Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2020 Nov 12;183(4):890-904.e29.
doi: 10.1016/j.cell.2020.10.015. Epub 2020 Nov 5.

A Dynamic 6,000-Year Genetic History of Eurasia's Eastern Steppe

Choongwon Jeong  1 , Ke Wang  2 , Shevan Wilkin  3 , William Timothy Treal Taylor  4 , Bryan K Miller  5 , Jan H Bemmann  6 , Raphaela Stahl  2 , Chelsea Chiovelli  2 , Florian Knolle  2 , Sodnom Ulziibayar  7 , Dorjpurev Khatanbaatar  8 , Diimaajav Erdenebaatar  9 , Ulambayar Erdenebat  10 , Ayudai Ochir  11 , Ganbold Ankhsanaa  12 , Chuluunkhuu Vanchigdash  8 , Battuga Ochir  13 , Chuluunbat Munkhbayar  14 , Dashzeveg Tumen  10 , Alexey Kovalev  15 , Nikolay Kradin  16 , Bilikto A Bazarov  17 , Denis A Miyagashev  17 , Prokopiy B Konovalov  17 , Elena Zhambaltarova  18 , Alicia Ventresca Miller  19 , Wolfgang Haak  2 , Stephan Schiffels  2 , Johannes Krause  20 , Nicole Boivin  3 , Myagmar Erdene  10 , Jessica Hendy  21 , Christina Warinner  22
Affiliations

A Dynamic 6,000-Year Genetic History of Eurasia's Eastern Steppe

Choongwon Jeong et al. Cell. .

Abstract

The Eastern Eurasian Steppe was home to historic empires of nomadic pastoralists, including the Xiongnu and the Mongols. However, little is known about the region's population history. Here, we reveal its dynamic genetic history by analyzing new genome-wide data for 214 ancient individuals spanning 6,000 years. We identify a pastoralist expansion into Mongolia ca. 3000 BCE, and by the Late Bronze Age, Mongolian populations were biogeographically structured into three distinct groups, all practicing dairy pastoralism regardless of ancestry. The Xiongnu emerged from the mixing of these populations and those from surrounding regions. By comparison, the Mongols exhibit much higher eastern Eurasian ancestry, resembling present-day Mongolic-speaking populations. Our results illuminate the complex interplay between genetic, sociopolitical, and cultural changes on the Eastern Steppe.

Keywords: Eastern Steppe; Mongol empire; Mongolia; Xiongnu empire; ancient DNA; human population history; migration; nomadic pastoralists.

PubMed Disclaimer

Conflict of interest statement

Declaration of Interests The authors declare no competing interests.

Figures

None
Graphical abstract
Figure 1
Figure 1
Overview of Ancient Populations and Time Periods (A) Distribution of sites with their associated culture and time period indicated by color: Pre-Bronze, purple; Early Bronze, red; Middle/Late Bronze, blue; Early Iron, pink; Xiongnu, green; Early Medieval, brown; Late Medieval, gold (see STAR Methods). See Figure S1A and Table S1B for site codes and labels. (B) Inset map of Eurasia indicating area of present study (box) and the locations of other ancient populations referenced in the text, colored by time period. The geographic extent of the Western/Central Steppe is indicated in light brown, and the Eastern Steppe is indicated in light green. (C) Timeline of major temporal periods and archaeological cultures in Mongolia. Site locations have been jittered to improve visibility of overlapping sites.
Figure S1
Figure S1
Archaeological Sites and Geographic and Ecological Features in Mongolia, Related to Figure 1 (A) Archaeological sites in Mongolia and neighboring regions analyzed in this study. (B) Mongolian regions and provinces (aimags). Provinces are indicated by gray lines and text. Regions are indicated by black dashed lines and text following the definitions of (Taylor et al., 2019). (C) Ecological zones of Mongolia. Map produced using QGIS software (v3.6) with ecological data from (Dorjgotiv, 2004).
Figure 2
Figure 2
Genetic Structure of Mongolia through Time PCA of ancient individuals (n = 214) from three major periods projected onto contemporary Eurasians (gray symbols). Main panels display PC1 versus PC2; insets display PC1 versus PC3. Inset tick marks for PC1 correspond to those for the main panels; PC3 accounts for 0.35% of variation. See Figure S3B for population, sample, and axis labels, and Tables S1B, S1C, and S2A for further site and sample details.
Figure 3
Figure 3
Genetic Changes in the Eastern Steppe across Time Characterized by qpAdm (A–F) Major time periods: (A) Pre-Bronze through Early Bronze Age, (B) Middle/Late Bronze Age, (C) Early Iron Age, (D) Xiongnu period, (E) Early Medieval, and (F) Late Medieval. Modeled ancestry proportions are indicated by sample size-scaled pie charts, with ancestry source populations shown below (see STAR Methods). The sample size range for each panel is indicated in the upper right. For (B) and (C), Baikal_EBA is modeled as light blue; in (D–F), Khövsgöl_LBA (purple) and the Sagly/Uyuk of Chandman_IA (pink) are modeled as new sources (Figure 4). Cultural groups are indicated by bold text. For (D–F), individuals are Late Xiongnu, Türkic, and Mongol, respectively, unless otherwise noted. Previously published reference populations are noted with white text; all others are from this study. Populations beyond the map borders are indicated by arrows. Burial locations have been jittered to improve visibility of overlapping individuals.
Figure S2
Figure S2
Uniparental Haplogroup Assignments by Group and Sex-Bias Z Scores, Related to Figure 5B and STAR Methods (A and B) Population structure from uniparentally inherited markers. (A) Distribution of Y haplogroups across each period. (B) Distribution of mitochondrial haplogroups across each period. (C) Sex-bias Z scores by evaluating the differences of WSH-/Iranian-/Han-related ancestry on the autosomes and the X chromosome. We calculated Z-score for each ancient individual who has genetic admixture with any of the three ancestries. Positive scores suggest more WSH-/Iranian-/Han-related ancestry on the autosomes, i.e., male-driven admixture.
Figure S3
Figure S3
PCA of Present-Day Eurasian Populations and Genetic Structure of Mongolia through Time, Related to Figure 2 (A) PCA of present-day Eurasian populations used as the background for Figure 2 and Figure S3B. Here we show the population labels for the 2,077 Eurasian individuals used for calculating PCs and plotted as gray dots in Figure 2. Each three-letter code in the plot represents a single individual. Population IDs matching to the three-letter codes are listed at the bottom. (B) Genetic structure of Mongolia through time. Principal component analysis (PCA) of ancient individuals (n = 214) from three major periods projected onto contemporary Eurasians (gray symbols). Projection and axis variance corresponds to Figure 2. Population labels are positioned over the mean coordinate across individuals belonging to each population.
Figure 4
Figure 4
Genetic Ancestry Changes in Chronological Order across All Newly Reported Genetic Groups Well-fitted modeling results for grouped-based population genetics analyses for (A) prehistoric periods and (B) historic periods. The number of individuals in each genetic group is given in Table S3A. Raw ancestry proportions and standard error estimates are provided in Table S5. Horizontal bars represent ± 1 standard error (SE) estimated by qpAdm.
Figure S4
Figure S4
Genetic Changes in the Eastern Steppe across Time Characterized by qpAdm with All Individuals Indicated, Related to Figures 3 and 4 (A) Pre-Bronze through Early Bronze Age; (B) Middle/Late Bronze Age; (C) Early Iron Age; (D) Xiongnu period; (E) Early Medieval; (F) Late Medieval. Modeled ancestry proportions are indicated by sample size-scaled pie charts, with ancestry source populations shown below. Cultural groups are indicated by bold text. For panels (D–F), individuals are Late Xiongnu, Türkic, and Mongol, respectively, unless otherwise noted. Previously published reference populations are noted with white text; all others are from this study. Populations beyond the map borders are indicated by arrows. Burial locations have been jittered to improve visibility of overlapping individuals. Zoom in to see individual labels. Here we report results from admixture models that include all ancestry components required to explain historic late Medieval individuals as a group for unbiased cross comparison between individuals. Individual results with simpler admixture models can be found in Table S5J. See modeling details in Section 7.
Figure S5
Figure S5
Outgroup f3-Statistics and Cladality Testing using f4-Statistics, Related to Figures 3 and 4 (A) Outgroup f3-statistics for the pre-Bronze Age to Early Bronze Age groups in the Eastern Steppe. We show top 15 outgroup f3-statistics of the form f3(Target, world-wide; Mbuti) out of 345 ancient and present-day populations for the six target groups: eastMongolia_preBA, centralMongolia_preBA, Fofonovo_EN, Afanasievo_Mongolia, Chemurchek_southAltai and Chemurchek_northAltai. Horizontal bars represent ± 1 standard error (SE) calculated by 5 cM block jackknifing. (B) Testing cladality of the four ANA populations using f4-statistics. We show top and bottom 15 symmetric f4-statistics of the form f4(ANA1, ANA2; world-wide, Mbuti) out of 345 ancient and present-day populations for the four ANA-related target groups: eastMongolia_preBA, centralMongolia_preBA, Fofonovo_EN, DevilsCave_N. Horizontal bars represent ± 1 standard error (SE) calculated by 5 cM block jackknifing. f4-statistics with Z-score > 3 are highlighted in red. (C) Testing cladality of Afanasievo and Chemurchek using f4-statistics. We show top and bottom 15 symmetric f4-statistics for the three target groups Afanasievo_Mongolia, Chemurchek_southAltai and Chemurchek_northAltai, in the form f4(Afanasievo_Mongolia, Afanasievo; world-wide, Mbuti), f4(Chemurchek_southAltai, Botai; world-wide, Mbuti), f4(Chemurchek _northAltai, Chemurchek_southAltai; world-wide, Mbuti), and f4(Chemurchek _northAltai, eastMongolia_preBA; world-wide, Mbuti) out of 345 ancient and present-day populations. Horizontal bars represent ± 1 standard error (SE) calculated by 5 cM block jackknifing. f4-statistics with Z-score > 3 are highlighted in red.
Figure S6
Figure S6
Dating Admixture in Prehistoric Individuals, Related to STAR Methods (A) Dating admixture in prehistoric individuals. We estimated admixture dates using the DATES program and converted it by adding the age of each ancient population (mean value of the center of the 95% confidence interval of calibrated 14C dates) and assuming 29 years per generation. Horizontal bars associated with the admixture dates (colored circles) are estimated by the square root of summing the variance of DATES estimate using leave-one-chromosome-out jackknifing method and the variance of the 14C date estimate, assuming that the two quantities are independent. Published groups are marked with an asterisk (∗). For the Chemurchek_northAltai, we used Baikal_EN as the representative of ANA ancestry for dating the admixture event, given the larger sample size of Baikal_EN. (B) Ancestry covariance in prehistoric individuals. We show the weighted ancestry covariance (y axis) calculated from DATES which is expected to decay exponentially along genetic distance (x axis) with a decay rate indicating the time since admixture, and fitted exponential curves (shown in red line). We start the fit at genetic distance at 0.45 centiMorgans, and estimate standard error by a weighted block jackknife removing one chromosome in each run.
Figure S7
Figure S7
Breakdown of Geography and Genetics among Xiongnu and Comparison of Mongol Period and Present-Day Populations, Related to Figure 3 and STAR Methods (A) Breakdown of the geographic-genetic correlation in Xiongnu. We show the proportions of West Eurasian ancestry on all individuals/groups from Xiongnu era (y axis) versus the longitude of archaeological site they come from (x axis). The raw numbers of individual estimates can be found in Table S5G for models using Sarmatian as the western Eurasian source. Unlike MLBA/EIA individuals (Figure 3), Xiongnu individuals from more western sites do not have higher proportion of western Eurasian ancestry than those from eastern sites. (B) Comparing genetic homogeneity between ancient Mongol individuals and seven present-day Mongolic-speaking populations using qpWave. We report the p-value for every individual-based qpWave {ancient Mongol individual; Mongolic group} using seven modern Mongolic-speaking populations: Buryat, Daur, Kalmyk, Khamnegan, Mongol, Mongola, and Tu in the Human Origins dataset. When the p-value from qpWave is > 0.05, it suggests that the ancient individual on the y axis is genetically indistinguishable from the modern Mongolic-speaking population shown on the x axis. Smaller p-values indicate that the ancient individual is significantly different from the modern group.
Figure 5
Figure 5
Functional Allele Frequencies and Sex-Biased Patterns of Genetic Admixture (A) Allele frequencies of five phenotypic SNP changes through time. For the effective allele, we show maximum likelihood frequency estimates and one standard error bar for each ancient group. The pre-MLBA category corresponds to the sum of all ancient groups before Mönkhkhairkhan. Xiongnu, Early Medieval, and Late Medieval correspond to the sum of all ancient groups in each period correspondingly. Horizontal dashed lines show allele frequency information from the 1000 Genomes Project’s five super populations. (B) Sex-biased patterns of genetic admixture by period and population. We calculated Z scores for every ancient individual who has genetic admixture with WSH-/Iranian-/Han-related ancestry. Positive scores suggest more WSH-/Iranian-/Han-related ancestry on the autosomes, i.e., male-driven admixture. See Figure S2C for individual Z scores.
See this image and copyright information in PMC

References

    1. Allentoft M.E., Sikora M., Sjögren K.-G., Rasmussen S., Rasmussen M., Stenderup J., Damgaard P.B., Schroeder H., Ahlström T., Vinner L. Population genomics of Bronze Age Eurasia. Nature. 2015;522:167–172. - PubMed
    1. Allsen T.T. Population Movements in Mongol Eurasia. In: Amitai R., Biran M., editors. Nomads as Agents of Cultural Change: The Mongols and Their Eurasian Predecessors. University of Hawai’i Press; Honolulu: 2015. pp. 119–151.
    1. Anthony D.W. Princeton University Press; 2010. The Horse, the Wheel, and Language: How Bronze-Age Riders from the Eurasian Steppes Shaped the Modern World.
    1. Bachrach B.S. U of Minnesota Press; 1973. A History of the Alans in the West.
    1. Bat-Oyun T., Erdenetsetseg B., Shinoda M., Ozaki T., Morinaga Y. Who is Making Airag (Fermented Mare’s Milk)? A Nationwide Survey of Traditional Food in Mongolia. Nomad. People. 2015;19:7–29.

Publication types