich locations. However, we also found a general practice of ‘extractive’ documentation of Noongar plant knowledge. We identify negative implications of such practice for Noongar people and SWAFR conservation outcomes and assert ways to avoid this going forward, reviving Noongar agency to care for traditional Country.
... ological evidence dated at 65,000 years BP34 . Molecular analyses suggest that individual language groups have had extremely strong fidelity to their clan estates ('country') since the Late Pleistocene34,35 . During the early twentieth century, the great majority of Aboriginal people on the Arnhem Plateau, including within presentday Kakadu, migrated to surrounding lowland settlements and cattle ranches, buffalo hunting camps, small mines and missions. ...
Article
Full-text available
Colonialism has disrupted Indigenous socioecological systems around the globe, including those supported by intentional landscape burning. Because most disruptions happened centuries ago, our understanding of Indigenous fire management is largely inferential and open to debate. Here, we investigate the ecological consequences of the loss of traditional Aboriginal fire management on fire-exposed savannas on the Arnhem Plateau, northern Australia, using the fire-sensitive conifer Callitris intratropica as a bio-indicator. We contrast Kakadu National Park, where traditional Aboriginal fire management was severely disrupted during the early twentieth century following Aboriginal relocation to surrounding settlements, and an adjacent Aboriginal estate where traditional Aboriginal fire management endures. Since 2006, traditional Aboriginal fire management at this site has been overlaid by a program of broad-scale institutionalized burning in the early dry season, designed to reduce greenhouse emissions. Using remote sensing, field survey, and dendrochronology, we show that on the Aboriginal estate, C. intratropica populations depend on the creation of a shifting patch mosaic of long unburned areas necessary for the recruitment of C. intratropica. However, the imposition of broad-scale fire management is disrupting this population patch dynamic. In Kakadu, there have been extreme declines of C. intratropica associated with widespread fires since the mid twentieth century and consequent proliferation of grass fuels. Fire management in Kakadu since 2007, designed to increase the size and abundance of patches of unburned vegetation, has not been able to reverse the population collapse of C. intratropica. Our study demonstrates that colonial processes including relocation of Indigenous people and institutional fire management can have deleterious consequences that are nearly irreversible because of hysteresis in C. intratropica population dynamics.
Article
Volume I of The Cambridge History of the Pacific Ocean provides a wide-ranging survey of Pacific history to 1800. It focuses on varied concepts of the Pacific environment and its impact on human history, as well as tracing the early exploration and colonization of the Pacific, the evolution of Indigenous maritime cultures after colonization, and the disruptive arrival of Europeans. Bringing together a diversity of subjects and viewpoints, this volume introduces a broad variety of topics, engaging fully with emerging environmental and political conflicts over Pacific Ocean spaces. These essays emphasize the impact of the deep history of interactions on and across the Pacific to the present day.
Preprint
Full-text available
We propose a transmission time investment model for integrating the tenets of human behavioral ecology and cultural evolutionary theory to investigate agency and optimality in the social transmission of lithic technologies. While the cultural transmission process is often overlooked in discussions of optimality, we view it as a critical area for the application of adaptive reasoning to further understand the mechanisms responsible for change in lithic technologies. The proposed model modifies a technological intensification model based on the marginal value theorem (Bettinger et al., 2006; Mohlenhoff and Codding, 2017) to explore how transmissibility may have affected the complexity of socially transmitted lithic production systems during the Pleistocene. This transmission investment model is contrasted with a passive demographic model derived from traditional explanations for changes in lithic technologies. To highlight how optimal considerations of transmissibility may have affected the long-term evolution of lithic technologies, we apply this model to three Pleistocene archaeological case studies investigating increases and decreases in lithic technological complexity. We propose that changes in each of these cases can be understood as the result of time management strategies related to the social transmission process.
Preprint
Full-text available
Pleistocene archaeology in Australia has focussed on the survival and behaviour of Indigenous populations across Sahul during the Last Glacial Maximum (28.6 ± 2.8 ka to 17.7 ± 2.2 ka). A long-standing conceptual model proposes people occupied ecological refugia while abandoning drier regions during extreme climatic conditions, with inferred patterns of subsequent recovery essential for describing the evolution of societies in the Holocene. Radiocarbon-derived population estimates partially support the conceptual model while genetic evidence does not, but how human populations were influenced by the Last Glacial Maximum remains untested. To test the refugia hypothesis, we developed a spatial-demographic model of human movement to project population patterns across Sahul from 40,000 years ago (ka) to the mid-Holocene (5 ka). The model predicts little population change in eastern Sahul or New Guinea during the Last Glacial Maximum. However, extensive movement and potential abandonment in the central-western deserts and the north-northwest coastal regions are predicted during the first half of the Last Glacial Maximum (~ 26–20 ka), with some recovery after 15 ka. The demographic implications to societies appear to have extended beyond the Last Glacial Maximum, with increasing populations not evident until the early Holocene in many regions. Our model describes a complex pattern where large areas of Sahul provided refugia that maintained populations throughout the Last Glacial Maximum, providing a possible explanation for the disparity between archaeological and genomic evidence. There was also a correlation between predicted rates of population change and those derived from radiocarbon dates, supporting the realism of applying dates to infer past demography.
Article
Full-text available
Elucidating the material culture of early people in arid Australia and the nature of their environmental interactions is essential for understanding the adaptability of populations and the potential causes of megafaunal extinctions 50-40 thousand years ago (ka). Humans colonized the continent by 50 ka, but an apparent lack of cultural innovations compared to people in Europe and Africa has been deemed a barrier to early settlement in the extensive arid zone. Here we present evidence from Warratyi rock shelter in the southern interior that shows that humans occupied arid Australia by around 49 ka, 10 thousand years (kyr) earlier than previously reported. The site preserves the only reliably dated, stratified evidence of extinct Australian megafauna, including the giant marsupial Diprotodon optatum, alongside artefacts more than 46 kyr old. We also report on the earliest-known use of ochre in Australia and Southeast Asia (at or before 49-46 ka), gypsum pigment (40-33 ka), bone tools (40-38 ka), hafted tools (38-35 ka), and backed artefacts (30-24 ka), each up to 10 kyr older than any other known occurrence. Thus, our evidence shows that people not only settled in the arid interior within a few millennia of entering the continent, but also developed key technologies much earlier than previously recorded for Australia and Southeast Asia.
Article
Full-text available
Significance Australia is the driest inhabited continent on earth, but humans dispersed rapidly through much of the arid continental interior after their arrival more than 47,000 y ago. The distribution and connectedness of water across the continent, and particularly in its arid core, played a pivotal role in facilitating and focusing early human dispersal throughout the continent. We analyze the distribution and connectedness of modern permanent water across Australia. The modelled least-cost pathways between permanent water sources indicate that the observed rapid occupation of the continental interior was possible along multiple, well-watered routes and likely was driven by the depletion of high-ranked resources in each newly occupied area over time.
Article
It is necessary to deduct an oceanic reservoir correction from radiocarbon dates of marine organisms in order to make them comparable with terrestrial dates. The magnitude of this correction varies latitudinally and is also affected by regional oceanographic factors. To determine the most appropriate oceanic reservoir correction for the northwestern coast of Australia, analyses were carried out to find the ‘apparent’ radiocarbon ages of seven nearshore marine shell specimens collected alive before 1950 AD from this part of the coast. After applying corrections for the date of live collection and the Suess Effect, the error-weighted pooled ages were found to closely match the average Australian oceanic reservoir correction of 450 ± 35 years (Gillespie and Polach 1979) . The study therefore verified the deduction of this correction from marine radiocarbon dates from the northwestern coast of the continent.
Article
OxCal is a widely used software package for the calibration of radiocarbon dates and the statistical analysis of ¹⁴ C and other chronological information. The program aims to make statistical methods easily available to researchers and students working in a range of different disciplines. This paper will look at the recent and planned developments of the package. The recent additions to the statistical methods are primarily aimed at providing more robust models, in particular through model averaging for deposition models and through different multiphase models. The paper will look at how these new models have been implemented and explore the implications for researchers who might benefit from their use. In addition, a new approach to the evaluation of marine reservoir offsets will be presented. As the quantity and complexity of chronological data increase, it is also important to have efficient methods for the visualization of such extensive data sets and methods for the presentation of spatial and geographical data embedded within planned future versions of OxCal will also be discussed.
Article
The population history of Aboriginal Australians remains largely uncharacterized. Here we generate high-coverage genomes for 83 Aboriginal Australians (speakers of Pama-Nyungan languages) and 25 Papuans from the New Guinea Highlands. We find that Papuan and Aboriginal Australian ancestors diversified 25-40 thousand years ago (kya), suggesting pre-Holocene population structure in the ancient continent of Sahul (Australia, New Guinea and Tasmania). However, all of the studied Aboriginal Australians descend from a single founding population that differentiated ∼10-32 kya. We infer a population expansion in northeast Australia during the Holocene epoch (past 10,000 years) associated with limited gene flow from this region to the rest of Australia, consistent with the spread of the Pama-Nyungan languages. We estimate that Aboriginal Australians and Papuans diverged from Eurasians 51-72 kya, following a single out-of-Africa dispersal, and subsequently admixed with archaic populations. Finally, we report evidence of selection in Aboriginal Australians potentially associated with living in the desert. © 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved
Article
The wide availability of precise radiocarbon dates has allowed researchers in a number of disciplines to address chronological questions at a resolution which was not possible 10 or 20 years ago. The use of Bayesian statistics for the analysis of groups of dates is becoming a common way to integrate all of the ¹⁴ C evidence together. However, the models most often used make a number of assumptions that may not always be appropriate. In particular, there is an assumption that all of the ¹⁴ C measurements are correct in their context and that the original ¹⁴ C concentration of the sample is properly represented by the calibration curve. In practice, in any analysis of dates some are usually rejected as obvious outliers. However, there are Bayesian statistical methods which can be used to perform this rejection in a more objective way (Christen 1994b), but these are not often used. This paper discusses the underlying statistics and application of these methods, and extensions of them, as they are implemented in OxCal v 4.1. New methods are presented for the treatment of outliers, where the problems lie principally with the context rather than the ¹⁴ C measurement. There is also a full treatment of outlier analysis for samples that are all of the same age, which takes account of the uncertainty in the calibration curve. All of these Bayesian approaches can be used either for outlier detection and rejection or in a model averaging approach where dates most likely to be outliers are downweighted. Another important subject is the consistent treatment of correlated uncertainties between a set of measurements and the calibration curve. This has already been discussed by Jones and Nicholls (2001) in the case of marine reservoir offsets. In this paper, the use of a similar approach for other kinds of correlated offset (such as overall measurement bias or regional offsets in the calibration curve) is discussed and the implementation of these methods in OxCal v 4.0 is presented.