Study area

The Baikal-Patom Highlands, part of the Baikal Mountain Region, are located north of the Stanovoy Ridge, to the east and southeast of the Central Siberian Plateau and west of the Olyokma-Chara Highlands (Zolotarev, Reference Zolotarev1974a). They extend in the northeastern direction and have a length of about 700 km and maximum width (in the northeast) of 330 km. The highest elevation is 2578 m above sea level in the southwest at Inyaptuk Peak. The uplands boundary is clearly defined by large erosion–tectonic valleys. The stepped slopes of the uplands are intensively eroded in some places, separating them from the Olyokma-Chara Highlands and the Siberian Plateau. The border with the Stanovoy Ridge is represented by the northwestern slopes of the ridges of Stanovoy Ridge (Kodar, Verhneangarsky, and Kichersky) towering over the adjacent sections of the Baikal-Patom Highlands at 150–200 m and the erosion–tectonic fragments of river valleys.

The main drainage system is the Vitim River and its tributaries. The Vitim River conventionally separates the Baikal-Patom Highlands into the Patom and Baikal parts. The study area includes the lower part of the Vitim valley. The average long-term water flow near Bodaybo city is 1521 m³/s. A vertical pattern in vegetation is well expressed from mountain tundra in the mountain ranges above 2100 m to larch (Larix) forests with cedar (Cedrus) and spruce (Picea) in mid-elevations, and wild rosemary (Rosmarínus officinalis), blueberry (Vaccínium myrtillus), and lingonberry (Vaccínium vitis-idaea) in the mountain taiga belt and along the bottoms of the valleys (Atlas of Irkutsk region, 1962).

Classic representation of terrace formation in the Vitim River valley and tributaries

The Late Pleistocene glaciation of the Baikal-Patom Highlands is one of the most important stages in the formation of the mountain landscape. Under these conditions, large glacier-dammed lakes and associated megafloods arose, developed, and degraded. The sources for paleogeographic reconstructions are glacial, erosion-accumulative topography and deposits, which are rich in faunal and anthracological materials.

In the Late Pleistocene, the Vitim valley was buried by alluvium of different thickness from the periodic discharge of the glacially dammed Lake Vitim (Margold et al., Reference Margold, Jansen, Codilean, Preusser, Gurinov, Fujioka and Fink2018), followed by erosional incisions and the formation of a series of terraces. For this reason, it is difficult to compare the terraces within both the Vitim valley and its neighboring river systems.

Endrikhinsky (Reference Endrikhinsky and Logachev1982) studied the middle part of the Vitim valley and documented a floodplain at 3–4 m and terraces at 8–10 m (T1) and 18–22 m (T2). According to other researchers (Zolotarev, Reference Zolotarev1974b; Tishchenko, Reference Tishchenko and Logachev1982), a large incision occurred after the Marine Isotope Stage (MIS) 4 glaciation, resulting in the formation of 30–40, 18–25, and 7–12 m terraces. There are sometimes deviations in the number and height of terraces depending on the position of the section in the valley and its remoteness from the local basis of erosion. The farther upriver the studied area is located, the lower the height and number of terraces (this is also relevant for the lateral tributaries of the Vitim River). The third and second terraces formed in MIS 3, the first in MIS 2 (Zolotarev, Reference Zolotarev1974a, Reference Zolotarev1974b).

According to Tishchenko (Reference Tishchenko and Logachev1982), the 15–25 m terrace of the Bodaibo River and the 50 m terrace of the Vitim River were formed in MIS 2. The accumulation of sediments at this time took place in a cold and humid climate with a clear trend of significant runoff during floods, which contributed to the accumulation of sediments on the high floodplain. This led to the formation of a 25 m terrace in the lower part of the Bodaibo River, which declines to 15 m in the middle section of the river. Its deposits are characterized by an increased thickness of the floodplain alluvium consisting mainly of fine-grained polymict light gray sands. In the middle stream of the Bodaibo River, the thickness of the flood silts decreases. In the Vitim valley, this terrace is represented downstream of Bodaibo city by surviving fragments almost 50 m above the waterfront (Tishchenko, Reference Tishchenko and Logachev1982). The floodplain alluvium is also represented by fine-grained polymict sand, and the channel alluvium is sandy–pebble–rock. On the surface of the gravel, there is a carbonate crust. It was also noted that the different heights of the same-age terraces can be explained by the prevalence of deep or lateral erosion, distance from the local basis of erosion, deformation by “young” tectonic processes, and association with different geomorphological regions (Endrikhinsky, Reference Endrikhinsky and Logachev1982).

The spore–pollen spectra of the first terrace of Vitim River, near the mouth of the Tamarak River (Fig. 1), contains dwarf birch (Betula nanae), Siberian selaginella (Selaginella sibirica), and burr selaginella (Lycopodium pungens) typical of periglacial vegetation (Kazakevich and Reverdatto, Reference Kazakevich and Reverdatto1972). The time of formation of these deposits is attributed to the Sartan glaciation (MIS 2) (Grichuk and Grichuk, Reference Grichuk and Grichuk1960; Grichuk and Zaklanskaya, Reference Grichuk and Zaklanskaya1985).

On the Baikal-Patom Highlands in MIS 2, the glaciation was most developed in the high mountains and had a mountain-valley character (Sinyugina, Reference Sinyugina1960; Zolotarev, Reference Zolotarev1974b; Kazakevich and Reverdatto, Reference Kazakevich and Reverdatto1972; Tishchenko, Reference Tishchenko and Logachev1982). It has been suggested that an ice cap more than 700 m thick existed within the Baikal part of the mountain range in the southern highlands (Levi et al., Reference Levi, Mats, Kusner, Kirillov, Alakshin, Tolstov, Osipov, Efimova and Bak1998). The center of glaciation in the northern highlands, was the Kropotkin Ridge (Zolotarev, Reference Zolotarev1974b, Reference Zolotarev1982; Tishchenko, Reference Tishchenko and Logachev1982).

Warming at the transition from the Late Pleistocene to Holocene led to the retreat of the glaciers and formation of end, lateral, and bottom moraines. Several attempts at paleogeographic reconstruction were made based on dating of a series of end moraines located at a considerable distance from each other (Levi et al., Reference Levi, Mats, Kusner, Kirillov, Alakshin, Tolstov, Osipov, Efimova and Bak1998; Ineshin, Reference Ineshin2003).

Early paleogeographic reconstructions based on archaeological data

There are previously published paleogeographic reconstructions of the Vitim valley based on archaeological research. The primary archeological sites used in these reconstructions were Bol’shoi Yakor’ I and Kovrizhka I–VI. These studies identified the stages of landscape formation in the Vitim River valley (Mamakansky site) (Belousov et al., Reference Belousov, Ineshin, Sulerzhitsky and Tetenkin2002; Ineshin and Tetenkin, Reference Tetenkin2017; Demonterova et al., Reference Demonterova, Ivanov, Ineshin and Tetenkin2014; Henry and Tetenkin, Reference Henry and Tetenkin2014; Henry et al., Reference Henry, Bezrukova, Teten’kin and Kuz’min2018).

The main points of this reconstruction are as follows:

1. 1. The Vitim paleochannel was located at an altitude of 50–60 m above the modern level of the river before MIS 4.

2. 2. In MIS 4 and subsequent times, erosional incision of the Vitim River occurs. The Mamakan VI site is located in the upper deposits of the Sartan age (MIS 2) at a height of 33 m above the Vitim River.

3. 3. In MIS 2, the formation of terraces (22, 18, and 14 m) occurred, on which the Avdeikha, Kovrizhka II–IV, and Bolshoy Yakor I sites are located.

4. 4. In the first half of the Holocene, the formation and lateral erosion of 9–11 m of the terrace occurred, on which archaeological sites Invalidny I–III, Mamakan I–V, Kovrizhka I, V, and Pavlova are located. The height of the terrace increased due to the formation of bars on its surface.

5. 5. The latest erosional incision of the Vitim River and the accumulation of 8–9 m of floodplain occurred in the Middle to second half of the Holocene (Ineshin and Tetenkin, Reference Tetenkin2017).

In the Vitim valley, the earliest stratified cultural remains at Kovrizhka IV (cultural horizon 6) are dated around 19–18 ka BP (Tetenkin et al., Reference Tetenkin, Henry and Klementyev2017). Radiocarbon-dated remains of snow sheep (Ovis nivicola), elk (Alces alces), and shrub willow (Salix) charcoal are interpreted as representing an open tundra landscape with small islands of forest vegetation, represented by a single piece of larch or spruce charcoal (Henry et al., Reference Henry, Bezrukova, Teten’kin and Kuz’min2018). Faunal remains of a horse (Equus sp.) from the Avdeikha site are also interpreted to represent the existence of the forest-free areas (Mochanov, Reference Mochanov1977, table D). Cultural horizon 6 at Kovrizhka IV contains the only remains of a mammoth (Mammuthus primigenius) found in the archaeological complexes of the Vitim valley, consisting of blades of tusk, from which anthropomorphic figures are made (Tetenkin et al., Reference Tetenkin, Zhmur, Demonterova, Kaneva and Salnaia2018b). The synchroneity of these remnants of mammoth to the episode of human habitation cannot be judged in the absence of direct dates of these remains. However, their presence in the layer is an indicator of the availability of material such as mammoth tusk to people occupying the site during this time.

The expanded paleoecological characteristic of the MIS 2 (15–13.5 ka) is provided from research of the Bol’shoi Yakor’ I site (Ineshin and Tetenkin, Reference Tetenkin2017). Here, 28 species of animals have been identified, including representatives of ornitho-, terio-, and ichthyofauna. Snow sheep, reindeer (Rangifer tarandus), horse, musk ox (Ovibos pallantis), arctic fox (Alopex lagopus), tundra partridge (Lagopus muta), gopher (Spermophilus cf. parryi), and marmot (Marmota sp.) remains are present and represent a tundra environment (Ineshin and Tetenkin, Reference Tetenkin2017). Bison (Bison sp.), wolf (Canis lupus), and red deer (Cervus elaphus) remains are indicators of mixed tundra–forest landscapes. The presence of sable (Martes zibellina) and elk remains indicates the existence of forest refugia. The mountain landscape of the Vitim valley determined the vertical zonal distribution of vegetation. With the general domination of the tundra–steppe landscape with shrubs, isolated locations of forest vegetation existed in the Vitim River Valley and its tributaries on the lower topographic levels: slopes, the first terrace, floodplain, and so on.

In the period 13.2–12.2 ka, the anthracological remains of Kovrizhka III cultural horizon 2 suggest vegetation is represented exclusively by larch, a cold-loving species (Henry and Tetenkin, Reference Henry and Tetenkin2014; Henry et al., Reference Henry, Bezrukova, Teten’kin and Kuz’min2018).

The expansion of heat-loving spruce, fir (Abies), and pine (Pinus) was recorded in the Early Holocene deposits of cultural horizon 1A at the Kovrizhka III site (9.2 ka) (Henry et al., Reference Henry, Bezrukova, Teten’kin and Kuz’min2018). Nevertheless, open spaces on the Baikal-Patom Plateau were still preserved, as evidenced by the remains of snow sheep in cultural horizons 4A and 2 at Kovrizhka II (Tetenkin, Reference Tetenkin2010). Based on the results of landscape and climatic reconstruction in the area of the Bols’shoe Inyaptukskoye Lake in the southern part of the Baikal Highlands, changes in vegetation in the inner parts of the Baikal-Patom Highlands were 1.5–2 millennia later when compared with the development of landscapes near the mouth of the river (Bezrukova et al., Reference Bezrukova, Anderson, Vinkovskaya, Kharinsky and Kulagina2012).

New geochronological data and ideas about the formation of the Vitim River Valley for the last 60 ka BP

Margold et al. (Reference Margold, Jansen, Gurinov, Codilean, Fink, Preusser, Reznichenko and Mifsud2016) reconstructed the last glaciation within the Baikal-Patom Highlands and the Kodar and Udokan ridges. The ages of terminal moraines and other glacial forms were obtained based on beryllium (¹⁰Be) and optically stimulated luminescence (OSL) dating, and a glaciation map was compiled for MIS 2. The exhibited ¹⁰Be age of end moraine of Lake Oron and the Basny River (in the Vitim catchment) indicates the existence of glaciers along the Vitim valley in the area of the Delyun-Oronsky knickpoint at ∼20 ka, during the LGM, and in the tributaries of Vitim River at ∼16 ka (Margold et al., Reference Margold, Jansen, Gurinov, Codilean, Fink, Preusser, Reznichenko and Mifsud2016).

The center of the glaciation was the Kodar and Udokan ridges. The movement of glaciers in the northeastern direction into the Mamakan River valley and part of the Vitim River valley on the section Nerpo–Shakhtol’nya–Amalyk was reconstructed. At the same time, despite the presence of traces of glaciation (e.g., sheepback at the mouth of the Bodaybo River and glacial hatching on boulders), the existence of a glacier in MIS 2 near Bodaybo city and the Mamakan Geoarchaeological Area is debatable (Margold et al., Reference Margold, Jansen, Gurinov, Codilean, Fink, Preusser, Reznichenko and Mifsud2016).

The Late Pleistocene glaciation on the Baikal-Patom Highlands had a significant impact on the paleolandscape situation, including periodically occurring dammed glacial lakes and catastrophic water discharges. According to Margold et al. (Reference Margold, Jansen, Codilean, Preusser, Gurinov, Fujioka and Fink2018), breakthroughs of the glacial-dammed Vitim Lake dated 56 ka, 37–29 ka, and 21.8 ± 1.5 to 14.9 ± 2.0 ka were catastrophic for the Vitim valley. These events eroded the terraces that existed at that time and formed thick layers of megaflood sediments. At present, megaflood deposits are recorded at a height of 100 m or more above the current water level of the Vitim River. Such transformations in the valley influenced the formation of a new complex of terraces. The decline of alluvium thickness downstream of the river influenced the maximum height of the terraces. It should also be noted that this type of sediment was formed in tributaries, for example, in the Bodaibo River valley (Margold et al., Reference Margold, Jansen, Codilean, Preusser, Gurinov, Fujioka and Fink2018). Filling of tributaries with sediments of a catastrophic flow could occur at a distance of tens of kilometers up the valley, gradually decreasing in thickness down the valley. These deposits blocked the mouth of the tributaries and created the conditions for the formation of dammed lakes in which silts and fine sands were deposited.

Kovrizhka IV

The Kovrizhka IV site is located on a 9–11 meter terrace (TI) with a rocky basement, which is a cuesta, expressed in the form of a crest. The site is located on the right bank of an unnamed creek flowing into the Vitim River at Cape Kovrizhka (Fig. 2). The total thickness of sediments at the site is more than 3.10 m (Figs. 8 and 9). In the northeastern part of the excavation, pebbles were uncovered at the base of the section at a depth of 0.90 m and in the southern part at a depth of 1.20 m. The order of the occurrence of sediments is shown in Figure 8 and described here.

Figure 8.

Generalized stratigraphic scheme of deposits and radiocarbon dates of the Kovrizhka IV site (A). A package of culture-containing floodplain sediments (FP Unit) bears traces of lateral erosion (B).

Figure 8 Long description

A stratigraphic scheme of the Kovrizhka IV site, illustrating various sediment layers and radiocarbon dates. At the top left, labeled ′Terrace (TI) 9-11 m′, the present soil is depicted in brown. Below, paleosol is marked, followed by two erosion episodes indicated by dashed lines. FP Unit II and FP Unit I are shown in yellow, with sand and silt layers above and below them. Radiocarbon dates are displayed in blue boxes, such as 18,167 to 18,314 years before present and 18,859 years before present. The base consists of sand and pebble, with bedrock at the bottom. A red square marks the location of OSL 39.4 plus or minus 4.1 ka. The image B shows a photograph of the excavation at the site, with visible layers corresponding to the diagram. The top layer is paleosol, followed by sand and silt and FP Unit II marked with an erosion line. Below, FP Unit I is visible, consisting of sand and silt and the bottom layer is sand and pebble. A measuring stick is placed vertically in the center, providing scale and depth reference.

Figure 9.

Detailed sections of culture-bearing sediments of the 9–11 m terrace (TI, Kovrizhka IV). The sediments in trench I reflect the dynamics of sedimentation over a period of about 40 ka. The section of sediments in trench II characterizes the sedimentation conditions over a period of about 19 ka. The photo shows the nature of the sedimentary strata of the upper part of the section of trench I.

Figure 9 Long description

The diagram consists of two main sections labeled FP deposits trench I and FP deposits trench II, both situated on a terrace labeled TI 9 to 11 m. The left section, FP deposits trench I, displays sediment layers numbered from 1 to 7, with specific years before present (BP) indicated for each layer, ranging from 17,364 to 19,022 years BP. The layers are visually represented with different patterns and colors, such as orange-brown silts, light-grey-green silts and alluvial sands. A red symbol labeled A1 indicates the presence of cultural horizons. The bottom of trench I shows a layer of alluvial pebbles with sandy matrix. The right section, FP deposits trench II, similarly shows sediment layers numbered from 1 to 3, with years BP ranging from 13,148 to 18,566. The layers are depicted with similar patterns and colors and cultural horizons are marked with red symbols. An explanation box at the bottom provides a legend for the sediment types, including modern soil, podsol and brown siltstones with charcoals. Below the diagram, a photograph shows the actual sediment layers in trench I, with markers indicating the cultural horizons and sediment layers. The photograph is oriented from southwest (SW) to northeast (NE), visually matching the diagram′s representation of sediment layers and cultural markers.

Layer 1. The soil–vegetation horizon, podzolized in the sole. Thickness: 0.10–0.15 m.

Layer 2. The sandy loam is orange-yellow and indistinctly layered, of subaerial colluvial–solifluction genesis. It contains one cultural horizon. The date of 10.1 ka (Poz–107434) was obtained from a ceramic fragment. Charcoal from a hearth yielded a date of 8.8 ka BP (SacA-75974). Thickness: 0.15–0.25 m.

Layer 3. The deformed horizon of the buried soil is a dense, bright, dark-brown sandy loam with a thickness of 0.02–0.05 m, containing charcoal in large quantities. Dates were obtained for charcoals: 13.1 ka (Poz-106967) and 13.2 ka (Poz-106963).

Layer 4. The sandy loam is light green, light brown, and indistinctly layered, of subaerial colluvial–solifluction genesis. It contains c.h. 2. Cryogenic cracks are generated from the sole, forming a polygonal network. Thickness: 0.05–0.20 m.

Layer 5. The sand horizon is a yellowish, fine-grained layer with rounded pieces of black slate. It is subject to cryogenic deformation, which is caused by solifluction processes. The horizon manifests itself as an independent layer or in the form of injection spots. Thickness: 0.01–0.30 m.

Layer 6. A package of sediments of floodplain alluvium—rhythmically and horizontally layered sands, sandy loams, and silts (Floodplain Unit II) that include cultural horizons 2/1, 2A, 2B, 2B, 2G, 2D, 3/1, and 3/2. Dates were obtained for c.h. 2B: 18.3 (Poz-106023), 18.5 (LTL-16563A), and 18.8 ka (Poz-106962); for c.h. 2G: 18.5 (Poz-111232), 18.6 (Poz-111356), 34.9 (Poz-106961), and 35.6 ka (Poz-106960); for c.h. 2D: 18.6 ka (Poz-106968); and for c.h. 3/2: 18.5 ka (Poz-106965) and 18.2 ka (SacA-75975). This package is embedded in the sediments of the previous cycle, partially destroyed by river erosion, observed as a layer of dark gray sandy loam containing c.h. 3/1 and c.h. 3/2. In the northern part of the excavation, these sediments are superimposed on the undisturbed part of layer 7; closer to the northern corner of the excavation and the outer edge of the ravine, they merge compressionally into the undifferentiated third cultural horizon. Thickness: 0.30–0.90 m.

Layer 7. A package of alternating light gray sand and dark gray sandy loams and silts (Floodplain Unit I) unconformably overlies layer 8. The thickness of the layers is 2–4 cm. In the lower part, there is an abundant inclusion of small, rounded shales in some horizons. In the upper part, in layers of dark gray sandy loam and silts, cultural horizons lie from top to bottom: 3A, 3B, 4, 5, 6, and 7. Dates were obtained for c.h. 6: 18.1 (UGAMS–27448), 18.8 (Ua-50437), 19.0 ka (LTL-16562A), and 19.0 (Beta–453119). Two age of 18.8 ka (Poz-131810 and Poz-131812) were obtained for c.h. 3B. Closer to the river, the package is eroded by flooding. The erosion line is oriented from west to east, to the tip of Cape Kovrizhka. The surface of the washout is clear, uneven, and stepped. Thickness: 0.70–1.50 m.

Layer 8. Ginger and grey sand-gravel deposits with inclusions of various boulders up to 30 cm in diameter. The deposits are poorly sorted. Small varieties are poorly rolled. Pebbles and boulders are well rolled. In the infill, the sand is well sorted, coarse-grained, and gray in color. An OSL date of 39.4 ± 4.1 ka was obtained from the upper part (Riso-208499). The apparent thickness is 0.6 m or more.

In alluvial deposits (layers 6 and 7), sedimentation blocks with vertical and horizontal displacement with an amplitude of up to 10 cm are noted.

Initially, the interpretation of the age of culture-containing deposits on Kovrizhka IV was based on general ideas about the Holocene age of the 9–11 m Vitim terrace. This was evidenced by all previous archaeological work on the sites located on this terrace: Invalidny III, excavation area 1–3; Mamakan I–V; and Kovrizhka I (Belousov et al., Reference Belousov, Ineshin, Sulerzhitsky and Tetenkin2002). The date 8.8 ka BP (SOAN-7294), obtained from charcoal found in a molehill from the layer containing c.h. 5, did not contradict this age estimate. A date of 9.3 ka (SOAN-8839), obtained by the oscillation method from a bulk charcoal sample from c.h. 6, confirmed this interpretation. However, unexpectedly, an AMS radiocarbon date of 18.8 ka BP (Ua-50437) on bone from c.h. 6 significantly increased the age. Further confirmation was obtained through the AMS dating of two samples of snow sheep tooth enamel from the cultural horizons 6 and 3: 18.1 (UGAMS-27448) and 17.4 ka (UGAMS-27447). The dates were made using bioapatite, and although bioapatite dating is not considered as reliable (Cherkinsky, Reference Cherkinsky2009; Zazzo and Saliège, Reference Zazzo and Saliège2011), these dates still suggest a Pleistocene age for the deposits. Moreover, the date of c.h. 3 suggested that the entire stack of cultural alluvial deposits may be of Pleistocene age. Then, with the help of anthracologist A. Henry, samples of willow (Salix) charcoal, which makes up a majority of wood charcoal in c.h. 6 and c.h. 2B, were dated to 19.0 (LTL-16562A) and 18.6 ka (LTL-16563A), respectively. Another AMS date of 19.0 ka (Beta-453119), which turned out to be almost identical with the previous one, was obtained on willow charcoal from the second hearth in the dwelling of c.h. 6. Thus, three dates (19.0, 19.0, and 18.8 ka) for c.h. 6 and a date of 18.5 ka for c.h. 2B have put an end to the question of the Pleistocene age of alluvial culture-bearing deposits. These horizons proved to be the oldest in the Vitim catchment with clearly stratified archaeological deposits.

The next contribution to radiocarbon chronology was a new series of 12 AMS dates on culture-bearing deposits at the site. Four dates for 2B, 2G, and 2D cultural horizons at ∼18.5 ka show a very small chronological gap between the horizons and agree satisfactorily with one another. The dates of 35.2 and 34.9 ka for c.h. 2G are older than expected and can be explained by the entry into the layer of MIS 3 wood (brought by the river) and used at a later time by people for household needs. The date of 18.5 ka BP for c.h. 3/2 should be older, because c.h. 3/2 lies stratigraphically below the 2A, B, V, G, and D cultural horizons. A bone date of 18.1 ka of c.h. 2B is somewhat younger than the dates for charcoal from the same horizon (18.5 ka BP). Similarly, the date 18.8 ka on a bone from c.h. 6 is younger than two charcoal dates ∼19.0 ka from the same context. Shrub willow was defined as the dominant taxon in charcoal accumulations, so the charcoal dates are accepted as a reference in the radiocarbon chronology of Kovrizhka IV site.

Based on the growth and wear patterns of snow sheep teeth, winter habitat seasons were determined for sites in c.h. 6 and c.h. 2B and spring habitat season for a site in c.h. 3B (Teten’kin et al., Reference Teten’kin, Smith and Henry2016; Tetenkin et al., Reference Tetenkin, Henry and Klementyev2017). Furthermore, the site’s location on a beach coastal strip was taken into account, as it was safe only during flood-free periods, that is, from autumn to spring.

Assemblages from c.h. 7 to c.h. 2A date to the early stage of the LUP (Tetenkin, Reference Tetenkin2025). The technical contexts have been determined: (1) production of microblades for insert hunting weapons, obtained by pressing from wedge-shaped microcores (Fig. 3.3, and 4.2, 4); (2) production of flakes by flaking from pebble cores—most of the tools are spalls with used, non-retouched edges; (3) production of macroblades up to 12 cm long (Fig. 3.2); and (4) production of coarse prismatic blades up to 5 cm long. Morphologically expressed tools are side-scrapers, end-scrapers, piѐce esquillée, and a median burin (Figs. 4.1, 3, 6–8 and 3.1, 2, 5, 6). An anthropomorphic figure made from a mammoth tusk blade was found (Fig. 3.9 and 4.5).

The age of c.h. 2 is calculated based on the fact that it is covered by a buried soil with dates of 13.1 and 13.2 ka. The lower age limit of c.h. 2 is controlled by the flood layer of light yellow sand (layer 5) underlying the subaerial package. This flood destroyed the upper part of the underlying package of floodplain alluvium. Radiocarbon and stratigraphic data establish the age of c.h. 2 to be in the range of about 13.2–13.1 ka. Cultural horizon 2 stone artifacts are LUP in appearance, displaying similarities with previous cultural horizons, for example, wedge-shaped microcores, end-scrapers, and side-scrapers. The artifacts were made primarily from locally available pebbles of volcanic rock and mudstone, as well as quartz.

A date of 10.1 ka (Poz-107434) was obtained on carbon directly on ceramic from c.h. 1. Unfortunately, the sample submitted for dating did not contain any carbon or organic matter. Therefore, the received date is assessed as invalid, namely, overdated. Therefore, the obtained date is considered to be too old. A more acceptable age is 8.8 ka (SacA-75974), obtained on charcoal from hearth 1 of the cultural horizon. The stone artifacts consist of prismatic microblade cores, flint microblades, chipped pebble adzes, and a hammer-pestle (Figs. 5 and 7.21–23). The ceramics bear the technical imprint of woven nets (Fig. 7.24).

The flood horizon (layer 5, erosion episode 3) destroyed deposits that could have accumulated after 18 ka. This episode, older than 13.2 ka, is logically associated with a sedimentation break on Bol’shoi Yakor’ I, where there is a break between the alluvial Pleistocene sediments and the Middle–Late Holocene subaerial package. The last cultural horizon in the floodplain alluvium on Bol’shoi Yakor’ I site—c.h. 3A—has a date of ∼13.6 ka (Ineshin and Tetenkin, Reference Tetenkin2017). Based on this correlation, the age of layer 5 at Kovrizhka is probably 13.5–13.2 ka.

Erosion episode 2 falls stratigraphically between the 3/2 and 3A cultural horizons. Its age is about 18.6 ka. The unconformity, fixed in this section, is an erosive shore edge. There are 14 cultural horizons associated with this feature; of which 6 are laid down before unconformity and 8 afterward, which indicates a preference for people to settle on the beach, near the water’s edge. On this landscape, safe seasonal long-term base camps could exist only in seasons with zero (minimal) flood activity, that is, from autumn to early spring.

The age difference of about 20,000 years between the ages of layer 8 and layer 7 indicates an incision or erosion occurred within this time, producing an unconformity identified as erosion episode 1. Layer 8 dates to MIS-3, and the variable roundness and poorly sorted boulder–pebble materials indicate the rapid nature of the water flow that deposited these sediments.

In summary, 16 cultural horizons have been identified at Kovrizhka IV: 14 horizons in alluvial sediments dating to the LUP, 19.2–18.5 ka; c.h. 2 dating to the Final Paleolithic, 13.4 ka BP; and c.h. 1 dating to the EN 8.5–8.0 ka BP.

Kovrizhka III

The Kovrizhka III site is located on the right bank of the Vitim River nearby Kovrizhka IV (Fig. 2). The main excavation is laid on a cape-shaped ledge of the 22 m Vitim terrace. It also includes an adjacent slope with a relative height of 24, 29, and 35 m from the edge of the Vitim River and up to 260 m from the shore. The slope topography has the character of terraced steps in the form of transverse ridges separated by small ravines.

The excavation is situated on the surface of the terrace at an altitude of 22.5 m from the shoreline of the Vitim and 176 m from the edge of the shore, oriented along the north-south line. The sediment sequence on the eastern wall is shown in Figure 10 and described here.

Figure 10.

(A) Diagram shows the main terrace deposits (TIII) formed by aquatic and subaerial sedimentation. The deposits are rich in artifacts reflecting the stages of occupation of the emerging surface. Photographs (B–D) show details of the deposit structure with cultural horizons and artifact inclusions.

Figure 10 Long description

Cross-sectional diagram of terrace deposits labeled Terrace TIII at 22 meters. The diagram is oriented from north to south, with the Vitim River indicated to the north. Various sediment layers are depicted, including yellowish-brown silts, brownish silts with lenses and grayish sands. The layers are marked with different patterns and colors and labeled with their respective ages in years before present, ranging from 12,194 to 13,263. The diagram includes labels for the upper and lower levels of cultural horizons, ch. 1 and ch. 2 and the positions of gneiss slabs. A scale bar is present at the bottom. Photograph of an excavation site with a measuring rod on the left, labeled 1A and 1. Another excavation view with a measuring rod in the center, showing the upper and lower levels of ch. 2, labeled 1 and 1A. Close-up of the sediment layers with a measuring rod on the right, labeled 1, 1A and 2, showing the stratification of the deposits.

Layer 1. Soil horizon. Thickness: 0.10–0.12 m.

Layer 2. Podzolic horizon. The sole of the layer has a broken appearance due to frost wedges entering the underlying layer and curving up the slope. Artifacts from the 0th cultural horizon were recovered in the roots of the trees and under the stumps. Thickness: 0.02–0.05 m.

Layer 3. Red-brown sandy loam, indistinctly layered. The sole is wavy, thickness is 0.08–0.20 m.

Layer 4. Greenish-light brown silt with scattered brown silt layers, to which the 1 and 1 A cultural horizons are sequentially confined. The age of 9.0 ka (SOAN-8262) was obtained for charcoal from c.h. 1. The ages of 9.08 (SOAN-7027) and 9.2 ka (SOAN-7965) were obtained for charcoals from c.h. 1A. Cultural horizon1A—in the middle part of the layer—is underlain by a light gray layer of fine-grained sand, schlier (horizontal permafrost rupture) with a thickness of up to 1.5 cm. The bending of this layer downward in the rear part of the excavation and its contact with the underlying sand of layer 5 has been traced. Below this interlayer, up to the sole of the layer, c.h. 2 is delineated. The lenses of the hearths in the upper and lower parts are fixed in it. For charcoals, the dates 12.2 (SOAN-7964) and 12.9 ka (SOAN-7966) were obtained from the upper and lower hearths, respectively. Hence the upper and lower levels of c.h. 2 are delineated. Cryogenic cracks are embedded from the sole of the layer, which is deformed by solifluction processes. The sands of the underlying layer are drawn in the solifluction tongues. In the lower, raised part, lies c.h. 3 in a brown lens with charcoals dating to 13.2 ka BP (SOAN-8261). Thickness: 0.24–0.28 m.

Layer 5. A package of alluvial layers of medium-grained sand, interlacing with layers of whitish-pale gray aleurite, cryogenically deformed, gleyed, with spots of oxidation, with inclusions of small pebbles. Thickness: 1.15–1.25 m.

Layer 6. Coarse-grained alluvium, consisting of rounded and unrounded pebbles, blocks, and boulders up to 40 cm in diameter, with a sand–gravel matrix. These deposits are unsorted and unstratified and can be interpreted as megaflood deposits. The deposits are heavily oxidized, especially in the upper part. A sample (Kov-24-1) was taken from the coarse-grained alluvium for OSL dating: 23 ± 2 ka BP (RGI-1363) (Arzhannikova et al., Reference Arzhannikova, Arzhannikov, Tetenkin, Chebotarev, Timofeeva, Efremova, Gladkochub and Bryanskiy2025). The apparent thickness is 0.6 m or more.

The total thickness of the exposed loose deposits is 1.90 m.

Cultural horizons 3 and 2, dated to 13.2–12.2 ka, lie in the lower part of layer 4, which is of subaerial genesis (Final Paleolithic, Bølling-Allerød–Younger Dryas). In the upper part, c.h. 1A and c.h. 1 with an age of 9.2–9.0 ka (Mesolithic, Early Holocene) were discovered in the buried soils scattered along the slope. The floodplain alluvium of this third terrace (TIII), lying below, is defined as deposits of the MIS 2. Its age of 23 ± 2 ka correlates with the OSL dating of sediments in the middle part of the quarry of the 19 m alluvial pack of Mamakan VI sediments (22 ± 1 ka) (RGI-1364) and is determined as the time of the last ice-dammed paleolake outburst flood (Ulan-Makit paleolake) ∼22.3–21 ka (Arzhannikova et al., Reference Arzhannikova, Arzhannikov, Tetenkin, Chebotarev, Timofeeva, Efremova, Gladkochub and Bryanskiy2025).

Cultural horizons 2 and 3 are close to each other; their tool kits consists of wedge-shaped microcores and microblades, pebble cores with parallel and radial flaking, choppers, side-scrapers, end-scrapers, transverse burins made of effusive rocks, and multifaceted burins made of quartz (Figs. 3.10, 11, 13-19 and 5.1-9). They are also quite similar to the collection from Kovrizhka IV, c.h. 2, and overall have an LUP appearance.

Cultural horizon 1A consists of end-scrapers and microblades made of flint and volcanic glass and has a Mesolithic appearance. The materials from c.h. 1 are few and inconspicuous.

Kovrizhka II

Kovrizhka II is located on the 17–19 m terrace (TII) (Fig. 2). The site was partially destroyed by a road quarry. The terrace edge has been preserved undisturbed. Riverbed deposits (pebbles and sands) have been uncovered by the quarry. In the edge part of the terrace, there is a rocky outlier, to which alluvial deposits are conjoined (Fig. 11).

Figure 11.

The general scheme of the structure of sediments and radiocarbon dates of the terrace (TII, Kovrizhka II), including cultural horizons saturated with traces of ancient human activity (A and B). Photo (C) shows the clearing of the hearth of c.h. 5, arranged by the inhabitants at the rock wall—a windbreak shelter.

Figure 11 Long description

The image consists of three parts labeled A, B and C. In A, a cross-section of Terrace TII is depicted, showing various sediment layers with radiocarbon dates labeled as 9132 years before present, 9200 years before present and 19,328 years before present. The section includes cultural horizons marked as c.h. 5, 4 and 3, with a legend explaining different soil types and features such as yellow loess, brown loam and pebbles with charcoal. The diagram indicates the direction towards the Vitim River. In B, a vertical profile of cultural horizons is shown, with similar radiocarbon dates and labeled horizons. The profile is marked with a scale in meters. In C, a photograph shows two individuals working at an excavation site, clearing a hearth at the rock wall, which serves as a windbreak shelter. The image provides a visual representation of the archaeological context and stratigraphy of the site.

The stratigraphy of the excavation site at the location of Kovrizhka II is shown in Figure 11A–C and described here.

Layer 1. Soil and vegetation horizon. Thickness: 0.13–0.24 m.

Layer 2. Podzolic horizon. The first cultural horizon lies in contact with the underlying layer. Thickness: 0.07–0.02 m.

Layer 3. A horizon of orange-yellow, clay, fine-sandy, deformed sediments with traces of podzolization. The sole of the layer is soliflucted. The cultural remains contained in it are allocated to c.h. 2; the main materials lie in the upper part of the layer, and few finds are found throughout the lower part of the layer. Thickness: 0.19–0.28 m.

Layer 4. Fragments of a ruptured deformed horizon of dark-pale fine-grained aleurite. The upper part of the horizon was destroyed by the overlying layer 3. Thickness: 0.03–0.12 m.

Layer 5. The horizon of a thinly micro-layered, silty, light-green aleurite. In the lower part of the horizon, there is a slight ferruginization. In the dried state of the layer, horizontal layering is observed along micro-layers with a thickness of up to 0.01 m. The upper part of the layer accommodates two levels of cultural remnants: c.h. 3 and c.h. 4. A radiocarbon age of 9.1 ka (SOAN-5277) was obtained from c.h. 3. In the middle part of the layer lies a layer saturated with charcoal particles with a thickness of up to 2–3 cm and containing c.h. 4A. It has a radiocarbon age of 9.2 ka (SOAN-5276). On the upper surface of the artifacts in c.h. 3, c.h. 4, and c.h. 4A, a clay sheet is marked. Small cryogenic cracks, nested one within the other, are laid from the upper part of the layer and from the base. The sinking of cultural remains into the crack was noted. The upper crack is poorly readable in the section, so it is difficult to determine whether the shoulders of the upper crack lie above or below relative to c.h. 3, c.h. 4, and c.h. 4A. Light gray sandy layers with a thickness of up to 2 cm are noted in the sole of the layer. Thickness: 0.60–0.65 m.

Layer 6. Horizontally layered light gray sand with interlayers of silty aleurite/silt similar to layer 5. The layer is ruptured by cryogenic cracks from the overlying layer 5. Thickness: 1.05–1.08 m.

Layer 7. Package of bluish-dark gray horizontally layered fine-grained sands up to 2 cm thick with layers of light gray medium-grained sand. In the lower part of the layer, a light gray sandy layer has a thickness of up to 6 cm. The deposits are ruptured by cracks from layer 5. Cultural horizon 5 lies in the upper part of the layer. A ¹⁴C date of 13.0 ka (SOAN-4543) was obtained from it using the scintillation method in the 2000s, and a ¹⁴C AMS date of 19.9 ka (SacA-75973) was recently obtained. Thickness: 0.24–0.70 m.

Layer 8. Light gray medium- and coarse-grained rhythmically layered sand with inclusions of gravel. Thickness: 0.60–0.70 m.

Layer 9. The boulder–pebble layer is unsorted, with gruss and sand in the filling. Thickness: 0.10–0.50 m.

Layer 10. A layer of coarse-grained sand with gruss. A distinct lens of thin-layered sand (layer 11) is marked in the middle part of the layer. Thickness: 0.60–0.80 m.

Layer 11. A lens of obliquely layered silt and fine-grained sands with layers of gray coarse-grained sand. Thickness: 0.01–0.50 m.

Layer 12. A layer of obliquely and rhythmically layered silt and fine-grained sands with layers of gray coarse-grained sands. Thickness: 0.55–0.70 m.

Layer 13. A layer of coarse-grained sand with gruss. Thickness: 0.05–0.20 m.

Layer 14. Gray rhythmically layered coarse-grained sands with layers of silty fine-grained sands. This layer has the opposite orientation of layers in 11 and 12. Thickness is up to 0.80 m or more.

The total thickness of the exposed deposits reaches 4.5 m.

Sediments in layers 14–8 belong to the channel alluvium, layers 7–6 represent the floodplain alluvium, sediments 5–3 were formed in subaerial conditions, experiencing the processes of pyrogenesis, cryoturbation, and sedimentation, and layer 1 is the modern soil layer. A stratigraphic unconformity is observed between layers 5 and 6, indicating a hiatus in sedimentation. This hiatus could have been caused by either flood erosion of some of the floodplain facies sediments or by subaerial denudation processes, which halted sedimentation in this area. Layer 9 (an unsorted boulder–pebble layer with debris and sand in the fill) is correlated with the pebble layer 6 of Kovrizhka III and the last megaflood of ∼22.3–21 ka.

Cultural horizon 5 is a sparse accumulation of cultural remains near a hearth located behind a rock outcrop (Fig. 11C). These remains include only one scraper and a few flakes (Tetenkin, Reference Tetenkin2010).

Cultural horizons 4A–1 are generally similar to one another and to assemblages from the preceding period, the end of MIS-2. This is a set of pebble cores with a parallel flaking principle, wedge-shaped microcores and microblades, side-scrapers, end-scrapers, perforators, and retouched flakes of an LUP appearance (Figs. 3.18 and 5.11-17).

Kovrizhka V

Kovrizhka V is situated on a 9–11 m terrace (TI) 30 m from the river’s edge, at the base of the ledge of a 17–19 m terrace (TII) (Fig. 2). Excavation revealed the sequence of deposits illustrated in Figure 12 and described here.

Figure 12.

The section of the terrace sediments (TI, Kovrizhka V site) reflecting the dynamics of sedimentation (A). Photo (B) shows buried soil horizons and culture-containing sediments.

Figure 12 Long description

Sediment layers at Terrace 9-11 m (TI), illustrating various geological strata. The layers are labeled with different textures and colors, representing modern soil, podzol, orange-brown silts, light-grey-green silts with lenses of brown silts, medium-grained sands and black silt lenses with charcoals. Each layer is marked with its corresponding age in years before present, ranging from 814 to 14138 plus or minus 138 years before present. The diagram includes arrows indicating the direction to the Vitim River and an explanation key at the bottom detailing the symbols used for artifacts, cultural horizons, sediment layers, cultural horizons and lithological layers. A photograph of an excavation site at the same terrace, with a vertical measuring stick marked with numbers 1 to 4, indicating different layers or horizons. The excavation reveals distinct layers of soil, with visible color changes and textures. The direction to the Vitim River is indicated at the top and the photograph is oriented from west to east. The measuring stick provides a scale for the depth of the excavation and the numbers correspond to the explanation key provided in image.

1. The subaerial slope pack of orange-yellow and light-green sandy loams has the same composition as that in the Kovrizhka IV section, but is less susceptible to solifluction. Contact with the underlying deposits of layer 1 is at a depth of 0.85 m. At the base of member 1, in a pyrogenic interlayer, lies c.h. 1 dating to 8.3 ka (SOAN-8838, date obtained by the oscillatory method) (Early Neolithic) (Table 1). Judging by the AMS age from the underlying level, the age of c.h. 1 should be closer to 10.5 ka.

2. The unit is up to 3 m thick or more and consists of yellow-gray, brownish, greenish-gray fine-grained sands with wavy bedding, interbedded with pyrogenic, intensely coal-saturated, eroded horizons. The nature of the deposits (periodic flooding of the surface exposed to fires) corresponds to the floodplain facies of alluvium. The age of 10.6 ka (SacA-75972) was obtained from the buried soil in the top of the unit underlying the first cultural horizon. Cultural horizon 2 was found at a depth of about 1.2 m and yielded dates of 11.5 ka (SOAN-8837) and 12.1 ka BP (SacA-75971) (Final Paleolithic). Cultural horizon 3 was found at a depth of about 2.0 m and yielded dates of 12.9 ka BP (GV-05544) and 12.8 ka (SacA-75970) (Final Paleolithic). Cultural horizon 4 was identified in the underlying layer. Dates of 13.2 ka BP (SacA-75969) and 14.1 ka (SacA-75968) were obtained from depths of 2.35 and 3.5 m, respectively.

The transition from the floodplain to the terrace in this section of the valley is noted at a depth of 0.85 m from the daylight surface and is dated between 10.6 ka BP (the date was obtained from the charcoal-rich horizon at the top of layer 2) and 8.3 ka (the age was obtained from the pyrogenic layer at the base of layer 1).

Cultural horizon 1 contains a few finds of prismatic flint blades and flakes. Charcoal from this layer yielded a date of 8.3 ka (SOAN-8838). Cultural horizon 2 includes a retouched blade, retouched flakes, and pebble debris from effusive rocks. This assemblage has an LUP appearance (Figs. 3.20 and 7.5). Finds from c.h. 3 are few in number. In terms of substrate and morphology, they are similar to those from c.h. 2.

Kovrizhka I

Kovrizhka I is located in the ravine part of the first terrace (TI) at a height of 10.9 m from the Vitim River, 60 m from the shore (Fig. 2). A description of the test pit is provided here.

Layer 1. Soil and vegetation horizon. Thickness: 0.14–0.26 m.

Layer 2. Charcoal, bright black medium sandy loam. Thickness: 0.03–0.16 m.

Layer 3. Pale orange-yellow, gray-yellow light sandy loam, ferrugination layered, cryogenically deformed, torn, with charcoal layers up to 0.04 m thick containing detritus lenses. The layering of the interlayers is indistinct due to blurring. Cultural horizon 1 is located in the middle part of the layer, within a charcoal layer. An end-scraper, representing c.h. 1A, was found 50 cm below c.h. 1. Thickness: 0.48–1.10 m.

Layer 4. Gray cryogenically deformed sandy loam, indistinctly layered with spots of calcification, charcoal layers up to 0.04 m thick, with inclusions of small flat pebbles of carbonaceous shale lying flat, up to 0.008 m in size. In the raised part, three flakes were found representing c.h. 1B. Thickness: 0.35–0.60 m.

Layer 5. Culture-containing (c.h. 2) cryogenically deformed grayish-pink sand. In the upper contact with layer 5, there is a charcoal layer, which also contains detritus. The thickness of the interlayer is 0.01 m. Thickness: 0.10–0.50 m.

Layer 6. Light gray sand, cryogenically deformed resulting in discontinuous stratigraphy. Thickness: up to 0.60 m or more.

Layer 7. Light and dark gray indistinctly layered cryogenically deformed sand with inclusions of charcoal, with lenses of dark gray charcoal-saturated sand. One of these lenses contained a core representing c.h. 3. Thickness: up to 0.35 m or more.

The total depth of the loose sediments uncovered by the test pit is 2.10 m. The permafrost is fixed at a depth of 1.25 m.

The age of the culture-containing carbonaceous layer of c.h. 1 is 2.9 ka (GIN-9003). Taking into account the sloping aeolian–diluvial nature of the deposits of lithological layers 3 to 4, the discontinuity of culture-containing layers, and the vertical and inclined position of the finds and their uniqueness, it should be concluded that the upper c.h. 1, c.h. 1A, and c.h. 1B were deposited by the removal of artifacts by sloping processes from the above levels. Lithological layer 5, containing c.h. 2, has a thickness of 0.08–0.20 m. Its deformation appears to be complex due to both slope drift and cryogenic disturbances. Charcoal from this horizon yielded dates of 6.9 (SOAN-4245) and 6.7 ka (SOAN-4545) (Table 1). The single finds of c.h. 3 lie at the top of the floodplain of the 9–11 m Kovrizhka terrace.

Cultural horizons 1, 1A, and 1B are represented by isolated artifacts, while c.h. 2 contains end-scrapers made of brown exotic mudstone blades (Figs. 3.25, 26 and 7.3, 4). X-ray fluorescence (XRF) studies revealed a near-identical compositional match between the stone and morphologically similar end-scrapers made of blades from the Ust-Karenga XVI site, located 800 km upstream of the Vitim River (Tetenkin et al., Reference Tetenkin, Vetrov, Demonterova, Pashkova and Kaneva2018a). The remaining artifacts are made of flint and effusive rocks. In general, here and in other archaeological assemblages of Kovrizhka I–VI, flint appears only in Mesolithic–Neolithic cultural assemblages in the Holocene. The materials from c.h. 3 are few in number and of little diagnostic value. They are made of effusive rocks.

Kovrizhka VI

The Kovrizhka VI site is located on a small cape of bedrock of the Vitim River at a height of 5 m above the water level. A test pit excavated to a depth of 1.90 m revealed three layers described here.

Layer 1. Soil and vegetation horizon. Thickness: 0.10–0.16 m.

Layer 2. A package of light gray indistinctly layered sands with inclusions of plant organic matter with a thickness of up to 0.40 m.

Layer 3. A package of dark and light gray rhythmically layered sands with inclusions of lenses of vegetable detritus and carboniferous wood with a thickness of up to 1.1 m, lying on boulder–pebble deposits. A prismatic core and sinker representing the first cultural horizon of the Neolithic were revealed in layer 1.