INTRODUCTION
Upper Paleolithic localities in Western Siberia are not numerous (Zenin Reference Zenin2002), and the Volchia Griva (Wolf’s Ridge) is today one of the best-studied sites in this vast region. It is also one of the largest accumulations of megafaunal remains in Asia, with in situ bones (Leshchinskiy and Burkanova Reference Leshchinskiy and Burkanova2022). While the Volchia Griva (hereafter—VG) site was discovered in the late 1950s, and excavated mainly in the 1960s–1970s (Derev’anko and Markin Reference Derev’anko, Markin, Derev’anko, Shimkin and Powers1998:82–83) and in 1991 (Zenin Reference Zenin2002), its chronological aspect was unclear until the mid-2000s when more data were obtained (Leshchinskiy et al. Reference Leshchinskiy, Kuzmin, Zenin and Jull2008). Recently, numerous 14C dates were generated for VG, and this allowed us to better understand the main temporal patterns of megafaunal presence in the southern part of the West Siberian Plain, and for the first time to firmly establish the timing and nature of the site’s occupation by humans in the Upper Paleolithic. In this paper, we present the progress with 14C dating of VG, and discuss more general issue of human presence in Siberia at the Last Glacial Maximum (LGM) that is still a subject of active debates (e.g., Kuzmin and Keates Reference Kuzmin and Keates2018 versus Buvit et al. Reference Buvit, Rasic and Izuho2022).
MATERIALS AND METHODS
The VG site is located within the modern Mamontovoe Village (Kargat County, Novosibirsk Province, Russia), with coordinates 54°39′55″N, 80°20′02″E (Figure 1). The site occupies the top and slopes of the griva (ridge; also the term “mound” was used, see Leshchinskiy and Burkanova Reference Leshchinskiy and Burkanova2022), the elongated landform ca. 11 km long and ca. 0.5–1 km wide, with an elevation of ca. 10–15 m above the surrounding flat surface of the Baraba Lowland (Figures 1–2). Two areas (a.k.a. sectors) in the spatial structure of VG can be determined: (1) northeastern, and (2) central (Figure 1, b). Up to 2015, the excavations were carried out in the former area, and since 2016—in the latter one. Overall, more than ca. 550 m2 of bone-bearing deposits were unearthed. Several thousand mammoth bones and teeth, and ca. 80 lithic tools and flakes were obtained before 2020 (Leshchinskiy et al. Reference Leshchinskiy, Zenin and Bukharova2021); unfortunately, most of the faunal material collected before 2015 was lost (Leshchinskiy and Burkanova Reference Leshchinskiy and Burkanova2022). Excavations using the three-dimensional plotting of bones and lithic artifacts were carried out in the northeastern sector in 1991 (main pit) and in 2015 (4 × 3 m pit next to it). In addition, two small test pits were examined in this part of the site in 2000–2001; the 2000 test pit was dug next to the 1991 main area of work (Orlova et al. Reference Orlova, Kuzmin, Zenin and Dement’ev2003) (Figure 1, b). In 2016–2022, investigations were shifted toward the center of the village (Figure 1, b) where by preliminary sounding the high concentrations of bones were detected. Until recently, all studies were conducted on the flat top surface of the ridge. In 2020–2021, a pit was dug on the gentle slope (2–3°), ca. 38 m distant from the 2016–2017 excavation.
Position of the Volchia Griva site: (a) in the West Siberian Plain; (b) on the Volchia Griva ridge (NE—northeastern sector; C—central sector), with excavation localities and years of work; background is based on satellite image (free access).
General stratigraphy of the Volchia Griva site (after Leshchinskiy et al. Reference Leshchinskiy, Zenin and Bukharova2021, modified): (a) 2016–2019 excavations, central sector; (b) 1957–2015 excavations, northeastern sector (1—bones from the upper level; 2—from the middle level; and 3—from the lower level); (c) schematic latitudinal cross-section. 1—sand; 2—silt; 3—clay; 4—loess-like loam; 5—soil; 6—layer boundaries: a—certain, b—assumed; 7—desiccation cracks; 8—faunal remains; 9—lithic artifacts; 10—boundaries of bone-bearing levels; 11—abbreviations: lbbl—lower bone-bearing level, mbbl—middle bone-bearing level, and ubbl—upper bone-bearing level.
Besides numerous bones of large mammals (mainly woolly mammoth, Mammuthus primigenius Blum.), until 2020 a few dozens of lithic artifacts were found at VG (Zenin Reference Zenin2002; Leshchinskiy et al. Reference Leshchinskiy, Zenin and Bukharova2021). This raised the issue of the relationship between prehistoric humans and the megafauna; it was not clear—did people kill mammoths or exploit the natural accumulation of bones (“mammoth cemetery”), or did the animals die naturally? Some observations in 1991 and 2015–2018, when excavations were carried out with strict stratigraphic and planigraphic controls, allowed us to suggest that the human impact on the VG megafauna was very small, almost negligible (see details: Leshchinskiy and Burkanova Reference Leshchinskiy and Burkanova2022). Prehistoric people could have hunted animals like bison, horse, wolf, and fox, but the amount of these species at VG is small, up to a few percent. The breakthrough in understanding the nature of VG took place in 2020–2021 when clear traces of human occupation, including ca. 10,500 lithic artifacts, slightly re-deposited hearths with small pieces of charcoal and abundant ash, and numerous fox bones (mainly of polar fox), were excavated.
The general stratigraphy of the VG site can be presented as follows (Figures 2–3). Below the modern soil (up to 0.5 m thick), there are subaerial deposits—loess-like loams (ca. 0.3–0.75 m thick) and underlying clayey sands and silts (up to 1–2 m thick; these are present mainly in the central sector), and they contain three levels (upper, middle, and lower) of bones and artifacts. Bone-bearing sediments are represented by lenses ca. 0.15–1 m thick, which are distributed throughout the VG site but are not continuous. It is sometimes difficult (especially in the northeastern sector; see Figure 2, b) to determine to which level the bones belong because of post-sedimentation disturbance caused by trampling of subfossil bones in antiquity by animals who were attracted to VG as a mineral lick. Beneath loess-like deposits, clayey sands, and silts, there are finely laminated sands, most probably of aquatic (lacustrine) origin, without any bones and artifacts.
The stratigraphy of the Volchia Griva site, 2020–2021 excavations (central sector) (numbers in circles are sediment layers: 1—Layer 1, modern soil; 2—Layer 2, loess-like loam; 3 and 4—Layers 3–4, thinly laminated clayey sand and silts; 5—Layer 5, thinly laminated sand). 1—sand; 2—silt; 3—loess-like loam; 4—soil; 5—boundaries: a—certain, b—assumed; 6—desiccation cracks; 7—krotovinas; 8—concentrations of ash and small charcoal pieces; 9—animal bones; 10—artifacts. Blue circles with black centers and their numbers correspond to samples in Table 1. (Please see online version for color figures.)
The 14C dating of materials from VG was conducted mainly using the AMS technique (AA, UGAMS, IGANAMS, and RICH laboratories), and also by LSC counting (GIN and SOAN labs). Details of the pretreatment for bones can be found in the relevant publications (Ovodov et al. Reference Ovodov, Crockford, Kuzmin, Higham, Hodgins and van der Plicht2011; Kuzmin et al. Reference Kuzmin, Kosintsev, Boudin and Zazovskaya2020, Reference Kuzmin, Boudin, Wojcieszak, Zazzo, van der Sluis, Stulova, Gavrilov, Veselovskaya and Vasilyev2022; Leshchinskiy and Burkanova Reference Leshchinskiy and Burkanova2022). In general, bone collagen was extracted by usual protocols (see Kuzmin and Orlova Reference Kuzmin and Orlova2004:144–145; Ovodov et al. Reference Ovodov, Crockford, Kuzmin, Higham, Hodgins and van der Plicht2011; Leshchinskiy and Burkanova Reference Leshchinskiy and Burkanova2022: 160; Leshchinskiy et al. Reference Leshchinskiy, Zenin, Burkanova and Kuzmin2023:7); charcoal was treated by standard acid-base-acid protocol (see Taylor and Bar-Yosef Reference Taylor and Bar-Yosef2014:93).
RESULTS AND DISCUSSION
The 14C values, available up to early 2023 for the VG site, are summarized in Tables 1–2. The first 14C dates were obtained in the late 1960s at the SOAN laboratory by the LSC method (Table 2). Unfortunately, the documentation about the sample locations is poor, and this is why we decided to label them as “unclear”. On the other hand, these three 14C values do not contradict the general chronological range of the site.
Reliable 14C dates of the Volchia Griva site (the youngest mammoth values are in bold).
a NE—northeastern; C—central.
b U—upper; M—middle; L—lower.
c Expressed with ± 2 σ, combined and rounded to the next 10 yrs, using Calib 8.1.0 software (http://calib.org/calib/).
d The same specimen was 14C-dated.
e Mammoth tusks.
f The same specimen was 14C-dated (the IGANAMS-8486 value is not plotted in Figure 4).
g These samples constitute an artificial bone assemblage (fox bones were placed in a cavity made in a mammoth’s femur; see Leshchinskiy et al. Reference Leshchinskiy, Zenin, Burkanova and Kuzmin2023).
Unclear 14C dates and outliers of the Volchia Griva site.
a NE—northeast; C—central.
b U—upper; M—middle; L—lower.
c Expressed with ± 2 σ, combined and rounded to the next 10 yrs, using Calib 8.1.0 software (http://calib.org/calib/).
The next stage in 14C dating the VG site came in the 2000s when the northeastern sector was additionally excavated (see Leshchinskiy et al. Reference Leshchinskiy, Kuzmin, Zenin and Jull2008, Reference Leshchinskiy, Zenin and Bukharova2021). This allowed us to better understand the overall age range, including the middle and upper bone-bearing stratigraphic levels. One 14C value, ca. 34,000 BP (AA-60770), was much older than the rest of dates, and we determined it as an “outlier” (Table 2). The reason for such a large deviation from the other ages is still unclear. The upper level of the northeastern sector was 14C-dated to ca. 11,090–11,815 BP; the combined upper and middle level—to ca. 12,520–17,840 BP, and the lower level—to ca. 18,840–19,130 BP (Table 1; Figure 4). On the other hand, the age of the upper/middle level is to some extent tentative due to the possibility of re-deposition (see below).
Distribution of calibrated 14C dates for the Volchia Griva site according to bone-bearing levels (see Table 1).
Since 2015, the amount of 14C dates for VG increased sharply, and the AMS method was employed; bones of woolly mammoth, Pleistocene horse and bison, wolf, and fox were selected for analysis (Leshchinskiy et al. Reference Leshchinskiy, Zenin and Bukharova2021; Leshchinskiy and Burkanova Reference Leshchinskiy and Burkanova2022). The excavations in 2020–2021 of the part of the central sector with clear traces of human habitation (i.e., presence of artifacts and hearths) and increased funding allowed us to generate even more 14C values on different material, including bones of woolly mammoth, Pleistocene horse, wolf, fox, and a bird (unidentified to species), and wood charcoal (Table 1). The overall ages for the middle level are determined as ca. 10,620–18,640 BP; and for the lower level—ca. 16,940–19,790 BP (Table 1; Figure 4). Once again, these ranges are tentative because of possible re-deposition occurred in the middle level. One 14C value on freshwater mollusk shell, ca. 29,550 BP (IGANAMS-8475), turned out to be much older than the rest of dates from the lower level. Probably, Upper Paleolithic people collected this unusual looking sample from the older deposits and brought it to the site as a kind of ‘souvenir’; we determined it as an “outlier” (Table 2). Another 14C date of ca. 15,000 BP (IGANAMS-9800), run on burnt bone, does not fit the overall sequence, and we also placed it into the “outliers” (Table 2).
There are several cases when particular 14C values do not correspond well to the rest of dates from the bone-bearing levels. For example, in the northeastern sector two 14C dates associated with the middle level—ca. 12,520–14,280 BP (SOAN-4292 and SOAN-4293)—are younger than the dates from the upper/middle (?) level, ca. 16,090–17,800 BP. Because in this sector of the VG the levels are located closely to each other (Figure 2, b), trampling of subfossil bones by mammoths and other large animals (and possibly even prehistoric humans) may have caused the re-deposition of these bones from the upper level to the middle one. Therefore, overlaps in the 14C dates for the upper and middle levels (Figure 4) can be related to trampling by animals and re-deposition by erosion.
In the central sector, a similar situation can be detected for the presumably upper level; the 14C date of ca. 16,750 BP (UGAMS-32286) is much older than the rest of the 14C values for this level in the northeastern sector. The relatively young 14C dates from the middle level—ca. 10,620 BP (UGAMS-40948) and ca. 13,700 BP (UGAMS-32275)—deviate from the rest of the 14C values, ca. 15,870–18,640 BP. In the lower level, the 14C date of ca. 17,375 BP probably belongs to the middle level. The most likely reason for that could be the re-deposition by trampling and by water transport and erosion in the depression with a large accumulation of bones, and disturbance by prehistoric humans who might have visited VG after the LGM, although this is not certain (Figure 2, a; Leshchinskiy and Burkanova Reference Leshchinskiy and Burkanova2022:165–169). It was noticed that some bones were found in almost vertical position (Leshchinskiy and Burkanova Reference Leshchinskiy and Burkanova2022:163). Nevertheless, there are several cases of bones in anatomical articulations (Leshchinskiy and Burkanova Reference Leshchinskiy and Burkanova2022) in the middle and lower levels, and this testify especially in favor of a minor disturbance of the lower level.
In the excavation pit of 2020–2021, the stratigraphy of VG is the best in terms of separation of the artifact-bearing levels (Figure 3). The majority of animal bones and artifacts (lower level) is associated with Layer 4 and the bottom of Layer 3. Several 14C values from the lower level, ca. 18,230–19,790 BP, were run on wood charcoal associated with slightly re-deposited hearths (Figure 5) and correspond to the timing of human occupation of this area at the VG. Out of two 14C dates of the same sample, 17,375 ± 50 BP (IGANAMS-8486) and 19,790 ± 70 BP (RICH-29414.1.1), we accept the older value because it fits better with the rest of dates from the lower level (Table 1). However, a disturbed stratigraphy even in central sector of VG can be assumed. For example, it is quite possible that two 14C values associated with the middle level—ca. 18,460 BP (UGAMS-32277) and ca. 18,640 BP (UGAMS-40949) (Table 1)—belong to the lower level. In this case, the age of the middle level can be determined as ca. 13,700–16,940 BP, and of the lower level—as ca. 18,230–19,790 BP.
Photo of a lens (view from the top) consists of slightly re-deposited hearth, the bottom of Layer 3, Grid A-6 (2020 pit) (small artifacts and bones and covered by foil). The contour of a lens, enriched with charcoal particles, bones (fragment of the mammoth’s rib; two halves of the mandible and small fragments of Vulpes sp.), and small artifacts, is highlighted.
An additional factor of disturbance of the original stratigraphy at VG is the burrowing activity of rodents that is quite common today and in the Holocene of the Baraba Lowland. One sample of rodent bones taken from a krotovina in the lower level was 14C-dated, with a late Holocene age, ca. 1590 BP (Table 1). This is why the areas with traces of numerous burrowing holes were treated carefully, in order to avoid the mixing of krotovina fills with materials from undisturbed strata (Leshchinskiy and Burkanova Reference Leshchinskiy and Burkanova2022).
By combining all available 14C dates for the VG site into three stratigraphic levels, the general age ranges can be established after critical re-assessment of the 14C values that do not correspond to initially assumed levels. The upper level can be dated to ca. 10,620–12,520 BP; the middle level—to ca. 13,700–17,800 BP; and the lower level—to ca. 18,230–19,790 BP (Figure 4). Due to re-deposition of some animal bones by trampling and erosional and colluvial processes, the ages for the upper and middle levels remain to some extent provisional.
There are two quite “late” 14C dates on woolly mammoth from the VG site, ca. 10,620–11,090 BP (or ca. 12,500–13,200 cal BP) (Table 1). Previously, similar 14C values of ca. 10,200–11,080 BP (or ca. 11,350–13,290 cal BP) were obtained for Western Siberia (Kuzmin Reference Kuzmin2010; Orlova et al. Reference Orlova, Kuzmin and Dementiev2004a). Even though the youngest 14C date, 10,210 ± 135 BP (SOAN-4752) from the Lugovskoe locality in central Western Siberia, could be an outlier (see Orlova et al. Reference Orlova, Zenin, Stuart, Higham, Grootes, Leshchinsky, Kuzmin, Pavlov and Maschenko2004b), now the presence of mammoth in the southern part of the West Siberian Plain at ca. 10,620 BP seems to be securely established. According to Puzachenko et al. (Reference Puzachenko, Markova, Kosintsev, van Kolfschoten, van der Plicht, Kuznetsova, Tikhonov, Ponomarev, Kuitems and Bachura2017), the fast fragmentation of mammoth habitat in Siberia occurred at the end of the Bølling–Allerød interstadial, ca. 11,000 BP (ca. 12,900 cal BP), with isolated refugia in northern Eurasia outside of the High Arctic (see also Stuart Reference Stuart2021; Stuart et al. Reference Stuart, Sulerzhitsky, Orlova, Kuzmin and Lister2002). The Baraba Lowland is one of such areas where mammoths survived until the Younger Dryas, ca. 10,100–11,000 BP (or ca. 11,700–12,900 cal BP).
Based on the results obtained, we can suggest that Paleolithic people occupied the VG site in the southern part of the West Siberian Plain during the second half of the LGM (see Lambeck et al. Reference Lambeck, Rouby, Purcell, Sun and Sambridge2014), ca. 18,200–19,800 BP (or ca. 22,000–24,000 cal BP), in the cold treeless environment sometimes called the ‘mammoth steppe’ (Leshchinskiy and Burkanova Reference Leshchinskiy and Burkanova2022). The latest data for the chronology of VG can shed a new light on the hotly debated issue of human presence in Siberia during the LGM. Some scholars (e.g., Hoffecker Reference Hoffecker2005:110–111; Graf Reference Graf2005, Reference Graf, Kaifu, Izuho, Sato and Ono2015; Buvit et al. Reference Buvit, Rasic and Izuho2022) are skeptical about the human occupation of northern Eurasia at the LGM, including Siberia. On the contrary, others (e.g., Kuzmin Reference Kuzmin2008, Reference Kuzmin2021; Kuzmin and Keates Reference Kuzmin and Keates2018) are positive about the existence of archaeological sites at least in southern Siberia, up to the 58° N latitude (Figure 6).
LGM-associated sites (dated to ca. 19,000–23,000 BP) in Siberia and the Russian Far East (after Kuzmin Reference Kuzmin2021; Seuru et al. Reference Seuru, Leshchinskiy, Auguste and Fedyaev2017; modified).
The progress in studying the LGM paleogeography of the Lake Baikal region in southern Siberia (Müller et al. Reference Müller, Tarasov, Hoelzmann, Bezrukova, Kossler and Krivonogov2014; Tarasov et al. Reference Tarasov, Ilyashuk, Leipe, Müller, Plessen, Hoelzmann, Kostrova, Bezrukova and Meyer2019; Kobe et al. Reference Kobe, Leipe, Shchetnikov, Hoelzmann, Gliwa, Olschewski, Goslar, Wagner, Bezrukova and Tarasov2022; see also Krivonogov et al. Reference Krivonogov, Takahara, Kuzmin, Orlova, Jull, Nakamura, Miyoshi, Kawamuro and Bezrukova2004), where several LGM-associated archaeological sites are situated, confirms the survivial of woody plants in the river valleys at that time. This is an additional argument in favor of a suitable environment for the LGM humans in this area, along with the presence of tailored clothes and dwellings among the Upper Paleolithic populations of northern Eurasia. Currently, we have at least 30 sites south of the 58°N, and even six sites north of it (Figure 6). It is obvious that ancient inhabitants of Siberia were able to cope with the LGM climate (e.g., Pitulko et al. Reference Pitulko, Pavlova and Nikolskiy2017; Kuzmin and Keates Reference Kuzmin and Keates2018; Pitulko and Pavlova Reference Pitulko, Pavlova, Goldstein and DellaSala2020; Pavlova and Pitulko Reference Pavlova and Pitulko2020).
CONCLUSIONS
Based on 20+ years of the 14C dating program at the VG site, today we have 42 values that can be accepted as reliable ones. Currently, in the deposits of VG three stratigraphic levels of mammoth fauna and artifacts can be determined, with preliminary 14C age ranges of the ca. 10,620–12,520 BP (upper level), ca. 13,700–17,800 BP (middle level), and ca. 18,230–19,790 BP (lower level). The latter level contains the major part of animal fossils and artifacts. The human presence can now be securely established in the second part of the LGM, ca. 18,200–19,800 BP. It is not clear at the current stage of research for how long people occupied the VG site. However, the rare presence of lithic artifacts in the middle and upper levels may indicate periodical visits of prehistoric people to VG until the end of the Pleistocene. Animals visited the VG site since ca. 19,600 BP until ca. 10,600 BP. Today, VG is one of the latest refugia of woolly mammoth in northern Eurasia beyond the Arctic, and it existed until ca. 10,600 BP. Because only a small part of the site (ca. 5%) has been excavated to date, the potential of VG for Late Quaternary paleontology and Upper Paleolithic archaeology of northern Eurasia is very high.
ACKNOWLEDGMENTS
We are grateful to numerous colleagues, who participated in the excavations of the VG site in 1991–2021, for their assistance. The RICH laboratory (Royal Institute for Cultural Heritage, Brussels, Belgium) generously 14C-dated one of the VG samples, and we are thankful for their help. Work by YV Kuzmin was supported by the State Assignment of the Sobolev Institute of Geology and Mineralogy, Siberian Branch of the Russian Academy of Sciences, with funding provided by the Ministry of Science and Higher Education of the Russian Federation. We are grateful to SG Keates for final polishing of the text; however, all mistakes are only ours.
