Initial Upper Palaeolithic Homo sapiens from Bacho Kiro Cave, Bulgaria

Abstract

The Middle to Upper Palaeolithic transition in Europe witnessed the replacement and partial absorption of local Neanderthal populations by Homo sapiens populations of African origin1. However, this process probably varied across regions and its details remain largely unknown. In particular, the duration of chronological overlap between the two groups is much debated, as are the implications of this overlap for the nature of the biological and cultural interactions between Neanderthals and H. sapiens. Here we report the discovery and direct dating of human remains found in association with Initial Upper Palaeolithic artefacts2, from excavations at Bacho Kiro Cave (Bulgaria). Morphological analysis of a tooth and mitochondrial DNA from several hominin bone fragments, identified through proteomic screening, assign these finds to H. sapiens and link the expansion of Initial Upper Palaeolithic technologies with the spread of H. sapiens into the mid-latitudes of Eurasia before 45 thousand years ago3. The excavations yielded a wealth of bone artefacts, including pendants manufactured from cave bear teeth that are reminiscent of those later produced by the last Neanderthals of western Europe4–6. These finds are consistent with models based on the arrival of multiple waves of H. sapiens into Europe coming into contact with declining Neanderthal populations7,8. Direct dates for human remains found in association with Initial Upper Palaeolithic artefacts at Bacho Kiro Cave (Bulgaria) demonstrate the presence of Homo sapiens in the mid-latitudes of Europe before 45 thousand years ago.

Full text

Nature | www.nature.com | 1
Article
Initial Upper Palaeolithic Homo sapiens from
Bacho Kiro Cave, Bulgaria
Jean-Jacques Hublin1,2 ✉, Nikolay Sirakov3, Vera Aldeias4, Shara Bailey1,5, Edouard Bard6,
Vincent Delvigne7,8, Elena Endarova9, Yoann Fagault6, Helen Fewlass1, Mateja Hajdinjak10,
Bernd Kromer1, Ivaylo Krumov11, João Marreiros4,12 , Naomi L. Martisius13, Lindsey Paskulin14,
Virginie Sinet-Mathiot1, Matthias Meyer10, Svante Pääbo10, Vasil Popov15, Zeljko Rezek1,16 ,
Svoboda Sirakova3, Matthew M. Skinner1,17, Geoff M. Smith1, Rosen Spasov18, Sahra Talamo1,19,
Thibaut Tuna6, Lukas Wacker20, Frido Welker1,21, Arndt Wilcke22, Nikolay Zahariev23,
Shannon P. McPherron1 & Tsenka Tsanova1
The Middle to Upper Palaeolithic transition in Europe witnessed the replacement and
partial absorption of local Neanderthal populations by Homo sapiens populations of
African origin1. However, this process probably varied across regions and its details
remain largely unknown. In particular, the duration of chronological overlap between
the two groups is much debated, as are the implications of this overlap for the nature
of the biological and cultural interactions between Neanderthals and H.sapiens. Here
we report the discovery and direct dating of human remains found in association with
Initial Upper Palaeolithic artefacts2, from excavations at Bacho Kiro Cave (Bulgaria).
Morphological analysis of a tooth and mitochondrial DNA from several hominin bone
fragments, identied through proteomic screening, assign these nds to H.sapiens
and link the expansion of Initial Upper Palaeolithic technologies with the spread of
H.sapiens into the mid-latitudes of Eurasia before 45thousand years ago3. The
excavations yielded a wealth of bone artefacts, including pendants manufactured
from cave bear teeth that are reminiscent of those later produced by the last
Neanderthals of western Europe4–6. These nds are consistent with models based on
the arrival of multiple waves of H.sapiens into Europe coming into contact with
declining Neanderthal populations7,8.
Fragmentary specimens from the sites of Kent’s Cavern (United King-
dom)
9
and Cavallo (Italy)
10
have been claimed to document the earliest
presence of our species in western Europe, between 44,200–41,500cali-
brated years before present (cal. ; taken as 1950) for the former
and between 45,000–43,000cal.  for the latter. However, these dates
are based on the archaeological contexts of the specimens rather than
direct dating, and—in both cases—the exact stratigraphic origin of the
fossils is debated
11,12
. In the absence of directly dated fossil remains,
reconstructing the timing of the expansions of H.sapiens into Europe
rests on hypotheses concerning the makers of various so-called ‘tran-
sitional’ artefact assemblages at the advent of the Upper Palaeolithic.
Bacho Kiro Cave is located 5km west of Dryanovo (Bulgaria), on
the northern slope of the Balkan mountain range (Stara Planina) and
about 70km south of the Danube River (Extended Data Fig.1b). The
site formed at the mouth of a large karstic system and its deposits
encompass late Middle Palaeolithic and early Upper Palaeolithic
occupations. Bacho Kiro Cave was excavated by D.Garrod in 1938,
but is best known from more extensive excavations (in 1971 to 1975) by
a team led by B.Ginter and J.Kozłowski13. The excavations in the 1970s
yielded fragmentary human remains13 that were subsequently lost. In
2015, the National Archaeological Institute with Museum in Sofia and
the Department of Human Evolution at the Max Planck Institute for
Evolutionary Anthropology resumed work at Bacho Kiro Cave with the
goals of clarifying the chronology (which had previously been based on
a handful of inconsistent radiocarbon ages
14
) and the biological nature
of the makers of the lithic assemblages. Two sectors with similar and
well-preserved sequences were re-excavated: the Main sector and the
previously unexcavated Niche1 sector, located on the south and east
https://doi.org/10.1038/s41586-020-2259-z
Received: 30 July 2019
Accepted: 24 February 2020
Published online: xx xx xxxx
Check for updates
1Department of Human Evolution, Max Planck Institute for Evolutionary Anthropology, Leipzig, Germany. 2Chaire Internationale de Paléoanthropologie, Collège de France, Paris, France.
3National Institute of Archaeology with Museum, Bulgarian Academy of Sciences, Soia, Bulgaria. 4Interdisciplinary Centre for Archaeology and the Evolution of Human Behaviour, Universidade
do Algarve, Faro, Portugal. 5Department of Anthropology, New York University, New York, NY, USA. 6CEREGE, Aix Marseille University, CNRS, IRD, INRAE, Collège de France, Aix-en-Provence,
France. 7Service de Préhistoire, University of Liège, Liège, Belgium. 8CNRS, UMR 5199 PACEA, University of Bordeaux, Pessac, France. 9National History Museum, Soia, Bulgaria. 10Department of
Evolutionary Genetics, Max Planck Institute for Evolutionary Anthropology, Leipzig, Germany. 11History Museum, Belogradchik, Bulgaria. 12TraCEr, Monrepos Archaeological Research Centre
and Museum for Human Behavioural Evolution, RGZM, Mainz, Germany. 13Department of Anthropology, University of California, Davis, Davis, CA, USA. 14Department of Archaeology, University
of Aberdeen, Aberdeen, UK. 15Institute of Biodiversity and Ecosystem Research, Bulgarian Academy of Sciences, Soia, Bulgaria. 16University of Pennsylvania Museum of Archaeology and
Anthropology, Philadelphia, PA, USA. 17School of Anthropology and Conservation, University of Kent, Canterbury, UK. 18Archaeology Department, New Bulgarian University, Soia, Bulgaria.
19Department of Chemistry ‘G. Ciamician’, University of Bologna, Bologna, Italy. 20Department of Earth Sciences, ETH Zurich, Zurich, Switzerland. 21Evolutionary Genomics Section, Globe
Institute, University of Copenhagen, Copenhagen, Denmark. 22Department of Cell Therapy, Fraunhofer Institute for Cell Therapy and Immunology, Leipzig, Germany. 23Archaeology
Department, New Bulgarian University, Soia, Bulgaria. ✉e-mail: hublin@eva.mpg.de
Content courtesy of Springer Nature, terms of use apply. Rights reserved

2 | Nature | www.nature.com
Article
sides, respectively, of the excavation from the 1970s (Extended Data
Fig.1a). At the base of the sequence (Supplementary Information sec-
tion1) and overlying the bedrock, layerK has a relatively low density of
Middle Palaeolithic artefacts. Sedimentologically, the contact of layerK
with the overlying layerJ is gradual and the artefact densities remain
low; however, the upper part of layerJ contains artefacts identical to
those in layerI. Based on the radiocarbon dates
3
, layerJ represents
more than 3,000years of accumulation. LayerI represents an inten-
sification of the trends seen in layerJ. LayerI is an easily recognized
and archaeologically rich organic deposit that spans from 45,820 to
43,650cal. 
3
(95% modelled range) and yields an assemblage that was
initially described as ‘Bachokirian’, but is now considered a variant of
the Initial Upper Palaeolithic (IUP) industry
15
(Extended Data Figs.2–4,
Supplementary Information section2). LayerI is capped by water-laid
deposits (layersH and G) that have little archaeological content. More
than 1.7m of deposits, containing low densities of Upper Palaeolithic
artefacts, overlays layer G in the Main sector.
We found a hominin second lower molar (specimen code F6-620)
(Extended Data Fig.5a) in the upper part of layerJ. The crown dimen-
sions of this tooth place it at the high end of both the Neanderthal
and the Upper Palaeolithic H.sapiens range (Extended Data Table1).
With the exception of a moderately expressed—but divided—middle
trigonid crest, all of the morphological trait expressions found in
F6-620 align the tooth with H.sapiens (Supplementary Information
section3). The expression of a middle trigonid crest observed on the
second lower molar from Bacho Kiro Cave is present in 10% of these
teeth in some groups of humans today
16
and in 8% of early H. sapiens
17
.
The pulp chamber is hypotaurodont18, a condition that is common in
some recent human groups
19
and is unlike the hypertaurodont molars
of Neanderthals
20
. The four-cusp configuration of the second lower
molar from Bacho Kiro Cave is absent in Neanderthals. Our geometric
morphometric analysis of the enamel–dentine junction also clearly
assigns the specimen to H.sapiens (Extended Data Fig.5b).
We screened 1,271non-identifiable bones and teeth using
matrix-assisted laser desorption–ionization time-of-flight mass
spectrometry (MALDI–TOF MS) collagen-peptide mass fingerprint-
ing (also known as ZooMS21) to identify hominin remains, with the
additional aims of providing accurate molecular identifications for
radiocarbon-dated specimens and of enlarging our understanding
of the species composition of the fauna. ZooMS screening identified
six hominin bone fragments (Extended Data Fig.6, Supplementary
Information section4), of which four come from layerI in Niche1, one
from layerB in the Main sector (Extended Data Fig.1) and one from the
interface of layers 6a and 7 of the excavations in the 1970s13. Including
the F6-620 tooth, we recovered five hominin specimens in total from the
IUP layers. The calibrated radiocarbon dates of the 4ZooMS-identified
human fragments range from 46,790to 42,810cal.  at 95.4% prob-
ability (Fig.1). These ages are in full agreement with the modelled
boundaries of layerI (45,820–43,650cal.  at 95.4%), which includes
the 4humans and 21other dates on modified fauna
3
. Therefore, to
our knowledge, these bones represent the oldest European Upper
Palaeolithic hominins recovered to date.
We extracted DNA22,23 from F6-620 and the six hominin bone frag-
ments identified using ZooMS. We performed library preparation
24
,
enrichment of human mitochondrial DNA (mtDNA)
25
and sequencing,
which enabled us to recover between 13,856 and 795,043unique mtDNA
fragments (Supplementary Information section5). The frequencies
of cytosine-to-thymine substitutions, which are characteristic of
ancient-DNA base damage, ranged from 13.5% to 54.9% at the 5 ends
and from 9.4% to 42.2% at the 3 ends of these fragments (Extended
Data Fig.7), which suggests that at least some of the fragments are of
ancient origin. After restricting analyses to putatively deaminated DNA
fragments to remove contamination by recent human DNA, sequence
coverage of the mitochondrial genome enabled us to reconstruct six full
mitochondrial genomes out of seven. The mtDNA sequences of F6-620
and one of the ZooMS-identified hominin bone fragments (AA7-738)
are identical, which indicates that these specimens belonged either to
the same individual or to two maternally related individuals. In a tree
relating these mtDNA genomes to the known mtDNA sequences of
54present-day humans, 12ancient H.sapiens, 22Neanderthals, 4Denis-
ovans and a hominin from Sima de los Huesos, all of the Bacho Kiro Cave
mtDNA genomes fall within the variation of H.sapiens (Fig.2, Extended
Data Fig.8). The specimens from layerI yielded mtDNA sequences that
fall close to the base of each of the three major macro-haplogroups of
present-day non-Africans (M, N and R). Although the mtDNA sequences
belong to different macro-haplogroups, they differ (at most) at 15posi-
tions from each other—which is lower than the differences observed
among 97.5% of contemporary European individuals who are not closely
related to one another26. The older Bacho Kiro population contains
early representatives of the macro-haplogroup M, which is not present
in Europe today
27
. Furthermore, the mtDNA genomes of the Bacho
Kiro Cave specimens accumulated fewer substitutions than those
of present-day humans. Using 10directly dated ancient H.sapiens
as calibration points28,29 (Supplementary Information section5), we
obtained genetic dates that range from 44,830 to 42,616yr  for the
layer-I hominins (Extended Data Table2), in good agreement with the
calibrated radiocarbon dates (Fig.1).
The fauna associated with these H.sapiens specimens (11,259 piece
plotted animal bone fragments from layersI andJ) includes 23species,
dominated by Bos or Bison, cervids and caprines, alongside equids
(Supplementary Information section6). The species composition com-
prises a mix of taxa adapted both to cold and to warmer environments,
characteristic of the faunal record during marine isotope stage3 in the
Balkans
30,31
. A variety of carnivores are also present, dominated by cave
bear (Ursus spelaeus). Zooarchaeological analyses strongly indicate that
the accumulation of the fauna is predominantly anthropogenic. One
notable aspect of the faunal assemblage is the presence of numerous
anthropogenically modified objects (Fig.3, Supplementary Informa-
tion section6): worked pieces include awls, lissoirs (‘smoothers’) and
incised pieces. Several of the artefacts have red staining that is consist-
ent with the use of ochre. We identified 1perforated ivory bead and
12perforated or grooved pendants, 11 of which were made from cave
bear teeth and 1 from an ungulate tooth (Fig.3).
The stone tools associated with H.sapiens in layerI were initially
assigned to the Bachokirian techno-complex because they did not fit
35,00040,00045,00050,000
Date (cal. BP)
Châtelperronian Late Neanderthals
Europe
Grotte du Renne AR-14 (MAMS-25149)
Saint-Césaire SP 28 (OxA-18099)
Upper Palaeolithic
H. sapiens
Europe
Bacho Kiro CC7-335 (ETH-86772)
Bacho Kiro CC7-2289 (ETH-86771)
BachoKiro BB7-240 (ETH-86770)
Bacho Kiro AA7-738 (ETH-86769)
Peũtera cu Oase 1 (GrA-22810)
Kostenki 14 (OxA-X-2395-15)
Kostenki 1 (OxA-15055)
Bacho Kiro F6-597 (ETH-86773/AIX-12025*)
Bacho Kiro BK 1653 (ETH-86768/AIX-12024*)
Siberia Ust'-Ishim 1 (OxA-25516/OxA-30190*)
China Tianyuan Cave (BA-03222)
Fig. 1 | Dire ct dates for ho minins of the M iddle to Uppe r Palaeolith ic
transition in Eurasia. Directly dated Châtelperronian Neanderthals (blue) and
H.sapiens (red or black) of the M iddletoUpper Palaeoli thic transitio n in
Eurasia. T he dates from Ba cho Kiro Cave (red) are repo rted in an ass ociated
study3, as part of a n extensive site chr onology. Aster isks mark the dat es that
were combine d using the R _Combine func tion in OxCal v.4.3 . Accelerated ma ss
spectr ometry (AM S) laboratory co des are shown in pa rentheses; all t he dates
shown here are i n Supplement ary Table16, with sample infor mation and
references.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Nature | www.nature.com | 3
comfortably with either the Middle Palaeolithic or Aurignacian-like
Upper Palaeolithic techno-complexes. We now know that these tools
fit within the IUP
15
. IUP assemblages—similar to that of Bacho Kiro Cave
(Supplementary Information section2)—are characterized by blades
and tool types typical of the Upper Palaeolithic, but with some Levallois
forms and faceted platforms that are reminiscent of the preceding Mid-
dle Palaeolithic and African Middle Stone Age
2
(Extended Data Figs.3,
4). IUP assemblages, which span Eurasia from central Europe to Mon-
golia, occur before the appearance of Upper Palaeolithic assemblages
characterized by bladelet production, and arguably have their origin
in southwest Asia (Extended Data Fig.2). For instance, the Bacho Kiro
Cave IUP is similar to the IUP from layers I–F at Üçağızlı Cave (Turkey) in
terms of lithic technology, typology, and the presence of shaped bone
tools and pendants, as well as with respect to ages32,33.
The Bacho Kiro Cave site clearly demonstrates that the IUP in this
region was made by H.sapiens, and is consistent with models that attrib-
ute the spread of the IUP to the dispersal of our species throughout large
parts of Eurasia. The presence of IUP assemblages documents a wave
of peopling that precedes the spread of the first Upper Palaeolithic
bladelet techno-complexes—such as the Early Ahmarian industry in the
Levant, the Early Kozarnikan industry in the eastern Balkans and the Pro-
toaurignacian industry in western and central Europe—by several mil-
lennia1,34. At Bacho Kiro Cave, the IUP starts before 45,000cal.  and,
as the assemblage of the upper part of layerJ is identical to that from
layerI, it may begin as early as 47,000cal. 
3
. We now have evidence
for H.sapiens in Eurasia spanning from Ust'-Ishim28 in western Siberia
to Bacho Kiro Cave in eastern Europe, directly dated to approximately
45,000cal. . Together, the behavioural and biological evidence
strongly suggest a relatively rapid dispersal of IUP assemblages from
southwest Asia
35
into mid-latitude Eurasia by groups that—contrary to
Aurignacian populations—seem unrelated to present-day European
populations28. Direct contact with Neanderthals must have occurred
much earlier in eastern Europe than in western Europe, where the latest
Neanderthals and their associated assemblages persisted until at least
about 40,000cal. 
1,5,6
. In Romania, the Petera cu Oase H.sapiens
L3
M
N
R
U
Georgian
Ket
Chinese 2
Australian 2
Chukchi
Australian 1
Mixteca Baja
Malay
Piman
Spain
Japan 2
Warao 1
Ewondo
Bacho Kiro CC7-2289
rCRS
Bamileke
Ust'Ishim*
Onge 1
Fumane 2*
English
Oase 1*
Kostenki 14*
BS11*
French
Dolni Vestonice 13*
Warao 2
Korea
Tatar
PNG high 1
Dolni Vestonice 14*
Saami
Native American 1
Dutch
Italian
SibInuit
Native American 2
Oberkassel 998*
PNG high 2
Bacho Kiro BK-1653
Bacho Kiro molar F6-620
Iceman*
Tianyuan*
Onge 2
Uzbek
Palaeo-Eskimo Saqqaq*
Lisongo
Yoruba
Bacho Kiro AA7-738
Great Andamanese 2
Buriat
Loschbour*
Great Andamanese 1
PNG coast
Chinese 1
Guarani
Japan 1
Bacho Kiro CC7-335
Bacho Kiro BB7-240
Neanderthals
Denisovans
Africans
Eurasians
26
6
5
8
4
4
2
2
10
5
9
4
3
0
5
2
11
7
2
0
1
09
11
2
2
1
10
57
2
12
1
4
4
12
4
3
2
11
1
2
7
3
2
3
1
12
5
3
0
1
0
3
8
1
0
4
3
6
0
4
0
3
2
9
5
3
5
9
8
6
3
4
2
7
6
8
14
0
1
8
1
0
8
15
0
2
5
1
1
3
7
6
4
20
1
8
0
10
2
6
2
1
6
19
4
0
6
8
1
1
3
Sima de los Huesos
Fig. 2 | Maxi mum parsimo ny tree. Maxi mum parsimony tre e relating Bach o
Kiro Cave mtDNA s to 54present-day hum ans, 12ancient H.sapiens,
22Neander thals, 4Denisova ns and 1individual f rom Sima de los Hue sos. The
insert s hows the part of th e tree closest t o the mtDNAs of th e specimens f rom
Bacho Kiro C ave. Bacho Kiro Cave mt DNAs are red. A sterisks de note mtDNA
from ancie nt H.sapiens (Supp lementar y Table9) other than the B acho Kiro
Cave specime ns. The numb er of inferred sub stitutions p ersequence is gi ven
above each bra nch. A chimpanze e mtDNA sequ ence was used to r oot the tree
(not shown). rCRS, r evised Cambr idge Reference Se quence. U, R, N, M a nd L3
refer to the mitochondrial haplogroups.
q
a
kl
m
n
o
p
bcde f
ghij
Fig. 3 | Bone t ools and per sonal ornam ents from Ba cho Kiro Cave layersI
and J (Nich e1 and Main sec tors). a–j, Pendants mad e from perforate d and
grooved tee th (a, ungulate; b–j, cave bear). k, l, o, Awls. m, Anthro pogenically
modifi ed piece. n, p, Lissoirs. q, Ivory bead . Further detai ls are provided in
Supplemen tary Table15. Scal e bars, 1cm (a–o, q), 3cm (p).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

4 | Nature | www.nature.com
Article
individual had a Neanderthal ancestor as recently as four-to-six genera-
tions back in his family tree
36
. In light of the Bacho Kiro Cave results,
the 42,000–37,000cal. yr  radiocarbon age of the Petera cu Oase
fossil implies an extended period of contact between Neanderthals and
H.sapiens in eastern Europe. Alternatively, it may be that the direct date
of Petera cu Oase—which was obtained before recent improvements in
pretreatment techniques—is an underestimate, and that local coexist-
ence was more ancient and ephemeral. The IUP pendants of Bacho Kiro
Cave (Fig.3) are notably similar to artefacts produced by late Nean-
derthals of the Châtelperronian layers at Grotte du Renne (France)
4
.
Whatever the cognitive complexity of the last Neanderthals might
have been, the earlier age of the Bacho Kiro Cave material supports the
notion that these specific behavioural novelties seen in declining Nean-
derthal populations resulted from contacts with migrant H.sapiens7.
Online content
Any methods, additional references, Nature Research reporting sum-
maries, source data, extended data, supplementary information,
acknowledgements, peer review information; details of author con-
tributions and competing interests; and statements of data and code
availability are available at https://doi.org/10.1038/s41586-020-2259-z.
1. Hublin, J.-J. The modern human colonization of western Eurasia: when and where? Quat.
Sci. Rev. 118, 194–210 (2015).
2. Kuhn, S. L. & Zwyns, N. Rethinking the initial Upper Paleolithic. Quat. Int. 347, 29–38 (2014).
3. Fewlass, H. etal. A 14C chronology for the Middle to Upper Palaeolithic transition at Bacho
Kiro cave, Bulgaria. Nat. Ecol. Evol. https://doi.org/10.1038/s41559-020-1136-3 (2020).
4. White, R. Personal ornaments from the Grotte du Renne at Arcy-sur-Cure. Athena Review
2, 41–46 (2001).
5. Welker, F. etal. Palaeoproteomic evidence identiies archaic hominins associated with the
Châtelperronian at the Grotte du Renne. Proc. Natl Acad. Sci. USA 113, 11162–11167 (2016).
6. Hublin, J.-J. etal. Radiocarbon dates from the Grotte du Renne and Saint-Césaire support
a Neandertal origin for the Châtelperronian. Proc. Natl Acad. Sci. USA 109, 18743–18748
(2012).
7. Hublin, J.-J., Spoor, F., Braun, M., Zonneveld, F. & Condemi, S. A late Neanderthal
associated with Upper Palaeolithic artefacts. Nature 381, 224–226 (1996).
8. Ruebens, K., McPherron, S. J. P. & Hublin, J.-J. On the local Mousterian origin of the
Châtelperronian: integrating typo-technological, chronostratigraphic and contextual
data. J. Hum. Evol. 86, 55–91 (2015).
9. Higham, T. etal. The earliest evidence for anatomically modern humans in northwestern
Europe. Nature 479, 521–524 (2011).
10. Benazzi, S. etal. Early dispersal of modern humans in Europe and implications for
Neanderthal behaviour. Nature 479, 525–528 (2011).
11. White, M. & Pettitt, P. Ancient digs and modern myths: the age and context of the Kent’s
Cavern 4 maxilla and the earliest Homo sapiens specimens in Europe. Eur. J. Archaeol. 15,
392–420 (2012).
12. Zilhão, J., Banks, W. E., d’Errico, F. & Gioia, P. Analysis of site formation and assemblage
integrity does not support attribution of the Uluzzian to modern humans at Grotta del
Cavallo. PLoS ONE 10, e0131181 (2015).
13. Kozłowski, J. K. Excavation in the Bacho Kiro Cave (Bulgaria): Final Report. 172 (Panstwowe
Wydawnictwo Naukowe, 1982).
14. Hedges, R. E. M., Housley, R. A., Bronk Ramsey, C. & Klinken, G. J. V. Radiocarbon dates
from the Oxford AMS system: archaeometry datelist 18. Archaeometry 36, 337–374
(1994).
15. Tsanova, T. & Bordes, J. G. in The Humanized Mineral World: Towards Social and Symbolic
Evaluation of Prehistoric Technologies in South Eastern Europe (Proceedings of the ESF
Workshop) (eds Tsonev, T. S. & Montagnari Kokclj, E.) 41–50 (ERAUL, 2003).
16. Bailey, S. E. A closer look at Neanderthal postcanine dental morphology: the mandibular
dentition. Anat. Rec. 269, 148–156 (2002).
17. Bailey, S. E., Skinner, M. M. & Hublin, J.-J. What lies beneath? An evaluation of lower molar
trigonid crest patterns based on both dentine and enamel expression. Am. J. Phys.
Anthropol. 145, 505–518 (2011).
18. Keith, A. Problems relating to the teeth of the earlier forms of prehistoric man. Proc. R.
Soc. Med. 6, 103–124 (1913).
19. Shaw, J. The Teeth, the Bony Palate and the Mandible in Bantu Races of South Africa (Bale
and Danielsson, London, 1938).
20. Kallay, J. in Dental Anthropology (ed. Brothwell, D.) 75–86 (Pergamon, 1963).
21. Buckley, M., Collins, M., Thomas-Oates, J. & Wilson, J. C. Species identiication by analysis
of bone collagen using matrix-assisted laser desorption/ionisation time-of-light mass
spectrometry. Rapid Commun. Mass Spectrom. 23, 3843–3854 (2009).
22. Dabney, J. etal. Complete mitochondrial genome sequence of a Middle Pleistocene cave
bear reconstructed from ultrashort DNA fragments. Proc. Natl Acad. Sci. USA 110,
15758–15763 (2013).
23. Korlević, P. etal. Reducing microbial and human contamination in DNA extractions from
ancient bones and teeth. Biotechniques 59, 87–93 (2015).
24. Gansauge, M.-T. etal. Single-stranded DNA library preparation from highly degraded DNA
using T4 DNA ligase. Nucleic Acids Res. 45, e79 (2017).
25. Fu, Q. etal. DNA analysis of an early modern human from Tianyuan Cave, China. Proc.
Natl Acad. Sci. USA 110, 2223–2227 (2013).
26. Lippold, S. etal. Human paternal and maternal demographic histories: insights from
high-resolution Y chromosome and mtDNA sequences. Investig. Genet. 5, 13 (2014).
27. Kivisild, T. Maternal ancestry and population history from whole mitochondrial genomes.
Investig. Genet. 6, 3 (2015).
28. Fu, Q. etal. Genome sequence of a 45,000-year-old modern human from western
Siberia. Nature 514, 445–449 (2014).
29. Fu, Q. etal. A revised timescale for human evolution based on ancient mitochondrial
genomes. Curr. Biol. 23, 553–559 (2013).
30. vander Made, J. in The Encyclopedia of Archaeological Sciences (ed. López Varela, S. L.)
1–4 (Wiley-Blackwell, 2018).
31. Guérin, C. Première biozonation du Pléistocène Européen, principal résultat
biostratigraphique de l’étude des Rhinocerotidae (Mammalia, Perissodactyla) du Miocène
terminal au Pléistocène supérieur d’Europe Occidentale. Geobios 15, 593–598 (1982).
32. Kuhn, S. L. etal. The early Upper Paleolithic occupations at Uçağizli Cave (Hatay, Turkey).
J. Hum. Evol. 56, 87–113 (2009).
33. Kuhn, S. L. Upper Paleolithic raw material economies at Üçagizh cave, Turkey. J.
Anthropol. Archaeol. 23, 431–448 (2004).
34. Müller, U. C. etal. The role of climate in the spread of modern humans into Europe. Quat.
Sci. Rev. 30, 273–279 (2011).
35. Hershkovitz, I. etal. Levantine cranium from Manot Cave (Israel) foreshadows the irst
European modern humans. Nature 520, 216–219 (2015).
36. Fu, Q. etal. An early modern human from Romania with a recent Neanderthal ancestor.
Nature 524, 216–219 (2015).
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional afiliations.
© The Author(s), under exclusive licence to Springer Nature Limited 2020
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Methods
No statistical methods were used to predetermine sample size.
Excavation methods
The site was excavated closely following existing protocols
37–42
. Layers
were defined first on lithological criteria, and second on archaeological
criteria. Our stratigraphy was determined and named independently of
previous excavations13. Additionally, we excavated in two unconnected
areas of the site, and thus separate naming conventions were used
between these two areas. The excavations in the south area (Extended
Data Fig.1) are known as the Main sector, and the layers are named with
letters for the large divisions and numbers for divisions within these
(for example, layerI or layerA0). The other area excavated is a niche
to the east of the previous excavations. We call this area Niche1 (or
N1), and all layer names from this area are prefixed with N1-. Where we
hypothesize a link between the two sectors, we use the same layer name
(for example, layers N1-I and I). Where we are unable to form a strong
hypothesis about the link between the two sectors, we use different
layer names (numbers in this case) in the Niche1 sector to denote this
(for example, N1-3), followed by letter for internal subdivisions (for
example, N1-3a). All finds were recorded by layer and 3D coordinates
(using an arbitrary grid established for the excavation and aligned to
the previous excavations) measured with Leica total stations (5 accu-
racy) using data collectors with self-authored software (EDM-Mobile).
All lithics and fauna >20mm in length and all specialists’ samples (for
example, ancient DNA, micromorphology, phytoliths and so on) were
provenienced and given unique identifiers (IDs). Complete bones,
identifiable teeth and human remains <20 mm in length (but larger
than microfauna) were also given coordinates and IDs. Natural stones
>10 cm in length were recorded with a single coordinate, and stones
>20 cm in length were measured with multiple coordinates to describe
their volume and orientation. The sediment, excluding recorded stones
and artefacts, was collected by 9-l buckets and wet-screened on-site
through 6- and 1.2-mm meshes to form two fractions. Buckets have
unique IDs. Their coordinates were measured first in the centre of the
area to be excavated and then again at the centre of the area excavated
at the completion of the bucket. Large (>20-mm-long) objects found in
the sediment in the buckets during wet-screening were given IDs and
assigned the coordinates of the bucket. All features were provenienced.
Digital photographs documenting the excavation were recorded daily,
and final sections were documented through a combination of digi-
tal photography, drawing, and total station measures. Additionally,
structure-from-motion models were made of all final sections and
excavation areas. These models were georeferenced to the excavation
grid using total station coordinates.
ZooMS
We screened 1,271fragmentary bone and tooth specimens from Bacho
Kiro Cave using ZooMS
21
. Eleven bone specimens were derived from
the previous excavations at the site13, 371 from our excavations in
the Main sector, and 889 bone specimens from the Niche 1 area. We
particularly focused on IUP layersI and N1-I (n=822). Extraction and
analytical protocols followed previously published work5. In brief,
a small bone sample (<20 mg) was taken from each bone or dentine
specimen. The sample was incubated at 65°C for an hour in 50 mM
ammonium-bicarbonate buffer, digested overnight using trypsin (Pro-
mega) at 37°C, acidified using 20% TFA, and cleaned on C18 ZipTips
(either from Sigma-Aldrich or Thermo Scientific). MALDI–TOF MS anal-
ysis was conducted at the IZI Fraunhofer in Leipzig43. MALDI–TOF MS
spectra were analysed in comparison to a reference database containing
collagen-peptide marker masses of all medium- to larger-sized genera
in existence in western Eurasia during the Late Pleistocene epoch5. In
cases in which ammonium-bicarbonate extraction failed, an attempt
was made to recover further informative collagen peptides through acid
demineralization of the same bone sample, as previously explained5.
Collagen deamidation in these spectra was assessed for two peptides
(P1105 and P1706)44,45.
Bone pretreatment and accelerator mass spectrometry dating
Small aliquots (80–110 mg) of the six ZooMS-identified hominin bone
fragments were sampled for dating to preserve as much material as
possible for further analyses. Collagen was extracted using a previously
described technique
46
for small bone sample sizes, based on a modified
Longin collagen-extraction protocol
47
followed by an ultrafiltration
step48. In brief, the outer surfaces of the bone samples were removed
with a sandblaster, and samples were removed using a rotary tool.
The bones were demineralized in 0.5 M HCl at 4°C until soft and CO
2
effervescence had stopped. Then, 0.1 M NaOH was added for 10 min at
room temperature to remove humic acid contamination, and samples
were re-acidified in 0.5M HCl. The collagen was gelatinized in acidic
water (HCl pH 3) at 70°C for several hours (4–6 h). The collagen sam-
ples were then passed through an Ezee Filter (Elkay Laboratories) to
remove large particles (>80 µm) and separated by molecular weight
with pre-cleaned Sartorius VivaSpin Turbo 15 ultrafilters (30kDa molec-
ular weight cut-off (MWCO))49,50. The samples were freeze-dried and the
large molecular fraction (>30 kDa) was graphitized using Automated
Graphitisation Equipment III
51
and measured using the latest model
of the MICADAS accelerator mass spectrometer
52
in the Laboratory
of Ion Beam Physics at ETH-Zurich (laboratory code ETH). Small ali
-
quots (66–89 mg) of a background cave bear bone (>50,000yr )
were extracted alongside the samples to monitor contamination intro-
duced in the laboratory53. These were measured in the same magazine
as the hominin samples and used in the age calculation. Oxalic acid
II standards were also measured in the same magazine and used for
normalization. Data reduction was performed using BATS software54.
An additional 1‰ was added to the error calculation of the samples, as
per standard practice. The dates were calibrated using the IntCal13
55
dataset in OxCal v.4.356.
Shape analysis of the molar enamel–dentine junction
Enamel and dentine tissues (Extended Data Fig.5) of lower second
molars were segmented using the 3D voxel value histogram and its dis-
tribution of greyscale values
57,58
. After segmentation, the enamel–den-
tine junction was reconstructed as a triangle-based surface model using
Avizo. Small enamel–dentine junction defects were corrected digitally
using the ‘fill holes’ module of Geomagic Studio. We then used Avizo
to digitize 3D landmarks and curve-semilandmarks on the enamel–
dentine junction surface57,58. Anatomical landmarks were placed on
the tip of the dentine horn of the protoconid, metaconid, entoconid
and hypoconid. A sequence of landmarks was also placed along the
marginal ridge connecting the dentine horns, beginning at the top of
the protoconid and moving in lingual direction; the points along this
ridge curve were then later resampled to the same point count on every
specimen using Mathematica. Likewise, we digitized and resampled a
curve along the cemento–enamel junction as a closed curve starting
and ending below the protoconid horn and the mesiobuccal corner
of the cervix. The resampled points along the two ridge curves were
subsequently treated as sliding curve semilandmarks and analysed
using geometric morphometrics together with the four anatomical
landmarks. Landmarks not preserved on the Bacho Kiro Cave specimen
were removed before principal component analysis. The specimens of
Homo erectus sensu lato include KNM-ER 1802, KNM-ER 992 and San-
giran 1b. Specimens of archaic Middle Pleistocene hominins include
Balanica 1, Mauer, Xiahe and KNM-ER BK 67. The Neanderthal sample
includes Abri Suard S36, Krapina 1, 6, 9, 53, 54, 55, 57, 59, 80, 86, 105 and
107, La Quina H9, Le Moustier 1, Regourdou, Scladina 4A1, El Sidron 540
and 755, and Vindija 11.39. The fossil H.sapiens sample includes Dar es
Soltane II H4, El Harhoura, Jebel Irhoud 3 and 11, Qafzeh 9, 10, 11 and 15,
and Temara. The recent H.sapiens sample includes clinical extractions
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Article
from dentists based in Germany, Neolithic specimens from Belgium
(Royal Belgian Institute of Natural Sciences) and specimens from the
Francisc J. Rainer Collection (Institutul de Antropologie ‘Francisc J.
Rainer’).
DNA extraction and library preparation
Samples of between 29.3 mg and 64.7 mg of tooth or bone powder were
removed from 7 Bacho Kiro Cave specimens (F6-620, AA7-738, BB7-
240, CC7-2289, CC7-335, F6-597 and BK-1653) using a sterile dentistry
drill (Supplementary Table4) after a thin layer of surface was removed
from the sampling areas. DNA was extracted from the powder using a
silica-based method22 as previously described23. Five single-stranded
DNA libraries24 were made from 10 µl from each extract on an automated
liquid handling platform (Bravo NGS workstation B, Agilent Technol-
ogies)59. A control oligonucleotide was spiked into each reaction to
determine the efficiency of library preparation
60
, and quantitative PCR
was used to determine the total number of unique library molecules
as well as the number of oligonucleotides that were successfully con-
verted
24,60
. The libraries were amplified into plateau with AccuPrime
Pfx DNA polymerase (Life Technologies)
61
and labelled with two unique
indices
23,62
. Half of the volume of the amplified libraries (50 µl) was puri-
fied using SPRI beads on the automated liquid handling platform
59
. The
concentrations of the purified DNA libraries were determined using a
NanoDrop Spectrophotometer (NanoDrop Technologies).
mtDNA capture and sequencing
An aliquot of each amplified library was enriched for human mtDNA
using a bead-based hybridization method29. Enriched libraries were
sequenced on an Illumina MiSeq platform in a double index configura-
tion (2×76 cycles)
62
and base-calling was done using Bustard (Illumina).
Overlapping paired-end reads were merged into single sequences and
the adapters were trimmed using leeHom
63
. The Burrows–Wheeler
Aligner (BWA, version: 0.5.10-evan.9-1-g44db244; https://github.
com/mpieva/network-aware-bwa)64, with parameters adjusted for
ancient DNA (‘-n 0.01 –o 2 –l 16500’)
65
, was used to align the data to the
revised Cambridge Reference Sequence (NC_01290). Only reads with
perfect matches to the expected index combinations were retained for
downstream analyses. PCR duplicates were removed using bam-rmdup
(version 0.6.3; https://bitbucket.org/ustenzel/biohazard). SAMtools
(version 1.3.1)66 was used to filter for fragments that were longer than
35base pairs and that had a mapping quality of at least 25. We merged
the libraries originating from the same extract (that is, the same speci-
men) using SAMtools merge to produce the final dataset.
Phylogenetic inferences
We reconstructed the mitochondrial genomes of the Bacho Kiro Cave
specimens once by using all mapped fragments longer than 35 base
pairs with a mapping quality of at least 25 and once using only fragments
with a cytosine (C) to thymine (T) difference to the reference genome at
the first three and/or last three terminal positions
36
(that is, putatively
deaminated fragments). We called a consensus base at each position
along the mtDNA that was covered by at least 3 DNA fragments and at
which at least 2/3 of fragments carried an identical base and the base
quality was 20 or higher67. To prevent deamination-induced substitu-
tions affecting the calling of a consensus base, we converted A on the
reverse strands and T on the forward strands in the first three and the
last three positions of a fragment into N.
The libraries prepared from the F6-597 specimen yielded too few
informative mtDNA fragments to reconstruct a complete mtDNA
using putatively deaminated fragments. We investigated the state of
F6-597 DNA fragments that overlapped positions ‘diagnostic’ for each
branch in a mtDNA tree relating present-day humans, Neanderthals,
Denisovans and the hominin from Sima de los Huesos68 (Supplementary
Table6). To diminish the influence of substitutions derived from deami-
nation, all forward strands were ignored if one of the possible states
at an informative state was a C and all reverse strands were ignored if
one of the possible states was G.
We aligned the reconstructed mitochondrial genomes of the Bacho
Kiro Cave individuals to the mtDNA genomes of 54 present-day humans
from a wide geographical distribution
69
, 12 ancient H.sapiens
25,28,41,70–73
(Supplementary Table9), 22 Neanderthals
69,74–78
, 4Denisovans
79–82
,
a Sima de los Huesos individual
67
and a chimpanzee
83
using MAFFT
v.7.27184. The number of pairwise differences among the genomes was
calculated using MEGA785 and a maximum parsimony tree was recon-
structed using Parsimony ratchet as implemented in the R package
phangorn
86
. We identified the haplogroup of each of the reconstructed
mitochondrial genomes with HaploGrep87 based on the PhyloTree
database (PhyloTree.org, build 17).
Contamination estimates
We used two complementary approaches to estimate levels of
present-day human mtDNA contamination in the libraries. We iden
-
tified positions at which each of the reconstructed Bacho Kiro Cave
mtDNAs differ from at least 99% of a world-wide panel of 311 present-day
human mtDNAs36,69 (Supplementary Tables7, 8). We then counted
DNA fragments that overlap these positions and did not match the
consensus base of the respective specimen, again taking into account
the strand orientation in cases in which one of the possible states at an
informative site was C or G. In the second approach, we used an iterative
probabilistic method, schmutzi88, which uses a nonredundant database
of human mitochondrial genomes to estimate levels of present-day
human DNA contamination (Supplementary Information section5)
(parameters: ‘--notusepredC --uselength’).
Molecular DNA dating
We estimated the tip dates of the reconstructed Bacho Kiro Cave
mtDNAs using the Bayesian phylogenetic method as implemented in
BEAST2 (version 2.4.8)
89
by aligning the reconstructed mitochondrial
genomes to 54 present-day humans and 10 directly radiocarbon-dated
ancient H.sapiens
28,29,72,73,80
, which were used for tip calibration. The
Neanderthal mtDNA genome of Vindija 33.16
69
was used as an outgroup.
The best-fitting substitution model was determined using jModel-
Test2
90
. We investigated a strict clock and an uncorrelated log-normal
relaxed clock as two models of rate variation and a constant population
size and a Bayesian skyline as tree priors28. For each model, we carried
out Markov chain Monte Carlo runs with 30,000,000 iterations and
sampling every 1,000 steps. After discarding 10% of the iterations as
burn-in, the output was analysed with Tracer v.1.5.0 (http://tree.bio.
ed.ac.uk/software/tracer/). A marginal likelihood estimation
91
analysis
was used for model comparison and best support assessment. Both
the maximum parsimony and the BEAST2 tree were visualized with
FigTree (version v.1.4.2) (http://tree.bio.ed.ac.uk/software/figtree/).
Micromorphology
Field observations of the sediments were complemented by archaeo-
logical micromorphology analyses. Micromorphological samples were
collected as undisturbed blocks by carefully carving and wrapping them
with either pre-plastered bandages or soft paper and tape. Thin sections
were manufactured by Spectrum Petrographics through a standard
procedure of drying the blocks in an oven for several days at about
60°C. The blocks were then impregnated with a mixture of polyester
resin and styrene, to which a catalyst was added. Thin sections were
ground to a thickness of 30 µm and observed under a petrographic
microscope in plane- and cross-polarized light at magnifications rang-
ing from 20× to 400×. Micromorphological nomenclature follows
previously published work92,93.
Reporting summary
Further information on research design is available in theNature
Research Reporting Summary linked to this paper.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Data availability
The data that support the findings of this study are available from the
corresponding author upon reasonable request. Genetic sequence
reads from all libraries and corresponding negative controls are depos-
ited at European Nucleotide Archive under the study accession number
PRJEB35466. The FASTA files of the mitochondrial genomes are depos-
ited in GenBank with the accession numbers MN706602–MN706607.
Details are as follows: Bacho Kiro AA7-738, MN706602; Bacho Kiro
BB7-240, MN706603; Bacho Kiro BK-1653, MN706604; Bacho Kiro
CC7-335, MN706605; Bacho Kiro CC7-2289, MN706606; and Bacho
Kiro molar F6-620, MN706607.
37. Dibble, H. L. & Lenoir, M. The Middle Paleolithic Site of Combe-Capelle Bas (France) (The
University Museum Press, 1995).
38. Turq, A. etal. in Les Sociétés du Paléolithique dans un Grand Sud-ouest dela France:
Nouveaux Gisements, Nouveaux Résultats, Nouvelles Méthodes (eds. Jaubert, J. etal.)
83–94 (Mémoire dela Société Préhistorique Française, 2008).
39. Chase, P. G., Debénath, A., Dibble, H. L. & McPherron, S. P. in The Cave of Fontéchevade:
Recent Excavations and their Paleoanthropological Implications (eds Chase, P. G. etal.)
28–62 (Cambridge Univ. Press, 2009).
40. Soressi, M. etal. Neandertals made the irst specialized bone tools in Europe. Proc. Natl
Acad. Sci. USA 110, 14186–14190 (2013).
41. Richter, D. etal. The age of the hominin fossils from Jebel Irhoud, Morocco, and the
origins of the Middle Stone Age. Nature 546, 293–296 (2017).
42. Sandgathe, D. M., Dibble, H. L., McPherron, S. J. P. & Goldberg, P. in The Middle Paleolithic
Site of Pech de l’Azé IV Cave and Karst Systems of the World (eds Dibble, H. L. etal.) 1–19
(Springer, 2018).
43. Sinet-Mathiot, V. etal. Combining ZooMS and zooarchaeology to study Late Pleistocene
hominin behaviour at Fumane (Italy). Sci. Rep. 9, 12350 (2019).
44. Wilson, J., van Doorn, N. L. & Collins, M. J. Assessing the extent of bone degradation using
glutamine deamidation in collagen. Anal. Chem. 84, 9041–9048 (2012).
45. Welker, F. etal. Variations in glutamine deamidation for a Châtelperronian bone
assemblage as measured by peptide mass ingerprinting of collagen. Sci. Technol.
Archaeol. Res. 3, 15–27 (2017).
46. Fewlass, H. etal. Pretreatment and gaseous radiocarbon dating of 40–100 mg
archaeological bone. Sci. Rep. 9, 5342 (2019).
47. Longin, R. New method of collagen extraction for radiocarbon dating. Nature 230,
241–242 (1971).
48. Brown, T. A., Nelson, D. E., Vogel, J. S. & Southon, J. R. Improved collagen extraction by
modiied Longin method. Radiocarbon 30, 171–177 (1988).
49. Bronk Ramsey, C., Higham, T., Bowles, A. & Hedges, R. Improvements to the pretreatment
of bone at Oxford. Radiocarbon 46, 155–163 (2004).
50. Brock, F., Bronk Ramsey, C. & Higham, T. Quality assurance of ultrailtered bone dating.
Radiocarbon 49, 187–192 (2007).
51. Wacker, L., Němec, M. & Bourquin, J. A revolutionary graphitisation system: fully
automated, compact and simple. Nucl. Instrum. Methods Phys. Res. B 268, 931–934
(2010).
52. Wacker, L. etal. MICADAS: routine and high-precision radiocarbon dating. Radiocarbon
52, 252–262 (2010).
53. Korlević, P., Talamo, S. & Meyer, M. A combined method for DNA analysis and radiocarbon
dating from a single sample. Sci. Rep. 8, 4127 (2018).
54. Wacker, L., Christl, M. & Synal, H. A. Bats: a new tool for AMS data reduction. Nucl.
Instrum. Methods Phys. Res. B 268, 976–979 (2010).
55. Reimer, P. J. etal. IntCal13 and Marine13 radiocarbon age calibration curves 0–50,000
years cal BP. Radiocarbon 55, 1869–1887 (2013).
56. Bronk Ramsey, C. Bayesian analysis of radiocarbon dates. Radiocarbon 51, 337–360
(2009).
57. Skinner, M. M., Gunz, P., Wood, B. A. & Hublin, J.-J. Enamel–dentine junction (EDJ)
morphology distinguishes the lower molars of Australopithecus africanus and
Paranthropus robustus. J. Hum. Evol. 55, 979–988 (2008).
58. Skinner, M. M., Gunz, P., Wood, B. A., Boesch, C. & Hublin, J.-J. Discrimination of extant
Pan species and subspecies using the enamel–dentine junction morphology of lower
molars. Am. J. Phys. Anthropol. 140, 234–243 (2009).
59. Slon, V. etal. Neandertal and Denisovan DNA from Pleistocene sediments. Science 356,
605–608 (2017).
60. Glocke, I. & Meyer, M. Extending the spectrum of DNA sequences retrieved from ancient
bones and teeth. Genome Res. 27, 1230–1237 (2017).
61. Dabney, J. & Meyer, M. Length and GC-biases during sequencing library ampliication: a
comparison of various polymerase-buffer systems with ancient and modern DNA
sequencing libraries. Biotechniques 52, 87–94 (2012).
62. Kircher, M., Sawyer, S. & Meyer, M. Double indexing overcomes inaccuracies in multiplex
sequencing on the Illumina platform. Nucleic Acids Res. 40, e3 (2012).
63. Renaud, G., Stenzel, U. & Kelso, J. leeHom: adaptor trimming and merging for Illumina
sequencing reads. Nucleic Acids Res. 42, e141 (2014).
64. Li, H. & Durbin, R. Fast and accurate long-read alignment with Burrows–Wheeler
transform. Bioinformatics 26, 589–595 (2010).
65. Meyer, M. etal. A high-coverage genome sequence from an archaic Denisovan individual.
Science 338, 222–226 (2012).
66. Li, H. etal. The Sequence Alignment/Map format and SAMtools. Bioinformatics 25,
2078–2079 (2009).
67. Meyer, M. etal. A mitochondrial genome sequence of a hominin from Sima de los
Huesos. Nature 505, 403–406 (2014).
68. Meyer, M. etal. Nuclear DNA sequences from the Middle Pleistocene Sima de los Huesos
hominins. Nature 531, 504–507 (2016).
69. Green, R. E. etal. A complete Neandertal mitochondrial genome sequence determined
by high-throughput sequencing. Cell 134, 416–426 (2008).
70. Benazzi, S. etal. The makers of the Protoaurignacian and implications for Neandertal
extinction. Science 348, 793–796 (2015).
71. Ermini, L. etal. Complete mitochondrial genome sequence of the Tyrolean Iceman. Curr.
Biol. 18, 1687–1693 (2008).
72. Gilbert, M. T. P. etal. Paleo-Eskimo mtDNA genome reveals matrilineal discontinuity in
Greenland. Science 320, 1787–1789 (2008).
73. Krause, J. etal. A complete mtDNA genome of an early modern human from Kostenki,
Russia. Curr. Biol. 20, 231–236 (2010).
74. Green, R. E. etal. A draft sequence of the Neandertal genome. Science 328, 710–722 (2010).
75. Posth, C. etal. Deeply divergent archaic mitochondrial genome provides lower time
boundary for African gene low into Neanderthals. Nat. Commun. 8, 16046 (2017).
76. Prüfer, K. etal. The complete genome sequence of a Neanderthal from the Altai
Mountains. Nature 505, 43–49 (2014).
77. Rougier, H. etal. Neandertal cannibalism and Neandertal bones used as tools in Northern
Europe. Sci. Rep. 6, 29005 (2016).
78. Skoglund, P. etal. Separating endogenous ancient DNA from modern day contamination
in a Siberian Neandertal. Proc. Natl Acad. Sci. USA 111, 2229–2234 (2014).
79. Krause, J. etal. The complete mitochondrial DNA genome of an unknown hominin from
southern Siberia. Nature 464, 894–897 (2010).
80. Reich, D. etal. Genetic history of an archaic hominin group from Denisova Cave in Siberia.
Nature 468, 1053–1060 (2010).
81. Sawyer, S. etal. Nuclear and mitochondrial DNA sequences from two Denisovan
individuals. Proc. Natl Acad. Sci. USA 112, 15696–15700 (2015).
82. Slon, V. etal. A fourth Denisovan individual. Sci. Adv. 3, e1700186 (2017).
83. Horai, S. etal. Man’s place in Hominoidea revealed by mitochondrial DNA genealogy.
J. Mol. Evol. 35, 32–43 (1992).
84. Katoh, K. & Standley, D. M. MAFFT multiple sequence alignment software version 7:
improvements in performance and usability. Mol. Biol. Evol. 30, 772–780 (2013).
85. Kumar, S., Stecher, G. & Tamura, K. MEGA7: molecular evolutionary genetics analysis
version 7.0 for bigger datasets. Mol. Biol. Evol. 33, 1870–1874 (2016).
86. Schliep, K. P. phangorn: phylogenetic analysis in R. Bioinformatics 27, 592–593 (2011).
87. Kloss-Brandstätter, A. etal. HaploGrep: a fast and reliable algorithm for automatic
classiication of mitochondrial DNA haplogroups. Hum. Mutat. 32, 25–32 (2011).
88. Renaud, G., Slon, V., Duggan, A. T. & Kelso, J. Schmutzi: estimation of contamination and
endogenous mitochondrial consensus calling for ancient DNA. Genome Biol. 16, 224
(2015).
89. Bouckaert, R. etal. BEAST 2: a software platform for Bayesian evolutionary analysis. PLOS
Comput. Biol. 10, e1003537 (2014).
90. Darriba, D., Taboada, G. L., Doallo, R. & Posada, D. jModelTest 2: more models, new
heuristics and parallel computing. Nat. Methods 9, 772 (2012).
91. Baele, G. etal. Improving the accuracy of demographic and molecular clock model
comparison while accommodating phylogenetic uncertainty. Mol. Biol. Evol. 29,
2157–2167 (2012).
92. Stoops, G. Guidelines for Analysis and Description of Soil and Regolith Thin Sections. 184
(Soil Science Society of America, 2003).
93. Courty, M. A., Goldberg, P. & Macphail, R. Soils and Micromorphology in Archaeology 344
(Cambridge Univ. Press, 1989).
Acknowledgements We thank the tourism association of Bacho Kiro Cave in the town of
Dryanovo, the History museum – Dryanovo, the Regional History museum in the city of Gabrovo,
Dryanovo town hall and V. Lafchiiski for their assistance with the ieldwork and in the laboratory;
N. Spassov from the National Museum of Natural History in Soia for cooperating and hosting
researchers of our project; H. Temming and J. Honeyford for their technical assistance and S.
Nagel, B. Nickel, B. Schellbach and A. Weihmann for their help with the ancient DNA laboratory
procedures and sequencing. Field operations were funded by the Max Planck Society.
AixMICADAS and its operation are funded by Collège de France and the EQUIPEX
ASTER-CEREGE (principal investigator, E.B.). S.T. is funded by the European Research Council
under the European Union’s Horizon 2020 Research and Innovation Programme (grant
agreement no. 803147-951 RESOLUTION). The ancient DNA part of this study was funded by the
Max Planck Society and the European Research Council (grant agreement no. 694707 to S.P.).
Author contributions J.-J.H. designed the study. T. Tsanova, N.S., V.A., S.S., R.S., E.E., Z.R. and
S.P.M. collected ield data; H.F., B.K., L.W., E.B., Y.F., T. Tuna and S.T. established the radiocarbon
dates; V.A. studied the micromorphology of the sediments; S.B., M.M.S. and J.-J.H. analysed
hominin dental morphology; V.S.-M., L.P., F.W. and A.W. performed ZooMS; M.H., M.M. and S.P.
performed mtDNA analysis; T. Tsanova, N.S., N.Z., S.S., I.K., V.D., J.M. and S.P.M. conducted the
study of the lithics; G.M.S., R.S., V.P. and N.L.M. analysed the faunal assemblages and the
osseous objects. J.-J.H. wrote the paper with contributions of all authors.
Competing interests The authors declare no competing interests.
Additional information
Supplementary information is available for this paper at https://doi.org/10.1038/s41586-020-
2259-z.
Correspondence and requests for materials should be addressed to J.-J.H.
Peer review information Nature thanks William Banks, Richard G. Klein and the other,
anonymous, reviewer(s) for their contribution to the peer review of this work.
Reprints and permissions information is available at http://www.nature.com/reprints.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Article
Extende d Data Fig. 1 | Excava tions at Bach o Kiro Cave, 2015–2018. a, Plan
view of the en trance and the excav ated areas of the c ave, with the gri d system of
our recent exc avations (let ters in the lef t column) and those of t he 1971–1975
excavations ( letters in the r ight column). b, Site loc ation in south eastern
Europe. c, Photo graph of the entr ance of the cave. T he floor is ar tificia lly
raised; the o riginal entra nce was several me tres lower than show n in this
photograph. d, Initial str atigraphic s ection drawi ng of the expose d profile
from the Main s ector in 201 5 (codes for the archae ological layers ar e on the left,
with the cor responding l ayers from the 1971–1975 excavations in p arenthese s).
e, Frontal view of t he Niche1 sector a nd its stratig raphic subdiv isions. f, Lower
part of the s tratigraphi c section draw ing of the Niche1 sec tor, in 2018. Note the
thickne ss and preser vation of the lower d eposits here i n comparison w ith the
Main sector profile. g, Photo graph of the Mai n sector tran sversal sect ion on the
line betwe en squares F5–F6 and sq uares G5–G6 be fore excavation in 201 5. CF,
combustion feature. h–n, Hominin remain s identifie d by ZooMS with th eir IDs:
BK-1653 (h) and F6-597 (j) f rom layerB, with h comin g from the 1971–1975
excavations (dash ed line); BB7-240 (k), CC7-2289 (l), CC7-335 (m) and A A7-738
(n) from layer N1-I. Con tinuous lines c onnect the fos sils with their f ind
locations. i, Second lowe r molar (F6-620) from layerJ in th e Main sector.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Extended Data Fig. 2 | Geographical distributions. Geographical d istribution o f the main IUP site s of western and c entral Eurasia ( black dots), directly d ated
early H.sapiens predatin g 37,000cal.  (empty bl ack dots) and direc tly dated late Ne anderthals a ssociated w ith Châtelper ronian assem blages (orange square s).
Bacho Kiro C ave is represente d by a red circle.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Article
Extende d Data Fig. 3 | Pho tographs of li thic artefa cts from layerI of Ba cho Kiro Cave. Poin ted retouche d blades and frag ments (1–4, 6, 7) and pie ce with
bifacial reto uch (5). Photograph s by V.S.-M. and T. Tsanova.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Extende d Data Fig. 4 | Draw ings of lithi c artefac ts from layerI of Bacho K iro
Cave. Poin ted retouche d blade with sligh tly oblique trun cation and bas e
modifi ed by inverse retouc h (1), pointed blade fra gments (2 and 5, wh ich has
an oblique tr uncation and sli ght notch on the le ft edge, and was p erhaps
intentio nally fragmen ted), pointed, sma ll blades frag ments (3, 7, 8 and 9),
pointed bl ade fragment w ith opposing p seudo-bur in blows on the apex and o n
the distal f racture edge (p erhaps indic ating use as a proje ctile) (4) and Levallois
flake (6). Drawings by I .K. and T. Tsanova).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Article
Extende d Data Fig. 5 | Hum an lower secon d molar (F6-62 0). a, Mesial, buccal
and distal v iews of the crown, ro ot and pulp chamb er (left) and oc clusal views
of the ename l and dentine crow n (right). b, A principal com ponent analysi s of
the shape of th e enamel–dent ine junction r idge and cerv ix places the Ba cho
Kiro Cave seco nd lower molar (F6-62 0) representedby a red st arwithin the
samples of r ecent (n=8) and Pleistoce ne (n=9) H.sapiens, and outsid e the
distribution of Neanderthals (n=20) and H.erectus (n=3).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Extende d Data Fig. 6 | MA LDI–TOF MS spectra fo r the six bone sp ecimens i dentif ied as homin ins through Zo oMS analysi s. a, B4-1653 (inter face of layers6a
and 7). b, AA7-738 (layerN1-I). c, BB7-240 (layerN1-I). d, CC7-2289 (layerN1-I). e, CC7-335 (layerN1-I). f, F6-597 (layerB).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Article
Extende d Data Fig. 7 | Frequ ency of nucle otide subs titutions at t he
beginn ing and the end s of mtDNA alig nments for the B acho Kiro Cave
specimens. Only fragme nts of at least 3 5base pairs in leng th that mappe d to
the revise d Cambridge Refere nce Sequenc e with a mapping qu ality of at leas t
25 were use d for this analysis . Solid lines in red d epict all frag ments and das hed
lines depi ct the fragm ents that have a C-to-T substi tution at the opp osing end
(‘conditional’ C-to-T substitutions). All other types of substitution are marked
in g rey.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Extende d Data Fig. 8 | Bayes ian phylogen etic tree rel ating Bacho K iro Cave
mtDNA to 54present-day humans, 10directly radiocarbon dated ancient
H.sapiens and the Vindija 33.16 Neanderthal. The Bacho Kiro C ave
specime ns are in red. Oth er ancient H.sapiens use d as calibration p oints to
estimate t he tip dates of Ba cho Kiro Cave speci mens are italic ized. The
posteri or probabilitie s are denoted ab ove the branches. T he mtDNA of Vin dija
33.16 was us ed to root the tre e (not shown).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Article
Extended Data Table 1 | Comparative dental metrics
BL, bucco-lingual width; MD, mesiodistal length; CI, crown index (BL/MD); CCA, calculated crown area (BL×MD).Values are in mm.
x
is the mean; minimum and maximum values are between
the brackets; σ is the standard deviation; n indicates sample size. The Upper Palaeolithic H.sapiens sample includes individuals from the sites of: Les Abeilles, Bacho Kiro Cave, Brno, Bruniquel,
Castenet, La Chaud, Dolní Vĕstonice, Farincourt, La Ferrassie, La Grèze, Les Rois, Isturitz, Kostenki, Kumchon, Laugerie-Basse, Lespugue, La Linde, Abri dela Madeleine, Nazlet Khater, Petera cu
Oase, Peche dela Boissiere, San Teodoro, St Germaine-la-Rivière, Sunghir, Les Vachons and Vindija. The early H.sapiens sample includes individuals from the sites of Border Cave, El Harhoura,
Cave of Hearths, Dar es Soltane, Die Kelders, Haua Fteah, Jebel Irhoud, Klaises River Mouth, Mumba, Qafzeh, Skhul, Témara and Zhiren. The Neanderthal sample includes individuals from the
sites of: Arcy-sur-Cure, Krapina, La Fate, Grotta Guattari, Hortus, Monte Fernera, Montmaurin, Ochoz, Petit-Puymoyen, La Quina, Le Regourdou, Spy, St Césaire, Subalyuk and Tabun. The recent
human sample includes archaeological specimens representing western Europe, eastern Europe, southern Europe, Japan, China, the Near East, India, the Andaman Islands, Australia, New
Guinea, northern Africa, southern Africa, eastern Africa and western Africa.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Extended Data Table 2 | mtDNA branch-shortening estimates
Estimates for Bacho Kiro Cave specimens as determined in a Bayesian framework imple-
mented in BEAST2, and by using 10radiocarbon-dated ancient H.sapiens as calibration points
(Supplementary Table9).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

1
nature research | reporting summary October 2018
Corresponding author(s): Jean-Jacques Hublin
Last updated by author(s): Feb 4, 2020
Reporting Summary
Nature Research wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency
in reporting. For further information on Nature Research policies, see Authors & Referees and the Editorial Policy Checklist.
Statistics
For all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section.
n/a Confirmed
The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement
A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly
The statistical test(s) used AND whether they are one- or two-sided
Only common tests should be described solely by name; describe more complex techniques in the Methods section.
A description of all covariates tested
A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons
A full description of the statistical parameters including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient)
AND variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals)
For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted
Give P values as exact values whenever suitable.
For Bayesian analysis, information on the choice of priors and Markov chain Monte Carlo settings
For hierarchical and complex designs, identification of the appropriate level for tests and full reporting of outcomes
Estimates of effect sizes (e.g. Cohen's d, Pearson's r), indicating how they were calculated
Our web collection on statistics for biologists contains articles on many of the points above.
Software and code
Policy information about availability of computer code
Data collection Avizo 6.3; Geomagic Studio 2015.1.3; E4 software (http://www.oldstoneage.com/software/e4.shtml) and Microsoft Access 2016;
Data analysis BATS 4.0; OxCal V4.3; Wolfram Mathematica 10; bam-rmdup 0.6.3; SAMtools 1.3.1; MAFFT v7.271; MEGA7, jModelTest2, mMass 5.5.0, R
3.4.2 and R 3.6.2, R package ggplot 2 3.2.1, R, package hexbin 1.28.0, RStudio 1.2.5033.
For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors/reviewers.
We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Research guidelines for submitting code & software for further information.
Data
Policy information about availability of data
All manuscripts must include a data availability statement. This statement should provide the following information, where applicable:
- Accession codes, unique identifiers, or web links for publicly available datasets
- A list of figures that have associated raw data
- A description of any restrictions on data availability
Sequence reads from all libraries and corresponding negative controls are deposited at ENA under the study accession number PRJEB35466.
The FASTA files of the mitochondrial genomes are deposited in GenBank with the accession numbers MN706602-MN706607.
Details are as follows:
BachoKiro_AA7_738 MN706602
BachoKiro_BB7_240 MN706603
BachoKiro_BK_1653 MN706604
BachoKiro_CC7_335 MN706605
BachoKiro_CC7_2289 MN706606
Content courtesy of Springer Nature, terms of use apply. Rights reserved

2
nature research | reporting summary October 2018
BachoKiro_molar_F6_620 MN706607
Other data that support the findings of this study are available from the corresponding authors upon reasonable request
Field-specific reporting
Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection.
Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences
For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pdf
Ecological, evolutionary & environmental sciences study design
All studies must disclose on these points even when the disclosure is negative.
Study description Analysis of the palaeontological and archaeological material discovered in the course of a new excavation at the site of Bacho Kiro
Cave (Bulgaria).
Research sample The archaeological material (artifacts and faunal remains) was extracted from two areas (Main Sector and Niche 1) in the Bacho Kiro
cave. Sediment samples were collected from each layer to perform micromorphological analyses. Collagen was extracted from
archaeological bones from all stratigraphic levels of Bacho Kiro Cave to perform ZooMS and radiocarbon analyses.
Sampling strategy For ZooMS, a random selection of morphologically unidentifiable bone specimens with a maximum length of over 2 centimeters was
conducted. For radiocarbon dating and ancient DNA analyses , all human remains identified by ZooMS were sampled. For the
morphological study of the human tooth F6-620, all comparative samples are detailed in the methods section.
Data collection M.H., T.T., N.S., VA, S.S., R.S. Z.R. and S.P.M. collected field data; S.B., M.S. and J.J.H. collected morphological data on the hominin
remains; V.S.M., L.P., F.W. and A.W. collected bone samples to perform ZooMS; M.H extracted mtDNA from hominin bone samples;
H.F. and S.T. collected bone samples to perform radiocarbon dating of hominin remains; G.S., R.S., V.P., N.M. collected morphological
data on the faunal assemblage using the faunal reference collection stored at the Bulgarian National Museum of Natural History was
used to accurately identify species and skeletal element; T.T. and S.P.M. collected metrical data on the lithic assemblage; V.A.
collected micromorphological and sedimentological data in the site.
Timing and spatial scale Bones were excavated from the Niche 1 and Main Sector areas of Bacho Kiro Cave during the 2015/2016/2017 field seasons. Bone
pretreatment, and ancient DNA analyses were carried out over the course of 2016-2018.
Data exclusions For the EDJ analysis, fossil teeth that were highly worn were excluded. For Radiocarbon dating two AMS dating methods across three
AMS labs were used to check reproducibility. 11 collagen extracts from different layers were dated with graphite targets on a
MICADAS AMS at two labs (ETH-Zurich and MAMS). Results were in statistical agreement for 8 of the extracts. Dates from the two
labs were outside 2 sigma for 3 of the oldest extracts (all >40,000 BP). These samples were excluded from further analysis in the
companion paper by Fewlass et al. Collagen from two human bones was dated with graphite targets at ETH-Zurich and in replicate
with the gas ion source of the Aix-MICADAS AMS at CEREGE. All measurements were in statistical agreement.
Reproducibility The mitochondrial genome sequences of Bacho Kiro Cave hominin specimens are deposited in GenBank.
Randomization ZooMS samples were randomly analyzed. For the EDJ study we cannot determine any reason to apply randomization.
Blinding For the EDJ study, blinding would be inappropriate given the small sample sizes and the relatively simple inferences made from the
results of the principal component analysis.
Did the study involve field work? Yes No
Field work, collection and transport
Field conditions Yearly excavation inside the Bacho Kiro Cave of about one month between 2015-2018
Location Bacho Kiro Cave, near Dryanovo (Bulgaria)
Access and import/export The archaeological material was studied in Bulgaria. Temporary exports of some items were organized between Bulgaria and
Germany. Permit delivered by National Museum of Natural History (Sofia) Nr. 4CH30/04.01.19.
Disturbance The samples were obtained through archaeological excavation of two sections in the cave. At the end of the project (summer
2020) measures will be taken to protect the stratigraphic profiles.
Reporting for specific materials, systems and methods
Content courtesy of Springer Nature, terms of use apply. Rights reserved

3
nature research | reporting summary October 2018
We require information from authors about some types of materials, experimental systems and methods used in many studies. Here, indicate whether each material,
system or method listed is relevant to your study. If you are not sure if a list item applies to your research, read the appropriate section before selecting a response.
Materials & experimental systems
n/a Involved in the study
Antibodies
Eukaryotic cell lines
Palaeontology
Animals and other organisms
Human research participants
Clinical data
Methods
n/a Involved in the study
ChIP-seq
Flow cytometry
MRI-based neuroimaging
Palaeontology
Specimen provenance Excavation of the Bacho Kiro Cave aurorized by the Bulgarian Ministery of the culture, delivered by NAIM-BAS: Nr124/11.05
2015; Nr225/28.04.2016; Nr47/02.05.2017; Nr99/17.04.2018/ Nr120/2019.
Specimen deposition The palaeontological material will be deposited at the National Museum of Natural History in Sofia and the lithic material at the
History Museum of Dryanovo (Bulgaria)
Dating methods Small aliquots (80-110 mg) of the six ZooMS identified hominin bone fragments were sampled to preserve as much material as
possible for further analyses. Collagen was extracted using a technique based on a modified Longin collagen extraction protocol
followed by an ultrafiltration step. The gelatinized collagen samples were then passed through an Ezee Filter (Elkay labs, UK) to
remove large particles (>80 μm) and separated by molecular weight with pre-cleaned Sartorius VivaSpin Turbo 15 ultrafilters (30
kD MWCO). The samples were freeze dried and the large molecular fraction (>30 kD) was graphitised using the Automated
Graphitisation Equipment III and measured using the latest model of the MICADAS AMS in the Laboratory of Ion Beam Physics at
ETH-Zurich (lab code: ETH). The dates were calibrated using the IntCal13 dataset in OxCal v4.3.
Tick this box to confirm that the raw and calibrated dates are available in the paper or in Supplementary Information.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

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