Pleistocene Mitochondrial Genomes Suggest a Single Major Dispersal of Non-Africans and a Late Glacial Population Turnover in Europe
Correspondence information about the author Cosimo Posth
Email the author Cosimo PosthCorrespondence information about the author Johannes KrauseEmail the author Johannes Krause
Article Info
Permitted
For non-commercial purposes:
- Read, print & download
- Text & data mine
- Translate the article
Not Permitted
- Reuse portions or extracts from the article in other works
- Redistribute or republish the final article
- Sell or re-use for commercial purposes
Elsevier's open access license policy
Figure 1
Late Pleistocene and Early Holocene Archeological Sites and Hunter-Gatherer mtDNA Haplogroups
(A) Pre-LGM dispersal of non-African populations, carrying both M and N lineages (hgs R, U, U5, and U2′3′4′7′8′9 belong to the N clade, distinct from the M clade).
(B) Post-LGM re-expansion in Europe while ice sheets retracted.
(C) Late Glacial shift in mtDNA hg frequency.
(D) Holocene hunter-gatherer mtDNA, mainly belonging to hg U5.
See also Table S1x
Download
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Table S1, Table S2xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Table S2, Table S4xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Table S4, and the Supplemental Experimental ProceduresxDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Supplemental Experimental Procedures.
Figure 2
Maximum Parsimony Tree of Present-Day Human and 55 Pre-Neolithic mtDNA Genomes
Pre-LGM samples are shown in blue, Post-LGM in green, Late Glacial in magenta, Holocene hunter-gatherers in red, and present-day individuals in black print. Average values of 14C dates are reported next to each specimen when available. Red arrows indicate divergence times of M and N clades. Hg M is almost absent in present-day individuals with European ancestry. Oase1 represents a pre-N lineage. The tree is rooted with one Neanderthal and 16 deeply divergent African mtDNAs (not shown). See also Figure S1xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Figure S1 and the Supplemental Experimental ProceduresxDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Supplemental Experimental Procedures.
Figure 3
Late Pleistocene and Early Holocene Climatic Fluctuations and European Hunter-Gatherer Demography
On the left is the NGRIP δ18O climate record, and on the right is an illustration of the best-supported demographic model (2b in Figure S2xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Figure S2). Each colored point gives the mtDNA hg of the 55 dated pre-Neolithic individuals used in the coalescent modeling analysis. West-East site locations for each sample are approximated. See also Figure S2xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Figure S2, Table S4xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Table S4, and the Supplemental Experimental ProceduresxDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Supplemental Experimental Procedures.
Highlights
- •
Newly generated pre-Neolithic European mtDNA genomes triple the number available
- •
Clade M found for the first time in Europe, prior to the Last Glacial Maximum bottleneck
- •
Rapid single dispersal of all non-Africans less than 55,000 years ago
- •
Previously unknown major population shift in Europe at the end of the Pleistocene
Summary
How modern humans dispersed into Eurasia and Australasia, including the number of separate expansions and their timings, is highly debated [1xRevising the human mutation rate: implications for understanding human evolution. Scally, A. and Durbin, R. Nat. Rev. Genet. 2012;
13: 745–753
Crossref | PubMed | Scopus (13)See all References, 2xRethinking the dispersal of Homo sapiens out of Africa. Groucutt, H.S., Petraglia, M.D., Bailey, G., Scerri, E.M., Parton, A., Clark-Balzan, L., Jennings, R.P., Lewis, L., Blinkhorn, J., Drake, N.A. et al. Evol. Anthropol. 2015;
24: 149–164
Crossref | PubMed | Scopus (30)See all References]. Two categories of models are proposed for the dispersal of non-Africans: (1) single dispersal, i.e., a single major diffusion of modern humans across Eurasia and Australasia [3xGenetic and archaeological perspectives on the initial modern human colonization of southern Asia. Mellars, P., Gori, K.C., Carr, M., Soares, P.A., and Richards, M.B. Proc. Natl. Acad. Sci. USA. 2013;
110: 10699–10704
Crossref | PubMed | Scopus (79)See all References, 4xSingle, rapid coastal settlement of Asia revealed by analysis of complete mitochondrial genomes. Macaulay, V., Hill, C., Achilli, A., Rengo, C., Clarke, D., Meehan, W., Blackburn, J., Semino, O., Scozzari, R., Cruciani, F. et al. Science. 2005;
308: 1034–1036
Crossref | PubMed | Scopus (425)See all References, 5xA single southern exit of modern humans from Africa: before or after Toba?. Oppenheimer, S. Quat. Int. 2012;
258: 88–99
Crossref | Scopus (30)See all References]; and (2) multiple dispersal, i.e., additional earlier population expansions that may have contributed to the genetic diversity of some present-day humans outside of Africa [6xTowards a theory of modern human origins: geography, demography, and diversity in recent human evolution. Lahr, M.M. and Foley, R.A. Am. J. Phys. Anthropol. 1998;
: 137–176
Crossref | PubMedSee all References, 7xMajor genomic mitochondrial lineages delineate early human expansions. Maca-Meyer, N., González, A.M., Larruga, J.M., Flores, C., and Cabrera, V.M. BMC Genet. 2001;
2: 13
Crossref | PubMed | Scopus (108)See all References, 8xGenomic and cranial phenotype data support multiple modern human dispersals from Africa and a southern route into Asia. Reyes-Centeno, H., Ghirotto, S., Détroit, F., Grimaud-Hervé, D., Barbujani, G., and Harvati, K. Proc. Natl. Acad. Sci. USA. 2014;
111: 7248–7253
Crossref | PubMed | Scopus (46)See all References, 9xThe southern route “out of Africa”: evidence for an early expansion of modern humans into Arabia. Armitage, S.J., Jasim, S.A., Marks, A.E., Parker, A.G., Usik, V.I., and Uerpmann, H.P. Science. 2011;
331: 453–456
Crossref | PubMed | Scopus (206)See all References]. Many variants of these models focus largely on Asia and Australasia, neglecting human dispersal into Europe, thus explaining only a subset of the entire colonization process outside of Africa [3xGenetic and archaeological perspectives on the initial modern human colonization of southern Asia. Mellars, P., Gori, K.C., Carr, M., Soares, P.A., and Richards, M.B. Proc. Natl. Acad. Sci. USA. 2013;
110: 10699–10704
Crossref | PubMed | Scopus (79)See all References, 4xSingle, rapid coastal settlement of Asia revealed by analysis of complete mitochondrial genomes. Macaulay, V., Hill, C., Achilli, A., Rengo, C., Clarke, D., Meehan, W., Blackburn, J., Semino, O., Scozzari, R., Cruciani, F. et al. Science. 2005;
308: 1034–1036
Crossref | PubMed | Scopus (425)See all References, 5xA single southern exit of modern humans from Africa: before or after Toba?. Oppenheimer, S. Quat. Int. 2012;
258: 88–99
Crossref | Scopus (30)See all References, 8xGenomic and cranial phenotype data support multiple modern human dispersals from Africa and a southern route into Asia. Reyes-Centeno, H., Ghirotto, S., Détroit, F., Grimaud-Hervé, D., Barbujani, G., and Harvati, K. Proc. Natl. Acad. Sci. USA. 2014;
111: 7248–7253
Crossref | PubMed | Scopus (46)See all References, 9xThe southern route “out of Africa”: evidence for an early expansion of modern humans into Arabia. Armitage, S.J., Jasim, S.A., Marks, A.E., Parker, A.G., Usik, V.I., and Uerpmann, H.P. Science. 2011;
331: 453–456
Crossref | PubMed | Scopus (206)See all References]. The genetic diversity of the first modern humans who spread into Europe during the Late Pleistocene and the impact of subsequent climatic events on their demography are largely unknown. Here we analyze 55 complete human mitochondrial genomes (mtDNAs) of hunter-gatherers spanning ∼35,000 years of European prehistory. We unexpectedly find mtDNA lineage M in individuals prior to the Last Glacial Maximum (LGM). This lineage is absent in contemporary Europeans, although it is found at high frequency in modern Asians, Australasians, and Native Americans. Dating the most recent common ancestor of each of the modern non-African mtDNA clades reveals their single, late, and rapid dispersal less than 55,000 years ago. Demographic modeling not only indicates an LGM genetic bottleneck, but also provides surprising evidence of a major population turnover in Europe around 14,500 years ago during the Late Glacial, a period of climatic instability at the end of the Pleistocene.
Results and Discussion
Genetic studies of human mitochondrial DNA (mtDNA) show that all present-day non-Africans belong to two basal mtDNA haplogroups (hgs), M and N [10xUpdated comprehensive phylogenetic tree of global human mitochondrial DNA variation. van Oven, M. and Kayser, M. Hum. Mutat. 2009;
30: E386–E394
Crossref | PubMed | Scopus (716)See all References10]. The time to the most recent common ancestor (TMRCA) of each of these two clades has been estimated independently at around 50,000 years ago (50 ka) (95% confidence interval [CI], 53–46 ka) and 59 ka (95% CI, 64–54 ka), respectively [11xA “Copernican” reassessment of the human mitochondrial DNA tree from its root. Behar, D.M., van Oven, M., Rosset, S., Metspalu, M., Loogväli, E.L., Silva, N.M., Kivisild, T., Torroni, A., and Villems, R. Am. J. Hum. Genet. 2012;
90: 675–684
Abstract | Full Text | Full Text PDF | PubMed | Scopus (138)See all References11]. However, whereas present-day Asians, Australasians, and Native Americans carry both M and N mtDNA hgs, modern individuals with European ancestry lack almost completely lineages of the M clade [12xMaternal ancestry and population history from whole mitochondrial genomes. Kivisild, T. Investig. Genet. 2015;
6: 3
Crossref | PubMed | Scopus (7)See all References12]. The different spatial distributions and TMRCA estimates of these two ancestral clades have been interpreted as evidence of an early spread of modern humans carrying hg M into Asia, perhaps via a southern route, followed by a later non-African diffusion of the N clade, perhaps via a northern route [7xMajor genomic mitochondrial lineages delineate early human expansions. Maca-Meyer, N., González, A.M., Larruga, J.M., Flores, C., and Cabrera, V.M. BMC Genet. 2001;
2: 13
Crossref | PubMed | Scopus (108)See all References7]. However, an alternative model proposes a rapid and single dispersal across Eurasia, with Asia being reached first, whereas Western Eurasia would have been settled only after a hiatus, during which hg M was lost [4xSingle, rapid coastal settlement of Asia revealed by analysis of complete mitochondrial genomes. Macaulay, V., Hill, C., Achilli, A., Rengo, C., Clarke, D., Meehan, W., Blackburn, J., Semino, O., Scozzari, R., Cruciani, F. et al. Science. 2005;
308: 1034–1036
Crossref | PubMed | Scopus (425)See all References4].
Little is known about the genetic makeup of the first European hunter-gatherers, who likely arrived ∼45 ka [13xEarly dispersal of modern humans in Europe and implications for Neanderthal behaviour. Benazzi, S., Douka, K., Fornai, C., Bauer, C.C., Kullmer, O., Svoboda, J., Pap, I., Mallegni, F., Bayle, P., Coquerelle, M. et al. Nature. 2011;
479: 525–528
Crossref | PubMed | Scopus (162)See all References13], or about the subsequent population dynamics during the nearly 40,000 years spanning from the Late Pleistocene to the Neolithic transition [14xAncient human genomes suggest three ancestral populations for present-day Europeans. Lazaridis, I., Patterson, N., Mittnik, A., Renaud, G., Mallick, S., Kirsanow, K., Sudmant, P.H., Schraiber, J.G., Castellano, S., Lipson, M. et al. Nature. 2014;
513: 409–413
Crossref | PubMed | Scopus (173)See all References14]. Here, we reconstructed 35 complete or nearly complete mtDNAs (from 11× to 1,860× average coverage) of ancient modern human individuals from Italy, Germany, Belgium, France, Czech Republic, and Romania, spanning in age from 35 to 7 ka (Figure 1Figure 1; Table S1xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Table S1). Hybridization capture of mtDNA in combination with high-throughput sequencing technologies [15xMultiplexed DNA sequence capture of mitochondrial genomes using PCR products. Maricic, T., Whitten, M., and Pääbo, S. PLoS ONE. 2010;
5: e14004
Crossref | PubMed | Scopus (151)See all References15] allowed us to evaluate typical DNA damage patterns and average fragment length [16xTargeted retrieval and analysis of five Neandertal mtDNA genomes. Briggs, A.W., Good, J.M., Green, R.E., Krause, J., Maricic, T., Stenzel, U., Lalueza-Fox, C., Rudan, P., Brajkovic, D., Kucan, Z. et al. Science. 2009;
325: 318–321
Crossref | PubMed | Scopus (275)See all References16] as criteria for authentication of ancient DNA (Supplemental Experimental ProceduresxDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Supplemental Experimental Procedures). Both features were taken into account in an iterative probabilistic approach [17xSchmutzi: estimation of contamination and endogenous mitochondrial consensus calling for ancient DNA. Renaud, G., Slon, V., Duggan, A.T., and Kelso, J. Genome Biol. 2015;
16: 224
Crossref | PubMed | Scopus (10)See all References17] that jointly estimates present-day human contamination and reconstructs mtDNA sequences (Table S2xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Table S2). Combining 311 modern and 66 ancient dated worldwide mtDNA genomes (both new and from the literature; Table S3xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Table S3), we used Bayesian phylogenetic methods [18xBEAST: Bayesian evolutionary analysis by sampling trees. Drummond, A.J. and Rambaut, A. BMC Evol. Biol. 2007;
7: 214
Crossref | PubMed | Scopus (6657)See all References18] to estimate the mutation rate and hg coalescence times. Further, we combined our 35 new mtDNA genomes with 20 previously published ancient European mtDNAs for a total of 55 pre-Neolithic sequences (Table S4xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Table S4) and explicitly tested scenarios of the early population history of Europe using coalescent demographic modeling paired with approximate Bayesian computation (ABC) [19xApproximate Bayesian computation in population genetics. Beaumont, M.A., Zhang, W., and Balding, D.J. Genetics. 2002;
162: 2025–2035
PubMedSee all References19] (Supplemental Experimental ProceduresxDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Supplemental Experimental Procedures).
Figure 1
Late Pleistocene and Early Holocene Archeological Sites and Hunter-Gatherer mtDNA Haplogroups
(A) Pre-LGM dispersal of non-African populations, carrying both M and N lineages (hgs R, U, U5, and U2′3′4′7′8′9 belong to the N clade, distinct from the M clade).
(B) Post-LGM re-expansion in Europe while ice sheets retracted.
(C) Late Glacial shift in mtDNA hg frequency.
(D) Holocene hunter-gatherer mtDNA, mainly belonging to hg U5.
See also Table S1xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Table S1, Table S2xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Table S2, Table S4xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Table S4, and the Supplemental Experimental ProceduresxDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Supplemental Experimental Procedures.
Hg assignment of the authenticated mtDNAs confirmed that the vast majority of Late Pleistocene and Early Holocene individuals belonged to the U lineage, which is a subgroup of the N clade [20xHuman paleogenetics of Europe--the known knowns and the known unknowns. Brandt, G., Szécsényi-Nagy, A., Roth, C., Alt, K.W., and Haak, W. J. Hum. Evol. 2015;
79: 73–92
Crossref | PubMed | Scopus (21)See all References20] (Figures 2Figures 2 and S1xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6S1). We also found a basal U lineage that had no derived position leading to known sub-hgs in a 33,000-year-old Romanian individual. Surprisingly, three hunter-gatherers from Belgium and France dating to between 35 and 28 ka carried mtDNA hg M, today found predominantly in Asia, Australasia, and the Americas, although it is almost absent in extant populations with European ancestry [12xMaternal ancestry and population history from whole mitochondrial genomes. Kivisild, T. Investig. Genet. 2015;
6: 3
Crossref | PubMed | Scopus (7)See all References12].
Figure 2
Maximum Parsimony Tree of Present-Day Human and 55 Pre-Neolithic mtDNA Genomes
Pre-LGM samples are shown in blue, Post-LGM in green, Late Glacial in magenta, Holocene hunter-gatherers in red, and present-day individuals in black print. Average values of 14C dates are reported next to each specimen when available. Red arrows indicate divergence times of M and N clades. Hg M is almost absent in present-day individuals with European ancestry. Oase1 represents a pre-N lineage. The tree is rooted with one Neanderthal and 16 deeply divergent African mtDNAs (not shown). See also Figure S1xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Figure S1 and the Supplemental Experimental ProceduresxDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Supplemental Experimental Procedures.
We used 66 ancient dated mtDNAs as tip calibration points in BEAST v1.8.1 [18xBEAST: Bayesian evolutionary analysis by sampling trees. Drummond, A.J. and Rambaut, A. BMC Evol. Biol. 2007;
7: 214
Crossref | PubMed | Scopus (6657)See all References18] in combination with 311 modern worldwide mtDNA sequences to reduce the possible impact of sample biases (Table S3xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Table S3 and Supplemental Experimental ProceduresxDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Supplemental Experimental Procedures) in estimating the mtDNA mutation rate and hg M and N divergence times. Strict and uncorrelated lognormal relaxed clocks were tested, under both a constant size and a Bayesian skyline tree prior. The four analyses returned mtDNA mutation rates (Table 1Table 1) consistent with previously published rates using similar methodology [21xA revised timescale for human evolution based on ancient mitochondrial genomes. Fu, Q., Mittnik, A., Johnson, P.L., Bos, K., Lari, M., Bollongino, R., Sun, C., Giemsch, L., Schmitz, R., Burger, J. et al. Curr. Biol. 2013;
23: 553–559
Abstract | Full Text | Full Text PDF | PubMed | Scopus (120)See all References, 22xImproved calibration of the human mitochondrial clock using ancient genomes. Rieux, A., Eriksson, A., Li, M., Sobkowiak, B., Weinert, L.A., Warmuth, V., Ruiz-Linares, A., Manica, A., and Balloux, F. Mol. Biol. Evol. 2014;
31: 2780–2792
Crossref | PubMed | Scopus (14)See all References]. The Bayesian skyline, in combination with strict rate variation among branches, performed best according to stepping-stone and path sampling methods [23xMake the most of your samples: Bayes factor estimators for high-dimensional models of sequence evolution. Baele, G., Lemey, P., and Vansteelandt, S. BMC Bioinformatics. 2013;
14: 85
Crossref | PubMed | Scopus (25)See all References23] and highest effective sample size (ESS) values, giving a best estimate of the mutation rate of 2.74 × 10−8 (95% highest posterior density [HPD], 2.44–3.01 × 10−8) mutation/site/year. This model allowed us to refine time estimates for the TMRCA of the basal non-African clades M and N of circa 49 ka (95% HPD, 54.8–43.6 ka) and 51 ka (95% HPD, 55.1–46.9 ka), respectively (Table 1Table 1; Figure 2Figure 2; Supplemental Experimental ProceduresxDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Supplemental Experimental Procedures).
| Tree Prior | Clock | Statistic | Divergence Time | Clock Rate Whole mtDNA | Log Marginal Likelihood | ||
|---|---|---|---|---|---|---|---|
| TMRCA hg M | TMRCA hg N | Stepping-Stone Sampling | Path Sampling | ||||
| Constant | strict | mean | 58,869 | 57,482 | 2.62 × 10−8 | −48,759 | −48,754 |
| median | 58,578 | 57,181 | 2.62 × 10−8 | ||||
| 95% HPD | 68,163–50,380 | 64,363–51,387 | 2.30–2.93 × 10−8 | ||||
| ESS | 585 | 445 | 651 | ||||
| Constant | relaxed | mean | 58,961 | 58,531 | 2.67 × 10−8 | −48,755 | −48,751 |
| median | 58,507 | 58,207 | 2.67 × 10−8 | ||||
| 95% HPD | 70,389–49,125 | 66,398–51,664 | 2.30–3.04 × 10−8 | ||||
| ESS | 354 | 416 | 431 | ||||
| Skyline | strict | mean | 49,106 | 50,562 | 2.74 × 10−8 | −48,577 | −48,571 |
| median | 48,837 | 50,317 | 2.74 × 10−8 | ||||
| 95% HPD | 54,780–43,598 | 55,138–46,892 | 2.44–3.01 × 10−8 | ||||
| ESS | 741 | 799 | 863 | ||||
| Skyline | relaxed | mean | 48,005 | 50,179 | 2.77 × 10−8 | −48,550 | −48,546 |
| median | 47,695 | 50,021 | 2.77 × 10−8 | ||||
| 95% HPD | 53,917–43,054 | 54,189–46,483 | 2.47–3.07 × 10−8 | ||||
| ESS | 251 | 285 | 348 | ||||
The values reported are obtained in BEAST [18xBEAST: Bayesian evolutionary analysis by sampling trees. Drummond, A.J. and Rambaut, A. BMC Evol. Biol. 2007;
7: 214
Crossref | PubMed | Scopus (6657)See all References18] using 377 worldwide mtDNAs, 66 of which come from ancient dated human remains. A Bayesian skyline tree prior in combination with strict rate variation between branches performed better than the other three tested models according to higher log marginal likelihood estimates (compared to the constant tree prior models) and effective sample size (ESS) values. HPD, highest posterior density. See also Table S3xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Table S3 and the Supplemental Experimental ProceduresxDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Supplemental Experimental Procedures.
The observed mtDNA hg variation through time, including the apparent loss of hg M in Europe, suggests a genetic bottleneck that may have been influenced by climatic events (Figure 3Figure 3). This period of European prehistory was accompanied by severe climatic fluctuations, such as the Last Glacial Maximum (LGM, 25 to 19.5 ka) and, at the end of the Pleistocene, the Bølling-Allerød interstadial followed by the stadial Younger Dryas—a period we refer to as the Late Glacial (14.5 to 11.5 ka) [24xValidation of climate model-inferred regional temperature change for late-glacial Europe. Heiri, O., Brooks, S.J., Renssen, H., Bedford, A., Hazekamp, M., Ilyashuk, B., Jeffers, E.S., Lang, B., Kirilova, E., Kuiper, S. et al. Nat. Commun. 2014;
5: 4914
Crossref | PubMed | Scopus (31)See all References, 25xClimate change and evolving human diversity in Europe during the last glacial. Gamble, C., Davies, W., Pettitt, P., and Richards, M. Philos. Trans. R. Soc. Lond. B Biol. Sci. 2004;
359: 243–253
Crossref | PubMed | Scopus (139)See all References]. These climatic changes have been proposed as a driver of the range contraction to refugia in many species [26xRefugia revisited: individualistic responses of species in space and time. Stewart, J.R., Lister, A.M., Barnes, I., and Dalén, L. Proc. Biol. Sci. 2010;
277: 661–671
Crossref | PubMed | Scopus (394)See all References26], including modern humans, for whom there is absence of evidence of northwestern European occupation during the LGM [25xClimate change and evolving human diversity in Europe during the last glacial. Gamble, C., Davies, W., Pettitt, P., and Richards, M. Philos. Trans. R. Soc. Lond. B Biol. Sci. 2004;
359: 243–253
Crossref | PubMed | Scopus (139)See all References, 27xHuman evolution out of Africa: the role of refugia and climate change. Stewart, J.R. and Stringer, C.B. Science. 2012;
335: 1317–1321
Crossref | PubMed | Scopus (89)See all References]. We used coalescent modeling paired with ABC [19xApproximate Bayesian computation in population genetics. Beaumont, M.A., Zhang, W., and Balding, D.J. Genetics. 2002;
162: 2025–2035
PubMedSee all References19] to test a range of explicit models of European hunter-gatherer demography (Figure S2xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Figure S2; Table S6xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Table S6), using the complete set of 55 pre-Neolithic ancient mtDNA genomes (Table S4xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Table S4). The best-fitting model (Figure 3Figure 3 and 2b in Figure S2xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Figure S2) strongly supports maternal population continuity through the LGM, albeit as a single genetic bottleneck, before being replaced by a new incoming population at the onset of the Late Glacial 14.5 ka (model posterior probability, P2b = 0.807). Based on the estimated parameter values of this model (Table S5xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Table S5), we infer that this surviving population diverged from the ancestral one around 29 ka (95% HPD, 36–25 ka), prior to the beginning of the LGM.
Figure 3
Late Pleistocene and Early Holocene Climatic Fluctuations and European Hunter-Gatherer Demography
On the left is the NGRIP δ18O climate record, and on the right is an illustration of the best-supported demographic model (2b in Figure S2xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Figure S2). Each colored point gives the mtDNA hg of the 55 dated pre-Neolithic individuals used in the coalescent modeling analysis. West-East site locations for each sample are approximated. See also Figure S2xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Figure S2, Table S4xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Table S4, and the Supplemental Experimental ProceduresxDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Supplemental Experimental Procedures.
The new hunter-gatherer mtDNA genomes reported here approximately triple the available sequences from pre-Neolithic Europe. One novel finding, that three out of 18 European pre-LGM hunter-gatherers carry a previously undescribed basal mtDNA lineage M (Figure 1Figure 1A), has important implications for the timing of the dispersal of modern humans into Eurasia. Excluding a ∼40,000-year-old Romanian individual known not to have contributed notably to the modern European gene pool [28xAn early modern human from Romania with a recent Neanderthal ancestor. Fu, Q., Hajdinjak, M., Moldovan, O.T., Constantin, S., Mallick, S., Skoglund, P., Patterson, N., Rohland, N., Lazaridis, I., Nickel, B. et al. Nature. 2015;
524: 216–219
Crossref | PubMed | Scopus (65)See all References28], our BEAST analyses give a TMRCA for clades M and N from 44 to 55 ka, respectively. Our estimated dates, together with the oldest accepted archeological evidence for the presence of early modern humans in Australia and Europe (both dated to at least 45 ka [13xEarly dispersal of modern humans in Europe and implications for Neanderthal behaviour. Benazzi, S., Douka, K., Fornai, C., Bauer, C.C., Kullmer, O., Svoboda, J., Pap, I., Mallegni, F., Bayle, P., Coquerelle, M. et al. Nature. 2011;
479: 525–528
Crossref | PubMed | Scopus (162)See all References, 29xNew ages for human occupation and climatic change at Lake Mungo, Australia. Bowler, J.M., Johnston, H., Olley, J.M., Prescott, J.R., Roberts, R.G., Shawcross, W., and Spooner, N.A. Nature. 2003;
421: 837–840
Crossref | PubMed | Scopus (346)See all References]), are consistent with a model of a single, late, and therefore rapid dispersal of a source population containing both M and N hgs, which contributed all the mitochondrial diversity of present-day non-Africans (cf. [7xMajor genomic mitochondrial lineages delineate early human expansions. Maca-Meyer, N., González, A.M., Larruga, J.M., Flores, C., and Cabrera, V.M. BMC Genet. 2001;
2: 13
Crossref | PubMed | Scopus (108)See all References7]). Human individuals whose ancestries trace back to potential earlier expansion(s) outside Africa [30xThe earliest unequivocally modern humans in southern China. Liu, W., Martinón-Torres, M., Cai, Y.J., Xing, S., Tong, H.W., Pei, S.W., Sier, M.J., Wu, X.H., Edwards, R.L., Cheng, H. et al. Nature. 2015;
526: 696–699
Crossref | PubMed | Scopus (31)See all References, 31xEarly modern human dispersal from Africa: genomic evidence for multiple waves of migration. Tassi, F., Ghirotto, S., Mezzavilla, M., Vilaça, S.T., De Santi, L., and Barbujani, G. Investig. Genet. 2015;
6: 13
Crossref | PubMed | Scopus (5)See all References] are thus unlikely to have left any surviving mtDNA descendants.
Phylogeographic inference based solely on mtDNA has limitations [2xRethinking the dispersal of Homo sapiens out of Africa. Groucutt, H.S., Petraglia, M.D., Bailey, G., Scerri, E.M., Parton, A., Clark-Balzan, L., Jennings, R.P., Lewis, L., Blinkhorn, J., Drake, N.A. et al. Evol. Anthropol. 2015;
24: 149–164
Crossref | PubMed | Scopus (30)See all References2], but information from single loci can provide meaningful constraints on models of human prehistory. In particular, the fact that hg M has never previously been found in Europe is generally interpreted as an important limitation for the proposed scenarios of non-African population dispersals [4xSingle, rapid coastal settlement of Asia revealed by analysis of complete mitochondrial genomes. Macaulay, V., Hill, C., Achilli, A., Rengo, C., Clarke, D., Meehan, W., Blackburn, J., Semino, O., Scozzari, R., Cruciani, F. et al. Science. 2005;
308: 1034–1036
Crossref | PubMed | Scopus (425)See all References, 7xMajor genomic mitochondrial lineages delineate early human expansions. Maca-Meyer, N., González, A.M., Larruga, J.M., Flores, C., and Cabrera, V.M. BMC Genet. 2001;
2: 13
Crossref | PubMed | Scopus (108)See all References]. According to the most popular model [4xSingle, rapid coastal settlement of Asia revealed by analysis of complete mitochondrial genomes. Macaulay, V., Hill, C., Achilli, A., Rengo, C., Clarke, D., Meehan, W., Blackburn, J., Semino, O., Scozzari, R., Cruciani, F. et al. Science. 2005;
308: 1034–1036
Crossref | PubMed | Scopus (425)See all References4], an early expansion occurred before the M and N diversification with a subsequent loss of M in only the population ancestral to Europeans. Our evidence for the existence of hg M in Late Pleistocene Europe revises this scenario. It suggests that the loss of hg M may be due to population dynamics that occurred later within Europe itself. The expansion either occurred before the diversification of M and N, with subsequent migration bringing both lineages into Europe, or the dispersal was later, occurring after their TMRCAs. Contrary to recent findings [11xA “Copernican” reassessment of the human mitochondrial DNA tree from its root. Behar, D.M., van Oven, M., Rosset, S., Metspalu, M., Loogväli, E.L., Silva, N.M., Kivisild, T., Torroni, A., and Villems, R. Am. J. Hum. Genet. 2012;
90: 675–684
Abstract | Full Text | Full Text PDF | PubMed | Scopus (138)See all References11], though similar to a previous study [32xPhylogenetic star contraction applied to Asian and Papuan mtDNA evolution. Forster, P., Torroni, A., Renfrew, C., and Röhl, A. Mol. Biol. Evol. 2001;
18: 1864–1881
Crossref | PubMedSee all References32], our two TMRCAs are almost identically dated, suggesting a single major dispersal after 55 ka for all non-African populations, including Europe. The genetic evidence of pre-LGM hg M indicates that this lineage reached Western Europe by at least 35 ka (GoyetQ116-1), either alongside the first arrival of N or later. The reconstructed phylogeny (Figure 2Figure 2) with both basal N and M lineages in Late Pleistocene Europe possibly mirrors the inferred back migration into Africa, which has been suggested by the existence of hgs U6 and M1 in modern-day North Africans [33xThe mtDNA legacy of the Levantine early Upper Palaeolithic in Africa. Olivieri, A., Achilli, A., Pala, M., Battaglia, V., Fornarino, S., Al-Zahery, N., Scozzari, R., Cruciani, F., Behar, D.M., Dugoujon, J.M. et al. Science. 2006;
314: 1767–1770
Crossref | PubMed | Scopus (146)See all References33]. Therefore, the major modern human dispersal described here after 55 ka might have affected not only non-Africans, but also African populations to some extent.
The potential impact of climatic events on the demography, and thus the genetic diversity of early Europeans, has previously been difficult to quantify, but it likely had consequences for the relative components of ancient ancestry in modern-day populations [14xAncient human genomes suggest three ancestral populations for present-day Europeans. Lazaridis, I., Patterson, N., Mittnik, A., Renaud, G., Mallick, S., Kirsanow, K., Sudmant, P.H., Schraiber, J.G., Castellano, S., Lipson, M. et al. Nature. 2014;
513: 409–413
Crossref | PubMed | Scopus (173)See all References14]. Our demographic modeling reveals a dynamic history of hunter-gatherers, including a previously unknown major population shift during the Late Glacial interstadial (the Bølling-Allerød, ∼14.5 ka). Under our best-fitting model (Figure 3Figure 3 and 2b in Figure S2xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Figure S2), the small initial founder population of Europe slowly grows up until ∼25 ka and survives with smaller size in LGM climatic refugia (25–19.5 ka) [25xClimate change and evolving human diversity in Europe during the last glacial. Gamble, C., Davies, W., Pettitt, P., and Richards, M. Philos. Trans. R. Soc. Lond. B Biol. Sci. 2004;
359: 243–253
Crossref | PubMed | Scopus (139)See all References25] before re-expanding as the ice sheets retract (Figure 1Figure 1B). Although this model supports population continuity from pre- to post-LGM, the genetic bottleneck is consistent with the apparent loss of hg M in the post-LGM. The subsequent Late Glacial period is characterized by drastic climatic fluctuations, beginning with an abrupt warming during the Bølling-Allerød interstadial and followed by a similarly drastic period of cooling during the Younger Dryas [24xValidation of climate model-inferred regional temperature change for late-glacial Europe. Heiri, O., Brooks, S.J., Renssen, H., Bedford, A., Hazekamp, M., Ilyashuk, B., Jeffers, E.S., Lang, B., Kirilova, E., Kuiper, S. et al. Nat. Commun. 2014;
5: 4914
Crossref | PubMed | Scopus (31)See all References24]. Globally, the early warming phases of the Late Glacial are strongly associated with substantial demographic changes, including extinctions of several megafaunal species [34xPALEOECOLOGY. Abrupt warming events drove Late Pleistocene Holarctic megafaunal turnover. Cooper, A., Turney, C., Hughen, K.A., Brook, B.W., McDonald, H.G., and Bradshaw, C.J. Science. 2015;
349: 602–606
Crossref | PubMed | Scopus (29)See all References34] and the first expansion of modern humans into the Americas [35xFirst Peoples in a New World: Colonizing Ice Age America. Meltzer, D.J.
See all References35]. In European hunter-gatherers, our model best explains this period of upheaval as a replacement of the post-LGM maternal population by one from another source. Although the exact origin for this later population is unknown, the inferred demographic history (Figure 3Figure 3 and 2b in Figure S2xDownload
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6Figure S2) suggests that it descended from another, separate LGM refugium. On the basis of mtDNA alone, we cannot rule out some degree of genomic continuity throughout the Late Pleistocene and early Holocene hunter-gatherer populations, and thus into present-day Europeans [14xAncient human genomes suggest three ancestral populations for present-day Europeans. Lazaridis, I., Patterson, N., Mittnik, A., Renaud, G., Mallick, S., Kirsanow, K., Sudmant, P.H., Schraiber, J.G., Castellano, S., Lipson, M. et al. Nature. 2014;
513: 409–413
Crossref | PubMed | Scopus (173)See all References14]. For this reason, we interpret our model as capturing the maternal signal of a major population shift, rather than a complete replacement. Ancient nuclear DNA data and additional geographically and temporally distributed specimens may provide a more comprehensive picture, possibly identifying the source and ancestry of these later incoming hunter-gatherers.
In conclusion, the large dataset presented here allowed us to provide a late upper bound on the major dispersal of all non-Africans and to uncover unexpected population dynamics of European hunter-gatherers. The Late Glacial event that we identify here is the oldest in an accumulating list of major European population turnovers revealed by ancient mtDNA [20xHuman paleogenetics of Europe--the known knowns and the known unknowns. Brandt, G., Szécsényi-Nagy, A., Roth, C., Alt, K.W., and Haak, W. J. Hum. Evol. 2015;
79: 73–92
Crossref | PubMed | Scopus (21)See all References20].
Author Contributions
D.G.D., H.R., C.C., F.V., C.T., M.F., M.M., M.B., M.L., E.G., G.C. I.C., C.B., D.F., M.G., J.v.d.P., R.C., B.G., A.R., K.W., H.B., D.G., J.S., D.C., P.S., K.H., and N.J.C. provided archeological material and related information. C.P., A.M., A.F., and C.W. performed laboratory work. C.P., G.R., W.H., A.P., and J.K. analyzed genetic data. C.P., W.H., A.P., and J.K. wrote the manuscript with input from all co-authors.
Acknowledgments
We are grateful for comments from Stephan Schiffels, Hugo Reyes-Centeno, David Reich, Maria Spyrou, Henrike Heyne, Heidi Colleran, and members of the Department of Archaeogenetics of the Max Planck Institute for the Science of Human History, as well as the three anonymous reviewers. We thank Pontus Skoglund, Qiaomei Fu, Viviane Slon, and Eppie Ruth-Jones for access to unpublished data, Martyna Molak, Alexander Peltzer, Marek Dynowski, and Judith Beier for technical support, and Annette Günzel for graphical support. We further thank the Soprintendenza Archeologia della Puglia, which authorized and supported the excavations at Grotta Paglicci, and Professor A. Palma di Cesnola for fieldwork and studies over the years. Part of this work was performed on the computational resource bwGRiD Cluster Tübingen funded by the Ministry of Science, Research and the Arts Baden-Württemberg and the Universities of the State of Baden-Württemberg, Germany, within the framework program bwHPC. J.K. and C.P. were supported by the Baden Württemberg Foundation, J.K. and A.M. by the DFG grant KR 4015/1-1, and K.H. by the European Research Council (ERC StG 283503). The Goyet project led by H.R. was funded by the Wenner-Gren Foundation (grant no. 7837), the College of Social and Behavioral Sciences of CSUN, and the RBINS.
Accession Numbers
Supplemental Information
Download
(1.48 MB
)
Document S1. Supplemental Experimental Procedures, Figures S1 and S2, and Tables S1–S6