The Kelteminar culture, located in modern Uzbekistan around the Kyzyl Kum desert, represents the colonisation of the Tugias forests and typical steppe close to river deltas and lakes, with a technical tradition derived from the local Mesolithic background. The early stage of the culture (ca. 7th-6th millennium BC), concentrated on the region of the Zeravshan Valley, shows several lithic production systems—including microblades, bladelets, and blades—with at least two techniques, the most common being controlled indirect percussion, the other being the bullet-shaped core method, but there is little evidence for pressure knapping techniques. Therefore, it was probably Mesolithic Ural groups in direct contact with this culture the ones who introduced the technique (Brunet 2012).
During the second stage (5th–4th millennium BC), a significant development of blade and bladelet production is seen, requiring more elaborated skills. Among the tools seen, the Kelteminar arrowhead and the horned trapeze show a wide distribution in parts of Central Asia (Kazakhstan, Russia, Uzbekistan, Turkmenistan, Northern Afghanistan), suggesting a symbolic value, and thus a way to define social identity. During that period, pressure knapping technique becomes prominent in blade production, with new relationships—evidenced by decoration in pottery—probably being developed in Chorasmia with agropastoral communities of southern Turkmenistan (Brunet 2012).
The Atbasar culture (ca. 5th–4th millennium BC) of the forest-steppe zone of Northern Kazakhstan also developed from the local Mesolithic microblade production using bullet-shaped cores as pressure knapping technique. The introduction of few regular blades and new formal tools can be seen during this period, such as points, trapezes, triangular arrowheads with a basal notch, bifacial pieces, and leaf-shaped bifacial points, possibly with different functional and socioeconomic contexts. This and the introduction of pottery with incised or combed decoration, and the domestic horse at the end of the period, suggest migrations in northern Kazakhstan at the same time as those seen among similar post-Mesolithic cultures of eastern Kazakhstan, Altai, and eastern Siberia (Brunet 2012).
The Botai-Tersek culture (ca. 3700–3000 BC) probably emerged from groups of Atbasar foragers in the steppes of northern Kazakhstan who developed a specialised economy as horse riders who hunted essentially horses. Their main diet consisted preferentially of horses, but it included also wild animals like large bovids, elks, deers, bears, etc. The small temporary settlements in the steppes, the evidence for herd-driving hunting techniques, as well as the management and transport of great quantities of horses, together with evidence for bitted and ridden horses prove the appearance of horseback riding ca. 3700–3500 in the northern Kazakh steppes (Brunet 2012).
The Hissar culture (ca. 7th–4th millennium BC) shows a continuation of Mesolithic material culture related to the Yubetsu tradition. As in the Atbasar culture, the introduction of new Neolithic components is seen in the late stage, with blade production using indirect percussion, flake production by direct percussion (hard hammer), and the presence of trapezes and polished axes, linked to domestic activities related to leather, skin, and woodworking (Brunet 2012).
To the east, Neosiberians—from which contemporary Siberians derive—replaced the previous Ancient Palaeosiberian-like populations ca. 9000–2000 BC, restricting their AP-like ancestry to north-east Siberia, represented by an individual from Ol’skaya, Magadan (ca. 1000 BC), who closely resembles present-day Koryaks and Itelmens. In the Cis-Baikal area, thirteen Early Neolithic hunter-gatherers from Shamanka (ca. 5200–4200 BC) are representatives of AEA ancestry, closely related to individuals from the Devil’s Gate Cave (ca. 6000–5500 BC). Among reported haplogroups, there are seven samples likely all N1a2-L666, and one C2a1a1a-F3918 (Sikora et al. 2018).
Cis-Baikal populations are replaced likely during the Late Neolithic (Moussa et al. 2016), with samples from Early Bronze Age (ca. 2200–1800 BC) evidencing an almost full population replacement with a resurgence of AP ancestry (up to 50%)—probably from a population migrating from the east and north—and influence from West Eurasian steppe ANE ancestry (ca. 10%) in the Altai region, represented by BA individuals from Afanasevo. Reported haplogroups are all Q1a2a-L712, with one Q1a2a1-L715 and one Q1a2a1c probably suggesting these subclades as those expanding in the EBA (de Barros Damgaard, Martiniano, et al. 2018).
AP ancestry is also found in modern Kets (ca. 40%), speaking a Yeniseian language, with genetic links to Palaeoeskimos, thus connecting Yeniseian genetically with Na-Dene-speaking peoples. The main reported haplogroup of Na-Dene peoples is Q1a2-M25 (ca. 90%), which suggests that its lineages expanded with ancestral Dene-Yeniseian speakers through north Eurasia, most likely during the Late Neolithic (Sikora et al. 2018). Yeniseians, on the other hand, belong to haplogroup Q1b1a-L54 (formed ca. 16100 BC, TMRCA ca. 14000 BC), which may point to the dispersal of certain Palaeosiberian languages with these particular lineages during the Palaeolithic (Huang et al. 2017).
It is unclear which lineage may have spread with AEA ancestry through northern Siberia, along the inner Asian Palaeolithic EHG–ANE–AEA cline, and when, although possibly some N1a1-Tat subclade (formed ca. 13900 BC, TMRCA ca. 9800 BC). In particular, the split into an eastern N1a1a2-Y23747 (TMRCA ca. 4500 BC), found among modern Japanese and Chinese, and a western N1a1a1-F1419 (TMRCA ca. 8800 BC), found among Khakassians and northern Indians, suggests a split around Lake Baikal.
Similarly, its subclade N1a1a1a-L708 (formed ca. 8800 BC, TMRCA ca. 5400 BC) shows a wide Northern Eurasian distribution in the regions east of the Urals. In particular, although basal N1a1a1a-L708 subclades can be found today around the Urals without a particular linguistic connection, its subclade N1a1a1a1a-L1026/L392 (formed ca. 4300 BC, TMRCA ca. 2900 BC) seems to represent expansions through Siberia from around Lake Baikal, often associated with Altaic-speaking peoples. This ancient Altaic connection is reflected in the finding of hg. N1a1a1a1a4-M2019 (formed ca. 4300 BC, TMRCA ca. 1700 BC) through Arctic populations, from Estonians to Tungusic speakers from Yakutia, from Chinese to Hungarians (see §viii.21.1. Yukaghirs).
Different expansions of these lineages include, among modern Northern Eurasian populations: N1a1a1a1a2-Z1936 (TMRCA ca. 2300 BC) connects N1a1a1a1a2a1c-Y13850 among peoples from the Trans-Urals region (see §viii.17. Ugrians and Samoyeds) with N1a1a1a1a2a-Z1934 in Palaeo-Arctic populations of the Cis-Urals (see §viii.16.1. Saami and Laplandic peoples), possibly through a Northern Eurasian forest–taiga route. SNP Y6058 (formed ca. 5300 BC, TMRCA ca. 2900 BC) connects hg. N1a1a1a1a3-Y16323 (TMRCA ca. 2900 BC) of Mongolic speakers and Chukchi of (see§viii.21.2. Turkic peoples and Mongols) with Mordvinic and later Balto-Finnic speakers of hg. N1a1a1a1a1-CTS10760 (TMRCA ca. 2100 BC), possibly through more southern forest-steppe–steppe routes of expansion, given the appearance of N1a1a1a1a1c-B479 among Tungusic speakers and Nenets (see §viii.15. Mordvins and Mari-Permians and §viii.16.2. Baltic Finns).
Late northern Siberian nomadic peoples close to the Arctic region are known to be easily subject to exogamy practices due to their mobility, and are thus associated with plurilingualism and eventual acculturation within few generations of admixture (Karafet et al. 2018). Therefore, it will remain unclear to what extent Palaeo-Laplandic from Lovozero (see §viii.16. Saami and Baltic Finns) or the language of northern nomadic peoples who adopted Mari-Permic, Ugric or Samoyedic languages remained related to Chukotko-Kamchatkan family thousands of years later, or spoke West Siberian languages like Yeniseian, or other Eurasiatic dialects.
Three samples from the west Siberian forest zone (ca. 6200–4000 BC) are representatives of a mixture of ancestry called “west Siberian hunter-gatherer” (WSHG) ancestry, made up of EHG (ca. 30%), ANE (ca. 50%), and AEA-related ancestry (ca. 20%). This ancestry was also present in the southern steppe and in Turan (BMAC), and formed ca. 80% of the ancestry of an early 3rd millennium BC agropastoralist from Dali, Kazakhstan, contributing to multiple outliers from 2nd millennium sites in Kazakhstan and Turan (Narasimhan et al. 2018).
The widespread presence of this ancestry in west Siberia is compatible with its association with hunter-gatherers of Kelteminar and other central Asian sub-Neolithic cultures (Narasimhan et al. 2018). The presence of an ancestral cline EHG–ANE–AEA ancestry in inner Asia is also supported by the west–to–east gradient formed in the PCA. This ancestral WSHG ancestry, separated from other ancient and present-day populations, is found in Botai (ca. 3600–3100 BC), Okunevo (ca. 2500–1800 BC), central steppe EMBA samples from Sjolpan (ca. 2550 BC), Takhirbai and Gregorievka (ca. 2150 BC), as well as Cis-Baikal EN and EBA populations (de Barros Damgaard, Martiniano, et al. 2018).
Of the three Botai samples published, one (ca. 3600–3100 BC) R1b1a1a-M73, while another (ca. 3300–3100 BC) shows haplogroup N-M231 (Narasimhan et al. 2018). Another individual of hg. R1b1a1-P297 is found in the Bol'shemysskaya culture (ca. 4500–3500 BC). The presence of R1b1a1-P297, and R1b1a1a-M73 in particular, is linked to the previous expansion of the North-Eastern Technocomplex, during the Early and Middle Mesolithic (see §ii.1. Eurasians), and thus likely associated with the creation of an ancestral Altaic-speaking population in inner Asia closely related to peoples with WSHG ancestry (see §viii.21.1. Yukaghirsand §viii.21.2. Turkic peoples and Mongols).