Post by Admin on Apr 25, 2021 2:49:54 GMT
From a population perspective, we performed a multi-dimensional scaling (MDS) analysis based on outgroup f3 distances (Figure 4A) and found the sample to fall within the broader European Western Hunter-Gatherer (WHG) genetic variation, pointing to an affinity to the previously described Villabruna Cluster. This group has been defined on the basis of genetic affinity among individuals known to have largely replaced previous European Hunter-Gatherer populations at least ∼14 ka ago. Notably, these individuals showed little or no trace of genetic contributions from pre-existing European groups, such as the ones genetically close to the Dolní Věstonice or to Goyet-Q116-1 human remains and known to have inhabited Europe until the LGM.9 One of the defining features of the Villabruna Cluster is a higher affinity with Near Eastern genetic components than that exhibited by pre-existing Palaeolithic West Eurasians. The significantly negative f4 test (Kostenki14, Tagliente2, Druze, Mbuti: f4 = −0.0037; standard error = 0.00063; Z = −5.88) further confirmed Tagliente2 to share genetic features with the Villabruna Cluster and to be in discontinuity with the preceding European genetic background. We followed up this observation using a series of f4 tests in the form Tagliente2, X; Y, Mbuti, where Y is a population of interest and X is either Villabruna (∼14 ka ago),9 Bichon (∼13.7 ka ago),9 or a Mesolithic Italian from Grotta Continenza (∼11.9 ka ago; Figure 4B).26 We chose three independent WHG samples to control for potential biases introduced by the genotyping strategy and indeed found small discrepancies when we compared results obtained using capture (Villabruna) or shotgun (Bichon and Continenza) data. To minimize this effect, we chose to put more weight on the interpretation of shotgun results, deemed to be more readily comparable with the shotgun data generated for Tagliente2 in this study. We also show that the data available are sufficient to achieve significance in a f4 test when the order of X and Y populations are inverted, as for Tagliente, Y, Grotta Continenza, Mbuti (Figure S2). The higher affinity of Continenza and Bichon to later WHG (Loschbour, Iberia_HG and Continenza, Bichon, and Villabruna themselves) when compared to Tagliente2 may be explained by the more ancient age of Tagliente2 or with the former individuals being genetically closer to the ancestry that reached central Europe at least by 14 ka ago. Alternatively, the higher affinity emerging among more recent WHG samples may also be ascribed to subsequent admixtures between the newly arrived Tagliente2 individuals and pre-existing Dolní Věstonice- or Goyet Q116-1-like (∼35 ka ago)9 genetic substrates, as already reported for Loschbour (∼8.1 ka ago).9,27 We then modeled the position of Tagliente2 within the tree proposed by Fu and colleagues9 (Figure S3A) and found that it may fit well within the Villabruna branch, confirming previous results (Figures S3B–S3D). To minimize the effects of the mismatch between capture and shotgun data due to attraction, we also explore the feasibility of the basic qpGraph (Figure S3B) using, where possible, shotgun samples (Figure S4).
Figure 4
Demographic inference from Tagliente2
(A) Multi-dimensional scaling based on Mbuti; X, Tagliente2 outgroup f3 statistics show Tagliente2 (golden star) to cluster within the Villabruna Cluster (in blue) and away from the pre-existing South European samples (in green).
(B) f4 tests (Tagliente2, X, Y, Mbuti ± 3 SE) where X is either a Mesolithic Italian, Villabruna, or BC WHG sample and Y, shown along the y axis, is a population of interest (Data S1C; Figures S2–S4).
With our work, we provide genomic, uniparental, and chronological evidence that backdates the presence of the so called Villabruna component in Northern Italy to as early as 17 ka ago, when it chronologically overlaps with major cultural transitions involving the region. The shift from Early to Late Epigravettian has not been abrupt, and despite the emergence of regionalism and environmental/cultural differences between Adriatic and Tyrrhenian contexts,12 change in the relative frequency of artifact types and reduction sequences, as well as in raw material procurement and settlement patterns, can be recorded since ∼17 ky cal BP.4,5,12,28, 29, 30, 31, 32, 33, 34 After 14 ky cal BP, a more-marked discontinuity is attested by the greater reliance on geometric microlithic pieces and a stronger presence of engraved and painted bones and stones bearing linear, geometric, zoomorphic, or anthropomorphic motives.28,30,34
The Early to Late Epigravettian transition is broadly coeval to a marked retreat of Alpine glaciers (ca. 18–17.5 ka ago)35 after they had reached their maximum between 26 and 24 ka ago13 and to a rapid rise in sea levels since 16.5 ka ago.1 These processes led to considerable change in the geomorphology of the Alpine sectors and stabilized large surfaces of the Great Adriatic/Po Region.12 A rapid forest recolonization of the Alpine foothills started about 17 ka cal BP, well before the major Bølling/Allerød warmup.6,36,37 Alpine forelands became (open) pine forests with tree Betula and Larix (Figure 1C)3 while open vegetations developed in the distal sector of the megafans.12,38 At the end of LGM, local faunal availability was limited and mostly consisted of species adapted to open environments that had been able to find their climatic optimum by moving to higher elevations in warmer interstadials (e.g., Alpine ibex and marmot) or taking refuge (e.g., elk) in areas with favorable microclimates. Most Central European, cold-adapted megafaunal species had entered northern Italy before the onset of LGM through the central Slovenian corridor—the same used by Gravettian hunter-gatherers to follow game and settle across the Adriatic.12 During LGM, both new arrivals and northward movements of these large mammals were hampered, and they either locally disappeared or became extinct in association with short LGM interstadials (as in the case of cave bears, between 24.2 and 23.5 ka ago).21 The reduction of the forest cover during colder phases is confirmed by the presence of Capra ibex, Rupicapra rupicapra, and Marmota marmota, both in the core of the Po Plain and in caves and shelters of Berici Hills.39,40 In the same area, archaeological records show the presence of palaeoarctic birds41 currently only found in high-latitude regions of the northern hemisphere.
Taken together, our results support two different—although not mutually exclusive—scenarios. The first one involves a broad network of refugia connecting Mediterranean and Eastern Europe during and immediately after the LGM. The network could have facilitated long-distance transmission through a stepwise exchange of both cultural and genetic information from the Black Sea all the way to the Iberian Peninsula. This scenario, which cannot be tested with the available evidence, would predict a relationship between the age of a sample, its geographic location, and the relative abundance of Villabruna genetic components shared with present-day Near Easterners. From a cultural point of view, the development of Early and Late Epigravettian material culture in Southern Europe would not be directly driven by abrupt, millennial-scale climatic events and could result from convergence, local adaptation, and cultural diffusion without entailing population movement. In this case, geographic distance between sites would also predict similarity in lithic assemblages.
The second scenario implies instead population movement and replacement, a more-abrupt genetic turnover, and a distribution of both genetic and cultural similarities that is not well predicted by geographic clines. This population shift could have taken place during the LGM, i.e., after ∼27 ka ago, when Věstonice-like components were still visible at Ostuni (Southern Italy) and before ∼17 ka ago. Groups bearing the Villabruna lineage may have exploited the Slovenian corridor and lower Adriatic Sea levels to occupy Adriatic Italy up to the Po plain and recolonized pre-Alpine valleys only at a later stage. According to this model, after ∼27 ka ago, genetic lineages found in Italy and only later in France and Spain, should show either or both of (1) genomic affinities with the Villabruna Cluster and (2) uniparental lineages belonging to U2′3′4’7’8’9 /U5b mtDNA and/or I2/R1a Y groups. According to this model, cultural change recorded in Italy across the Epigravettian sequence may have been at least in part triggered by demic processes linked to population replacement.
From a cultural perspective, the biased spatial and temporal distribution of the available archaeological record can hardly be used to directly discriminate between these two models, and there is still considerable uncertainty on the temporal dynamics underlying the beginning of the Late Epigravettian across the whole of Italy.12 Since ∼18 to 17 cal kBP, there is evidence of a shift from Solutrean to Magdalenian material culture in Southwestern Europe and from Early to Late Epigravettian material culture in a vast area ranging from the Rhone river to the Southern Russian plain.12,30,42,43 Environmental pressure conditioned the movement of megafauna during the LGM and limited the movement of human groups to corridors connecting Southern Europe, the Balkans, and Eastern Europe. In this period, human groups inhabiting Southern Europe were exposed to limited ecological risk compared to other regions of Central and Northern Europe.2,44, 45, 46 Variability in Sr isotope composition of individuals uncovered at Grotta Paglicci (Apulia, Southern Italy) shows a conspicuous change in residential mobility patterns and adaptive strategies between Gravettian and Early Epigravettian hunter-gatherers. Given the lack of any contextual evidence for change in climate, these differences are imputed to cultural factors and possibly linked to population replacement that may have taken place already at an early stage of the Epigravettian sequence.47
Similarity in the distribution of artifact types between Italy and the Balkans, on the other hand, supports the possible expansion of techno-complexes from Central Europe via eastern/Slovenian routes and hint at long-distance mobility for Epigravettian hunter-gatherers.12 The same pattern, however, has been used to challenge this view in favor of a social network hypothesis.31 Similarity between Balkan and Italian contexts is documented from the Gravettian to the Mesolithic, and contacts involved raw lithic materials, marine molluscs, ornamental beads, clay figurines, decorative motives, and lithic technology.5,30,32,48 At the same time, the reliability of some cultural markers (such as shouldered points) as proxies of human movement/interaction has been recently questioned.12
Uniparental genetic markers might be useful to disentangle the two proposed scenarios. The majority of samples attributed to the Villabruna Cluster (Figure 3) on a solid genetic and chronological basis share mtDNA and chrY belonging to a limited number of lineages. Reduced diversity within the uniparental Villabruna landscape is consistent either with a bottleneck interrupting the network or with a founder event in a broader scenario of population shift.10 This uniparental landscape, paired with evidence for seamless cultural exchange across the Adriatic12 and increased affinity with eastern genomic components, makes genetic replacement the most likely explanation for our results. The presence of Paglicci71 (∼18.5 ka ago)10 within the Villabruna maternal lineage advocates for a founder event in Southern Italy as early or even before ∼18.5 ka ago, followed by a later expansion in Northern Italy at the beginning of deglaciation. In this light, Tagliente2 may be seen as an early settler of the Southern Alpine region, perhaps explaining the basal mtDNA lineage of this sample. The possibility of an earlier connection between Eastern and Western Europe through southern corridors, either in the form of an extended LGM network or as an early arrival of Villabruna-like individuals in Western Europe, is also supported by El Mirón, an Iberian sample dated to ∼18.7 ka ago9,12,27 presenting with an admixture of Magdalenian, Goyet-2-like ancestry and ancestry related to the Villabruna cluster. The most parsimonious interpretation of the emerging genetic scenario suggests that cumulative cultural change observed in Southern Europe from the end of LGM to the end of the Younger Dryas (∼11.7 ka ago) was at least in part triggered by gene flow from southeastern refugia into Italy. This process, in its early stage and to the south of the Alps, was independent of the later Bølling-Allerød event and contributed to the gradual replacement of pre-LGM ancestry across the Italian peninsula49 and beyond. Further genetic evidence from Southern European contexts dated between ∼27 and 19 ka ago and analyses of cultural similarity between Italy and putative sources of population movement, however, will be needed in the future to test this hypothesis.
In conclusion, Tagliente2 provides evidence that the major migrations which strongly affected the genetic background of all Europeans7, 8, 9, 10, 11 started considerably earlier in Southern Europe than previously reported and were already in place in this region during the cold phase following the LGM peak, possibly favored by stepwise reductions of glacier extent and forest expansion preceding the Bølling-Allerød rapid warming. At this stage, Southern Europe, the Balkans, and Eastern Europe/Western Asia were already connected into the same network of potential LGM refugia and exchanged both genes and cultural information, posing the basis for the observed population replacement. This finding also backdates previous conclusions concerning a plausible demic component to change over time in the coeval material culture of Southern Europe7, 8, 9, 10, 11 and temporally locates this process at the transition between Early and Late Epigravettian or even possibly at the very beginning of the Epigravettian sequence.
Figure 4
Demographic inference from Tagliente2
(A) Multi-dimensional scaling based on Mbuti; X, Tagliente2 outgroup f3 statistics show Tagliente2 (golden star) to cluster within the Villabruna Cluster (in blue) and away from the pre-existing South European samples (in green).
(B) f4 tests (Tagliente2, X, Y, Mbuti ± 3 SE) where X is either a Mesolithic Italian, Villabruna, or BC WHG sample and Y, shown along the y axis, is a population of interest (Data S1C; Figures S2–S4).
With our work, we provide genomic, uniparental, and chronological evidence that backdates the presence of the so called Villabruna component in Northern Italy to as early as 17 ka ago, when it chronologically overlaps with major cultural transitions involving the region. The shift from Early to Late Epigravettian has not been abrupt, and despite the emergence of regionalism and environmental/cultural differences between Adriatic and Tyrrhenian contexts,12 change in the relative frequency of artifact types and reduction sequences, as well as in raw material procurement and settlement patterns, can be recorded since ∼17 ky cal BP.4,5,12,28, 29, 30, 31, 32, 33, 34 After 14 ky cal BP, a more-marked discontinuity is attested by the greater reliance on geometric microlithic pieces and a stronger presence of engraved and painted bones and stones bearing linear, geometric, zoomorphic, or anthropomorphic motives.28,30,34
The Early to Late Epigravettian transition is broadly coeval to a marked retreat of Alpine glaciers (ca. 18–17.5 ka ago)35 after they had reached their maximum between 26 and 24 ka ago13 and to a rapid rise in sea levels since 16.5 ka ago.1 These processes led to considerable change in the geomorphology of the Alpine sectors and stabilized large surfaces of the Great Adriatic/Po Region.12 A rapid forest recolonization of the Alpine foothills started about 17 ka cal BP, well before the major Bølling/Allerød warmup.6,36,37 Alpine forelands became (open) pine forests with tree Betula and Larix (Figure 1C)3 while open vegetations developed in the distal sector of the megafans.12,38 At the end of LGM, local faunal availability was limited and mostly consisted of species adapted to open environments that had been able to find their climatic optimum by moving to higher elevations in warmer interstadials (e.g., Alpine ibex and marmot) or taking refuge (e.g., elk) in areas with favorable microclimates. Most Central European, cold-adapted megafaunal species had entered northern Italy before the onset of LGM through the central Slovenian corridor—the same used by Gravettian hunter-gatherers to follow game and settle across the Adriatic.12 During LGM, both new arrivals and northward movements of these large mammals were hampered, and they either locally disappeared or became extinct in association with short LGM interstadials (as in the case of cave bears, between 24.2 and 23.5 ka ago).21 The reduction of the forest cover during colder phases is confirmed by the presence of Capra ibex, Rupicapra rupicapra, and Marmota marmota, both in the core of the Po Plain and in caves and shelters of Berici Hills.39,40 In the same area, archaeological records show the presence of palaeoarctic birds41 currently only found in high-latitude regions of the northern hemisphere.
Taken together, our results support two different—although not mutually exclusive—scenarios. The first one involves a broad network of refugia connecting Mediterranean and Eastern Europe during and immediately after the LGM. The network could have facilitated long-distance transmission through a stepwise exchange of both cultural and genetic information from the Black Sea all the way to the Iberian Peninsula. This scenario, which cannot be tested with the available evidence, would predict a relationship between the age of a sample, its geographic location, and the relative abundance of Villabruna genetic components shared with present-day Near Easterners. From a cultural point of view, the development of Early and Late Epigravettian material culture in Southern Europe would not be directly driven by abrupt, millennial-scale climatic events and could result from convergence, local adaptation, and cultural diffusion without entailing population movement. In this case, geographic distance between sites would also predict similarity in lithic assemblages.
The second scenario implies instead population movement and replacement, a more-abrupt genetic turnover, and a distribution of both genetic and cultural similarities that is not well predicted by geographic clines. This population shift could have taken place during the LGM, i.e., after ∼27 ka ago, when Věstonice-like components were still visible at Ostuni (Southern Italy) and before ∼17 ka ago. Groups bearing the Villabruna lineage may have exploited the Slovenian corridor and lower Adriatic Sea levels to occupy Adriatic Italy up to the Po plain and recolonized pre-Alpine valleys only at a later stage. According to this model, after ∼27 ka ago, genetic lineages found in Italy and only later in France and Spain, should show either or both of (1) genomic affinities with the Villabruna Cluster and (2) uniparental lineages belonging to U2′3′4’7’8’9 /U5b mtDNA and/or I2/R1a Y groups. According to this model, cultural change recorded in Italy across the Epigravettian sequence may have been at least in part triggered by demic processes linked to population replacement.
From a cultural perspective, the biased spatial and temporal distribution of the available archaeological record can hardly be used to directly discriminate between these two models, and there is still considerable uncertainty on the temporal dynamics underlying the beginning of the Late Epigravettian across the whole of Italy.12 Since ∼18 to 17 cal kBP, there is evidence of a shift from Solutrean to Magdalenian material culture in Southwestern Europe and from Early to Late Epigravettian material culture in a vast area ranging from the Rhone river to the Southern Russian plain.12,30,42,43 Environmental pressure conditioned the movement of megafauna during the LGM and limited the movement of human groups to corridors connecting Southern Europe, the Balkans, and Eastern Europe. In this period, human groups inhabiting Southern Europe were exposed to limited ecological risk compared to other regions of Central and Northern Europe.2,44, 45, 46 Variability in Sr isotope composition of individuals uncovered at Grotta Paglicci (Apulia, Southern Italy) shows a conspicuous change in residential mobility patterns and adaptive strategies between Gravettian and Early Epigravettian hunter-gatherers. Given the lack of any contextual evidence for change in climate, these differences are imputed to cultural factors and possibly linked to population replacement that may have taken place already at an early stage of the Epigravettian sequence.47
Similarity in the distribution of artifact types between Italy and the Balkans, on the other hand, supports the possible expansion of techno-complexes from Central Europe via eastern/Slovenian routes and hint at long-distance mobility for Epigravettian hunter-gatherers.12 The same pattern, however, has been used to challenge this view in favor of a social network hypothesis.31 Similarity between Balkan and Italian contexts is documented from the Gravettian to the Mesolithic, and contacts involved raw lithic materials, marine molluscs, ornamental beads, clay figurines, decorative motives, and lithic technology.5,30,32,48 At the same time, the reliability of some cultural markers (such as shouldered points) as proxies of human movement/interaction has been recently questioned.12
Uniparental genetic markers might be useful to disentangle the two proposed scenarios. The majority of samples attributed to the Villabruna Cluster (Figure 3) on a solid genetic and chronological basis share mtDNA and chrY belonging to a limited number of lineages. Reduced diversity within the uniparental Villabruna landscape is consistent either with a bottleneck interrupting the network or with a founder event in a broader scenario of population shift.10 This uniparental landscape, paired with evidence for seamless cultural exchange across the Adriatic12 and increased affinity with eastern genomic components, makes genetic replacement the most likely explanation for our results. The presence of Paglicci71 (∼18.5 ka ago)10 within the Villabruna maternal lineage advocates for a founder event in Southern Italy as early or even before ∼18.5 ka ago, followed by a later expansion in Northern Italy at the beginning of deglaciation. In this light, Tagliente2 may be seen as an early settler of the Southern Alpine region, perhaps explaining the basal mtDNA lineage of this sample. The possibility of an earlier connection between Eastern and Western Europe through southern corridors, either in the form of an extended LGM network or as an early arrival of Villabruna-like individuals in Western Europe, is also supported by El Mirón, an Iberian sample dated to ∼18.7 ka ago9,12,27 presenting with an admixture of Magdalenian, Goyet-2-like ancestry and ancestry related to the Villabruna cluster. The most parsimonious interpretation of the emerging genetic scenario suggests that cumulative cultural change observed in Southern Europe from the end of LGM to the end of the Younger Dryas (∼11.7 ka ago) was at least in part triggered by gene flow from southeastern refugia into Italy. This process, in its early stage and to the south of the Alps, was independent of the later Bølling-Allerød event and contributed to the gradual replacement of pre-LGM ancestry across the Italian peninsula49 and beyond. Further genetic evidence from Southern European contexts dated between ∼27 and 19 ka ago and analyses of cultural similarity between Italy and putative sources of population movement, however, will be needed in the future to test this hypothesis.
In conclusion, Tagliente2 provides evidence that the major migrations which strongly affected the genetic background of all Europeans7, 8, 9, 10, 11 started considerably earlier in Southern Europe than previously reported and were already in place in this region during the cold phase following the LGM peak, possibly favored by stepwise reductions of glacier extent and forest expansion preceding the Bølling-Allerød rapid warming. At this stage, Southern Europe, the Balkans, and Eastern Europe/Western Asia were already connected into the same network of potential LGM refugia and exchanged both genes and cultural information, posing the basis for the observed population replacement. This finding also backdates previous conclusions concerning a plausible demic component to change over time in the coeval material culture of Southern Europe7, 8, 9, 10, 11 and temporally locates this process at the transition between Early and Late Epigravettian or even possibly at the very beginning of the Epigravettian sequence.