Abstract
The Epigravettian human remains from Riparo Tagliente in northern Italy represent some of the earliest evidence of human occupation in the southern Alpine slopes after the Last Glacial Maximum. Genomic analyses of the 17,000-year-old Tagliente 2 mandible revealed the oldest presence of a genetic profile with affinities to the Near East in the Italian peninsula, which later became the most widespread hunter-gatherer ancestry across Europe. However, a comparable biomolecular characterization of the Tagliente 1 burial remains unavailable, preventing us from defining its biological relationships with Tagliente 2. Here, we apply paleogenomic, isotopic, and radiocarbon dating analyses on a femur fragment of Tagliente 1 and compare the reconstructed data with previously reported results from Tagliente 2. Despite their different isotopic signatures and non-overlapping radiocarbon dates, we reveal that the two human remains belong to the same male individual. We determine that the distinct isotopic values can be explained by different dietary practices during lifetime, whereas the non-overlapping radiocarbon dates can be caused by minimal radiocarbon contamination, possibly deriving from chemical treatments for conservation purposes. These findings highlight the importance of interdisciplinary biomolecular studies in offering new perspectives on the Palaeolithic fossil record and addressing long-standing bioarchaeological questions.
Similar content being viewed by others
Introduction
In recent years, research in biomolecular archeology is rapidly expanding, owing to major methodological advancements both at the laboratory and computational levels in the fields of genetics, proteomics, radiocarbon dating and stable isotopes, among others. Despite major progress, biomolecular preservation remains a limiting factor for this type of research. This is likely influenced by many microclimatic factors, but also by time i.e., biomolecular preservation tends to be poorer for older periods1,2,3. Investigations focusing on the European Palaeolithic have found that archeological caves and rock shelters often offer environmental conditions that are favorable for biomolecular preservation, greatly aiding our investigation of early modern human populations4. In addition, in those sites where human skeletal remains have been discovered, it is possible to perform biomolecular assays on the retrieved specimens providing direct insights on the date, diet and genomic profile of the analyzed individuals. However, most of these caves and rock shelters were excavated prior to major technological advancements and can now be re-examined with an interdisciplinary approach in an attempt to answer questions that were previously unfeasible to address.
Riparo Tagliente, located in the province of Verona in northeast Italy, is an example of a rock shelter with an abundant Paleolithic zooarchaeological and osteological collection, whose recent re-investigation with radiocarbon dating, isotopic analysis and ancient DNA (aDNA), has proven to be highly informative many years after its initial discovery5,6. The rock shelter was first discovered in 1958 by Francesco Tagliente. The site underwent preliminary investigations between 1962 and 1964 under the Museo Civico di Storia Naturale of Verona. In 1967, the University of Ferrara took charge of the excavations and is the main institution to conduct research at the site ever since. The rock shelter’s strategic ___location at the convergence of diverse topographic formations has yielded extensive knowledge about the interaction between the populations who inhabited the site and their environment, as well as insights into their subsistence strategies5,6,7,8,9,10,11,12,13. The remarkable assemblage of archeological remains found at the site has provided invaluable temporal information, spanning from the Middle Palaeolithic onwards, including the presence of Neanderthal skeletal remains14. In the Upper Palaeolithic layers of Riparo Tagliente, other skeletal remains associated with the Late Epigravettian were found. In 1963, a human mandible was discovered in the disturbed external area of the shelter15. Ten years later, an incomplete burial, referred to as Tagliente 1, was unearthed within the rock shelter in the southernmost sector of the excavated area (Fig. 1). Bartolomei et al.7, conducted a comprehensive descriptive analysis of the Tagliente 1 burial, only consisting of postcranial skeletal elements positioned in a supine orientation and partially covered by calcareous limestone blocks. Morphologically, both findings (the initially discovered mandible and Tagliente 1) have been identified as males with an estimated age at death ranging between 22 and 30 years. However, due to the different stratigraphic locations and the fragmentary nature of the remains, previous investigations attributed the mandible to a distinct individual, referred to as Tagliente 215. This assignment to two individuals is consistent with existing chronological analyses. A rib of Tagliente 1 and a molar of Tagliente 2 have been radiocarbon dated to 16,130–15,570 cal BP (2σ) and 16,980–16,500 cal BP (2σ), respectively5,6 indicating a time separation between the two individuals of 890 years using the medians of both calibrated dates (Fig. 2).
A Topographic map (geospatial data was downloaded using the giscoR package in R) of Italy where the ___location of the Riparo Tagliente site is indicated with a black dot; (B) Photograph of the rock shelter entrance; (C) Human hemimandible Tagliente 2; (D) Human burial Tagliente 1.
Recent aDNA analysis on Tagliente 2 has revealed that this individual carried a genetic profile related to the “Villabruna cluster”, an ancestry that was widespread in Epigravettian-associated individuals from the Italian peninsula and that largely contributed to the vast majority of European populations after ~14,000 years ago5,16,17. The presence of this genetic profile in the Alpine region soon after the end of the Last Glacial Maximum (LGM) indicates that it reached Italy thousands of years before spreading into central Europe. In addition to the aDNA work, isotopic analysis conducted on teeth from the Tagliente 2 mandible showed that the individual had a very high protein diet, similar to what is observed for contemporaneous individuals13. For Tagliente 1, apart from the younger radiocarbon date, isotopic values pointed to a substantial amount of aquatic protein sources in this individual’s diet6. A paleogenetic study has, however, never been conducted on this specimen. In this study, our objective is to align the biomolecular information obtained from Tagliente 1 with that of Tagliente 2. We applied state-of-the-art protocols on a femur fragment of Tagliente 1 to obtain aDNA data from this specimen, alongside newly generated isotopic and radiocarbon dating results. We then compared the genomic, temporal, and stable isotopic profiles of Tagliente 1 and Tagliente 2, to clarify the relationships between these two Late Epigravettian human remains within an interdisciplinary framework.
Results
The new AMS date on the femur of Tagliente 1 yielded an age of 16,360-16,210 cal BP at 68.3% and 16,460-16,130 cal BP at 95.4% probability (Supplementary Data 1 and Fig. 2). This date has a median value which is ~450 years older than the previous date on a rib of Tagliente 16. However, the newly generated date is still ~450 years younger than what obtained for a first molar of Tagliente 2 and not overlapping at 2σ probability5. We then investigated the possibility that this temporal difference could be influenced by different dietary practices during the individual’s lifespan. In fact, the dentine of the first molar does not remodel after its completed growth, between the third and fifth year of life, and it retains an early dietary signal that may be impacted by milk consumption before complete weaning. It can thus be suspected that the dentine of the first molar has an elevated δ13C and δ15N value compared to continuously remodeling bones. In contrast, ribs are among the bones with the fastest collagen turnover, meaning that they would reflect the last years of the dietary life of the individual. Femurs show a slower turnover and thus isotopically record more than the last ten years of the individual’s diet18. We then performed a new stable isotope measurement on the collagen extracted from the femur of Tagliente 1. The δ13C and δ15N values (–18.5‰, +13.5‰, respectively) are consistent with what previously obtained from a rib of the same burial (δ13C = –18.4‰, δ15N = + 13.0‰)6 (Fig. 3, Supplementary Data 2). When compared with the stable isotopic measurements obtained from the fauna in Riparo Tagliente6, the results from both Tagliente 1 bones confirm the higher than expected positive shift of ca. 1‰ δ13C and 3 to 5‰ δ15N between the collagen of the consumer and those of its supposed prey19. This was interpreted as Tagliente 1 having consumed aquatic resources6. Importantly, an aquatic based diet could lead to a reservoir effect on the radiocarbon dating resulting in artificially older dates20. However, since the isotopic values from the first molar of Tagliente 2 (δ13C = −19.5‰, δ15N = 11.5‰)13 are lower than those obtained from the bone fragments of Tagliente 1 (rib and femur), a possible reservoir effect on the radiocarbon date of Tagliente 2 should also be lower. Taken together, the fact that the newly generated radiocarbon date of Tagliente 1 is significantly younger than Tagliente 2 cannot be explained by a higher fresh-water reservoir effect, which would instead make the specimen’s date artificially older.
Isotopic values of carbon and nitrogen isotopes (δ13C and δ15N) from the collagen of Epigravettian human and/or animal remains from Riparo Tagliente (RT) in northern Italy, Grotta del Romito (GR) in southern Italy, and Šandlja II (SD) in Croatia.
To assess if the Epigravettian-associated mandible and postcranial remains from Riparo Tagliente are biologically related, we generated genomic data from the femur of Tagliente 1. The initial shallow shotgun screening of the genomic library yielded ~4.5 M raw reads, of which 0.51% were mapped to the human reference genome (hg19) after quality filtering. The average fragment length of the mapped reads was ~58 bp with a CtoT substitution rate of 8.1% at the 5’ end of partially UDG-treated DNA, suggesting the presence of aDNA in this sample. We then used an in-solution capture approach to enrich both the complete mitochondrial DNA and ~1.24 M SNPs across the human genome (mtDNA and 1240K captures, respectively). The targeted enrichment produced an average mtDNA coverage of 5.3X and a nuclear DNA coverage on the 1240K capture panel of 0.14X, resulting in 150,334 SNPs. We determined the individual to be genetically male and estimated 6 ± 4% and 26.9 ± 7.3% mtDNA and nuclear DNA contamination, respectively. After utilizing PMDtools to reduce contamination by filtering for deaminated reads, we retained a total of 6317 SNPs (see Methods section).
We determined the mtDNA haplogroup of Tagliente 1 to U2′3′4’7’8’9, which is the same haplogroup obtained from Tagliente 25. This mtDNA haplogroup is commonly observed in the Palaeolithic European record, with particularly high frequency in the Italian peninsula16,21,22,23,24. We then assigned the Y-chromosome haplogroup to I2, which is also found in Tagliente 25. This Y-chromosome haplogroup rose in frequency during the Upper Palaeolithic to become the dominant paternal marker in younger Epigravettian and Mesolithic Europeans16,17.
Subsequently, we calculated the pairwise mismatch rate (PMR) by comparing the genotype of Tagliente 1 to Tagliente 2 and ten other Epigravettian and Mesolithic individuals from Italy belonging to the “Villabruna cluster” (Fig. 4A and Supplementary Data 3). The PMR between Tagliente 1 and the other Villabruna-related individuals (excluding Tagliente 2) has an average value of 0.212. In contrast, the PMR between the two Tagliente remains is determined to be 0.116. Notably, when PMR is calculated between Tagliente 2 and the genotype of Tagliente 1 after PMD filtering or between both Tagliente samples after PMD filtering, the resulting values approach even closer to half of the PMR baseline for unrelated individuals (Fig. 4A and Supplementary Data 3). Since the genotype for each individual is inferred in pseudo-haploid form with equal probability of selecting one allele for each heterozygous SNP, the expected PMR value for identical or twin individuals is half of the PMR baseline25. Therefore, based on this analysis, we conclude that Tagliente 1 and Tagliente 2 are either twin individuals or different skeletal elements originating from the same individual.
A Pairwise Mismatch Rate (PMR) between Tagliente 1 and other Italian hunter-gatherers carrying “Villabruna” ancestry or Tagliente 2 only. In the box plot the orange line describes the PMR median and box limits the upper and lower quartiles (n = 10). B Pairwise f3-outgroup statistics between Tagliente 1 and other Italian hunter-gatherers associated with the “Villabruna” ancestry (Mbuti used as outgroup). Error bars represent one standard error.
To confirm this conclusion via a method that provides statistical uncertainty, we calculated f3-outgroup statistics in the form f3(Mbuti; X1, X2), where X1 and X2 are Epigravettian and Mesolithic individuals from Italy carrying “Villabruna” ancestry (Fig. 4B and Supplementary Data 4). When Tagliente 2 is compared to our newly generated data from Tagliente 1, we obtain a f3 value of 0.571 ± 0.006. This value increased (0.619 ± 0.024) when we used the PMD-filtered data from Tagliente 1 (Supplementary Data 4). The resulting shared drift between Tagliente 1 and Tagliente 2, being much higher than the baseline among hunter-gatherer individuals associated with the “Villabruna” ancestry (0.378 on average) (Fig. 4B), can be explained by these two skeletal elements indeed belonging to identical twins or to the same individual. In addition, the fact that the f3-outgroup value between the two Tagliente samples is substantially higher than 0.5 suggests that the population these samples derive genetic ancestry from had low heterozygosity. This likely resulted from a small effective population size, as previously reported for other Epigravettian-associated individuals from Italy16.
Discussion
Inferences using nuclear DNA (sex determination, PMR, f-statistics) and uniparental markers (mtDNA and Y-chromosome) all converge on the conclusion that the skeletal remains of Tagliente 1 and Tagliente 2 either belong to identical twins or to the same individual. To distinguish between these two scenarios, we re-evaluate the morphological analyses previously performed on both human remains.
The determination of Tagliente 1 age at death was achieved through the examination of the full fusion of the epiphysis in long bones, indicative of an adult status, as well as an assessment of the pelvic region. The analysis of symphyseal facets revealed the persistence of transverse grooves and distinct characteristics in the anterior and posterior edges of the symphyseal regions26. The presence of horizontal crests within the pelvic region and the absence of clear demarcation between the upper and lower boundaries of these facets suggest that the age at the time of death likely fell within the range of 22–24 years26. Regarding Tagliente 2, limited information is available due to the fragmentary nature of the mandible and the scarcity of diagnostic morphological features. However, it is plausible to attribute this individual to an age above 25 years but below 30 years, as evidenced by the eruption of the third molar and the limited dental wear. Through morphological analyses, both samples have been identified as males. The strong indication of male sex for Tagliente 1 is inferred from the proportions and shape of the sacrum, which is further supported by the overall morphology of the pelvis. The pelvis exhibits a relatively tall and narrow shape with robust and massive characteristics. Distinctive features include the sharp angulation of the pelvic brim (linea arcuata), the triangular appearance of the body of the pubis (lamina quadrilatera), and its subpubic angle, which is concave rather than convex downward. The obturator foramen is nearly oval rather than triangular. Additionally, the long bones exhibit robust features, reinforcing the determination to male sex26. For Tagliente 2, morphological sex determination is less straightforward due to the absence of several key characteristics typically used for such assessment. However, measurable dimensions, such as the height of the mandibular body and its thickness at the mental foramen level, are more consistent with a male rather than a female attribution15. Overall, the newly obtained genomic results, alongside previously performed morphological investigations indicating a male sex and similar age at death, allow us to most parsimoniously conclude that Tagliente 1 and Tagliente 2 belong to one single individual. We thus retract the previous individual names and, to avoid future inaccuracies, we rename the entire fossil as “Tagliente 1/2” (postcranium and mandible).
This result requires explanation of the reasons why the two human remains were for long considered as belonging to two different individuals. The Tagliente burial came from the inner area of the rock shelter, but it was found partially incomplete, with only postcranial elements preserved. In addition, there is evidence that some of the deposits, which had formed in this area during the Late Epigravettian, were displaced before the start of any archeological excavation. This activity was most likely carried out during historical periods in order to create room inside of the shelter, although it is also documented that in the years following the discovery of the site, the Late Epigravettian deposits continued to be dug by local people searching for bones and chert items. The mandible was not found in association with the human burial, but it was accidentally identified within disturbed deposits in the external area of the shelter. In the light of the genomic results presented here, it is likely that the mandible was initially part of the burial, but it was later displaced from the rock shelter through the non-archeological excavation activities mentioned above. Therefore, a probable explanation is that only one individual was buried in the Late Epigravettian deposits and, for this reason, the hypothesis that the postcranium and the mandible belong to the same individual can also be reconciled with the archeological evidence.
Regarding the discrepancy in the isotopic results between the two human remains, it is important to consider the isotopic composition of the faunal baseline. Despite a relatively low number of specimens (n = 10), the herbivores from the same site can be separated in two groups based on δ15N values: one ranging from +4.2 to +5.8‰, values close to the fauna dated to the same period from the Late Epigravettian layers of another Italian site (Grotta del Romito in Fig. 3)27, and another one ranging from +1.2 and +2.5‰ (Supplementary Data 2). Since δ15N values of plants and herbivores decrease with decreasing temperature, aridity, and soil microbial activity, the second group, which includes red deer and ibex in contrast to wild boar and large bovine, can be interpreted as the result of a high elevation habitat28. The fox specimen analyzed at Riparo Tagliente shows δ13C and δ15N values consistent with the consumption of high δ15N prey (Fig. 3), interpreted here as low-elevation prey. Interestingly, both human remains, Tagliente 1 and Romito 9, have likely consumed aquatic resources, based on their elevated δ15N values. Indeed, the archeological pikes from the Istrian site of Šandalja II29 illustrate well the typical lower δ13C and higher δ15N values of freshwater resources compared with terrestrial resources (Fig. 3). Since the rib of Tagliente 1 and the humerus/femur of Romito 9 did not provide depleted δ13C values, it was concluded that an additional marine contribution could not be ruled out6,27. In all cases, increased consumption of aquatic resources will result in increased δ15N values, which is not observed in the first molar of the Tagliente 2 mandible13 in comparison with both the rib and femur of Tagliente 1. Ribs and limb bones are considered to reflect the dietary signal of the last years of the life of a mature individual as their collagen is continuously remodeled. The difference in δ13C and δ15N values between the rib and the femur is indeed less than 0.5‰ and 0.9‰, respectively, as expected from two bones belonging to the same individual30. In contrast, the dentine of a first molar, whose sequential formation starts during the first year and is completed until the ninth year of life, will retain an early dietary signal that may be impacted by milk consumption before complete weaning. It is thus expected to observe higher δ13C and δ15N values in dentine compared with bone, independently of any other dietary change. Since the contrary is observed here, we can conclude that a higher aquatic consumption is observed in the collagen extracted from the rib and femur than from the molar suggesting a larger reliance on aquatic resources in adulthood rather than in childhood of the Tagliente 1/2 individual.
As mentioned before, an aquatic-based diet would lead to a reservoir effect on radiocarbon dating that could result in artificially older dates. However, a younger date was obtained for the rib and the femur compared to the mandible’s first molar (Fig. 2). Since both postcranial and dental elements belong to the same individual, a distinct aquatic consumption in different phases of life is not a plausible explanation for the difference in radiocarbon ages.
We thus investigated the possibility that 14C contamination could be the reason for the inconsistencies in the radiocarbon dates of different skeletal elements of the same individual. Compared to the 14C date obtained for the mandible, the previously and newly generated dates on the postcranium are 600 and 280 14C years younger, respectively (Fig. 2). The newly generated date on the femur was produced by following the exact same laboratory procedures as for the dating of the mandible, but two different MICADAS AMS spectrometers were implemented in the dating process. Bartolini et al.31, utilized an older generation machine, whereas our new date was obtained with a newer generation AMS spectrometer with enhanced transmission, as well as reduced and more stable blank control values32,33,34,35. However, despite these two dates being temporally closer compared to the date reported by Gazzoni et al.6, they still do not overlap and cannot be statistically combined after a 2-σ calibration (Fig. 2).
Given that the collagen extracted from the mandible’s tooth provided an older radiocarbon date and that 14C contamination is more likely to cause an artificially younger date36, we consider the date from Bortolini et al.5, to be the most reliable one. Thus, assuming the age obtained from the mandible as the true age of Tagliente 1/2, we calculated which levels of modern 14C contamination would be necessary to explain such chronological shifts. It was estimated that a modern 14C contamination of just 1.4% and 0.7% for the Gazzoni et al.6, date and the newly reported date, respectively, would be sufficient to make them artificially younger than the one in Bortolini et al.5, (Supplementary Data 1). The most likely source of collagen contamination is post-excavation treatments of the skeletal material for preservation purposes1. Further investigations revealed that the restoration work in the 1970s on Tagliente 1 involved the use of acetone for cleaning and the application of glue in areas with fractures (Benedetto Sala, pers. comm.), potentially influencing the dating accuracy. Based on visual inspection, a consolidant treatment of Tagliente 2 is also likely, but the post-excavation taphonomic history of the mandible is less certain and possibly different from Tagliente 1. Nevertheless, the older date obtained for Tagliente 2 implies that even if a consolidant was applied to the mandible for preservation purposes, it might have not infiltrated the tooth root that was used for radiocarbon dating. As far as stable isotopes are concerned, it has been shown that treatments with acetone not exceeding 48 hours do not impact the original collagen δ13C and δ15N values37. Only glue made with modern collagen may substantially influence carbon and nitrogen stable isotope values. However, to explain a positive shift on both δ13C and δ15N values, a marine origin of such a modern consolidant would be necessary (modern terrestrial specimens would provide lower δ13C values due to the Suess effect), and in a proportion that would much more strongly affect the radiocarbon dating results.
It is concerning that such minimal amounts of 14C contamination possibly introduced during restoration procedures could cause three skeletal elements belonging to the same individual to be radiocarbon-dated several generations apart. Given how common consolidation procedures have been performed for skeletal collections deriving from older excavations38, it is important to take specific radiocarbon dating results with great caution, especially when the post-excavation taphonomic history of the remains is not known or poorly recorded. Here, we provide an example of how assessing biological relationships through direct paleogenetic investigations of multiple isolated skeletal remains from the same archeological site is a powerful way to provide a reliable temporal constraint.
In summary, we re-investigated the Late Epigravettian human remains from Riparo Tagliente in Northern Italy. We examined the relationship between the Tagliente 1 human burial and the Tagliente 2 isolated mandible using an interdisciplinary approach. Integrating newly generated data through paleogenetics, radiocarbon dating, and stable isotope analyses with previously published archeological and paleoanthropological results, we conclude that Tagliente 1 and Tagliente 2 belong to the same human fossil that was renamed here “Tagliente 1/2”. We further provide compelling explanations to reconcile the apparent discrepancies that exist between the multiple lines of evidence. The separate ___location where the mandible was recovered can be explained by disturbance at the rock shelter during historical times and provide support for the conclusion that the mandible was initially part of a single human burial. Instead, the distinct stable isotopic signatures likely derived from different dietary practices during the lifetime of the individual while the non-overlapping radiocarbon dates might have been caused by low amounts of modern 14C contamination due to the use of consolidants. We conclude that it is crucial to apply paleogenetic techniques alongside other biomolecular disciplines in future bioarchaeological investigations of old osteological collections to provide novel insights into the existing Paleolithic fossil record.
Methods
Sampling
Biomolecular analyses of Tagliente 1 were performed after collecting an osteological fragment from the posterior distal region of the right femur. The specimen is curated at the Natural Science Museum in Verona and sampling permits were provided by the Soprintendenza Archeologia Belle Arti e Paesaggio for the Provinces of Verona, Rovigo, and Vicenza (prot. 0016605 23/06/2021, MIC_SABAP-VR_016). The procedure involved creating a space around the specimen to minimize the introduction of external contamination during the sampling procedures. After a thorough tomographic examination to precisely locate the targeted femur area, we proceeded with sampling. The operation started by marking the bone piece, cutting the cortex with a 20 mm diameter and 0.35 mm thickness diamond disc mounted on a red ring contra-angle handpiece, and a micro-motor set at 18 V (150,000 rpm). Once an area of ~2 × 0.5 cm was marked, the grooves were deepened using a 25 mm long and 0.5 mm diameter conical slotted diamond bur. The cuts were then finalized using a No. 12 surgical scalpel and a Buck 5/6 periodontal scalpel, detaching the bone piece completely. This controlled abrasion process allowed for the extraction of a bone fragment of ~1 g.
Radiocarbon dating
A sub-sample of around 500 mg was pretreated following the method described in Talamo et al.1. The sample underwent an ‘acid-base-acid’ (ABA) sequence designed to achieve decalcification, decontamination, and gelatinization of the bone chunks. This sequence was initiated with an initial step using hydrochloric acid (HCl 0.5 M), which was employed to dissolve mineral substances and certain organic impurities.
Once the CO2 effervescence ceased, the demineralized sample was rinsed once with ultrapure Milli-Q water. Subsequently, it was subjected to treatment with NaOH 0.1 M for 30 minutes at room temperature. Afterwards, the NaOH was replaced with ultrapure water for another rinsing step. Finally, the water in glass tubes was replaced with HCl 0.5 M, effectively re-acidifying the bone for an additional 15 minutes at room temperature.
Using a heater block, the resultant collagen was dissolved, turning into gelatin in acidic water (HCl pH 3) at 70°C for a duration of 20 hours. The gelatin obtained was initially filtered using Ezee-filter™ separators (Elkay Laboratory, UK) to eliminate particles smaller than 80μm. Following this step, ultrafiltration was performed to effectively separate low molecular weight contaminants and degraded proteins (<30 kDa) from the larger molecules (>30 kDa) (Sartorius VivaSpin® Turbo 15). Ezee-filters and ultrafilters were meticulously pre-cleaned to prevent any contamination risks associated with the filter membranes. After the ultrafiltration process, only the fractions with a molecular weight greater than 30 kDa were frozen for 24 hours and subsequently freeze-dried for 48 hours.
The collagen obtained was then subjected to graphitization at the BRAVHO lab39 using the Elemental vario ISOTOPE select coupled to the AGE 3 (Automated Graphitization Equipment, IonPlusAG, Switzerland)34. The resulting graphite target was sent to the Curt-Engelhorn-Center for Archaeometry in Mannheim, Germany (CEZA) measured on a MICADAS AMS31.
For quality control, an aliquot of a background bone sample (with a radiocarbon age exceeding 50,000 years) was subjected to pre-treatment and dating procedures together with the sample. This was done to monitor and account for any contamination that might have occurred during laboratory processes. The data reduction was carried out using the BATS software34. In accordance with standard practice, an additional 1‰ was incorporated into the error calculation of the sample. Radiocarbon date (reported as 14C years before present, 14C BP) was calibrated using OxCal 4.440 with the IntCal20 calibration curve41. The calibrated date ranges (calibrated years before present, cal BP) are reported at 1 and 2σ ranges, corresponding to a 68.3 and 95.4% probability level respectively. Uncalibrated 14C dates are presented with their associated 1σ errors (Supplementary Data 1).
We then estimated the level of potential 14C contamination in Tagliente 1. Rather than considering the age of the specimens, the Fraction of Modern Carbon-14 (F14C) value was used. F14C expresses the ratio of 14C to 12C in an ancient sample compared to the amount of 14C to 12C in a modern reference standard. A potential contamination in the sample causes a linear shift in the F14C value, which is expressed as a fraction or a percentage. Therefore, the percentage of 14C contamination can be calculated based on the deviation of the F14C value from the expected sample age1.
Stable isotope measurement
The elemental analyses (C, N) and isotopic measurements (δ13C, δ15N) were carried out at the CEZA lab in triplicate on the collagen extracted in the BRAVHO lab. The measurement of C and N content and isotopic composition was conducted using a vario ISOTOPE select elemental analyzer coupled to an Isoprime visION isotope ratio mass spectrometer (Elementar GmbH, Langenselbold). The international standards are a marine carbonate (V-Pee Dee Belemnite) for 13C and atmospheric nitrogen (AIR) for 15N. Isotopic measurements were calibrated using IAEA-CH-6 (δ13C:–10.45‰), IAEA-CH-7 (δ13C:–32.15‰), USGS40 (δ15N:–4.52‰) and USGS41a (δ15N: + 47.55‰). Measurement errors were smaller than ±0.1‰ for 13C and ±0.2‰ for 15N values (1σ) based on triplicate analysis of reference material (sulfanilamide, USGS-43, USGS-40, and USGS-41a). Reliability of the δ13C and δ15N values can be established by measuring the chemical composition of collagen, with C:N atomic ratio ranging from 2.9 to 3.642 and percentage of C and N above 8% and 3%, respectively43 (Fig. 3 and Supplementary Data 2).
Paleogenetic data generation and processing
DNA was extracted from 53 mg of bone powder following a protocol developed for ancient DNA44 and a genetic library was built from 25 µl of extract following the UDG-half protocol, where Uracil-DNA-Glycosidase partly reduces the damage on the aDNA molecules45. Negative and positive controls were taken along the entire workflow. Following indexing PCR and subsequent amplifications, the DNA library was both shotgun sequenced and enriched for the whole mitochondrial DNA (mtDNA) and 1.24 million SNPs (1240K capture) across the human genome46 and sequenced on Illumina platforms.
After demultiplexing of the raw sequences allowing for a maximum of one mismatch in each index, the EAGER (v1.92.55) pipeline was followed for data processing47. The pipeline incorporates AdapterRemoval (v2.2.0) to remove the adapters from both ends of the reads and fragments with a length below 30 bp48. Afterwards, BWA (v0.7.12) was used to align the sequences to the hg19 human reference genome49 with parameters -l 16500 and -q 30, as well as CircularMapper to realign mitochondrial sequences to the NC_012920.1 mtDNA reference47. After the alignment step, the pipeline utilizes DeDup (v0.12.2) to eliminate duplicate sequences that arise due to PCR amplification47. Finally, a preliminary authentication control is performed with mapDamage2.0 that quantifies average post-mortem damage using a Bayesian approach in both molecule termini50.
The 1240K capture data was trimmed at each end of the reads by two base pairs to avoid post-mortem DNA damage to introduce errors in our downstream analyses using the trimBam function of bamUtils (1.0.15)51. Using the trimmed data, we inferred the genetic sex of the individual by normalizing the coverages on the X and Y chromosomes with the autosomal coverage17.
We used ANGSD-based nuclear contamination estimate tool52 that makes use of the heterozygosity rate on the X-chromosome to quantify the possible human DNA contamination. However, due to low SNP coverage (45 SNPs covered twice on the X-chromosome), the nuclear contamination estimation should be treated with caution17. Instead, we used schmutzi53 with -notusepredC –uselength as parameters to jointly assemble a mtDNA consensus sequence and quantify the contamination in the mtDNA. This tool surveys the consensus endogenous mtDNA sequence for possible modern human mtDNA contaminants using a Eurasian database of 256 mtDNA sequences. We determined the mtDNA haplogroup using Haplogrep254, while Y-chromosome haplogroup was assigned with Yleaf55.
Genotyping involved randomly selecting an allele call from high-quality bases (with a Phred-scaled base quality score of at least 30) aligned to 1240K SNP positions that were covered at least once. This so-called pseudo-haploid calling is achieved via the pileupCaller tool (https://github.com/stschiff/sequenceTools). Following this, the genotyped data were combined with the relevant ancient populations5,16,17,24,56,57.
To ensure the reliability of our conclusions in subsequent analyses, we considered the presence of possible modern-day DNA contamination in our sample. We repeated all analyses on genotype data generated after using PMDtools. This tool specifically retains the deaminated DNA reads with CtoT substitutions towards the 3’ end or GtoA substitutions towards the 5’ end58. These filtered reads are regarded as originating from the ancient individual, thus minimizing possible influence of DNA contamination.
In order to quantify the background relatedness among individuals sharing similar ancestries with Tagliente 1, we computed a pairwise PMR calculation25 (Fig. 4 and Supplementary Data 3). To obtain a normalized allele-sharing estimate for a pair of samples, we computed f3-outgroup statistics using AdmixTools (version 7.0.1)59. For each statistic, we used Mbuti as the outgroup (Fig. 4 and Supplementary Data 4).
Statistics and reproducibility
Detailed information of the statistical analyses carried out as described in the methods section. All analyses can be reproduced by accessing the associated data linked in the Data Availability statement. Sample sizes were not determined in advance. In Fig. 4A, boxplots extend from the lower to upper quartile values of the PMR data with a line at the median. In Fig. 4B, standard errors for f3-outgroup statistics are calculated using the jackknife method outlined in the AdmixTools paper59.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
References
Talamo, S., Fewlass, H., Maria, R. & Jaouen, K. Here we go again”: the inspection of collagen extraction protocols for 14C dating and palaeodietary analysis. STAR. Sci. Technol. Archaeol. Res. 7, 62–77 (2021).
Olalde, I. & Posth, C. Latest trends in archaeogenetic research of west Eurasians. Curr. Opin. Genet. Dev. 62, 36–43 (2020).
Kistler, L., Ware, R., Smith, O., Collins, M. & Allaby, R. G. A new model for ancient DNA decay based on paleogenomic meta-analysis. Nucleic Acids Res. 45, 6310–6320 (2017).
Slon, V. et al. Neandertal and denisovan DNA from pleistocene sediments. Science 356, 605–608 (2017).
Bortolini, E. et al. Early Alpine occupation backdates westward human migration in Late Glacial Europe. Curr. Biol. 31, 2484–2493.e7 (2021).
Gazzoni, V. et al. Late Upper Palaeolithic human diet: first stable isotope evidence from Riparo Tagliente (Verona, Italy). Bull. Mem Soc. Anthropol. Paris 25, 103–117 (2013).
Bartolomei, G. et al. Una sepoltura epigravettiana nel deposito pleistocenico del Riparo Tagliente in Valpantena (Verona). Riv. Sci. Preistoriche 29, 101–152 (1974).
Bartolomei, G. et al. I depositi würmiani del Riparo Tagliente. Ann. DellUniv. Ferrara 15, 61–105 (1982).
Berto, C., Luzi, E., Canini, G. M., Guerreschi, A. & Fontana, F. Climate and landscape in Italy during Late Epigravettian. The Late Glacial small mammal sequence of Riparo Tagliente (Stallavena di Grezzana, Verona, Italy). Quat. Sci. Rev. 184, 132–142 (2018).
Cavallo, G. et al. Sourcing and processing of ochre during the late upper Palaeolithic at Tagliente rock-shelter (NE Italy) based on conventional X-ray powder diffraction analysis. Archaeol. Anthropol. Sci. 9, 763–775 (2017).
Fontana, F., Cilli, C. & Cremona, M. G. Recent data on the Late Epigravettian occupation at Riparo Tagliente, Monti Lessini (Grezzana, Verona): a multidisciplinary perspective. Preistoria Alp. 44, 49–57 (2009).
Fontana, F. et al. Advanced Studies On Early Human Adaptation In The Apennine Peninsula. p. 287–310 (Sidestone Press, 2018).
Oxilia, G. et al. Exploring late Paleolithic and Mesolithic diet in the Eastern Alpine region of Italy through multiple proxies. Am. J. Phys. Anthropol. 174, 232–253 (2021).
Arnaud, J. et al. A reexamination of the Middle Paleolithic human remains from Riparo Tagliente. Italy Quat. Int. 425, 437–444 (2016).
Corrain, C. Atti Del Reale Istituto Veneto Di Scienze, Lettere Ed Arti. p. 23–26 (Nabu Press, 1966).
Posth, C. et al. Palaeogenomics of Upper Palaeolithic to Neolithic European hunter-gatherers. Nature 615, 117–126 (2023).
Fu, Q. et al. The genetic history of Ice Age Europe. Nature 534, 200–205 (2016).
Hedges, R. E. M., Clement, J. G., Thomas, C. D. L. & O’Connell, T. C. Collagen turnover in the adult femoral mid‐shaft: modeled from anthropogenic radiocarbon tracer measurements. Am. J. Phys. Anthropol. 133, 808–816 (2007).
Bocherens, H. & Drucker, D. Trophic level isotopic enrichment of carbon and nitrogen in bone collagen: case studies from recent and ancient terrestrial ecosystems. Int. J. Osteoarchaeol. 13, 46–53 (2003).
Stuart, A. & Lister, A. Extinction chronology of the cave lion Panthera spelaea. Quat. Sci. Rev. 30, 2329–2340 (2011).
Posth, C. et al. Pleistocene mitochondrial genomes suggest a single major dispersal of Non-Africans and a late glacial population turnover in Europe. Curr. Biol. 26, 827–833 (2016).
Diroma, M. A. et al. New insights into mitochondrial DNA reconstruction and variant detection in ancient samples. Front. Genet. 12, 619950 (2021).
Modi, A. et al. More data on ancient human mitogenome variability in Italy: new mitochondrial genome sequences from three Upper Palaeolithic burials. Ann. Hum. Biol. 48, 213–222 (2021).
Yu, H. et al. Genomic and dietary discontinuities during the Mesolithic and Neolithic in Sicily. iScience 25, 104244 (2022).
Kennett, D. J. et al. Archaeogenomic evidence reveals prehistoric matrilineal dynasty. Nat. Commun. 8, 14115 (2017).
Corrain, Cleto. I resti scheletrici umani della sepoltura epigravettiana del Riparo Tagliente in Valpantena (Verona). Boll Mus Civ Storia Nat Verona 4, 35–79 (1977).
Craig, O. E. et al. Stable isotope analysis of Late Upper Palaeolithic human and faunal remains from Grotta del Romito (Cosenza), Italy. J. Archaeol. Sci. 37, 2504–2512 (2010).
Männel, T. T., Auerswald, K. & Schnyder, H. Altitudinal gradients of grassland carbon and nitrogen isotope composition are recorded in the hair of grazers. Glob. Ecol. Biogeogr. 16, 583–592 (2007).
Richards, M. P., Karavanić, I., Pettitt, P. & Miracle, P. Isotope and faunal evidence for high levels of freshwater fish consumption by Late Glacial humans at the Late Upper Palaeolithic site of Šandalja II, Istria, Croatia. J. Archaeol. Sci. 61, 204–212 (2015).
Clauzel, T. et al. Climate conditions and dietary practices during the Second Iron Age studied through the multi-isotope analysis of bones and teeth from individuals of Thézy-Glimont, Picardie, France. Archaeol. Anthropol. Sci. 14, 61 (2022).
Kromer, B., Lindauer, S., Synal, H.-A. & Wacker, L. MAMS – a new AMS facility at the Curt-Engelhorn-Centre for Achaeometry, Mannheim, Germany. Nucl. Instrum. Methods Phys. Res. Sect. B Beam Interact. Mater. 294, 11–13 (2013).
Schulze-König, T., Seiler, M., Suter, M., Wacker, L. & Synal, H.-A. The dissociation of 13CH and 12CH2 molecules in He and N2 at beam energies of 80–250keV and possible implications for radiocarbon mass spectrometry. Nucl. Instrum. Methods Phys. Res. Sect. B Beam Interact. Mater. 269, 34–39 (2011).
Maxeiner, S., Seiler, M., Suter, M. & Synal, H.-A. Charge state distributions and charge exchange cross sections of carbon in helium at 30–258keV. Nucl. Instrum. Methods Phys. Res. Sect. B Beam Interact. Mater. 361, 541–547 (2015).
Wacker, L., Němec, M. & Bourquin, J. A revolutionary graphitisation system: fully automated, compact and simple. Nucl. Instrum. Methods Phys. Res. Sect. B Beam Interact. Mater. 268, 931–934 (2010).
Wacker, L. et al. MICADAS: routine and high-precision radiocarbon dating. Radiocarbon 52, 252–262 (2010).
Prüfer, K. et al. A genome sequence from a modern human skull over 45,000 years old from Zlatý kůň in Czechia. Nat. Ecol. Evol. 5, 820–825 (2021).
Moreiras Reynaga, D. K., Munizzi, J. S., McMillan, R., Millaire, J.-F. & Longstaffe, F. J. Effects of consolidants and their removal by polar solvents on the stable isotope compositions of bone. Quat. Int. 660, 31–41 (2023).
Schellmann, N. C. Animal glues: a review of their key properties relevant to conservation. Stud. Conserv. 52, 55–66 (2007).
Tassoni, L. et al. Safe preparation and delivery of graphite targets for 14C analysis: procedures of BRAVHO lab at Bologna University. Radiocarbon https://doi.org/10.1017/RDC.2023.43 (2023).
Ramsey, C. B. Dealing with outliers and offsets in radiocarbon dating. Radiocarbon 51, 1023–1045 (2009).
Reimer, P. J. et al. The IntCal20 Northern Hemisphere radiocarbon age calibration curve (0–55 cal kBP). Radiocarbon 62, 725–757 (2020).
DeNiro, M. J. Postmortem preservation and alteration of in vivo bone collagen isotope ratios in relation to palaeodietary reconstruction. Nature 317, 806–809 (1985).
Ambrose, S. H. Preparation and characterization of bone and tooth collagen for isotopic analysis. J. Archaeol. Sci. 17, 431–451 (1990).
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).
Rohland, N., Harney, E., Mallick, S., Nordenfelt, S. & Reich, D. Partial uracil–DNA–glycosylase treatment for screening of ancient DNA. Philos. Trans. R. Soc. B Biol. Sci. 370, 20130624 (2015).
Fu, Q. et al. An early modern human from Romania with a recent Neanderthal ancestor. Nature 524, 216–219 (2015).
Peltzer, A. et al. EAGER: efficient ancient genome reconstruction. Genome Biol. 17, 60 (2016).
Schubert, M., Lindgreen, S. & Orlando, L. AdapterRemoval v2: rapid adapter trimming, identification, and read merging. BMC Res. Notes 9, 88 (2016).
Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinform. Oxf. Engl. 25, 1754–1760 (2009).
Jónsson, H., Ginolhac, A., Schubert, M., Johnson, P. L. F. & Orlando, L. mapDamage2.0: fast approximate Bayesian estimates of ancient DNA damage parameters. Bioinformatics 29, 1682–1684 (2013).
Jun, G., Wing, M. K., Abecasis, G. R. & Kang, H. M. An efficient and scalable analysis framework for variant extraction and refinement from population-scale DNA sequence data. Genome Res. 25, 918–925 (2015).
Korneliussen, T. S., Albrechtsen, A. & Nielsen, R. ANGSD: analysis of next generation sequencing data. BMC Bioinform. 15, 356 (2014).
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).
Weissensteiner, H. et al. HaploGrep 2: mitochondrial haplogroup classification in the era of high-throughput sequencing. Nucleic Acids Res. 44, W58–W63 (2016).
Ralf, A., Montiel González, D., Zhong, K. & Kayser, M. Yleaf: software for human Y-chromosomal haplogroup inference from next-generation sequencing data. Mol. Biol. Evol. 35, 1291–1294 (2018).
Antonio, M. L. et al. Ancient Rome: a genetic crossroads of Europe and the Mediterranean. Science 366, 708–714 (2019).
Mathieson, I. et al. The genomic history of southeastern Europe. Nature 555, 197–203 (2018).
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).
Patterson, N. et al. Ancient admixture in human history. Genetics 192, 1065–1093 (2012).
Acknowledgements
The authors thank Maria A. Spyrou, Vitale S. Sparacello and the Archeo- and Paleogenetics group at the University of Tübingen for comments; Ronny Friedrich for discussion of the AMS results; Corina Knipper for detailed information about isotopic measurements; the Natural Science Museum in Verona and Prof. Antonio Guerreschi who assisted the sampling operations; the Italian Ministry of Culture and Soprintendenza Archeologia Belle Arti e Paesaggio for the Provinces of Verona, Rovigo, and Vicenza for granting access to the human remains of Tagliente 1; O.E.Y. and C.P. have received funding from the Deutsche Forschungsgemeinschaft (DFG 497783672). S.T. has received funding from the European Research Council under the European Union’s Horizon 2020 Research and Innovation Programme (grant agreement No. 803147 RESOLUTION).
Funding
Open Access funding enabled and organized by Projekt DEAL.
Author information
Authors and Affiliations
Contributions
F.F., S.B., and C.P. designed the project. G.O. and S.S. sampled the material, O.E.Y., E.R., and C.P. performed aDNA laboratory work and analyses, S.T. and L.T. conducted and analyzed radiocarbon dates, D.G.D. interpreted the isotopic results, G.O. revisited the physical anthropological assessments, F.F. provided archeological descriptions. O.E.Y. and C.P. wrote the manuscript with substantial input on the associated sections from G.O., D.G.D., S.T., F.F., and all other co-authors. All authors reviewed and approved the manuscript.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Communications Biology thanks the anonymous reviewers for their contribution to the peer review of this work. Primary Handling Editors: Luke Grinham and Dario Ummarino.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Yavuz, O.E., Oxilia, G., Silvestrini, S. et al. Biomolecular analysis of the Epigravettian human remains from Riparo Tagliente in northern Italy. Commun Biol 7, 1415 (2024). https://doi.org/10.1038/s42003-024-06979-9
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s42003-024-06979-9
