An agricultural complement to ancient northern Central Asian pastoralism on the steppe: a regional analysis of ceramic impressions

Abstract

The adaptive economies that allowed ancient people to, not only survive on the arid steppe of northern Central Asia, but also form early polities, elaborate burial practices, and specialized craft production, have attracted the attention of archaeologists for more than a century. Traditionally, these scholars have emphasized the first millennium BC as a key period of cultural change, labeling it the period when highly specialized nomadic pastoralism develop in this part of the world. While specialized nomadism has recently been challenged, it remains difficult to identify further evidence of agriculture at open-air sites because of deflation. We bypass the limitations in the archaeobotanical data by studying grain impressions on ceramic sherds from 24 archaeological sites across southern Kazakhstan, spanning more than two millennia. We identified impressions of hulled barley, naked barley, both broomcorn and foxtail millet, and rice at sites across this steppe belt, allowing us to piece together a more nuanced view of the ancient economy.

1 Introduction

As the intersection or fusion zone between East and West Eurasian cultures, interest in the Bronze (ca. 2500–800BC) and Iron ages (800BC-AD200) of northern Central Asia is growing. In particular, archaeobotanical and isotopic approaches are shedding light on the timing and routes of dispersal for crops, such as barley and wheat, originally cultivated in southwest Asia, and broomcorn millet, foxtail millet, and rice first brought under cultivation in East Asia. This research includes macrobotanical studies (Frachetti et al., 2010; Motuzaite Matuzeviciute et al., 2020; Spengler et al., 2014), stable isotope analysis of human and animal bones (Hermes et al., 2019; Itahashi et al., 2020), miliacin biomarkers as an indicator of broomcorn millet (Murakami et al., 2022), and ancient DNA analysis (Kumagai and Sakai, 2024; Silva et al., 2018). These studies have demonstrated the presence of farming economies across the mountain foothills in the second millennium BC (Spengler et al., 2016), with possible intensification in the cultivation system leading to the establishment of villages across much of the montane zone in the first millennium BC (Spengler et al., 2017). These revelations are not new, Akishev (1970; see also Baipakov, 2008) argued that first millennium BC sites were farming villages in the Tien Shan Mountains and Di Cosmo (1994) argued for ancient agriculture in Xinjiang and neighboring zones based on finds of stone tools, notably grinding stones and querns. Despite the immense wave of data – new and old – from the mountainous spine and southern oasis regions, Central Asian archaeologists have been slow to step past the nomadic narratives and to accept that various forms of agropastoralism were practiced throughout the second and first millennia BC. The role of farming among ancient people in the steppe remains particularly in question given that archaeobotanical data are still almost completely lacking from the wide expanse (ca. 3,500 km latitudinally) of this ecozone. This vast chasm in the data is also problematic when discussing prevailing ancient routes of connectivity, depicted in either the Steppe Highway Model (Christian, 2000) or the prehistoric Silk Road Model (Frachetti, 2012; Kuzmina, 2008).

The longstanding view of a sharp dichotomy in the ancient economy of Central Asians has been characterized by specialized mobile pastoralism in the northern steppe and agriculture in the southern regions (Beckwith, 2023; Frachetti, 2012). However, recent research has challenged the strict polarized division in the economy, in many ways resulting in a merging of these two modes into a hybrid system, sometimes referred to as agropastoralism (Spengler et al., 2021a). It is well-recognized now that people in northern Central Asia, as far back as the third millennium BC, were engaged in farming (Spengler, 2015; Zhou et al., 2020), and that there were pastoralists in the southern regions even earlier (Arbuckle and Hammer, 2019; Taylor et al., 2021). Nonetheless, it is also evident that there were cultural divisions between the north and south, which persisted through the second and first millennia BC (Di Cosmo, 2015). Broadly speaking, these divisions include more intensive agriculture and village or proto-urban settlements, which were largely restricted to the montane foothills from Ferghana southward prior to the turn of the common era. Rouse and Cerasetti (2014:34) have pointed out that:

“Eurasian steppe pastoralists and the farmers of southern Central Asia were physically separated by the Karakum and Kyzylkum deserts. However, these deserts, which seemingly kept subsistence strategies from overlapping for millennia, were never cultural barriers, nor devoid of human occupation. Mobile pastoralism in the desert was a viable subsistence strategy, and seasonal movements would have encouraged the sharing of ideas and goods throughout the Bronze Age”.

A loose tripartite division would be more effective than a bifurcate cultural division, but such an approach would still obscure the fact that these economies all fell along a spectrum and readily changed their adaptive strategies through time. We seek to highlight the importance of the intermediate zone, whereas farming peoples occupied a string of arable pockets stretching along the Syrdarya into the mountain foothills and Ferghana. Archaeologists working in Ferghana through the mid-twentieth century have recognized the presence of a late second millennium BC farming people, which they labeled the Chust Culture. Archaeologists have further acknowledged similarities between these peoples and those along the southern steppe peripheries, as well as eastward into Xinjiang (see An, 1992; Chen and Hiebert, 1995). It is also well accepted that people throughout the oases of Xinjiang and the Tien Shan foothills were engaged in a mixed farming and herding system since the second millennium BC (Li et al., 2024). The cultural dynamics of these southern steppe populations remain largely unexplored, and with the exception of a few isotope analyses of human bone collagen (Itahashi et al., 2020; Kurmaniyazov et al., 2023), there is still very little data available for reconstructing the diet of people within this vast ecoregion. The dearth of data is mainly due to the difficulty in accessing archaeological sites so far away from any large cities and because wind deflation has destroyed the stratigraphic integrity of most open-air sites (Oralbay et al., 2016). In lieu of preserved archaeobotanical remains, we present data about plant cultivation in this region, both temporally and spatially, by identifying seed impressions left on ceramic sherds.

2 Archaeological sites under study

2.1 A culturally diverse realm

We analyzed ceramic sherds from 24 archaeological sites spanning southern Kazakhstan, associated with: 1) the Andronovo Culture (second millennium BC) (Figure 1 and Table 1); 2) the onset of the Iron Age (traditionally thought of as the era of early nomads; the Saka, ninth century BC – fifth century AD); and 3) the Kaunchin Culture (first century BC – fourth century AD). All of the peoples studied here developed an economy well-adapted to an arid continental climate, with annual precipitation in the Turkestan region is roughly 200 mm, whereas in the Karatau Mountains it reaches 200–400 mm, and along the Syrdarya it ranges between 100 and 350 mm (Mustafayev et al., 2023); any large-scale cultivation system across much of this region would have depended on irrigation water from rivers fed by mountain meltwater. This region is characterized by a mosaic of grasslands, hilly terrain, and shallow incised valleys; therefore, in specific ecological pockets, low-investment plots of millet and possibly hulled barley could have been cultivated as a complement to herding (see discussion of ethnographic evidence in Spengler, 2015). Hence, the presence of millets and barley grains in sites across this region could either represent low-investment complements to a pastoral-dominant system or more intensive investment in collective irrigation projects. While many models for ancient cultural adaptation, which mingle livestock farming and agriculture, has been drawn from ethnographic analogies and promoted for this region by certain scholars (Akishev, 1970; Baipakov, 2008; Chang, 2018; Di Cosmo, 1994; Spengler et al., 2017), reliable archaeological evidence regarding the actual character of farming has remained scarce from the heart of the steppe.

Figure 1

Most paleoclimate models suggest that the modern climate is loosely similar to that of late prehistory, given that the steppe tends to buffer the effects of climate change (Spate et al., 2022). In southern Kazakhstan, particularly in the Turkestan area and the foothills of the Karatau Mountains, there is evidence of ancient irrigation systems. However, these features have not been sufficiently investigated, and at present we cannot conclusively confirm or reject their existence and extent. On the other hand, although no definitive ancient irrigation structures have been discovered in this region, analysis of hydrological and climatic conditions with AMS dating at Otrar indicates that there would have been “chronic environmental challenges to floodwater farming over the last two millennia, punctuated by multicentennial-length periods with favorable hydromorphic and hydroclimatological conditions that enabled irrigation agriculturists to flourish” (Toonen et al., 2020). Currently, wheat, barley, sunflowers, cotton, rice, vegetables, melons, flax, corn, and potatoes are all grown in this region under irrigation (Pachikin et al., 2022). Nearly all of this cultivation occurs within a short distance from the course of the Syrdarya, leaving much of the region as uncultivatable arid steppe.

2.2 Andronovo-type sites

While aspects of both small-scale sedentary and nomadic lifestyles have variably been linked to the late third and second millennia BC peoples of this region, archaeologists generally agree that these Andronovo-type populations were characterized by animal husbandry, bronze production, and small-scale settlements (homesteads) or seasonal encampments (Frachetti, 2012). The existence of agriculture along this southern swath is largely assumed based on its presence in the Aral Sea region to the west (Brite, 2018; Endo et al., 2025) as well as the mountainous region to the east (Endo et al., 2023; Spengler et al., 2017). It is well recognized by all archaeologists working in this region that sites are concentrated along microenvironmental pockets and rivers fed by montane meltwaters in these arid steppe and foothill ecozones, reflecting localized adaptation to arid conditions (Frachetti et al., 2017; Spengler et al., 2013). Major Bronze Age sites in this ecotonal belt include the burial complexes at Baganaly, Sherbay, Zhasazhyryk, and Karasuiyr (Eleuov et al., 2005; Murgabayev et al., 2016; Murgabayev et al., 2018; Murgabayev et al., 2020).

2.3 Saka-type sites

Traditional interpretations have regarded the Scythians, Saka, and Wusun peoples of the ninth to second centuries BC as nomads (Beckwith, 2009; Grousset, 1970), but we interpret the growing data as suggesting that agriculture gained greater prominence in Central Asia during this period. The broadly applied name of Scythian is used in popular discourse to debate the wide-spread similarities in material culture - notably artifacts - represented in this time period across the steppe (Demaitre and Demaitre, 1981). While the term, Scythian, is largely a misappropriation of Classical texts, which use the term to refer to people near the Caspian littoral, Saka and Wusun are taken from an early Chinese text from the Han dynasty. Often scholars have posited that climate change further promoted a transition to a nomadic or mobile pastoralist life. Nonetheless, many seemingly sedentary or semisedentary villages have been explored from this time period (e.g., Chang et al., 2002). The Syrdarya River basin is surrounded by arid steppe, and it is generally assumed that any ancient agriculture in this region would have depended on rivers; hence, the settled areas are almost always restricted to riparian zones (Chang et al., 2002; Taleev et al., 2019).

2.4 Kaunchin-type sites

The burial sites of Myntobe, Karatobe, Shaga, and the settlement of Kultobe are located in the middle reaches of the Syrdarya and its tributaries. While three of our four sites ascribed to this cultural complex are funerary, these mortuary complexes were rich in ceramics, which appear to reflect forms used for domestic purposes, as opposed to the creation of unique forms made specifically for burial offerings. These sites have been associated with the population of the Kaunchin Culture and the early Kangly (Kangyu) state. This cultural complex covered the middle reaches of the Syrdarya, including the valleys of the Kurkeles, Keles, Shyrshyk (Chirchik), and Akhangaron (Angren/Akhangaran) rivers, and included ancient groups of people inhabiting the Tashkent Oasis (Sizdikov et al., 2025). The Kaunchin Culture (100 BC-AD 400) existed in the lower reaches of the Syrdarya, forming an independent trajectory of development within the unified historical and cultural space of the region. The Kaunchin Culture demonstrated a common regional affiliation with people of the Middle Syrdarya zone, and these similarities are often used to argue for interactions between sedentary and steppe populations in the early historical period (Levina, 1971). These ancient people were singled out by archaeologists decades ago as an exception to the nomadic narratives for the region. For example, Levina (1971) theorized that their economy was characterized by a stable sedentary agricultural tradition and a developed urban structure. Kuzmina (2008) and Smagulov and Baratov (2017) present a model whereas these archaeological sites reflect early adaptations to arid environments, foreshadowing what they see as a split between pastoralism – largely focused on the steppe – and urbanism along the rivers. In this way, they envision the people in this region bridging the expansive cultures that are often labeled as nomadic and agricultural.

The Myntobe and Kultobe sites are located at the intersection of the Keles and Syrdarya rivers (Turkestan region). Modern cultivation of rice and cotton in this interfluvial relies on irrigation from the Syrdarya. Radiocarbon analyses conducted at the TÜBİTAK Carbon Dating laboratory yielded the following results: for the Karatobe site, Kurgan 1 dated to 125–235 cal AD (95.4%), Kurgan 3 dated to 153–253 cal AD (81.9%); and for a Kurgan from the Myntobe site, which is where one of the two sets of rice impressions was detected (discussed below), teeth taken from a human skeleton dated to 126–248 cal AD (94.4%) (Gursoy and Sizdikov, 2024). Additionally, charcoal fragments collected from the second construction horizon of the Kultobe settlement, where the other rice impressions were identified, was dated to 354–400 cal AD (73.5%) (Sizdikov et al., 2025).

3 Method

3.1 Sampling protocol for ceramic impression analysis

The procedure for collecting replicas from ceramic impressions follows the steps outlined by Ushino and Tagawa (1991): (1) use a magnifying glass (10× magnification) to examine the inner and outer surfaces and cross-sections of the sherds to locate impressions left by seeds. Impressions are not always fully exposed; many are embedded within the ceramic body, requiring observation through small surface holes to see deeper inside. (3) Carefully remove sand and debris from the hole using a soft brush. (4) Coat the impression and its surrounding area with acetone containing 5% Paraloid B72 to protect the ceramic surface. (5) Fill the impression with silicone (Tokuyama Fit Tester, Tokuyama, Japan). (6) After hardening, carefully remove the cast from the cavity. (7) Remove the coating solution using 100% acetone and clean the surface. And, (8) observe and identify using SEM (in this case, KEYENCE VE-8800) with low acceleration voltage observation (0.5–20k).

3.2 Benefits of this method in Central Asia

This method is simple, but useful for collecting and analyzing plant data from a wide geographic area and over a long period of time, and allows us to examine what plants were used by analyzing ceramics that have already been excavated and stored or curated. The method also has the advantage of enabling the estimation of the age of seeds by borrowing the chronological position within the site stratigraphy of the specific ceramic sherds of interest, and, in turn, verifying the stratigraphic positioning by ceramic seriation, or by linking the style of the sherd to its time period of fabrication. Seeds readily move through the sediment column, but a ceramic type can be relatively diagnostic to a time period. Additionally, by using a microscope to observe the surface structure of impressions that were made by fresh and uncharred seeds, we can achieve a high degree of accuracy in seed identification. Often these impressions include the glumes, which rarely preserve in a carbonized state, and the presence of unprocessed seeds and chaff allows us to theorize about the role of grains, as we do in the discussion below, in ways that we could not have done with an archaeobotanical assemblage alone. In other words, by utilizing ceramic sherds already excavated and in storage, we can expect to gradually fill in prominent gaps in the archaeobotanical data. The gap that we seek to deal with in the current study covers the central part of the Eurasian supercontinent.

In addition, plants detected using the flotation method represent a palimpsest of activities spanning the entire duration of existence of structures. In contrast, plants identified from ceramic impressions only represent the short period when the clay used to make the ceramic was still damp. In other words, they can be considered snapshots taken within a very short period of time, allowing us a glimpse into a kitchen on a given day. In Kazakhstan, results using this method have been recently accumulating (Endo et al., 2023, 2025; Sasaki et al., 2022).

3.3 Seed identification criteria

One advantage of the method for seed identification using SEM observation of replicas collected from pottery impressions is the ability to observe even the fine detail of the surface of uncharred seeds. This is especially significant in cases where the husks or glumes were still articulated at the time the impression was made, traits that are rarely preserved in conventional archaeobotanical assemblages. As one example of the significance, this unique characteristic of the method illustrates how complicated it can be to differentiate between wild and domesticated forms of foxtail millet – possibly challenging the prevailing criteria used today (discussed below). All of the identification criteria we used match those utilized in studies of basic archaeobotanical approaches. In addition, for the convenience and reference of the readers, SEM images of all the impressions examined in this study are available at 10.5281/zenodo.21659719 as Supplementary Information.

3.3.1 Hordeum vulgare

The morphological criteria for differentiating barley grains from other species of Pooid grasses are well established and need not be rehashed here (Motuzaite Matuzeviciute et al., 2021). The distinction between hulled barley and naked barley was made based on the presence of a visible husk attached to the grain and a slightly angular cross-sectional shape (Figures 2a–c). Furthermore, the ventral furrow of hulled barley is wide, deep, and rigid. In contrast, grains that we identified as naked barley had no husk attached – readily falling off after harvesting – and the cross-sectional views are elliptical with smoothed edges (Figures 2d,e).

Figure 2

3.3.2 Panicum miliaceum

Caryopses of broomcorn millet lack papilla-like surface structures on both the outer and inner glumes, which possess a smooth surface. The overall shape is elliptical with a pointed apex. In lateral view, both the outer and inner glumes exhibit rounded bulges, with the outer glume or palea enveloping the inner glume (lemma) (Figures 2f–h, 4t,v). Figure 3s shows seed impressions observed across the outer surface of the base of pottery excavated from the Shymkent site, and Figure 3t is an SEM image of a replica of some of these impressions. Figure 3u shows seed impressions observed across the outer surface of the base of a pottery vessel excavated from the Kainar site. Figure 3v depicts a replica taken from one of these impressions, clearly showing a grain of broomcorn millet that has been removed from the palea and lemma with a small fragment of glume still remaining articulated. Figure 3w shows the bottom of a pottery vessel with many similar impressions excavated from the Bugunskoye site.

Figure 3

3.3.3 Setaria italica

Caryopses of foxtail millet were distinguished based on their lateral view, showing the outer glume rounded and the inner glume somewhat flattened. Additionally, there are small papillate projections on both the inner and outer glumes, and a smooth, crescent-shaped region at the boundary between the inner and outer glumes, which is lacking papillate projections. Figure 2i shows a caryopsis of foxtail millet, with visible papillate projections. In comparing Figure 2i with those of Figures 2j,k, which are what we have identified as Setaria viridis, it is clear that in the latter two examples the papillate projections form ridges. Typical criteria used for the identification of the wild relative of foxtail millet include both: 1) the presence of these ridge-like structures and 2) thinner and smaller seeds (Nasu et al., 2007). In our study, specimens with these ridge structures did not correlate with smaller sizes or narrower seeds, as can be seen in the cluster plot in Figure 4. While we are cautious about any interpretations regarding these specimens, given that we only have two grains for the S. italica category and seven measurable specimens of S. viridis there does not appear to be any clear difference in size between the specimens in each category (Figure 4).

Figure 4

In our data table, we confidently identified four specimens as domesticated foxtail millet, and we left another sixteen specimens labeled as S. viridis. That said, given the large size and rounded shape of these specimens, we suspect that they are domesticated. It is likely that the criteria used to divide wild and domesticated forms needs to be reconsidered when dealing with ancient populations at an earlier stage along the domestication spectrum. Our data is not sufficient to accommodate such a reconsideration, but it is likely the ceramic impression studies will provide such data in the future. Wild Setaria is a common companion weed in foxtail millet fields and the two are often found together in Central Asian archaeobotanical assemblages. We theorize that, at the time these pots were formed, the two were not clearly distinguished as a crop and a weed by local farmers. Ultimately, interpreting the smaller specimens as weeds in cultivated field, arable weeds growing on the edges of villages, or crops that were either not fully domesticated or still immature all seem like plausible possibilities.

3.3.4 Oryza sativa

Caryopses of rice are characterized by their spindle-shaped form, ridges and grooves formed by vascular bundles in the grain and rows of granular protrusions on the husk surface. These surface structures result in a distinct gridded pattern (Figure 3m; note that the replica came from the impression in Figure 3n). Figures 3o,r show rice husks, with clearly visible granular protrusions. Figures 3p,q also show rice husks, but the granular protrusions are not observable on the inner surface, but the ridges and grooves formed by the vascular bundles converging at the base are clearly visible.

4 Results

The data presented here covers twenty four archaeological sites, fifteen of which have never been published before, spanning more than two millennia, from 1700 BC to 400 AD. The sites cover a swath of land that spans the southern rim of Kazakhstan, representing one of the most expansive extant gaps in data for the ancient spread of farming. Of the six Bronze Age sites that we examined, domesticated grains were recovered at all of them, and, out of nine Iron Age sites ascribed to the Saka Culture – which is traditionally thought to be nomadic – eight of them provided domesticated grain impressions. Additionally, of the nine sites that were from more sedentary-looking Iron Age peoples, all but one of them had evidence for domesticated grains. We should, of course, not take the sites lacking evidence as evidence for absence of agriculture, nor should we make any efforts to quantify these findings into an interpretation of the significance in the diet. There is no way to determine how readily grains were lost and inadvertently ended up in wet ceramic clay.

Additionally, of the nine sites that were from more sedentary-looking Iron Age peoples, all but one of them had evidence for domesticated grains. We should, of course, not take the sites lacking evidence as evidence for absence of agriculture, nor should we make any efforts to quantify these findings into an interpretation of the significance in the diet. There is no way to determine how readily grains were lost and inadvertently ended up in wet ceramic clay. The lack of any impressions from either wheat or any legumes is also noteworthy. The logical explanation for a lack of wheat would be that barley and the millets are less water-demanding and easily grow in low-investment plots along riverways, but the presence of rice is hard to rectify with this argument.

Although four caryopses of foxtail millet were identified, a total of 16+ caryopses of Setaria viridis were identified from two sites. It is worth pointing out that wild Setaria is essentially absent from archaeological sites in Central Asia from periods prior to the introduction of the domesticated form, and the wild form is almost always found in ancient sites that have the domesticated form. It is possible that the wild form represents a field crop that is not fully – on a population scale – transitioned to domestication. Additionally, from two Iron Age sites, one rice grain and four rice husks were identified. Although a full measurable specimen only occurred on one sherd, the single rice grain from Kultobe is morphologically more similar to japonica rice. Differentiation between japonica and indica varietal clades can be loosely made based on the criteria that rice with an aspect ratio >2.2 is indica and rice with an aspect ratio <2.0 is japonica (Chen et al., 2020). However, unfortunately, comparing the size of ceramic impressions—which are subject to various biases, such as pottery firing temperature and clay moisture content—to the size of carbonized seeds is inherently flawed and requires further experimentation.

5 Discussion

5.1 Changes in cereal cultivation during the Bronze and Iron Age interface

The presence of grains at an ancient settlement does not necessarily mean local cultivation, and we contend that we cannot prove that these were not precious exotic imports. That said, they were readily dropped and incorporated accidentally into ceramics; hence, it seems more plausible that they were common items. Additionally, some of the grains continue to remain adhered to their glumes or palea and lemma, and chaff is often used by archaeobotanists as an indicator of local cultivation (Fuller et al, 2010; Stevens, 2003). The reasoning suggests that transporting grain would have been a costly enterprise, as a horse, camel, or donkey can only transport a few sacks of grain at a time and the pack animals and their operators would have needed food and water along the way. In the rare cases when grain was transported over land in the ancient world, it would logically have been done after threshing to remove all the inedible weight. While we cannot rule out the possibility that small amounts of grain were transported over the arid steppe for several days from the mountain foothills or several weeks, if the argument is made that they came from the south, it is far more likely that they were grown locally on a small scale.

We believe that the incorporation of millet grains, both in a hulled and dehulled condition (i.e., processed for human consumption), into the pottery clay used to produce domestic cooking vessels, is an indication that these grains were meant for human consumption. Based on SEM observations, almost all caryopses of millet had not been threshed, leaving only the palea and lemma, or inner and outer glumes. The presence of the chaffing material likely suggests local cultivation. That said, many of the barley grains have been removed for their glumes, which indicates that labor-intensive processing had been conducted to prepare them for human consumption rather than for animal feed (Endo et al., 2025).

While we recognize the limitations in our sample sizes, comparing the Bronze Age to the Iron Age may suggest potential changes in crop cultivation. For example, it has already been suggested that there was a transition from the predominant morphotype of barley in the second millennium BC being naked to most of the first millennium BC barley being hulled (Motuzaite Matuzeviciute et al., 2021), whereas in some regions, such as Tibet, naked forms persist through time (Tang et al., 2021). Indeed, all but one of our naked specimens come from the earlier contexts. It has also been well argued previously that, during the late third millennium BC, barley spread up the mountain corridor from earlier sedentary sites to the south. At these early sites in northern Central Asia, barley was the primary crop, accompanied by small amounts of millet, which dispersed into the region from East Asia around this time (Spengler, 2015). However, broomcorn millet appears to have only become a prominent crop at more urban sites during the first millennium BC, at which time foxtail millet appears to have spread across the broader region.

Scholars have already pointed out that, ethnographically, millets are more suited for small-scale cultivation and possibly more mobile lifestyles (Motuzaite Matuzeviciute et al., 2021; Spengler, 2015). Millet has often been cited as having advantages, such as being able to be cultivated in dry environments and having a short cultivation period (Miller et al., 2016). They also have a low sowing seed to harvesting ratio, whereas fewer seeds are needed to plant a small plot of the grain. Millets can easily be planted on a small scale near water sources during seasonal camp movements, and, given the short growing season, can be harvested before herders move to a new camp. In addition, the culms and leaves of millet, when it is sown in the summer and harvested in autumn, can be used as winter fodder for herd animals (Hermes et al., 2019).

Some scholars have been suggested that an increase of broomcorn millet in the economy of Central Asia during the later Iron Age may be tied to the adoption of crop-rotation cycles, with winter wheat alternating with summer-sown irrigated millets (Miller et al., 2016). Hence, the potential increase of millet in the diet of people after the Bronze Age could either be an indication of more intensive use of land close to riverways with the goal of maximizing yields or a more expansive and low-investment system, whereas millets were opportunistically sown close to water sources reducing the need for irrigation. Of the 19 Iron Age sites presented here, broomcorn millet was identified at 15, representing an occurrence rate of 79%, compared to 33% for the Bronze Age. Therefore, we speculate that, in this region during the Iron Age, broomcorn millet may have begun to overshadow barley as the main staple crop.

While we reiterate the inability to clearly quantify the role of these crops in the diet using the ceramic impression method, at the nine Saka-Culture sites, only three barley grains were identified, while more than forty broomcorn millet grains were found. As noted above, the Saka people have long been described as nomads, based largely on the fact that they are primarily known from burial complexes, often containing horse-riding tackle. Three of the four Saka-type sites that we present in this paper are burial complexes; however, as we stated above, the ceramic types are clearly domestic forms and appear to have been used domestically, possibly linking the seeds to domestic, as opposed to ritual use. Additionally, there is no reason to believe that people were carrying heavy ceramic pots across the arid steppe at this time period, and it is logical to assume that the pots were all locally produced. Indeed, we speculate that millet, in the cultural contexts of the Saka-type sites, represents a low-investment complement to a largely pastoralist system.

In contrast to the material culture at the Saka-type sites, the other Iron Age sites are thought to represent more sedentary people, who resided in small villages and focused more on grain cultivation. At these sites too, there is a greater quantity of millet impressions compared to barley, as is evidenced in the results from the five sites of the Chirik-Rabat Culture in the Kyzylorda Desert, dating to the first millennium BC (Endo et al., 2025). We further speculate that these data suggest that a transition towards a greater reliance in the cultivation system on millets had already begun in this region by the first half of the first millennium BC. Further research is needed to either refute or support the idea that people started to cultivate irrigated summer crops of millet at this time, but these interpretations present enticing directions to take future research.

Furthermore, there is ongoing debate over whether the charred millets reported in Central Asia from the Bronze to the Iron ages were used as livestock feed (Hermes et al., 2019), ritual offerings (Frachetti, 2014), or whether they were also consumed by humans (Motuzaite Matuzeviciute et al., 2022). The multi-site dataset that we present here, rather conclusively illustrates that these grains were a regular component in the diet. The fact that many of the millet grains had been removed from their bracts (upper glumes) suggests that threshing had been conducted. If the grains were to be used solely as animal feed, threshing would not be necessary. Therefore, we hypothesize that millet seeds were used for human consumption. In addition, many impressions of broomcorn millet were discovered on the outer surfaces of the bases of ceramics from four Iron Age sites (Figures 3s,u,w). Similar concentrations of millet grains were observed on ceramics from the Chilik Rabat sites, which date back to the first millennium BC. As we show Figures 3t,v, the inner husk (lemma) and outer husk (palea) of the millet grains have been beaten off and removed. Similar examples of numerous millet impressions concentrated on the outer surface of bases has been confirmed not only in Central Asia, but also in ceramics excavated from the Novokiyivka site of the Sabatynivka Culture (second millennium BC) in Ukraine, and on ceramics from the Middle Yayoi period (late first millennium BC) in Japan (Kakinuma et al., 2026). The interesting case of Togolok 21 (early second millennium BC), could also be mentioned here, whereas seed impressions were originally and rather infamously identified as cannabis seeds in the 1970s, but reanalyzed as dozens of millet impressions on the bottom of a pot (Bakels, 2003).

5.2 The presence of Oryza sativa in the fusion zone

At the Myntobe and Kultobe sites along the Syrdarya River, associated with artifacts ascribed to the Kaunchin Culture (100 BC–400 AD), five impressions of rice were identified, consisting of one rice grain and four husks. Three of the specimens came from impressions on sherds from Myntobe and two came from Kultobe (Table 2). While most charred ancient rice in Asia lacks its husks (Chen et al., 2020; Huang et al., 2023, Spengler et al., 2021b), these pottery impressions permit us to clearly see the surface structure. The retention of the protective casing suggests that the de-husking of the rice grains was likely performed near the pottery-making area within the dwelling – possibly in a multipurpose domestic area. Spengler et al. (2021b) have speculated that rice during the early medieval period in Central Asia was a rare or exotic food, but our finds suggest that it may have been more ubiquitous in ordinary households than previously recognized. Unfortunately, direct dating is impossible, so the exact period cannot be narrowed down any further. Nevertheless, two AMS radiocarbon dates from teeth excavated from burials at the Myntobe site are 184 ± 25 cal AD and 318 ± 27 cal AD, the later of these two dates being the exact burial where the pot with the rice impressions was recovered (see Gursoy and Sizdikov, 2024 and Sizdikov et al., 2025). We believe these dates are likely to be reliable indications for the rice are as well, given the discreet nature of burials. Additionally, charcoal fragments collected from the second construction horizon of the Kultobe settlement, where a rice impression was also detected, was dated to 354–400 cal AD (73.5%) (Sizdikov et al., 2025). The grains date to roughly the period of the earliest recorded spread of rice into Central Asia, making them some of the oldest rice ever discovered outside East Asia.

Archaeobotanists have recently compiled all historical and archaeobotanical data for ancient rice in Central Asia (Spengler et al., 2021a), permitting a rough sketch of the dispersal, adoption, and intensification of this crop in cultivation systems outside East Asia. While a few finds of grains have been reported from urban sites further south, there are no reported claims of rice along the Syrdarya or northern mountain foothills prior to the establishment of fortified medieval centers. Notably, in southern Kazakhstan, there are reports of rice dating back to the seventh century AD from Konyr-tobe, the fourth to fifth centuries AD at Karaspan-tobe, and at Djuvan-tobe in the seventh century AD (Bashtannik, 2008). Additionally, two charred rice grains from the Khalchayan site in southeastern Uzbekistan have been reported, with direct radiocarbon dates of 236–194 cal AD (Chen et al., 2020). Recently, a single fragmented specimen of an older rice grain was reported from the Sym-Ota 1 site in Uzbekistan (Brite et al., 2026). The specimen only represents the proximal end of a grain, but assuming the identification is accurate, it is directly dated to 409–211 cal BC (calibrated 95% CI) – now making it the oldest specimen outside East Asia. Two other rice grains (339–46 cal BC and 196 cal BC–cal AD 4 at 95.4% confidence) have recently been reported from the Ghal e-Ben site on the southern coast of the Caspian Sea in northern Iran (Huang et al., 2023:10).

While most discussions of crop dispersal through Central Asia have focused on the mountains, what has been coined the IAMC (Frachetti, 2012), there is currently a lack of definitive rice evidence from this period across northern Central Asia. However, there are very early finds of rice in the valleys of the Hindu Kush, which seem to have served as a cultural contact zone between South and Central Asia, suggesting a southern dispersal for rice (Spengler et al., 2021b). Therefore, it is highly likely that the rice identified at the two sites in our study further supports the argument of a spread from the south. If so, this “fusion zone” was not merely a route for the exchange of grains between east and west, but rather a crossroads where grains also arrived from the south with other cultural elements. It is likely that the Eurasian spread of rice was not a gradual diffusion from neighboring regions, like it appears to have been for broomcorn millet, but rather a leapfrogging pattern that crossed harsh environments and focused on wetland areas in oases and river valleys. Further research is needed to better understand the ways that rice was cultivated and whether paddy systems spread rapidly with the grains.

6 Conclusion

We examined assemblages of ceramic sherds from 24 archaeological sites (15 of the assemblages are newly analyzed for this paper) spanning a period from 1700 BC to AD 400. Using impressions of agricultural grains identified in ceramics, we clarify the role of farming in subsistence practices from the Bronze Age to the Iron Age along the arid steppe of southern Kazakhstan. This key region was, geographically, a cultural fusion zone between the more pastoral steppe regions of northern Central Asia and the more intensive agricultural regions to the south. We show that people across Central Eurasia found ways to fit agriculture into their diverse economic systems and to adapt specific types of farming to more arid regions or less arable zones.

Despite the fact that the ceramic impression method does not permit a quantification of the grains in the diet or assessments of their significance as part of the economy, our discoveries are significant, as they provide the first clear evidence for domesticated crops in this region from this early time. While our data alone cannot prove local cultivation, we think local cultivation is more likely than long-distance transport of grain over the arid steppe (a claim that merits further investigation). We also suggest that a shift unfolded from a greater focus on cereals (notably barley) during the second millennium BC towards a greater incorporation of low-investment millet cultivation, particularly of broomcorn millet, during the first millennium BC. We also reason, based on the culturally modified nature of the grains in our impressions, that millets were used for human consumption; ethnographic analogy would also suggest that the stems and leaves were utilized as fodder. The discoveries of rice and foxtail millet are particularly significant, not just because our discoveries of rice are some of the oldest remains of this grain ever discovered outside of East Asia, but its presence suggests a cultural diffusion from the south (it is likely that foxtail millet followed a similar trajectory). We demonstrate that the ecotonal belt between the steppe and desert areas served as a fusion zone for cultural memes spreading latitudinally from either direction. Our study illustrates how culturally diverse and dynamic Central Asia was through time, further drawing the prevailing scholarly views away from the nomadic narratives and placing the people of this region at the center of an interconnected ancient world.

References

• 1

  AkishevK. A. (1970). “Origins of sedentism and farming of the ancient Wusun of Semirech’ye (in Russian),” in About the Ancient Cultures of Kazakhstan, (Alma-Ata: Nauka), 69–77.

  • Google Scholar
  • View reference in article

• 2

  AnZ. (1992). “The bronze age in eastern parts of Central Asia,” in History of Civilizations of Central Asia, vol. 1 of the Dawn of Civilization: Earliest Times to 700 B.C., eds. DaniA. H.MassonV. M. (Paris: UNESCO), 319–336.

  • Google Scholar
  • View reference in article

• 3

  ArbuckleB. S.HammerE. (2019). The rise of pastoralism in the ancient near east. J. Archaeol. Res.27, 391–445. 10.1007/s10814-018-9124-8

  • CrossRef
  • Google Scholar
  • View reference in article

• 4

  BaipakovK. M. (2008). The city and steppe in antiquity: sedentism and agriculture in the Saka and Wusun of Semirech’ye. Izvestia1, 3–25. (In Russian).

  • Google Scholar
  • View reference in article

• 5

  BakelsC. C. (2003). The contents of ceramic vessels in the bactria-Margiana archaeological complex, Turkmenistan. Electron. J. Vedic Stud.9, 49–52. 10.11588/ejvs.2003.1.785

  • CrossRef
  • Google Scholar
  • View reference in article

• 6

  BashtannikS. (2008). Archaeobotanical research in medieval sites in the Arys River Valley. Archaeol. Ethnol. Anthropol. Eurasia33 (1), 85–92. 10.1016/j.aeae.2008.04.009

  • CrossRef
  • Google Scholar
  • View reference in article

• 7

  BeckwithC. I. (2009). EMPIRES oF tHE SILK ROAD A History of Central Eurasia from the Bronze Age to the Present. Princeton, New Jersey: Princeton University Press.

  • Google Scholar
  • View reference in article

• 8

  BeckwithC. I. (2023). The Scythian Empire: Central Eurasia and the Birth of the Classical Age from Persia to China. New Jersey: Princeton University Press: Princeton.

  • Google Scholar
  • View reference in article

• 9

  BriteE. (2018). The hydrosocial empire: the Karakum river and the soviet conquest of Central Asia in the 20th century. J. Anthropol. Archaeol.52, 123–136. 10.1016/j.jaa.2018.08.003

  • CrossRef
  • Google Scholar
  • View reference in article

• 10

  BriteE. B.ToreniyazovA.HunterS. A.AmmermanS.BekbaulievA.KhozhalipesovI.et al (2026). Climate change and world history: evidence from the site of sym-ota 1 in the Aral Sea basin. Quat. Sci. Adv.21 (2026), 100313. 10.1016/j.qsa.2025.100313

  • CrossRef
  • Google Scholar
  • View reference in article

• 11

  ChangC.TourtellotteP. A.BaipakovK. M.GrigorievF. G. (2002). The Social Evolution of Eurasian Steppe Communities in Southeastern Kazakhstan: The Kazakh-American Talgar Project 1994-2001. Almaty: Institute of Archaeology, Kazakh Academy of Sciences.

  • Google Scholar
  • View reference in article

• 12

  ChangC. (2018). Rethinking Prehistoric Central Asia: Shepherds, Farmers, and Nomads. New York: Routledge.

  • Google Scholar
  • View reference in article

• 13

  ChenG.ZhouX.WangJ.MaJ.KhasannovM.KhasanovN.et al (2020). Kushan period rice in the Amu Darya basin: evidence for prehistoric exchange along the southern Himalaya. Sci. China Earth Sci.63, 841–851. 10.1007/s11430-019-9585-2

  • CrossRef
  • Google Scholar
  • View reference in article

• 14

  ChenK.HiebertF. T. (1995). The late prehistory of Xinjiang in relation to its neighbors. J. World Prehist.9 (2), 243–300. 10.1007/BF02221840

  • CrossRef
  • Google Scholar
  • View reference in article

• 15

  ChristianD. (2000). Silk roads or steppe roads? The silk roads in world history. J. World. Hist.11 (1), 1–26. 10.1353/jwh.2000.0004

  • CrossRef
  • Google Scholar
  • View reference in article

• 16

  DemaitreE.DemaitreA. (1981). The five avatars of the scythian. Hist. Eur. Ideas.2 (4), 315–337. 10.1016/0191-6599(81)90044-9

  • CrossRef
  • Google Scholar
  • View reference in article

• 17

  Di CosmoN. (1994). Ancient inner Asian nomads: their economic basis and its significance in Chinese history. J. Asian Stud.53 (4), 1092–1126. 10.2307/2059235

  • CrossRef
  • Google Scholar
  • View reference in article

• 18

  Di CosmoN. (2015). “China-Steppe relations in historical perspective,” in Complexity of Interaction Along the Eurasian Steppe Zone in the First Millennium CE, eds. BemmannJ.SchmauderM. (Bonn: University of Bonn), 49–72.

  • Google Scholar
  • View reference in article

• 19

  EndoE.ShodaS.FrachettiM.KaliyevaZ.KiyasbekG.ZhuniskhanovA.et al (2023). Pottery impressions reveal earlier westward dispersal of foxtail millet in inner Asian moun­tain corridor. Agronomy2023, 1706. 10.3390/agronomy13071706

  • CrossRef
  • Google Scholar
  • View reference in article

• 20

  EndoE.KaliyevaZ.UtubayevZ.KasenovaA.SuyundikovaM.OnggarulyA.et al (2025). Grain impressions in ceramic vessels illustrate the extent of 1st millennium BC agriculture across the northern Central Asian steppe. Veg. Hist. Archaeobot.35, 363–376. 10.1007/s00334-025-01058-3

  • CrossRef
  • Google Scholar
  • View reference in article

• 21

  EleuovM.KaliyevS.AkymbekE.MurgabayevS.AlzhanovR. (2005). Report on the 2005 archaeological studies of the Turan expedition in south Kazakhstan region. Turkistan (in Kazakh). HKITU Archive5, 107.

  • Google Scholar
  • View reference in article

• 22

  FrachettiM.SpenglerR. N.FritzG. J.Mar’yashevA. N. (2010). Earliest direct evidence for broomcorn millet and wheat in the central Eurasian steppe region. Antiquity.84, 993–1010. 10.1017/S0003598X0006703X

  • CrossRef
  • Google Scholar
  • View reference in article

• 23

  FrachettiM. (2012). Multiregional emergence of mobile pastoralism and nonuniform institutional complexity across Eurasia. Curr. Anthropol.53 (1), 2–38. 10.1086/663692

  • CrossRef
  • Google Scholar
  • View reference in article

• 24

  FrachettiM. D. (2014). “Seeds for the soul: ideology and diffusion of domesticated grains across inner Asia,” in Reconfiguring the Silk Road: New Research on East-West Exchange in Antiquity, eds. MairV. H.HickmanJ. (Philadelphia, Pennsylvania: University of Pennsylvania Press), 41–54.

  • Google Scholar
  • View reference in article

• 25

  FrachettiM. D.SmithC. E.TraubC. M.WilliamsT. (2017). Nomadic ecology shaped the highland geography of Asia’s silk roads. Nature543, 193–198. 10.1038/nature21696

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 26

  FullerD. Q.AllabyR. G.StevensC. (2010). Domestication as innovation: the entanglement of techniques, technology and chance in the domestication of cereal crops. World. Archaeol.42 (1), 13–28. 10.1080/00438240903429680

  • CrossRef
  • Google Scholar
  • View reference in article

• 27

  GroussetR. (1970). The Empire of the Steppes A History of Central Asia. New Brunswick, New Jersey: Rutgers University Press.

  • Google Scholar
  • View reference in article

• 28

  GursoyM.SizdikovB. (2024). Относительное и Абсолютное Датирование Курганов в Некрополе Мынтобе (in Russian) (relative and absolute dating of mounds in the mintobe Necropolis), Вестник КазНПУ имени Абая. Серия: Исторические и социально-политические науки3, 82. 10.51889/2959-6017.2024.82.3.018

  • CrossRef
  • Google Scholar
  • View reference in article

• 29

  HermesT. R.FrachettiM. D.Doumani DupuyP. N.Mar'yashevA.NebelA.CherylA.et al (2019). Early integration of pastoralism and millet cultivation in bronze age Eurasia. Proc. Biol. Sci.286 (1910), 20191273. 10.1098/rspb.2019.1273

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 30

  HuangY.DengZ.NashliH. F.FullerD. Q.Xiaohong WuX.SafariM. (2023). The early adoption of east Asian crops in west Asia: rice and broomcorn millet in northern Iran. Antiquity.2023, 1–16. 10.15184/aqy.2023.42

  • CrossRef
  • Google Scholar

• 31

  ItahashiY.AnanyevskayaE.YonedaM.Ventresca MillerA. R.NishiakiY.Motuzaite MatuzeviciuteG. (2020). Dietary diversity of Bronze-Iron Age populations of Kazakhstan quantitatively esti­mated through the compound-specific nitrogen analysis of amino acids. J. Archaeol. Sci. Rep.33, 102565. 10.1016/j.jasrep.2020.102565

  • CrossRef
  • Google Scholar

• 32

  KakinumaM.EndoE.TakumaK.FukadaY.YoshikawaT. (2026). Survey of seed impressions on pottery excavated from the Shittapara site (in Japanese). Sat. Archaeol.44, 95–115.

  • Google Scholar

• 33

  KumagaiM.SakaiH. (2024). "Genomic studies on the origin, hybridization and dispersal of Asian rice," in Phylogeographic History of Plants and Animals Coexisting with Humans in Asia, Evolutionary Studies. eds. OsadaN.et al, (Springer Nature), 131–134. 10.1007/978-981-97-6887-5_9

  • CrossRef
  • Google Scholar

• 34

  KurmaniyazovY.TazhekeyevA.BolelovS.DarmenovR.SultanzhanovZ.SharapovaK.et al (2023). Animal stable isotope data from Dzhetyasar and Oguz culture sites of the lower Syr Darya in Kazakhstan. Archaeol. Res. Asia35, 100450. 10.1016/j.ara.2023.100450

  • CrossRef
  • Google Scholar

• 35

  KuzminaE. (2008). The Prehistory of the Silk Road: Encounters with Asia. V. H. Mair (Transl.). Philadelphia, Pennsylvania: University of Pennsylvania Press.

  • Google Scholar

• 36

  LevinaL. M. (1971). Keramika Nizhnei I Srednei Syrdar’i v I tys. n.e.(in Russian). Moscow: Trudy Khorezmskoi arkheologo-etnograficheskoi ekspeditsii; VII. Nauka.

  • Google Scholar

• 37

  LiK.QinX.PlunkettG.PressleeS.SpenglerR.XuB.et al (2024). Multi-proxy investigations of bronze age diet and environment in the hyper-arid eastern Tarim basin (lop Nur), northwest China. J. Archaeol. Sci.171 (1), 106089. 10.1016/j.jas.2024.106089

  • CrossRef
  • Google Scholar

• 38

  MillerN. F.SpenglerR. N.IIIFrachettiM. (2016). Millet cultivation across eurasia: origins, spread, and the influence of seasonal climate. Holocene26, 1566–1575. 10.1177/0959683616641742

  • CrossRef
  • Google Scholar

• 39

  Motuzaite MatuzeviciuteG.HermesT. R.Mir-MakhamandB.TabaldievK. (2020). Southwest Asian cereal crops facilitated high-elevation agriculture in the central Tien Shan during the third millennium. PLoS One15 (5), e0229372. 10.1371/journal.pone.0229372

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 40

  Motuzaite MatuzeviciuteG.Mir-MakhamadB.TabaldievK. (2021). The first comprehensive archaeobotanical analysis of prehistoric agriculture in Kyrgyzstan. Veget Hist Archaeobot30, 743–758. 10.1007/s00334-021-00827-0

  • CrossRef
  • Google Scholar

• 41

  Motuzaite MatuzeviciuteG.AnanyevskayaE.SakalauskaiteJ.SoltobaevO.TabaldievK. (2022). The integration of millet into the diet of Central Asian populations in the third millennium BC. Antiquity.96 (387), 560–574. 10.15184/aqy.2022.23

  • CrossRef
  • Google Scholar

• 42

  MurakamiN.OnggarulyA.RakhimzhanovaS.StandallE. A.TalbotH. M.LucquinA.et al (2022). Lipid residues in ancient pastoralist pottery from Kazakhstan reveal regional differences in cooking practices. Front. Ecol. Evol.10, 1032637. 10.3389/fevo.2022.1032637

  • CrossRef
  • Google Scholar

• 43

  MurgabayevS. S.BakhtybayevM. M.JoldasbayevS. J.JetibayevK. M.MaldybekovaL. J.ArynovK. S. (2016). Archeological research at the Bağanalı Bronze Age burial site. Anthropol. Archeol. Eurasia.55 (3–4), 269–300. 10.1080/10611959.2016.1317558

  • CrossRef
  • Google Scholar

• 44

  MurgabayevS. S.BakhtybayevM. M.MaldybekovaL. (2018). “Research at the zhasazhyryk I burial site (in Kazakh),” in International Scientific-Methodical Conference Proceedings, ed. OrazbaevX. (Almaty: Kazakh University), 73–80.

  • Google Scholar

• 45

  MurgabayevS. S.BakhtybayevM. M.MaldybekovaL. D. (2020). Research at the Karasuiyr III bronze age burial ground (in Russian). Put’ Nauki (Way of Science)10 (80), 41–47.

  • Google Scholar

• 46

  MustafayevZ. S.TuletayevA. B.AldazhanovaG. B. (2023). Climate change in the natural zones of the turkestan region of the republic of Kazakhstan (in Russian). Geografiya I Vodnye Resursy3, 40–48.

  • Google Scholar

• 47

  NasuH.MomoharaA.YasudaY.HeJ. (2007). The occurrence and identification of Setaria italica (L.) P. Beauv. (foxtail millet) grains from the chengtoushan site (ca. 5800 cal B.P.) in central China, with reference to the domestication centre in Asia. Veg. Hist. Archaeobot.16 (6), 481–494. 10.1007/s00334-006-0068-4

  • CrossRef
  • Google Scholar

• 48

  OralbayY.TazhekeyevA.MeirmanovaG.DarmenovR.BilalovS.AmirgalinaG. (2016). Monuments of the early iron age of the Aral-Caspian region as a historical source. Anthropologist26 (1,2), 137–144. 10.1080/09720073.2016.11892141

  • CrossRef
  • Google Scholar

• 49

  PachikinK. M.ErokhinaO. G.SongulovE. E.YershibulovA. K.AdaminG. K.YakovlevaN. A. (2022). Ecological state of the soils within the Karatau ridge foothill plains (Kentau city surroundings) (in Russian). Soil Sci. Agrochem.2, 22–32.

  • Google Scholar

• 50

  RouseL. M.CerasettiB. (2014). Ojakly: a late bronze age mobile pastoralist site in the Murghab region, Turkmenistan. J. Field Archaeol.39 (1), 32–50. 10.1179/0093469013Z.00000000073

  • CrossRef
  • Google Scholar

• 51

  SasakiY.RakhimzhanovaS.OnggarA.KairmagambetovA.EndoE.Doumani DupuyP.et al (2022). Broomcorn mil­let impressions on early iron age ceramics from tortoba burial mound in western Kazakhstan. Kazakhstan Archaeol4, 116–132.

  • Google Scholar

• 52

  SizdikovB. S.GursoyM.BakhtybayevM. M.ZhetibayevKМ (2025). Результаты археологических исследований на городище Культобе в 2023 году (result of archaeological research work, conducted in the ancient settlement of Kultobe in 2023) (in Russian). J. Hist.113 (2), 234–244. 10.26577/JH.2024.v113i2-020

  • CrossRef
  • Google Scholar

• 53

  SilvaF.WeisskopfA.CastilloC.MurphyC.Kingwell-BanhamE.QinL.et al (2018). A tale of two rice varieties: modelling the prehistoric dispersals of japonica and proto-indica rices. The Holocene28 (11), 1745–1758. 10.1177/0959683618788634

  • CrossRef
  • Google Scholar

• 54

  SmagulovE. A.BaratovS. R. (2017). Preliminary results of archaeological investigations at the Sherbay Bronze Age burial ground (in Russian). Izvestiya MON RK – NAN RK. Soc. Sci. Ser.1, 38–41. Available online at:https://archeo.kz/ru/library/izvestiya-an-kazahskoj-ssr/izvestiya-ministstva-obrazovaniya-i-nauki-respub/read/(Accessed July 30, 2026).

  • Google Scholar

• 55

  SpateM.LeipeC.Motuzaite MatuzeviciuteG. (2022). Reviewing the palaeoenvironmental record to better understand long-term human-environment interaction in inner Asia during the late holocene. Front. Ecol. Evol.10, 939374. 10.3389/fevo.2022.939374

  • CrossRef
  • Google Scholar

• 56

  SpenglerR. N.IIIMichaelD.FrachettiM. D.FritzG. J. (2013). Ecotopes and herd foraging practices in the bronze and Iron Age, steppe and mountain ecotone of Central Asia. J. Ethnobiol.33 (1), 125–147. 10.2993/0278-0771-33.1.125

  • CrossRef
  • Google Scholar

• 57

  SpenglerR. N.FrachettiM. D.DoumaniP. N.RouseL. M.CerasettiB.BullionE.et al (2014). Early agriculture and crop transmission among bronze age mobile pastoralists of central Eurasia. Proc. R. Soc. B281, 20133382. 10.1098/rspb.2013.3382

  • CrossRef
  • Google Scholar

• 58

  SpenglerR. N.III (2015). Agriculture in the Central Asian bronze age. J. World Prehist.28 (3), 215–253. 10.1007/s10963-015-9087-3

  • CrossRef
  • Google Scholar

• 59

  SpenglerR. N.IIIRyaboginaN.TarasovP.WagnerM. (2016). The spread of agriculture into northern Central Asia: timing, pathways, and environmental feedbacks. Holocene26, 1527–1540. 10.1177/0959683616641739

  • CrossRef
  • Google Scholar

• 60

  SpenglerR. N.IIINaomiF.MillerN. F.NeefR.TourtellotteP. A.ChangC. (2017). Linking agriculture and exchange to social developments of the Central Asian Iron Age. J. Anthropol. Archaeol.48, 295–308. 10.1016/j.jaa.2017.09.002

  • CrossRef
  • Google Scholar

• 61

  SpenglerR. N.IIIStarkS.ZhouX.FuksD.TangL.Mir-MakhamadB.et al (2021a). A journey to the west: the ancient dispersal of rice out of east Asia. Rice14, 83. 10.1186/s12284-021-00518-4

  • CrossRef
  • Google Scholar

• 62

  SpenglerR. N.IIITangL.NayakA.BoivinN.OlivieriL. M. (2021b). The southern Central Asian mountains as an ancient agricultural mixing zone: new archaeobotanical data from Barikot in the Swat Valley of Pakistan. Veg. Hist. Archaeobot.30, 463–476. 10.1007/s00334-020-00798-8

  • CrossRef
  • Google Scholar

• 63

  StevensC. J. (2003). An investigation of agricultural consumption and production models for prehistoric and Roman Britain. Environ. Archaeol.8, 61–76. 10.1179/env.2003.8.1.61

  • CrossRef
  • Google Scholar

• 64

  TaleevD. A.SizdikovB. S.SeraliyevA. A. (2019). History оf the keless region (in Kazakh). in bulletin of the L.N. Gumilyov Eurasian National University. Hist. Sci. Philos. Relig. Ser.4 (129), 48–62.

  • Google Scholar

• 65

  TangL.LuH.SongJ.WangdueS.ChenX.ZhangZ.et al (2021). The transition to a barley-dominant cultivation system in Tibet: first millennium BC archaeobotanical evidence from Bangga. J. Anthropol. Archaeol.61, 101242. 10.1016/j.jaa.2020.101242

  • CrossRef
  • Google Scholar

• 66

  TaylorW. T. T.PruvostM.PosthC.RenduW.KrajcarzM. T.AbdykanovaA.et al (2021). Evidence for early dispersal of domestic sheep into Central Asia. Nat. Hum. Behav.5 (9), 1169–1179. 10.1038/s41562-021-01083-y

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 67

  ToonenW. H.MacklinM. G.DawkesG.DurcanJ. A.LemanM.NikolayevY.et al (2020). A hydromorphic reevaluation of the forgotten river civilizations of Central Asia. PNAS117 (52), 1–7. 10.1073/pnas.2009553117

  • CrossRef
  • Google Scholar

• 68

  UshinoT.TagawaH. (1991). Replication method of impressions on the pottery surface (in Japanese). Archaeol. Nat. Sci.24, 13–36.

  • Google Scholar

• 69

  ZhouX.YuJ.SpenglerR. N.ShenH.ZhaoK.GeJ.et al (2020). 5,200-year-old Cereal grains from the eastern Altai mountains redate the trans-Eurasian crop exchange. Nat. Plants.6, 78–87. 10.1038/s41477-019-0581-y

  • Pubmed Abstract
  • CrossRef
  • Google Scholar