|
Post by Admin on Nov 28, 2020 19:42:53 GMT
The mixed genetic origin of the first farmers of Europe 41 Summary 42 While the Neolithic expansion in Europe is well described archaeologically, the genetic origins 43 of European first farmers and their affinities with local hunter-gatherers (HGs) remain unclear. 44 To infer the demographic history of these populations, the genomes of 15 ancient individuals 45 located between Western Anatolia and Southern Germany were sequenced to high quality, 46 allowing us to perform population genomics analyses formerly restricted to modern genomes. 47 We find that all European and Anatolian early farmers descend from the merging of a European 48 and a Near Eastern group of HGs, possibly in the Near East, shortly after the Last Glacial 49 Maximum (LGM). Western and Southeastern European HG are shown to split during the LGM, 50 and share signals of a very strong LGM bottleneck that drastically reduced their genetic 51 diversity. Early Neolithic Central Anatolians seem only indirectly related to ancestors of 52 European farmers, who probably originated in the Near East and dispersed later on from the 53 Aegean along the Danubian corridor following a stepwise demic process with only limited (2- 54 6%) but additive input from local HGs.
55 Our analyses provide a time frame and resolve the genetic origins of early European farmers. 56 They highlight the impact of Late Pleistocene climatic fluctuations that caused the 57 fragmentation, merging and reexpansion of human populations in SW Asia and Europe, and 58 eventually led to the world's first agricultural populations.
59 Introduction 60 The origins and spread of agriculture in Southwest (SW) Asia, often described as the ‘Neolithic 61 transition’, have been under research for well over a century1. While early sedentary 62 communities emerged at the end of the Pleistocene2, crop cultivation and ungulate management 63 developed in the Fertile Crescent after the Younger Dryas cold spell (12.9-11.7 kya) with the 64 onset of warmer conditions at the beginning of the Holocene3–5. Starting 10.6 kya, shifts 65 towards small-scale agriculture with imported cultivars and livestock management are 66 observed among sedentarizing communities of Central Anatolia6,7 (Fig. 1). Around 8.7-8.6 kya, 67 farming traditions expanded into the wider Aegean region, including the western half of the 68 Anatolian Peninsula8,9. The new subsistence economy reached Crete and the Greek mainland 69 shortly thereafter10,11. From the Aegean, farming spread into the Central Mediterranean Basin 70 and the Danubian corridor, reaching the Central Balkans by about 8.2 kya, and Austria and 71 Southern Germany by 7.5 kya12.
72 Even though agriculture was first invented in SW Asia, the genetic origins of Europe’s first 73 farmers remain elusive. Recent palaeogenetic findings revealed that most European farmers 74 are genetically closer to Central and Northwestern (NW) Anatolian farmers than to Pre-Pottery 75 Neolithic (PPN) farmers of the Southern Levant or the Zagros region of Western Iran, who 76 were genetically well differentiated13–18. However, these findings mostly rely on a set of 77 ascertained genomic sites19 that cannot easily be used for demographic reconstruction, and the 78 temporal framework they provide depends on the dating of tested samples16.
79 In order to characterize the demographic history and origin of European and Anatolian farmers, 80 we generated high quality palaeogenomes from two Mesolithic hunter-gatherers (HGs) and 13 81 Early Neolithic farmers (mean depth between 10.55X and 15.21X, Table 1). These individuals 82 were chosen on a regular spatial and temporal gradient along the main expansion axis of the 83 Neolithic from the Near East into Central Europe (Fig. 1). We combined these data with nine 84 ancient genomes of similar high quality available for this region and period14,17,18,20–24 85 (Fig. 1, S11a, Table S4). We used these complete ancient genomes to perform model-based 86 demographic inference based on the site frequency spectrum (SFS) at neutral sites25. We thus 87 obtained a precise scenario of the colonization of Europe by early farmers and their interactions 88 with local HGs, and estimated population size changes, interactions and split times with high 89 accuracy.
|
|
|
Post by Admin on Nov 29, 2020 0:17:05 GMT
Figure 1 - Spatial and temporal distribution of the palaeogenomes analyzed in this study. 90 Results 91 Patterns of genomic diversity along the Danubian corridor 92 The genetic structure and affinities of ancient individuals 93 Both multidimensional scaling (MDS) performed on the neutral portion of ancient and modern 94 genomes (Fig. 2a) and an admixture analysis (Fig. S21) revealed three main clusters of ancient 95 samples, which are found overall much more differentiated than modern individuals of SW 96 Asia and Europe: i) a cluster of European HGs, ii) a cluster of Early Neolithic individuals from 97 Iran, here represented by a single genome from Wezmeh Cave, WC1, and iii) a cluster with all 98 other Holocene individuals. In keeping with previous analyses based on a restricted set of 99 SNPs26,27, this MDS analysis (Fig. 2a) suggests strongest affinities of European and NW 100 Anatolian Neolithic samples with modern Sardinians, with the exception of the Early Neolithic 101 NW Anatolian individual Bar8 found to be closer to modern individuals from Greece, Albania 102 and other individuals from Southern Europe. However, a MDS analysis performed on the 103 whole genome including sites potentially affected by selection (Fig. S19) rather suggests that 104 early farmers are closer to Southern Europeans other than Sardinians. The Early Neolithic 105 individual from Iran (WC1) shows strongest genetic affinities with modern Iranians (Fig. 2a), 106 and to a lesser extent with individuals from the Northern Caucasus, suggesting some genetic 107 continuity in Iran since Neolithic times. Finally, Palaeolithic and Mesolithic HGs are generally 108 distinct from all modern SW Asians and Europeans, the closest of whom are Baltic Sea 109 individuals, Russians and Scandinavians.
110 Early farmers are genetically more diverse and decline in stature over time 111 While genetic diversity as quantified by the heterozygosity at neutral sites was much reduced 112 in HGs, most Early Neolithic farmers show diversity levels only slightly lower than those of 113 modern humans (Fig. 2b), with genomes from NW Anatolia at the lower end of the 114 distribution17. We note a slight reduction of diversity in modern humans with distance from 115 Anatolia to the West of Boncuklu (Spearman’s ⍴ = -0.344, p-value = 0.028), but not to the East 116 (Spearman’s ⍴ = 0.019, p-value = 0.929), while no such simple trend is observed among early 117 European farmers 118 Compared to the other samples, the HG genomes, and in particular Bichon and SF12, show a 119 larger proportion of short (2-10Mb) Runs of Homozygosity (ROHs, Fig. 2c), in keeping with 120 previous results17,28,29. This is indicative of higher levels of remote inbreeding within European 121 HGs, likely due to smaller population sizes as corroborated by MSMC2 analyses (Fig. S25). 122 Among early farmers, WC1, Bon002, AKT16, STAR1 and Stuttgart also show a high 123 proportion of short ROHs and seem to be drawn from small populations, too. Furthermore, 124 WC1, Stuttgart, LEPE52, Bichon and Loschbour, as well as several modern individuals from 125 the Near East, carry some very long ROHs (>10Mb), indicative of recent inbreeding between 126 close relatives (potentially second cousins or closer30, Fig. S20).
127 We find that the vast majority of early farmers in our dataset had 128 complexion, while HGs had a darker skin tone (Supp. Table 3). A dark (brown to black) hair 129 color was inferred for all but two samples, LEPE52 and VC3-2, who likely had light brown 130 hair. Eye color variation was similarly low, with all samples showing high probabilities for 131 brown eyes, except for two individuals of the Starčevo culture (STAR1 and VC3-2) which 132 were likely blue-eyed.
133 Based on polygenic scores, we show that early farmers are shorter than HGs (Student t-test, t 134 = -2.427, p-value = 0.027), and their stature declined between 8,300 and 7,000 BP (Pearson's r 135 = 0.6537, p-value < 0.008, Fig. S24), suggesting that selection for short stature occurred during 136 the Neolithic expansion along the Danubian corridor.
137 The allele associated with lactase persistence was not found in any of the analyzed ancient 138 samples, consistent with an increase in frequency of these alleles at a later stage31. However, 139 early farmers already show allele frequencies similar to contemporary Europeans for 6 out of 140 7 SNPs of the FADS1/2 gene complex, potentially selected in populations with plant-based 141 diet32,33 (see Suppl. Information - Section 5, Table S7).
|
|
|
Post by Admin on Nov 29, 2020 20:27:37 GMT
90 Results 91 Patterns of genomic diversity along the Danubian corridor 92 The genetic structure and affinities of ancient individuals 93 Both multidimensional scaling (MDS) performed on the neutral portion of ancient and modern 94 genomes (Fig. 2a) and an admixture analysis (Fig. S21) revealed three main clusters of ancient 95 samples, which are found overall much more differentiated than modern individuals of SW 96 Asia and Europe: i) a cluster of European HGs, ii) a cluster of Early Neolithic individuals from 97 Iran, here represented by a single genome from Wezmeh Cave, WC1, and iii) a cluster with all 98 other Holocene individuals. In keeping with previous analyses based on a restricted set of 99 SNPs26,27, this MDS analysis (Fig. 2a) suggests strongest affinities of European and NW 100 Anatolian Neolithic samples with modern Sardinians, with the exception of the Early Neolithic 101 NW Anatolian individual Bar8 found to be closer to modern individuals from Greece, Albania 102 and other individuals from Southern Europe. However, a MDS analysis performed on the 103 whole genome including sites potentially affected by selection (Fig. S19) rather suggests that 104 early farmers are closer to Southern Europeans other than Sardinians. The Early Neolithic 105 individual from Iran (WC1) shows strongest genetic affinities with modern Iranians (Fig. 2a), 106 and to a lesser extent with individuals from the Northern Caucasus, suggesting some genetic 107 continuity in Iran since Neolithic times. Finally, Palaeolithic and Mesolithic HGs are generally 108 distinct from all modern SW Asians and Europeans, the closest of whom are Baltic Sea 109 individuals, Russians and Scandinavians.
Figure 3 - Demographic scenario inferred from the sampled genomes and underlying genetic data. a: This demographic history was obtained by summarizing the best models of all tested scenarios. b: zoom-in on the red-square area in panel a. The X symbols indicate very strong bottlenecks that occurred on the HG ancestral branch before the divergence between Bichon-Loschbour and central European HGs and some 200 kya in the ancestral population. Only admixtures with point estimates ≥ 5% are represented with arrows (≥ 10% when arrows have a dark outline). c: MDS analyses performed on the neutrally evolving portion of the 17 ancient whole-genomes used in the demographic models (left) or on data simulated (right) according to the inferred ML parameters of the global scenario shown in panes a and b. 110 Early farmers are genetically more diverse and decline in stature over time 111 While genetic diversity as quantified by the heterozygosity at neutral sites was much reduced 112 in HGs, most Early Neolithic farmers show diversity levels only slightly lower than those of 113 modern humans (Fig. 2b), with genomes from NW Anatolia at the lower end of the 114 distribution17. We note a slight reduction of diversity in modern humans with distance from 115 Anatolia to the West of Boncuklu (Spearman’s ⍴ = -0.344, p-value = 0.028), but not to the East 116 (Spearman’s ⍴ = 0.019, p-value = 0.929), while no such simple trend is observed among early 117 European farmers.
118 Compared to the other samples, the HG genomes, and in particular Bichon and SF12, show a 119 larger proportion of short (2-10Mb) Runs of Homozygosity (ROHs, Fig. 2c), in keeping with 120 previous results17,28,29. This is indicative of higher levels of remote inbreeding within European 121 by MSMC2 analyses (Fig. S25).
122 Among early farmers, WC1, Bon002, AKT16, STAR1 and Stuttgart also show a high 123 proportion of short ROHs and seem to be drawn from small populations, too. Furthermore, 124 WC1, Stuttgart, LEPE52, Bichon and Loschbour, as well as several modern individuals from 125 the Near East, carry some very long ROHs (>10Mb), indicative of recent inbreeding between 126 close relatives (potentially second cousins or closer30, Fig. S20).
127 We find that the vast majority of early farmers in our dataset had intermediate to light skin 128 complexion, while HGs had a darker skin tone (Supp. Table 3). A dark (brown to black) hair 129 color was inferred for all but two samples, LEPE52 and VC3-2, who likely had light brown 130 hair. Eye color variation was similarly low, with all samples showing high probabilities for 131 brown eyes, except for two individuals of the Starčevo culture (STAR1 and VC3-2) which 132 were likely blue-eyed.
133 Based on polygenic scores, we show that early farmers are shorter than HGs (Student t-test, t 134 = -2.427, p-value = 0.027), and their stature declined between 8,300 and 7,000 BP (Pearson's r 135 = 0.6537, p-value < 0.008, Fig. S24), suggesting that selection for short stature occurred during 136 the Neolithic expansion along the Danubian corridor.
137 The allele associated with lactase persistence was not found in any of the analyzed ancient 138 samples, consistent with an increase in frequency of these alleles at a later stage31. However, 139 early farmers already show allele frequencies similar to contemporary Europeans for 6 out of 140 7 SNPs of the FADS1/2 gene complex, potentially selected in populations with plant-based 141 diet32,33 (see Suppl. Information - Section 5, Table S7).
|
|
|
Post by Admin on Nov 29, 2020 22:26:04 GMT
142 Demographic inference 143 A step-wise expansion of Neolithic farmers into Central Europe 144 We contrasted eight scenarios of the spread of farmers into Europe, using four Early Neolithic 145 populations from Northern Greece, Central Serbia, Lower Austria and Southern Germany, and 146 one HG population from Serbia, each represented by at least two individuals (Fig. S30, 147 Table S9). We find that strict stepwise scenarios are better supported than scenarios allowing 148 for a long-distance migration from the Aegean directly to Lower Austria (Fig. S30e, Supp. 149 Table 4). Importantly, scenarios without HG introgression into early farmer populations are 150 clearly rejected. It implies that early farmer communities incorporated a few HG individuals 151 (2-6%, Fig. S31) at all major stages of the dispersal along the Danubian corridor. However, the 152 total amount of HG contribution into the farmer gene pool did not necessarily increase along 153 the expansion as the input of HG genes was almost matched by the input from other farming communities (2-5%). This complex pattern of gene flow might 154 explain the apparent lack of 155 genetic structure among early farmer individuals shown in the MDS plot (Fig. 2a), as well as 156 an absence of increasing HG ancestry along the Danubian corridor in our admixture analysis 157 (Fig. S21).
158 A mixed ancestry of all European and Anatolian farmers originating just after the LGM 159 In our initial model, the population ancestral to all European farmers is surprisingly found to 160 be the product of a substantial post-LGM admixture between a HG population, potentially from 161 Anatolia or the Near East, and a HG population closely related to the genomes from Vlasac, 162 Central Serbia called hereafter east and central (Fig. S31). To shed more light on this 163 admixture, we progressively added individuals of other populations. We started with a 164 population from NW Anatolia represented by an individual from Barcın (Bar25), which we 165 found to have diverged from the other Aegean population (Northern Greece) very recently at 166 the beginning of the Holocene (~9.9 kya, 95% CI 10.8-9.1). We then added the Neolithic 167 genome from Aktopraklık in NW Anatolia, which we estimate to have split very recently from 168 Barcın (9.2 kya, 95% CI 9.5-9.1; Fig. S36d, S37, Supp. Table 4). However, this individual 169 received massive genetic contributions from both surrounding farmers (25%, 95% CI 28-18) 170 and surrounding HGs (16%, 95% CI 22-14) (Fig. S37, Supp. Table 4), in line with the 171 admixture analysis (Fig. S21), f-statistics (Fig. S52), and its affiliation to the ‘coastal Fikirtepe 172 horizon’ thought to have been influenced by both Epipalaeolithic and Neolithic traditions8. 173 Importantly, these extended analyses confirm the old (~19.4 kya, 95% CI 23.3-10.4) and 174 massive (41% central HG contribution, 95% CI 38-50) admixture between the two HG 175 populations, which are found to have diverged during the LGM (23.4 kya; 95% CI 31.5-21.2) 176 (Fig. S31, Supp. Table 4).
177 To further study the spread of Neolithic people into Europe, we added two Early Neolithic 178 individuals from Lepenski Vir (Fig. S40), a site in the Danube Gorges with long pre-Neolithic 179 traditions of fishing, hunting and gathering and without ecological conditions for agriculture. 180 These two individuals previously shown to resemble Neolithic farmers from NW Anatolia16,34 181 are found to be tightly connected to the Northern Greek early farmers, and could thus be part 182 of an early migration of farmers into the Balkans.
183 Adding an early farmer from Boncuklu in the Konya plain of Central Anatolia (Bon002, 184 Fig. S42a) revealed that the Boncuklu population also shows a mixed ancestry, and that it 185 diverged ~13.4 kya (95% CI 14.6-11.5) from the branch leading to the Aegeans. In addition, it 186 would have received quite large (8%, 95% CI 1-17) and recent (~11.8 kya, 95% CI 13.0-10.3) 187vHG admixture, and relatively little input from surrounding farmers (2%, 95% CI 0-9) (Fig. S43, 188 Supp. Table 4).
189 The genetically distinct (Fig. 2a) early farmer from Wezmeh Cave in the Iranian Zagros region 190 is inferred to have diverged from the HG population ancestral to Aegeans and Central 191 Anatolians (Fig. S42b-c) during the LGM (~20.1 kya, 95% CI 20.9-19.6, Fig. S44, Supp. 192 Table 4), before it received the massive admixture from the central HGs observed in all other 193 investigated early farmers. Its genetic proximity with the pre-admixed HG population 194 (Fig. S50, Fig. S51c) suggests that the latter was located in the Near East.
195 Finally, we investigated the relationship between two western European HGs from Bichon and 196 Loschbour and our newly-sequenced Mesolithic individuals from Serbia. We find that Bichon 197 and Loschbour have a common ancestor branching off the central HG ancestral population 198 23.3 kya (95% CI 23.3-20.0) (Fig. S47c, S48, Supp. Table 4), and that they diverged from 199 each other soon after this split. In contrast, the Danube Gorges Mesolithic population from 200 Vlasac diverged from the central HG group relatively late about 10.2 kya (95% CI 9.0-21.3) 201 and remained well isolated afterwards with very little later admixture (<1%) (Fig. S31, Supp. 202 Table 4). Altogether, this suggests that the LGM led to a fragmentation of HG populations in 203 SW Asia and Europe with at least four genetically distinct groups: one related to Loschbour 204 and Bichon (called west, subdivided into west1 and west2, based on the old divergence between 205 Loschbour and Bichon branches), one related to the Danube Gorges Mesolithic samples 206 (central), another one that later received the massive Central HG introgression (east1 then 207 admixed), and a last one potentially further East (east2) related to WC1 (Fig. 3a, 4a, S34, S48, 208 Supp. Table 4).
|
|
|
Post by Admin on Nov 30, 2020 4:29:06 GMT
210 Discussion 211 The LGM shaped Holocene genetic diversity in SW Asia and Europe 212 We find that Holocene human genetic structure in SW Asia and Europe emerged briefly before 213 or during the LGM with the initial separation 32-21 kya of a western-central European and an 214 eastern group of HGs. Right after this initial split, the western-central European branch 215 experienced a very strong bottleneck (equivalent to a single human couple for one generation) 216 that decreased the diversity of all descending populations. Then, these HGs further divided 217 23.3-20.0 kya, leaving us with three genetically distinct groups in western-central Europe that 218 potentially differentiated in separate LGM refuge areas (Fig. 4a). The ancestors of Loschbour 219 and Bichon could have resided in separate refugia in South Western Europe, and the ancestors 220 of the Mesolithic Vlasac population could have lived in a geographically distinct central 221 refugium likely located around the Balkans and the Aegean. Broadly speaking, these refugial 222 populations coincide later on with what archaeologists have identified as the areas of 223 distribution of Magdalenian and Epigravettian traditions in Europe35,36. In contrast, the eastern 224 group of HGs, which does not show any sign of a strong bottleneck and was potentially 225 genetically more diverse, diverged further into at least three groups of Near Eastern HGs during 226 the LGM: one that later massively admixed with central HGs to become the ancestors of later 227 Anatolian and Aegean farmers, one leading to the ancestors of Iranian Neolithic farmers, and 228 one to Neolithic populations in the southern Levant (respectively east1, east2 and east3 on 229 Fig 4a). After the LGM, these HG populations re-expanded from their southern refugia 230 probably due to improving climatic conditions37, allowing previously separated central and 231 east1 refugial populations to overlap and admix 19 kya (Fig. 4b), and then to become a distinct 232 population from which Northwestern, Central Anatolian and European farmers would later 233 descend.
234 Even though Central Anatolia has previously been proposed to have hosted admixture 235 events15,17, the exact geographic location of the massive post-LGM admixture event is difficult 236 to pinpoint, and even though we modeled a single pulse of gene flow, admixture could also 237 have consisted in extensive gene flow over several generations and over a relatively large area.
238 We can nevertheless envision two alternative scenarios of admixture and later migrations. 239 1) A demic diffusion scenario: the admixture took place mainly in the Fertile Crescent 240 (Fig. 4b), implying separate migrations from the western Fertile Crescent to Central Anatolia 241 and the broader Aegean region including NW Anatolia. Given the genetic proximity between 242 Epipalaeolithic Central Anatolian foragers and Early Neolithic farmers15 and our inferred early split of the Boncuklu population, an initial migration into the Central 243 Anatolian Plateau could 244 have occurred already before the Younger Dryas and thus well before the Neolithic transition 245 (Fig. 4c). In contrast, the migration to NW Anatolia would have occurred at the time of the 246 fully developed, ceramic Neolithic (Fig. 4c), characterized by the establishment of widespread 247 mixed farming across large parts of Anatolia38. Archaeological observations suggest two 248 separate routes of neolithization towards the broader Aegean region. The first one would be a 249 land-route across the Anatolian plateau, with Barcın showing clear cultural but only remote 250 genetic affinities with Central Anatolia. The second one would be a maritime route connecting 251 seafaring communities of the Eastern Mediterranean and the Aegean region39,40 (Fig. 4c). 252 2) A cultural diffusion scenario: under this scenario, the admixture event at the origin of the 253 ancestors of Anatolian and European Neolithic farmers occurred further west, i.e. closer to the 254 inferred location of the Aegean refugial population. This scenario, which is plausible given 255 technological interactions between HG communities in the Eastern Mediterranean and the 256 Aegean41, assumes a pre-Neolithic expansion of Near Eastern refugial populations into NW 257 Anatolia. It would also explain the appearance of Near Eastern-like genetic signals in post- 258 LGM European HGs, which has been postulated for the period 14-17 kya16,42,43.
259 Despite some continuity in flake-based lithic industries across the Mesolithic-Neolithic 260 transition in Greece, the abrupt appearance of fully developed Neolithic lifeways in that region 261 involving dozens of innovations at hundreds of newly-founded sites44,45 seems difficult to be 262 explained by cultural diffusion alone and appears to be more compatible with demic diffusion 263 from the Fertile Crescent. In contrast, based on our genetic data, adoption of agriculture by 264 indigenous HG communities is more likely in Central Anatolia, where early aceramic sites like 265 Boncuklu and Aşıklı show experiments in crop cultivation and caprine management, with 266 increasing dependence on food-production, including a heavy emphasis on caprines after 9.7 267 kya6,46,47.
268 Further support for the demic diffusion scenario comes from f-statistics showing Levantine 269 populations to share more drift with Aegeans than with Central Anatolian Neolithic individuals 270 (Fig. S57). This signal could either be due to some long distance gene flow between the 271 Aegeans and the Levant, a higher level of central HG admixture observed in Boncuklu 272 (Fig. S56), or a combination of i) an early migration of the Boncuklu HG ancestors from the 273 Fertile Crescent to Central Anatolia before the Younger Dryas (Fig. 3a, 4c), ii) some gene flow 274 between people from the Levant and the ancestors of Aegeans, who would have remained in 275 the Fertile Crescent and only later migrated to the West. However, early farmers from the 276 Aegean are rather heterogeneous in their levels of shared drift with several populations, 277 suggesting that the peopling of the Aegean was a complex process. 278 279 A demic diffusion of Neolithic farmers along the Danubian corridor 280 From an archaeological point of view, there have been a large number of proposed explanations 281 for the introduction of Neolithic lifeways in Europe48. Our explicit modelling supports the 282 simplest of all demic models, namely a gradual spread/progressive migration of early farmers 283 originating in the wider Aegean region (NW Anatolia or Greece) and extending to Serbia along 284 the Balkans and the so-called Danubian corridor, then to Hungary (usampled) and Austria, and 285 eventually up to the Rhine valley in southwestern Germany (Fig. 1a, 4c). While this study 286 focused on the Danubian or continental route of Neolithic expansion, we expect Impresso- and 287 Cardial-related farmers who spread along the Mediterranean shoreline to have shared a similar 288 genetic background in the Aegean14.
289 Low levels of admixture with local Mesolithic populations (2-6%) seem to have occurred at 290 each of the four modelled migration steps, suggesting that early farming communities were not 291 completely genetically isolated49. The inferred rates of admixture are slightly lower than 292 previously reported (3-9%14,49,50). Even though we have modelled this hybridization process to 293 have occurred from the same Mesolithic metapopulation to which the Danube Gorges 294 Mesolithic individuals from Vlasac belong, we cannot exclude that admixture in Austria and 295 Southern Germany occurred with a Western European Mesolithic metapopulation, to which 296 Loschbour and Bichon are connected, as previous work has suggested that different Mesolithic 297 backgrounds could have introgressed early farmer gene pools in different regions50.
|
|