PAVEL PAVLOV*†, JOHN INGE SVENDSEN†‡ & SVEIN INDRELID§

* Institute of Language, Literature and History, Komi Scientific Center, Russian Academy of Sciences, Ural Division, Kommunisticheskaya st. 26, 167000, Syktyvkar, Komi Republic, Russia
‡ Centre for Studies of the Environment and Resources, University of Bergen, Høyteknologisenteret (HIB), N-5020 Bergen, Norway
§ Bergen Museum, University of Bergen, Harald Hårfagresgt.1, N-5020 Bergen, Norway
† These authors contributed equally to the work

Correspondence and requests for materials should be addressed to J.I.S. (e-mail: john.svendsen@smr.uib.no).

The transition from the Middle to the Upper Palaeolithic, approximately 40,000–35,000 radiocarbon years ago, marks a turning point in the history of human evolution in Europe. Many changes in the archaeological and fossil record at this time have been associated with the appearance of anatomically modern humans^{1, 2}. Before this transition, the Neanderthals roamed the continent, but their remains have not been found in the northernmost part of Eurasia. It is generally believed that this vast region was not colonized by humans until the final stage of the last Ice Age some 13,000–14,000 years ago^{3, 4}. Here we report the discovery of traces of human occupation nearly 40,000 years old at Mamontovaya Kurya, a Palaeolithic site situated in the European part of the Russian Arctic. At this site we have uncovered stone artefacts, animal bones and a mammoth tusk with human-made marks from strata covered by thick Quaternary deposits. This is the oldest documented evidence for human presence at this high latitude; it implies that either the Neanderthals expanded much further north than previously thought or that modern humans were present in the Arctic only a few thousand years after their first appearance in Europe.

The Mamontovaya Kurya site is located on the southern bank of the Usa river at the Arctic circle (66° 34' N; 62° 25' E), close to the polar Urals . The riverbed at this site has been known as a place for finding mammoth tusks and bones since the end of the 18th century, but finds of artefacts have not been reported. In order to clarify the stratigraphic context of these bones and to find out if they could be related to early human activities, archaeological and geological field investigations were carried out during the summer seasons of 1992, 1994, 1996 and 1997.

Figure 1 Map showing the location of the Palaeolithic sites Mamontovaya Kurya and Byzovaya discussed in the text and the maximum extent of the Eurasian ice sheets during the Last Glacial Maximum (21,000–18,000 yr BP)10.   Full legend   High resolution image and legend (58k)

A rich faunal assemblage and several stone artefacts were uncovered for the basal layers of a 12–13 m high river bluff which is cut into the terrace along a bend in the river (Fig. 2). The finds, which were scattered throughout the excavated area (48 m²) without any clear concentrations, were incorporated in cross-bedded gravel and sand that accumulated on the floor of an old river channel. Many of the bones uncovered were encapsulated in silt and we also noticed frequent mud clasts within the basal part of the find-bearing channel deposit, which probably reflects slumping from an ancient river terrace covered by over-bank mud. In all, 123 mammalian bones, primarily mammoth (114), but also horse (2), reindeer (5) and wolf (2), were collected (Table 1). The most important find was a 1.3-m-long tusk from a young, 6–8-year-old female mammoth which exhibits a series of distinct grooves (Figs 3 and 4).

The marks are 1–2 mm deep, 0.5–1 cm long and appear as densely spaced rows of lines lying crosswise along the tusk. Microscopic analysis reveals that the grooves were made by chopping with a sharp stone edge, unequivocally the work of humans. It is uncertain whether the marks were formed during processing while using the tusk as an anvil, or if they reflect intentional marks with artistic or symbolic meaning. The stone artefacts that were excavated from the same strata comprise five unmodified stone flakes, a straight side-scraper on a massive cortical blade and a bifacial tool (Fig. 3). The edges of the stone artefacts are sharp and the tusks and bones show minimal signs of wear, indicating a very short transportation and that the material were swiftly buried by alluvial deposits. The few artefacts are not diagnostic and resemble Middle Palaeolithic Mousterian as well as the earliest Upper Palaeolithic assemblages in eastern Europe⁵, a time interval which is also in accordance with the radiocarbon dates discussed below. Similar bifaces are reported for Late Mousterian sites on Crimea, for instance Zaskalnaya V (ref. 6), but they are also known from early Upper Palaeolithic complexes in Eastern Europe, among them Kostenki XII at the Don river⁷. However, we are not able to determine the cultural affiliation on the basis of the sparse material found.

Figure 2 The excavated sediment section at Mamontovaya Kurya on the southern bank of the Usa river.   Full legend   High resolution image and legend (55k)

Figure 3 Drawings of the mammoth tusk with human-made marks, a side-scraper and a bifacial stone tool (knife?) that were uncovered from the excavated river channel deposits at the Mamontovaya Kurya section.   Full legend   High resolution image and legend (35k)

Figure 4 Photograph of the mammoth tusk from Mamontovaya Kurya.   Full legend   High resolution image and legend (23k)

The bones and tusks were in good condition, well suited for radiocarbon dating. The tusk with incision marks was radiocarbon dated to 36,660 ¹⁴C years before present (yr BP) and three other bones from the same unit yielded similar ages in the range of 34,400–37,400 yr BP (Table 2). This time interval is close to the maximum limit for obtaining accurate radiocarbon dates and the calculated standard deviations for age determinations using conventional dating techniques are normally larger than for accelerator mass spectrometry (AMS) dates.

Considering that relatively large amounts of contamination by 'old' inactive carbon is needed to significantly affect the radiocarbon dates, it seems unlikely that the animal remains are significantly younger than the obtained ages. All five radiocarbon dates of various animal remains from the same strata indicate very similar ages. We think it very likely that the artefacts from this layer are of the same age as the tusk and the bones, because the find-bearing strata were buried by several metres of sediment soon after their deposition. Terrestrial plant remains from a slumped mud clast within the find-bearing sand and gravel were dated to 31,380 and 30,160 yr BP by using an AMS technique, indicating that the alluvial formation is younger than the bones.

The find-bearing strata is covered by thick layers of cross-bedded sand followed by ripple- and planar-laminated mud, which together are interpreted as a point-bar sequence (arcuate ridge deposit) that accumulated along the inner bend of a meandering river by the addition of individual accretion accompanying migration of the channel. Then follows a 6–10-m-thick formation of diffusely laminated aeolian (wind-driven) silt and sand, in contrast to the pronounced stratified strata below. A series of eight AMS dates of terrestrial plant remains from the alluvial sediments covering the find-bearing strata yielded ages ranging between 31,420 and 23,860 yr BP whereas optical stimulated luminescence (OSL) dates from the aeolian sediments above give consistently younger ages ranging from 19,900 to 13,800 calendar years BP.

The sedimentological and stratigraphic evidence suggests the following geological history for Mamontovaya Kurya: (1) The refuse of the human occupation was left on the flood plain at around 36,000 yr BP and was shortly thereafter covered by sediments. (2) Slightly before 27,000 yr BP the meandering river undercut these strata and bones and artefacts slumped into the river where they were concentrated in the channel gravel. (3) The bone-bearing gravel was quickly buried by alluvial point-bar deposits as the meander-loop migrated across the site. (4) Aeolian loess-like sediments accumulated on top of the alluvial deposits during the final stage of the Ice Age from 20,000 to 13,000 calendar years ago. (5) Finally the Usa river incised into the terrace during the Holocene and exposed the bones and artefact-bearing layer.

The bone material from Mamontovaya Kurya indicates that humans preyed on, or at least utilized, large herbivorous animals, mostly mammoths. Pollen analysis of the alluvial silt clasts that were found in association with the bones reflects a treeless steppe environment dominated by herbs and grasses, presumably with local stands of willow scrubs (Salix spp) along the river banks⁸. Human occupation probably occurred during a relatively mild interlude of the last Ice Age, although the climate at this time was probably considerably colder and more continental than today. This mild interlude may correspond with the Hengelo interstadial (39,000–36,000 yr BP) in western Europe⁹. A palaeo-environmental reconstruction⁹ suggests that the landscapes in The Netherlands and northern Germany and eastwards were then covered by a shrub tundra. The northern rim of the Eurasian continent was evidently not glaciated¹⁰ and probably only small mountain glaciers existed in the Ural Mountains^{11, 12}. The Scandinavian ice sheet was probably much smaller than during the Last Glacial Maximum some 20,000 yr BP (Fig. 1).

The fact that humans were present in this area as early as around 36,000 yr BP leads us to reassess the history of the earliest human occupation in the Arctic. Until now, the oldest known Palaeolithic sites in the Eurasian Arctic are dated to 13,000–14,000 yr BP^{3, 4, 13}. However, there is an early Upper Palaeolithic site close to the Byzovaya village along the Pechora river, approximately 300 km to the southwest of Mamontovaya Kurya (Fig. 1). At this site nearly 300 artefacts and more than 4,000 animal bones (mainly of mammoth) have been unearthed during several excavations^{12, 14-17}. The lithic industry of Byzovaya is classified as eastern Szeletien with Aurignacian traits^{15, 17}, which is typical for many sites of the early Upper Palaeolithic in Eastern Europe^{5, 18}. An early Upper Palaeolithic age has recently been supported by 13 radiocarbon dates on bones from the find-bearing layer which have yielded ages in the range of 26,000–29,000 yr BP with a mean of 28,000 yr BP¹².

We believe that survival of humans in this arctic environment on a year-round basis would have required long-term planning and an extended social network, qualities that are generally associated with modern human behaviour¹. A pressing question is whether the pioneers who lived in these northern landscapes were members of the ancient Neanderthal population (Homo sapiens neanderthalensis) or newcomers from the south. Most scholars associate the Aurignacian industry—the more advanced stone-tool technology that appeared in Europe at around 40,000 yr BP—with the emergence of modern humans¹⁹. However, the earliest indisputable remains of humans with a fully modern morphology (Homo sapiens sapiens) date to 30,000–35,000 yr BP²⁰; that is, well after the archaeologically defined transition from the Middle to the Upper Palaeolithic. In European Russia, well preserved skeletons from the famous Palaeolithic site of Sungir, northeast of Moscow (Fig. 1), show that anatomically modern humans were present there not later than 28,000 yr BP^{21, 22}.

At the Kostenki IV site on the west bank of the Don river, bones of modern humans have been uncovered from strata dated to 30,000 yr BP²². The stone-working technology reflected in the Byzovaya material is similar to that of Sungir and other early Upper Palaeolithic sites of the eastern Szeletien tradition, indicating that these artefacts were manufactured by modern humans. However, whether the person who inflicted the marks on the tusk from Mamontovaya Kurya, as much as 8,000-9,000 years earlier, belonged to the same human lineage as the residents at Byzovaya and other Palaeolithic sites further to the south is more uncertain. If this person was a modern human who descended from temperate areas, as predicted by the 'Out of Africa' hypothesis², then the Russian Arctic was occupied by Homo sapiens sapiens shortly after the first newcomers entered Europe^{23, 24}. On the other hand, if the person was a Neanderthal, then these humans expanded much further north than hitherto assumed, implying that their stage of cultural development was not a barrier to colonization of this Arctic habitat. Whoever she or he was, the findings from Mamontovaya Kurya provide evidence that the European part of the Arctic was inhabited by humans long before the Neanderthals vanished from the continent soon after 28,000 yr BP^{20, 25, 26}.

Received 27 February 2001;accepted 27 June 2001

References

1.	Gamble, C. Paleolithic Societies of Europe 268-426 (Cambridge Univ. Press, Cambridge, 1999).
2.	Stringer, C. B. & Mackie, R. African Exodus: the Origin of Modern Humanity 84-111 (Cape, London, 1996).
3.	Hoffecker, J. F., Powers, W. R. & Goebel, T. The colonization of the Beringia and the peopling of the New World. Science 259, 46-53 (1993).
4.	Mochanov, Yu. A. Initial Settling of the Territory of North-Eastern Asia (Nauka, Novosibirsk, 1977) (in Russian).
5.	Allsworth-Jones, P. The Szeletian 83-198 (Clarendon, Oxford, 1986).
6.	Kolosov, Yu. G. Mousterian Sites of the Belogorsk Area, Crimea (Nauka, Kiev, 1983) (in Russian).
7.	Praslov, N. D. & Rogachev, A. N. (eds) Paleolit Kostenkovsko-Borschevskogo raiona na Donu 1879-1979 (Palaeolithic of the Kostenki-Borshevo Area on the Don River) 16-66 (Nauka, Leningrad, 1982) (in Russian).
8.	Halvorsen, L. S. Palaeovegetation and Environment during Weichselian Stadials and Interstadials at Mamontovaya Kurja and Sokolova in the Pechora Basin, Northern Russia. Thesis, Univ. Bergen (2000).
9.	Van Andel, T. H. & Tzedakis, P. C. Palaeolithic landscapes of Europe and environs, 150,000-25,000 years ago: An overview. Quat. Sci. Rev. 15, 481-500 (1996). | Article |
10.	Svendsen, J. I. et al. Maximum extent of the Eurasian ice sheet in the Barents and Kara Sea region during the Weichselian. Boreas 28, 234-242 (1999). | Article |
11.	Astakhov, V. I. et al. Marginal formations of the last Kara and Barents ice shelves in northern European Russia. Boreas 28, 23-45 (1999). | Article |
12.	Mangerud, J., Svendsen, J. I. & Astakhov, V. I. Age and extent of the Barents and Kara ice sheets in Northern Russia. Boreas 28, 46-80 (1999). | Article |
13.	Powers, W. R. in Humans at the End of the Ice Age--The Archaeology of the Pleistocene-Holocene Transition (eds Straus, L. G., Eriksen, B. V., Erlandson, J. M. & Yesner, D. R.) 229-253 (Plenum, New York, 1996).
14.	Kanivets, V. I. The Paleolithic of the Extreme North-East of Europe (Nauka, Moscow, 1976) (in Russian).
15.	Pavlov, P.-Yu. The excavation of Byzovaya Upper Palaeolithic site in 1983-1985 in Pamiatniki materialnoj kultyury na Evropeiskom severo-vostoke 7-16 (Russian Acad. Sci., Syktyvkar, 1986) (in Russian).
16.	Pavlov, P. Yu The Palaeolithic Archaeology of the Komi Republic 44-91 (DiK, Moskva, 1996) (in Russian).
17.	Pavlov, P. Yu. & Indrelid, S. Initial settling of North-Eastern Europe. Bull. Inst. Biol., Syktyvkar 2, 24-26 (1999)
18.	Anikovich, M. V. Early Upper Palaeolithic in Eastern Europe (AN SSSR, St Petersburg, 1991) (in Russian).
19.	Straus, L. G. Age of Modern Europeans. Nature 342, 476-477 (1989).
20.	Smith, F. H., Trinkhaus, E., Pettitt, P. B., Karavanic, I. & Paunovic, M. Direct radiocarbon dates for Vindija G1 and Velika Pecina Late Pleistocene hominid remains. Proc. Natl Acad. Sci. USA 96,12281-12286 (1999). | Article | PubMed |
21.	Bader, O. N. Sungir-Upper Palaeolithic Site (Nauka, Moskva, 1978) (in Russian).
22.	Sinitsyn, A. A. & Praslov, N. D. Radiocarbon chronology of the Paleolithic of Eastern Europe and Northern Asia. Problems Perspectives (Russian Acad. Sci., St Petersburg, 21-66 1997).
23.	Bocquet-Appel, J.-P. & Demars, P. Y. Neanderthal contraction and northern human colonization of Europe. Antiquity 74, 544-552 (2000).
24.	Straus, L. G., Bicho, N. & Winegardner, A. C. The Upper Palaeolithic settlement of Iberia: first-generation maps. Antiquity 74, 553-566 (2000).
25.	Ovchinnikov, I. V. et al. Molecular analysis of Neanderthal DNA from the northern Caucasus. Nature 404, 490-493 (2000). | Article | PubMed |
26.	d'Ericco, F., Zilhao, J., Julien, M., Baffier, D. & Pelegrin, J. Neanderthal acculturation in Western Europe? A critical review of evidence and its interpretation. Curr. Anthrop. 39 (Suppl.), 1-44 (1998).
27.	Bar-Yosef, O. in Neanderthals and Modern Humans in Western Asia (eds Akazawa, T., Aoki, K. & Bar-Yosef, O.) 39-56 (Plenum, New York, 1998).

Acknowledgements. This work is a contribution to the Russian–Norwegian research project Paleo Environment and Climate History of the Russian Arctic (PECHORA), which forms part of the European Science Foundation Program Quaternary Environment of the Eurasian North (QUEEN). We thank E. Giria, Institute of History of the Material Culture, St Petersburg University, for carrying out microscopic analysis of the marks on the mammoth tusk. We thank J. Mangerud for his comments on this manuscript and for discussions. The bond material was identified by I. Kuzmina and D. Ponomarev. The drawing of the tusk and the stone artefacts were done by N. Pavlov and the figures by E. Bjørseth. We thank the Norwegian Research Council for financial support.

---

Archaeology: Out in the cold

JOHN A. J. GOWLETT

John A. J. Gowlett is in the Department of Archaeology, University of Liverpool, Liverpool L69 3BX, UK.
e-mail: gowlett@liverpool.ac.uk

Humans are very adaptable: during the last ice age, they apparently lived within the Arctic Circle. The discovery suggests that, although cold, the region was probably not covered in ice at the time.

Archaeological finds described by Pavlov and colleagues on page 64 of this issue¹ show for the first time that humans were present north of the Arctic Circle almost 40,000 years ago, in the last ice age. The idea of people living in a land gripped by an ice age goes back to nineteenth-century France, but the new finds extend both the geographic and the temporal range of the phenomenon. The results should also rekindle debate about the effects of the climate on the movements of early human populations.

Pavlov et al.¹ carried out fieldwork at a site in the Russian Arctic known as Mamontovaya Kurya, which dates to Middle to Upper Palaeolithic times, some 35,000–40,000 years ago. Their finds comprise various stone tools and over a hundred mammalian bones, as well as a mammoth tusk bearing cut marks that were apparently made by tools. The age of the tusk was determined by a radiocarbon-dating technique known as accelerator mass spectrometry, illustrating the power of this technique for dating artefacts directly rather than by the age of the sediments in which they are found.

It is not possible from these finds to determine whether they were left by Neanderthals or by some of the first modern humans in Europe, but this is equally true of most contemporary artefacts further south. In the broader scheme of things, knowing who made the tools is less important than simply knowing that someone was adapted to the cold conditions. This is significant because all evidence from recent foragers (such as Inuit or Siberian Yukaghir) suggests that adaptation to northern climes requires high levels of technological and social organization.

That said, it would be interesting to know whether these people were Neanderthals or early 'anatomically modern' humans. If they were Neanderthals, this provides further support — along with their anatomical adaptations and the height and remoteness of many of the sites at which Neanderthal artefacts have been found — of the Neanderthals' rugged durability and extensive capabilities². Their high degree of meat-eating, indicated by recent studies of stable carbon and nitrogen isotopes in bone³, also suggests a specialized socioeconomic adaptation, perhaps developed over a long period in environments rich in animals but limited in plant resources.

If, on the other hand, these ice-age people were modern humans, then this is evidence of a remarkably rapid advance to the north — modern humans had only just set foot in the southeast of Europe. Pavlov et al.¹ incline towards this view — the nearest archaeological finds to the south, along the Pechora River near the Urals^{4, 5}, are allied more closely with those Upper Palaeolithic traditions associated with modern humans than with the Middle Palaeolithic toolkits more commonly associated with Neanderthals.

The bones and artefacts found at Mamontovaya Kurya suggest that the northeast must have been relatively dry and ice-free in this period of the ice age. These finds are one outcome of a major interdisciplinary study that has also shown that, for most of the time, the ice sheets of the last glaciation were far more restricted on their eastern flank than is sometimes suggested. Support for the existence of large ice-free areas also comes from Finland⁶, where direct radiocarbon dating of key evidence — this time a series of mammoth teeth — establishes the presence of large animals between 22,000 and about 40,000 years ago. The existence of large animals implies that the environment was steppe-like, consisting of open grassland.

Icy cold, however, it was. Temperature estimates derived from variations in the ¹⁸O content of Greenland ice cores such as GRIP2 (at a latitude of about 72° north) show wild fluctuations throughout the middle of the last glaciation⁷ (Fig. 1), but it was always at least 10 °C colder than today. The dominant feature of the time from 60,000 to 30,000 years ago was a series of saw-toothed temperature fluctuations of up to 15 °C. Similar temperatures were found in intensive research across north-central Europe, using indicators such as beetle remains and pollen as proxy evidence⁸. These indicate an average annual temperature of -1 °C in the Netherlands from 50,000 to 41,000 years ago, with the coldest month being at least 10 °C below this⁹.

Figure 1 Climate and early human populations.   Full legend   High resolution image and legend (56k)

Yet, despite this newly detailed backdrop, recent archaeology-based discussion about the Neanderthals has not — with certain exceptions¹⁰ — been concerned primarily with climate. The emphasis has been on chronology, population movements, and the nature of cultural contact (if indeed there was any) between Neanderthals and incoming modern humans (Box 1). There has been good reason to focus on the cultural changes that occurred in the past 40,000 years, as this time period includes the more rapid developments of the Upper Palaeolithic era and falls within the range of radiocarbon dating. Climate change has seemed less important because different Neanderthal populations successfully made distinct adaptations to different regions¹¹, and these adaptations remained roughly the same for 200,000 years¹².

But climate may have its moment again. The dramatically spiky record from ice cores in the interval from 60,000 to 40,000 years ago, together with pollen evidence, implies that steppe environments moved up and down rapidly from southeast Europe to the far north, and suggests that climate change could have been crucial in promoting population movement and cultural change. In 'warmer' parts of the ice age, as Pavlov et al.¹ show, fauna-rich steppe environments and humans apparently reached the Arctic. During colder intervals, wooded environments gave way to steppe even in Greece¹³. In the Last Glacial Maximum, 20,000 years ago, conditions were so ferociously cold that even modern humans were driven down towards the south of France¹⁴. Indirectly, such responses may help to explain the southward expansion of Neanderthals into the Middle East around 60,000 years ago, and (perhaps) the similar spread of Upper Palaeolithic Aurignacian human populations around 30,000 years ago.

The new finds¹ show that humans had a hold on the north, if only for a short time. Although there are questions to be answered, the artefacts illustrate both the capacity of early humans to do the unexpected, and the value of archaeologists researching in unlikely areas.

 

References

1.	Pavlov, P., Svendsen, J. I. & Indrelid, S. Nature 413, 64-67 (2001). | Article |
2.	Hayden, B. J. Hum. Evol. 24, 113-146 (1993). | Article |
3.	Bocherens, H. et al. J. Archaeol. Sci. 26, 599-607 (1999).
4.	Ivanova, I. K. L'Anthropologie 73, 5-48 (1972).
5.	Allsworth-Jones, P. in The Emergence of Modern Humans (ed. Mellars, P.) 160-242 (Edinburgh Univ. Press, 1990).
6.	Ukkonen, P., Lunkka, J. P., Jungner, H. & Donner, J. J. Quat. Sci. 14, 711-714 (1999).
7.	Johnsen, S. J. et al. J. Quat. Sci. 16, 299-307 (2001).
8.	Vandenberghe, J., Kasse, K. & Coope, R. (eds) J. Quat. Sci. 13, 361-497 (1998).
9.	Huijzer, B. & Vandenberghe, J. J. Quat. Sci. 13, 391-417 (1998).
10.	Bar-Yosef, O. in Paleoclimate and Evolution, with Emphasis on Human Origins (eds Vrba, E. S., Denton, G. H., Partridge, T. C. & Burckle, L. H.) 507-523 (Yale Univ. Press, New Haven, CT, 1995).
11.	Gamble, C. The Palaeolithic Societies of Europe (Cambridge Univ. Press, 1999).
12.	Mellars, P. et al. Curr. Anthropol. 40, 341-364 (1999).
13.	Okuda, M., Ysuda, Y. & Setoguchi, T. Boreas 30, 73-82 (2001).
14.	Bocquet-Appel, J.-P & Demars, P.-Y J. Archaeol. Sci. 27, 551-570 (2000).
15.	Bilsborough, A. in The Hominids and their Environment during the Lower and Middle Pleistocene of Eurasia: Proceedings of the International Conference of Human Palaeontology, Orce, 1995 (eds Gibert, J., Sanchez, F., Gibert, L. & Ribot, F.) 311-315 (Museo de Prehistoria y Paleontologia, Orce, 1999).
16.	Fox, R. G. (ed.) Curr. Anthropol. 39 (suppl.), S1-S189 (1998).
17.	Churchill, S. E. & Smith, F. H. Yb. Phys. Anthropol. 43, 61-115 (2000).