SCIENCE ADVANCES | RESEARCH ARTICLE

AGRICULTURE Copyright © 2019

The Authors, some
Discontinuous spread of millet agriculture in eastern rights reserved;
exclusive licensee
Asia and prehistoric population dynamics American Association
for the Advancement
of Science. No claim to
C. Leipe1,2*, T. Long3, E. A. Sergusheva4, M. Wagner5*, P. E. Tarasov2 original U.S. Government
Works. Distributed
Although broomcorn and foxtail millet are among the earliest staple crop domesticates, their spread and impacts under a Creative
on demography remain controversial, mainly because of the use of indirect evidence. Bayesian modeling applied Commons Attribution
to a dataset of new and published radiocarbon dates derived from domesticated millet grains suggests that after NonCommercial
their initial cultivation in the crescent around the Bohai Sea ca. 5800 BCE, the crops spread discontinuously across License 4.0 (CC BY-NC).
eastern Asia. Our findings on the spread of millet that intensified during the fourth millennium BCE coincide with
published dates of the expansion of the Sino-Tibetan languages from the Yellow River basin. In northern China,
the spread of millet-based agriculture supported a quasi-exponential population growth from 6000 to 2000 BCE.
While growth continued in northeastern China after 2000 BCE, the Upper/Middle Yellow River experienced
decline. We propose that this pattern of regional divergence is mainly the result of internal and external anthro-
pogenic factors.

Downloaded from http://advances.sciencemag.org/ on September 28, 2019

INTRODUCTION Studies based on phytolith (6) and starch grain (7) analyses claim
Broomcorn (Panicum miliaceum) and foxtail (Setaria italica) millet, a much earlier appearance at, respectively, ca. 8500 to 7500 BCE
often summarized as the East Asian millet cultigens, are two of the (broomcorn millet) and 9500 BCE (foxtail millet) within the Lower
world’s oldest crops. Both were important in many prehistoric food Yellow River and Liao River region. However, this evidence is
production systems across Asia and other parts of the world over controversial because of methodological limitations and/or ambiguous
several millennia (1). Especially in northern China, where they were chronologies (8). Recent studies indicate an early onset of wild grasses
domesticated, broomcorn and foxtail millet played a fundamental role (Paniceae) use and thus a long-term shift to agricultural lifeways that
in the formation of one of the world’s early agricultural systems and might have already started around 20,000 years ago (8). This incomplete
civilizations (2). Although having largely lost significance in worldwide knowledge is mainly due to the absence of documents of a continuous in
agriculture over the last century, they are currently regaining attention situ foraging-farming trajectory and limited insights into pre-Neolithic
in developing strategies for global food security in light of advancing societies owing to a scarce archaeological record (4). Uncertainty about
global warming and a growing world population (3). the spatiotemporal spread of millet cultivation across eastern Asia
Despite their significance as a food source in the ancient Old World, also exists. Existing studies only focus on certain regions and/or are
the process and timing of millet domestication and the transition to based, at least partly, on ambiguous evidence (1, 9).
agriculture in northern China remain unclear (4). Records of directly It has been recognized that plant domestication was a cornerstone
14C-dated carbonized caryopses suggest that cultivation of domesticated for the Neolithic agricultural revolution (10), which led to major
millet first occurred more or less contemporaneously during the sixth transformations in human societies (11), particularly linked with
millennium BCE in the regions of three different Early Neolithic population growth (12), and disturbance of natural environments
cultural complexes (Peiligang, Houli, and Xinglongwa) (2) in the (13). However, there are ongoing debates, on the one hand, about
fertile crescent surrounding the Bohai Sea, i.e., stretching across the the pattern of this population growth including whether it was
Lower Yellow River and Liao River region. The claim that the Upper accelerated and whether it was different from those of foraging
Wei River is also a center of early millet domestication exists (5). societies (12, 14) and, on the other hand, about the intensity of early
This claim relies on eight undated broomcorn millet grains collected agricultural activities and whether they had a global-scale impact on
at the Dadiwan site from a single pit assigned to the occupation environments, which induced the Anthropocene era (15). Previous
phase 1 (5800 to 5300 BCE). All other archaeological remains, however, studies have shown that China, especially the northern part, was
point to a hunter-gatherer subsistence rather than food production already a dynamic global population hot spot since the Neolithic era
(5). This and the fact that Dadiwan is known for intensive millet (16). Although agricultural practices based on domesticated millet
cultivation during the fourth millennium BCE makes contamination and pigs have been identified as a major control on population
from younger layers highly probable. dynamics in this region (4, 17), systematic knowledge about their
spatiotemporal relationship remains poor.
1
Institute for Space-Earth Environmental Research (ISEE), Nagoya University, Research To address these issues, we compiled a set of 14C dates directly
Institutes Building II, Furo-cho, Chikusa-ku, Nagoya 464-8601, Japan. 2Institute of derived from remains of domesticated millet (Fig. 1). Available datings
Geological Sciences, Section Paleontology, Freie Universität Berlin, Malteserstraße extracted from the literature (n = 170; table S1) were supplemented
74-100, Building D, 12249 Berlin, Germany. 3School of Geographical Sciences,
University of Nottingham Ningbo China, 199 Taikang East Road, Yinzhou Qu, Ningbo
by a set of newly obtained data (table S2) from the Khanka-Ussuri
Shi 315100, Zhejiang Sheng, China. 4Institute of History, Archaeology and Ethnography region between China and Russia (n = 7), one of the less well-represented
of the Peoples of the Far East, Far Eastern Branch of the Russian Academy of Sciences, regions in far-eastern continental Asia, and from the Chinese province
Pushkinskaya 89, Vladivostok 690001, Russia. 5Eurasia Department and Beijing Branch of Shandong (n = 7), one of the centers of millet domestication (8).
Office, German Archaeological Institute, Im Dol 2-6, Building II, 14195 Berlin, Germany.
*Corresponding author. Email: c.leipe@fu-berlin.de (C.L.); mayke.wagner@dainst. This dataset was subdivided into eight geographical regions from
de (M.W.) eastern central Asia to the Japanese archipelago (see Materials and

Leipe et al., Sci. Adv. 2019; 5 : eaax6225 25 September 2019 1 of 9

SCIENCE ADVANCES | RESEARCH ARTICLE

Longitude (°E)
60 80 100 120 140
A
Mongolia
Begash

Hokkaido
Latitude (°N)
40

Xiaohe

Ojakly Qinghai
Lake Honshu
Yellow
Sea

Tibet Kyushu
30

300 km
Median calibrated age (BCE/CE × 103)
CE

Downloaded from http://advances.sciencemag.org/ on September 28, 2019

6 4 2 BCE 0

80 100 120 140

Lower Yellow River/Liao River Upper Yangtze River Korean peninsula
Middle Yellow River Central Asia Japanese archipelago
Upper Yellow River Khanka-Ussuri plain
Fig. 1. Spatiotemporal distribution of directly dated millet remains from central and eastern Asia. (A) Topographic map with the locations of directly 14C-dated
foxtail and broomcorn millet remains (n = 184) contained in the compiled datasets (tables S1 and S2) organized into eight geographical regions (see Materials and Methods).
(B) Longitudinal distribution of calibrated median millet ages.

Methods) and analyzed using Bayesian modeling to reconstruct the mainly to eastern and western directions (Figs. 1 and 2). Spreading
spatiotemporal pattern of millet dispersal across eastern Asia. To toward the west and southwest, the crops arrived in the Middle Yellow
test the effect of millet-pig–based agricultural activities on the River region 5000–4400 BCE (median, 4600 BCE), but it took another
development of population size, we compiled and reanalyzed a subset thousand years or more for establishing in the higher elevated re-
(n = 40,696) of an archaeological site distribution database (16) covering gions along the Upper Yangtze River (3700–3100 BCE; median,
13 provinces in northern China (see Materials and Methods) from 3400 BCE) and the Upper Yellow River (3400–3000 BCE; median,
the Early Neolithic to early Iron Age (ca. 8000 to 500 BCE). 3200 BCE). For the region of central Asia, which includes the data
from Xinjiang in the east and from Kazakhstan and Turkmenistan in the
west, the model suggests appearance of millet since 3200–2100 BCE
RESULTS AND DISCUSSION (median, 2400 BCE). However, the 95% probability ranges of the
Millet spread across eastern Asia earliest available millet-based date from the archaeological site of
The results of Bayesian modeling (Fig. 2 and fig. S2) applied to the Begash (2461–2154 BCE; median, 2304 BCE) in southeastern
compiled set of published and newly obtained millet 14C dates suggest Kazakhstan (Fig. 1) indicate that millet likely appeared in the west-
that the crop first appeared in the fertile region stretching from ern central Asian steppes before it was introduced into the Xinjiang
the Lower Yellow River in the south to the Liao River in the north region, where the earliest date comes from the Xiaohe site (2011–1756
around 6100–5700 BCE (95% probability range, as for all time in- BCE; median, 1886 BCE).
tervals for the defined regions presented in the following) with a The reversal in the east-west spread of millet suggested by the data
median age of 5800 BCE. This provides robust support for the common in hand deserves a discussion. The most probable route connecting
assumption of earliest domesticated millet appearance in different the Upper Yellow River and the western central Asian steppes
sedentary cultures within this region around the beginning of the sixth runs along the southwestern slopes of the Mongolian Altai and the
millennium BCE (1). From this domestication core zone, millet spread southern Mongolian Plateau. This agrees with suggested routes for

Leipe et al., Sci. Adv. 2019; 5 : eaax6225 25 September 2019 2 of 9

SCIENCE ADVANCES | RESEARCH ARTICLE

140
Japanese
130 Khanka-Ussuri plain (2889) archipelago
Korean peninsula (3736) (1044)
120

Longitude (°E)
Lower Yellow River/Liao River (5847)
110 Middle Yellow River (4594)
Upper Yellow River (3202)
100 Modeled age boundaries Upper Yangtze River (3364)
95.4% range
90 68.2% range
Central Asia (2448)
80 Median age

6 5 4 3 2 1 0
Age (years BCE × 103)
Fig. 2. The modeled age ranges and medians for the appearance of millet in different regions of central and eastern Asia. Numbers in parentheses show modeled
calibrated median ages BCE. The longitudinal position of a defined region represents the center of the longitudinal range of all dated millet remains contained in the
respective dataset.

the dispersal of West Asian crops such as wheat and barley and extent during the Late Neolithic (3000/2700 to 1500 BCE) (23).

Downloaded from http://advances.sciencemag.org/ on September 28, 2019

animal domesticates (sheep/goat, cattle, and horse) into eastern Asia Currently discussed reasons for this decrease in Chulmun populations,
(18, 19), linked to the spread of agropastoral populations from the whose subsistence economy was largely based on foraging, are declining
central Asian steppe (20). We propose that it was also these populations natural resources related to overexploitation or pressure from
moving into Xinjiang oases along the Tarim, Ili, and other rivers, immigrating farmers related to the early Bronze Age Mumun culture
that introduced not only wheat and barley, which simultaneously (1500 to 800 BCE), which could have pushed parts of the population
appeared there around 1900 BCE (i.e., several centuries later than in to northern regions, such as the fertile and less populated Khanka-­
the eastern part of China) (18), but also millet. The recorded use of Ussuri region (23, 26).
both western and eastern domesticates including wheat, millet, sheep/ Because of the limited number (n = 4) of available direct 14C millet
goat, and cattle (21) and assemblages of Andronovo culture origin datings, the 95% probability range (2900–500 BCE) for the earliest
or at least affinity (22) since the early second millennium BCE in presence of the crop in the Japanese archipelago comprises a com-
different parts of Xinjiang provides evidence for the presence of paratively long time interval. The derived median age of 1000 BCE
central Asian steppe agropastoral populations in this currently is thus more useful for discussion. It is believed that foxtail and
extremely arid region. In addition, increased population numbers are broomcorn millet were introduced to Japan together with rice
also indicated by the substantial rise in archaeological site numbers around the Final Jomon–Initial Yayoi transition (ca. 1000 BCE) and
in Xinjiang from 44 to 153 between the two time slices 2350 and spread northeastward after arriving on Kyushu Island from the
1750 BCE (16), which may point to immigration and/or intensified Korean Peninsula (27). The Yayoi period (ca. 1000 BCE to 300 CE)
food production as a result of changes in subsistence economy. is characterized by the mixing of immigrants from China and/or the
Spreading toward the east, millet appeared on the Korean Peninsula Korean Peninsula and the indigenous Jomon population. It seems
4500–3300 BCE (median, 3700 BCE) during the late Early Neolithic that the introduction of rice and millet into the Japanese archipelago
(6000 to 3500 BCE) associated with the Chulmun culture. Recent reflects a continuation of the spread of both crops to the Korean
evidence from seed impressions on pottery suggests an earlier exis- Peninsula, where they were intensively cultivated since the Bronze
tence since the Initial Neolithic (6000 to 5000/4500 BCE), although no Age (Mumun culture).
direct dates are available for the related sherds (23). Archaeological Although millet seems to have appeared in northern Japan (i.e., the
findings suggest that millet was adopted by existing populations Hokkaido region) relatively late around 689–1015 CE (95% probabi­
rather than introduced by immigration and did not play a substantial lity range of calibrated direct date TO-1998; median, 862 CE), there
role until the Middle Neolithic (3500 to 3000/2700 BCE) (23) when is evidence for an earlier presence by carbonized seed assemblages from
millet cultivation as part of the mixed (foraging-farming) subsistence Epi-Jomon cultural (300 BCE to 700 CE) layers in Hokkaido (28)
economy was enhanced. Full-scale agriculture is documented only and Late Jomon layers in northern Honshu (29). A rice grain asso-
since the Bronze Age associated with the Mumun culture (1500 to ciated with the latter millet assemblage, which was dated (29) and
800 BCE), during which rice, wheat, and barley were also cultivated. recalibrated (95% probability range, 1683–382 BCE; median, 1018
The earliest presence of millet in the agriculturally fertile Khanka-­ BCE), indicates a Late–Final Jomon context. Nondirectly dated fox-
Ussuri region is suggested to be ca. 3700–2500 BCE (median, 2900 tail and broomcorn millet grains were also found in Okhotsk cul-
BCE) and is related to the Late Neolithic Zaisanovka culture (3300/3200 tural layers (ca. 500 to 1000 CE) of six archaeological sites on the
to 1400 BCE), which seems to have spread into the region in several Hokkaido island (30). This culture has also used other domesticated
waves bringing along millet cultivation (24). A broad cultural interaction crops like naked barley, which was not introduced from southern
sphere (“south Manchurian Neolithic sphere”) comprising the regions Japan but from the Russian Far East (30). The so far oldest well-­
of the Liao River/Liaoning, Korean Peninsula, and Khanka-Ussuri documented use of naked barley in the Hokkaido region dates be-
has been identified on the basis of conformity in pottery styles (25). tween 375 and 203 BCE (95% probability range; median, 279 BCE)
There is evidence for a rise in the number of settlement sites in the (31). It is conceivable that, like naked barley, millet spread into northern
western Korean Peninsula by 3000 BCE due to increased farming Japan from the Russian Far East. This hypothesis is supported by
productivity, which was followed by a decrease in site numbers and evidence from DNA analysis of modern landraces, which suggests

Leipe et al., Sci. Adv. 2019; 5 : eaax6225 25 September 2019 3 of 9

SCIENCE ADVANCES | RESEARCH ARTICLE

that barley and broomcorn millet spread to Japan from two different This suggests that substantial population growth in Neolithic
directions, including a southern route via Kyushu and a northern northern China started in the late seventh millennium BCE, which was
route via Hokkaido (32). likely initiated by the onset (ca. 6100–5700 BCE) of millet-based
The Bayesian modeling results demonstrate a discontinuous agriculture. Our results demonstrate that the population growth
westward spread of millet, with an interval of ca. 1200 years (on controlled by millet cultivation was exponential and accompanied
average) between its appearance in the neighboring regions (Fig. 2). by the spread of millet across the northern part of China from the
However, no clear temporal pattern can be found for its eastward Upper Yellow River in the west to the Khanka-Ussuri region in the
dispersal. Although the Korean Peninsula is geographically close to east. The temporal millet dispersal suggests that population growth
the Lower Yellow River and Liao River regions, millet did not was, until 2900 BCE, driven by both intensification and spread of
appear there until ca. 2100 years later, followed by the Khanka-Ussuri millet cultivation and, between 2900 and 1900 BCE, controlled
and Japan regions (another ca. 800 and 2700 years, respectively). exclusively by intensification of agricultural activities. The site
This shows that adoption of agriculture by prehistoric populations number curve, which shows stable low values between 6000 and
is a complex process. Similar complexity appears in the dispersal of 5000 BCE, suggests that the temporal gap in millet datings (ca. 5700
early agriculture from West Asia to Europe along the Danube valley to 4500 BCE; Fig. 1B) and its associated minimum in the estimated
(33). After appearance in southeast Europe (6500–6000 BCE), probability distributions (Fig. 3B) may be an artifact caused by a
agriculture spread relatively quickly across the fertile central European general scarcity of millet cultivation sites and relatively little direct-­
loess regions (5500–5000 BCE) but subsequently paused on its way dated archaeological information available for this period (4). It
to Scandinavia and the British Isles, where it did not arrive before ca. seems plausible that this period represents the early stage of the

Downloaded from http://advances.sciencemag.org/ on September 28, 2019

4000 BCE (33). transition to agriculture, which is often attributed to deteriorating
Further evidence against the long-standing hypothesis of a pro- health conditions and declining life expectancy (36). These unfavorable
gressive spread of agriculture (and thus, “civilization”) from the region conditions at the onset of a fully agricultural lifestyle may have
of the Middle Yellow River, traditionally seen as the core area of also constrained population growth and millet spread in northern
Chinese Neolithic tradition and Chinese civilization (7), is provided China.
by the modeled spread of agriculture across Britain and Ireland The decrease in probability density of millet datings at the turn
(34). Starting in southeast England, it reached northwest Scotland of the third and second millennium BCE indicates reduced millet
earlier than south Scotland. The latter region, for a complex of reasons, cultivation, which may explain the apparently rapid and substantial
did not become the first choice for agriculture spread to the north, depopulation by 1900 BCE (Fig. 3). Both curves suggest that diminished
as did Xinjiang not become for the millet spread to the west (Fig. 2) site numbers prevailed until the middle second millennium when
and for the wheat spread to the east (18). agricultural activities and population size started to reincrease. We
Disregarding its mode, the inferred long-term spread of millet demonstrate, however, that this trend was not universal throughout
agriculture and associated population dynamics adds to ongoing northern China (16). The archaeological site dataset shows (Fig. 4)
debates about the evolutionary history and divergence of the Sino-­ that it was mainly bounded to the region between 94°E and 115°E
Tibetan language family (35). The reconstructed patterns of millet and associated with the coeval decline of several Late Neolithic
spread in China that intensified during the fourth millennium BCE cultures (Majiayao, Longshan, Qijia, and Laohushan) flourishing in
concur with results of a recent Bayesian phylogenetic modeling study the Middle and Upper Yellow River regions. At the same time, the
(35), which dates the major dispersal of the Sino-Tibetan language region between 115°E and 131°E demonstrates a generally steady
populations from their core area in the Yellow River basin between population growth interrupted by only a weak reduction in population
ca. 5850 and 2250 BCE (average value, 3950 BCE). Further progress size around 2000 BCE.
in assessing the relationship (that is, coincidence or causality) between Numerous studies have tried to identify the driving forces behind
the spread of millet farming and dispersal of languages in northern this population/cultural decline at the Neolithic–Bronze Age transition
China and eastern Asia can be achieved through interdisciplinary in different parts of northern China. The most frequently argued
projects, including phylogenetic modeling and a systematic analysis causes are changing climate conditions characterized by higher
of archaeological and paleoenvironmental records. variability leading to more unstable environments and an enhanced
trend toward lower temperatures and moisture availability related
Population dynamics in millet-based farming societies to the 4.2-kiloyear event and/or the continuous middle-late Holocene
of northern China decline in Asian summer monsoon intensity (37–40). Although the
Changes in archaeological site numbers contained in a dataset from importance of climatic conditions in the life of prehistoric farmers
China were interpreted as qualitative estimates for changes in population in China is frequently acknowledged (5, 16, 40), geoscientists and
size (16). Here, we reanalyze this dataset to address large-scale changes archaeologists still have difficulties in determining where, when,
in the core area of millet-based agriculture and Chinese civilization and how changes in climate affected the ancient population (41).
(see Materials and Methods). The results (Fig. 3A) show the following Many of the published archaeological and paleoclimatic records
trends: an increasing growth from 21 to 2220 sites per century from often allow controversial interpretations due to the lack of high enough
6400 to 1900 BCE, best explained by an exponential regression resolution and accurate dating (16). Recent high-resolution studies
curve; a period of decreased site numbers (to 1195 sites per century, also reveal a rather complex and even asynchronous precipitation/
1900 to 1600 BCE); and a period of discontinuous reincrease to humidity trends in different parts of China (42). This further
2420 sites per century until 1000 BCE. The calibrated probability complicates the assessment of the potential impact of Holocene
distribution of the millet-based datings (Fig. 3B) stretching from climate change on China’s population, agricultural development, and
the Upper Yellow River region to the Khanka-Ussuri region (ca. subsistence strategies and requires the highest standards of proof in
98°E to 133°E) corroborates this trend. each particular case.

Leipe et al., Sci. Adv. 2019; 5 : eaax6225 25 September 2019 4 of 9

SCIENCE ADVANCES | RESEARCH ARTICLE

8 7 6 5 4 3 2 1 0
25
A

Site number per 100 years (× 102)

20

15

10

Probability density (× 10−2)

B
0

Downloaded from http://advances.sciencemag.org/ on September 28, 2019

6

8 7 6 5 4 3 2 1 0
Age (years BCE × 103)
Fig. 3. Archaeological site numbers in northern China, longitudinal spread of millet in eastern Asia, and probability distribution of available millet datings.
(A) Development of archaeological site numbers per 100 years in northern China (ca. 94°E to 131°E) between 8000 and 500 BCE. (B) Probability density distribution of
direct millet radiocarbon datings located between 98°E and 133°E based on kernel density estimation (KDE) modeling using the KDE_Model approach implemented in
OxCal v4.3.2. The dark gray silhouette is the sampled KDE estimated distribution. The blue line and lighter blue band show the mean ± 1 for snapshots of the KDE distribution
generated during the Markov chain Monte Carlo process. The light gray silhouette represents the stacked (summed) distribution of all datings plotted using the Sum
function in OxCal v4.3.2 (see Materials and Methods).

The results summarized in Fig. 4 also require careful consideration climate cooling enhanced after 2000 BCE had a negative impact on
before giving priority to climatic or nonclimatic factors. Both the Middle the yields of millet and led agropastoralists to diversify their crops,
and Upper Yellow River regions west of 115°E and northeastern regions introducing more cold-resistant wheat and barley. Although crop
east of 115°E are located at the modern limits of the Asian summer diversification after 2000 BCE is also documented in the archaeological
monsoon (42). Thus, the environments there must have been similarly record from northeastern China (5, 18), the probability density of
vulnerable to precipitation change. Moreover, climate instability millet-based 14C dates does not indicate a decrease in millet cultivation
would be even more pronounced in the generally colder northeast. but suggests that millet continued to be a staple in the region during
Compared to other cereals, millet is well adapted to short growing the second millennium BCE (Fig. 4B).
seasons, dry and infertile soils with poor water-holding capacity, The drop in archaeological site numbers and in the probability
irregular precipitation, and droughts (1), making aridification and/ density of millet-based 14C dates in north-central China (Fig. 4A)
or shorter summers less likely a main driver of the rapid and strong and the synchronous increase in site numbers in the northeastern
agriculture/population decline. A recent review of evidence from region (Fig. 4B) during the early second millennium BCE, only at
paleoenvironmental reconstructions illustrated the complex signature first glance, appear to be the retreat of farmers from the Middle Yellow
of the 4.2-kiloyear event across the Northern Hemisphere, showing River catchment that became progressively drier compared to the rela-
variations in timing and intensity among the different regions (43). tively humid Liao River region. However, such a straightforward
Regarding eastern Asia, several authors argued that a century-scale climatic interpretation cannot explain why the still warm and humid
oscillation toward colder/drier climate around 4200 years ago strongly middle and lower reaches of the Yangtze River shared the same fate
affected large parts of China. However, more recent high-resolution (i.e., habitation collapse and emigration of agricultural population) as
and well-dated records from northeastern and southern China (42) the Middle Yellow River region between ca. 2350 and 1750 BCE (16).
and from northern Japan (31) corroborate only a short-term phase We suggest that not climate change but economic development
of weak cooling accompanied by a slight decrease in generally high and associated social stratification, high population density, and
precipitation values. In line with this finding, it seems that the concentration in the growing settlements and their intensified
4.2-kiloyear event had no (or very little) impact on the archaeological contacts with the Asian steppes could have played a major role in
site and millet data presented in Fig. 4. the population and cultural decline at the end of the third millennium
A crop niche modeling study (41) demonstrates that in parts of BCE. Hosner et al. (16) were the first who argued for this relationship
the Tibetan Plateau and in higher-elevated regions of central Asia, and discussed as the possible main cause the spread of plagues by

Leipe et al., Sci. Adv. 2019; 5 : eaax6225 25 September 2019 5 of 9

SCIENCE ADVANCES | RESEARCH ARTICLE

16 River might also reflect rising tensions and conflicts between indig-

Site number per 100 years (× 102)

A North-central China
enous farmers and western pastoralists. Further support for com-
(94°E to 115°E) petition and conflict lies in the replacement of different agricultural
12
societies by agropastoralists in the eastern Gansu-Qinghai region
(Upper Yellow River) and southern central Inner Mongolia (Middle
8 Yellow River) (40). This context provides an alternative explanation
for the increasing number of fortifications and site clustering since
the Late Neolithic and the development of urbanism and formation
4 of early states during the Bronze Age. Instead of marked climate
Probability density (× 10−2)

changes (40), which are most widely considered as main and often
Middle Yellow R.
6 the only triggers (5), we suggest that it might have been pressure
0
Upper Yellow R. from internal population growth and external anthropogenic forcing
4
that initialized the changes in the cultural landscape expressed by
2 the development of centralized, hierarchical settlement systems. There
is increasing evidence for broad expansion processes of Eurasian
0 steppe populations during the third millennium BCE toward the west
12 and east from archaeological records and genetic studies (44–46).

Site number per 100 years (× 102)

Westward migration of so-called Yamnaya steppe herders from

Downloaded from http://advances.sciencemag.org/ on September 28, 2019

B Northeastern China
eastern Europe in the early Bronze Age (around 2500 BCE) was
(115°E to 133°E)
8 even so massive that it led to the large-scale replacement of populations
in central Europe (47). Our study argues that a similar scenario
could be applied to northern China.
For northeastern China, our results suggest prolonged population
4
growth underpinned by continued millet-based agriculture beyond
the end of the Neolithic period ca. 2000 BCE (Fig. 4B), which might
Probability density (× 10−2)

have been augmented by immigration from the Yellow River region.

3 A population shift is also indicated by the appearance of massive
0
2 fortifications after 2000 BCE, which originated in the Middle Yellow
River region (48). Our findings contradict claims of a Late Neolithic
1
(third millennium BCE) population decline in northeastern China
0 and the perception of a general transformation from sedentary
8 7 6 5 4 3 2 1 0
agriculture to pastoral nomadism around 2000 BCE (40, 49).
Age (years BCE × 103)
Fig. 4. Archaeological site numbers and probability distribution of available
CONCLUDING REMARKS
millet datings for north-central and northeastern China. (A) Development of
archaeological site numbers per 100 years in north-central China between 8000 and
The results presented in this study shed new light on the relationship
500 BCE and KDE probability density distribution of direct millet radiocarbon among millet-based agricultural activities, the spread of crops and
datings (west of 115°E). (B) As in (A) for northeastern China (east of 115°E). Error bars technologies, and population dynamics in large parts of Neolithic
and vertical black strokes in (A) indicate, respectively, the 95% probability range and Bronze Age eastern Asia. Our findings suggest that millet-based
and median age of the estimated onset of wheat cultivation in the Middle and agricultural systems appeared around 6000 BCE in different parts of
Upper Yellow River regions (18). See caption of Fig. 3 for further details. the Lower Yellow River and Liao River reaches and do not provide
evidence for early millet cultivation in the Wei River region tradi-
tionally seen as one of the core areas of Chinese civilization. Comparable
Afanasievo and/or Andronovo cultural populations and its possible with the dispersal of West Asian crops across Europe and China,
impact on the densely populated village clusters and fortified centers millet spread was discontinuous across eastern Asia. This suggests
along the Yellow River and possibly also in the Middle and Low that adaptation of agriculture by prehistoric populations is a nonlinear
Yangtze River regions. It seems conceivable that expansion of process depending on the interaction of various anthropogenic and
semisedentary agropastoral populations may have involved both environmental factors.
increasing cultural pressure in the form of competition for natural The shift to agricultural practices induced an exponential population
resources (e.g., copper and pastoral grounds) as well as the intro- growth in northern China, which continued until ca. 2000 BCE. We
duction and spread of plague epidemics. The advance of steppe hypothesize that the following decline in millet cultivation and
cultural elements is documented during the Late Neolithic by the population size in north-central China is mainly the result of enhanced
appearance of domesticates (such as sheep/goat, cattle, barley, and expansion of agropastoralists leading to competition with indigenous
wheat) and technologies (such as metallurgy) of West Asian origin. farmers for natural resources (e.g., copper and pastoral grounds)
The estimated (95% probability range) onset of wheat cultivation in likely in combination with the spread of plague epidemics. It seems
the Upper and Middle Yellow River regions ca. 2000–1700 BCE and plausible that the Neolithic farmers of northern China faced similar
ca. 1800 –1500 BCE, respectively (18), is contemporaneous with the profound immigration pressure at about the same time as their
observed population gap there (Fig. 4A). This population decrease counterparts in central Europe, who were largely replaced by expanding
in the “core areas of Chinese civilization” (5) along the Yellow eastern European pastoralists in the early Bronze Age around 2500 BCE

Leipe et al., Sci. Adv. 2019; 5 : eaax6225 25 September 2019 6 of 9

SCIENCE ADVANCES | RESEARCH ARTICLE

(45, 47). These migration processes exemplify the important role of in the respective region, and the lower boundary of that modeled
the steppe cultural sphere in Eurasian prehistory at least since the phase (18) was adopted as its age estimate. Outliers were detected
third millennium BCE. The recognized expansion of agropastoral on the basis of the OxCal agreement index calculation. The modeled
population might have been promoted by improving moisture ages are presented as both 95% probability range and median (point
availability in the central Asian and eastern European steppe regions estimate).
over the middle and late Holocene (50). One reason for the construction
14
of massive fortifications by populations in northern China may be C probability distribution
the need for protection from this “migration threat.” If this hypothesis Probability density distribution was calculated based on a selection
is true, these defense works might represent precursors of the walls (n = 144) of the compiled set (n = 184) of millet-based 14C datings
built some 2000 years later by early feudal states during the pre-imperial representing five of the defined geographical regions (ca. 98°E to
late Eastern Zhou dynasty, also called the “Warring States” period 133°E) including the Upper Yellow River, Middle Yellow River, Lower
(475 to 221 BCE), in response to interstate conflict and for protecting Yellow River and Liao River, Korean Peninsula, and Khanka-Ussuri,
against hostile nomadic societies, which later developed into what is roughly representing the region covered by the archaeological site
today known as the Great Wall of China, the largest military structure dataset used. For analyzing spatial differences, the subset of 14C
in world history. dates was divided along 115°E, resulting in a western (n = 90), rep-
resenting north-central China, and eastern (n = 54), representing
northeastern China, dataset. 14C dates <2000 uncalibrated years be-
MATERIALS AND METHODS fore the present were not considered. The probability density distri-

Downloaded from http://advances.sciencemag.org/ on September 28, 2019

Calibration of radiocarbon dates bution and stacked (summed) distribution were derived from kernel
Newly obtained 14C dates and the reported 14C date of a rice grain density estimation (KDE) modeling using the KDE_Model approach
from northern Honshu (29) were calibrated to calendar ages using (52) implemented in OxCal v4.3.2.
OxCal v4.2.3 (https://c14.arch.ox.ac.uk/oxcal.html) and the IntCal13
curve (51). Archaeological site data
We used archaeological site data extracted from an existing database
Compilation of millet-based 14C dates (16) to estimate the Neolithic–Bronze Age population development
In addition to the newly obtained datings from millet caryopses in the core region of early millet-based (dry-land) agriculture and
(n = 14), we compiled 170 directly dated millet-based 14C dates from Chinese civilization in northern China. The original database contains
available publications. Reported millet finds with ages derived from a total of 51,074 archaeological sites from the Early Neolithic to the
typologies or stratigraphically correlated material (e.g., carbon-rich early Iron Age (ca. 8000 to 500 BCE) and covers most regions of China
soil and charcoal) were not included in the dataset. These correlations (i.e., site data from 25 Chinese provinces, autonomous regions, and mu-
may introduce additional errors, including those resulting from nicipalities, published in the series Atlas of Chinese Cultural Relics). The
postdepositional processes or the “old wood effect,” which may lead entire dataset is also available in the open access database PANGAEA
to age shifts of hundreds or even thousands of years (18). Additional Data Publisher for Earth & Environmental Science (https://doi.pangaea.
care should be taken when equating radiocarbon dates with demography. de/10.1594/PANGAEA.860072). It should be acknowledged that
The decreasing numbers of millet-based 14C dates in the Middle overreliance on archaeological site numbers as population proxy
and Upper Yellow River regions of north-central China after 2000 can be critical, given the heterogeneous nature of the base data (16).
BCE (Fig. 4A) could reflect a decrease in site numbers and population This issue was addressed in detail by Hosner et al. (16), to which the
densities or could be an artifact of sampling strategy, for example, reader is referred to for further discussion and references. They also
preferentially selecting for dating grains of wheat and barley when performed a comparison of the estimated population (53) with the
they begin to appear in cultural layers during this period of time. archaeological site numbers obtained in their study and reported
Such a bias, if it exists, should more or less equally affect the probability rather similar trends in both curves, thus suggesting a reasonably
density of millet-based 14C dates in both parts of northern China good correlation between number/density of archaeological sites
shown in Fig. 4 and would contradict the independently collected and prehistoric population estimates in China. Being aware of the
archaeological site dataset (16) and interpretations based on archae- existing biases in data collection, we put accent in the discussion not
ological materials (5). Since this is not the case, we assume that the on the absolute values but on the major trends (i.e., on qualitative
data used in this study are suitable for qualitative interpretation and rather than quantitative parameters).
justify the reliability of our conclusions. The compiled subset of 40,696 sites represents 13 Chinese provinces
including Beijing, Gansu, Hebei, Henan, Inner Mongolia Autonomous
Bayesian chronological modeling Region, Jilin, Liaoning, Ningxia Hui Autonomous Region, Qinghai,
We constructed a Bayesian chronological model (fig. S1) to estimate Shaanxi, Shandong, Shanxi, and Tianjin (ca. 32°N to 51°N, 90°E to
the appearance of millet in different parts of eastern Asia. Our approach 131°E), covering the period 8000 to 500 BCE. The subset represents
generally follows the one used by Long et al. (18), which the reader the areas with directly dated millet grains that originate from a clearly
is referred to for further references. The employed model uses the millet-based agricultural context. Each contained archaeological site
overlapping multiphase model for the overall structure and the in this subset is assigned to one or more well-dated cultural periods.
phase model as a building block (i.e., a submodel) implemented in This implies that the sites represent different time ranges, which
OxCal v4.3.2. The compiled millet-based 14C datings (n = 184) were vary between 100 and 6000 years. To temporally normalize the site
organized into eight geographical regions (listed in Fig. 1), each of data eliminating influence of the length of the defined cultural periods,
which is represented by a submodel. Each submodel conceptually the site numbers are presented by time intervals of equal length, which
represents the archaeological phase during which millet appeared we randomly set to 100 years. In addition, to allow determination of

Leipe et al., Sci. Adv. 2019; 5 : eaax6225 25 September 2019 7 of 9

SCIENCE ADVANCES | RESEARCH ARTICLE

longitudinal differences in site number development in response to 19. J. R. Dodson, X. Li, X. Zhou, K. Zhao, N. Sun, P. Atahan, Origin and spread of wheat
in China. Quat. Sci. Rev. 72, 108–111 (2013).
the evolution of millet-based agriculture, the compiled dataset was
20. L. Barton, C.-B. An, An evaluation of competing hypotheses for the early adoption
divided in concert with the set of millet 14C datings into two subsets of wheat in East Asia. World Archaeol. 46, 775–798 (2014).
along 115°E representing north-central China and northeastern 21. Z. Qiu, Y. Yang, X. Shang, W. Li, Y. Abuduresule, X. Hu, Y. Pan, D. K. Ferguson, Y. Hu,
China, respectively. It should be mentioned that we are only relying C. Wang, H. Jiang, Paleo-environment and paleo-diet inferred from Early Bronze Age cow
on radiocarbon dating when discussing the probability density of dung at Xiaohe Cemetery, Xinjiang, NW China. Quat. Int. 349, 167–177 (2014).
22. P. W. Jia, A. Betts, D. Cong, X. Jia, P. D. Dupuy, Adunqiaolu: New evidence
directly dated millet remains, while the age determinations in the for the Andronovo in Xinjiang, China. Antiquity 91, 621–639 (2017).
archaeological site database are made by the respective regional 23. G.-A. Lee, in Handbook of East and Southeast Asian Archaeology, J. Habu, P. V. Lape,
teams of Chinese archaeologists, who used all available dating J. W. Olsen, Eds. (Springer, 2017), pp. 451–481.
approaches (e.g., absolute dates, pottery typology, and other sources 24. E. A. Sergusheva, Y. E. Vostretsov, in From Foragers to Farmers: Papers in Honour of Gordon
C. Hillman, A. S. Fairbairn, E. Weiss, Eds. (Oxbow Books, 2009), pp. 205–219.
of information), which makes the two datasets involved in the analysis
25. S. V. Alkin, Ancient Cultures of North-East China: The Neolithic Epoch of South Manchuria,
(Figs. 3 and 4) independently and securely dated. This is particularly V. Y. Larichev, Ed. (The Institute of Archaeology and Ethnography Press, Novosibirsk,
important, keeping in mind the tendency of Chinese archaeologists 2007).
to have more confidence in the chronological usefulness of ceramics 26. G.-A. Lee, The transition from foraging to farming in prehistoric Korea. Curr. Anthropol. 52,
and other chronologically sensitive materials when dating sites and S307–S329 (2011).
27. H. Nasu, A. Momohara, The beginnings of rice and millet agriculture in prehistoric Japan.
cultural layers from first millennium BCE contexts. Quat. Int. 397, 504–512 (2016).
28. A. C. D’Andrea, Archaeobotanical evidence for Zoku-Jomon subsistence at

Downloaded from http://advances.sciencemag.org/ on September 28, 2019

SUPPLEMENTARY MATERIALS the Mochiyazawa site, Hokkaido, Japan. J. Archaeol. Sci. 22, 583–595 (1995).
Supplementary material for this article is available at http://advances.sciencemag.org/cgi/ 29. A. C. D’Andrea, G. W. Crawford, M. Yoshizaki, T. Kudo, Late Jomon cultigens
content/full/5/9/eaax6225/DC1 in northeastern Japan. Antiquity 69, 146–152 (1995).
Fig. S1. Bayesian model structure and OxCal coding. 30. G. Yamada, Y. Tsubakisaka, in Prehistoric and Ancient Cultigens in the Far East,
Fig. S2. OxCal modeling results. H. Obata, Ed. (University of Kumamoto, 2008), vol. 3, pp. 303–310.
Table S1. Available 14C dates directly derived from remains of domesticated millet extracted 31. C. Leipe, S. Müller, K. Hille, H. Kato, F. Kobe, M. Schmidt, K. Seyffert, R. Spengler III,
from the literature. M. Wagner, A. W. Weber, P. E. Tarasov, Vegetation change and human impacts on Rebun
Table S2. Newly obtained 14C dates directly derived from remains of domesticated millet from Island (Northwest Pacific) over the last 6000 years. Quat. Sci. Rev. 193, 129–144
prehistoric cultural layers in the Khanka-Ussuri region and the Chinese province of Shandong. (2018).
32. H. V. Hunt, M. G. Campana, M. C. Lawes, Y.-J. Park, M. A. Bower, C. J. Howe, M. K. Jones,
Genetic diversity and phylogeography of broomcorn millet (Panicum miliaceum L.) across
REFERENCES AND NOTES Eurasia. Mol. Ecol. 20, 4756–4771 (2011).
1. N. F. Miller, R. N. Spengler, M. Frachetti, Millet cultivation across Eurasia: Origins, spread, 33. J.-P. Bocquet-Appel, S. Naji, M. Van der Linden, J. Kozlowski, Understanding the rates
and the influence of seasonal climate. Holocene 26, 1566–1575 (2016). of expansion of the farming system in Europe. J. Archaeol. Sci. 39, 531–546 (2012).
2. Z. Zhao, New archaeobotanic data for the study of the origins of agriculture in China. 34. A. Whittle, F. Healy, A. Bayliss, Gathering Time: Dating the Early Neolithic Enclosures of
Curr. Anthropol. 52, S295–S306 (2011). Southern Britain and Ireland (Oxbow Books, 2015).
3. R. Saxena, S. Vanga, J. Wang, V. Orsat, V. Raghavan, Millets for food security in the context 35. M. Zhang, S. Yan, W. Pan, L. Jin, Phylogenetic evidence for Sino-Tibetan origin in northern
of climate change: A review. Sustainability 10, 2228 (2018). China in the Late Neolithic. Nature 569, 112–115 (2019).
4. G. Shelach-Lavi, The Archaeology of Early China (Cambridge Univ. Press, 2015). 36. M. N. Cohen, G. J. Armelagos, Paleopathology at the Origins of Agriculture (Academic Press,
5. L. Liu, X. Chen, The Archaeology of China: From the Late Paleolithic to the Early Bronze Age 1984).
(Cambridge World Archaeology Series, Cambridge Univ. Press, 2012). 37. C. C. Huang, J. Pang, X. Zha, H. Su, Y. Jia, Extraordinary floods related to the climatic event
6. H. Lu, J. Zhang, K.-b. Liu, N. Wu, Y. Li, K. Zhou, M. Ye, T. Zhang, H. Zhang, X. Yang, L. Shen, at 4200 a BP on the Qishuihe River, middle reaches of the Yellow River, China.
D. Xu, Q. Li, Earliest domestication of common millet (Panicum miliaceum) in East Asia Quat. Sci. Rev. 30, 460–468 (2011).
extended to 10,000 years ago. Proc. Natl. Acad. Sci. U.S.A. 106, 7367–7372 (2009). 38. Y. Wang, H. Cheng, R. L. Edwards, Y. He, X. Kong, Z. An, J. Wu, M. J. Kelly, C. A. Dykoski,
7. X. Yang, Z. Wan, L. Perry, H. Lu, Q. Wang, C. Zhao, J. Li, F. Xie, J. Yu, T. Cui, T. Wang, M. Li, X. Li, The Holocene Asian monsoon: Links to solar changes and North Atlantic climate.
Q. Ge, Early millet use in northern China. Proc. Natl. Acad. Sci. U.S.A. 109, 3726–3730 (2012). Science 308, 854–857 (2005).
8. G. W. Crawford, in Handbook of East and Southeast Asian Archaeology, J. Habu, P. V. Lape, 39. W. Wu, T. Liu, Possible role of the “Holocene Event 3” on the collapse of Neolithic Cultures
J. W. Olsen, Eds. (Springer, 2017), pp. 421–435. around the Central Plain of China. Quat. Int. 117, 153–166 (2004).
9. C. J. Stevens, C. Murphy, R. Roberts, L. Lucas, F. Silva, D. Q. Fuller, Between China 40. F. Liu, Z. Feng, A dramatic climatic transition at ~4000 cal. yr BP and its cultural responses
and South Asia: A Middle Asian corridor of crop dispersal and agricultural innovation in Chinese cultural domains. Holocene 22, 1181–1197 (2012).
in the Bronze Age. Holocene 26, 1541–1555 (2016). 41. J. d’Alpoim Guedes, R. K. Bocinsky, Climate change stimulated agricultural innovation
10. J. L. Weisdorf, From foraging to farming: Explaining the neolithic revolution. and exchange across Asia. Sci. Adv. 4, eaar4491 (2018).
J. Econ. Surv. 19, 561–586 (2005). 42. M. Stebich, K. Rehfeld, F. Schlütz, P. E. Tarasov, J. Liu, J. Mingram, Holocene vegetation
11. J. Diamond, Evolution, consequences and future of plant and animal domestication. and climate dynamics of NE China based on the pollen record from Sihailongwan Maar
Nature 418, 700–707 (2002). Lake. Quat. Sci. Rev. 124, 275–289 (2015).
12. J.-P. Bocquet-Appel, When the world’s population took off: The springboard 43. T. P. Roland, C. J. Caseldine, D. J. Charman, C. S. M. Turney, M. J. Amesbury, Was there
of the Neolithic Demographic Transition. Science 333, 560–561 (2011). a ‘4.2 ka event’ in Great Britain and Ireland? Evidence from the peatland record.
13. A. Goudie, The Human Impact on the Natural Environment: Past, Present, and Future Quat. Sci. Rev. 83, 11–27 (2014).
(John Wiley & Sons, ed. 7, 2013). 44. M. E. Allentoft, M. Sikora, K.-G. Sjögren, S. Rasmussen, M. Rasmussen, J. Stenderup,
14. H. J. Zahid, E. Robinson, R. L. Kelly, Agriculture, population growth, and statistical analysis P. B. Damgaard, H. Schroeder, T. Ahlström, L. Vinner, A.-S. Malaspinas, A. Margaryan,
of the radiocarbon record. Proc. Natl. Acad. Sci. U.S.A. 113, 931–935 (2016). T. Higham, D. Chivall, N. Lynnerup, L. Harvig, J. Baron, P. D. Casa, P. Dąbrowski, P. R. Duffy,
15. A. M. Bauer, E. C. Ellis, The Anthropocene divide: Obscuring understanding of  A. V. Ebel, A. Epimakhov, K. Frei, M. Furmanek, T. Gralak, A. Gromov, S. Gronkiewicz,
social-environmental change. Curr. Anthropol. 59, 209–227 (2018). G. Grupe, T. Hajdu, R. Jarysz, V. Khartanovich, A. Khokhlov, V. Kiss, J. Kolář, A. Kriiska,
16. D. Hosner, M. Wagner, P. E. Tarasov, X. Chen, C. Leipe, Spatiotemporal distribution I. Lasak, C. Longhi, G. McGlynn, A. Merkevicius, I. Merkyte, M. Metspalu, R. Mkrtchyan,
patterns of archaeological sites in China during the Neolithic and Bronze Age: V. Moiseyev, L. Paja, G. Pálfi, D. Pokutta, Ł. Pospieszny, T. D. Price, L. Saag, M. Sablin,
An overview. Holocene 26, 1576–1593 (2016). N. Shishlina, V. Smrčka, V. I. Soenov, V. Szeverényi, G. Tóth, S. V. Trifanova, L. Varul,
17. X. Li, J. Dodson, J. Zhou, X. Zhou, Increases of population and expansion of rice agriculture M. Vicze, L. Yepiskoposyan, V. Zhitenev, L. Orlando, T. Sicheritz-Pontén, S. Brunak,
in Asia, and anthropogenic methane emissions since 5000 BP. Quat. Int. 202, 41–50 (2009). R. Nielsen, K. Kristiansen, E. Willerslev, Population genomics of Bronze Age Eurasia.
18. T. Long, C. Leipe, G. Jin, M. Wagner, R. Guo, O. Schröder, P. E. Tarasov, The early history Nature 522, 167–172 (2015).
of wheat in China from 14C dating and Bayesian chronological modelling. Nat. Plants 4, 45. D. W. Anthony, The Horse, the Wheel, and Language: How Bronze-Age Riders from the
272–279 (2018). Eurasian Steppes Shaped the Modern World (Princeton Univ. Press, 2007).

Leipe et al., Sci. Adv. 2019; 5 : eaax6225 25 September 2019 8 of 9

SCIENCE ADVANCES | RESEARCH ARTICLE

46. C. Gaunitz, A. Fages, K. Hanghøj, A. Albrechtsen, N. Khan, M. Schubert, A. Seguin-Orlando, Acknowledgments: The current study contributes to the PAGES-sponsored program
I. J. Owens, S. Felkel, O. Bignon-Lau, P. de Barros Damgaard, A. Mittnik, A. F. Mohaseb, “LandCover6k: Global anthropogenic land-cover change and its role in past climate” and to the
H. Davoudi, S. Alquraishi, A. H. Alfarhan, K. A. S. Al-Rasheid, E. Crubézy, N. Benecke, MEXT Grant-in-Aid project “Cultural History of PaleoAsia” (MEXT Grant-in-Aid no. 1802). We
S. Olsen, D. Brown, D. Anthony, K. Massy, V. Pitulko, A. Kasparov, G. Brem, M. Hofreiter, express our special thanks to G. Jin (Shandong University, China and Reading University, UK),
G. Mukhtarova, N. Baimukhanov, L. Lõugas, V. Onar, P. W. Stockhammer, J. Krause, T. Goslar (Poznan Radiocarbon Laboratory), and the Russian archaeologists N. A. Klyuev and
B. Boldgiv, S. Undrakhbold, D. Erdenebaatar, S. Lepetz, M. Mashkour, A. Ludwig, A. V. Garkovik (Institute of History, Archaeology and Ethnography, Russian Academy of Sciences,
B. Wallner, V. Merz, I. Merz, V. Zaibert, E. Willerslev, P. Librado, A. K. Outram, L. Orlando, Far Eastern Branch, Vladivostok, Russia), E. V. Krutykh (Scientific and Production Centre of
Ancient genomes revisit the ancestry of domestic and Przewalski’s horses. Science 360, Historical and Cultural Expertise, Vladivostok, Russia), and S. A. Kolomiets (Pacific State Medical
111–114 (2018). University, Vladivostok, Russia) for providing various help in collecting the 14C-dated plant
47. W. Haak, I. Lazaridis, N. Patterson, N. Rohland, S. Mallick, B. Llamas, G. Brandt, remains from archaeological sites in the Russian Far East. We also extend our thanks to the
S. Nordenfelt, E. Harney, K. Stewardson, Q. Fu, A. Mittnik, E. Bánffy, C. Economou, anonymous reviewers who made valuable suggestions, which allowed the improvement of the
M. Francken, S. Friederich, R. G. Pena, F. Hallgren, V. Khartanovich, A. Khokhlov, manuscript. Funding: Obtaining the dated plant remains from archaeological sites in the
M. Kunst, P. Kuznetsov, H. Meller, O. Mochalov, V. Moiseyev, N. Nicklisch, S. L. Pichler, Russian Far East was possible because of the financial support of the Russian Foundation for
R. Risch, M. A. Rojo Guerra, C. Roth, A. Szécsényi-Nagy, J. Wahl, M. Meyer, J. Krause, Basic Research (project no. 13-06-12027-офи-м). C.L. acknowledges financial support by a
D. Brown, D. Anthony, A. Cooper, K. W. Alt, D. Reich, Massive migration from the  long-term Research Fellowship (grant LE3508/2-1) granted by the German Research
steppe was a source for Indo-European languages in Europe. Nature 522, 207–211 Foundation. The work of T.L. was supported by the FoSE New Researchers Grant by the
(2015). University of Nottingham Ningbo China. Author contributions: C.L., E.A.S., M.W., and P.E.T.
48. G. J. Tian, in Proceedings of the International Conference to Commemorate the 60th designed the research. E.A.S. performed botanical identification and contributed millet samples
Anniversary of the Excavation of Chengziya Site, X. H. Zhang, Ed. (Shandong Univ. Press, 1993), from field campaigns for 14C dating. C.L., T.L., M.W., and P.E.T. constructed the 14C dataset. T.L.
pp. 119–135. and P.E.T. designed the Bayesian chronological model. C.L. modeled the probability density
49. X. Li, Development of Social Complexity in the Liaoxi Area, Northeast China distribution of the 14C dates and reanalyzed the archaeological site data. T.L., E.A.S., and M.W.

Downloaded from http://advances.sciencemag.org/ on September 28, 2019

(BAR International Series 1821, Archaeopress, 2008). contributed to the Introduction and Results and Discussion sections. C.L. and P.E.T. wrote the
50. T. Kleinen, P. Tarasov, V. Brovkin, A. Andreev, M. Stebich, Comparison of modeled manuscript and compiled the figures. Competing interests: The authors declare that they have
and reconstructed changes in forest cover through the past 8000 years: Eurasian no competing interests. Data and materials availability: All data needed to evaluate the
perspective. Holocene 21, 723–734 (2011). conclusions in the paper are present in the paper and/or the Supplementary Materials.
51. P. J. Reimer, E. Bard, A. Bayliss, J. W. Beck, P. G. Blackwell, C. B. Ramsey, C. E. Buck, Additional data related to this paper may be requested from the authors.
H. Cheng, R. L. Edwards, M. Friedrich, P. M. Grootes, T. P. Guilderson, H. Haflidason,
I. Hajdas, C. Hatté, T. J. Heaton, D. L. Hoffmann, A. G. Hogg, K. A. Hughen, K. F. Kaiser, Submitted 8 April 2019
B. Kromer, S. W. Manning, M. Niu, R. W. Reimer, D. A. Richards, E. M. Scott, J. R. Southon, Accepted 28 August 2019
R. A. Staff, C. S. M. Turney, J. van der Plicht, IntCal13 and Marine13 radiocarbon age Published 25 September 2019
calibration Curves 0–50,000 years cal BP. Radiocarbon 55, 1869–1887 (2013). 10.1126/sciadv.aax6225
52. C. Bronk Ramsey, Methods for summarizing radiocarbon datasets. Radiocarbon 59,
1809–1833 (2017). Citation: C. Leipe, T. Long, E. A. Sergusheva, M. Wagner, P. E. Tarasov, Discontinuous spread of
53. J. Biraben, The rising numbers of humankind. Popul. Soc. 394, 1–4 (2003). millet agriculture in eastern Asia and prehistoric population dynamics. Sci. Adv. 5, eaax6225 (2019).

Leipe et al., Sci. Adv. 2019; 5 : eaax6225 25 September 2019 9 of 9

Discontinuous spread of millet agriculture in eastern Asia and prehistoric population
dynamics
C. Leipe, T. Long, E. A. Sergusheva, M. Wagner and P. E. Tarasov

Sci Adv 5 (9), eaax6225.

DOI: 10.1126/sciadv.aax6225

Downloaded from http://advances.sciencemag.org/ on September 28, 2019

ARTICLE TOOLS http://advances.sciencemag.org/content/5/9/eaax6225

SUPPLEMENTARY http://advances.sciencemag.org/content/suppl/2019/09/23/5.9.eaax6225.DC1
MATERIALS

REFERENCES This article cites 40 articles, 7 of which you can access for free
http://advances.sciencemag.org/content/5/9/eaax6225#BIBL

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Use of this article is subject to the Terms of Service

Science Advances (ISSN 2375-2548) is published by the American Association for the Advancement of Science, 1200 New
York Avenue NW, Washington, DC 20005. 2017 © The Authors, some rights reserved; exclusive licensee American
Association for the Advancement of Science. No claim to original U.S. Government Works. The title Science Advances is a
registered trademark of AAAS.