Evidence for early dispersal of domestic sheep into Central Asia

ARTICLES https://doi.org/10.1038/s41562-021-01083-y Evidence for early dispersal of domestic sheep into Central Asia William T. T. Taylor 1,2 ✉, Mélanie Pruvost 3, Cosimo Posth 4,5, William Rendu 3,6, Maciej T. Krajcarz7, Aida Abdykanova 8, Greta Brancaleoni 7, Robert Spengler 2, Taylor Hermes 4, Stéphanie Schiavinato9, Gregory Hodgins 10, Raphaela Stahl4, Jina Min11, Saltanat Alisher kyzy 12,13, Stanisław Fedorowicz14, Ludovic Orlando8, Katerina Douka 2, Andrey Krivoshapkin12,13, Choongwon Jeong 11, Christina Warinner 4,15 and Svetlana Shnaider 6,12 ✉ The development and dispersal of agropastoralism transformed the cultural and ecological landscapes of the Old World, but little is known about when or how this process first impacted Central Asia. Here, we present archaeological and biomolecular evidence from Obishir V in southern Kyrgyzstan, establishing the presence of domesticated sheep by ca. 6,000 BCE. Zooarchaeological and collagen peptide mass fingerprinting show exploitation of Ovis and Capra, while cementum analysis of intact teeth implicates possible pastoral slaughter during the fall season. Most significantly, ancient DNA reveals these directly dated specimens as the domestic O. aries, within the genetic diversity of domesticated sheep lineages. Together, these results provide the earliest evidence for the use of livestock in the mountains of the Ferghana Valley, predating previous evidence by 3,000 years and suggesting that domestic animal economies reached the mountains of interior Central Asia far earlier than previously recognized. T he early Holocene domestication of crops and livestock in the Fertile Crescent is among the earliest in the world, with the first traits of domestication appearing in plants and animals by 7,500 BCE1–6. The food-producing economy of this region, based on sheep, goats, cows, cereals and legumes, launched humanity’s first agricultural demographic transition6,7, which would eventually reshape human populations, both genetically and culturally, across the ancient world8. Recent human genomic studies have clearly illustrated that the Neolithization of western Eurasia involved a demic wave of expansion as peoples of southwest Asian ancestry expanded and admixed with local populations in Europe and West Asia9–11. Understanding the dynamics of early Neolithic dispersals informs the processes that shaped the trajectory of human societies in Eurasia. Archaeozoological evidence suggests that the first progenitors of domesticated goats and sheep came under sustained, multigenerational human control by ca. 11,000–9,000 years ago in a region stretching from eastern Anatolia to the Zagros Mountains of Iran and Iraq12–14. Following their initial domestication, animal livestock (including sheep and goats, as well as cattle) and plant crops from this region dispersed across Eurasia and Africa in one of the most important globalization processes in human prehistory15. While these three livestock species sometimes moved together into new regions as part of a ‘Neolithic package’, domesticated sheep and goats became particularly widespread in the ancient world, reaching areas across Europe and the Mediterranean by ca. 6,000 BCE and North Africa by 5,000 BCE15. Although the impact of Western Asian livestock dispersals on the ancient cultural landscape of western Eurasia and Africa has been substantial, the spread of animal domesticates to the Eurasian interior is poorly understood16,17. Domesticated sheep and goat are not found in the archaeological record of the eastern steppe region of Inner Asia until the arrival of the pastoralist Afanasievo culture ca. 3,000 BCE18,19, and agropastoralism has only been traced to the early third millennium BCE in Central Asia20–22. However, Soviet archaeologists have long hypothesized a much earlier arrival of agropastoralism from southwest Asia between the seventh and fourth millennia BCE associated with the Kelteminar culture (Fig. 1)23–27. This poorly understood culture, concentrated in the Khoresm region beyond the southern edge of the Aral Sea, is characterized by their microlith technology, distinctive arrowheads with a prominent notch cut out at their base, handmade coarseware pottery with pointed bases and geometric incisions and a mixed pastoralist–hunting–fishing economy.26,28–31 Cattle and ovicaprid (sheep/goat) remains dating to at least the fifth millennium BCE have been reported from Kelteminar sites, but they have not been subjected to detailed zooarchaeological analysis27,32, and thus it is not clear whether they represent wild or domesticated populations. Museum of Natural History, University of Colorado-Boulder, Boulder, CO, USA. 2Department of Archaeology, Max Planck Institute for the Science of Human History, Jena, Germany. 3De la Préhistoire à l’Actuel: Culture, Environnement et Anthropologie (PACEA), Université de Bordeaux, Pessac, France. 4 Department of Archaeogenetics, Max Planck Institute for the Science of Human History, Jena, Germany. 5Institute for Archaeological Sciences, University of Tübingen, Tübingen, Germany. 6ArchaeoZOOlogy in Siberia and Central Asia – ZooSCAn, CNRS – IAET SB RAS International Research Laboratory, IRL 2013, Institute of Archaeology SB RAS, Novosibirsk, Russia. 7Institute of Geological Sciences, Polish Academy of Sciences, Warszawa, Poland. 8American University of Central Asia, Bishkek, Kyrgyzstan. 9Faculté de Médecine Purpan, Université Paul Sabatier, Toulouse, France. 10Accelerator Mass Spectrometry Laboratory, University of Arizona, Tucson, AZ, USA. 11School of Biological Sciences, Seoul National University, Seoul, Republic of Korea. 12Institute of Archaeology and Ethnography SB RAS, Novosibirsk, Russia. 13Novosibirsk State University, Novosibirsk, Russia. 14Department of Geomorphology and Quaternary Geology, University of Gdańsk, Gdańsk, Poland. 15Department of Anthropology, Harvard University, Cambridge, MA, USA. ✉e-mail: william.taylor@colorado.edu; sveta.shnayder@gmail.com 1 NATuRE HuMAN BEHAviOuR | www.nature.com/nathumbehav Content courtesy of Springer Nature, terms of use apply. Rights reserved ARTICLES NATURE HUMAN BEHAVIOUR Kazakhstan Kyrgyzstan Uzbekistan Obishir V Tajikistan China Afghanistan Pakistan Kelteminar culture C S E A A B L A C K S P I A N A S E Ca. 8,500 BCE KOPE T D AG M T NS Ca. 6,500 BCE M E D I T E R R A N E A N Obishir Djeitun Hissar culture SE A Before ca. 7,000 BCE Fig. 1 | Modeled dispersal of domestic animals into Central Asia. Proposed centre of sheep and goat domestication (dark red) and early Holocene dispersals out of western Eurasia (information from Vigne16 and Murphy90) alongside relevant archaeological cultures and newly hypothesized dispersal event(s) (red dashed arrow). Inset map of Obishir V and the Ferghana Valley situated within the mountainous zone of Central Asia. More recently, extensive research at the Djeitun site in southern Turkmenistan (Fig. 1) has revealed a complex agropastoral system, relying on harvesting tools, grinding stones, irrigation and a mixed crop assemblage, including glume wheats and barley, dated to ca. 6500 BCE. Based on the predominance of Ovis and Capra remains at Djeitun, along with their small size and some distinct morphological traits, domesticated sheep and goat are also thought to have been an important part of the subsistence economy at Djeitun33–36. Soviet scholars have also hypothesized the presence of another Neolithic culture known as the Hissar in the Hissaro-Alay Mountains of Tajikistan (Fig. 1). In its classic formulation, the Hissar culture is dated to between the sixth and second millennia BCE and is characterized by a pebble and microblade lithic technology, which includes geometric microliths37. Importantly, scholars have also argued that the Hissar economy included goat and sheep pastoralism37. However, as at Djeitun, the inference of domestic animals at Hissar sites has been based largely on the high frequency of Ovis and Capra remains and indirect lines of evidence such as site location, rather than direct evidence of human management. Taken together, these archaeological finds provide evidence for the presence of Neolithic material culture and domestic plants, along with tentative evidence for domestic animals, along the margins of Central Asia during the seventh millennium BCE. However, most of the relevant scholarly research on these cultures was performed prior to the widespread availability of scientific methods, including radiocarbon dating via accelerator mass spectrometry. Moreover, no direct archaeobotanical or zooarchaeological evidence for food production has been found east of the Kopet Dag piedmont of Turkmenistan prior to ca. 3,500 BCE22. With the advent of powerful new methods in archaeological science over the past decade, it is now possible to revisit these Soviet-era cultural designations and Neolithic dispersal models and to test these hypotheses in a robust scientific framework. Here we investigate animal husbandry at the early Holocene site of Obishir V in southern Kyrgyzstan. Located in the heart of Central Asia within the Inner Asian Mountain Corridor, Obishir V is situated along the southern margin of the Ferghana Valley, a historically significant crossroads for the exchange of people and animals across eastern and western Eurasia21. Through new excavations, we identified stone tools and faunal remains suggestive of grain processing and livestock pastoralism, dating to the late seventh millennium BCE. Zooarchaeological analysis and collagen fingerprinting reveal a faunal assemblage consisting primarily of Ovis and Capra, which based on cementum annulation appear to have been killed during the autumn season. Based on DNA sequences from five well-preserved specimens, we further identify at least four of these animals as O. aries within the diversity of domestic sheep, and showing genetic affinity to modern Anatolian and South Asian breeds. The presence of these domestic sheep at Obishir V ca. 6,000 BCE suggests that pastoral livestock species reached the foothills of the Pamiro-Alay mountains millennia earlier than previously recognized. These findings warrant a re-evaluation of the timing and routes of the earliest eastward agropastoral expansions and the role of domestic plants and animals in the development of social complexity in this crossroads region. Results Site excavation. The site of Obishir V (39° 57′ N, 71° 16′ E) is located near the Town of Aidarkyen, in Batken Prefecture of Osh Province, along the northern edge of the Pamiro-Alay range in the southern Ferghana Basin in Kyrgyzstan (Fig. 1). The area has an arid, montane climate with cold winters and hot summers, situated at an elevation of roughly 1,700 m above sea level. The site itself consists of a small rockshelter, originally identified by U.I. Islamov in 1965 and excavated throughout the 1960s and 1970s38. These initial excavations explored an area of 141 m2. Renewed excavations began at the site in 2015, revealing a stratigraphic sequence more than 4 m in depth, with five primary layers (Fig. 2). The sedimentary sequence was formed largely through colluvial rockfall and scree accumulation from Palaeozoic limestone and shale, which form the steep hillside. The lowermost layer, layer 5, consists of an aeolian loess-like deposit, NATuRE HuMAN BEHAviOuR | www.nature.com/nathumbehav Content courtesy of Springer Nature, terms of use apply. Rights reserved ARTICLES NATURE HUMAN BEHAVIOUR S N X (m) 8 9 10 11 13 12 14 15 –2 0 0 2.1 1 2.1 2.1 2.2 1 –3 2.3 2.3 4 2.4 3 Z (m) –4 4 5 –5 5 –6 Legend: Plan of 2015–2019 excavation area N Debris (limestone and shale angular clasts) X Riverine rounded pebbles Western profile Bedrock (carboniferous shale) Layer numbers Y 1 Layer boundaries Intra-layer sedimentary structures 1m Fig. 2 | Stratigraphic profile of Obishir v. The key layers are layer 0 (modern topsoil), layer 1 (Bronze through Middle Ages), layer 2 (early Holocene strata, including occupation surfaces and palaeosol), layers 3 and 4 (underlying colluvium containing contemporaneous and earlier cultural material) and layer 5 (a sterile final late Pleistocene/initial early Holocene loess deposit). Table 1 | Bayesian radiocarbon model (uniform prior) for Obishir v, by stratigraphic series, produced in OxCal using the iNTCAL20 calibration curve65 Boundary Modelled date (1σ) Modelled date (2σ) Layer 1 end 447–1372 cal CE 424 cal CE–present Layer 1 start 1,783–939 cal BCE 2,499–902 cal BCE Layer 2–3 end 2,897–2,342 cal BCE 2,914–1,624 cal BCE Layer 2–3 start 7,349–6,876 cal BCE 7,781–6,817 cal BCE Layer 4 end 8,129–7,555 cal BCE 8,160–7,161 cal BCE Layer 4 start 8,608–7,875 cal BCE 9,827–7,819 cal BCE Layer 5 end 10,124–8,276 cal BCE 11,581–7,993 cal BCE Layer 5 start 11,958–8,905 cal BCE 14,200–8,300 cal BCE For radiocarbon dates and model distributions, see Supplementary Information. which passes upward into layer 4 with a weakly developed palaeosol. Layer 5 was dated to approximately 10,000 BCE using thermoluminescence (Table 1 and Supplementary Information). Layers 2, 3 and 4 contain cultural deposits formed through complex sedimentary processes, involving some downslope relocation. One radiocarbon date on unidentified animal bone suggests possible cultural activity in layer 4 beginning as early as ca. 8,198–7,820 cal BCE. Layer 2 is a dark organic palaeosol and occupation surface with cultural activity clustering around ca. 6,000 cal BCE, underlain by several metres of colluvium that contain roughly contemporaneous cultural material, including from layer 3 (Supplementary Information). Dates from animal bones and charcoal recovered from layer 1 suggest that this layer formed much later and dates from the late Bronze Age through the Early Middle Ages (Supplementary Information). Comparing the lithic assemblage of Obishir V with earlier materials from Central Asia reveals an important economic transition characterized by a decline in the use of projectiles and a concomitant increased use of microblades for carcass processing, as well as an increased emphasis on grinding stones for food production. This lithic transition is reflected across much of mid-Holocene Central Asia39. Stone tools from layers 2–4 reveal a lithic industry based on pressure knapping of microblades, with important similarities to the Hissar culture, including a technological emphasis on retouched bladelets, trapezoids, end-scrapers and grinding stones, and evidence of processing of wild or domestic grain33. In addition, as seen in the Kelteminar culture23, the Obishirian industry also utilized pressure knapping from prismatic and cis-prismatic cores, but only yielded a handful of bullet cores and lacked other diagnostic features of the Kelteminar stone tool assemblage, such as scalene triangles, sickle blades and ‘Kelteminar points’. Finally, excavations at Obishir V recovered ornamental stone pendants and grinding stones, perhaps associated with food production, that have close parallels in both Hissar and Kelteminar cultures31,40. NATuRE HuMAN BEHAviOuR | www.nature.com/nathumbehav Content courtesy of Springer Nature, terms of use apply. Rights reserved ARTICLES NATURE HUMAN BEHAVIOUR a b Cervidae 287 1 Non-mammalian 4 250 Cervidae Bovidae 4 5 Deer Gazella sp. Saiga sp. Ovinae Ovis sp. Saiga sp. 1 200 8 Bovidae Count Deer Gazella sp. Saiga sp. Ovis sp. 150 35 100 Insufficient collagen for identification 88 7 Ovis sp. Capra sp. Pantholops sp. 6 50 Capra sp. 22 1 0 1 1 d. ul U ni Canis sp. Vulpes sp. 1 Lagomorph (unknown sp.) 1 Rodentia (unknown sp.) U ng ul at at e e (M (L ) ) pr a U ng O vi s/ ca ra C ap O vi s sp . sp . 1 Taxonomic category Fig. 3 | identification of animal remains at Obishir v by ZooMS. a, Morphology-based identifications of animal remains from layers 2–4 at Obishir V, by number of identified specimens (NISP) and taxonomic category. b, Collagen peptide-based identifications for layers 2–4, square R8 at Obishir V, grouped by general phylogenetic relationships. Each colored circle represents a different identified taxon. Specimens with a grey circle are missing necessary peptide markers for more detailed identification, and silhouettes depict a range of possible taxa from which specimens in that category may derive. Icons created or modified by Hans Sell. Unid., unidentified. In addition to lithics, Obishir V also yielded a small archaeofaunal assemblage, which has allowed an investigation of subsistence changes underlying the technological shift towards microblades. Although the faunal material (Supplementary Data A) is highly fragmented, layers 2–4 contained a total of 400 animal bone fragments. Of these, 24 could be morphologically identified as Ovis (n = 1), Capra (n = 1) or Ovis/Capra (n = 22). An additional 89 were identified as ungulate (88 medium-sized ungulates and 1 large ungulate), and 287 fragments were not identifiable to a taxonomic class (Fig. 3a). While the assemblage displayed many specimens with burn marks (n = 36), some with cut marks (n = 5) and others with spiral fracture (n = 3), only a small number of these anthropogenically modified specimens were taxonomically identifiable (Ovis, n = 1; ovicaprid, n = 1). A single human tooth was identified in the assemblage (a deciduous incisor shed from a child of approximately 6 years of age); however, this specimen was not recovered in stratigraphic context but was found during screening of sediments at the contact between layer 1 and 2. Radiocarbon date estimates for this tooth suggest that it belongs to the late Holocene cultural level 1 (Supplementary Information). Despite the small sample size and heavy fragmentation of the skeletal assemblage, the archaeofaunal material suggests that early Holocene animal economies at Obishir incorporated meaningful exploitation of sheep and goat. We next applied biomolecular methods to test whether these animals represent wild or domestic individuals, and to assess their antiquity directly. Archaeofaunal remains. Zooarchaeology by mass spectrometry (ZooMS) analysis41 of 74 archaeofaunal skeletal fragments from square R8, layers 2–4 (Supplementary Data B) confirms the presence of Ovis and Capra in the early Holocene layers at Obishir V, and among identifiable specimens suggests a nearly complete economic emphasis on these two taxa. Although nearly half of the assemblage from square R8 no longer had identifiable collagen (n = 35), the largest number of identifiable specimens were assigned as sheep (n = 8) or goat (n = 6), or ovicaprid lacking the necessary peptide marker to distinguish between sheep and goat (n = 7; Fig. 3b). We also identified one deciduous tooth belonging to a canine (Canis, likely a domestic dog or fox), one rodent bone, one lagomorph bone and one bone broadly assigned to deer/Saiga/ gazelle. On the basis of observed peptide markers, four specimens appeared non-mammalian, four were assigned only to Cervidae/ Bovidae and six additional specimens were consistent with either sheep or deer/Saiga/gazelle. Comparing the faunal assemblages of the Neolithic layers 2–4 at Obishir V with an Early Bronze Age assemblage from the nearby Alay Valley21 shows that an emphasis on Ovis, augmented by Capra, is characteristic of local Neolithic/ Bronze Age pastoral subsistence strategies. Later, during the first millennium BCE and first millennium CE, the Obishir V fauna also includes additional domesticated taxa, such as Bos sp. and Equus sp. (identified through DNA analysis as a male domesticated donkey, E. asinus; Supplementary Information). Dietary focus on domestic sheep, supplemented by goat and other taxa, characterizes modern herding lifeways in the region42, and may reflect cultural preferences or herd compositions optimized to the dry montane environment. Cementum analysis and dental eruption/wear. Cementum analysis of intact tooth specimens can provide insights into the age structure and management of animal herds, including the season during which animals are culled. Cementum is a connective calcified tissue that is deposited on the outer surface of the dental root, linking it with the fibres of the periodontal ligament. Consisting of a collagen matrix within which hydroxyapatite crystals form, its growth is continuous throughout an animal’s life43–45 and follows an annual cycle that typically produces light-coloured bands of low-mineral-density growth during the late spring to fall followed by dark banding of high mineral density during the non-growth, winter season46–48. On the basis of these patterns, the season of death for the animal can be established by identifying the growth status of the final mineral deposit44,49–51. Four faunal teeth (OB20-01, OB20-06, OB20-07 and OB20-09) were sufficiently intact to allow cementum analysis, and it was possible to identify at least one region of interest (ROI) per specimen based on annulation patterns (Supplementary Information). The four teeth likely represent four distinct individuals: three Ovis (as indicated by ZooMS and ancient DNA (aDNA)), and one Ovis/Capra (identified only as sheep/cervid by ZooMS from poor collagen preservation, but Ovis/Capra based on morphology). Across these four specimens, the cementum was globally well preserved. Limited recrystallizations were identified, but not concerning the NATuRE HuMAN BEHAviOuR | www.nature.com/nathumbehav Content courtesy of Springer Nature, terms of use apply. Rights reserved ARTICLES NATURE HUMAN BEHAVIOUR Table 2 | Osteological information and wear, age (cementum) and sex (DNA) estimates for teeth recovered from early Holocene cultural layers at Obishir v, along with taxonomic identifications from peptide fingerprinting Layer iD number Species iD (ZooMS) DNA sex/ Element species Side Wear stage (Payne 1973) 1–2 contact DA-OBI-1518-20.4 Ovis O.aries, male Lower incisor (permanent) R — 2 DA-OBI-1518-20.7 Ovis — Lower M1 (permanent) R C-I 5 years Late cementogenenesis (late fall?) 2 DA-OBI-1518-20.6 Ovis O.aries, female Lower I3 (deciduous) L — 2 years Late cementogenenesis (late fall?) 2 DA-OBI-1518-20.5 Ovis — Lower P4 (deciduous) L B-C 2.1 DA-OBI-1518-20.9 Deer/Saiga/ sheep/gazelle — Upper M1 (permanent) R B 1 year Late cementogenenesis (late fall?) 2.2 DA-OBI-1518-20.1 O.aries, female Upper M3 (permanent) L G-I 5 years Late cementogenenesis (late fall?) 2.2 DA-OBI-1518-20.2 Deer/Saiga/ sheep/gazelle — Upper P2 (permanent) R — 2.3 DA-OBI-1518-20.8 Capra — Upper P3 (permanent) R — Ovis/Capra Age (cementum) Estimated season of Notes death (cementum) Direct radiocarbon date 4,260 ± 36 7,012 ± 45 Deciduous (<18 months) Hyper-cementosis 6,989 ± 45 advanced wear Grey shading indicates DNA-based identification as domestic O. aries. b a m ntu me Ce Epoxy Dentine Dentine m tu en m Ce Epoxy 50 μm 200 μm Fig. 4 | Cementum analysis of sheep and goat remains from Obishir v. a, OB20-01. A thick deposit of mixed cementum is followed by incremental acellular cementum with extrinsic fibrils. Within it, four pairs of growth zones + annuli can be seen. The last increment is a complete growth zone, marking the end of the seasonal record. The cementum is well preserved and is not affected by post mortem modification in the ROI. b, OB20-07. Annuli are underlain by a line of cementocytes, showing five complete growth zones (the last one constituting the outermost increment). Observations were conducted in cellular cementum because no extrinsic fibrils of acellular cementum were present on the tooth. For both teeth, observations were conducted under polarized light with the insertion of a lambda plate under high magnification (500×). Dots show the location of annuli (winter bands); black arrow indicates the direction of cementum growth. outermost deposits, and no microbial damage was identified. Localized weathering influenced the cementum exterior on two specimens, but in both cases at least one tooth root was sufficiently preserved to allow reliable observation. Two teeth exhibited evidence of an older age-at-death (4 and 5 years, respectively), while the remaining two belonged to juvenile animals (1 and 2 years; Table 2). For all four teeth, the final increment of growth was nearly equivalent in width to the previous increment, suggesting the animals were at the end of their growth phase (fall or early winter) at the time of death (Fig. 4). Late-fall livestock culling patterns are a typical feature of pastoralist herd management52. Animal DNA. Five tooth fragments from Obishir V, square R8, layer 2 were selected for genetic analysis, on the basis of preservation and identification as Ovis (n = 4) or Capra (n = 1) through ZooMS (Table 3). Ancient DNA was successfully extracted from all five fragments, built into Illumina libraries and sequenced using a shotgun strategy. We observed high variability in endogenous DNA preservation across the fragments, ranging from 0.2% to 8.3%, and all five specimens exhibited patterns of post mortem DNA damage consistent with ancient DNA (Table 3 and Supplementary Information). We next compared genome-wide sequences of the Obishir V specimens with published genomes of modern NATuRE HuMAN BEHAviOuR | www.nature.com/nathumbehav Content courtesy of Springer Nature, terms of use apply. Rights reserved ARTICLES NATURE HUMAN BEHAVIOUR Table 3 | Whole mitogenome DNA results for specimens analysed in this study Specimen iD Species Reads sequenced Endogenous DNA No. of genomewide SNPs Damage First base 3′ Second base 3′ First base 5′ Sex Second base 5′ Avg frag. length No. of mt reads mt genome mt haplogroup coverage OB20-01 O. aries 18,506,454 8.32% 494,366 0.1242 0.0372 0.1628 0.0309 59.7 XX 718 1.9× A OB20-06 O. aries 17,858,515 6.25% 316,536 0.0797 0.0291 0.0787 0.0157 55.86 XX 6,846 13.6× A OB21-06 O. aries 25,946,522 0.23% 16,784 0.0758 0.0372 0.0688 0.0165 54.49 XX 399 0.69× n.d. OB20-04 O. aries 23,830,365 7.61% 473,283 0.1621 0.0352 0.1683 0.0212 48.97 XY 1589 3.6× A OB21-04 Capra sp. 27,141,767 0.27% 19,313 0.0975 0.0419 0.1155 0.0316 59.35 XY 37 0.3× n.d. Notes n.d., not determined Genome-wide SNPs are those from the ISGC SNP chip. HVRI and MT-CYTB numbering for sheep is relative to the sheep reference genome NC001941; numbering for goat is relative to the goat reference genome KR059154. O. orientalis 0.2 0.1 O. aries, N Europe O. aries, C Europe O. aries, SW Europe PC 3 (2.24 %) PC 2 (6.09 %) 0.1 C. hircus Obishir Obishir 21-4 0.0 O. aries (modern) O. orientalis Obishir 21–4 O. canadensis O. dalli 0.0 O. aries, Africa C. hircus O. aries, SW Asia –0.1 Obishir O. aries, Tibet –0.1 O. aries, Indonesia O. canadensis O. dalli –0.2 O. aries, S Asia –0.2 0.0 0.1 0.2 0.3 0.0 0.1 0.2 0.3 PC 1 (11.49 %) PC 1 (11.49 %) Fig. 5 | PCA results of Obishir v sheep and goat samples. a,b The results for principal component (PC) 1 versus PC 2 (a) and PC 3 (b), projected onto a reference database of wild sheep taxa as well as domestic sheep and goat specimens from across Eurasia and Africa. domestic sheep and goats from Eurasia and Africa, as well as wild Asiatic mouflon using principal component analysis (PCA). Four of the Obishir V specimens (OB20-01, OB20-06, OB21-06 and OB20-04) clustered together with present-day domesticated sheep breeds (O. aries) while being clearly separated from wild sheep species (Fig. 5). Based on the autosomal-to-sex chromosome ratio in the samples, three sheep (OB20-01, OB20-06 and OB21-06) were identified as female (Table 3 and Supplementary Information) and one sheep (OB20-04) was identified as male. Specimens OB20-01 and OB20-06 were directly dated through ultrafiltration to 6,989 ± 45 and 7,012 ± 45 14C years BP, or between 5,983 and 5,751, and 5,991 and 5,771 cal BCE, respectively (2σ calibrated range). Both specimens showed good measures of quality control (fraction of modern carbon 0.4189 ± 0.00231 (1σ) for OB20-1 and 0.4178 ± 0.0023 (1σ) for OB20-06). No material remained for direct dating of the third Ovis specimen with poor coverage (OB21-06), while specimen OB20-04 was dated to the Early Bronze Age (ca. 3,003–2,701 cal BCE, 2σ). The single Capra tooth (OB21-04), recovered from the contact between layers 1 and 2, was revealed to be Late Bronze Age in origin (Supplementary Information). In PCA space, the Obishir V sheep genomes fall close to modern southwest and Central Asian sheep breeds (Fig. 5). In addition, the Obishir V sheep carry hypervariable region I (HVRI) polymorphisms that are shared with domestic sheep breeds but that are not found in modern wild sheep (Supplementary Information). For three of the four Obishir V sheep (OB20-01, OB20-04 and OB20-06), we obtained sufficient mitogenome coverage (1.9–13.6x) to perform phylogenetic analysis comparing the Obishir V samples with 160 modern mitogenomes (Supplementary Data C) of O. aries (domestic sheep), O. ammon (wild argali), O. vignei (wild urial), O. orientalis ophion (wild Cyprus mouflon) and O. aries musimon (wild European mouflon). We found that Neolithic Obishir V sheep, as well as the single Early Bronze Age specimen, all clustered within O. aries haplogroup A, a major mitochondrial DNA subclade of domestic sheep (Fig. 6 and Table 3). Discussion Our identification of O. aries belonging to mitochondrial haplogroup A at Obishir V ca. 6,000 BCE has critical implications for our understanding of transcontinental connections and domestic animal dispersals. Although our assemblage is small and badly fragmented, ZooMS analysis enabled us to confirm inferences based on morphological comparisons that the Obishir V economy included both sheep and goat (Ovis and Capra sp.). Wild Ovis and Capra in this region include the large argali sheep (O. ammon) and Siberian ibex (C. sibirica), but a ratio of roughly 3:1 between Ovis and Capra has characterized pastoralist archaeological assemblages in Central and Inner Asia since the Bronze Age22,52, and is also typical of other early western Eurasian pastoral assemblages53. Based on the ZooMS results, Ovis and Capra appear to account for the majority of the unidentifiable fragments in the layer 2 faunal assemblage. NATuRE HuMAN BEHAviOuR | www.nature.com/nathumbehav Content courtesy of Springer Nature, terms of use apply. Rights reserved ARTICLES NATURE HUMAN BEHAVIOUR KF938359 KF938349 KU681202 B1a12 KF938329 KF938339 86 KU6811 1183 KU68 1185 KU68 76 811 KU6 90 811 KU6 192 681 KU 182 681 80 KU 1 a7 681 5 B1 KU 8117 5 6 9 KU 11 68 55 KU 024 4 3 5 KF 24 3 a9 30 22 KF 681 14 B1 2 KU 681 224 KU 681 344 3 KU 938 34 8 KF 93 KF European mouflon O. orientalis musimon HM HM 23 2 6 KF 36 185 1 KF 302 84 4 3 KF 024 51 3 5 0 0 KU 2 68 449 12 KR B1 04 8 a6 KF 6867 B2 93 83 8 KU 681 54 2 KF 938 03 348 MF 004 246 KF9 383 50 KF9 3835 7 KF93 8352 KF938 335 KF9383 B1b 33 KF3024 47 KU681200 KU681199 KU681196 KP702285 KF938328 KU681212 NC_001941 .1 AY858379 B1a5 KF9383 56 KF938 355 EF49 0451 KF9 7784 6 KF3 024 53 KF3 0 KF9 2452 383 KF 51 3 KF 02448 9 MF 3834 EF 0042 0 4 4 KF 904 5 KF 302 55 4 KU 302 61 KU 681 4 6 2 KF 68 217 KF 93 121 97 834 3 78 7 45 Haplogroup B 76 61 7 23 617 HM 23 341 M H 938 346 KF 938 358 KF 938 52 KF 904 6 4 2 5 a EF 904 B1 0 4 EF 0246 3 KF 38353 9 KF 456 2 0 3 KF 57 024 KF3 458 302 F K 8 59 024 B1a 3 F K 1379 8 9 KP 1216 KU68 05 KU6812 EF490453 B1a10 KU681210 EF490454 KR059146 HM236189 HM236186 HM236187 HM236188 NC_02065 6_1 HM2361 88_2 HM23 6188_1 NC_0 2065 6_2 KT7 8168 9_2 KT7 8 KP9 1689_2 813 KP 8 0 98 HM 1378 2 HM 3618 KF 2361 2 83 31 HM 22 38 HM 236 1 KF 236 78 1 9 7 KF 38 9 KP 938 318 KT 99 327 14 847 89 1 68 KF977847 KF938336 KF938332 98 KU6811 94 KU6811 1193 KU68 84 11 8 6 KU 7 8119 KU6 81 811 KU6 11 812 6 U K 215 681 79 KU 1 8 6 1 7 KU 8117 9 6 KU 8120 6 22 U K 12 68 207 U K 81 74 6 KU 2361 21 2 HM 681 206 9 KU 681 21 1 KU 68 KU b A1 O. vignei O. ammon East Asian O. aries Obishir 20-4 KF938331 KU KU 68 1 KF 681 22 0 KF 938 218 3 9 KF 38 24 3 9 KF 383 21 9 2 KF 383 2 2 9 KF 3831 3 9 9 KF 3833 93 83 4 25 KF3 A2 024 44 KF3 024 4 3 KF3 024 42 KF30 2441 KF30 2440 KF3024 45 KU6812 08 KF302446 KF938345 KF938326 Haplogroup E 87 11 8 68 118 9 KU 68 118 KU 68 191 KU 681 201 KU 681 20 83 0 KU 3 9 18 KF 36 81 2 HM 2361 HM 243 004 MF 472 998 KP 8337 3 KF9 6175 3 HM2 0 3833 KF9 4 0424 MF0 242 MF004 73 KP9984 KP998470 KU575248 KF938342 KF938338 Haplogroup A a A1 C. hircus Cypriot mouflon O. orientalis ophion Haplogroup C Haplogroup D Obishir 20-1 Obishir 20-6 Fig. 6 | Phylogenetic tree produced using maximum parsimony. The tress shows the whole mitochondrial genome relationships of the Obishir V specimens (stars) versus modern reference genomes of the domestic sheep O. aries (haplogroups A–E), the European mouflon (O. aries musimon), the Cypriot mouflon (O. orientalis ophion), the wild urial (O. vignei) and the wild argali (O. ammon), with C. hircus as an outgroup. Major haplogroups A–E are labelled following previous published studies91. For full details on tree generation, see Supplementary Information. Further genetic analysis suggests that the Obishir V includes domestic animals. Mitochondrial DNA data recovered from sheep teeth confirms that these specimens fall within the present-day genetic diversity of domestic O. aries, and at least three (two Neolithic and one Early Bronze Age specimen) are members of haplogroup A. This haplogroup has been hypothesized to be linked with the earliest domestication of sheep that widely dispersed throughout Eurasia, as it is the only haplogroup found among the local sheep populations of North Africa, Britain and Northern Europe54,55. Significantly, haplogroup A is also found at high frequency in early domestic sheep populations in East Asia56. Information available for domesticated sheep suggests that domesticated sheep are monophyletic with five distinct haplogroups (haplogroups A–E)57,58, with the sole exception of the Cypriot mouflon (O. orientalis ophion) and the European mouflon (O. aries musimon). These subspecies are often considered to be relict populations of early domesticated European sheep. The identification of Obishir V specimens as members of haplogroup A does not rule out the possibility that this haplogroup was present in a previously undocumented wild species of Ovis exploited by ancient Central Asian hunters. However, the fact that all Obishir V specimens with complete mitochondrial genomes are nested within the range of established diversity in domesticated sheep but do not form a basal lineage or cluster with other wild sheep taxa (Fig. 6) casts doubt on this possibility. Moreover, our PCA comparison of Obishir V sheep places them between modern South Asian, East Asian and European populations (Fig. 5). Based on their co-occurence with O. aries, the Capra sp. specimens recovered at Obishir V may also represent domestic animals. However, no haplogroup assignment could be made for the Capra specimen due to the poor preservation of the recovered DNA, and the only directly dated Capra specimen, recovered from the contact between layers 1 and 2, was dated to the early first millennium BCE. As a result, future research is necessary to assess whether the Capra specimens at Obishir V are indeed reliably associated with the early occupation layer and how they relate to known ancient populations of wild and domestic animals. Careful excavation and direct radiocarbon dating of analysed specimens (including two teeth identified through aDNA as O. aries: specimens OB20-01 and OB20-06) confirms the deep antiquity of the Obishir V sheep specimens. Radiocarbon dating of faunal remains at Obishir V provides a robust chronology, dating the initial occupation of layers 2 and 3 to before 6,000 BCE. Bayesian model estimates for the onset of occupation in this layer range between 7,800 and 6,800 cal BCE at the 2σ range (Table 1). The earliest radiocarbon date in the early Holocene cultural layer comes from unidentified wood charcoal (Obishir V, R-11, depth 352, layer 2), which may be influenced by old wood effect and may not reliably date cultural activity at the site. During the early formation of this layer, geogenic material and archaeological material may have been re-deposited and mixed together over a period of two or three millennia. Thus, while human activity at the site might have begun as early as 9,800–7,800 cal BCE with unidentified faunal materials in layer 4, specimens identified as Ovis and Capra (as well as those of NATuRE HuMAN BEHAviOuR | www.nature.com/nathumbehav Content courtesy of Springer Nature, terms of use apply. Rights reserved ARTICLES NATURE HUMAN BEHAVIOUR similar size class but identified only to lower-resolution taxonomic categories) produced radiocarbon dates clustering around ca. 6,000 BCE (Supplementary Information). Sample sizes are too small to demonstrate clear patterns of management by humans, but age and seasonality data from the Obishir V teeth provide important context supporting the genetic findings, and strengthen the interpretation of Obishir V sheep as domestic animals. Incremental cementum banding on four intact teeth from Obishir V shows that two animals died between 1 and 2 years of age, a pattern consistent with early pastoral culling practices across southwest Asia after 7,500 BCE and probably representing a focus on meat or dairy rather than wool59. The slaughter of mature females (older than 2 years) that have failed to reproduce or are small in body size is another hallmark of pastoral management of a breeding herd12,59,60. Based on cementum analysis, the female O. aries specimens identified through DNA sequencing were 2+ years of age and 5+ years of age (Table 2). Another deciduous premolar (<18 monoths) and several unfused metapodial shafts reinforce the impression of the regular slaughter of juvenile individuals at Obishir V. Analysis of banding patterns indicates seasonal slaughter of Obishir V sheep and goat in the fall or early winter in all analysed teeth. Among contemporary pastoral sheep and goat dietary economies, it is common to cull animals that are unlikely to survive the winter in late autumn, so as to improve winter herd survival and store meat more easily over winter months61. The consistency in the apparent seasonality of death among all four teeth inspected for cementum analysis implies either: (A) seasonal occupation of Obishir V during the fall months paired with hunting of juveniles and mature females or (B) regular late-season slaughter of domestic animals. Although our previous research21 demonstrated that pastoral economies had been present in the region since the early Bronze Age, until now, no direct evidence had been found for domestic animals in montane Central Asia during the early Holocene. Our findings indicate that the subsistence behaviours that underpinned these nascent interaction networks could date from at least three millennia earlier than previously demonstrated. Assuming a single centre of domestication in the Near East, the evidence from Obishir V would appear to directly demonstrate the movement of domestic animals deep into the Eurasian interior, either through human movements or cultural exchange with western Eurasia. The timing of probable domesticated animal use at Obishir V chronologically aligns with the first dispersal of domesticated bovids into Mediterranean North Africa15 and the arrival of domestic sheep and goat in the Kopet Dag of southern Turkmenistan34. Therefore, Obishir V may form an extended outward branch of animals from southwest Asia by before ca. 6,000 BCE (Fig. 1). In comparison with the Iranian Plateau and the Levant, the relatively more challenging ecological setting of montane Central Asia (with its high altitude, hard frosts and cold winters) implies that such conditions were not an impermeable ecological barrier for the dispersal of domestic animals during the mid-Holocene, and raise the possibility that similar assemblages may exist across other areas of Central Asia. The presence of probable domestic sheep in the Ferghana Basin by the seventh millennium BCE may warrant reconsideration of the timing of other components of the ‘Neolithic package’, including goat, cattle and crops such as wheat and barley, in the Inner Asian Mountain Corridor. Our identification of O. aries dated to the early Bronze Age (OB20-04, dating to ca. 3,003–2,701 cal BCE at the 2σ level) from layer 2, along with the faunal remains from layer 1 (Supplementary Data A, B) point to continuous use of domestic animals in the Ferghana region from their Neolithic introduction throughout the Middle Ages. In later history and prehistory, the southern Ferghana region became linked to northwest China through networks of pastoral mobility that drove exchange across the interior62. Our research shows that such networks could date from at least three millennia earlier than previously demonstrated. Based on the material culture links between Obishir V and other important Neolithic horizons (such as Hissar and Kelteminar), as well as the paucity of scientific study on highly fragmented early faunal assemblages from the Pamiro-Alay and Tian Shan region, we suggest that Obishir V may form part of a much broader phenomenon linked with early pastoralist or agropastoralist economies (Fig. 1). If so, the region may have been a key node in an early dispersal system for southwest Asian domesticated bovids and, possibly, plants into other areas of East and South Asia. No archaeobotanical data are yet available for Obishir V. Future investigation of this and other mid-Holocene sites in Central Asia should assess the possibility that domestic plants were also dispersed into the region during the mid-Holocene, and clarify whether this newly identified Neolithic expansion was demic (accompanied by human dispersals, as in much of Eastern Europe) or took place through transfer of organisms and technologies, as in much of non-Mediterranean Africa63 to better understand the cultural and ecological dynamics of the Neolithic transition across the ancient world. Until now, archaeological data have pointed to a relatively late arrival of the agropastoral economy to interior Central Asia, around 3,500 BCE. Results from Obishir V suggest that the chronology of Neolithization in this region is in need of revision. Archaeological investigation of the early Holocene deposits at the rockshelter of Obishir V have revealed major changes in material culture, along with fragmented faunal material potentially linked to an early pastoral economy. Using a combination of collagen fingerprinting, ancient DNA and thin-section cementum analysis, we propose an early dispersal of domestic sheep into the Ferghana Valley by at least 6,000 BCE, concurrent with other out-migrations from Western Asia into the Old World. These results support Soviet scholarship hypothesizing a Neolithic dispersal of food-producing economies into Central Asia through connections with Western Asia, and raise the possibility that domesticated sheep, perhaps in association with other early domesticated plants and animals, could have dispersed into the continental interior millennia earlier than previously recognized. Methods Excavation. Excavations at the Obishir V site were carried out between 2015 and 2019 by a joint Russian–Kyrgyzstani archaeological expedition. In 2015–2016, researchers working at the site excavated an area of 8 m2 adjoining the 1968–1969 excavation area (Supplementary Information). Each artefact (lithic and faunal remains) greater than 1 cm in size was individually piece-plotted with 1 mm accuracy using a Leica Total Station TS02 plus in tandem with the Trimble software EDM Mobile, and documented along with contextual information (orientation, dip, inclination, find category, stratigraphic level, etc.). Fragments smaller than this were collected in bulk for each 0.25 m2 area. Features (natural or anthropogenic) were recorded with the total station using outlines, surface points and breaklines. Large surfaces, such as walls of old excavation pits, sections and archaeological horizons with numerous artefacts, were also documented using photogrammetry and three-dimensional (3D) scanning. Digital records were supplemented by field drawings and notes. All sediment obtained during the excavation was wet-screened with mesh of size of 2.0 mm, and finds were documented according to excavated square, layer and depth. Stratigraphic units were defined during the excavation based on grain size composition, consistency, colour, presence of erosional surfaces and accumulation of limestone clasts. Radiocarbon dating. Nine radiocarbon samples of identified and unidentified animal bone and tooth (n = 7) and charcoal (n = 2) were selected from various depths across the stratigraphic unit of layers 2–4 and submitted for radiocarbon dating at the Oxford Radiocarbon Accelerator Laboratory, the Center for Isotope Research at the University of Groningen, the Golden Valley Laboratory at Novosibirsk and the University of Arizona AMS Laboratory. Portions of domestic sheep teeth OB20-01, OB20-06 and OB20-04 from layer 2 used in genomic analysis were submitted to the University of Arizona Accelerator Mass Spectrometry Laboratory. Collagen was obtained from dentin using acid–base–acid (ABA) pretreatment, gelatinization, 0.45 micron filtration and ultrafiltration. The quality control parameters, collagen yield of 7.1%, 6.4% and 7.2%, carbon yield of 42.9%, 36% and 36% and d13C values of −19.7, −19.3 and −19.6 ± 0.1 per mil, indicated good preservation64. The atomic C:N ratio for OB20-01 (3.2) and OB20-04 (3.2) also fell within ranges indicating good preservation, although no material remained from tooth OB20-06 for C:N measurement after 14C analysis. All dates were calibrated in OxCal using the INTCAL20 calibration curve65. NATuRE HuMAN BEHAviOuR | www.nature.com/nathumbehav Content courtesy of Springer Nature, terms of use apply. Rights reserved ARTICLES NATURE HUMAN BEHAVIOUR We performed Bayesian stratigraphic phase modelling (using ordered phases and combining layers 2 and 3 due to their apparent mixing) in the software OxCal with a uniform prior using all available radiocarbon dates from Obishir V, and two thermoluminescence dates from layer 5, following methods outlined by Ramsey66. Because charcoal dates, particularly those from layer 2, suggest an old wood effect that produced unsuccessful models in OxCal, only radiocarbon dates from bone and teeth were included in the final chronology. The models provided good agreement, and repeating the analysis using a general outlier model did not significantly alter estimated parameters. The OxCal code used in the analysis, along with the human tooth dating methods, are outlined in Supplementary Information. Thermoluminescence dating. During the 2017 field season, two sediment samples were collected from the southern portion of the western excavation profile of layer 5 at a depth of 295 and 335 cm below the surface. We measured the deposit moisture in each sample and, after drying, determined the dose rate (DR) using a MAZAR gamma spectrometer. Concentrations of 226Ra, 228Th and 40 K in each sample were obtained from 20 measurements lasting 2,000 s each. We established equivalent dose (ED) on the 63–80 mm polymineral fraction, after 10% HCl and 30% H2O2 washing and ultraviolet optical bleaching. The samples were irradiated with 20, 30, 40, 50 and 100 Gy doses from 60Co gamma source. Before measurement, we heated the samples at 140 °C for 3 h. A sample pre-treated in this way was used to determine the ED by the thermoluminescence multiple-aliquot regenerative technique67, according to the description published by Fedorowicz et al.68. Curve registration was performed on a RA’94 (Mikrolab) thermoluminescence reader coupled with an EMI 9789 QA photomultiplier. Finally, we calculated thermoluminescence age according to Frechen69. A detailed description of the preparation and the equipment used in the study is provided in Fedorowicz et al.68. Zooarchaeology and ZooMS. For each specimen, we conducted taphonomic analysis and morphology-based species identification using comparative faunal collections held by the Russian Academy of Sciences-Siberia in their field station at Aidarkyen, Kyrgyzstan, and attempted refits for all specimens to control for issues of fragmentation. We recorded taphonomic indicators such as rodent and carnivore damage, root etching, and weathering70, along with evidence for anthropogenic modification such as spiral fracturing, cut marks and burning via visual inspection (Supplementary Data A). In addition, we used collagen fingerprinting (ZooMS) to taxonomically identify a subset of this material, all derived from 2017 excavations. We analysed all bone specimens from square R8, layers 2.2, 3 and 4, as well as all bone specimens from square R8, layer 1.3 using ZooMS. To do so, we demineralized 10 mg of bone using 50 mM ammonium bicarbonate buffer (Sigma-Aldrich) produced in UHQ water and pH adjusted to 8.0 using ammonium hydroxide (1 M), following the protocol outlined by van Doorn et al.71. The extracted collagen was digested into component peptides using trypsin (Pierce), and was purified using Pierce C18 tips (Thermo Scientific), eluting in 50% acetonitrile (Sigma-Aldrich) and 0.1% trifluoroacetic acid (Sigma-Aldrich). Samples were then spotted on Bruker AnchorChip with Bruker Peptide Calibration Standard in calibration spots directly neighbouring the samples. Mass spectrometric analysis was conducted using a Bruker Autoflex Speed LRF matrix-assisted laser desorption/ionization–time of flight (MALDI-TOF) device in the dedicated ZooMS Laboratory at the Max Planck Institute for the Science of Human History, Jena, Germany. The acquisition used the following parameters: 4,000 laser shots at 50–60% intensity (50 shots per spot), mass range 600–3,500 Da and reflector mode. Identifications were made using published reference spectra from the Eurasian mammals database72 and are reported according to the level of taxonomic specificity. In some cases (for example, Ovis versus muskox and chamois), species that were not necessarily separable on the basis of observed peptide markers were inferred on the basis of known habitat distribution. All peptide marker data are provided in Supplementary Data B. Cementum analysis. First, all analysed teeth were 3D scanned using a MicroCT scanner, and washed in 90% ethanol solution. A cast of the occlusal surface was retained for later micro-wear analyses. Thin sections were prepared following recommendations by Rendu73. Teeth were embedded in epoxy within a vacuum chamber (at 0.2 bar pressure). They were then cut at thickness of 250 µm using an Isomet5000 automatic saw, and ground to thickness of 100 µm with a PressiCube grinding machine. For each tooth, four independent thin sections were prepared and analysed independently. Observations were conducted under polarized light both with and without the insertion of a lambda plate74, using a Leica 2500P microscope equipped with a Leica DC 120 camera. Post-depositional modifications were analysed systematically following Geusa et al.75 and Stutz74. Measurements of the increments were conducted using ImageJ software following methods outlined by Lieberman et al.76. DNA analysis. Extraction. DNA was extracted from six teeth using silica-based purification in a dedicated laboratory for aDNA at PACEA (UMR 5199, University of Bordeaux). We incubated 100–250 mg of bone powder in 1 mL extraction buffer (0.5 M EDTA, 0.25 M Na2HPO4, 50 μg mL−1 proteinase K) for 48–72 h at 37 °C. DNA was purified from the collected supernatant using a method adapted from the QIAquick MinElute purification kit protocol (Qiagen). PCR amplification. Purified DNA was amplified by quantitative polymerase chain reaction (qPCR) using a LightCycler 96 Instrument (Roche Life Science). Primer pairs were designed to amplify two regions of the mitochondrial DNA, a 130 bp region from cytB and 170 bp from the hypervariable region (see Supplementary Information for a list of the primers used and product sizes). We obtained each PCR product at least twice, conducting standard negative control to ensure the authenticity of the results. We used 1 µL of extract in 10 µL reaction with 1× FastStart Essential DNA Green Master and 1 µM of each primer. Products were sequenced by capillary electrophoresis at Genewiz (Leipzig, Germany). Sequences were visually inspected and manually curated, assembled and aligned using Geneious software suite v9.1.8 (https://www.geneious.com). Library construction and sequencing. We used DNA extracts to build double-stranded Illumina libraries following the protocol described by Gorge et al.77. We sequenced whole-genome shotgun libraries for ~500,000 reads on an Illumina NextSeq 500 at the Institut de Recherche Biomédicale des Armées in Brétigny-sur-Orge, France, to assess DNA preservation. DNA extracts of five samples produced in Bordeaux were sent to the dedicated cleanroom facilities of the Max Planck Institute for the Science of Human History. There, 20 μL of the extract was used to build double-stranded libraries with partial treatment of the uracil-DNA glycosylase enzyme (UDG-half)78. The resulting genetic libraries were double indexed and further amplified79. Finally, all five samples were shotgun sequenced for ~20 M reads on an Illumina HiSeq run with a 75-cycle single-end configuration. Sequence authenticity was estimated on libraries subjected to either partial or no USER treatment using MapDamage80. We trimmed adapter sequences from generated sequences, merging overlapping paired-end reads and filtering out reads shorter than 30 bases using Clip and Merge. We then mapped our filtered reads to both the O. aries genome assembly (International Sheep Genome Consortium build Oarv3.1) and the C. hircus genome assembly (ARS1) using BWA81, and against the full mitochondrial genome of O. aries (NC_001941.1) with a duplication of its first 500 bases at the end to ensure mapping of the reads overlapping the junction resulting from the virtual linearization of the circular mitogenome. PCA of genome-wide sequences. Genotypes were called for the 38.5M single-nucleotide polymorphisms (SNPs) from the International Sheep Genomics Consortium (ISGC) SNP project from 60 domesticated sheep along with three specimens of North American bighorn sheep (O. canadensis) and two Dall sheep (O. dalli). We drew a single allele at random for each position (minimum mapping and base quality of 30) using PileupCaller (https://github.com/stschiff/ sequenceTools), rendering the individuals from the dataset homozygous for each locus. The two files were then merged using Plink v1.982. To augment the ISGC SNP discovery panel, we added published genome data of 16 Asiatic mouflons (O. orientalis)83. For this, we downloaded FastQ files for the 16 individuals from the National Center for Biotechnology Information Sequence Read Archive under the accession number PRJNA24020 and aligned reads to the oviAri4 reference genome using BWA-mem v0.7.1781. We removed PCR duplicates using the MarkDuplicates module of the picardtools program v2.20.0 (https://broadinstitute.github.io/ picard/). We then retained properly paired reads with mapping quality score 30 or higher using SAMtools v1.984. For the SNPs from the ISGC SNP project, we calculated genotype likelihoods using GATK UnifiedGenotyper (v3.8.1.0) with ‘–genotype_likelihoods_model SNP–output_mode EMIT_ALL_SITES –allSitePLs’ options85. Then, we used an in-house python script to calculate posterior genotype probability with a non-default prior [0.4995, 0.0010, 0.4995] to reduce reference bias, following the approach taken in the Simons Genome Diversity Project86. We took genotypes with posterior probability 0.9 or higher and kept the remaining ones as missing. Finally, we calculated sequencing coverage with Qualimap v2.2.1, assigning biological sex to each individual based on the X to autosome coverage ratio (males around 0.5, females around 1.0). Modern wild and domestic sheep samples used from the ISGC, along with 16 Asiatic mouflons, and the three goat genomes are indicated in Supplementary Data D. We then conducted PCA using smartPCA implemented in EIGENSOFT, projecting ancient sequences from Obishir onto the first three components of the PCA defined by the full dataset (including goat, wild sheep and domestic sheep reference specimens) to visualize diversity among sheep species. Mitochondrial genome analysis. We produced a multiple genome alignment of our newly reconstructed sequences (excluding OB21-04 and OB21-06 because of low coverage) along with 160 present-day sequences from various breeds of modern O. aries, O. ammon, O. vignei, O. orientalis ophion and O. aries musimon (Supplementary Data C), using a modern C. hircus as outgroup, with Geneious software suite v9.1.8. Gaps were removed from the alignment. Phylogenetic trees were inferred using both the maximum-parsimony (MP) method (Fig. 6) and maximum likelihood (ML) with Tamura–Nei model (Supplementary Information) in MEGA version X87–89. We visualized and manipulated phylogenetic trees in FIGTREE v1.4.1 (http://tree.bio.ed.ac.uk/software/figtree/). NATuRE HuMAN BEHAviOuR | www.nature.com/nathumbehav Content courtesy of Springer Nature, terms of use apply. Rights reserved ARTICLES NATURE HUMAN BEHAVIOUR Reporting summary. Further information on research design is available in the Nature Research Reporting Summary linked to this article. Data availability Shotgun sequencing raw files are available at the European Nucleotide Archive (ENA) database under accession number PRJEB41594. Received: 7 September 2020; Accepted: 17 February 2021; Published: xx xx xxxx References 1. Asouti, E. & Fuller, D. Q. A contextual approach to the emergence of agriculture in southwest Asia: reconstructing early Neolithic plant-food production. Curr. Anthropol. 54, 299–345 (2013). 2. Conolly, J. et al. Meta-analysis of zooarchaeological data from SW Asia and SE Europe provides insight into the origins and spread of animal husbandry. J. Archaeol. Sci. 38, 538–545 (2011). 3. Larson, G. et al. Current perspectives and the future of domestication studies. Proc. Natl Acad. Sci. U. S. A. 111, 6139–6146 (2014). 4. Willcox, G. Measuring grain size and identifying Near Eastern cereal domestication: evidence from the Euphrates Valley. J. Archaeol. Sci. 31, 145–150 (2004). 5. Willcox, G. & Stordeur, D. Large-scale cereal processing before domestication during the tenth millennium cal BC in northern Syria. Antiquity 86, 99–114 (2012). 6. Larson, G. & Fuller, D. Q. The evolution of animal domestication. Annu. Rev. Ecol. Evol. Syst. 45, 115–136 (2014). 7. Bocquet‐Appel, J. Paleoanthropological traces of a Neolithic demographic transition. Curr. Anthropol. 43, 637–650 (2002). 8. Bellwood, P. First Farmers: the Origins of Agricultural Societies (Wiley, 2004). 9. Omrak, A. et al. Genomic evidence establishes Anatolia as the source of the European Neolithic gene pool. Curr. Biol. 26, 270–275 (2016). 10. Olalde, I. et al. The genomic history of the Iberian Peninsula over the past 8000 years. Science 363, 1230–1234 (2019). 11. Brace, S. et al. Ancient genomes indicate population replacement in early Neolithic Britain. Nat. Ecol. Evol. 3, 765–771 (2019). 12. Zeder, M. A. & Hesse, B. The initial domestication of goats (Capra hircus) in the Zagros mountains 10,000 years ago. Science 287, 2254–2257 (2000). 13. Daly, K. G. et al. Ancient goat genomes reveal mosaic domestication in the Fertile Crescent. Science 361, 85–88 (2018). 14. Alberto, F. J. et al. Convergent genomic signatures of domestication in sheep and goats. Nat. Commun. 9, 813 (2018). 15. Zeder, M. A. in Human Dispersal and Species Movement: from Prehistory to the Present (eds Petraglia, M.D., Crassard, R. & Boivin, N.) p. 261 (Cambridge Univ. Press, 2017). 16. Vigne, J.-D. Early domestication and farming: what should we know or do for a better understanding? Anthropozoologica 50, 123–151 (2015). 17. Pereira, F. & Amorim, A. Encyclopedia of Life Sciences https://doi. org/10.1002/9780470015902.a0022864 (2010). 18. Hermes, T. R. et al. Mitochondrial DNA of domesticated sheep confirms pastoralist component of Afanasievo subsistence economy in the Altai Mountains (3300–2900 cal BC). Archaeol. Res. Asia 24, 100232 (2020). 19. Wilkin, S. et al. Dairy pastoralism sustained eastern Eurasian steppe populations for 5,000 years. Nat. Ecol. Evol. 4, 346–355 (2020). 20. Hermes, T. R. et al. Early integration of pastoralism and millet cultivation in Bronze Age Eurasia. Proc. Biol. Sci. 286, 20191273 (2019). 21. Taylor, W. et al. Early pastoral economies along the Ancient Silk Road: biomolecular evidence from the Alay Valley, Kyrgyzstan. PLoS ONE 13, e0205646 (2018). 22. Spengler, R. N. & Willcox, G. Archaeobotanical results from Sarazm, Tajikistan, an early Bronze Age settlement on the edge: agriculture and exchange. Environ. Archaeol. 18, 211–221 (2013). 23. Korobkova, G. F. Tools and Economy of the Neolithic Populations in Central Asia (Nauka, 1969). 24. Masson, V. M. The Jeitun Settlement: the Emergence of a Productive Economy (Nauka, 1971). 25. Itina, M.A. History of Steppe Tribes of Southern Aral Sea Region (2–1 Ka BP) (Nauka, 1977). 26. Lamberg-Karlovsky, C. C. The Bronze Age khanates of Central Asia. Antiquity 68, 398–405 (1994). 27. Brunet, F. Pour une nouvelle étude de la culture néolithique de Kel’teminar. Ouzbékistan. Paléorient 31, 87–105 (2005). 28. Masson, V. M. The Bronze Age in Khorasan. Hist. Civiliz. Cent. Asia 1, 225 (1999). 29. Ranov, V. A. Hissar culture - Neolithic mountain regions of Central Asia. in Stone Age of Northern, Middle and Eastern Asia 27–28 (Nauka, 1985). 30. Yablonsky, L. T. Kelteminar craniology. Intra-group analysis. Sov. Ethnogr., Mosc., USSR Acad. Sci. 2, 127–140 (1985). 31. Vinogradov, A. V. Drevnie okhotniki i rybolovy Sredneaziatskogo Mezhdurechija (Former hunters and fishermen of Central Asian Mesopotamia) (Nauka (Science), 1981a). 32. Vinogradov, A. V. in Kul ‘tura i iskusstvo drevnego Khorezma (Culture and Art of Ancient Khoresam) (eds Itina, M. A., Raporort, J. A. et al.) pp. 88–98 (1981b). 33. Harris, D. R., Origins of Agriculture in Western Central Asia https://doi. org/10.9783/9781934536513 (Univ. of Pennsylvania, 2010). 34. Dolukhanov, P. M. The ecological prerequisites for early farming in southern Turkmenia. Sov. Anthropol. Archeol. 19, 359–385 (1981). 35. Lisitsina, G. N. in The Bronze Age Civilization of Central Asia: Recent Soviet Discoveries (ed. Kohl, P. L.) pp. 350–358 (M. E. Sharpe, 1981). 36. Larkum, M. in Origins of Agriculture in Western Central Asia: An Environmental–Archaeological Study (ed. Harris, D.) pp. 142–149 (Univ. of Pennsylvania Museum of Archaeology and Anthropology, 2010). 37. Ranov, V. A. & Korobkova, G. F., Tutkaul – multilayered settlement site of the Gissar culture in southern Tajikistan. Sov. Archaeol. 133–147 (1971). 38. Islamov, U. I., Obishirian Culture (FAN, 1980). 39. Fedorchenko, A. Y. et al. Personal ornament production technology in the early Holocene complexes of western Central Asia: insights from Obishir-5. Archaeol., Ethnol. Anthropol. Eurasia 46, 3–15 (2018). 40. Szymczak, K. & Khudzhanazarov, M., Exploring the Neolithic of the Kyzyl-Kums. Ayakagytma “The Site” and other collections. Swiatowit Suppl. Ser. P: Prehistory and Middle Ages 11. Central Asia–Prehist. Stud (2006). 41. Buckley, M., Collins, M., Thomas-Oates, J. & Wilson, J. C. Species identification by analysis of bone collagen using matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry. Rapid Commun. Mass Spectrom. 23, 3843–3854 (2009). 42. Kerven, C., Steimann, B., Ashley, L., Dear, C. & ur Rahim, I. Pastoralism and Farming in Central Asia’s Mountains: a Research Review (Univ. of Central Asia, 2011). 43. Lieberman, D. E. Life history variables preserved in dental cementum microstructure. Science 261, 1162–1164 (1993). 44. Klevezal, G. A. Recording Structures of Mammals: Determination of Age and Reconstruction of Life History (A. A. Balkema, 1996). 45. Diekwisch, T. G. The developmental biology of cementum. Int. J. Dev. Biol. 45, 695–706 (2001). 46. Klevezal, G. & Kleinenberg, S. Age determination of mammals from annual layers in teeth and bones of mammals. Trans. from Russian by the Israel Program for Scientific Translations, Jerusalem (1969). 47. Grue, H. & Jensen, B., Review of the formation of incremental lines in tooth cementum of terrestrial mammals [age determination, game animal, variation, sex, reproductive cycle, climate, region, condition of the animal]. Dan. Rev. Game Biol. 11 (1979). 48. Stock, S. R. et al. Cementum structure in Beluga whale teeth. Acta Biomater. 48, 289–299 (2017). 49. Gordon, B. C. Of Men and Reindeer Herds in French Magdalenian Prehistory https://doi.org/10.30861/9780860545040 (BAR Publishing, 1988). 50. Pike-Tay, A. Red Deer Hunting in the Upper Paleolithic of South-West France: a Study in Seasonality (BAR Oxford, 1991). 51. Burke, A. & Castanet, J. Histological observations of cementum growth in horse teeth and their application to archaeology. J. Archaeol. Sci. 22, 479–493 (1995). 52. Frachetti, M. D. Multiregional emergence of mobile pastoralism and nonuniform institutional complexity across Eurasia. Curr. Anthropol. 53, 2–38 (2012). 53. Stiner, M. C. et al. A forager–herder trade-off, from broad-spectrum hunting to sheep management at Aşıklı Höyük, Turkey. Proc. Natl Acad. Sci. U. S. A. 111, 8404–8409 (2014). 54. Demirci, S. et al. Mitochondrial DNA diversity of modern, ancient and wild sheep (Ovis gmelinii anatolica) from Turkey: new insights on the evolutionary history of sheep. PLoS ONE 8, e81952 (2013). 55. Meadows, J. R. S., Hiendleder, S. & Kijas, J. W. Haplogroup relationships between domestic and wild sheep resolved using a mitogenome panel. Heredity 106, 700–706 (2011). 56. Cai, D. et al. Early history of Chinese domestic sheep indicated by ancient DNA analysis of Bronze Age individuals. J. Archaeol. Sci. 38, 896–902 (2011). 57. Meadows, J. R. S., Cemal, I., Karaca, O., Gootwine, E. & Kijas, J. W. Five ovine mitochondrial lineages identified from sheep breeds of the Near East. Genetics 175, 1371–1379 (2007). 58. Bruford, M. W. & Townsend, S. J., in Documenting Domestication: New Genetic and Archaeological Paradigms (eds Zeder, M. A., Bradley, D. G. & Emshwiller, E. A.) pp 306–316 (Univ. of California Press, 2006). 59. Arbuckle, B. S. & Atici, L. Initial diversity in sheep and goat management in Neolithic south-western. Asia. Levant. 45, 219–235 (2013). 60. Hesse, B. Slaughter patterns and domestication: the beginnings of pastoralism in western Iran. Man 17, 403–417 (1982). 61. Fijn, N., Living with Herds: Human–Animal Coexistence in Mongolia (Cambridge Univ. Press, 2011). NATuRE HuMAN BEHAviOuR | www.nature.com/nathumbehav Content courtesy of Springer Nature, terms of use apply. Rights reserved ARTICLES NATURE HUMAN BEHAVIOUR 62. 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 (2017). 63. Jerardino, A., Fort, J., Isern, N. & Rondelli, B. Cultural diffusion was the main driving mechanism of the Neolithic transition in Southern Africa. PLoS ONE 9, e113672 (2014). 64. van Klinken, G. J. Bone collagen quality indicators for palaeodietary and radiocarbon measurements. J. Archaeol. Sci. 26, 687–695 (1999). 65. Reimer, P. J. et al. The IntCal20 Northern Hemisphere radiocarbon age calibration curve (0–55 cal kBP). Radiocarbon 62, 725–757 (2020). 66. Ramsey, C. B. Bayesian analysis of radiocarbon dates. Radiocarbon 51, 337–360 (2009). 67. Wintle, A. G. & Prószyńska, H. TL dating of loess in Germany and Poland. PACT 9, 547–554 (1983). 68. Fedorowicz, S. et al. Loess–paleosol sequence at Korshiv (Ukraine): chronology based on complementary and parallel dating (TL, OSL), and litho-pedosedimentary analyses. Quat. Int. 296, 117–130 (2013). 69. Frechen, M. Systematic thermoluminescence dating of two loess profiles from the Middle Rhine Area (FRG). Quat. Sci. Rev. 11, 93–101 (1992). 70. Behrensmeyer, A. K. Taphonomic and ecologic information from bone weathering. Paleobiology 4, 150–162 (1978). 71. van Doorn, N. L., Hollund, H. & Collins, M. J. A novel and non-destructive approach for ZooMS analysis: ammonium bicarbonate buffer extraction. Archaeol. Anthropol. Sci. 3, 281 (2011). 72. Welker, F. et al. Palaeoproteomic evidence identifies archaic hominins associated with the Châtelperronian at the Grotte du Renne. Proc. Natl Acad. Sci. U. S. A. 113, 11162–11167 (2016). 73. Rendu, W. Hunting behavior and Neanderthal adaptability in the Late Pleistocene site of Pech-de-l’Azé I. J. Archaeol. Sci. 37, 1798–1810 (2010). 74. Stutz, A. J. Polarizing microscopy identification of chemical diagenesis in archaeological cementum. J. Archaeol. Sci. 29, 1327–1347 (2002). 75. Geusa, G. et al. in Osteodental Biology of the People of Portus Romae (Necropolis of Isola Sacra, 2nd-3rd Cent. AD) (eds Bondioli, L. & Macchiarelli, R.) (Roma, 1999). 76. Lieberman, D. E., Deacon, T. W. & Meadow, R. H. Computer image enhancement and analysis of cementum increments as applied to teeth of Gazella gazella. J. Archaeol. Sci. 17, 519–533 (1990). 77. Gorgé, O. et al. Analysis of ancient DNA in microbial ecology. Methods Mol. Biol. 1399, 289–315 (2016). 78. Rohland, N., Harney, E., Mallick, S., Nordenfelt, S. & Reich, D. Partial uracil-DNA-glycosylase treatment for screening of ancient DNA. Philos. Trans. R. Soc. Lond. B Biol. Sci. 370, 20130624 (2015). 79. Kircher, M., Sawyer, S. & Meyer, M. Double indexing overcomes inaccuracies in multiplex sequencing on the Illumina platform. Nucleic Acids Res. 40, e3 (2012). 80. Ginolhac, A., Rasmussen, M., Gilbert, M. T. P., Willerslev, E. & Orlando, L. mapDamage: testing for damage patterns in ancient DNA sequences. Bioinformatics 27, 2153–2155 (2011). 81. Li, H. & Durbin, R. Fast and accurate long-read alignment with Burrows– Wheeler transform. Bioinformatics 26, 589–595 (2010). 82. Purcell, S. et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am. J. Hum. Genet. 81, 559–575 (2007). 83. Li, X. et al. Whole-genome resequencing of wild and domestic sheep identifies genes associated with morphological and agronomic traits. Nat. Commun. 11, 2815 (2020). 84. Li, H. et al. The sequence alignment/map format and SAMtools. Bioinformatics 25, 2078–2079 (2009). 85. DePristo, M. A. et al. A framework for variation discovery and genotyping using next-generation DNA sequencing data. Nat. Genet. 43, 491 (2011). 86. Mallick, S. et al. The Simons Genome Diversity Project: 300 genomes from 142 diverse populations. Nature 538, 201–206 (2016). 87. Kumar, S., Stecher, G., Li, M., Knyaz, C. & Tamura, K. MEGA X: molecular evolutionary genetics analysis across computing platforms. Mol. Biol. Evol. 35, 1547–1549 (2018). 88. Stecher, G., Tamura, K. & Kumar, S. Molecular evolutionary genetics analysis (MEGA) for macOS. Mol. Biol. Evol. 37, 1237–1239 (2020). 89. Tamura, K. & Nei, M. Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Mol. Biol. Evol. 10, 512–526 (1993). 90. Murphy, D. J. People, Plants and Genes: The Story of Crops and Humanity (Oxford Univ. Press, 2007). 91. Lv, F.-H. et al. Mitogenomic meta-analysis identifies two phases of migration in the history of eastern Eurasian sheep. Mol. Biol. Evol. 32, 2515–2533 (2015). Acknowledgements The authors thank D. Paul and S. Palstra for performing radiocarbon dating of tooth enamel, and E. Rannamäe for assistance with manuscript preparation. Cementum analyses were funded through the CemeNTAA project, via the French National Agency for Research (ANR-14-CE31-0011). Geological investigations were supported by the National Science Center, Poland (grant no. 2018/29/B/ST10/00906). Sampling for ZooMS, DNA and radiocarbon analysis (Golden Valley Laboratory) and lithic analysis of Obishir V were supported by RSF project no. 19-78-10053, ‘The emergence of food-producing economies in the high mountains of interior Central Asia’. Ancient DNA analyses were conducted with the support of the palaeogenomic platform from the UMR5199 PACEA Universite de Bordeaux and the European Research Council under the European Union’s Horizon 2020 research and innovation programme under grant agreement no. 804884-DAIRYCULTURES. The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript. Author contributions W.T.T.T. and S.Sh. designed the research, collected data, conducted analysis and wrote the manuscript. M.P., C.P., A.A., W.R., C.J., T.H., and C.W. collected data, conducted analysis and helped to write the manuscript. M.T.K., G.B., S.Sc., G.H., R.Sp., R.St., J.M., A.S., S.F., L.O., K.D. and A.K. collected data, conducted analysis and assisted in data interpretation. All authors reviewed the manuscript. Competing interests The authors declare no competing interests. Additional information Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s41562-021-01083-y. Correspondence and requests for materials should be addressed to W.T.T.T. or S.S. Peer review information Nature Human Behaviour thanks Suzanne Birch, Laurent Frantz and Eve Rannamae for their contribution to the peer review of this work. Reprints and permissions information is available at www.nature.com/reprints. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. © The Author(s), under exclusive licence to Springer Nature Limited 2021 NATuRE HuMAN BEHAviOuR | www.nature.com/nathumbehav Content courtesy of Springer Nature, terms of use apply. Rights reserved Last updated by author(s): Sep 11, 2020 Reporting Summary Nature Research wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency in reporting. For further information on Nature Research policies, see our Editorial Policies and the Editorial Policy Checklist. nature research | reporting summary Corresponding author(s): William Taylor Statistics For all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section. n/a Confirmed The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly The statistical test(s) used AND whether they are one- or two-sided Only common tests should be described solely by name; describe more complex techniques in the Methods section. A description of all covariates tested A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons A full description of the statistical parameters including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient) AND variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals) For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted Give P values as exact values whenever suitable. For Bayesian analysis, information on the choice of priors and Markov chain Monte Carlo settings For hierarchical and complex designs, identification of the appropriate level for tests and full reporting of outcomes Estimates of effect sizes (e.g. Cohen's d, Pearson's r), indicating how they were calculated Our web collection on statistics for biologists contains articles on many of the points above. Software and code Policy information about availability of computer code Data collection No software used Data analysis Radiocarbon dates Bayesian analysis of radiocarbon dates was performed in OxCal v. 4.4 (https://c14.arch.ox.ac.uk/oxcal.html) ZooMS Mass spectrometry identifications were performed using the open-source software MMass (http://www.mmass.org/) Cementum analysis Measurements of the increments were conducted using ImageJ (https://imagej.nih.gov/ij). PCR amplification Sequences were visually inspected and manually curated, assembled and aligned using Geneious software suite v9.1.8 (https:// www.geneious.com). April 2020 Library construction and sequencing Sequence authenticity was estimated on libraries subjected to either a partial or no USER treatment using MapDamage. We drew alleles at random using PileupCaller (https://github.com/stschiff/sequenceTools), rendering the individuals from the dataset homozygous for each locus. The two files were then merged using Plink v1.9. We then conducted principal component analysis using smartPCA implemented in EIGENSOFT. We mapped our filtered reads to both the Ovis aries genome assembly (international Sheep Genome consortium build Oarv3.1) and the Capra hircus genome assembly (ARS1) using BWA. We produced a multiple genome alignment of our newly reconstructed sequences with Geneious software suite v9.1.8. Gaps were removed from the alignment and a Maximum Parsimony tree was built using MEGA version X.Phylogenetic trees were visualized and manipulated in FIGTREE v1.4.1 (http://tree.bio.ed.ac.uk/software/figtree/). 1 Content courtesy of Springer Nature, terms of use apply. Rights reserved For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors and reviewers. We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Research guidelines for submitting code & software for further information. Data Policy information about availability of data All manuscripts must include a data availability statement. This statement should provide the following information, where applicable: - Accession codes, unique identifiers, or web links for publicly available datasets - A list of figures that have associated raw data - A description of any restrictions on data availability Sequencing data can be found on the European Nucleotide Archive: [to be added upon upload] nature research | reporting summary Graphs All graphs were produced and edited in R (https://cran.r-project.org/) and Adobe Photoshop Field-specific reporting Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection. Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pdf Behavioural & social sciences study design All studies must disclose on these points even when the disclosure is negative. Study description This study is a mixed-methods, interdisciplinary analysis of archaeological materials from the site of Obishir V, in Kyrgyzstan. All available archaeological materials (as of 2017) were selected for the analysis, of which the best preserved specimens (n=6) were selected for genomic sequencing. Research sample All available skeletal remains from 2017 and early excavations were chosen for study. For ZooMS analysis, all materials from one square of the excavation (R8) were selected for study. Sampling strategy All available materials from the archaeological assemblage were chosen for study. Data collection Researchers were blind to any hypothesis during data collection, which did not involve human participants. Excavation Excavations at Obishir V site carried out between 2015-2019 by a joint Russian-Kyrgyzstan archaeological expedition. In 2015–2016, researchers working at the site excavated an area of 8 m2 adjoining the 1968–1969 excavation area (Supplementary Appendix A). Each artifact (lithic and faunal remains) greater than 1 cm in size was individually piece-plotted with 1 mm accuracy using a Leica Total Station TS02 plus in tandem with the Trimble software EDM Mobile, and documented along with contextual information ( orientation, dip, inclination, find category, stratigraphic level, etc.). Fragments smaller than this were collected in bulk for each 0.25 m2 area. Features (natural or anthropogenic) were recorded with the total station using outlines, surface points and breaklines. Large surfaces, such as walls of old excavation pits, sections and archaeological horizons with numerous artefacts were also documented using photogrammetry and 3D-scanning. Digital records were supplemented by field drawings and notes. All sediment obtained during the excavation was wet-screened with mesh of a size of 2.0 mm, and finds were documented according to excavated square, layer and depth. Stratigraphic units were defined during the excavation based on grain size composition, consistency, color, presence of erosional surfaces and accumulations of limestone clasts. Zooarchaeology and Zooarchaeology by Mass Spectrometry For each specimen, we conducted taphonomic analysis and morphology-based species identifications using comparative faunal collections held by the Russian Academy of Sciences-Siberia in their field station in Aidarkyen, Kyrgyzstan, and attempted refits for all specimens to control for issues of fragmentation. We recorded taphonomic indicators such as rodent and carnivore damage, root etching, and weathering (70), along with evidence for anthropogenic modification such as spiral fracturing, cut marks, and burning via visual inspection (Supplementary Appendix C). In addition, we used collagen fingerprinting (or Zooarchaeology by Mass Spectrometry, also known as ZooMS) to taxonomically identify a subset of this material, all derived from 2017 excavations. We April 2020 Thermoluminescence dating During the 2017 field season, two sediment samples were collected from the southern portion of the western excavation profile of layer 5 at a depth of 295 and 335 cm below the surface. We measured deposit moisture in each sample, and after drying determined dose rate (DR) using a MAZAR gamma spectrometer. Concentrations of 226Ra, 228Th, 40K in each sample were obtained from twenty measurements lasting 2000 s each. We established equivalent dose (ED) on the 63-80 mm polymineral fraction, after 10% HCl and 30% H2O2 washing and UV optical bleaching. The samples were irradiated with 20 Gy, 30 Gy, 40 Gy, 50 Gy and 100 Gy, doses from 60Co gamma source. Before measurement, we heated the samples at 140 degrees for 3 hours. A sample pre-treated in this way was used to determine the equivalent dose (ED) by the TL multiple-aliquot regenerative technique (67), according to the description published by Fedorowicz et al. (68). Curve registration was performed on RA'94 (Mikrolab) thermoluminescence reader, coupled with EMI 9789 QA photomultiplier. Finally, we calculated TL age according to Frechen (69). 2 Content courtesy of Springer Nature, terms of use apply. Rights reserved Cementum analysis First, all analyzed teeth were 3D scanned using a MicroCT scanner, and washed in a 90% ethanol solution. A cast of the occlusal surface was retained for later micro-wear analyses. Thin sections were prepared following recommendations by Rendu (73). Teeth were embedded in epoxy within a vacuum chamber (at 0.2 bars pressure). They were then cut at a thickness of 250μm using an Isomet5000 automatic saw, and ground to a thickness of 100μm with a PressiCube grinding machine. For each tooth, four independent thin sections were prepared and analyzed independently. Observations were conducted under polarized light and polarized/analyzed light both with and without the insertion of a lambda plate (74), using a Leica 2500P microscope equipped with a Leica DC 120 camera. P DNA Extraction DNA was extracted from 6 teeth using silica-based purification in a dedicated laboratory for ancient DNA at PACEA (UMR 5199, University of Bordeaux). We incubated between 100-250 mg of bone powder in a 1 mL extraction buffer ( 0.5 M EDTA, 0.25 M Na2HPO4, 50 μg/mL proteinase K) for 48-72 h at 37°C. DNA was purified from the collected supernatant using a method adapted from the QIAquick MinElute purification kit protocol (Qiagen). nature research | reporting summary analyzed all bone specimens from Square R8, Layers 2.2, 3, and 4, as well as all bone specimens from Square R8, Layer 1.3 using ZooMS. To do so, we demineralized 10 mg of bone using 50mM ammonium bicarbonate buffer (Sigma-Aldrich) produced in UHQ water and pH adjusted to 8.0 using Ammonium Hydroxide (1M), following the protocol outlined by (71). The extracted collagen was digested into component peptides using trypsin (Pierce), and was purified using Pierce C18 tips (Thermo Scientific), eluting in 50% acetonitrile (Sigma-Aldrich) and 0.1% TFA (Sigma-Aldrich). Samples were then spotted on Bruker AnchorChip with Bruker Peptide Calibration Standard in calibration spots directly neighbouring the samples. Mass spectrometric analysis was conducted using a Bruker Autoflex Speed LRF MALDI-TOF in the dedicated ZooMS Laboratory at the Max Planck Institute for the Science of Human History in Jena, Germany. PCR amplification Purified DNA was amplified by qPCR using a LightCycler® 96 Instrument (Roche Life Science). Primer pairs were designed to amplify 2 regions of the mitochondrial DNA, a 130 bp region from cytB and 170bp from the hypervariable region (Supplementary Appendix F: list of primers used and product sizes). We obtained each PCR product at least twice, conducting standard negative control to ensure authenticity of the results. We used 1 μL of extract in 10 μL reaction with 1X FastStart Essential DNA Green Master and 1μM of each primer. Products were sequenced by capillary electrophoresis at Genewiz (Leipzig, Germany). Library construction and sequencing We used DNA extracts to build double-stranded Illumina libraries following the protocol described in Gorge et al (77). We sequenced whole genome shotgun libraries for ~ 500 000 reads on an Illumina NextSeq 500 at the Institut de Recherche Biomédicale des Armées in Brétigny-sur-Orge, France, to assess DNA preservation. DNA extracts of five samples produced in Bordeaux were sent to the dedicated clean room facilities of the Max Planck Institute for the Science of Human History. There, 20 ul of the extract were used to build double stranded UDG-half libraries (78). The resulting genetic libraries were double indexed and further amplified (79) Finally, all five samples were shotgun sequenced for ~20M reads on an Illumina HiSeq run with a 75-cycles single-end configuration. Timing Archaeological research was conducted in the summer of 2017, with ZooMS conducted in Fall 2017, and radiocarbon dating and DNA sequencing from early 2018 through spring of 2020. Data exclusions No data were excluded from the analyses. Non-participation No human participants were used in the study Randomization No human participants were used in the study, and no data were allocated into groups Reporting for specific materials, systems and methods We require information from authors about some types of materials, experimental systems and methods used in many studies. Here, indicate whether each material, system or method listed is relevant to your study. If you are not sure if a list item applies to your research, read the appropriate section before selecting a response. Materials & experimental systems Methods n/a Involved in the study n/a Involved in the study Antibodies ChIP-seq Eukaryotic cell lines Flow cytometry Palaeontology and archaeology MRI-based neuroimaging Animals and other organisms Human research participants Dual use research of concern April 2020 Clinical data Palaeontology and Archaeology Specimen provenance Specimens all come from the archaeological site of Obishir V in Aidaryken, Kyrgyzstan, and research was conducted under Permit #0040/02-05 issued by the The Field Committee of the Institute of History and Cultural Heritage of the National Academy of Sciences, Kyrgyz Republic on 19 April, 2017. Specific provenance for each object is provided in the supplementary material. 3 Content courtesy of Springer Nature, terms of use apply. Rights reserved All specimens are stored in the collections at the Institute for Archaeology and Ethnography, Siberian Branch of the Russian Academy of Sciences, Novosibirsk, Russian Federation Dating methods Nine radiocarbon samples on identified and unidentified animal bone and tooth samples (n = 7) and charcoal (n = 2) were selected from various depths across the stratigraphic unit of layers 2-4, and submitted for radiocarbon dating at the Oxford Radiocarbon Accelerator Laboratory, the Center for Isotope Research at the University of Groningen, and the Golden Valley Laboratory at Novosibirsk. A portion of tooth 20-1 was submitted to the University of Arizona Accelerator Mass Spectrometry Laboratory. Collagen was obtained from 20-1 dentin using Acid-Base-Acid (ABA) pretreatment, gelatinization, 0.45 micron filtration, and ultrafiltration. Quality control parameters: collagen yield (7.1%); carbon yield (42.9%); CN ratio (3.2); and d13C (19.7 +/- 0.1 per mil), indicated good preservation (64). All dates were calibrated in OxCal using the INTCAL13 calibration curve (65). We performed Bayesian stratigraphic phase modeling (using ordered phases and combining Layers 2-4) in the software OxCal with a uniform prior using all available radiocarbon dates from Obishir V, and two thermoluminescence dates from Layer 5 (using OxCal’s C_date function), following methods outlined by Ramsey (66). We also modeled the cultural occupation of Layer 2 separately, using only bone and tooth samples using a uniform prior. The models provided good agreement, and repeating the analysis using a general outlier model did not significantly alter estimated parameters. OxCal code used in the analysis, along with human tooth dating methods are outlined in Supplementary Appendix B. Tick this box to confirm that the raw and calibrated dates are available in the paper or in Supplementary Information. Ethics oversight nature research | reporting summary Specimen deposition All research was conducted in concurrence with ethical guidelines of the Max Planck Institute for the Science of Human History, Jena, Germany. Note that full information on the approval of the study protocol must also be provided in the manuscript. April 2020 4 Content courtesy of Springer Nature, terms of use apply. Rights reserved Terms and Conditions Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”). Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for smallscale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial. These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription (to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will apply. We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as detailed in the Privacy Policy. While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may not: 1. use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access control; 2. use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is otherwise unlawful; 3. falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in writing; 4. use bots or other automated methods to access the content or redirect messages 5. override any security feature or exclusionary protocol; or 6. share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal content. In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue, royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any other, institutional repository. These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved. To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law, including merchantability or fitness for any particular purpose. Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed from third parties. If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not expressly permitted by these Terms, please contact Springer Nature at onlineservice@springernature.com