|
Post by Admin on Sept 6, 2023 18:42:15 GMT
Figure 2 Cemetery maps of Fonyód, Hács, and Balatonszemes Burials are identified based on age, sex (osteological and genetic), and sampling status as well as the presence of amber beads, brooches, double- and single-sided combs, and weapons. Dotted outlines indicate biological relatedness. Pedigrees of the biologically related individuals are overlaid on the cemetery maps with colors corresponding to fastNGSadmix results from Figure 4. (A) Cemetery map of Fonyód. Due to the dimensions of the cemetery, graves sizes are not to scale with the size of the cemetery. This figure was modified from Gallina and Straub.26 (B) Cemetery map of Hács. This figure was modified from Kiss.5 (C) Cemetery map of Balatonszemes. Bal_26 is not included in this figure, as this individual was buried 200 m west of the main site. This figure was modified from Miháczi-Pálfi.27 See also Tables S1, S3, and S4. 5th–8th century Italy: Bardonecchia and Torino Lavazza The two penecontemporary Italian sites of Bardonecchia and Torino Lavazza were included in this study, so as to increase the number of available high-quality Medieval Italian sequenced individuals. These data supplement the often lower-coverage Italian data from Rome from Antonio et al.7 (Table S1). Genome sequencing From Fonyód, we obtained genomic data from 13 of 14 graves with human remains, from Hács from 14/15 preserved burials (the 15th, Hacs_23, was not accessible for sampling), and from Balatonszemes 11 of 13 graves with human remains. Altogether 38 samples were newly sequenced and analyzed (STAR Methods). Ten samples were used in whole-genome sequencing (WGS) (coverage mean 8.36×, range 5–11×) and the remaining 28 in genome-wide SNP capture sequencing for ∼1.2 million SNPs (1240K)28,29,30 (coverage mean 1.47×, range 0.02–3.46×) (Table S1). Among WGS samples, virtually all of the 1,126,140 autosomal 1240K SNPs were covered (1,098,348 to 1,120,721 SNPs), while for captures the number of SNPs covered ranged from 24,558 to 859,675 (Table S1). Mitochondrial contamination rate medians31 were 1%–3% for all individuals, while nuclear contamination point estimates32 from the 14 males were between 0.4% and 2%. In addition, 14 new samples from Bardonecchia and Torino Lavazza (5th–8th century Italy) underwent 1240K capture with coverages from 0.39 to 3.85× (covering 306,226 to 845,032 SNPs), while one from Bardonecchia underwent WGS (Bard_T1, coverage 6.54×, 1,121,223 SNPs).
|
|
|
Post by Admin on Sept 7, 2023 19:20:57 GMT
Temporal variation in population genetic structure in the 5th century We conducted a principal component analysis (PCA) of 561 ancient individuals (with coverage ≥ 0.1×); ancient individuals were transformed onto a background of modern POPRES33 reference individuals using a Procrustes-transformation technique6,34 using smartPCA.35,36 We found that all 5th/6th century Lake Balaton individuals (n = 69 due to the coverage cut-off) reflected primarily modern European genetic diversity (Figures 3A and S1). In addition, we analyzed a penecontemporary (c. 4th–8th century CE) set of 492 comparative individuals (477 previously published alongside 15 newly produced from Bardonecchia and Torino Lavazza) sampled from across Europe (Table S1; STAR Methods).6,7,8,9,10,11,12,37 Confidence ellipses of PC1 and PC2 coordinates demonstrate that while the penecontemporary regions cluster primarily with modern individuals from the same geographic regions (i.e., they are localized) (Figure 3B), our 5th–6th century Lake Balaton individuals have a much more diverse genomic ancestry, encompassing a range reflecting the entire north to south axis of modern European genetic variation in the POPRES dataset33 (though not western or eastern Europe) (Figure 3A). While all Lake Balaton individuals were sampled from a single region of 200 km2 over ∼100 years, the penecontemporary populations generally encompass much larger geographic regions and wider time frames (and thus we would expect more genetic diversity): this suggests that this part of central Europe experienced particularly high rates of gene flow during the 5th–6th centuries. We explicitly modeled gene flow in a spatial context across Europe using our Lake Balaton and penecontemporary individuals using FEEMS38 and indeed found a strong rate of gene flow between the Lake Balaton region and populations across Northern Europe ∼1,000 km away (Figures 3D and S2). Figure 3 Principal component analysis and spatial gene flow analysis Procrustes PCA of 561 individuals6,7,8,9,10,11,12,37 transformed onto a PCA with 1,385 modern European individuals from the POPRES dataset33 using pseudohaploid genotype calls from 328,670 SNPs with individuals. Penecontemporaneous reference individuals are colored in pastel colors based on their region, while modern individuals from POPRES are colored in gray. (A) PCA results from the Lake Balaton communities are plotted with covariance confidence ellipses with radii corresponding to 1.5 standard deviations overlaying the PCA. The dashed box identifies the northern clusters from the Lake Balaton communities. (B) PCA results from the penecontemporary populations are plotted with covariance confidence ellipses with radii corresponding to 1.5 standard deviations. (C) A zoom-in of the area within the dashed box from (A). This plot focuses on the northern cluster of the PCA with polygons identifying the lack of overlap between the 5th century Lake Balaton communities. (D) A FEEMS plot estimating gene flow between the 4th and 8th century communities analyzed in the PCA with all four Lake Balaton communities analyzed together. Bluer colors indicate more migration, while brown indicates less/no migration. To further explore the structure within our Lake Balaton cemeteries, we also used a model-based clustering method (fastNGSadmix)39 to characterize the genomic ancestry of each individual using 1000 Genomes Project (1000G)40 populations as possible sources. In addition, given the localized geographic structure observed in the PCA for the penecontemporary Europeans, we constructed a panel of ancient c. 4th–8th century penecontemporary reference individuals with a similar geographic distribution to the 1000G populations to form eight ancient panels (EASIA, MEDEU, NAFRICA, NGBI, SASIA, SCAND, and SUBSAHARAN) (STAR Methods).7,9,10,11,12,14,15,16,17,18,19 The modern and ancient reference sets yielded similar patterns (Figures 4A , 4B, and S3), with almost all individuals possessing some combination of primarily Tuscan (TSI), Central European and Great Britain (CEU+GBR), and Finnish (FIN) modern ancestry or Mediterranean (MEDEU), northern Germany and British (NGBI), and Scandinavian/Estonian (SCAND) penecontemporary ancestry. We conducted linear regressions comparing ancestry proportions of MEDEU versus TSI, NGBI versus CEU+GBR, and SCAND versus FIN (for 441 individuals not in a reference panel) and found R2 values of 0.79, 0.54, and 0.36, respectively (all comparisons p < 0.001). However, all three 5th century cemeteries have distinct profiles with regard to the relative proportions of these components, with the penecontemporary analysis showing these differences most clearly. Most notably, there is a clear progressive significant increase in SCAND ancestry and corresponding decrease in NGBI ancestry over time for the 5th century sites (Fonyód, Hács, Balatonszemes; SCAND 0.07 < 0.26 < 0.38; NGBI 0.40 > 0.31 > 0.14; Figure 4C). See also Figures S1 and S2 and Table S1.
|
|
|
Post by Admin on Sept 8, 2023 20:12:29 GMT
Figure 4 Supervised ancestry proportions from individuals from Fonyód, Hács, Balatonszemes, and Szólád Individuals with <0.1× coverage were excluded from these analyses. The ♂ and ♀ symbols identify genetic males and genetic females, respectively, while triangles identify individuals buried with dress accessories (in the case of Fonyód, Hács, and Balatonszemes) or brooches specifically (in the case of Szólád). (A) Proportions were estimated using fastNGSadmix and using 1000 Genomes populations as references40 (CEU+GBR, Northern Europeans from Utah [CEU] and British in England and Scotland [GBR]; FIN, Finnish in Finland; IBS, Iberian populations in Spain; TSI, Tuscans from Italy; EAS, the East Asian super-population; SAS, the South Asian super-population; YRI, Yoruba in Ibadan, Nigeria). (B) Proportions were estimated using penecontemporaneous individuals to form reference populations7,9,10,11,12,14,15,16,17,18,19 (MEDEU, Italy and Iberia [Mediterranean Europe]; NGBI, what are now northern Germany and Britain; SCAND, what are now Scandinavia/Estonia; EASIA, what is now Hanben, Taiwan; NAFRICA, what is now Sudan; SASIA, Roopkund Lake in what is now India; SUBSAHARAN, sites from sub-Saharan Africa). Individuals are sorted based on increasing MEDEU and then by decreasing NGBI. (C) A line graph showing the change in ancestry proportions through time over the four sites. Error bars correspond to 95% confidence intervals calculated based on bootstrap analyses. See also Figure S3 and Table S1. Fonyód, which dates to the peak of the Hunnic power in the region, differs markedly from the later two 5th century sites from both an archaeological and genomic perspective (STAR Methods). Its unique spatial structure suggests that it is a burial site of a short-lived coexistence of Hunnic period groups, part of them practicing ACD, a tradition that is considered a foreign phenomenon in Pannonia and less prevalent in the later sites.41 Genomically, Fonyód also has significantly (as assessed by non-overlapping 95% CIs) more Mediterranean ancestry than the other Lake Balaton sites and is somewhat southern shifted on the PCA (Figures 3 and 4). Overall, the site shows more non-European diversity, with two individuals possessing ∼6% South and/or East Asian ancestry (Fonyod_278 and Fonyod_316) and one individual 12% African ancestry (Fonyod_469) (Figure 4). Thus, the genomic profile of Fonyód may reflect heterogeneity of the late Roman period and/or recent influx from the East. Both Hács and Balatonszemes are enriched for individuals with very high proportions of northern European ancestry (NGBI+SCAND) that are not as prominent at Fonyód (Figure 3C), potentially indicating the arrival of new groups to the region consistent with written records describing the emergence of new “barbarian” powers after the fall of the Hunnic empire, in the second half of the 5th century. The high level of gene flow between Lake Balaton individuals and northern Europe is only seen in our FEEMS analysis when examining these later two communities, while the use of the earlier Fonyód site results in a barrier to gene flow to this region and greater gene flow with southern Europe (Figure S2). However, because of variation in the relative SCAND/NGBI components, individuals with higher PC1 values (i.e., more northern European ancestry) show no overlap in the PCA between the two sites (Figure 3C), with Hács angling toward the modern northwestern Europe and Balatonszemes northeastern Europe. These differences could indicate that groups from similar, yet distinguishable, sources from northern Europe arrived in the region in multiple waves during the second half of the 5th century. On the other hand, the presence of individuals with high amounts of MEDEU ancestry is a constant in all four sites, and they show overlap in the PCA, suggesting this may represent a more stable local genomic signature during this entire period. To test our findings of northern gene flow into the Balaton region subsequent to Fonyód, we also conducted qpAdm29 analyses; these findings are consistent with our described findings from FEEMS and fastNGSadmix, though we caution that the qpAdm method is not ideal for analyzing closely related Late Antique/Early Medieval source populations (STAR Methods; Figure S4; Table S2). Interestingly, Szólád demonstrates a profile that encompasses genomic variation observed in all three 5th century sites, albeit it has a much larger sample size. Northern European ancestry is even more prominent than at Hács and Balatonszemes, driven primarily (but not totally) by a large nine-member pedigree identified previously6 (Figure 4). The individuals in this pedigree along with most other northern-like individuals lack the high SCAND component observed in Balatonszemes (qualitatively resembling those from Fonyód and Hács instead; the three notable exceptions being 2 second-degree relatives, Sz_41 and Sz_42, with 92% and 73% SCAND ancestry, respectively, and Sz_4 with 59%), suggesting no major direct continuity between these two sites. Szólád is notable for strontium isotope data, suggesting most adults at the site were non-local, regardless of genomic ancestry.6 However our data make clear that the major patterns of genetic ancestry observed at Szólád were already established during the second half of the 5th century in and around the Lake Balaton area. Thus, this community could have formed from the existing diverse regional pool of genetic variation established ∼50 years earlier, rather than just being the result of the arrival of a new population group, i.e., the Langobards, as interpreted by both historical and archaeological research.42,43,44,45
|
|
|
Post by Admin on Sept 9, 2023 20:09:01 GMT
Biological relatedness and social relationships The small burial groups at Fonyód were previously interpreted as signs of strong social ties,26 and Hács and Balatonszemes have been described as family cemeteries.5,25,46 We used lcMLkin47 to identify close biological relatives in all three 5th century sites. At all three sites, biological kindreds consisted of only a few very close first- and second-degree relatives, mostly on the mother’s side (Figure 2; Table S3). These analyses were also validated by a READ analysis that found the same kinship relationships (STAR Methods).48 This is in stark contrast to Szólád, where the cemetery was organized largely around male biological relatives with a large extended pedigree.6 Biologically related individuals at Fonyód and Balatonszemes were buried in close proximity to each other, and in the latter case their connection is also clearly reflected by the similarities in the burial customs, suggesting that these connections held meaningful social values (STAR Methods). However, the relatively low number of biologically related individuals, the small size of the kindreds, and the lack of biological relations in most burial groups—probably also influenced by the short occupation period of the sites and the low number of buried individuals—suggest that other factors may have had a major influence on the formation of these 5th century communities. To examine to what extent the broader biological backgrounds of individuals (as determined by genomic ancestry) beyond close biological kinship were acknowledged as meaningful social ties, we also compared burial customs demonstrating variation with the POPRES PCA PC1 and PC2 coordinates using a logistic regression framework (Table S4). We observed a strong significant association between the presence of various jewelry types and dress accessories (i.e., polyhedral earrings, various brooch types, bracelets, amber beads, etc.) characteristic of the 5th century female burials49,50 and genetic variation (p = 0.008, n = 20 for adults; p = 0.002, n = 23 for adults and nonadults). This appears to be primarily driven by most individuals with dress accessories possessing high amounts of northern European ancestry (13/16 are majority SCAND/NGBI) (Figures 4 and S1), and as such, site-specific regressions are only significant for Hács (p = 0.01, n = 8) and Balatonszemes (p = 0.008, n = 7) but not Fonyód (p = 0.08, n = 8) (despite all three tests having similar power; Table S4). These results point to individuals with northern European genetic backgrounds being treated differently in death by their respective communities, perhaps marking cultural and/or social differences between locals and more recent migrants in the region. We also found a strong association between the PCA results and ACD for the individuals at Fonyód (with preserved crania) (p = 0.001, n = 10) (STAR Methods; Table S4). Though close biological relatives were rare, it is also noteworthy that all biological kindreds involved individuals with predominantly northern genomic ancestry, maybe reflecting another aspect of this biologically structured social organization. While Szólád shows a similar genomic profile to the three 5th century sites, it is strikingly different in terms of funerary practices, spatial organization, and demographics. There was no significant association between genetic variation and brooches—an artifact that in various forms and types is present in the 5th century sites as well as the 6th century Szólád—from genetically female burials (including and excluding nonadults) (p < 0.56, n = 11), suggesting that social and economic differences were no longer formulated along fault lines of genetic background in this community. Whether this weakening of the association between this artifact and genomic ancestry reflects a more general shift in social kinship practices (at least with respect to women) since the 5th century and the establishment of Langobard rule is difficult to conclude with just our data, as even among other 6th century Lake Balaton sites Szólád is considered somewhat unique, with the vast majority of adult and adolescent males (15/19) being buried with weapons.42,51 The importance of a northern genomic background appears to still be manifested in this community socially, but rather through the dominance of the cemetery by a spatially clustered and primarily male kindred. Conclusion A common assumption of ethnoarchaeology posited a common ancestral, ethnic, and cultural heritage of Early Medieval migratory communities following the dissolution of the Roman Empire in the West.52,53 Recent historical, archaeological, and anthropological studies have found this to be a vast oversimplification, with the material culture demonstrating significant complexity both on site and on the regional level. This study adds striking evidence of this complexity. Rather than homogeneity, our three post-Roman 5th century sites from Lake Balaton exhibit considerable genomic diversity compared to penecontemporaneous Europe. This region experienced particularly high levels of gene flow during this period, and there are significant shifts in genomic ancestry through our 5th century time transect over a period of only ∼50 years that indicate migration into the region from various sources, probably from areas in northern Europe, consistent with the continuously changing political landscape of the period described in the historical record.54,55 As these post-Roman communities formed from mixtures of locals and migrants, social ties between individuals of similar genomic backgrounds appear to have still been maintained to some extent, but these links could shift rapidly rather than being enduring states of social organization, and the importance of close biological relatedness varied considerably. Given the immense complexity observed in just these four small cemeteries, it is clear that comprehensive fine-grained spatiotemporal genomic sampling will be critical to unpack the processes that underlie the subsequent development of modern Europe.
|
|
|
Post by Admin on Nov 16, 2023 19:31:57 GMT
Local population structure in Cambridgeshire during the Roman occupation Abstract The Roman period saw the empire expand across Europe and the Mediterranean, including much of what is today the United Kingdom. While there is written evidence of high mobility into and out of Britain for administrators, traders and the military, the impact of imperialism on local population structure is invisible in the textual record. The extent of genetic change that occurred in Britain before the Early Medieval Period and how closely linked by genetic kinship the local populations were, remains underexplored. Here, using genome-wide data from 52 ancient individuals from Cambridgeshire, we show low levels of genetic ancestry differentiation between Romano-British sites and lower levels of runs of homozygosity over 4 centimorgans (cM than in the Bronze Age and Neolithic. We find fourteen cases of genetic relatedness within and one between sites without evidence of patrilineal dominance and one case of temporary mobility within a family unit during the Late Romano-British period. We also show that the modern patterns of genetic ancestry composition in Modern Britain emerged after the Roman period. Introduction The most visible individuals from Roman Britain through texts are soldiers and administrators, many of whom came from other parts of the Empire (Eckardt and Müldner 2014). The Roman army was recruited from across the whole empire, and, in general, soldiers were posted to areas away from their homelands to avoid conflicts of loyalty (Haynes 2013)(pgs. 121–29). Migration from the Empire into Britain was likely dominated by these groups, as well as traders. Although local populations are archaeologically very well documented, the movement is less well understood and is invisible in the textual record. The extent of mobility in this period has been the subject of recent debate, with work largely focusing on the use of isotope data (Eckardt and Müldner 2014). However, as sampling has been dominated by the examination of unusual burials, our knowledge of the scale of migration and its impact on the overall population is impossible to assess. While the subsequent Early Medieval Period (5th - 10th centuries CE) arguably resulted in major genetic shift towards higher affinities to Dutch, Danish, and other continental North Sea zone ancestries in eastern England, at the scale of 38–75% on average (Schiffels et al. 2016; Gretzinger et al. 2022), it is not clear whether this is due to migration only during the Early Medieval Period or if any of this change could be ascribed to gene flow during the Roman Period and before (Oosthuizen 2017); however, the long-standing ties between Britain and Gaul (Champion 2016)(pg. 155), both prior to and during the Roman period, may obscure the genetic distinction between local, indigenous Britons and incoming individuals. In contrast to recent genomic studies on demographic changes in Bronze and Iron Age (Patterson et al. 2021) and Early Medieval (Gretzinger et al. 2022) periods, to date, few genomes from Great Britain from the Roman period have been published. A study of seven individuals from a cemetery in York with decapitations showed most individuals had a higher affinity to the modern Welsh than modern English, yet also highlighted the cosmopolitan nature of the Roman empire by identifying an individual with Middle Eastern / North African ancestry (Martiniano et al. 2016). However, York was a cosmopolitan urban centre and cannot be taken as typical of the province as a whole (Ottaway 2004). The population of Roman Britain totalled 2–4 million and was dominated by rural communities, who accounted for about 90% of the people (Millett 1990)(pg. 185). Although extensively networked, these communities were arguably little affected by migration (A. Smith, M. Allen, T. Brindle, M. Fulford 2016)(pg 416–17). However, this hypothesis has not yet been tested. The area that is today Cambridgeshire provides an extensively researched rural, agricultural region, not atypical of the province as a whole, so genetic information from farmstead communities in this region provides a key opportunity to improve our understanding of the local population(s) of Roman Britain. www.biorxiv.org/content/10.1101/2023.07.31.551265v1.full
|
|