Article | Open | Published:

Urban and nomadic isotopic niches reveal dietary connectivities along Central Asia’s Silk Roads

Scientific Reportsvolume 8, Article number: 5177 (2018) | Download Citation


The ancient ‘Silk Roads’ formed a vast network of trade and exchange that facilitated the movement of commodities and agricultural products across medieval Central Asia via settled urban communities and mobile pastoralists. Considering food consumption patterns as an expression of socio-economic interaction, we analyse human remains for carbon and nitrogen isotopes in order to establish dietary intake, then model isotopic niches to characterize dietary diversity and infer connectivity among communities of urbanites and nomadic pastoralists. The combination of low isotopic variation visible within urban groups with isotopic distinction between urban communities irrespective of local environmental conditions strongly suggests localized food production systems provided primary subsistence rather than agricultural goods exchanged along trade routes. Nomadic communities, in contrast, experienced higher dietary diversity reflecting engagements with a wide assortment of foodstuffs typical for mobile communities. These data indicate tightly bound social connectivity in urban centres pointedly funnelled local food products and homogenized dietary intake within settled communities, whereas open and opportunistic systems of food production and circulation were possible through more mobile lifeways.


Medieval Central Asia (ca. AD 2nd–16th c.) was a locus of extensive cultural and economic interaction between East Asia, the Middle East, and Europe through a vast network of overland trade routes, commonly called the ‘Silk Roads’1,2,3,4. Urban centers, located in fertile oases and often described by archaeologists and historians alike as cosmopolitan cities5,6, helped anchor Silk Road exchange and foster early globalization across Asia7,8. Although settled communities provided a substantial economic foundation through agricultural output and the manufacture of valuable craft commodities such as metals, ceramics, glass, and textiles9,10,11,12,13,14, mobile pastoralists also had strong influence on the trade system as operators of highland pathways that were based on seasonal movements for herding livestock15.

While transfers of objects and materials have long represented the intensity and scope of Silk Road exchange16, we still lack detailed data about the way food systems, which reflect sustained engagements with the environment and broader community dynamics, were influenced by these far-reaching economic networks. Medieval Central Asia was defined by unusually diverse multicultural intersections, sudden social upheavals, and frequent demographic movements (Supplementary Information 1), but fundamental and perhaps durable dealings with food remain unclear. Food is culturally expressive of ecological adaptation, social relations, ideology, and economy17,18,19,20,21. Resolving dietary diversity in ancient Central Asian foodways provides an opportunity to investigate subsistence models of ‘nomadic’ and ‘urban’ communities that, together, played key roles in transcontinental interaction across the Silk Roads.

In Central Asia, which is characterized by strong seasonal climate and uneven distribution of resources on the landscape22,23, there is high potential for dietary diversity among pastoralist communities. Mobile pastoralists heavily subsist on livestock herding but also draw from a variety of food resources beyond domesticated animal products, including cultivated cereals, wild plants, fishes, and hunted game, which are exploited and consumed at varying intensities depending on environment, seasonal availability, mobility, and social networks24,25,26,27,28. On the other hand, medieval urban food systems were strongly invested in cereal agriculture, food storage, and sedentism29,30, and were subject to powerful political and religious institutions7, which may have generated less diverse dietary repertoires. However, dynamic commercial, political, and social activities between population centers and peripheral settlements, in addition to transactions with pastoralists, could have greatly expanded food availability and choice, as people and provisions, such as grains and live animals, regularly moved between urban and nomadic domains31,32,33,34,35.

We consider community-level dietary breadth over long periods of human life history to be a marker of dietary connectivity, which represents the cultural integration of food production, distribution, and consumption among individuals. Through globalization processes, which involve growing economic networks between increasingly distant communities, cultural differences diminish as groups cooperate and synchronize their consumption patterns, whether of foods, styles, or ideas36. In contemporary societies, globalized food economies expand with production standardization and increased dietary uniformity37,38. Along these lines, high intra-community dietary variability indicates that community members maintained divergent connections to food resources that express individual dietary preferences and group partitioning. Conversely, low dietary variability within communities signals converged trajectories of foods that reflect shared dietary practices and socio-economic coordination.

In order to establish diversity in human dietary intake in Central Asian urban and nomadic communities, we analysed the carbon and nitrogen stable isotopic composition of bone collagen from human remains of 74 individuals (Table 1). Carbon and nitrogen stable isotope ratios (δ13C and δ15N) provide an integrative measure of dietary intake39,40, and human bone collagen reflects a 10–15 year rolling average of protein consumption41,42,43. We model isotopic niches in bi-variate isotopic space (δ-space) to estimate the breadth and structure of community-level diets based on intra-group variation across both dimensions of isotopic ratios simultaneously44,45. Food resources with distinct isotopic content, which are incorporated into individual diets in various proportions, drive community dietary breadth46. The isotopic niche modelling uses Bayesian inference to fit standard ellipses to data points that are then expressed as probability distributions of area (‰2) and position in δ-space47. Crucially, we performed redundancy analysis to unravel isotopic diversity driven by culturally defined food choices from environmental variables that influence the isotopic composition of food resources.

Table 1 Geographic and archaeological information about analysed sites, including number of human samples. Detailed archaeological information for each site is provided in Supplementary Information 3. *See methods.

Isotopic variation in Central Asia

We sampled 14 cemeteries associated with medieval Silk Road communities that span a long transect of Central Asian geography to include present-day Kazakhstan, Uzbekistan, and Turkmenistan (Fig. 1). The human individuals analysed in this study represent urban and non-urban consumers, who potentially had access to substantial food options that were available through prolific agricultural systems and marketplaces that drew in people and foodstuffs from oasis, desert, steppe, and highland environments (Supplementary Information 2). Diverse food remains were recovered from the sites represented in this study, which included cereals, legumes, fruits, fish, and livestock (Supplementary Information 3).

Figure 1
Figure 1

Map of Central Asia showing sites and regions with human stable isotopic data (δ13C and δ15N) analysed in this paper. Uzbekistan: 1) Tok-kala, 2) Uturlik, 3) Chor Dona, 4) Chartok, 5) Tashbulak, 6) Altyntepe, 7) Frinkent; Turkmenistan: 8) Geoktchik Depe, 9) Misrijan; Kazakhstan: 10) Konyr-Tobe, 11) Temirlanovka, 12) Turgen, 13) Butakty, 14) Karatal. Map generated with Quantum GIS, version 2.18.2 (, using public domain data from Natural Earth (

Pronounced variation in the regional environments and topographies, combined with strong seasonality, in which these sites are situated, confers high isotopic variation in Central Asian foodwebs accessed by people. Vegetation communities range from cool montane meadows and forests consisting of largely C3 taxa to hot lowland deserts that support both C3 and C4 plants48,49,50,51, which exhibit (pre-modern52,53) δ13C values of −25 ± 2–5‰ and −11 ± 1‰, respectively54,55,56,57. Food crops exploited in medieval Central Asia included the full spectrum of Eurasian domesticates including C3 taxa, such as wheat, barley, rice, nuts, and fruits, and C4 taxa, such as millets58. Although nitrogen isotopic variation in vegetation communities in Central Asia is under-characterized, research in ecosystems comparable to those of Central Asia, such as the Gobi steppe-desert59, the steppe deserts of the Caspian Depression60, and semi-arid western Loess Plateau61, demonstrate wide variation in plant δ15N values, ranging from −5 to 14‰, due to differences in local soil nitrogen pools and animal stocking rates62,63,64,65. In general, fish exhibit high δ15N values relative to terrestrial fauna66,67, and in Central Asia, fish exhibit an apparent continuum of δ13C values from ca. −11.5‰ to −27‰68.


Human remains in this study represent two chronological intervals of the medieval period. The bulk of the dataset (n = 63) dates to a ‘mid-late’ period of 6th–13th c., which were recovered from sites in southern Kazakhstan, Uzbekistan, and western Turkmenistan. A small sample (n = 11) dates to an ‘early’ period of nomadic occupation at sites in southern Kazakhstan, which allows for a diachronic comparison of nomadic dietary intake within this region. Overall, human isotopic values range from ca. −20‰ to −10.5‰ for δ13C and 9‰ to 15‰ for δ15N (Fig. 2a). Isotope values cluster in both δ13C and δ15N for urban communities on a regional basis, while pastoralist communities from later periods exhibit wide distributions of δ13C values and relatively narrow ranges of δ15N values. Subsequent analyses using Bayesian inference clarified isotopic differences among communities while factoring in uncertainty due to small sample sizes. Summary statistics of isotopic data are provided in Supplementary Figure S1 and Table S1; raw isotopic data are provided in Supplementary Tables S2S4.

Figure 2
Figure 2

(a) Human carbon and nitrogen isotopic ratios from medieval Central Asia; (b–c) Posterior probability distributions of isotopic means obtained by Bayesian bootstrapping (meanb) from medieval urban communities in Uzbekistan and Turkmenistan and (d-e) from medieval nomadic communities in southern Kazakhstan.

Human isotopic variation across medieval Central Asia

Significant distinctions in the dietary intake of medieval urban and nomadic communities across Central Asia are revealed by Bayesian means (meanb) of δ13C and δ15N values. Sharp dissimilarities are observed between urban communities located south of the Syr-Darya river, which provides a rough environmental boundary between the semi-arid/mountain steppe landscapes in southern Kazakhstan and sandy desert landscapes interspersed with oases and foothill zones to the east. Significant differences in the 95% CIs of isotopic values from each southerly community, with the exception of West Pamir-Alay and Tashkent, are present with meanb δ13C values of −19.9‰ for Dehistan, −18.1‰ for West Pamir-Alay, −17.5‰ for Tashkent, −16.3‰ for Ferghana, and −13.3‰ for Khoresm (Fig. 2b; Information 4). Regional communities in Uzbekistan and Turkmenistan fall into in three trophic groups, each separated by ca. 2‰ in meanb δ15N (Fig. 2c).

A diachronic shift in the dietary intake of nomadic communities located in southern Kazakhstan is indicated by a significant 3‰ decrease in meanb δ13C values between early and late Otrar (Fig. 2d; Supplementary Information 4) and an increase of ca. 2‰ in meanb δ13C between early and late Zhetysu. However, meanb δ13C for late Zhetysu has a wide 95% CI between −15.9 and −12.3‰, which precludes a reliable estimate of the change. Overall, meanb δ15N for communities in southern Kazakhstan are more mutually similar to each other than that in southerly regions (Fig. 2c,e). Early Otrar displays meanb δ15N that is spaced apart by slightly less than ca. 1.7‰ from that of early Zhetysu, while in the later period the regions have identical meanb δ15N of ca 12‰.

Environment and isotopic variation

Isotopic patterns observed at individual archaeological sites are not driven by environmental inputs. Redundancy analysis between mean δ13C and δ15N values per cemetery site and 25 environmental parameters of ecologically relevant rainfall and temperature variations, in addition to elevation and soil properties, did not result in statistically significant relationships. Multiple linear regressions were also performed, which further failed to generate statistically significant relationships. (See Supplementary Information 5 for methods and results.)

Isotopic niche modelling

Central Asian medieval communities display isotopic niches that cluster in two distinct size ranges (Fig. 3; Supplementary Information 6). Group 1 exhibits small areas between 0.1 and 3.7‰2 for urban communities in Uzbekistan and Turkmenistan, as do nomadic communities in southern Kazakhstan from the early medieval period. Over the next several hundred years, dietary diversity among individuals in Zhetysu and the Otrar Oasis (Group 2) appears to have radically increased, with isotopic niches estimated between 2.7 and 15.7‰2. Notably, the greatest intra-community dietary diversity is visible in late Zhetysu, indicated by a 95% CI exceeding that from Group 1 communities, except early Otrar, which slightly overlaps by 0.7‰2.

Figure 3
Figure 3

Community-level dietary diversity of medieval humans represented by posterior distributions of core isotopic niche area (‰2) by sites and regions in Central Asia. Isotopic niches were calculated by fitting standard ellipses to cover ca. 39% of the δ13C and δ15N data points using Bayesian inference. Black dots indicate area means, and the shaded boxes, from dark to light, represent the 50%, 75%, and 95% credible intervals.

While all narrow, modelled isotopic niches for medieval urbanites in Uzbekistan are highly unique in orientation and position in δ-space (Fig. 4a). For late Zhetysu and Otrar, the diffuse spacing of isotopic values indicates exceedingly varied dietary intake. Individuals in late Zhetysu, which represent two archaeological sites, fall within the isotopic niches for Khoresm, Ferghana and Tashkent. Likewise, individuals in late Otrar, which represent one site, span all four urban isotopic niches from Uzbekistan. In contrast, there is substantial proportional overlap between the small isotopic niches from early Zhetysu and Ferghana and also early Otrar and Khoresm, respectively (95% CI: 0.13–0.64 and 0.17–0.69; Supplementary Fig. S6), which indicates a high likelihood of dietary parity.

Figure 4
Figure 4

(a) Medieval urban isotopic niches from Uzbekistan are displayed as probability clouds, and individual isotopic values from southern Kazakhstan (nomadic communities) and western Turkmenistan (urban community) are represented as points. (b) Isotopic niche overlap analysis for urban communities in Uzbekistan. Standard ellipses covering 95% of δ13C and δ15N values were modelled using Bayesian inference. Overlapping areas for each pairwise comparisons in δ-space were visualized as probability clouds with underlying isotopic data points superimposed (lower left). Area overlap of total isotopic niche area for each pair was plotted as probability distributions (upper right).

Isotopic niches from urban communities in Uzbekistan illustrate strong dissimilarity from each other (Fig. 4b). The highest proportion of niche overlap occurs between West Pamir-Alay and Ferghana (95% CI: 0.05–0.36) and between West Pamir-Alay and Tashkent (95% CI: 0.01–0.32), suggesting that a maximum of one-third of individuals in these communities had similar dietary intake. However, this niche overlap could be as low as 1–5%. The remaining pairwise comparisons of urban communities show negligible occurrence of overlap (Supplementary Information 7).


Taken together, urban and nomadic communities display distinctive but wide-ranging isotopic values that are strongly suggestive of diverse dietary intake across medieval Central Asia. A large overall range of human δ13C values from ca. −20 to −11‰ is likely due to individuals consuming mostly C3 crops (wheat, barley, and rice) or C4 millets at sustained intensities. Among Eurasian cultigens, millets are isotopically distinct with high 13C concentration. Millets thrive in hot and arid climate and exhibit fast-growing and drought-tolerant adaptations69, traits which would have provided Central Asian farmers and mobile pastoralists opportunities for low risk, low-investment cultivation in marginal agricultural areas70,71. The natural abundance of C4 vegetation is substantially higher in the desert zones of Central Asia50, where livestock, as part of urban or nomadic subsistence, could have accessed enriched δ13C biomass and provided human consumers with protein-dense foods (meat and milk) with high δ13C values. A lack of correlations between human isotopic values and site environmental parameters suggests that dietary intake across the region was shaped primarily by food choice. Medieval agriculture in Central Asia, which used complex crop schedules and large irrigation works71,72,73,74,75,76, likely enabled productivity to overcome environmental constraints on crop variety in order to meet the inter-connected dietary demands of consumers.

Modelled isotopic niches indicate that nomadic communities exploited a wide variety of dietary resources, while urban communities engaged in more limited dietary repertoires. The small isotopic niche sizes documented among urban communities in Uzbekistan and Turkmenistan suggest food channels that were shaped via consistent and insular dietary connectivity. On the other hand, large isotopic niches in southern Kazakhstan from a relatively contemporaneous period indicate that a multitude of food acquisition strategies were in use, by which communities tied more closely with pastoral nomadic lifeways comprised individuals with assorted dietary relationships that led to sustained differences in isotopic variability. The dietary connectivity for these nomadic groups may have fostered group partitioning through unsynchronized food interactions among different community members. The small isotopic niches for nomadic communities in early Otrar and Zhetysu, which in this case substantially overlap with that for Khoresm and Ferghana, suggesting more restricted dietary intake by pastoralists, emphasize the subsistence plasticity of pastoral nomads who readily contour their own food production and interaction networks in response to dynamic social and natural landscapes24,26,77.

The diachronic shift in isotopic niche size between early and late nomadic communities also highlights two distinct scales of dietary variability that illustrate the importance of multi-resource pastoralism to Silk Road interactions. In southern Kazakhstan, the early medieval period is marked by a growth of urban centres, villages, and agricultural economies78,79,80,81, which is also historically associated with frequent conflict among nomadic confederacies that instigated socio-political turmoil1,82. In order to mitigate risk and take advantage of economic opportunities presented by these newly founded centres, nomadic communities likely participated in coordinated subsistence interactions with settled populations over short distances, which would have effectively limited access to diverse food resources and thus narrowed their dietary breadth. During the strengthening of Turkic empires several centuries later1,3,31,83, Silk Road trans-regional trading expanded to include bulk commodities and raw materials32, and nomadic communities across southern Kazakhstan expressed wide dietary breadth, as indicated by large isotopic niche sizes.

One explanation for greater inter-individual dietary diversity during this later medieval period is that nomadic communities tapped into growing trade economies as agents of food exchange and broke out of insular urban subsistence channels. Recent excavations of nomadic encampments in the foothill zones of West Pamir-Alay and Zhetysu illustrate highly variable levels of economic interaction between pastoralists and urban centres. At these sites, cultural materials associated with 8th–13th c. radiocarbon chronologies include cotton fabrics84 and variable mixtures of ceramics ranging between standardized wheel-spun food storage vessels from distant oases communities and locally produced ‘handmade’ coarsewares, which are rare in urban contexts85,86. The presence of hybrid ceramic assemblages and cotton, a woven trade good associated with oasis production centers87,88, suggests that complex and non-uniform relationships with urban economies coincided with intra-group dietary diversity in mobile pastoralist communities25.

Accordingly, at the community level, a second scale of dietary variability that is representative of multi-resource pastoralism is observed at late Otrar and Zhetysu, which display wide ranges of δ13C values (Fig. 4a). Some individuals in late Zhetysu have δ13C values similar to those commonly observed in humans from prehistoric millet-based farming societies in China, where millets were domesticated89,90,91, while individuals in late Zhetysu and Otrar had low δ13C values, typical of Neolithic and early Bronze Age humans before millet spread to the C3-dominant Central Asian steppe92. Both communities in late Otrar and Zhetysu display similar δ15N distributions, though individuals exhibit differences in dietary intake of relative proportions of meat and dairy products (Fig. 2e). Together, these findings suggest that dietary connectivity at the steppe margins was associated with an ecumene of diverse food consumption, in which individuals maintained separate subsistence strategies as they simultaneously participated in a common nomadic ethos. Compared to urbanites, mobile pastoralists likely maintained closer control of food production and distribution, allowing them to eat according to food preferences, which may have been less important for maintaining social ties than in urban contexts.

The combination of narrow dietary niches with isotopic distinction in human remains from medieval urban communities south of the Syr-Darya river is due to trophic-level variation in the intensity of meat and cereal consumption between communities as well as differences in the contribution of millet to urban diets, either eaten directly or indirectly obtained from meat and dairy of animals foddered with millet. Isotopic niche overlap was highest for urban communities in close geographic proximity to one another, a pattern that suggests neighbouring communities either participated in similar food traditions or agricultural practices that were confined to small catchments. These communities may have participated in limited inter-regional trading of staple foods, which likely would have moved through established subsistence channels as if locally produced. There is also the possibility that dietary connectivity in urban contexts was subject to bureaucratic intermediaries, which exerted influence through land tenure and taxation1,32,93. Alternatively, in the absence of top-down control, food exchange networks may have steadily channelled provisions to urbanites as a reflection of other economic networks that inevitably developed to be streamlined towards cultural insiders in cosmopolitan contexts.

While the consensus among historians and archaeologists is that urbanites in medieval Central Asia dwelled in rich multicultural settings5,7,16, there appears to be a limit to this diversity in dietary intake as revealed through isotopic niche modelling. Distinctions in food choice and diet between urban communities suggest regional food repertoires were narrowly circumscribed, at least between C3 and C4 crops and also between animal and plant protein. Regional patterns in diet imply that cultural differences surrounding foods may have been surprisingly diminished within urban communities, which runs counter to the notion of collective cosmopolitanism in medieval Central Asia. Medieval urbanites in Central Asia maintained inward-focused dietary connectivity that likely generated a localized social cohesion through culturally integrated supply chains for consumers.

Scholars also associate Silk Road activity with early globalization processes7,8, in which urban centres are viewed as the main drivers of cultural influence and outward economic connectivity, while ‘nomads’ are interpreted as antagonists to ancient civilization94,95,96. Yet, through multi-resource subsistence strategies, nomadic communities likely wielded flexible economic engagements that traversed open landscapes of contact with people who facilitated far-reaching connectivity. In this sense, nomadic individuals may have been more culturally interoperable and able to participate in, disengage from, and influence cultural spheres more easily than urban populations. Indeed, many of the pan-regional turnovers in religion, language, and political authority that resulted in changes in architecture, technologies, and other commodity classes in the medieval period are historically described as nomadic innovations1,2,31, and essential routes that connected Silk Road sites in the highland regions of Central Asia were likely shaped by nomadic mobility15. This study takes a new step toward resolving the complex interplay between urban and nomadic societies that are rarely available through archaeological datasets. Establishing dietary diversity provides an emerging understanding of food and connectivity along Central Asia’s Silk Roads that highlights the significance of ancient nomadic pastoralists in bridging seemingly insulated urban centres.

Materials and Methods

Human remains

We performed new analyses on human remains from Uzbekistan and Kazakhstan, which were selected based on 1) medieval chronology from ca. 8th–13th c., 2) information on archaeological context, and 3) minimum age estimation of young adult. The majority of human remains analysed from Uzbekistan (n = 38) were excavated at various times over the past 80 years and do not include associated post-cranial elements. Human remains from Tashbulak (n = 4) were excavated in 2015 and are represented by complete inventories of skeletal elements. Human remains from Uzbekistan are stored in the Institute of Archaeology of the Uzbek Academy of Sciences in Samarkand, under the auspices of the Archaeology of the Qarakhanids Project (Co-PIs: Farhod Maksudov and Michael Frachetti). Human remains from Karatal (n = 3) were excavated in 2006 and 2008 and include incomplete skeletal elements due to ancient burial looting and modern erosion (Supplementary Information 3). Human remains from Karatal are stored at the Central State Museum of Kazakhstan in Almaty, under the auspices of the Dzhungar Mountain Archaeology Project (Co-PIs: Alexei Mar’yahshev and Michael Frachetti). A list of human samples is provided in Supplementary Tables S2S4.

Isotopic analysis

Approximately 1–2 cm3 of sample were cut from dense cortical bone. Collagen extraction was performed in the Archaeological Stable Isotope Laboratory of Kiel University following Tuross et al.97. Mass spectrometry for δ13C and δ15N was performed at the Boston University Stable Isotope Laboratory using a EuroVector Euro EA elemental analyser coupled with a GVI IsoPrime in continuous flow mode with an analytical error of 0.1‰ and 0.2‰ for δ13C and δ15N, respectively. Isotopic values are reported in permil (‰) relative to the Vienna Pee Dee Belemnite (VPDB) standard for δ13C and atmospheric nitrogen (AIR) for δ15N. Collagen samples with an elemental C:N ratio less than 2.9 or greater than 3.6 were considered diagenically altered and unsuitable for inclusion98,99,100. Failed samples (n = 3) are reported in Supplementary Table S5.

Previously published δ13C and δ15N values of human bone collagen from southern Central Asia were analysed (Supplementary Tables S3-S4). Isotopic data from medieval southern Kazakhstan included 13 samples from the Otrar Oasis (3rd–7th c.) and 13 samples from southern Zhetysu (2nd–12th c.)101. Data from the Dehistan Plain in western Turkmenistan included two samples from the Iron Age (ca. 1300 BC) and one sample from the medieval period (ca. 11th–12th c.)102. These data were lumped together irrespective of chronology due to low sample size for the region and similar isotopic values, which were unique in our dataset.

Human collagen is δ13C enriched by ca. 1–3‰ relative to the carbon isotope composition of consumed foods103, whereas herbivores are enriched in δ13C by ca. 5‰ relative to consumed vegetation104,105. Human nitrogen isotope values reflect the intensity of cereal versus meat consumption, as there is a 2–5‰ trophic enrichment in δ15N with each step up in the food web40,106,107, but also reflect variation in food production systems that introduce exogenous nitrogen, usually in the form of manure, to the floral base of the food web, potentially imparting considerable nitrogen isotopic variation in agricultural and livestock food62,63,65,108,109. In attempts to estimate the relative contribution of specific foods in diets, a common approach in archaeological dietary studies is to sample ancient and modern plants and animals to assess possible isotopic variability in food sources that result from various environmental and anthropogenic factors110. For the purposes of establishing human dietary diversity at the community level, however, knowledge of the isotopic content of ancient food remains is unnecessary.

Statistical analysis

All statistical analyses were performed using R, version 3.4.0111. Due to low sample sizes per site (1 ≤ n ≤ 11), δ13C and δ15N values were analysed by geographic region (3 ≤ n ≤ 15) using Bayesian techniques to quantify uncertainty and overcome issues with non-parametric data distributions, which cannot be reliably analysed with frequentist methods. The means of isotopic values for each region were calculated with a Bayesian bootstrapping method at 5000 iterations using the package bayesboot112. Posterior distributions of means for each group were considered statistically different from one another if the 95% credible intervals (CI) did not overlap.

Isotopic niche modelling

Isotopic niches were analysed using SIBER (Stable Isotope Bayesian Ellipses in R), version 2.1.347. The Markov chain Monte Carlo simulation was run with uniform priors 2,000,000 times, with the first 10,000 results discarded (burn-in), followed by a 1:100 thinning. The fitted ellipses express a posterior probability distribution of area (‰2) and position in δ-space. This technique is statistically advantageous for analysing communities of consumers, as fitted ellipses are less sensitive to sample size than other spatial metrics, such as convex hulls47, and uncertainty is factored into estimates, such as sample size, which has previously challenged studies using a finite number of specimens from the archaeological record. Isotopic niches were compared among regional groups by using the Bayesian standard ellipse areas (SEAb) and the proportion of SEAb overlap as total SEAb pairwise for each region. Overlapping areas of isotopic niches indicate similarity in isotopic inputs and thus comparable resource exploitation113,114,115.

Environmental modelling and GIS

Interactive effects between human stable isotope values and environmental parameters at each site in 10 km and 50 km spatial buffers were explored using redundancy analysis (RDA) with the R package vegan, version 2.4–5116. Environmental parameters included elevation, bioclim data117, which are derived from monthly temperature and rainfall values from 1970–2000, and soil parameters obtained from the Soil Grid Project118. (See Supplementary Information 5 for RDA methods and results.) Data were mapped using Quantum GIS, version 2.18.2119. Base imagery in Fig. 1 was obtained from Natural Earth (

Additional information

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


  1. 1.

    Barisitz, S. Central Asia and the Silk Road: Economic Rise and Decline over Several Millennia. (Springer, 2017).

  2. 2.

    Beckwith, C. Empires of the Silk Road a History of Central Eurasia from the Bronze Age to the Present. (Princeton University Press, 2009).

  3. 3.

    Biran, M. The Qarakhanids’ Eastern Exchange: Preliminary Notes on the Silk Roads in the Eleventh and Twelfth Centuries. In Complexity of Interaction along the Eurasian Steppe Zone in the First Millennium CE (eds Bemmann, J. & Schmauder, M.) 7, 575–595 (Vor- und Frühgeschichtliche Archäologie Rheinische Friedrich-Wilhelms-Universität Bonn, 2015).

  4. 4.

    Rezakhani, K. The Road That Never Was: The Silk Road and Trans- Eurasian Exchange. Comparative Studies of South Asia, Africa, and the Middle East 30, 420–433, (2010).

  5. 5.

    Hansen, V. The Silk Road: A New History. (Oxford University Press, 2012).

  6. 6.

    Starr, S. F. Lost Enlightenment: Central Asia’s Golden Age from the Arab Conquest to Tamerlane. (Princeton University Press, 2013).

  7. 7.

    Foltz, R. Religions of the Silk Road: Premodern Patterns of Globalization. (Palgrave Macmillan, 2010).

  8. 8.

    Grataloup, C. Géohistoire de la mondialisation: Le temps long du monde. (Armand Colin, 2010).

  9. 9.

    Allsen, T. T. Commodity and Exchange in the Mongol Empire: A Cultural History of Islamic Textiles. (Cambridge University Press, 1997).

  10. 10.

    Anarbaev, A. Srednevekovoe zhilische Akhsikenta (XI-XII vv.). In Istoriya Material’noj Kul’tury Uzbekistana 35 220–229 (Izd-vo ‘FAN’, 2006).

  11. 11.

    Belennitskij, A. M., Bentovich, I. B. & Bol’shakov, O. G. Srednebekovyj gorod Srednej Azii. (Izd-vo ‘Nauka’, 1973).

  12. 12.

    Borell, B. Ancient Glass from the Silk Road. SPAFA Journal (Old series 1991–2013) 20, (2010).

  13. 13.

    Ivanov, G. Excavations at Kuva (Ferghana Valley, Uzbekistan). Iran 41, 205–216 (2003).

  14. 14.

    Rehren, T. & Papakhristu, O. Cutting edge technology: the Ferghana Process of medieval crucible steel smelting. Metalla 7, 55–69 (2000).

  15. 15.

    Frachetti, M. D., Smith, C. E., Traub, C. M. & Williams, T. Nomadic ecology shaped the highland geography of Asia’s Silk Roads. Nature 543, 193–198, (2017).

  16. 16.

    Liu, X. The Silk Road in World History. (Oxford University Press, 2010).

  17. 17.

    Mills, B. J. Identity, Feasting, and the Archaeology of the Greater Southwest. (University Press of Colorado, 2004).

  18. 18.

    Mintz, S. W. & Bois, C. M. D. The Anthropology of Food and Eating. Annual Review of Anthropology 31, 99–119, (2002).

  19. 19.

    Pearson, M. P. Food, Culture and Identity in the Neolithic and Early Bronze Age. (BAR International Series, 2003).

  20. 20.

    Smith, M. L. The Archaeology of Food Preference. American Anthropologist 108, 480–493, (2006).

  21. 21.

    Stone, D. J. The Consumption of Field Crops in Late Medieval England. In Food in medieval England: diet and nutrition (eds. Woolgar, C. M., Serjeantson, D. & Waldron, T.) 11–26 (Oxford University Press, 2006).

  22. 22.

    Spengler, R. N., Frachetti, M. D. & Fritz, G. J. Ecotopes and Herd Foraging Practices In the Steppe/Mountain Ecotone of Central Asia During the Bronze and Iron Ages. Journal of Ethnobiology 33, 125–147, (2013).

  23. 23.

    Spengler, R. N. Niche Dwelling vs. Niche Construction: Landscape Modification in the Bronze and Iron Ages of Central Asia. Human Ecology 42, 813–821, (2014).

  24. 24.

    Frachetti, M. D. Pastoralist Landscapes and Social Interaction in Bronze Age Eurasia. (University of California Press, 2008).

  25. 25.

    Frachetti, M. D. Multiregional Emergence of Mobile Pastoralism and Nonuniform Institutional Complexity across Eurasia. Current Anthropology 53, 2–38, (2012).

  26. 26.

    Honeychurch, W. & Amartuvshin, C. Hinterlands, Urban Centers, and Mobile Settings: The ‘New’ Old World Archaeology from the Eurasian Steppe. Asian Perspectives 46, 36–64, (2007).

  27. 27.

    Rouse, L. M. & Cerasetti, B. Ojakly: A Late Bronze Age mobile pastoralist site in the Murghab Region, Turkmenistan. Journal of Field Archaeology 39, 32–50, (2014).

  28. 28.

    Salzman, P. C. Multi-Resource Nomadism in Iranian Baluchistan. Journal of Asian and African Studies 7, 60–68 (1972).

  29. 29.

    Glick, T. F., Livesey, S. & Wallis, F. Medieval Science, Technology, and Medicine: An Encyclopedia. (Routledge, 2014).

  30. 30.

    Trépanier, N. Foodways and Daily Life in Medieval Anatolia: A New Social History. (University of Texas Press, 2014).

  31. 31.

    Golden, P. The Karakhanids and early Islam. In The Cambridge History of Early Inner Asia (ed. Sinor, D.) 343–370 (Cambridge University Press, 1990).

  32. 32.

    Negmatov. The Samanid State. In History of civilizations of Central Asia, Volume 4 (Part 1) (eds Asimov, M. E. & Bosworth, C. E.) 77–94 (Unesco Publishing, 1998).

  33. 33.

    Brykina, G. A. Karabulak. (Izd-vo ‘Nauka’, 1974).

  34. 34.

    Buryakov, Y. F. Genezis i etapi razvitiya gorodskoj kul’tury Tashkentskogo oazisa. (Izd-vo Nauka Uzbekskoj SSR, 1982).

  35. 35.

    Yakubov, Y. Pargar v VII-VIII vekakh nashej ery. (Izd-vo ‘Donish’, 1979).

  36. 36.

    Kuran, T. & Sandholm, W. H. Cultural Integration and Its Discontents. The Review of Economic Studies 75, 201–228 (2008).

  37. 37.

    Kusaka, S. et al. Homogeneous diet of contemporary Japanese inferred from stable isotope ratios of hair. Scientific Reports 6, 33122, (2016).

  38. 38.

    Phillips, L. Food and Globalization. Annu. Rev. Anthropol. 35, 37–57, (2006).

  39. 39.

    DeNiro, M. J. & Epstein, S. Influence of diet on the distribution of nitrogen isotopes in animals. Geochimica et Cosmochimica Acta 45, 341–351, (1981).

  40. 40.

    Schoeninger, M. J. & DeNiro, M. J. Nitrogen and carbon isotopic composition of bone collagen from marine and terrestrial animals. Geochimica et Cosmochimica Acta 48, 625–639, (1984).

  41. 41.

    Hedges, R. E. M., Clement, J. G., Thomas, C. D. L. & O’Connell, T. C. Collagen turnover in the adult femoral mid-shaft: Modeled from anthropogenic radiocarbon tracer measurements. Am. J. Phys. Anthropol. 133, 808–816, (2007).

  42. 42.

    Szulc, P., Seeman, E. & Delmas, P. D. Biochemical Measurements of Bone Turnover in Children and Adolescents. Osteoporos Int 11, 281–294, (2000).

  43. 43.

    Tsutaya, T. & Yoneda, M. Quantitative Reconstruction of Weaning Ages in Archaeological Human Populations Using Bone Collagen Nitrogen Isotope Ratios and Approximate Bayesian Computation. PLOS ONE 8, e72327, (2013).

  44. 44.

    Layman, C. A. et al. Applying stable isotopes to examine food-web structure: an overview of analytical tools. Biological Reviews 87, 545–562, (2012).

  45. 45.

    Newsome, S. D., M del Rio, C., Bearhop, S. & Phillips, D. L. A niche for isotopic ecology. Frontiers in Ecology and the Environment 5, 429–436, (2007).

  46. 46.

    Bearhop, S., Adams, C. E., Waldron, S., Fuller, R. A. & MacLeod, H. Determining trophic niche width: a novel approach using stable isotope analysis. Journal of Animal Ecology 73, 1007–1012, (2004).

  47. 47.

    Jackson, A. L., Inger, R., Parnell, A. C. & Bearhop, S. Comparing isotopic niche widths among and within communities: SIBER – Stable Isotope Bayesian Ellipses in R. Journal of Animal Ecology 80, 595–602, (2011).

  48. 48.

    Akhani, H., Trimborn, P. & Ziegler, H. Photosynthetic pathways in Chenopodiaceae from Africa, Asia and Europe with their ecological, phytogeographical and taxonomical importance. Pl Syst Evol 206, 187–221, (1997).

  49. 49.

    Pyankov, V., Voznesenskaya, E., Kondratschuk, A. & Black, C. A comparative anatomical and biochemical analysis in salsola (Chenopodiaceae) species with and without a Kranz type leaf anatomy: a possible reversion of C4 to C3 photosynthesis. American Journal of Botany 84, 597–597, (1997).

  50. 50.

    Toderich, K. et al. C3/C4 plants in the vegetation of Central Asia, geographical distribution and environmental adaptation in relation to climate. In Climate Change and Terrestrial Carbon Sequestration in Central Asia 33–63, (Taylor & Francis, 2007).

  51. 51.

    Winter, K. C4 plants of high biomass in arid regions of asia-occurrence of C4 photosynthesis in Chenopodiaceae and Polygonaceae from the Middle East and USSR. Oecologia 48, 100–106, (1981).

  52. 52.

    Suess, H. E. Radiocarbon Concentration in Modern Wood. Science 122, 415–417, (1955).

  53. 53.

    Keeling, C. D. The Suess effect: 13Carbon-14Carbon interrelations. Environment International 2, 229–300, (1979).

  54. 54.

    Farquhar, G. D., Ehleringer, J. R. & Hubick, K. T. Carbon Isotope Discrimination and Photosynthesis. Annual Review of Plant Physiology and Plant Molecular Biology 40, 503–537, (1989).

  55. 55.

    Marino, B. D. & McElroy, M. B. Isotopic composition of atmospheric CO2 inferred from carbon in C4 plant cellulose. Nature 349, 127–131, (1991).

  56. 56.

    O’Leary, M. H. Carbon Isotopes in Photosynthesis: Fractionation techniques may reveal new aspects of carbon dynamics in plants. BioScience 38, 328–336, (1988).

  57. 57.

    Tieszen, L. L. Natural Variations in the Carbon Isotope Values of Plants: Implications for Archaeology, Ecology, and Paleoecology. Journal of Archaeological Science 18, 227–248, (1991).

  58. 58.

    Watson, A. M. Agricultural Innovation in the Early Islamic World: The Diffusion of Crops and Farming Techniques, 700–1100. (Cambridge University Press, 1983).

  59. 59.

    Makarewicz, C. & Tuross, N. Foddering by Mongolian pastoralists is recorded in the stable carbon (δ13C) and nitrogen (δ15N) isotopes of caprine dentinal collagen. Journal of Archaeological Science 33, 862–870, (2006).

  60. 60.

    Shishlina, N., Sevastyanov, V. & Kuznetsova, O. Seasonal practices of prehistoric pastoralists from the south of the Russian plain based on the isotope data of modern and archaeological animal bones and plants. Journal of Archaeological Science: Reports (2017).

  61. 61.

    Ma, J.-Y., Sun, W., Liu, X.-N. & Chen, F.-H. Variation in the Stable Carbon and Nitrogen Isotope Composition of Plants and Soil along a Precipitation Gradient in Northern China. PLOS ONE 7, e51894, (2012).

  62. 62.

    Makarewicz, C. A. Winter pasturing practices and variable fodder provisioning detected in nitrogen (δ15N) and carbon (δ13C) isotopes in sheep dentinal collagen. Journal of Archaeological Science 41, 502–510, (2014).

  63. 63.

    Makarewicz, C. A. Winter is coming: seasonality of ancient pastoral nomadic practices revealed in the carbon (δ13C) and nitrogen (δ15N) isotopic record of Xiongnu caprines. Archaeological and Anthropological Sciences 9, 405–418, (2017).

  64. 64.

    Makarewicz, C. & Tuross, N. Finding Fodder and Tracking Transhumance: Isotopic Detection of Goat Domestication Processes in the Near East. Current Anthropology 53, 495–505, (2012).

  65. 65.

    Szpak, P. Complexities of nitrogen isotope biogeochemistry in plant-soil systems: implications for the study of ancient agricultural and animal management practices. Front Plant Sci 5, 1–19, (2014).

  66. 66.

    Schoeninger, M. J. & DeNiro, M. Stable nitrogen isotope ratios of bone collagen reflect marine and terrestrial components of prehistoric human diet. Science 220, 1381–1383, (1983).

  67. 67.

    Schwarcz, H. P., Melbye, J., Anne Katzenberg, M. & Knyf, M. Stable isotopes in human skeletons of Southern Ontario: reconstructing Palaeodiet. Journal of Archaeological Science 12, 187–206, (1985).

  68. 68.

    Svyatko, S. V., Reimer, P. J. & Schulting, R. Modern Freshwater Reservoir Offsets in the Eurasian Steppe: Implications for Archaeology. Radiocarbon 59, 1597–1607, (2017).

  69. 69.

    Nesbitt, M. G. In The Cultural History of Plants (eds Prance, G. & Nesbitt, M.) 45–60 (Routledge, 2005).

  70. 70.

    Brite, E. B., Kidd, F. J., Betts, A. & Negus Cleary, M. Millet cultivation in Central Asia: A response to Miller et al. The Holocene  27, 1415–1422, (2017).

  71. 71.

    Miller, N. F., Spengler, R. N. & Frachetti, M. Millet cultivation across Eurasia: Origins, spread, and the influence of seasonal climate. The Holocen26, 1566–1575, (2016).

  72. 72.

    Brite, E. B. Irrigation in the Khorezm oasis, past and present: a political ecology perspective. Journal of Political Ecology 23, 1–25, (2016).

  73. 73.

    Clarke, D., Sala, R., Deom, J.-M. & Meseth, E. Reconstructing irrigation at Otrar Oasis, Kazakhstan, AD 800–1700. Irrig. and Drain. 54, 375–388, (2005).

  74. 74.

    Groshev, V. A. Irrigatsiya yuzhnogo Kazakhstana v sredniye veka. (Izd-vo ‘Nauka’ KSSR, 1985).

  75. 75.

    Malatesta, L. C. et al. Dating the Irrigation System of the Samarkand Oasis: A Geoarchaeological Study. Radiocarbon 54, 91–105, (2012).

  76. 76.

    Adrianov, B. A. Ancient Irrigation Systems of the Aral Sea Area. (Oxbow Books, 2016).

  77. 77.

    Wright, J. & Makarewicz, C. Perceptions of pasture: the role of skill and networks in maintaining stable pastoral nomadic systems in Inner Asia. In Ancient Society and Climate (eds Kerner, S., Dann, R. & Jensen, P. B.) 267–288 (Museum Tusculanum Press, 2015).

  78. 78.

    Bashtannik, S. V. Archaeobotanical Studies at Medieval Sites in the Arys River Valley. Archaeology, Ethnology and Anthropology of Eurasia 33, 85–92, (2008).

  79. 79.

    Goryachev, A. A. Arkheologicheskij kompleks Turgen’. Evolyutsiya drevnikh kul’tur. in Archeologiya Kazakhstana v Epochu Nezavisimosti: Itogi, Perspektivy (ed. Bajtanaev, B. A.) 1, 256–266 (Institut arkheologii im A.Kh. Margulana, 2011).

  80. 80.

    Goryachev, A. A., Caraev, V. V. & Egorova, T. A. K voprosu o khozyajstvenno-kul’turnom pazvitii drevnego naseleniya Almaty. Tsentral’no-Aziatskij Iskusstvovedcheskij Zhurnal 3, 18–30 (2016).

  81. 81.

    Baipakov, K. M., Smagulov, E. A. & Erzhigitova, A. A. Rannesrednevekovye Nekropoli Yuzhnogo Kazakhstana. (BAUR, 2005).

  82. 82.

    Bartold, V. V. Ocherk istorii Semirech’ ya. (Kigizgosizdat, 1943).

  83. 83.

    Davidovich, E. A. The Karakhanids. In History of civilizations of Central Asia, Volume 4 (Part 1) (eds Asimov, M. E. & Bosworth, C. E.) 125–149 (Unesco Publishing, 1998).

  84. 84.

    Doumani Dupuy, P. N., Spengler, R. N III. & Frachetti, M. D. Eurasian textiles: Case studies in exchange during the incipient and later Silk Road periods. Quaternary International (2017).

  85. 85.

    Frachetti, M. D., Benecke, N., Mar’yashev, A. N. & Doumani, P. N. Eurasian pastoralists and their shifting regional interactions at the steppe margin: settlement history at Mukri, Kazakhstan. World Archaeology 42, 622–646, (2010).

  86. 86.

    Frachetti, M. D. & Maksudov, F. The landscape of ancient mobile pastoralism in the highlands of southeastern Uzbekistan, 2000 B.C.–A.D. 1400. Journal of Field Archaeology 39, 195–212, (2014).

  87. 87.

    Bartold, V. V. Four studies on the history of Central Asia. (E.J. Brill, 1962).

  88. 88.

    Brite, E. B. & Marston, J. M. Environmental change, agricultural innovation, and the spread of cotton agriculture in the Old World. Journal of Anthropological Archaeology 32, 39–53, (2013).

  89. 89.

    Dai, L. et al. Cattle and sheep raising and millet growing in the Longshan age in central China: Stable isotope investigation at the Xinzhai site. Quaternary International 426, 145–157, (2016).

  90. 90.

    Ma, M. M. et al. Stable Isotope Analysis of Human and Faunal Remains in the Western Loess Plateau, Approximately 2000 cal BC. Archaeometry 56, 237–255, (2014).

  91. 91.

    Pechenkina, E. A., Ambrose, S. H., Xiaolin, M. & Benfer, R. A. Jr. Reconstructing northern Chinese Neolithic subsistence practices by isotopic analysis. Journal of Archaeological Science 32, 1176–1189, (2005).

  92. 92.

    Motuzaite-Matuzeviciute, G. et al. Climatic or dietary change? Stable isotope analysis of Neolithic–Bronze Age populations from the Upper Ob and Tobol River basins. The Holocene 26, 1711–1721, (2016).

  93. 93.

    Golden, P. B. An Introduction to the History of the Turkic Peoples: Ethnogenesis and State-Formation in Medieval and Early Modern Eurasia and the Middle East. (Otto Harrassowitz, 1992).

  94. 94.

    Di Cosmo, N. Ancient China and its Enemies: The Rise of Nomadic Power in East Asian History. (Cambridge University Press, 2002).

  95. 95.

    Rogers, D. J. The Contingencies of State Formation in Eastern Inner Asia. Asian Perspectives 46, 249–274 (2007).

  96. 96.

    Sneath, D. The Headless State: Aristocratic Orders, Kinship Society, and Misrepresentations of Nomadic Inner Asia. (Columbia University Press, 2007).

  97. 97.

    Tuross, N., Fogel, M. L. & Hare, P. E. Variability in the preservation of the isotopic composition of collagen from fossil bone. Geochimica et Cosmochimica Acta 52, 929–935, (1988).

  98. 98.

    Ambrose, S. H. Preparation and characterization of bone and tooth collagen for isotopic analysis. Journal of Archaeological Science 17, 431–451, (1990).

  99. 99.

    DeNiro, M. J. Postmortem preservation and alteration of in vivo bone collagen isotope ratios in relation to palaeodietary reconstruction. Nature 317, 806–809, (1985).

  100. 100.

    DeNiro, M. J. & Weiner, S. Chemical, enzymatic and spectroscopic characterization of “collagen” and other organic fractions from prehistoric bones. Geochimica et Cosmochimica Acta 52, 2197–2206, (1988).

  101. 101.

    Motuzaite Matuzeviciute, G. et al. The extent of cereal cultivation among the Bronze Age to Turkic period societies of Kazakhstan determined using stable isotope analysis of bone collagen. Journal of Archaeological Science 59, 23–34, (2015).

  102. 102.

    Bocherens, H., Mashkour, M., Drucker, D. G., Moussa, I. & Billiou, D. Stable isotope evidence for palaeodiets in southern Turkmenistan during Historical period and Iron Age. Journal of Archaeological Science 33, 253–264, (2006).

  103. 103.

    Kellner, C. M. & Schoeninger, M. J. A simple carbon isotope model for reconstructing prehistoric human diet. Am. J. Phys. Anthropol. 133, 1112–1127, (2007).

  104. 104.

    Lee-Thorp, J. A., Sealy, J. C. & van der Merwe, N. J. Stable carbon isotope ratio differences between bone collagen and bone apatite, and their relationship to diet. Journal of Archaeological Science 16, 585–599, (1989).

  105. 105.

    Passey, B. H. et al. Carbon isotope fractionation between diet, breath CO2, and bioapatite in different mammals. Journal of Archaeological Science 32, 1459–1470, (2005).

  106. 106.

    Minagawa, M. & Wada, E. Stepwise enrichment of 15N along food chains: further evidence and the relation between δ15N and animal age. Geochimica et cosmochimica acta 48, 1135–1140, (1984).

  107. 107.

    O’Connell, T., Kneale, C., Tasevska, N. & Kuhnle, G. The diet-body offset in human nitrogen isotopic values: A controlled dietary study. Am J Phys Anthropol 149, 426–434, (2012).

  108. 108.

    Bogaard, A. et al. Combining functional weed ecology and crop stable isotope ratios to identify cultivation intensity: a comparison of cereal production regimes in Haute Provence, France and Asturias, Spain. Veget Hist Archaeobot 25, 57-73, (2015).

  109. 109.

    Fraser, R. A. et al. Manuring and stable nitrogen isotope ratios in cereals and pulses: towards a new archaeobotanical approach to the inference of land use and dietary practices. Journal of Archaeological Science 38, 2790–2804, (2011).

  110. 110.

    Ventresca Miller, A. & Makarewicz, C. A. Isotopic approaches to pastoralism in prehistory: Diet, mobility, and isotopic reference sets. In Isotopic Investigations of Pastoralism in Prehistory (eds. Ventresca Miller, A. & Makarewicz, C. A.) 1–14 (Routledge, 2018).

  111. 111.

    R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria Available at: (2017).

  112. 112.

    Bååth, R. bayesboot: An Implementation of Rubin’s (1981) Bayesian Bootstrap. R package version 0.2.1 Available at: (2016).

  113. 113.

    Flaherty, E. A. & Ben-David, M. Overlap and partitioning of the ecological and isotopic niches. Oikos 119, 1409–1416, (2010).

  114. 114.

    Jackson, M. C. et al. Population-Level Metrics of Trophic Structure Based on Stable Isotopes and Their Application to Invasion Ecology. PLOS ONE 7, e31757, (2012).

  115. 115.

    Rossman, S., Ostrom, P. H., Gordon, F. & Zipkin, E. F. Beyond carbon and nitrogen: guidelines for estimating three-dimensional isotopic niche space. Ecol Evol 6, 2405–2413, (2016).

  116. 116.

    Oksanen, J. et al. vegan: Community Ecology Package. Available at: (2017).

  117. 117.

    Fick, S. E. & Hijmans, R. J. WorldClim 2: new 1-km spatial resolution climate surfaces for global land areas. Int. J. Climatol 37, 4302–4315, (2017).

  118. 118.

    Hengl, T. et al. SoilGrids250m: Global gridded soil information based on machine learning. PLOS ONE 12, e0169748, (2017).

  119. 119.

    Quantum GIS Development Team. Quantum GIS Geographic Information System. Open Source Geospatial Foundation Project Available at: (2016).

  120. 120.

    Maksudov, F. et al. Nomadic Urbanism at Tashbulak: A New Highland Town of the Qarakhanids. In Central Asian Urbanism (eds Baumer, C. & Ecklin, S.) (in press).

  121. 121.

    Lunina, S. B. Goroda Yuzhnogo Sogda v VIII-XII vv. (Izd-vo ‘FAN’, 1984).

  122. 122.

    Grenet, F. Zoroastrianism in Central Asia. In The Wiley Blackwell Companion to Zoroastrianism (eds Stausberg, M. & Vevaina, Y. S.-D.) 129–146 (Wiley Blackwell, 2015).

  123. 123.

    Grigor’ev, G. V. Zoroastrijskoe kostekhranilische v kishlake Frinkent pod g. Samarkandom. Vestnik drevnej istorii 2, 144–150 (1939).

  124. 124.

    Abdulgazieva, B. Issledovanie poseleniya Chordona. In Istoriya material’noj kul’tury Uzbekistana 25, 132–137 (Izd-vo ‘FAN’, 1991).

  125. 125.

    Abdullaev, B. M., Ivanov, G. P. & Matbabaev, B. K. Arkheologicheskiye raboty v g. Andizhane. In Arkheologicheskye issledovaniya v Uzbekistanie- 2000 god 18–25 (Institut arkheologii, 2001).

  126. 126.

    Khodzhajov, T. K. Ethnicheskie Protsessy v Srednej Azii v Epokhu Srednevekov’ ya. (Izd-vo Nauka Uzbekskoj SSR, 1987).

  127. 127.

    Buryakov, Y. F., Rostovtsev, O. M., Perevozchikov, I. V., Khodzhajov, T. K. & Khalilov, K. Mogil’nik Uturlik-Tepe. In Uspekhi Sredneaziatskoj Arkheologii (ed. Bochever, V. T.) 4, 91 (Izd-vo ‘Nauka’, 1979).

  128. 128.

    Gudovka, A. V. Tok-kala. (Izd-vo Nauka Uzbekskoj SSR, 1964).

  129. 129.

    Smagulov, E. A. & Erzhigitova, A. A. K izucheniyu pogrebal’nykh sooruzhenij Otyrarskogo oazisa. In Voprosy Arkheologii Kazakhstana (ed. Bejsenov, A. Z.) 142–164 (Institut arkheologii im A.Kh. Margulana, 2011).

  130. 130.

    Goryachev, A. A. Arkheologicheskie pamyatniki kompleksa Butakty I na yugo-vostochnoj okraine goroda Almaty. Izvestiya NAN RK 1, 45–59 (2006).

  131. 131.

    Goryachev, A. A. & Motov, Y. A. Rezul’taty issledovanij arkheologicheskogo kompleksa Butakty-I na yugo-vostochnoj okranie goroda Almaty v 2007 gody. Izvestiya NAN RK 1, 67–82 (2008).

  132. 132.

    Atagarryev, E. Srednevekobyj Dekhistan. (Nauka, 1986).

  133. 133.

    Mashkour, M. The Subsistence Economy in the Rural Community of Geoktchik Depe in Southern Turkmenistan: Preliminary Results of the Faunal Analysis. In Archaeozoology of the Near East III, Proceedings of the Third International Symposium on the Archaeozoology of Southwestern Asia and Adjacent Areas (eds Buitenhuis, H., Bartosiewicz, L. & Choyke, A. M.) 200–220 (ARC, 1998).

Download references


We especially thank Amridin Berdimuradov, the director of the Institute of Archaeology in Samarkand (National Academy of Sciences of Uzbekistan) for his continued support and enthusiasm for this project and international collaborations in general. We also thank Dima Voyakin for providing access to archaeological reports on Temirlanovka and for his help facilitating research in Kazakhstan. Recognition is also due to Marc Meyer and Mellissa Murphy for providing osteological data for the human individuals excavated from the Karatal cemetery complex. A special thank you goes to Malte Rühlemann for assisting with the redundancy analysis. Funding for this research was provided by the Graduate School ‘Human Development in Landscapes’ at Kiel University for the doctoral research of Taylor Hermes. Research at Tashbulak was funded by the National Geographic Society, the Max van Berchem Foundation, and Washington University in St. Louis. We are grateful for support from the Open Access Publikationsfonds of Land Schleswig-Holstein. Finally, we thank two anonymous reviewers for their suggestions on improving this paper.

Author information


  1. Institute for Prehistoric and Protohistoric Archaeology, Christian-Albrechts-Universität zu Kiel, Johanna-Mestorf-Straße 2-6, 24118, Kiel, Germany

    • Taylor R. Hermes
    •  & Cheryl A. Makarewicz
  2. Graduate School ‘Human Development in Landscapes’, Christian-Albrechts-Universität zu Kiel, Leibniz Straße 3, 24118, Kiel, Germany

    • Taylor R. Hermes
    •  & Cheryl A. Makarewicz
  3. SAIE Laboratory, Department of Anthropology, Washington University in St. Louis, CB 1114, One Brookings Drive, St. Louis, MO, 63130, USA

    • Michael D. Frachetti
    •  & Elissa A. Bullion
  4. Institute for Archaeological Research, Academy of Sciences of Uzbekistan, Yahya Gulamov Street No. 70, Tashkent, 100000, Uzbekistan

    • Farhod Maksudov
  5. Afrasiyab Museum, Islam Karimov Street No. 7, Samarkand, 114151, Uzbekistan

    • Samariddin Mustafokulov


  1. Search for Taylor R. Hermes in:

  2. Search for Michael D. Frachetti in:

  3. Search for Elissa A. Bullion in:

  4. Search for Farhod Maksudov in:

  5. Search for Samariddin Mustafokulov in:

  6. Search for Cheryl A. Makarewicz in:


T.R.H. designed and initiated the research. T.R.H., E.A.B., F.M., and S.M. conducted sample collection. E.A.B. performed the osteological analysis of human remains and, with T.R.H., described sites in Uzbekistan for the supplementary information. T.R.H. wrote the remaining supplementary information. T.R.H. prepared specimens for stable isotope analysis and conducted the data analysis. T.R.H. and C.A.M., and M.D.F. interpreted the results. T.R.H., M.D.F., and C.A.M. wrote the manuscript.

Competing Interests

The authors declare no competing interests.

Corresponding authors

Correspondence to Taylor R. Hermes or Michael D. Frachetti or Cheryl A. Makarewicz.

Electronic supplementary material

About this article

Publication history






By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.