Australia: The Land Where Time Began

A biography of the Australian continent 

Aboriginal Australia – a Genomic History

The population history of Aboriginal Australians has remained largely uncharacterised. In this study Malaspinas et al. generate high-coverage genomics for 83 Australian Aboriginals, speakers of Pama-Nyungan languages, and 25 Papuans from the New Guinea Highlands. They found the that the ancestors of Australian Aboriginals and Papuans diversified 25-40 thousand years ago, which suggests that the pre-Holocene population structure of Sahul (Australia, New Guinea and Tasmania, at a time of lower sea level when all 3 landmasses were joined together in the single continent of Sahul. However, the results of the study suggest all of the Australian Aboriginals taking part in the study descended from a single population that differentiated about 10-32 ka. Malaspinas et al. infer that in the Holocene past 10,000 years) an expansion of the population in northeast Australia associated with limited gene flow from this region to the rest of Australia, which is consistent with the spread of the Pama-Nyungan languages. It was estimated by Malaspinas et al. that Australian Aboriginals and Papuans diverged from Eurasians 51-72 ka following a single dispersal from Africa, and subsequently admixed with archaic populations. Malaspinas et al. also report that there is evidence that selection among Australian Aboriginal populations occurred associated with desert living.

Sahul was a single continent comprised of Australia and New Guinea for most of the last 100,000 years which was separated from Sunda, the landmass that included continental landmass and western island Southeast Asia, by a series of deep oceanic troughs that have never been exposed by sea level changes. It has been thought that colonisation of Sahul required at least 8-10 sea crossings between islands, which could, potentially constrain the occupation of Australia and New Guinea by earlier hominins (Davidson, 2010). It is suggested by recent assessments that Sahul was settled by 47-55 Ka (Clarkson et al., 2015; O’Connell & Allen, 2015). According to Malaspinas et al. these align with those for the earliest evidence of modern humans in Sunda (Barker et al., 2007).

The suggestion that ancestral Aboriginal Australians and Papuans (‘Australo-Papuans’) migrated out of Africa earlier than those of the Eurasians of the present (Lahr & Foley, 1994) was led to by the distinctiveness of the Australian archaeological and fossil record. Some genetic studies support such multiple dispersals from Africa (Reyes & Centeno, 2014), though others favour a single migration from Africa, comprised of 1 or 2 founding waves that were independent of each other into Asia, the earlier of which contributed to the ancestry of Australians and Papuans (Wollstein et al., 2010; Rasmussen et al., 2011). Also, it has been shown by recent studies that the ancestors of both Australian Aboriginals (Rasmussen et al., 2011) and Papuans (Reich et al., 2010) admixed with Neanderthal and Denisovan archaic humans following their migration form Africa.

During the last glacial maximum between 26.r and 19 Ka Australia underwent increased desertification (Reeves et al., 2013) which affected the number and density of human populations (Hiscock & Wallis, 2005). Unique adaptations, morphological and physiological, have been identified in this context in Australian Aboriginal people that live in the desert areas at present (Birdsell, 1993). Desert groups, in particular, were hypothesised to have been capable of withstanding night temperatures as low as sub-zero C while displaying the increased metabolic rates that were observed in Europeans under the same conditions.

There were more than 250 distinct languages that were spoken by Aboriginal people at the time of European contact, of which 66% belonged to the Pama-Nyungan family, which covered 90% of the Australian mainland (Bowern & Atkinson, 2012). The pronounced similarity among Pama-Nyungan languages, as well as shared sociocultural patterns, have been interpreted as indicating an expansion in the Middle Holocene (Evans & McConvell, 1999), though there have been debates concerning the place of origin of this language family and the effects of its extensive diffusion on its internal phylogenetic structure (Dixon, 2002). In the middle Late Holocene about 4 Ka there were other changes that included the proliferation of backed blades and the dingo was introduced (Hiscock, 2008). Pama-Nyungan languages, dingos and backed blades have been suggested to all reflect the same recent migrations into Australia (Bellwood, 2013). Dingoes were introduced, most likely by way of Southeast Asia, though an external origin for backed blades has been rejected. Evidence has been found by a recent genetic study of Indian gene flow into Australia at approximately the same time as these changes in the Holocene (Pugach et al., 2013), which suggests a possible association, and in historical times there has been substantial mixtures with Asians and Europeans that is well documented (Ellinghaus, 2003).

There have been only 3 whole genome sequences of Australian Aboriginals that have been described to date, 1 of which derived from an historical tuft of hair from a man in the Western Desert (Rasmussen et al., 2011), and 2 others from cell lines with limited provenance information (Prüfer et al., 2014). Malaspinas et al. have reported in this study the first extensive investigation of genome diversity of Australian Aboriginals obtained by analysing the high-coverage genomes of 83 Australian Aboriginals who speak Pama-Nyungan and 25 Papuans from the highlands.

Dataset

Colonisation of Sahul

There has been much debate concerning the origin of Australian Aboriginals, as is the nature of the relationships among Australian Aboriginals, and between Australian Aboriginals and Papuans. Estimates of genomic ancestry proportions and classical multidimensional scaling (MDS) analyses, and was used by ƒ₃ statistics (Patterson et al., 2012), Malaspinas found that Australian Aboriginals and Papuans are closer to each other than to any other world-wide population of the present that were considered in the study (Fig. 2b, c, supplementary information section S05). Malaspinas et al. suggest this is consistent with a common ancestral population, who were the initially colonised Sahul, being shared by Australian Aboriginals and Papuans. Also it is not revealed by outgroup ƒ₃ statistics that there were any significant differences between Papuan populations (People from Papuan Highland groups sampled in this study and the Human Diversity Genome Study (HGDP-Papuans)) in their genetic affinities to Australian Aboriginals (Extended Data fig.2a), which suggests that Papuan populations diverged from one another after or at the time of their Divergence from Australian Aboriginals.

In order to investigate the number of  waves of settlers into Australia, this study alternative models of settlement history were contrasted through a composite likelihood method that compares the joint site frequency spectrum (SFS) that are observed to that predicted under specific demographic models (Excoffier et al., 2013) (Fig. 3, Supplementary Information section S07). HGDP-Papuans were compared with 4 Australian Aboriginal population samples that had low levels of European admixture (Extended Data Fig. 1) from both northeastern (CAI and WPA) and southwestern desert (WON and WCD) Australia. 1- and 2- wave models were compared, where each region in Australia was either colonised independently, or by descendants of a single founding population after it diverged from Papuans. The 1-wave model provides a better fit to the SFS that was observed, which suggests the ancestors of the sampled Aboriginal Australians diverged from a single ancestral population. MDS analysis also supports this conclusion, as well as by ancestry proportion (Cheng et al., 2016) where a cluster is formed by all Aboriginal Australians that is distinct from Papuan populations. Also, outgroup ƒ₃ analysis supports it, where all Aboriginal Australians are of equal distance from Papuans when adjusted for recent admixture. Therefore, the results of this study, based on 83 Pama-Nyungan speakers, fail to support earlier claims of multiple ancestral migrations into Australia which gave rise to the diversity of contemporary Aboriginal Australians (Thorne, 1976).

It is indicated by the SFS analysis that there was a bottleneck in the ancestral Australo-Papuan population about 50 ka (95% confidence intervals (CI) 35-54 ka), which overlaps with evidence from archaeology for the earliest occupation of both Sunda and Sahul 47-55 ka (2-4). It was further inferred by Malaspinas et al. that the ancestors of Pama-Nyungan speakers and Papuans from the Highlands diverged about 37 ka (95% CI 25-40 ka) which agrees closely with results of multiple sequentially Markovian coalescent (MSMC) analyses, a method that estimates rates of cross coalescence between pairs of populations based in the haplotypes of individuals (Shiffels & Durbin, 2014). This result also agrees with previous estimates, e.g., based on SNP array data (Pugach et al., 2013).

Archaic admixture

The number, timing and intensity of archaic gene flow events were characterised by Malaspinas et al. by the use of 3 complementary approaches: SFS-based, a goodness of fit analysis that combined D-statistics, and a method that infers putatively derived archaic ‘haplotypes’. An excess of putative Denisovan introgressed sites, as well as a substantially more putative Denisovan-derived haplotypes (PDHs) than other non-Africans is shown by Australian Aboriginal and Papuan genomes. There is considerable variation of the number and total length of those putative haplotypes across samples. The estimated number of PDHs correlates, however, almost perfectly (r²=0.96) with the proportion that has been estimated of Australo-Papuan ancestry in each individual. No significant difference in distribution of the number of PDHs or the average length of PDHs between Australian Aboriginals and Papuans was found (Mann-Whitney U-test, P>0.05). The genetic differentiation between WCD and Papuans was, moreover, also similar for both autosomal SNPs and PDHs, with F_ST values around 0.12. Evidence is provided for admixture with Denisovans that predates the split in the population between Aboriginal Australians and Papuans (Qin & Stoneking, 2015; Skoglund & Jakobsson, 2011), when these analyses are taken together, and widespread recent admixture of Eurasian in Australian Aboriginals. The SFS-based approach, results in an admixture estimate of about 4.0% (95% CI 3.3-5.0%), similar to that obtained by D-statistics (about 5%), by constraining Denisovan admixture as having occurred prior to divergence between Australian Aboriginals and Papuans. It is further suggested by the SFS analyses that the Denisovan/Australo-Papuan took place about 44 ka (95% CI 31-50 ka), a date that overlaps with an estimate from a more recent study (Sankararaman, Mallick, Patterson & Reich, 2016).

Evidence for a primary Neanderthal admixture event (about 2.3%, 95% CI 1.1-3.5%) that took place in the ancestral population of all non-Africans about 60 ka (95% CI 55-84 ka). A relative date can be obtained, though it has not been passible to estimate absolute dates of archaic admixture from the length of PDHs and haplotypes that are putative Neanderthal-derived in the samples, The study found that the average lengths for PDH is 33.8 kb and for the 37.4 kb for the PDH . These differ significantly from each other (P= 3.65 x 10⁶ by the use of the conservative sign test), suggesting that the time since the admixture of Neanderthal was about 11% greater than the time since Denisovan admixture, which is roughly in line with estimates based on SFS that were obtained in the study for the Denisovan pulse (31-50 ka), versus the primary pulse of Neanderthal admixture (55-84 ky). It is also indicated by the SFS analysis that the main Neanderthal pulse was followed by a 1.1% (95% CI 0.2-2.7%) pulse of the gene flow of Neanderthal into the ancestors of Eurasians. Finally, additional models that involved complex structure among the archaic populations were explored by SFS- and haplotype-based approaches. The study found evidence that was suggestive that the archaic contribution could be more complex than the model that involved the discrete Denisovan and Neanderthal pulses of admixture (Rasmussen et al., 2011; Reich et al., 2010).

Out of Africa

In order to investigate the relationship of the ancestors of the Australo-Papuans with other populations around the world, D statistics (22) were computed of the form:

·        H1 = Australian Aboriginal

·        H2 = Eurasian

·        H3 = African

And

·        H1 = Australian Aboriginal

·        H2 = Eurasian

·        H3 = Ust’-Ishim.

Several of these were significantly positive, which suggests that Africans and Ust’-Ishim, the remains of a human dating to 45 ka from Asia (27), are both closer to Eurasians than to Australian Aboriginals. According to Malaspinas et al. these findings agree with a model of Eurasians and Australo-Papuan ancestors dispersing from Africa in 2 independent waves. When correcting for a moderate amount of Denisovan admixture, the Australian Aboriginals and Eurasians become equally close to Ust’-Ishim, as would be expected in a single Out of Africa scenario. Similarly, the D-statistics for:

·        H1 = Australian Aboriginals

·        H2 = Eurasians

·        H3 = Africans

became such smaller after correcting for the admixture from Denisovans. Also, it is indicated by a goodness-of-fit-approach that combined D-statistics across worldwide populations indicates stronger support for 2 waves OoA, but when Denisovan admixture is taken into account, a 1-wave scenario fits the D-statistics that is observed equally well.

Malaspinas et al. used the observed SFS in a composite likelihood framework that was model-based (Excoffier et al., 2013), in order to investigate the timing and number of OoA events that gave rise to Australo-Papuans and Eurasians of the present further. Evidence was found for 2 waves of OoA, when only modern genomes were considered, with a dispersal of Australo-Papuans about 14,000 years before Eurasians. When Neanderthal and Denisovan introgression into modern humans, however (Ellinghaus, 2003; Prüfer et al., 2014), the SFS analysis supports a single origin for the OoA populations that is marked by a bottleneck about 72 ka (95% CI 60-104 ka). The ancestors of Australo=Papuans and Eurasians share an admixture pulse of 2.3% (95% CI 1.1-3.5%) from Neanderthals. Also, modern humans have a linkage disequilibrium decay rate and a number of deleterious homozygous mutations that had been predicted (recessive genetic load), that correlate with distance from Africa, which is again consistent with a single origin in Africa.

SFS analysis best supports the model, and also suggests an early divergence of Australo-Papuans from the ancestors of all non-Africans, with agrees with 2 waves of colonisation across Asia (Rasmussen et al., 2011; Pugach et al., 2013). Australo-Papuans began to diverge from Eurasians about 58 Ka (95% CI 51-72 Ka), whereas the divergence of Europeans from East Asians occurred about 42 Ka (95% CI 29-55 Ka), which agrees with previous estimates (Wollstein et al., 2010; Pugach et al., 2013; Gutenkunst, Hernandez, Williamson & Bustamante, 2009). Evidence was found by Malaspinas et al. for high levels of gene flow between the ancestors of Eurasians and Australo-Papuans, which suggests that following the fragmentation of the OoA population 57-58 Ka, the groups remained in close geographical proximity for some time prior to the ancestors of Australo-Papuans dispersing to the east. They also found evidence for gene flow between sub-Saharan Africans and Western Eurasians after about 42 Ka, which agrees with previous findings (Gutenkunst, Hernandez, Williamson & Bustamante, 2009).

It is suggested by MSMC         analysis that the Yoruba/Australo/Papuans and the Yoruba/Eurasians cross-coalescence rates are distinct, which implies that the gene trees across the genome of the Yoruba and Eurasians have, on average, more recent common ancestors. It was shown by simulations by Malaspinas et al. that typical amounts of archaic mixture (<20%) cannot explain these differences. Also, the differences that are expected in phasing quality among genomes is insufficient to fully explain this pattern. Malaspinas et al. found that the cross coalescence curves overlap when comparing Australo-Papuan and Eurasians with San, while a similar separation in cross coalescence curves is obtained when comparing Australo-Papuan and Eurasians with Dinka. They also found that effective changes in population size that were inferred over time of Australian Aboriginals, Papuans, and East Asians are very similar until about 50 Ka, which includes a deep bottleneck about 60 Ka. These MSMC results, when taken together, are consistent with a split of both the Australo-Papuans and Eurasians from a single African population, combined with gene flow between ancestors of Yoruba or Dinka, though not San, and the ancestors of Eurasians, that is not found among Australo-Papuans. According to Malaspinas et al. these results are in line, qualitatively, with the SFS-based analyses. The results of this study indicate that any such event left little trace in the genomes of modern Australo-Papuans, which is in line with conclusions from related work that appeared alongside this study (Mallick et al., 2016; Pagani et al., 2016), though the results of this study do no exclude the possibility of an earlier OoA expansion.

Australian Aboriginal – Genetic structure

Diversity of uniparental haplogroups in this dataset is consistent with previous studies using mitochondrial DNA (mtDNA) and variation of Y chromosomes in Australia and Oceania (Bergström et al., 2016), which included the presence of typically European, Southeast Asian and East Asian lineages (Hudjashov et al., 2007). Important insights into the social structure of Australian Aboriginal societies are provided by the combined results. Australian Aboriginals exhibit greater variation between groups for mtDNA, 16.8%, than for the Y chromosome, 11.3%, which contrasts with the pattern for most human populations (Lippold et al., 2014). Higher levels of male-mediated migration than female-mediated migration are suggested by this result, and may reflect the complex marriage and post-marital residence patterns among Pama-Nyungan Aboriginal groups (Radcliffe-Brown, 1930). The inferred European ancestry for the Y chromosome is much greater than that for mtDNA, 31.8% versus 2.4%, which reflects male biased gene flow into Australian Aboriginal groups during the colonial era.

It was found by this study that on an autosomal level the genetic relationships within Australia reflect the geography, and there is a significant correlation (R_GEN,GEO = 0.77, P<0.0005), between the first 2 dimensions of an MDS analysis on genomes that are masked and the geographical location. Genetically, populations from the continental centre are at intermediate positions. A population division between the groups from the northeast and the southwest is suggested by a similar result that is observed with an F_ST-based tree for the masked genomic data as well as in analyses of genetic affinity based on ƒ₃ statistics. SFS analyses further support this structure show the populations from the southwestern desert and northeastern regions diverges as early as about 31 Ka (95% CI 10-32 Ka), which was followed by limited gene flow (estimated scaled migration rate (2Nm) about 0.01, 95 % CI 0.00-1.25). A model, in which the interior of the Australian continent acted as a barrier to migration, was supported by an analysis of the major routes of gene flow within the continent. According to Malaspinas et al., they inferred that gene flow occurred preferentially along the coasts of Australia, based on a model that was inspired by the principles of electrical engineering where current flowing through the Australian continent that represents gene flow, and by the use of observed values of F_ST as a proxy for resistance. These findings are consistent with an expansion model followed by fragmentation of population when barriers formed as a result of the extreme aridity in the interior of Australia to movements of population during the LGM (Veth, 1989).

MSMC on autosomal data and mtDNA Bayesian skyline plots (Drummond et al., 2005) (BSP) was used to estimate changes in effective population size within Australia. Evidence of an expansion of population beginning about 10 ka in the northeast was provided by MSMC analyses, while a bottleneck in the populations of the southwestern desert took place over the past about 10,000 years. Malaspinas et al. suggest this is consistent with archaeological evidence for an expansion of population that was associated with significant changes in the socio-economic and subsistence strategies in Holocene in Australia (Lourandos & David, 2002).

Malaspinas et al. considered it almost certain that the European admixture had not occurred any earlier than the 18^th century, though earlier East Asian and/or New Guinean gene flow into Australia could have occurred. They used 3 different approaches to characterise the mode and tempo of the gene flow into Australia. They used Bayesian computation (ABC) for the mean and variance in the proportion of European, East Asian and Papuan admixture among Australian Aboriginal individuals to those that were computed from simulated datasets under various gene flow models. They estimated that admixture from European and East Asians to have occurred approximately 10 generations ago, which is consistent with historical and ethnographic records. Consistent with this, a local ancestry approach suggests that European and East Asian admixture is more recent than Papuan admixture. Also, it is indicated by both ABC and SFS analyses that the model of best fit for the Australian Aboriginal-Papuan data is one in which there is continuous, though modest, that is mostly unidirectional from Papuans to Australian Aboriginals, and is restricted geographically to northeast Australian Aboriginals (2Nm=0.41, 95% CI 0.00-20.35).

In order to further investigate gene flow from New Guinea, analyses were conducted on the Papuan ancestry tracts that were obtained from the analysis of local ancestry. It was inferred that local ancestry was the result of admixture between 4 components: European, East Asian, Papuan and Australian Aboriginals. It was shown by the Papuan tract length distribution that there was a clear geographic pattern; a significant correlation of Papuan tract length variance with distance from WDC to other Australian Aboriginal groups (r = 0.64, P<0.0001. It was suggested by the prevalence of short ancestry tracts of Papuan origin, compared to longer tracts of East Asian and European origin, that a large fraction of gene flow from Papua is much older than that from Europe and Asia, which is consistent with the ABC analysis. Possible South Asian (Indian-related) gene flow into Australian Aboriginals, as reported recently (Pugach et al., 2013), was also investigated. No evidence was found, however, of a component that could be uniquely assigned to Indian populations in the Australian Aboriginal gene pool by the use of either admixture analyses of ƒ₃ and D-statistics, even when including the original Australian Aboriginal genotype data from Arnhem Land. Malaspinas et al. suggest the nature of comparative datasets may account for this discrepancy.

Pama Nyungan languages and genetic structure

A Bayesian phylogenetic tree was inferred for 28 different Pama-Nyungan languages that were represented in this sample (Bowern & Atkinson, 2012) in order to investigate if linguistic relationships reflect genetic relationships among Australian Aboriginal populations. Several partitions that are well supported are shared by the resulting linguistic and F_ST-based genetic trees. E.g., it is indicated by both trees that the northeastern (CAI and WPA) and southwestern groups (ENY, NGA, WCD and WON) form 2 distinct clusters, while PIL, BDV and RIV are intermediate. A distance between language pairs, that were computed from the language-based tree, is correlated significantly with geographic distance (r_{GEO, LAN} = 0.83, Mantel test 2-tail P on 9,999 permutations = 0.0001). It is suggested by these that among Pama-Nyungan languages in Australia differentiations follow geographic patterns, as is observed in other language families from other parts of the world (36). Also, it was found that a correlation between linguistics and genetics (r_{GEO, LAN} = 0.43, Mantel test P < 0.0005), remains significant when geography is controlled for (r_{GEO, LAN} = 0.26, partial Mantel test P <0.0005). After populations (genetic) lost contact with one another, this is consistent with language differentiation. It is all the more notable that the correlation between the linguistic and genetic trees, given the differences in time scales: it is generally accepted that the Pama-Nyungan family diversified within the last 6,000 years (Evans & Jones, 1997), while the genetic estimates are 2 to 5 times that age. Therefore, the linguistic tree cannot simply reflect the initial dispersal of populations; rather, it reflects a genetic structure that has a complex history, with initial differentiation 10-32 Ka, population expansions that were localised (northeast) and bottlenecks (southwest) about 10 Ka, and gene flow from the northeast to the southwest that was subsequently limited. Malaspinas et al. suggest the latter may be the genetic signature that tracks the divergence of the Pama-Nyungan.

Selection in Australian Aboriginals

Malaspinas et al. used 2 different methods that were based on identification of SNPs with high allele frequencies differences between Australian Aboriginals and other groups, similar to the population branch statistics (Yi et al., 2010) (PBS), to identify selection signatures that were specific to Australian Aboriginals.

1)    the Australian Aboriginal genomes were scanned for loci that had changes in the frequency that were unusually large since divergence from Papuans, while taking into account admixture with Europeans and Asians that were recent (‘global scan’).

2)    Genomic regions that showed high differentiation that was associated with different ecological regions within Australia (‘local scan’). Among the peaks that were top ranked genes were found that were associated with the thyroid system (NETO1, 7^th peak in the global scan, and the (KCNJ2, the first peak in the global scan) and urate levels in the serum (8^th peak in the global scan).

Thyroid hormone levels are associated with adaptations to desert cold specific to Australian Aboriginals (Qi et al., 2014) and serum urate levels that are associated with dehydration (Tin et al., 2011). Therefore these genes are candidates for potential adaptation to life in the desert. According to Malaspinas et al., further studies are required to associate putative selected genetic variants with specific phenotypic adaptations in Australian Aboriginals. 

Discussion

Australia has one of the longest continuous occupation histories by modern humans outside Africa, which raises questions of origins, relatedness to other populations, differentiation and adaptation. Some answers have been provided by these large-scale genomic data and analyses, though other questions have also been raised. This study found that Australian Aboriginals share genomic signatures of an Out of Africa dispersal with Eurasians, a common African ancestor, a bottleneck and a primary pulse of Neanderthal admixture. The population history of Australian Aboriginals diverged, however, from that of other Eurasians shortly after the OoA event, and included private admixture with another archaic hominin.

The genetic-based estimates of timeline of this study are relative, and 2 rescaling parameters were relied upon to obtain absolute dates: the mutation rates and generation time of humans (assumed to be 1.25 x 10^-8 per generation per site and 29 years, respectively, based on recent estimates (41,42). The current values can be the starting point for research in the future and should be contextualised, though the absolute estimates reported would need to be revised if the parameters were to change.

A relatively old divergence between the ancestors of Pama-Nyungan speakers and Highlands Papuans, that are only about 10% younger than the European-East Asian split time, were found by the study. This corresponds, with the assumed rescaling parameters, this corresponds to about 37 Ka  (95% CI 25-40 Ka), which implies that the divergence between the Papuans that were sampled and the Australian Aboriginals predates the disappearance of the land bridge between New Guinea and Australia about 7-14.5 ka, which therefore suggests ancient genetic structure in Sahul. Malaspinas et al. suggest such a structure may be related to palaeoenvironmental changes that occurred in the lead up to the LGM. It has been shown by sedimentary studies that the large Lake Carpentaria, with dimensions of 500 x 250 km) formed about 40 ka when levels fell below the 53m deep Arafura Sill (43). The flooding of the Carpentaria basin and its increasing salinity (Holt, 2005) may have therefore promoted population isolation, though Australia and New Guinea remained connected until the Early Holocene.

The Australian Aboriginal groups are implied by the results of this study to have descended from the ancestral population of the first settlers that colonised Australia (Rasmussen et al., 2011, Heupink et al., 2016). The results also indicate that the population which diverged from Papuans about 37 Ka was ancestral to all Australian Aboriginal groups that were sampled in this study; yet, it is shown by archaeological evidence that humans were widespread throughout Australia by 40-45 ka. Malaspinas et al. suggest 3 non-exclusive scenarios could account for this observation:

1)    The ancestral population of Australian Aboriginals was widespread prior to the divergence from Papuans, maintaining gene flow across the continent;

2)    It was structured deeply, with only 1 group surviving to give rise to modern Aboriginal Australians; and

3)    Other groups survived, but their descendants are not represented in the sample used in this study.

Additional genomes, especially from Tasmania and non-Pama-Nyungan regions of the Northern Territory and Kimberly, and also ancient genomes that predate European contact in Australia and other expansions across Southeast Asia (Bellwood, 2013), may help resolve these questions in the future.

See Almost all living people outside of Africa trace back to a single migration more than 50,000 years ago

Sources & Further reading

1.     Malaspinas, A.-S., M. C. Westaway, C. Muller, V. C. Sousa, O. Lao, I. Alves, A. Bergström, G. Athanasiadis, J. Y. Cheng, J. E. Crawford, T. H. Heupink, E. Macholdt, S. Peischl, S. Rasmussen, S. Schiffels, S. Subramanian, J. L. Wright, A. Albrechtsen, C. Barbieri, I. Dupanloup, A. Eriksson, A. Margaryan, I. Moltke, I. Pugach, T. S. Korneliussen, I. P. Levkivskyi, J. V. Moreno-Mayar, S. Ni, F. Racimo, M. Sikora, Y. Xue, F. A. Aghakhanian, N. Brucato, S. Brunak, P. F. Campos, W. Clark, S. Ellingvåg, G. Fourmile, P. Gerbault, D. Injie, G. Koki, M. Leavesley, B. Logan, A. Lynch, E. A. Matisoo-Smith, P. J. McAllister, A. J. Mentzer, M. Metspalu, A. B. Migliano, L. Murgha, M. E. Phipps, W. Pomat, D. Reynolds, F.-X. Ricaut, P. Siba, M. G. Thomas, T. Wales, C. M. r. Wall, S. J. Oppenheimer, C. Tyler-Smith, R. Durbin, J. Dortch, A. Manica, M. H. Schierup, R. A. Foley, M. M. Lahr, C. Bowern, J. D. Wall, T. Mailund, M. Stoneking, R. Nielsen, M. S. Sandhu, L. Excoffier, D. M. Lambert and E. Willerslev (2016). "A genomic history of Aboriginal Australia." Nature advance online publication.