http://orcid.org/0000-0001-5826-1260
Figures
Abstract
Based on chronological and archaeobotanical studies of 15 Neolithic and Bronze Age sites from the northern Chinese Loess Plateau and southern Inner Mongolia—the agro-pastoral zone of China–we document changes in the agricultural system over time. The results show that wheat and rice were not the major crops of the ancient agricultural systems in these areas, since their remains are rarely recovered, and that millet cultivation was dominant. Millet agriculture increased substantially from 3000 BC–2000 BC, and foxtail millet evidently comprised a high proportion of the cultivated crop plants during this period. In addition, as the human population increased from the Yangshao to the Longshan periods, the length and width of common millet seeds increased by 20–30%. This demonstrates the co-evolution of both plants and the human population in the region. Overall, our results reveal a complex agricultural-gardening system based on the cultivation of common millet, foxtail millet, soybeans and fruit trees, indicating a high food diversity and selectivity of the human population.
Citation: Bao Y, Zhou X, Liu H, Hu S, Zhao K, Atahan P, et al. (2018) Evolution of prehistoric dryland agriculture in the arid and semi-arid transition zone in northern China. PLoS ONE 13(8): e0198750. https://doi.org/10.1371/journal.pone.0198750
Editor: John P. Hart, New York State Museum, UNITED STATES
Received: November 23, 2017; Accepted: May 24, 2018; Published: August 3, 2018
Copyright: © 2018 Bao et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: We thank the NSFC (grants 41371003, 41772371) and the National Key Basic Research Fund (grant 2015CB953800), and the Youth Innovation Promotion Association for their financial support. This paper is also supported by the Strategic Priority Research Program of Chinese Academy of Sciences, Grant No. XDB 26000000. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The development of agriculture is fundamental to the large-scale transition of human societies from a hunting and gathering to a settled lifestyle, in both the Old and New World [1–4]. Settled communities permitted humans to observe and experiment with plants and to understand how they grew, which greatly reduced food shortages and directly resulted in both population increases [5–8] and the co-evolution of both human society and crop production [9]. In recent years, the increasing amount of research on prehistoric agriculture has significantly improved our understanding of the origins of agriculture in the Old World [7–18]. Specifically, ancient agriculture has emerged as a major area of interest in Asian archaeology [19]. Chinese agricultural traditions have two independent areas of origin: rice agriculture in the lower Changjiang River Basin, and dryland agriculture in the Yellow River Basin [20–23]. Dryland agriculture was dominated by two principal crops, foxtail millet (Setaria italica) and common millet (Panicum miliaceum), which were both domesticated probably as early as 8000 BC in North China [7,20,21]. The spread of these two crop plants into Western Asia and eventually Europe occurred mainly from 2000 BC-1000 BC, when wheat was widely planted in northern China (no later than 2100 BC); in addition, common millet spread to Central Asia along the Silk Road at about the same time [24]. The cultivation of both plants directly reflects increasing socioeconomic development and technology proficiency [7]. The main advantages of foxtail and common millet, which are C4 plants, is that they have high drought tolerance and can be grown in nutrient-poor soils. In most cases, rain-fed millet produces good harvests under conditions of low summer rainfall and it can grow in almost any soil type [7]. Nevertheless, foxtail and common millet cultivation first developed in the North China Plain, where rainfall was plentiful, during the early and middle Holocene. However, in the context of research into crop domestication and evolution [25–30], how millet agriculture first expanded from its center of origin to the adjacent dryland area and became adapted to an arid environment is poorly known.
The agro-pastoral zone in North China marks the border between traditional agricultural societies and nomadic societies, which has shifted backwards and forwards between agricultural and nomadic nations during the last several thousand years. It is a key area for studying the early civilization of ancient China. In the middle Neolithic, farmers of the Yangshao culture moved from the humid and sub-humid monsoon area of central China to occupy the agro-pastoral zone, and the local agricultural society became extremely prosperous during the late Neolithic and early Bronze Age transition [31–33]. In addition to possible climatic factors, the prosperity of the ancient agricultural societies of the arid agro-pastoral zone may have been due to innovations in the local agricultural system. These innovations probably facilitated the expansion of millet cultivars across the steppe and deserts of Eurasia, which eventually provided the agricultural basis of the nomadic pastoralists from Asia to Eurasia, and resulted in their becoming global crops by around 2000 BC [7,24,34]. However, the Neolithic and Bronze age agricultural systems in the agro-pastoral zone of China are still poorly documented.
Here, we present the results of geochronological and archaeobotanical studies of 15 Neolithic and Bronze Age sites from the northern Chinese Loess Plateau (CLP) and southern Inner Mongolia (Fig 1). Our principal aim was to shed light on the pattern of agricultural change in the drylands of the East Asian temperate zone, using quantitative analysis of seeds and a study of seed morphology of flotation samples from Neolithic and Bronze Age sites in the region. In addition, we integrate our findings with published archaeobotanical data from the agro-pastoral zone to the east to provide a comprehensive picture of the process of agricultural development from 4000 BC-1000 BC in the agro-pastoral zone of China.
1. Shihushan, 2. Xiaojiamao, 3. Xiyuan, 4. Zhaimao, 5. Wangmushan (overlapping with Shihushan), 6. Huoshiliang, 7. Ashan, 8. Dakou, 9. Zhukaigou, 10. Xinhua, 11. Laohushan, 12. Erliban, 13. Shimao, 14. Lijiaya, 15. Xuejiaqu, 16. Xiajiadian, 17. Chengzishan. The precipitation contours are cited from China Meteorological Data Service Center, using the data for 2017 (http://data.cma.cn/site/index.html).
Background setting and study sites
The central agro-pastoral zone of China mainly comprises the southern part of Inner Mongolia and the northern part of the CLP and is an important pathway between northeast and northwest China. The central agro-pastoral zone has a continental monsoon climate, with mean annual temperature of 7 to 12 °C and mean annual precipitation from 300 to 450 mm. The modern vegetation is mainly Artemisia-Bothriochloa grassland and shrubland, with Zizyphus, Vitex, Rosa and Hippophae, and woodland consisting of Platycladus and Pinus [35].
The regional rainfall is relatively low and was sensitive to past climatic oscillations [36–38]. Paleoclimatic studies indicate a higher precipitation and high lake levels in the area during the Neolithic and Bronze Ages, caused by strong summer monsoon circulation during the early and middle Holocene [39]. However, the vegetation distribution during the middle Holocene was still dominated by shrub-grassland and steppe, with sparse woodland near riverbeds, which may reflect the high temperatures and high evaporation [39].
In recent decades, thousands of sites belonging to a series of Neolithic and Bronze Age cultures dated to 6000 BC-1500 BC have been discovered in the region. In sequence, the cultures are as follows: The first period of Hougang culture, Shihushan type; Yangshao culture, Lower Wangmushan type; Haishengbulang culture; Xinhua culture; Ashan culture; Laohushan culture; Dakou culture; Zhaimao culture; Zhukaigou culture; Shimao culture; Guifang culture (Table 1). Most of the sites belong to the Late Longshan or Zhukaigou cultures (2500 BC-1800 BC) [40,41]. Shimao (2036 BC-1519 BC) is the best known and most extensive site in the region, with an area of over 4.25 km2. It is surrounded by three stone walls that were built for defense against invasion [39] and is the largest prehistoric walled city found in China. In terms of social structure and residential characteristics, Shimao consisted of several settlement clusters; by 2000 BC it was highly organized with a well-developed arid-land agricultural-pastoral economy [40]. Other sites, such as Zhukaigou (2147–1190 BC), Dakou (2342–1903 BC), Xuejiaqu (1449–1031 BC), Ashan (2582–2121 BC) and Zhaimao (2135–1939 BC), were subsequently discovered[42–44]. Agricultural development and technological exchange within the northern steppe region are reflected by the consistency of the composition and abundance of agricultural remains at these sites, including the carbonized remains of crop plants and grains, and farm implements [45–47].
The locations of the 15 study sites (109°20′~112°43′E, 37°5′~40°35′N), together with Chengzishan (1783–1504 BC) and Xiajiadian (816–405 BC) [48], in the western Liaoning area, are shown in Fig 1 and listed in Table 1. Most of the sites are in riparian locations in the steppe and shrub grassland zone in southern Inner Mongolia and northern Shaanxi Province.
Methods
From among the 15 sites in the Chinese Loess Plateau area, 52 soil samples were collected from sections containing cultural horizons that were deposited within stratigraphic layers or from ancient ash pits which are filled with gray soil and contained numerous plant remains and pottery fragments [49–52]. The volume of each sample was 40 L in the cultural layers and 20 L in the ash pits. The cell number per square inch of flotation sieve was 50 (0.3 mm mesh). Analysis of the coarse fraction revealed only coarse rock fragments, animal bones and occasional pottery shards. The floated samples were separated and identified under a stereoscopic microscope in the laboratory of the Institute of Vertebrate Paleontology and Paleoanthropology, CAS, Beijing (Fig 2). In addition, we conducted morphological studies based on measurements of the length and width of the seeds under the microscope (Table 2). A total of 268 foxtail millet seeds and 123 common millet seeds from six sites (Xiaojiamao (3500–3000 BC), Xinhua (2890–2278 BC), Dakou, Zhaimao, Zhukaigou, Shimao) were selected for the morphological studies. At least five individual seeds of each millet type were measured from each individual site.
All results are reported with 2 sigma error. 14C ages are calibrated to calendar years using Calib Rev 5.1 and OxCal v4.3.6, with the r:5 IntCal 13 atmospheric curve.
Samples from the 15 sites were selected for AMS 14C dating (Table 1). The samples included 3 animal bones from Shihushan (4696 BC-4441 BC), 1 foxtail millet seed from Xiyuan (3028 BC-2858 BC), 4 animal bones from Ashan, 3 animal and human bones from Dakou, 1 common millet seed from Zhaimao, 4 animal bones from Huoshiliang (2878–1768 BC), 7 samples of charcoal, foxtail millet and human bones from Zhukaigou, 4 human bones and 1 common millet seed from Shimao, and 1 foxtail millet seed from Lijiaya (1233 BC-1011 BC) (Table 1).
The radiocarbon ages were measured by accelerator mass spectrometry in the laboratory of the Australian Nuclear Science and Technology Organization (ANSTO) in Sydney, Australia, and by Beta Analytic in Miami, USA. The samples were pretreated by washing in acid/alkali solutions and rinsing to neutral pH.
Results
The radiocarbon dating results show that the oldest site is Shihushan and the youngest is Lijiaya. Only three sites (Shihushan, Xiaojiamao, Wangmushan) are of Yangshao age, which is usually dated to 5000 BC-3000 BC. Most of the AMS 14C ages are from 2500 BC-1500 BC, corresponding to a prosperous period of dryland agricultural societies in the north Chinese Loess Plateau. Xuejiaqu and Lijiaya are dated to 1500 BC-1000 BC, which is contemporaneous with the Shang Dynasty.
From the 52 flotation samples taken from the 15 study sites, we selected and identified 3110 charred grains, belonging to 14 genera or species, (Table 3). By comparison with sites from traditional farming areas, seeds are scarce in the cultural layers of sites in the steppe region. No seeds or fruit stones were recovered from Wangmushan and Xiyuan, and no crops seeds were recovered from Shihushan, Erliban and Lijiaya.
The 3110 cultivated plant seeds consist primarily of three species: foxtail millet (Setaria italica), common millet (Panicum miliaceum) and soybean (Glycine soja). Foxtail millet and common millet account for the highest proportion, with foxtail millet comprising 60.39% and common millet 32.06% of the total (Table 2). The percentage of foxtail millet seeds increased from 30.79% in the Yangshao Period (Xiaojiamao) to 50–83.79% in the late Longshan Period (Dakou, Shimao, Zhaimao), indicating a trend of increasing importance of millet cultivation in the agricultural system.
Several seeds of Glycine soja were found at Shimao, Zhukaigou and Xinhua. In addition, several seeds of typical weed species of Fabaceae, including Melilotus, Lespedeza and Medicago, occurred sporadically in the samples. A high proportion of Chenopodium was present at the sites; however, whether Chenopodium grew as a weed or was derived from human or animal excreta could not be determined. In addition, seeds, several nut shells and drupe pits, including Castanea, Prunus persica, Corylus, Prunus armeniaca and Pinus, were found in the samples. Prunus armeniaca occurred at the highest frequency and was found at nine sites (Figs 3 and 4).
a. Setaria italica, Zhukaigou; b Panicum miliaceum, Zhukaigou; c. Glycine soja, Zhukaigou; d. Hibiscus trionum, Erliban; e. Chenopodium, Xinhua; f. Prunus armeniaca, Shimao; g. Prunus persica, Zhukaigou. a and b are cereal crops, c-e are the seeds of wild plants, and f and g are probably from cultivated trees.
These suspected fruit shells have radians and ornamentation on the surface. The shell remains were too fragile to be identified accurately; however, we could identify three relatively intact specimens: a. Corylus, Huoshiliang; b. Pinus, Huoshiliang; c. Castanea, Zhaimao.
Statistical analysis of the length and width of the foxtail and common millet seeds from six sites revealed an increasing trend from the early to the late period (Fig 5). From Xiaojiamao to Zhukaigou, the length and width of common millet seeds increased significantly, by about 20% and 35%, respectively. However, the increase was less significant in the case of foxtail millet.
Discussion
Foxtail millet and common millet were the principal crops in the early stage of the development of dryland agriculture in the area; they were first domesticated in the middle and lower reaches of the Yellow River, within the East Asian monsoon region, at about 6000 BC [38]. The average annual rainfall of the area is estimated to have been as high as 800–1000 mm in the early and middle Holocene [53]. During subsequent millennia, millet agriculture expanded stepwise to the marginal regions of the monsoon area where rainfall is lower and the climate less stable. In addition, rice was introduced from the lower Changjiang River basin, and was cultivated in the southern and western parts of the Loess Plateau, adjacent to the present study area, before 3000 BC [54]. Wheat was introduced to China at around 2000 BC from Western and Central Asia and during the next several hundred years its cultivation resulted in a rapid agricultural transformation in the Hexi Corridor [55–57]. At the same time, millet agriculture continued to flourish in the arid and semi-arid zone, near the Hexi Corridor [7,37]. Thus, there is a consistent picture of a broadening agricultural system, with millet, rice, wheat and barley, in the sites of late Neolithic to Bronze Age in central China [54,57].
Our results reveal that the agricultural system was continuously dominated by foxtail and common millet cultivation from the early Yangshao period (5000 BC-3000 BC) to the Bronze Age, and no wheat or rice grains were found at any of the 15 sites. This suggests that the inhabitants developed a cereal-based agricultural system that was adapted to a relatively arid climate. This inference agrees with conclusions based on δ13C analyses of human bones from Zhukaigou, Dakou and Xinhua, which show that C4-type plants were the main food source within the region [58]. This somewhat narrowly focused system in the agro-pastoral zone during the Neolithic and Bronze Age likely was mainly due to the aridity and soil infertility of the region [54,56,57]. The modern mean annual rainfall of the study region is only 300–500 mm, and thus it belongs to the arid and semi-arid climate zone (Fig 1). Therefore, drought-tolerant crops, foxtail millet and common millet, which are C4 plants with a short maturation time, were more suitable for cultivation than C3 crops such as wheat and rice, which need more water for yield maintenance [59–61]. In fact, highly drought-tolerant foxtail and common millet were probably the main crops for the last several thousand years until large-scale cultivation of high-yielding maize was adopted in the agro-pastoral zone in 1960s.
In addition to foxtail and common millet, both wild and cultivated soybean were found at three sites (Xinhua, Zhukaigou, Shimao) (Table 2); however, their size is small compared to modern cultivated varieties [62]. The finding of soybean in the arid agro-pastoral zone, outside the humid monsoon area which is the natural habitat of wild soybean, indicates that the species was domesticated in China since the Yangshao period (Fig 3). Chenopodium seeds were found at five sites, and were well represented at Xinhua and Erliban, which may indicate that several species of edible wild Chenopodium were used for food and forage in the Neolithic and early Bronze Age in the area. Several drupe pits and nut shell fragments were also found at the sites, including Prunus armeniaca, Prunus persica, Castanea and Corylus (Fig 2, Table 2). Among them, apricot was present in the highest proportion, showing that fruit trees occupied an important place in the agricultural system. Like soybean, the natural distribution of these fruit trees is within the humid and semi-humid zones of the monsoon region. Given the evidence for possible forest management and horticultural technology of late Neolithic peoples reported in several studies [54,63], the presence of the various nut shells and drupe pits is strong evidence for early fruit-tree cultivation and management during the Neolithic to Bronze Ages in the agro-pastoral zone of China.
Our excavation record shows that the studied sites date mainly from the Longshan period to the Western Zhou Dynasty (Fig 2). For example, Shimao is one of the best-known sites and the largest town sites discovered to date in the prehistoric cultural region in the northern China grassland zone[64–66]. The evident prosperity of Shimao during this period shows that settlement of the northern grassland region was developing rapidly and that the population of the sites was increasing substantially. However, as was likely the case for Shimao city, the communities occupying these sites must have experienced significant pressure from the rapid growth in population—with implications for the viability of the overall economy, provision of habitations, and agricultural productivity. Thus, the early agricultural practices at these sites likely reflect an adaptation to population growth (Fig 6). The first adaptation reflects a change in the structure of dryland agriculture, with an emphasis on the cultivation of foxtail millet and common millet.
The four taxa are crops and plants characteristic of rain-fed agriculture. From bottom to top, the vegetation zones range from shrub grassland in the northwest, to steppe in the southeast; they represent different monsoon regimes within the study area (see text for explanation).
The relative proportion of foxtail and common millet seeds from the sites varied significantly over time and space. In the early stage, common millet comprised only 30% of the total; subsequently, during 5000 BC-1000 BC, the proportion of foxtail millet increased substantially. Moreover, there were also significant spatial differences (Fig 6). A higher proportion of foxtail millet generally occurred in the southern shrub-grassland area, while there was a higher proportion of common millet in the northern steppe area.
Environmental differences were the likely cause of the differences between the principal crop plants in the northern and southern parts of the agro-pastoral zone. Common millet is easy to cultivate and it appears to have been better suited to primitive agricultural practices than other crops. Under favorable environmental conditions, however, it may be a poor choice because of its limited response to high soil fertility and abundant water availability [59]. In contrast, foxtail millet has a high yield, typically twice that of common millet, and is better adapted to low temperatures. The germination temperature of foxtail millet is 15–25°C, and the jointing temperature is about 22-25°C. By comparison, common millet needs temperatures of 15–20°C to germinate and 25–30°C to joint [40,60,61]. However, foxtail millet requires higher soil fertility and more abundant water than common millet, as well as a longer growing period [40].
The study area experienced an interval of cold and humid climate during 3000 BC-1000 BC [39] and we divided it into three regions reflecting the trajectory of the summer monsoon (Fig 6). The southern part was occupied by shrub grassland with higher precipitation and soil land fertility, while the northern steppe land was drier and less fertile. Thus, foxtail millet must have been selected and cultivated on an increasing scale through the Neolithic and Bronze Ages in the agro-pastoral zone. However, common millet, which is more tolerant of drought and low soil fertility, continued to be important in the northern steppe area. This may indicate that the early farmers were able to adapt continuously to the environment as the human population increased [39].
The second category of evidence for agricultural adaptation to increasing human population is crops yield. There were two aspects to changing crop yield: changes in grain morphology and biomorphic selection [61]. The morphological data show that the size of common millet seeds increased during 3500 BC-2000 BC. The length and width of the seeds of both common millet and foxtail millet were similar at Xiaojiamao site, 3000 BC (Fig 3). However, from 3000–1800 BC, conscious selection probably resulted in the observed increase in the size of common millet seeds which may also have been accompanied by increasing yields [67,68]. From Xiaojiamao to Zhukaigou, the length of common millet seeds increased by ~20%, and the width by ~35%, on average. It can be estimated that both average seed volume and the one-thousand grain weight of common millet in the Longshan period were twice that in the Yangshao period. Together with a possible increase in the average number of grains per common millet plant, we speculate that the average annual common millet yield per unit area may have doubled from 3000 BC-1800 BC in the agro-pastoral zone of China.
It is noteworthy that the percentage representation of foxtail millet increases sharply (Table 3); however, the seed size of foxtail millet is inversely proportional to its representation among the total number of seeds. For example, at Zhaimao (Table 1), very small seeds comprise 86% of the total of foxtail millet seeds, which is the second highest value at all the sites. In contrast to common millet, rice, and sorghum, foxtail millet cultivars can produce 1–3 spike ears, so that the number of grains per plant has a wide range of variability [40]. Therefore, there is a significant possibility that although the yield of foxtail millet may also have increased within a similar range to common millet, the increase in the number of seeds per plant may be the main reason for the increase. Moreover, this may be the cause of the increase in the proportion of foxtail millet within the total seed number of seeds.
Conclusions
Based on an analysis of fossil seeds at archaeological sites, we have studied changes in the proportions of crops cultivated in the arid and semi-arid transition zone of northern China during 5000 BC-1000 BC, spanning the Yangshao period to the Western Zhou dynasty. Due to the regional environment, climate and soil fertility, wheat and rice were not the main crops, and the typical rain-fed cereals, foxtail millet and common millet, occupied the dominant position in the agricultural system.
The proportion of foxtail millet increased substantially from the Yangshao to the Longshan period. In the southern shrub-grassland region of the agro-pastoral zone, foxtail millet, rather than common millet, became the primary food crop in the late Longshan period. In contrast, in the arid northern steppe region, common millet remained important in the Longshan and Bronze Age. Soybean played an important role in the dryland agricultural system, probably representing the earliest cultivated soybean species. Fruit trees, such as apricot, peach, and chestnut, may have been cultivated at the same time. These characteristics may reflect an adaption of the agricultural system to the increasing human population size during 3000 BC-1000 BC.
Analysis of the changes in seed morphology indicates an increase in the size of the seeds of common millet, especially, which may reflect an adaptation to the rapidly increasing human population from 3000 BC-2000 BC. These changes likely resulted in a significant improvement in the dryland agricultural productivity in the agro-pastoral zone. If we assume that the cultivated area was relatively constant, the average yield per unit area in the late Longshan may have been twice that of earlier periods. In addition, the increase in the productivity of millet cultivation may have promoted its spread from north China westwards to Eurasia with the result that it became a global crop by around 2000 BC.
Supporting information
S1 File. Lab calibration data (original).
This is the raw data we obtained from ANSTO and BETA laboratory, including a variety of measurement results. And in this paper, we only selected the 14C dating part.
https://doi.org/10.1371/journal.pone.0198750.s001
(XLSX)
S2 File. Lab calibration data (after arrangement).
https://doi.org/10.1371/journal.pone.0198750.s002
(XLSX)
S5 File. Seed identification under stereoscope.
https://doi.org/10.1371/journal.pone.0198750.s005
(RAR)
Acknowledgments
We thank the NSFC (grants 41371003, 41772371) and the National Key Basic Research Fund (grant 2015CB953800), and the Youth Innovation Promotion Association for their financial support. This paper is also supported by the Strategic Priority Research Program of Chinese Academy of Sciences, Grant No. XDB 26000000. We are grateful to the reviewers for their professional comments and useful suggestions. We thank Dr. Jan Bloemendal for improving the English. We are also grateful to Ji Ming, Gao Qiang, Du Xiaolong, Hu Songmei, Sun Zhouyong, Shaojing, Shen Hui, Wang Jian and Zhang Guilin for their assistance with the fieldwork.
References
- 1. Purugganan MD, Fuller DQ. The nature of selection during plant domestication. Nature. 2009; 457: 843–848. pmid:19212403
- 2. Hancock JF. Contributions of domesticated plant studies to our understanding of plant evolution. Annals of Botany. 2005; 96(6): 953–963. pmid:16159942
- 3.
Barker G. The Agricultural Revolution in Prehistory: Why Did Foragers Become Farmers? Oxford University Press. 2006.
- 4. Rindos D, Aschmann H, Bellwood P, Ceci L, Cohen MN and Hutchinson J et al. Symbiosis, instability, and the origins and spread of agriculture: a new model. Current Anthropology 1980; 21(6): 751–72.
- 5. Gignoux CR, Henn BM, Mountain JL. Rapid, global demographic expansions after the origins of agriculture. Proceedings of the National Academy of Sciences of the United States of America. 2011; 108: 6044–6049. pmid:21444824
- 6. Lev-Yadun S, Gopher A, Abbo S. The Cradle of Agriculture. Science. 2000; 288: 1602–1603. pmid:10858140
- 7. Miller NF, Spengler RN, Frachetti M. Millet cultivation across Eurasia: Origins, spread, and the influence of seasonal climate. The Holocene. 2016; 26: 1566–1575.
- 8. Fuller DQ. New archaeobotanical information on plant domestication from macro-remains: Tracking the evolution of domestication syndrome traits. Supervisor Nurse. 2012; 4(3–4): 110–135.
- 9. Fuller DQ, Denham T, Arroyokalin M et al. Convergent evolution and parallelism in plant domestication revealed by an expanding archaeological record. Proceedings of the National Academy of Sciences of the United States of America, 2014; 111(17): 6147–6152. pmid:24753577
- 10.
Rosen AM. Civilizing climate: social responses to climate change in the ancient Near East. Plymouth: Altamira Press. 2007.
- 11. Franchetti MD, Spengler RN, Fritz GJ, Mar’yashev AN. Earliest direct evidence for broomcorn millet and wheat in the central Eurasian steppe region. Antiquity. 2010; 84: 993–1010.
- 12.
Harlan JR. Origins of African Plant Domestication. World Anthropology. 1976.
- 13. Smith BD. Origins of Agriculture in Eastern North America. Science. 1989; 246: 1566–1571. pmid:17834420
- 14.
Zohary D, Hopf M, Weiss E. Domestication of Plants in the Old World: The origin and spread of domesticated plants in Southwest Asia, Europe, and the Mediterranean Basin. Oxford: Clarendon Press. 1988.
- 15. Crawford GW, Shen C. The origins of rice agriculture: recent progress in East Asia. Antiquity. 1998; 72: 858–866.
- 16. Crawford GW, Lee GA. Agricultural origins in the Korean Peninsula. Antiquity. 2003; 77: 87–95.
- 17.
Crawford GW. East Asian plant domestication. Archaeology of Asia. 2006.
- 18. Liu X, Lister DL, Zhao Z, Staff RA, Jones PJ, Zhou LP, et al. The virtues of small grain size: Potential pathways to a distinguishing feature of Asian wheats. Quaternary International. 2016; 426: 107–119.
- 19.
Liu L, Chen XC. The archaeology of China: From the late Paleolithic to the Early Bronze Age. Cambridge University Press. 2012.
- 20. Fuller DQ. Contrasting patterns in crop domestication and domestication rates: Recent archaeobotanical insights from the old world. Annals of Botany. 2007; 100: 903–924. pmid:17495986
- 21. Zhao ZJ. Paleoethnobotany and its new achievements in China. Archaeology. 2005; 7:42–49.
- 22. Fuller DQ, Denham T, Arroyo-Kalin M, Lucas L, Stevens CJ, Qin L, et al. Convergent evolution and parallelism in plant domestication revealed by an expanding archaeological record. Proceedings of the National Academy of Sciences of the United States of America. 2014; 111(17): 6147–6152. pmid:24753577
- 23. Purugganan MD, Michael D, Fuller DQ. The nature of selection during plant domestication. Nature. 2009; 457: 843–851 pmid:19212403
- 24. Spengler R, Frachetti M, Doumani P, Rouse L, Cerasetti B, Bullion E, et al. Early agriculture and crop transmission among Bronze Age mobile pastoralists of Central Eurasia[J]. Proceedings of the Royal Society of London B: Biological Sciences, 2014, 281(1783): 20133382.
- 25. Smith BD. Documenting plant domestication: The consilience of biological and archaeological approaches. Proceedings of the National Academy of Sciences of the United States of America. 2001; 98(4): 1324–1326. pmid:11171946
- 26. Ibarra JR, Morrell PL, Gaut BS. Plant domestication, a unique opportunity to identify the genetic basis of adaptation. Proceedings of the National Academy of Sciences of the United States of America. 2007; 104: 8641–8648. pmid:17494757
- 27. Diamond J. Evolution, consequences and future of plant and animal domestication. Nature. 2002; 418: 700–707 pmid:12167878
- 28. Lu HY, Zhang JP, Liu KB, Wu NQ, Li YM, Zhou KS, et al. Earliest domestication of common millet (Panicum miliaceum) in East Asia extended to 10,000 years ago. Proceedings of the National Academy of Sciences of the United States of America. 2009; 106: 7367–7372.
- 29. Yang X, Wan Z, Perry L. Early millet use in northern China. Proceedings of the National Academy of Sciences of the United States of America. 2012; 109: 3726–3730. pmid:22355109
- 30. Zhao ZJ. New data and new issues for the study of origin of rice agriculture in China. Archaeological and Anthropological Sciences. 2010; 2: 99–105.
- 31. Han M. The forming and climate changing of agro-pastoral Zone, Northern China. Archaeology. 2005; 10: 57–67.
- 32. Dong GH, Zhang SJ, Yang YS, Chen JH, Chen FH. Agricultural intensification and its impact on environment during Neolithic Age in northern China. Chin Sci Bull. 2016; 61: 2913–2925.
- 33. Yan HD. Research on the Potteries Unearthed from Shimao site, Shenmu county. Relics and Museolgy. 2010; 6: 3–8.
- 34. Hunt HV, Linden MV, Liu XY, Matuzeviciute GM, Colledge S, Jones MK. Millets across Eurasia: chronology and context of early records of the genera Panicum and Setaria from archaeological sites in the Old World. Vegetation History and Archaeobotany. 2008; 17(1): 5.
- 35. Qian CS, Wu ZY, Chen CD. The vegetation type of China. Acta Geographica Sinica. 1956; 1: 70–80.
- 36. Feng ZD, An CB, Wang HB. Holocene climatic and environmental changes in the arid and semi-arid areas of China: a review. The Holocene. 2006; 16: 119–130.
- 37. Xiao J, Nakamura T, Lu H, Zhang G. Holocene climate changes over the desert/loess transition of North-Central China. Earth & Planetary Science Letters. 2002; 197, 11–18.
- 38. Xiao J, Xu Q, Nakamura T. Holocene vegetation variation in the Daihai Lake region of north-central China: a direct indication of the Asian monsoon climatic history. Quaternary Science Reviews. 2004; 23: 1669–1679.
- 39. Li XQ, Zhou J, Shen J, Weng CY, Zhao HL, Sun QL. Vegetation history and climatic variations during the last 14 ka BP inferred from a pollen record at Daihai Lake, north-central China. Review of Paleobotany and Palynology. 2004; 132: 195–205.
- 40.
Institute of Crops Genetic Resources, Chinese Academy of Agricultural Sciences, Shanxi Academy of Agricultural Sciences. Foxtail millet cultivars in China. Beijing, China Agriculture Press. 1985.11.
- 41.
China State Bureau of Cultural Relics. An atlas of Chinese cultural relics (Shaanxi fascicule). Beijing, China: Sino Maps Press. 1998.
- 42.
China State Bureau of Cultural Relics. An atlas of Chinese cultural relics (Inner Mongolia fascicule). Beijing, China: Sino Maps Press. 2003.
- 43. Shaanxi Provincial Institue of Archaelolgy. The Excavation of Zhaimao Site in Shenmu County. Archaeology and Cultural Relics. 2003; 00: 319–334.
- 44. Lu ZR. The Excavation of Zhaimao Site in Shenmu County. Archaeology and Cultural Relics.2002; 3: 3–17.
- 45. Wang LW. Analysis on the Remains in Zhukaigou Site. Northern Cultural Relics. 2004; 3: 17–27.
- 46. Ji FX, Ma XQ. Survey and Excavating of Dakou site in Jungar Banner, Inner Mongolia Province. Archaeology. 1979; 4: 310–318.
- 47. Choi X, Siqin , Liu HZ, Lin H. A brief report on the excavation of Ashan site in Baotou City, Inner Mongolia Province. Archaeology. 1984; 2: 99–108.
- 48. Zhao KL, Li XQ, Shang X, Zhou XY, Sun N. Agricultural Characteristics of Middle-late Bronze Age in Western Liaoning Province. Chinese Bulletin of Botany, 2009. 6: 718–724
- 49. Fu YX. A Brief View of the “Ash-Pit”. HuaXia Archaeology, 2014; 2: 126–132.
- 50. Jiang XC. The Conception of Ash-Pit and Rediscuss of the Fieldwork Method. Jianghan Archaeology, 2009; 112: 41–44.
- 51. Yan SD. Several Questions about the Stratigraphic Deposit and the Ash-Pit. Huaxia Archaeology, 2008; 1: 128–133
- 52. Tuo G, Wen D. The discussion about the Ash-Pit. Jianghan Archaeology, 1995; 3: 94–96
- 53. Huang CC, Pang J, Zha X, Su H, Jia Y, Zhu Y. Impact of monsoonal climatic change on Holocene overbank flooding along Sushui River, middle reach of the Yellow River, China. Quaternary Science Reviews. 2007; 26(17): 2247–2264.
- 54. Li XQ, Dodson J, Zhou XY, Zhang HB, Masutomoto R. Early cultivated wheat and broadening of agriculture in Neolithic China. The Holocene. 2007; 17(5): 555–560
- 55. Dodson J, Li XQ, Ji M, Zhao K, Zhou X, Levchenko V. Early Bronze in two Holocene archaeological sites in Gansu, NW China. Quaternary Research. 2009; 72(3):309–314.
- 56. Zhou XY, Li XQ, Dodson J, Keliang Z. Rapid agricultural transformation in the prehistoric Hexi corridor, China, Quaternary International. 2016; 426:33–41.
- 57. Zhou XY, Li XQ, Zhao KL, Dodson J, Sun N, Yang Q. Early agricultural development and environmental effects in the Neolithic Longdong basin (eastern Gansu). Chinese Science Bulletin. 2011; 56(4–5): 318–326.
- 58. Atahan P, Dodson J, Li XQ. Temporal trends in millet consumption in northern China. Journal of Archaeological Science. 2014; 50: 171–177.
- 59.
Baltensperger DD. Foxtail and Proso millet. 1996.
- 60. Guedes JD, Manning SW, Bocinsky RK. A 5500 year model of changing crop niches on the Tibetan Plateau. Current Anthropology. 2016; 57(4): 517–522.
- 61. Guedes JD, Butler EE. Rethinking the Spread of Agriculture to the Tibetan Plateau. The Holocene. 2015; 25(9): 1498–1510.
- 62. Lee GA, Crawford GW, Liu L, Sasaki Y, Chen X. Archaeological Soybean (Glycine max) in East Asia: Does size matter? Plos One. 2011; 6(11): 1–12
- 63. Kitagawa J, Yasuda Y. Development and distribution of Castanea and Aesculus culture during the Jomon period in Japan, Quaternary International. 2008; 184(1): 41–55.
- 64.
Liu HB. Early Agricultural Characteristic and its impact in Erdos Region during the Middle and Late Holocene. Graduate. 2012.
- 65. Xu F. A Preliminary Comparison of the Archaeological discovery in Shimao site and Taosi site. Relics and Museolgy. 2014; 1: 18–22.
- 66. Sun ZY, Shao J, Shao AD, Kang NW, Qu FM, Liu XM. Shimao Site in Shenmu county, Shaanxi Province. Archaeology. 2013; 7: 15–23.
- 67. Purugganan MD, Fuller DQ. 2009. The nature of selection during plant domestication. Nature 457:843–848. pmid:19212403
- 68. Weber S, Kashyap A, Harriman D. Does size matter: the role and significance of cereal grains in the Indus civilization. Archaeological and Anthropological Sciences. 2010; 2: 35–43.