The paleoanthropological characteristics of the Los Canes fossils place them within the variability described for the last hunter-gatherers of the Late Upper Paleolithic/Early Mesolithic of Western Europe []. All adult individuals from Los Canes showed marked muscle insertions, and especially I-A and III-A individuals had oral pathologies including caries, ante-mortem tooth loss and dental calculus; marked wear of the crown surfaces was also observed []. Because of the availability of petrous bones, individuals I-A and II-A were selected for genetic analysis. After estimating the percentage of endogenous DNA, only I-A was selected for genome sequencing.

Burial III: This grave contained the remains of a complete adult male (III-A), dated to 6,930 ± 95 yBP (AA-6071 []; 7,944–7,609 cal yBP). Earlier skeletal remains from an infant (III-B) associated with coeval bones of the Iberian chamois (Rupicapra pyrenaica), red deer (Cervus elaphus) and wild boar (Sus scrofa) were found over III-A’s knees, indicating the removal of a previous burial [].

Burial II: Remains of two individuals interred at different times were found in this grave []. Most of the skeleton of the earlier one (individual II-B, an adult), dated to 6,860 ± 65 yBP (AA-5295 []; 7,826–7,583 cal yBP) was probably removed when the later individual (II-A) was buried, and only the feet in anatomical connection and some scattered bones and teeth were preserved. The skeleton of individual II-A, a subadult male (laboratory ID: Canes2; archaeological ID: II-A), was found in anatomical connection, lying on his left side in flexed position. It was dated by AMSC to 6,770 ± 65 yBP (AA-5296; []), 7,025 ± 80 yBP (AA-11744; []) and 7,208 ± 38 yBP (OxA-23185; []). The weighted mean of these three measurements is 7,092 ± 31 yBP (R-combine, Oxcal 4.2) and the 2-sigma calibrated age range is 7,974–7,850 cal yBP. Several grave goods were associated with the II-A skeleton: two frontlets of female ibex (Capra pyrenaica), a long cobble with traces of red colorant, a long, pointed bone, a perforated antler, and a pecked cobble, possibly representing a human head. Pendants (most of them made of Trivia sp. shells) were found around the head and shoulders of this individual, suggesting that they were sewn to a dress [].

Burial I: This corresponds to a woman of advanced age (laboratory ID: Canes1_Meso; archaeological ID: I-A) that was AMSC dated to 6,265 ± 75 yBP (AA-5294; []) and 6,160 ± 55 yBP (OxA-7148; []). The weighted mean of the two measurements is 6,197 ± 45 yBP (R-combine, Oxcal 4.2), giving a 2-sigma calibrated age calculated using OxCal 4.2 and the IntCal13 dataset of 7,245–6,985 cal yBP. Some objects have been interpreted as grave goods: a red deer (Cervus elaphus) scapula, an ungulate rib and three perforated red deer canines. Also, a large quantity of shells of Cepaea nemoralis was found all around the skeleton, probably an intentional deposit [].

Los Canes cave is located on the southern slope of Sierra de Cuera, a mountain chain in the eastern part of the Asturias region in northern Spain (43.3550277, −4.7166763). There is evidence of human occupation from the Solutrean until the Bronze Age, and several human remains have been found associated with these periods [].

Chan_Meso (archeological ID: Elba) is a female individual of small size (around 153 cm tall and 56 kg in weight, according to []) but shows marked muscle scars and several pathologies indicating hard physical work from youth []. The only tooth recovered shows a large caries lesion. IndependentC dates are available from the skull (7,995 ± 70C years yBP; Ua-13398; []) and the tibia (8,236 ± 51C yBP, Ua-38115; []). The weighted mean of the twoC ages is 8,155 ± 42 yBP (R-combine, Oxcal 4.2) and the 2-sigma calibrated age range calculated using OxCal 4.2 and the IntCal13 dataset is 9,255–9,007 cal yBP ( Table S1 ).

No other human remains or goods or lithic tools were found. The only other findings at the site are the remains of three small-sized aurochs. The sedimentological analysis suggests that the collapse occurred in a single episode, and the radiocarbon ages for the human and the aurochs overlap [], suggesting that all these remains might be somehow related.

The Chan do Lindeiro karstic system (42.7343714,-7.0305368, Pedrafita, Lugo, Spain) is a large vertical fracture associated with a small doline. The human remains from this site belong to a single individual (laboratory ID: Chan_Meso; archaeological ID: Elba) and were found with no funerary context in a deep cave gallery associated with a collapsed sinkhole []. The skeletal fragments were scattered among the collapse debris and completely disarticulated. Only a few bones were recovered: a partial neurocranium, a single tooth, some vertebrae, both clavicles, fragments of ribs, a partial ulna, both femora, a partial tibia and one metatarsal.

The individual from Gura Baciului included in this study (laboratory ID: GB1_Eneo) comes from grave no. 1 (archaeological ID: M1). This grave was a chance find when the section of an older trench collapsed. Discovered in 1962 near a pit-house, it is a primary inhumation of a single individual (GB1_Eneo) in anatomical connection, deposited in the crouched position on the left side, with ESE–WNW orientation (head to east). The reported grave goods comprise ten flint flakes in the feet area and a bone awl and two ochre fragments in the cheek and hip areas. Eighty-three fragments of animal bones (of Bos taurus, Ovis aries/Capra hircus, Cervus elaphus, Bos primigenius), mollusca, broken stones and numerous ceramic fragments supposedly formed a “bed” underneath the body []. Among this material were three “loose” human bones of an 11–13 year old child []. The skeleton from grave no. 1 is that of an adult female, around 155 cm tall []. The AMSC date of 4,621 ± 28 yBP (5,456–5,299 cal BP, MAMS-28614, this study) ( Table S1 ) places this burial in the Eneolithic period, and is further evidence of the post-abandonment use of the Starčevo culture settlement as a formal burial area. The δN value of +12.7‰ implies the individual had a high trophic level diet that may have included significant amounts of fish, which in turn would likely result in aC age that is older than the archaeological context (i.e., that includes a reservoir offset). Thus, theC date should be regarded as a maximum age for this individual. The Eneolithic period date for burial no. 1 is not entirely surprising, since in early survey work in the Gura Baciului locality (prior to 1942 — items from a private collection donated to the museum in 1942), traces of Eneolithic and EBA activity were recorded [].

Skeletal remains from grave no. 6 were dated to 6,400 ± 90 yBP (7,495–7,159 cal yBP; Lv-2157; []). Recently, two other graves from Gura Baciului were dated: grave no. 2 – 6,350 ± 40 yBP (7,415–7,174 cal yBP) and grave no. 3 – 6,370 ± 40 yBP (7,420–7,182 cal yBP; []). However, there are no associated carbon and nitrogen stable isotope values from which to assess the individuals’ diet. Assuming no freshwater reservoir effect, this date places the burial around the end of the Starčevo culture period and it is distinctly possible that the burial was emplaced after the living area had been abandoned. The onlyC date from Gura Baciului from a non-burial context is an AMS measurement on animal bone from a pit feature (possibly a dwelling structure) of 7,140 ± 45 yBP (8,035–7,860 cal BP; GrA-24137), which predates burial no. 6 by c. 600 years.

Seven primary inhumation burials (no. 1, 2, 3, 4, 5, 6, and 9), a cremation grave (no. 7) and a secondary burial (no. 10) were found within the area of the Starčevo culture settlement, while ‘loose’ human bones (a complete skull [M8] or skull fragments) were discovered in domestic contexts (e.g., pits and houses) belonging to the ‘cultural layer’. Most of the primary inhumations were buried in crouched positions, on the left or right sides, with varying orientations. These crouched inhumations were assumed to be contemporaneous with the Starčevo culture occupation, although few if any chronologically diagnostic items (‘burial offerings’) were recovered from the graves. Moreover, across Southeast Europe, Neolithic and later prehistoric people tended to bury their dead in formal disposal areas on the periphery of (or some distance away from) their living areas, and it was not unusual for graves to be dug into abandoned living areas.

The Gura Baciului site (46.7877247, +23.5208770) is located on a terrace of the Suceag creek, in Transylvania, near Cluj Napoca city (Cluj county). Excavations by N. Vlassa (in 1960, 1962, 1965, 1967-1971) and Gh. Lazarovici (in 1990–1993) uncovered the remains of a Starčevo-Criş settlement with huts or houses, pits and concentrations of domestic refuse []. Food remains recovered in the excavations indicate a typical Early Neolithic farming economy based on crop (cereals, etc.) and livestock (cattle, sheep/goat and pig) husbandry. The material culture remains included pottery, lithic artifacts (e.g., flint, obsidian, ground stone axes, seed grinding equipment), anthropomorphic and zoomorphic figurines, and various kinds of personal ornaments (clay bracelets and buttons, bone rings, Spondylus shell bracelets and pendants). Based on analysis of the pottery, the site was considered to have been occupied more-or-less continuously throughout the greater part (stages IB–IVB) of the Starčevo culture time range []. There are very fewC dates on finds from the Gura Baciului excavations, butC results from other Starčevo culture sites in Transylvania indicate a time range for phases I–IV of c. 7950–7350 cal BP (6,000–5,400 cal BCE) [].

The individual from Ostrovul Corbului included in this study (laboratory ID: OC1_Meso) comes from burial no. 24. Only the upper part of the skeleton was preserved, with the bones in anatomical position. The lower part of the skeleton appears to have been destroyed by a later pit feature. From the surviving portion of the skeleton, burial 24 was interpreted as an extended, supine inhumation oriented with the head toward the east []. The skeleton is that of an adult male, with age at death estimated at 30-35 year and stature at 172 cm []. The AMSC date obtained for this study was 8,277 ± 34 yBP (MAMS-28615). After applying a FRE correction using Method 1 of [], this converts to a 2-sigma calibrated age range of 8,972–8,435 cal BP ( Table S1 ), which overlaps with the calibrated age of SC1 from Schela Cladovei.

The Ostrovul Corbului site is also situated in the Iron Gates region of southwestern Romania (44.5154854, +23.52087725) on a former island in the Danube River, 28km downstream of Schela Cladovei. Settlement remains, individual graves and a cemetery belonging to various prehistoric periods (Mesolithic, Neolithic, Eneolithic, Bronze Age, and Iron Age) were identified during several excavation campaigns between 1973 and 1984 []. Seven inhumation burials (no. 2, 9, 18, 24, 25, 30 and 32) were found in an area with Mesolithic and Early Neolithic settlement remains at the SW (downstream) end of the island. These were previously interpreted as Early Neolithic in date, but AMSC dating has shown that burials no. 2, 25 and 30 belong to the Middle Mesolithic between about 9.7–9.3 kya, while burial no. 32 dates to the Late Mesolithic around 8.6 kya [].

SC2_Meso: Child, 5-7 years of age at death. There is noC date for SC2_Meso, but it belongs to the same (Late Mesolithic) burial cluster as SC1_Meso and all dates for those burials are statistically indistinguishable at the 2-sigma level []; thus SC2_Meso can be expected to date to the same time period as SC1_Meso.

SC1_Meso: Adult male, age-at-death 35-45 (dental attrition). The skeleton was lying on the right side, with the legs slightly flexed. The burial was truncated by an Early Neolithic pit, which removed the mid-section of the skeleton. The distal ends of both femurs and the lower legs were missing, possibly removed by another pit feature. The skeletal remains were dated to 8,380 ± 80 yBP (OxA-8583) and corrected to 7,960 ± 96 yBP (9,075-8,553 cal yBP) after considering the FRE [] ( Table S1 ). The FRE is related to fish consumption; since fish from the Danube are relatively depleted inC, radiocarbon dates for fish bones and the bones of animals (including humans) that consumed fish are older than their archaeological context. This age offset can be quantified and corrected for based on the δN ratio, as described by [].

At least 75 burials, containing the remains of over a hundred individuals, have been excavated from the Schela Cladovei site so far, most of them dated to the Late Mesolithic. The two individuals from Schela Cladovei included in this study were from burials M95/2 (Laboratory ID: SC1_Meso) and M96/3 (Laboratory ID: SC2_Meso), both found among 21 burials uncovered in an area c. 25 m x 4 m immediately adjacent to the Danube riverbank between 1991 and 1996. Of those 21 burials, which included adults and children, 11 (all adults) have single-entity AMS 14 C dates from the Oxford Radiocarbon Accelerator Unit (ORAU). The dating was done prior to the use of ultrafiltration by ORAU. The calibrated ages (after correction for the Danube Freshwater Reservoir Effect, FRE) range between 8,950 and 8,550 cal yBP.

Archaeological remains in the areas investigated relate mainly to the Late Mesolithic and Early Neolithic, with sporadic evidence of later (Iron Age and Medieval) occupation. A large series of single-entity AMSC dates on animal and human remains [] places the Late Mesolithic occupation between c. 9,150 and 8,250 cal yBP, and the Early Neolithic occupation between 7,950 and 7,550 cal yBP.

Schela Cladovei (Romania), is a large, open-air site on an Early Holocene terrace adjacent to the River Danube (44.6258333, +22.6066666), c. 7 km downriver from the Iron Gates I dam. Discovered in 1964, the first excavations were undertaken by the Romanian archaeologist Vasile Boroneanţ. From 1992 onward, the excavation became a joint Romanian–British research project, co-directed by V. Boroneanţ, A. Boroneant and C. Bonsall.

The sampling was agreed in order to minimize damage to the archaeological remains and all samples were collected and analyzed in accordance with the research ethics policies and procedures of the academic institutions involved in the research.

Sample preparation, DNA extraction and library building

32 Gamba C.

Jones E.R.

Teasdale M.D.

McLaughlin R.L.

Gonzalez-Fortes G.

Mattiangeli V.

Domboróczki L.

Kővári I.

Pap I.

Anders A.

et al. Genome flux and stasis in a five millennium transect of European prehistory. 68 Pinhasi R.

Fernandes D.

Sirak K.

Novak M.

Connell S.

Alpaslan-Roodenberg S.

Gerritsen F.

Moiseyev V.

Gromov A.

Raczky P.

et al. Optimal ancient DNA yields from the inner ear part of the human petrous bone. All pre-amplification DNA procedures were carried out in dedicated aDNA laboratories at the University of York (UK). All laboratory tools used to process the samples were either sterile and disposable or decontaminated with bleach (concentration 1:5) and exposed to UV light for 24h before being used. DNA was extracted from the walls forming the channels of the inner ear within the petrous bone, which tend to preserve comparatively high percentages of endogenous DNA []. In total, 7 petrous bones were processed at this step: 6 from human remains associated with Mesolithic sites in Romania (SC1_Meso, SC2_Meso and OC1_Meso) and Spain (Chan_Meso, Canes1_Meso and Canes2_Meso); and 1 from a Neolithic/Eneolithic site in Romania (GB1_Eneo). Before extraction, the fragments of petrous bones were treated for decontamination: they were first exposed to UV light for 10 min on each side, followed by physical removal of the surface with a Dremel drill, and finally, once the surfaces were clean, they were exposed again to UV light for another 10 min.

69 Rohland N.

Siedel H.

Hofreiter M. A rapid column-based ancient DNA extraction method for increased sample throughput. 70 Dabney J.

Knapp M.

Glocke I.

Gansauge M.T.

Weihmann A.

Nickel B.

Valdiosera C.

García N.

Pääbo S.

Arsuaga J.L.

Meyer M. Complete mitochondrial genome sequence of a Middle Pleistocene cave bear reconstructed from ultrashort DNA fragments. 69 Rohland N.

Siedel H.

Hofreiter M. A rapid column-based ancient DNA extraction method for increased sample throughput. The bone fragments were ground to powder using a mortar and pestle and one DNA extract was obtained from each sample. Because the samples from Romania and Spain were sampled and extracted at different times, we followed two different protocols ([] and [], respectively). From the Romanian samples, DNA was extracted following [], starting from approximately 250 mg of bone powder. Each sample was incubated in rotation with 5 ml of extraction buffer, consisting of 0.45 M EDTA (pH 8.0) and 0.25 mg/ml of proteinase K. After overnight incubation, the samples were centrifuged and the supernatant was transferred to a new tube with 2.5 ml of binding buffer (5 M guanidinium thiocyanate (GuSCN) and 300 mM sodium acetate) and 100 μl of silica suspension. The mix was incubated for 3 hr in gentle rotation at room temperature. After incubation, the tubes were centrifuged, the supernatant discarded and the silica pellet was resuspended in 400 μl of binding buffer and transferred into a clean Mobicol column with an extra 10 μm filter (MoBiTec GmbH). The columns were centrifuged and then washed twice with 450 μl of washing buffer (50% ethanol, 125 mM NaCl, 1 × TE) in order to remove salts and PCR-inhibitors. Finally, the silica pellet was dried and the DNA eluted into a clean tube using 50 μl of TE.

70 Dabney J.

Knapp M.

Glocke I.

Gansauge M.T.

Weihmann A.

Nickel B.

Valdiosera C.

García N.

Pääbo S.

Arsuaga J.L.

Meyer M. Complete mitochondrial genome sequence of a Middle Pleistocene cave bear reconstructed from ultrashort DNA fragments. The Spanish samples were extracted following a more recent protocol from [], which optimizes the recovery of short DNA fragments from small quantities of bone powder. From each sample, 50 mg of bone powder were digested in 1 ml of an extraction buffer consisting of 0.45 M EDTA (pH 8.0) and 0.25 mg/ml of proteinase K. After overnight incubation, 1 ml of supernatant was added to 13 ml of binding buffer (5M Guanidine hydrochloride (MW 95.53), 40% Isopropanol, 0.05% Tween-20, 9 mM Sodium Acetate) and poured into a binding apparatus consisting of an extension reservoir (Zymo Research) fitted to a MinElute silica spin column (QIAGEN). The binding apparatus was placed into a 50-ml falcon tube and centrifuged. During the centrifugation, the silica based membrane in the MinElute column will retain the DNA molecules. This filter is washed by adding 650 μl of PE buffer (QIAGEN). Finally, the filters are dried by centrifugation and the DNA molecules are eluted by adding a total of 25 μl of TET buffer (1 M Tris-HCL, pH 8.0, 0.5 M EDTA, 10% Tween-20).

Both extraction experiments were performed including two blank controls each.

36 Meyer M.

Kircher M. Illumina sequencing library preparation for highly multiplexed target capture and sequencing. 71 Fortes G.G.

Paijmans J.L.A. Analysis of whole mitogenomes from ancient samples. Illumina libraries were built following the protocol described by [], with the modifications suggested by [], which minimize the number of purification steps by introducing heat inactivation of the enzymes between consecutive reactions. We used 20 μl of DNA extract as template to build one sequencing library per sample. As first step, we performed the blunt end repair of overhang strands with enzymes T4 Polynucleotide kinase and T4 polymerase in the following reaction mix: 1x Buffer Tango, dNTP (100 μM each), 1 mM ATP, 0.1 U/μl T4 PNK, 0.1 U/μl T4 Polymerase. After 20 min of incubation at 25°C, the mix was heat up to 72°C for 20 min, in order to inactivate the enzymes. Following, the adaptor mix (1x T4 ligation buffer, 5% PEG-4000 and a mix 0.5 μM of the paired end adapters P5 and P7) was added to the former reaction, without any purification step through columns. Just before starting the incubation, 1.25 μl of T4 ligase (final concentration 5 U/μl) were added to each tube to complete the adaptor mix reaction. The tubes were incubated for 30 min at 22°C and then the volume was filtered through silica-based MinElute columns (QIAGEN). In the next step, the Bst polymerase was used to fill the nicks and to complete the double strand sequence of the adapters (reaction mix: 1x Thermopol buffer, dNTP (250 μM) and 0.3 U/μl Bst polymerase). The mix was incubated at 37°C for 20 min and then heated up to 80°C for 20 min.

After the incubation, each library was indexed and amplified in three parallel reactions, without any filtration step between the adaptor fill-in and the amplification. We used 5 μl of library as template for each reaction and the primers IS4 (5′-AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTT-3′) and P7 indexing (5′-CAAGCAGAAGACGGCATACGAGATxxxxxxxxGTGACTGGAGTTCAGACGTGT-3′). The P7 indexing primer includes a barcode sequence of 8 nucleotides (denoted by the eight x). We used a different barcode for each sample, so the libraries could be pooled and sequenced together. Library amplifications were carried out using Accuprime Supermix I (ThermoFisher Scientific), which includes a polymerase that is able to read over uracils. The PCR mix was prepared for a total volume of 25 μl following the instructions from the manufacturer. Amplification conditions were as follows: 95°C for 5 min; 12 cycles of: 95°C for 15 s, 60°C for 30 s, 68°C for 30 s and, finally, an extension of 68°C for 5 min. The three PCR reactions from each sample were pooled and purified on a MinElute column (QIAGEN), with a final elution in 15 μl of EB buffer. The amplified libraries were visualized on agarose gels and quantified on a Bioanalyzer 2100 with the High Sentitivity DNA chip (Agilent).