ARTICLE IN PRESS
Quaternary International xxx (2010) 1e9
Contents lists available at ScienceDirect
Quaternary International
journal homepage: www.elsevier.com/locate/quaint
Relations between climatic changes and prehistoric human migrations during
Holocene between Gissar Range, Pamir, Hindu Kush and Kashmir: The
archaeological and ecological data
Anne Dambricourt Malassé*, Claire Gaillard
UMR 7194 CNRS, Department of Prehistory, National Museum of Natural History, Institut de Paléontologie Humaine, 1, rue René Panhard, 75013 Paris, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online xxx
In the beginning of the Holocene, hunteregatherer populations settled in the Pamir plateau, especially in
the tectonic depression of Markansu near the Kara Kul Lake (Osh Kona site, 4100 m, 9580 BP and 7145
BP). These populations were well adapted to high altitude life conditions in summer. They were using
stone tools related to a cobble tool tradition that Ranov considered as the local continuation of an Asian
Middle Palaeolithic (Ferghana, Darwaz or Punjab). However, they disappeared during the second half of
the Holocene. As well, huntersegatherers characterised by a tradition close to Markansu Culture, namely
Hissar Culture, appeared during the Holocene in the northern mountains of the TajikeAfghan depression
(Amu Darya upper basin) occupying the middle altitude valleys (Gissar Range) when Neolithic populations began to develop at lower altitudes. For Ranov, origin and destiny of both Markansu and Hissar
cultures were the most important questions to solve in this region. The Department of Prehistory,
National Museum of Natural History (Paris) conducted field work in collaboration with the University of
Peshawar between 1996 and 1998 in the northernmost valley of the Indus basin giving access to the
Wakhan corridor. The upper Yarkhun valley (Chitral district, Pakistan), accessible only by foot, was
surveyed from 2500 to 4000 m altitude. This allowed the discovery of 6 sites yielding lithic artefacts,
among which half are cobble tools. They attest to human activity later than 5500 BP, suggesting population movements between the Amu Darya and Indus basins. The consequences of the Quaternary
climate oscillations between South Asian monsoon and Northern hemisphere influences on the one
hand, and the impact on biotopes of the increasing pastoral activity on the other hand, are discussed as
probable causes for the disappearance of the mountainous nomadic hunters of Central Asia.
Ó 2010 Elsevier Ltd and INQUA. All rights reserved.
1. Introduction
between the Indian and the Eurasian tectonic plates. Recent
isotopic analysis from the enamel of herbivores suggests that the
Tibetan plateau was 2700 m below its present elevation 2e3
million years ago (Wang et al., 2008). The elevation has increased
faster since the Pliocene, and transformed the landscapes and the
climate during the Pleistocene. Surveys in the sub-Himalayan and
TajikeAfghan foothills as well as in higher regions such as eastern
Pamir, have shown evidence of prehistoric occupations, but
a distinction is made between low-middle altitude and high altitude zones (above 2000 m). The first human occupations are
known since the Lower Pleistocene: 2 Ma in the Potwar plateau in
the western Siwaliks (Rendell and Dennell, 1985; Rendell et al.,
1989) and Tajik loess (Ranov, 1995, 2005; Ranov and Dodonov,
2003, Dennel 2004, 2008), whereas in the Pamir, settlements
seem to be rare before the Holocene (Fig. 2).
One specific feature of the Holocene populations in the foothills
and high plateaus is the persistence of a stone knapping technology
contemporary to the Neolithic culture, that recalls the Soanian
The sharp contrast between the highest mountains in the world
with peaks exceeding 7000 m altitude in Pamir, and the two Amu
Darya and Indus upper basins, confers specificity to the human
settlements in this region since the Late Pleistocene and particularly during the Holocene. Due to the strong gradients the climate
changes made the biotopes shift more vertically than horizontally.
As far as early human populations are concerned, this leads to
different adaptations (behaviour and biology), since it is not
necessary to move on long distances for escaping drastic environmental changes (Fig. 1).
The Himalayas, Hindu Kush, Pamir, and Tibetan Plateau are
young mountainous formations emerging from the collision
* Corresponding author. Tel.: þ(33) 01 55 43 27 39; fax: þ(33) 01 43 31 22 79.
E-mail addresses: [email protected] (A. Dambricourt Malassé), [email protected]
(C. Gaillard).
1040-6182/$ e see front matter Ó 2010 Elsevier Ltd and INQUA. All rights reserved.
doi:10.1016/j.quaint.2010.04.001
Please cite this article in press as: Dambricourt Malassé, A., Gaillard, C., Relations between climatic changes and prehistoric human migrations
during Holocene..., Quaternary International (2010), doi:10.1016/j.quaint.2010.04.001
ARTICLE IN PRESS
2
A. Dambricourt Malassé, C. Gaillard / Quaternary International xxx (2010) 1e9
the Pamir. This paper will discuss the hypothesis of the development and the disappearance of a vast hunting territory during the
Holocene between Amu Darya and Indus upper basins in the light
of the Pleistocene human settlements and Quaternary climate
changes.
2. Southern Tajikistan
The southern part of Tajikistan comprises two contrasted
regions: the Pamir with high valleys and plateau on the one hand,
and the hills of the the Gissar Range covered with loess on the other
hand (Fig. 2). Exploration in this sector began in 1956 with the
archaeologist V. A. Ranov and later developed in collaboration with
J. Schäfer in Gissar Range (Schäfer et al., 1998) and V.A. Zhukov in
Pamir (especially the Mesolithic Istyk cave).
2.1. AfghaneTajik depression and Gissar Range
Fig. 1. Localisation of Hindu Kush-Himalayan Region (Yarkhun, white arrow).
technical tradition frequently found on the surface of the subHimalayan terrain, and is marked by the large utilisation of
quartzite cobbles for making cobble tools and flakes. The question
was raised of a common background or technical template between
the mountainous regions of Pamir and the Indus upper basin
including the Himalayan foothills and Sub-Himalayan ranges
(Siwaliks). Three expeditions in 1996, 1997 and 1998 in the high
valleys of Hindu Kush, close to the Pamir plateau, investigated the
possibility of finding traces of such a local tradition (Dambricourt
Malassé, 2008; Gaillard and Dambricourt Malassé, 2008). Lithic
assemblages have been discovered in the upper Yarkhun valley
(southern slopes of the Hindu Kush Range, Pakistan), in six localities between 2500 and 4000 m altitude (Gaillard et al., 2002), the
last one close to the Boroghol Pass (3800 m of altitude) which
allows communication between the Hindu Kush high valleys and
The sites located in southern Dushanbe are in a clear stratigraphical context, in well-studied long sequences of loess including
many pedocomplexes, which reflect warmer climatic phases during
the Pleistocene (Ranov, 1982, 1984, 1995, 2001; Ranov and Schäfer,
2000; Ranov et al., 1999). The earliest occupation appears at Kul’dara, in the pedocomplexes 12 and 11, dated to about 900e950 ka
(Ranov and Dodonov, 2003; Ranov, 2005). At the end of the
Palaeolithic and preceding the Neolithic, two technical traditions
co-existed in Central Asia, the Mesolithic with geometric microliths
and an Epipalaeolithic resulting from a local evolution of the
Palaeolithic cobble tool tradition (Ranov, 1993, 2003, 2005).
Then, contemporary to the Neolithic, a series of sites discovered
in the 1960s by Okladnikov revealed the presence of a hunting
territory associated with a specific environment. These sites usually
occur on the lower river terraces (20e30 m), at altitudes ranging
from 700 to 1500 m, always in the context of loess deposits, surrounded by mountains. They are mostly open air sites where fireplaces are not common (Ranov, 1982; Amosova et al. 1993). The
lithic assemblages, attributed to the Hissar Culture, include
numerous tools on cobbles from a large variety of rocks. However,
flint was also used, and in this case the knapping technology is well
mastered and oriented towards the production of flakes and small
blades. There is no pottery, but some polished axes and a few
ornaments in bone are found. Among the bone remains, domesticated as well as wild animals are represented. In the later stages,
grinders and pestles increase in number, suggesting increasing
consumption of grains. The economy of these populations observed
during the VIIIeVII millennia was based on hunting and gathering.
While the neolithisation clearly occurred in some localities during
the VIeIV millennia in vast areas (1e1.5 ha) (Ranov 1984; Amosova
et al., 1993), the destiny of those hunteregatherer tribes remains
unclear. Two possible evolutions have been proposed: shift towards
either farmingebreeding economy or pastoralism (Ranov, 1984,
2005). Nevertheless nothing allows systematic conclusion of
a transformation driven by an inner trend. Hunteregatherers were
living in the same environment as farmers; some of the tribes could
collect cultivated grains and kill some animals from the livestock,
all the more easily as they were domesticated. A third hypothesis
which had not been taken into account is the extinction of some
tribes by endemism and disorganisation of their ecosystem,
because of interferences between climate changes and anthropic
activities such as deforestation and overgrazing.
2.2. Pamir
Fig. 2. Map of Western Himalaya, Pamir, Hindu Kush; Southern Tadjikisan: mountainous systems (arrow), river basins (star), archaeological stations (bubble) and
glaciers (circles) (satellite view from Google Earth).
Above 2000 m altitude, prehistoric evidence is scarce before the
early Holocene. In the northeastern region, a Late Pleistocene
Please cite this article in press as: Dambricourt Malassé, A., Gaillard, C., Relations between climatic changes and prehistoric human migrations
during Holocene..., Quaternary International (2010), doi:10.1016/j.quaint.2010.04.001
ARTICLE IN PRESS
A. Dambricourt Malassé, C. Gaillard / Quaternary International xxx (2010) 1e9
occupation is known from the Djamanta cave on the Karasu river,
the archaeological layer yielded broken bones and fires-places but
no tools (Ranov and Sidorovo, 1979). Similar occupations are also
known in south Pamir from many open-air sites dated to between
20 and 12 ka BP in the Alichur Valley (Ranov, 1984).
Ranov did not rule out the possibility of earlier human occupations, even as early as the Middle Pleistocene when the Pamir
was below its present altitude. Although rare, some isolated
Mousterian or Upper Palaeolithic artefacts, dating from 40 to 20 ka
were actually collected (Ranov, 1984). PlioePleistocene river
terraces are still preserved and they have not recorded any trace of
ice sheets (Bondarev et al., 1997). A partial isolated temporal bone,
dated to 30 ka and excavated from the Darra-i-Kur rock-shelter in
Badakhshan (Afghanistan) at 1300 m altitude clearly shows biological adaptation (see below). Therefore, altitude is not a sufficient
argument to explain the scarcity of archaeological evidence before
the early Holocene. The severely low temperatures during the Last
Glacial Maximum seem to be a satisfying hypothesis.
In contrast, numerous localities with pebble tools were recorded
on Holocene terraces in the entire Eastern Pamir. Most remarkable
are two open-air sites. The first one, in the north, is the Epipalaeolithic site of Osh-Khona in the Markansu tectonic depression,
near the Kara Kul Lake, at the altitude of 4100 m (Ranov, 1984, 1988,
1993). The second one, in the southeastern Pamir, is Karatumshuk
(Tokhtamych) near the Sarykol Range and the Chinese border. In
the southeastern Pamir, the sites are located along the Alichur and
Murgab Rivers (Figs. 2 and 3).
Today, the Markansu depression (Margansu valley or Death
Valley) offers a desolated mineral landscape drastically contrasting
with the early Holocene conditions (Fig. 4). Half a dozen open air
sites (on moraines and river terraces) provide lithic assemblages
characterized by two stone knapping processes, one producing
pebble tools and the other one producing microliths and microblades. This particular tradition, the “Markansu Culture”, mixes
Epipalaeolithic and Mesolithic. While hunting activity is attested
from the Epipalaeolithic sites of Markansu, a proto-pastoralist
economy cannot be excluded in the other sites.
Fig. 3. Localisation of Holocene archaeological stations in eastern Pamirs (in Frumkin,
1970).
3
Fig. 4. UjsuRiver, Osh-Khona archaeological site (arrow) and Kara Kul Lake (satellite
view adapted from Google Earth).
Such a co-existence is observed during the same period in the
northern Tibetan plateau, around Quingkai Lake (Western China).
Jiangxigou 2, dating to ca. 9000e5000 BP with microlithic tools and
a small number of ceramic shards, “appears to be connected to early
Neolithic agricultural settlements along the upper Yellow River (Huang
He) drainage during the middle Holocene, and may provide insights
into foragereagriculturalist interactions that lead to the development
of pastoralist systems in the region”, whereas Heimahe 3 is, “a brief
hunter’s camp dating ca. 8450 cal BP, with evident affinities to late
Upper Paleolithic camps in the same region that date several thousand
years older” (Rhode et al., 2007).
Osh-Khona archaeological site is located on the right bank of the
Ujsu River, at its confluence with the Oshkhona Zhilga River, 15 km
from the modern Ujsu glacier terminus (Ranov and Sidorovo, 1979).
The excavations have exposed a cultural layer typical of hunters’
summer campsites, with three 14C dates: 9530, 7380 and 7095 BP
(Velitchko and Lebedeva, 1974; Ranov and Davis, 1979). All the bone
remains belong to wild fauna: large mammals (Ovis, Capra), rodents
(chiefly marmots and hares) and birds. Charcoal remains from fireplaces represent three taxa: Betuta (birch), Juniperus (juniper) and
Eurotia ceratoides (winterfat) (Ranov and Sidorovo, 1979). The
assemblage includes a large number of cobble tools, small blades in
selected flint, end and side scrapers, etc.
The most difficult question to solve is the origin of these
hunters. Ranov thought they could not originate from Xinjiang,
where a quite distinct SiberoeMongolian tradition occurs. He
suggested they arrived from the neighbouring terrain of Fergana,
easier to access but with important cultural adaptations. In the
southeast, the hypothesis of a North-West Indian or Far West
Tibetan origin for the hunters’ open-air site of Karatumshuk cannot
be dismissed, even though few archaeological data are available.
Osh-Kona is the only human settlement for which climate
changes are well documented, through multidisciplinary investigations leading to reconstruct oscillations during the final Pleistocene and Holocene (Serebryanny and Solomina, 1996; Bondarev
et al., 1997). The age of the Last Glacial Maximum (LGM) is
disputed, as it occurs either at ca. 44 ka or at 14e10 ka. Nevertheless, in either case the Kara Kul basin was free of ice during the
interval 15,660 BP to 3240 BP, and the major deglaciation occurred
ca. 10,000 years ago, at the end of Batura glacial stage
(10.8e9.0 ka BP) (Lewis et al., 2002). Three lake terraces have been
identified: T1: 8e7 m above river level of Markansu valley dated
between 12,200e11,800 BP, T2: 6e5 m, 9000e8500 BP and T3:
1.5e1 m (Bondarev et al., 1997). The summer campsite is associated
with the second terrace, but the settlement could have been
Please cite this article in press as: Dambricourt Malassé, A., Gaillard, C., Relations between climatic changes and prehistoric human migrations
during Holocene..., Quaternary International (2010), doi:10.1016/j.quaint.2010.04.001
ARTICLE IN PRESS
4
A. Dambricourt Malassé, C. Gaillard / Quaternary International xxx (2010) 1e9
interrupted between 9500 and 7400 BP by moister climatic
condition ca. 8700 BP, when the glaciers re-advanced and the lake
level increased. A new phase of occupation occurred until 7000 BP
then the hunters abandoned the site.
Zech et al. (2005) concluded that, “Increasing aridity in the Pamir
is most likely responsible for the progressively limited extent of the
glaciers during the Late Pleistocene”. The causes of this increasing
aridity cannot be imputed to a rapid uplift of the Pamir, as the
tectonic rise in the last ten thousand years did not exceed 2 or 3 cm/
y (Agakhanyantz, 1996).
The culture directly overlying the hunters’ settlements is the
Bronze Age and especially the pastoral Kajrat-Kum Culture: no
transitional Eneolithic has ever been observed. The hunters abandoned the Pamir plateau during the second period of the Holocene,
probably because of increasing aridity generating an impoverishment of the biodiversity until the present. The Kara Kul Lake is the
largest of the High Pamir (tectonic origin, area: 38,000 ha,
maximum depth: 242 m), but today it is remarkably poor in fish,
and just one endemic species (Noemacheilus lacusnigri) reproduces.
For comparison, Yashikul Lake (moraine and landslide origin, area:
4800 ha, maximum depth: 50 m) shelters 3 species (Savvaitova and
Pietr, 1999). Prehistoric populations hunted large numbers of birds:
in 1976, a maximum of 230 Tibetan Sandgrouses were identified in
the huge Kara Kul ecosystem (Yakovlev, 1976). Kara Kul Lake is
considered a dead lake: no study bears on the evolution of the fish
species during the Holocene, and it is difficult to know whether the
biodiversity was reduced quickly during the last centuries or
progressively since the disappearance of the last nomadic hunteregatherers of Central Asia.
Increasing aridity provoked greater salinity of the Kara Kul Lake
but also changes in air circulation. The French surgeon Marie Jeanne
Koffman, a member of the 1958 Pamirian Expedition headed by B.A.
Lidvintchskij, describes the Markansu depression not only as
a desolated landscape but as a huge rocky desert with dust devils
able to project stone blocks (personal communication). The dust
devil is an ascent of warm and dry air created by a very powerful
thermal phenomenon (zone of warm air surrounding by cold air).
The Kyrgyz people precisely name this depression “creek of
tornadoes” or “creek of death”, “Markansu”.
Ranov considered the Hissar culture as an enigma, nevertheless
observing similarities with the Markansu Culture (Ranov, 1984). In
comparing the high valleys around the Kara Kull depression, it
clearly appears that communications were easy with the Gissar
Range (Fig. 1). Moreover, archaeologists recorded the same microlithic points in Istyk cave (Istykskaja, South-Eastern Pamir, 37440 N,
74 230 E) and in the AfghaneTajik depression as in Tutkaul (Gissar
Range) and on the other side of the Amu Darya, in the dunes of
Northern Afghanistan (Ranov, 2003). So, Ranov considered the
Hissar Culture as a component of the Eastern Pamirian EpipalaeolithiceNeolithic related to hunting economy, and based on
seasonal activities between the middle-low valleys and the Pamir
plateaus. Unfortunately, all the Pamirian valleys have not been
explored, especially the Gorny Badakhshan and the Darvaz Range
respectively in Central and South Western Pamir (Fig. 1) where
archaeological records could provide more precise observations.
Moreover, movements between South Eastern Pamir and India
cannot be excluded as Ranov and Sidorovo concluded, “On the south
and south-west of the Pamir we find the well-known rich Palaeolithic
sites of Punjab (.) H. de Terra’s previous works have led to the
discovery of much later Mesolithic cultures. These industries, in our
opinion, show a definite similarity with the Gissar culture” (works of
de Terra et al. 1934, 1936; Teilhard de Chardin and de Terra 1936;
Ranov and Sidorovo, 1979, p. 176).
During the Holocene, the Southern Hindu Kush highest ranges
were exposed to continental monsoons generating asynchronies
well described for the Nanga Parbat (North Pakistan, Karakoram
range, 440 km south of Kara Kul Lake; Fig. 2). When the subcontinental temperatures were increasing along with humidity, the
high massifs of the South Karakoram and South Hindu Kush were
exposed to strong monsoons and the glaciers advanced (William
et al., 2000) whereas in the opposite, the Tibet plateau was influenced by the, “Northern Hemisphere climate oscillations (rapid
climate changes), with minor influences from the South Asian
monsoon” (Seong et al., 2009). So, the coldest phases of the Indian
Sub-continent were not necessarily restricting conditions for the
North-Western Indian hunters to migrate to the Indus high valleys
and reach the Wakhan Corridor.
2.3. Anthropological features
Only two Pleistocene human fossils are known in this huge
palaeozone from sub-Himalayan borderlands to Gissar Range. The
first one is the famous Neandertal child excavated from the Teshik
Tash rock-shelter (180 m) in Uzbekistan, and associated with
a Mousterian lithic industry (Movius,1953a,b). However, this Middle
East or European technical tradition has never been recorded in the
High Pamir. The second human fossil is the petrous bone from Darrai-Kur rock-shelter (1300 m; Afghanistan). This fragmented bone is
quite similar to the Middle East Protocromagnoïds (Angel, 1972).
Nevertheless the skull base developed particular features, for
instance the important size of the foramen that corresponds to the
vascularisation of the middle ear. The internal acoustic meatus (a
neuro-vascular orifice) is thrice the normal size of that of Homo
sapiens in cross section. This implies more oxygenised blood.
Besides, the imprint of an important muscle of the vocal tract, the
fossa digastrica, is especially deep and indicates a powerful muscular
system sustaining the larynx. The styloïd bones are also robust as
a part of this system sustaining the throat; some tendons and
muscles fixed on the styloïd bones help in opening the Eustachian
tube to equalize middle-ear pressure with the outside atmospheric
pressures in altitude. Therefore the biology was clearly adapted to
the high altitude conditions and to important changes of the
atmospheric pressure. The corpse was not discovered in its natural
environment (1300 m) and the most frequent life conditions for this
person were higher biotopes in Pamir or Hindu Kush (Dambricourt
Malassé, 2008; Gaillard and Dambricourt Malassé, 2008). Only
four Mesolithic skeletons are known from the excavated site of
Tutkaul: two women and two children. The skulls alone have been
described. The adults had developed robust features and they cannot
be related to any geographical type (Kiyatkna and Ranov; 1971;
Dambricourt Malassé, 2008).
Fig. 5. Yarkhun valley, Zesht-O-Gust archaeological site (arrow) in 1996 (picture C.
Gaillard).
Please cite this article in press as: Dambricourt Malassé, A., Gaillard, C., Relations between climatic changes and prehistoric human migrations
during Holocene..., Quaternary International (2010), doi:10.1016/j.quaint.2010.04.001
ARTICLE IN PRESS
A. Dambricourt Malassé, C. Gaillard / Quaternary International xxx (2010) 1e9
3. Hindu Kush: the Yarkhun Valley
The Gissar ‘enigma’ (Ranov, 1984) actually raises the question of
the extent of hunting territories in the high massifs from the Final
Pleistocene (major deglaciation) to the Middle Holocene, when
fauna exposed to deterioration of the climatic conditions, was
forced to withdraw in the lower valleys. The southward expansion
of this territory was unknown, and that is why in 1996, 1997 and
1998 the authors explored one of the highest valleys of the Hindu
Kush Range, between 2500 and 4000 m high (FrenchePakistani
missions funded by the French Ministry of Foreign Affairs).
5
The Yarkhun valley was selected for its upper reaches are
directly connected with the Pjandzh River (Wakhan Corridor)
through one of the lowest passes of the sector, the Boroghol Pass
(3800 m; Fig. 1). The geology consists of metamorphic series
including mainly marble, quartzite, micaschist, schist, gneiss, and
amphibolites. After several days of walking, the first sites were
discovered on the Yarkhun river terraces, 10e25 m above water
level and about 400e1000 m away from the main river bed. All
are open-air sites and yield only lithic industry. There are
neither any organic remains, fire-places or settlement structures
(Fig. 5).
Fig. 6. Stone tools of the Yarkhun valley (drawings by C. Gaillard).
Please cite this article in press as: Dambricourt Malassé, A., Gaillard, C., Relations between climatic changes and prehistoric human migrations
during Holocene..., Quaternary International (2010), doi:10.1016/j.quaint.2010.04.001
ARTICLE IN PRESS
6
A. Dambricourt Malassé, C. Gaillard / Quaternary International xxx (2010) 1e9
From downstream to upstream along the upper Yarkhun River,
six spots and sites were identified: Nakht Sir Dem (NAK): 3 artefacts, Shusht (SHT): 3 artefacts, Lasht-Savalior (SAV): 24 artefacts,
Zeshtsh-o-Gutsh (ZTG), altitude 3350 m: 42 artefacts, BorogholChilmarâbâd (BOR), altitude 3650 m: 7 artefacts, Thinioupk (THI),
altitude 3900 m: 18 artefacts.
Most of the tools are made in amphibolite (80%). Only in the
upper reaches of the Yarkhun the quartzite is preferred. There are
also a few tools in marble, but crystalline rocks such as gneiss and
schist, the most common in the local landscape, were apparently
not used. The large majority of the tools are core tools of rather
large size; the smaller items, particularly cutting tools, are rare in
these series and one can wonder whether there was no small
cutting tool at all or whether they were made in other raw materials, such as schist for instance, of which all the natural fragments
are cutting, or even wood. Many of the tools (56% at ZTG) are on
cobbles, usually flat, others being on different types of blanks,
especially flakes and small slabs. The flakes are always slightly
abraded and it is possible that they were selected among the
components of the gravel. The shaping of the tools is very light and
does not modify much the original morphology of the blanks.
Typologically, they are not well defined and only a few of them
display standard morphologies occurring in different sites. These
more standardised tools may deserve the names of axes or adzes,
with triangular or rectangular outline (Fig. 6). In some cases the
lateral sides are concave, either naturally or resulting from a large
flake scar; these recall one type of the Aceramic Neolithic of
Kashmir (“waisted tool”). Besides, there are a few choppers and
chopping tools characterised by large and deep flake scars.
As far as chronology is concerned, it is difficult to assign an age
to these sites, since the tools are usually lying on the surface or
hardly buried in the deposit of the lowest terrace. They are slightly
rolled and none of them is in mint condition, therefore they cannot
be very recent (Fig. 7).
The only chronological reference is the Quaternary geomorphology. During the Last Glacial Maximum, the snowline in Hindu
Kush was located at 900e1100 m high (Porter, 1985) and the
bottom of the Yarkhun valley was covered by a glacier. The profile of
the Boroghol Pass shows a typical glacial valley profile. Communication between both sides of the Hindu Kush Range (Wakhan and
Fig. 8. Terraces in Lasht sector, with the brown boulders (arrow).
Chitral) was possible only by walking on the glaciers. Necessarily
the artefacts are much more recent than Late Pleistocene.
When the valley is broad enough, a few terraces can be observed
along the river (2e4). All along the Yarkhun, in the sector of Shusht
and Lasht, the alluvium includes a conspicuous level of boulders
(30e50 cm) bearing a dark brown patina. These brown boulders
usually lie close to the Yarkhun water level. At the site of Zeshtsh-oGutsht, this level is well exposed on the bank and appears as a boulder
conglomerate hardened by cementation of the sandy matrix (Fig. 8).
Tentatively, this dark patination may be related to the “desert
varnish” of the Batura glacial stage observed in the nearby Hunza
valley (Derbyshire et al. 1984) dating to 10.8e9.0 ka BP (Owen et al.,
2002; Lewis et al., 2002; Waragai, 2004). At Zeshtsh-o-Gutsht, lithic
artefacts are resting on a loose grey gravel of unpatinated cobbles and
devoid of any large boulders; this gravel covers the brown boulders.
The major deglaciation with the retreat of the ChitraleYarkhun
glacier generated huge terraces only visible in cross section in the
lower course of the Yarkhun River. The faults between Pamir and
Hindu Kush, two tectonic complexes characterized by intra-continental subductions between Hindu Kush and Pamirian plates
(Negredo et al., 2007), could have been active and the tectonics
Fig. 7. Stone tools from Lasht (picture A. Dambricourt Malassé).
Please cite this article in press as: Dambricourt Malassé, A., Gaillard, C., Relations between climatic changes and prehistoric human migrations
during Holocene..., Quaternary International (2010), doi:10.1016/j.quaint.2010.04.001
ARTICLE IN PRESS
A. Dambricourt Malassé, C. Gaillard / Quaternary International xxx (2010) 1e9
Fig. 9. The Yarkhun valley and connexion with the Wakhan corridor (Boroghol Pass) in
2009 (satellite view adapted from Google Earth).
could have contributed to change the regime of the Yarkhun,
especially when the water level increased in the Himalayan fringes,
as shown by a recent FrancoeNepali research program (Bettinelli
et al., 2008). Nevertheless, considering the climate factors only,
this occupation probably occurred after the major deglaciation
ending the Batura glacial phase attested by the brown varnish on
the boulders. The second terrace was formed during a climatic
phase not easy to date. For instance advances of Nanga Parbat
glacier appear to have resulted from an increase in precipitations
during winter ca. 7000e5500 BP. In Bhutan the same phenomenon
is observed ca. 6710e4680 BP (Meyer et al., 2009). This period
corresponds to the Atlantic climatic optimum. If extrapolation is
reasonable to the North-Western Hindu Kush, this implies a new
little glacial advance in the high valleys, which ended 5500 years
ago when a cooling phase started again implying drier conditions,
drop of snowfall and therefore reducing ice masses. In the Tibetan
plateau “an abrupt shift to dry climatic conditions between 5000 and
4500 BP coincided with glacial decay” (Meyer et al., 2009). So it
seems reasonable to date the grey cobble gravel terrace of the
Yarkhun River to the Atlantic phase or to a later humid or warmer
phase after 5500 years BP. The communication between the Amu
Darya and Indus upper basins, separated by the Hindu Kush Range,
again became theoretically possible since the second half of the
Holocene and is attested for the first time by the discovery of
archaeological stations in the Yarkhun Valley dating no more than
the 6th millennium (Gaillard et al., 2002; Fig. 9). The same pattern
has been recently observed in Bhutan “the sudden disappearance of
juniper and rhododendron pollen, the immediate onset of pollen input
from cereals and a clear pattern of over-grazing, trampling and peat
deterioration can be linked to human arrival in the valleys at ca.
4280 130 cal BP” (Meyer et al., 2009).
To conclude, nomadic hunters occupied not only the Eastern
Pamir as in Alichur or Murgab Valleys, but also the Hindu Kush as in
the Yarkhun valley, with possible access between the two through
passes such as Boroghol Pass. From where did these populations
come and where did they live during winter, is the actual question.
There were no reasons for them to limit their territories to the
Pamir: Epipalaeolithic and Mesolithic activities are equally attested
in the lower valleys of the southern Hindu Kush.
4. The Siwaliks and Kashmir
The same pattern as the Hissar cultural transition or acculturation is observed in the nearby Swat Valley in the rock shelter site of
Ghaligai (Fig. 1). This site offers a sequence of seven major periods
7
of which the earliest provides an industry mostly comprising
pebbles and flakes (Stacul, 1984). The overlying level yields fine
wheel-thrown pottery attesting a complete cultural change. Three
14
C dates give an age of 4970e4930 BP. “The stone artefacts are all
made from pebbles (.) there are also tools made from animal bones.
Animal remains include antler and boar tusks, and lead us to think that
in this period those who frequented the shelter practised hunting”
(Sharif and Thapar, 1999). The authors conclude that the Ghaligai
Neolithic belongs to a single complex, also including Sarai Kala near
Taxila and Burzahom in Kashmir, both located on the eastern side of
the Indus.
A few stratified Neolithic settlements are known in the Srinagar
Valley of Kashmir, especially at Burzahom (Indian Archaeology,
a Review (I.A.R.), 1961e1962, 1964e1965) and Gufkral (Sharma,
1982a,b; Fig. 1). These sites are imbedded in the top of a loess
sequence, developed in the valley since 300 ka (Singhvi et al., 1987),
but believed to be devoid of any Palaeolithic remains. Worth noting
in the Srinagar Valley is the occurrence of an Aceramic Neolithic
phase before the Neolithic proper. This cultural phase without
pottery is identified in stratigraphy at Gufkral (Sharma, 1982a,b),
Kanshipur and in several small localities. The lithic industry of this
phase includes different types of stone tools and at some places
large split cobbles that were first attributed to the Palaeolithic but
actually are Neolithic (Pant et al., 1982).
5. Discussion and conclusion
Human occupation on the Western Himalayan fringes is attested
at least 2 Ma ago, as shown by the flakes from Riwat in the Potwar
plateau (Siwaliks, Pakistan). Human occupation is known in Eastern
China around 2 Ma in the Renzi caveeRenzidong (Huang, 1999;
Dong, 2006) and future investigations will probably discover
equally old human settlements in early Indian Pleistocene formations (Dambricourt Malassé, 2009). The Himalayas and Tibet
Plateau were 2700 m below the present altitude when the first
human populations began to move between India and Western
China. Later, migrations depended on the climate because of the
elevation and formation of new natural barriers such as the glaciers
during the maximum of glacial periods. In the loess deposits of
Southern Tajikistan (pedocomplex 12e11, 900e950 ka), human
settlements were favoured by a warmer/wetter climatic stage and
the associated lithic technology was mainly based on the production of small flakes. In the Siwaliks, lithic assemblages composed of
flakes and cores on cobbles indicate human activity much before 1
Ma (Pabbi Hills). Then, around 1 Ma the rich fauna started disappearing (Nanda, 2002, 2008) while the first conglomeratic fans
accumulated down slope as a consequence of an important uplift of
the mountain range. Soon after, the severe global cooling down and
change in the glacialeinterglacial periodicity disturbed the
ecosystems.
The Acheulian technical tradition occurring in peninsular India
since 1.2 Ma (Paddayya et al., 2002) probably arrived in the subHimalayan lands during this Pleistocene ecosystemic crisis (Rendell
and Dennell, 1985), but sites are few and no archaeological record
indicates a long and local evolution. The Acheulian is totally missing
in the loess sequences of Tajikistan where native populations
continued to mainly produce flakes and cobble tools. The Acheulian
technology did not cross the mountainous massif. Therefore the
Hindu Kush seems to have been a natural barrier during Lower and
Middle Pleistocene, even in the interglacial periods. However, as in
southern Tajikistan, the cobble tool component never disappeared
from the Northwestern Indian lithic industries before the Middle
Holocene: it occurs in the Acheulian assemblages (Gaillard et al.,
2008) as well as in the probably much later Soanian assemblages
(Gaillard et al., this issue). Obviously the increasing abundance of
Please cite this article in press as: Dambricourt Malassé, A., Gaillard, C., Relations between climatic changes and prehistoric human migrations
during Holocene..., Quaternary International (2010), doi:10.1016/j.quaint.2010.04.001
ARTICLE IN PRESS
8
A. Dambricourt Malassé, C. Gaillard / Quaternary International xxx (2010) 1e9
quartzite cobbles resulting from the dismantlement of the Himalayas and spreading in vast alluvial fans, were encouraging the
production of cobbles tools and flakes.
The major deglaciation occurred during Early Holocene, as
attested in the high massifs, Himalayas, Hindu Kush, Pamir and
Tibetan Plateau. “Deglaciation in the mountain ranges of the Himalayas and the consequent meltwater discharge into the adjacent seas
was negligible immediately after the global LGM and it was therefore
of minor importance as a forcing factor for global climate change
during the end of the last glacial cycle. However, our studies show that
glaciation and particularly deglaciation is important in controlling the
deposition of thick valley fills and the landscape evolution of the high
mountain environments” (Owen et al., 2002). During the glacial
phases the Hindu Kush Range was probably an “ice barrier”
between the Amu Darya and Indus upper basins. Nevertheless,
whereas exogenous traditions reached the foothills on both sides at
different periods, Acheulian in the South, Mousterian (naming of
Ranov) in the North, none of them developed. Only the cobble tool
tradition (at least the cobble tool element) persisted beyond the
neolithisation, eventually crossing the barrier of the Hindu Kush as
Gupta (1979) suggested previously, when passes were ice-free.
The data suggest that the hunting territory in high plateaus was
a biotope exploited during summer, since the Late Pleistocene, by
Central Asian hunters and that a huge territory opened from the
second half of the Holocene, including lower valleys not only such
as Gissar and Afghani Badakhshan, but also Chitral, Swat, Indus and
maybe other regions awaiting further investigations in Himalayas
and Western China. Without those movements which allowed
interbreeding between the tribes, the genetic variability would
have declined.
But the most important event is that, after having for at least 2
million years, exposed to predators, climate changes, faunal
extinction, earthquakes, high altitude conditions, living in the
lower valleys of the highest massifs of the earth, crossing the
highest plateaus, hunting in the forest, knapping cobbles to make
their tools, this way of life disappeared suddenly about 5000e4000
years ago. The high valleys were their summer hunting territories
but in winter they probably returned in the lower valleys. The new
event which occurred around 5000 years ago was pastoralism and
the extensive exploitation of wood. Observing the ecosystem in the
south of the Yarkhun valley (Chitral), Miehe et al. (2009) raise the
question of the deforestation and ask which Neolithic factors have
generated a deregulation of the mountainous climate. Besides the
possibilities of shifting to agricultural or pastoral economies, a third
hypothesis regarding the disappearance of the last central Asian
hunters-gatherers is their extinction caused by endemism in relation with the forest retreat in the low altitude zones.
References
Agakhanyantz, O., 1996. The Pamir is rising. In: Shoumatoff, Nicholas,
Shoumatoff, Nina (Eds.), Around the Roof of the World. The University of
Michigan Press, pp. 103e109.
Amosova, A.G., Ranov, V.A., Filimonova, T.G., 1993. Les énigmes de la culture de
Hissar. In: Découverte des civilisations d’Asie Centrale. Les Dossiers d’Archéologie 185, 14e21.
Angel, J.L., 1972. A middle Palaeolithic temporal bone from Dara-i-Kur, Afghanistan.
Transactions of the American Philosophical Society 62 (4), 54e56.
Bettinelli, P., Avouac, J.P., Flouzat, M., Bollinger, L., Ramillien, G., Rajaure, S.,
Sapkota, S., 2008. Seasonal variations of seismicity and geodetic strain in the
Himalaya induced by surface hydrology. Earth and Planetary Science Letters
266, 332e344.
Bondarev, L.G., Gobedzhishivli, R.G., Solomina, O.N., 1997. Fluctuations of local
glaciers in the southern ranges of the former USSR: 18,000e8000 B.P. Quaternary International 38/39, 103e108.
Dambricourt Malassé, A., 2008. Le peuplement humain en Eurasie: la Haute Asie et
les piémonts sous-himalayens du Plio-Pléistocène à l’Holocène, origines,
évolution humaine et migrations. L’Anthropologie 112 (3), 370e403.
Dambricourt Malassé, A., 2009. Embryogeny and Human phylogeny. Chapitre 8. In:
Sankhyan, A.R. (Ed.), Asian Perspectives on Human Evolution. Serial Publications, New Delhi, pp. 103e121.
Dennell, R.W., 2004. Early hominid landscapes in Northern Pakistan. In: BAR
International Series, vol. 1265. J & E. Hedges Ltd, Oxford, 454 p.
Dennell, R.W., 2008. The taphonomic record of Upper Siwalik (Pinjor stage) landscapes in the Pabbi Hills, Northern Pakistan, with consideration regarding the
preservation of hominin remains. In: Chauhan, P., Patnaik, R., (Guest Eds.).
Multiple Approaches to South Asian Paleoanthropology: A Tribute to Gudrun
Corvinus, Quaternary International 192 (1), 62e77.
Derbyshire, E., Li, J.J., Street-Perrott, F.A., Xu, S.Y., Waters, R.S., 1984. Quaternary
glacial history of the Hunza Valley, Karakoram mountains, Pakistan. In:
Miller, K.J. (Ed.), The International Karakoram Project, vol. 2, Cambridge
University Press, Cambridge, pp. 456e495.
Dong, W., 2006. Récentes découvertes de restes d’Hominidés fossiles en Chine du
Sud. (Recent discoveries of hominid fossils in southern China). L’anthropologie
110, 241e257.
Frumkin, G., 1970. Archaeology of Soviet Central Asia. Brill, Leiden, 218 pp.
Gaillard, C., Dambricourt-Malassé, A., 2008. Les principales étapes de l’occupation
humaine en bordure de l’Himalaya occidentale. L’Anthropologie 112 (3),
404e422.
Gaillard, C., Dambricourt-Malassé, A., Magraner, J., Maîtrerobert, A., Taj, A., Voisin, J.L., Nasir, A., 2002. Discovery of recent lithic industries with archaic features in
the Hindu Kusch range (Chitral district, North Pakistan). Bulletin of the IndoPacific Prehistory Association 22, 25e33.
Gaillard, C., Singh, M., Dambricourt, A. Late Pleistocene to early Holocene industries
in the southern fringes of the Himalaya. Quaternary International, in this issue.
Gaillard, C., Singh, M., Rishi, K.K., 2008. Technological analysis of the Acheulian
assemblage from Atbarapur in the Siwalik Range (Hoshiarpur district, Punjab).
Man and Environment 33 (2), 1e14.
Gupta, S.P., 1979. Archaeology of Soviet Central Asia, and the Indian Borderlands. In:
Prehistory, vol. 1. B.R. Publishing Corporation, Delhi, 194 pp.
Huang, W., 1999. A message from early man: new finds from research on the lower
and middle Paleolithic of China. Nihon Kôkogaku (Archaeological Studies in
Japan) 8, 1e17.
Indian Archaeology - A Review 1961e1962, 1964e1965 Annual Publication on
Archaeological Reports, Archaeological Survey of India, New Delhi.
Kiyatkina, T.P., Ranov, V., 1971. Perviy anthropologiskie naxodki kamennogovea
Tadjikictan (neolithicekhe erepa yz Tutkaula). Voproc Anthropologii 37,
149e156.
Lewis, A.O., Finkel, R.C., Caffee, M.W., Gualtieri, L., 2002. Timing of multiple late
Quaternary glaciations in the Hunza Valley, Karakoram mountains, northern
Pakistan: defined by cosmogenic radionuclide dating of moraines. Geological
Society of America bulletin 114 (5), 593e604.
Meyer, M.C., Hofmann, Ch.-Ch., Gemmell, A.M.D., Haslinger, E., Häusler, H.,
Wangda, D., 2009. Quaternary Holocene glacier fluctuations and migration of
Neolithic yak pastoralists into the high valleys of northwest Bhutan. Science
Reviews 28 (13e14), 1217e1237.
Miehe, G., Miehe, S., Schlütz, F., 2009. Early human impact in the forest ecotone of
southern High Asia (Hindukush, Himalaya). Quaternary Research 71, 255e265.
Movius, H.L., 1953a. Paleolithic and Mesolithic site in Central and Soviet Asia.
Proceedings of the Philosophical American Society 97, 383e421.
Movius, H.L., 1953b. The Mousterian cave of Teshik-Tash, southeastern Uzbekistan,
central Asia. Bulletin of the American School of Prehistoric Research. 17:11e17.
Nanda, A.C., 2002. Upper Siwalik mammalian faunas of India and associated events.
Journal of Asian Earth Sciences 21, 47e58.
Nanda, A.C., 2008. Comments on the Pinjor mammalian fauna of the Siwalik group
in relation to the post-Siwalik faunas of Peninsular India and Indo-Gangetic
Plain. In: Multiple Approaches to South Asian Paleoanthropology: A Tribute to
Gudrun Corvinus, vol.1, pp. 6e13. http://www.sciencedirect.com/science/
journal/10406182192.
Negredo, A.M., Replumaz, A., Villaseñor, A., Guillot, S., 2007. Modeling the evolution
of continental subduction processes in the PamireHindu Kush region. Earth and
Planetary Science Letters 259, 212e225.
Owen, L.A., Finkel, R.C., Caffee, M.W., Gualtieri, L., 2002. Timing of multiple late
Quaternary glaciations in the Hunza Valley, Karakoram mountains, northern
Pakistan: defined by cosmogenic radionuclide dating of moraines. Geological
Society of America Bulletin 114, 593e604.
Paddayya, K., Blackwel, B.A.B., Jhaldiyal, I.R., Petraglia, M.D., Fevrier, S., Chaderton, D.
A., Blickstein, J.I.B., Skinner, A.R., 2002. Recent findings on the Acheulian of the
Hunsgi and Baichbal valleys, Karnataka, with special reference to the Isampur
excavation and its dating. Current Science 83 (5), 641e647.
Pant, R.K., Gaillard, C., Nautiyal, V., Gaur, G.S., Shali, S.L., 1982. Some new lithic and
ceramic industries from Kashmir. Man and Environment 6, 37e40.
Porter, S.C., 1985. Extent of Late-Pleistocene glaciers in Afghanistan based on interpretation of landsat imagery. In: Agrawal, D.P., Kusumgar, S., Krishnamurty, V.
(Eds.), Climate and geology of Kashmir and central Asia: the last four million years.
Current Trends in Geology, vol. 4, pp. 191e195. New Delhi.
Ranov, V.A., 1982. The Hissar neolithic culture of Soviet Central Asia. Man and
Environment 6, 63e71.
Please cite this article in press as: Dambricourt Malassé, A., Gaillard, C., Relations between climatic changes and prehistoric human migrations
during Holocene..., Quaternary International (2010), doi:10.1016/j.quaint.2010.04.001
ARTICLE IN PRESS
A. Dambricourt Malassé, C. Gaillard / Quaternary International xxx (2010) 1e9
Ranov, V.A., 1984. L’exploration archéologique du Pamir. Bulletin de l’Ecole française
d’Extrême-Orient 73 (73), 67e97.
Ranov, V.A., 1988. Oshkhona, Tadjikistan. In: Leroi-Gourhan, A. (Ed.), Dictonnaire de
la Préhistoire. Presses Universitaires de France, p. 780.
Ranov, V.A., 1993. Tout commence au Paléolithique. Découvertes des civilisations de
l’Asie Centrale, Les Dossiers de l’Archéologie 185, 4e13.
Ranov, V.A., 1995. The “Loessic Palaeolithic” in South Tajikistan, Central Asia: its
industries, chronology and correlation. Quaternary Science Reviews 14, 731e745.
Ranov, V.A., 2001. Loess-Paleosoil formation of Southern Tajikistan and the loess
palaeolithic. Praehistoria 2, 7e27.
Ranov, V.A., 2003. De l’ancien et du neuf dans l’étude du Mésolithique en Asie
moyenne. Paléorient 29 (29-1), 157e165.
Ranov, V.A., 2005. L’énigme de la diffusion des artefacts sur les sites paléolithiques
du loess au sud du Tadjikistan. In: Molines, N., Moncel, M.H., Monnier, J.L. (Eds.),
Données récentes sur les modalités de peuplement en Europe au Paléolithique
inférieur et moyen. John and Erica Hedges Ltd, Oxford, pp. 387e397.
Ranov, V.A., Davis, R.S., 1979. Toward a new outline of the Soviet Central Asian
Paleolithic. Current Anthropology 20 (2) 249e170.
Ranov, V.A., Dodonov, A.E., 2003. Small instruments of the lower Palaeolithic site
Kuldara and their geoarchaeological meaning. In: Burdukiewicz, J.M., Ronen, A.
(Eds.), Lower Palaeolithic small tools in Europe and the Levant. BAR International Series, vol. 1115. Archaeopress, Oxford, pp. 133e147.
Ranov, V.A., Dorj, D., Lü Zun, -E., 1999. Lower Palaeolithic cultures. In: Dani, A.H.,
Masson, V.M. (Eds.), History of civilizations of Central Asia, vol. 1. UNESCO/
Motilal Banarsidass Publ., pp. 45e64.
Ranov, V.A., Schäfer, J., 2000. Loessic Palaeolithic. Archaeology. Ethnology and
Anthropology of Eurasia 2 (2), 20e32.
Ranov, V.A., Sidorovo, I.F., 1979. The Pamir as Man's habitat. In: Gupta, S.P. (Ed.),
Archaeology of Soviet Central Asia and the Indian Borderlands. B.R. Publishing
Corporation, Delhi, vol 1, p. 176.
Rendell, H., Dennell, R.W., 1985. Dated lower Palaeolithic artefacts from northern
Pakistan. Current Anthropology 26 (3), 393.
Rendell, H.M., Dennell, R.W., Halim, M.A., 1989. Pleistocene and Palaeolithic
investigations in the Soan Valley, Northern Pakistan. In: B.A.R. International
Series, vol. 544, 346 pp., (Oxford).
Rhode, D., Zhang, H., Madsen, D.B., Xing, G., Brantingham, P.J., Haizhou, M., Olsen, J.
W., 2007. Epipaleolithic/early Neolithic settlements at QinghaiLake, Western
China. Journal of Archaeological Science 34 (4), 600e612.
Savvaitova, K.A., Pietr, T., 1999. Fish and fisheries in Lake Issik-Kul (Tien Shan), River
Chu and PamirLakes. In: Petr, T. (Ed.), Fish and fisheries at higher altitudes: Asia.
FAOeFisheries Technical Paper, vol. 385, p. 304.
Schäfer, J., Ranov, V.A., Sosin, P.M., 1998. The cultural evolution of man and the
chronostatigraphical background of changing environments in the loess paleosoil sequences of Obi-Mazar and Khonako (Tajikistan). Anthropologie (Brno)
36 (2), 121e135.
Seong, Y.B., Owen, L.A., Yi, C., Finkel, R.C., 2009. Quaternary glaciation of Muztag Ata
and Kongur Shan: Evidence for glacier response to rapid climate changes
9
throughout the Late Glacial and Holocene in westernmost Tibet. Geological
Society of America Bulletin 121 (3e4), 348e365.
Serebryanny, L., Solomina, O., 1996. Glaciers and climate of the mountains of the
former USSR during the Neoglacial. Mountain Research and Development 16
(2), 157e166.
Sharif, M., Thapar, B.K., 1999. Food-producing communities in Pakistan and North
India. In: Dani, A.H., Masson, V.M. (Eds.), History of civilizations of Central Asia,
vol. 1. UNESCO/Motilal Banarsidass Publ., pp. 127e152.
Sharma, A.K., 1982a. Excavations at Gufkral, 1981. Puratattva 11, 19e25.
Sharma, A.K., 1982b. Gufkral 1981: an Aceramic Neolithic site in the Kashmir valley.
Asian Perspectives 25, 25e41.
Singhvi, A.K., Bronger, A., Pant, R.K., Sauer, W., 1987. Thermoluminescence dating
and its implications for the chronostratigraphy of loess palaeosols sequences in
the Kashmir valley, India. Chemical Geology 65, 45e56.
Stacul, G., 1984. Cultural change in the Swât valley and beyond, c. 3000e1400 BC.
In: Allchin, B. (Ed.), South Asian Archaeology 1981. Cambridge University Press,
pp. 205e211.
Teilhard de Chardin, P., de Terra, H., 1936. Observations on the upper Siwalik
formation and later Pleistocene deposits in India. Proceeding of the American
Philosophical Society 76 (6), 791e822.
Terra (de), H., Teilhard de Chardin, P., Hawkes, J.C., 1934. Palaeolithic industries in
Western Punjab and Kashmir and their geological significance. Memoirs of the
ConnecticutAcademy of Science and Art 3.
Terra (de), H., Teilhard de Chardin, P., Paterson, T.T., 1936. Joint geological and
prehistoric studies in the late Cenozoic in India. Science 83, 233e236.
Velitchko, A.A., Lebedeva, I.M., 1974. An attempted of Palaeoglaciological
reconstruction for the Eastern Pamirs. Data of Glaciological Studies 23,
109e117.
Wang, Y., Wang, X., Xu, Y., Zhang, C., Li, Q., Tseng, Z.J., Takeuchi, G., Deng, T., 2008.
Stable isotopes in fossil mammals, fish and shells from Kunlun Pass Basin,
Tibetan Plateau: Paleo-climatic and Paleo-elevation implications. Earth and
Planetary Science Letters 270 (1e2, 15), 73e85.
Waragai, T., 2004. Calcrete crust formation on the lateral moraine of Batura glacier,
Northern Pakistan. 19th Himalaya-Karakoram-Tibet workshop, Niseko, Japan.
Himalayan Journal of Sciences 2 (4), 270e272.
William, M.P., Sloan, V.F., Shroder, J.F., Sharma Jr., P., Clarke, M.L., Rendell, H.M.,
2000. Asynchronous glaciation at Nanga Parbat, northwestern Himalaya
Mountains, Pakistan. Geology 28 (5), 431e434.
Yakovlev, E.P., 1976. Kratkie soobshcheniya o vstrechakh Tibetskoi saji [Brief
communications on records of the Tibetan Sandgrouse]. Na Pamire [In Pamir].
Redkie, ichezayushchie i maliozuchennye ptitsy SSSR [Rare, endangered and
little-studied birds of the USSR]. (Trudy Okskogo gos. Zapovednika [Proc. of Oka
[Okskiy] State Nature Reserve]), 13, 188e189.
Zech, R., Abramowski, U., Glaser, B., Sosin, P., Kubik, P., Wolfgang, Z., 2005.
Late Quaternary glacial and climate history of the Pamir Mountains
derived from cosmogenic 10Be exposure ages. Quaternary Research 64 (2),
212e220.
Please cite this article in press as: Dambricourt Malassé, A., Gaillard, C., Relations between climatic changes and prehistoric human migrations
during Holocene..., Quaternary International (2010), doi:10.1016/j.quaint.2010.04.001