Paper presented in SEAEAA conference in Sigtuna, Sweden 2002 and published in "SEA Archaeology" W.Solheim festschrift book, Manila, 2004

 

 Hoabinhian food strategy

in Vietnam

 

Nguyen Viet

 

Since 1982,  influenced by Chester Gorman’s (Professor Solheim’s first graduate student) excavation in Spirit Cave (Thailand) I began micro studies of archaeological remains aimed at studying Hoabinhian food strategies[1]. Beside traditional digging methods, all sediment in Xom Trai and Con Moong caves were carefully recovered through seaving and sorting using a microscope. These methods brought us many thousands of different meal remains of animals as well as plants, which allowed us to carry out a quantitative research to determine Hoabinhian food strategies in Vietnam.

 

Some new Interpretations on Hoabinhian concerning to Food Exploitation

 

Chronology and Evolution

Radiocarbon dating for classic typical Hoabinhian

In some of my recent publications (Viet 1989, 2000, 2001a-b) I have shown believable evidence of early  Hoabinhian sites ca. 20,000 BP.  This is mainly based on the results of my re-excavating in 1986-1987 of the most important Hoabinhian sites in Vietnam i.e., Xom Trai, Lang Vanh, Hang Muoi, Con Moong and Sung Sam. Nearly a hundred  samples for radiocarbon dating were collected, and half of these were processed in the radiocarbon dating laboratory in former East Berlin. Radiocarbon samples in many different Hoabinhian sites in Vietnam were re-examined and re-measured. Working with the laboratory in Berlin, we dated and identified a classic and typical Hoabinhian system existed about ten thousand years older than the traditionally accepted time depth. The valley Muong Vang in province Hoa Binh, where three typical Hoabinhian sites (Xom Trai, Lang Vanh, Hang Muoi) are located, is recognized as the homeland of Hoabinhian in Vietnam. The tool-kits of Hoabinhan here are very typical for the so called classic Hoabinhian, which were recovered and determined by Madalene.Colani more than 70 years ago.

At Xom Trai cave, we collected a total of 32 organic samples, of which 15 are charred fragments of Juglans spp. nut covers. The rest are shells of stream snail Antimelania spp. Almost all radiocarbon dates confirm the first Hoabinhian occupation at ca.18,500 BP. In the same valley, other Hoabinhian caves ( Xom Tre and Lang Vanh), those located around 2 km to 4 km southwards from Xom Trai existed contemporarily with Xom Trai. Their ¹⁴C dates for the upper stratigraphic layers were at 16 to 17,000 BP (Viet 1989, 2001). The radiocarbon dates for Hoabinhian sites in Muong Vang valley supports a very early occupation of Hoabinhian inhabitants in this area. They represent the early typical Hoabinhian.  

 

Radiocarbon dating for untypical Hoabinhian

Almost all un-typical Hoabinhian sites in Vietnam have radiocarbon date ranges of 13,000 to 8,000 BP. The re-excavation at Con Moong in 1986-1987 made by Nguyen Viet brought in light many reliable radiocarbon dates. Those ¹⁴C ages confirmed the existence of  a deeper Hoabinhian chronological sequence.

Except for some sites, for example at Hang Muoi, most cases sites contain a poor and untypical Hoabinhian tool-kit. They represented a Hoabinhian developing in more difficult conditions at later phases, when the Hoabinhian under the presure of  increasing population enlarged their food exploiting areas, further from their homelands. Because of this, two Hoabinhian subsistence systems developed. One was richer and earlier as we could see, for example, in the homeland Muong Vang valley. It existed in a moist and cooler environment. Another one was poorer, particulary in tool-kit development,  and lived in a warmer, even hot condition. The relationship between the two Hoabinhian systems will be discussed again below.

 

Hoabinhian Evolution

The earliest Hoabinhian evidence in Vietnam might be found in Tham Khuong cave (prov. Lai Chau), where we have ¹⁴C and ESR dates of ca.32,000 BP. However we have relatively few materials on this group. We believe that such earlier Hoabinhian sites could be found also in the Hoabinh province. Till now, in this province, the earliest Hoabinhian occupations were discovered only in the Muong Vang valley with. three Hoabinhian typical sites (Xom Trai, Lang Vanh, Xom Tre). Their ¹⁴C dates have a range of 18,500 to 16,000 BP. The Muong Vang valley was recognized as the best Hoabinhian locality, where with a padadisiac condition man could get easyly naturaly formed pebbles suitable for tool making, and could collect abundant food at all seasons of the year. They belong to the richer Hoabinhian and their living area became the homeland of Hoabinhian culture (Viet 2001a).

The Hoabinhian distribution shows that later Hoabinhian sites representing the poor Hoabinhians (Con Moong, Hang Dang, Sung Sam, Dong Cang, Mai Da Dieu and Hang Chua), are located outside the center of the Hoabinhian stone source. There is increasing evidence of Hoabinhian sites at the transition period from the Pleistocene to the Holocene.

The paradisiac condition in the “gold valley” at Muong Vang may have caused populations to increase and this increase resulted to the need to enlarge the food exploitation areas. The case of Hang Muoi is a good example for later Hoabinhian increasing in the homeland region. The enlarging food exploitation into the more difficult areas created the existences of poor Hoabinhian sites such as Con Moong, Sung Sam, Dong Cang. They belonged to the enlarging Hoabinhian later system. This enlarging process began at ca. 13,000 BP.

Of the Hoabinhian enlarging movements there were some Hoabinhian groups, whose food strategy was mainly gathering mollusc in swamp and lake condition. Their meal remains built the open air shell middens nearby traditional Hoabinhian valley regions. That is the case of Da But site. The first Hoabinhian occupation at Da But happened perhaps at 8000 to 9000 years BP. The Hoabinhian tried here a new way of living in a new environmental condition, that corresponded to warming and raining Holocene climate.However, the Hoabinhian in Da But did not live in caves or rockshelter, but in cottages or simple houses. Wood or bamboo working became more intensive than before. Edge polishing tools and ceramic were used more popularly in Da But way of life. The nomadic life in the mountains resulted in the absence of  burial fields. In Da But, however, the dead were buried usually nearby cottages or houses. This is evidence for a settled way of life. Since the Da But period,  the Hoabinhian culture had an open door to transform towards a Neolithic way of life (Viet 1984, 2002).

    

Palaeoenvironment

Sea level changing during Hoabinhian

         The Hoabinhian existed for around 14,000 years (from ca. 20,000 to 6,000 BP). This time corresponds to the changes of the sea level and the global climate from the Last Glacial Maximum (LGM) to the present climate and sea level.

                As the Hoabinhian occupied the Muong Vang valley, the sea level was at ca. –150m to -120 m. The temporal shoreline was ca 200 m to 400 m far from the Hoabinhian homeland (Chen et al 1985; Ty et al 1985; Viet 2001a). 

            According to Nakata & Lambeck (1988) the sea level arose relatively slow during 18,000 - 20,000 BP, but became rapid during 12,000 to 6000 BP. This is supported by new data of Fairbanks curve. In the Malacca Strait, Gegh et al (1979) recognized the sea level at 12,000 BP corresponding to - 80m and at 6,000 arose to the highest level (+ 6 m). The same situation had been recorded also in Japan, New Zealand and Australia (Nakata & Lambeck 1988) as well as in East China Sea (Chen et al 1985). The 14C- dates of some inundated shorelines in Hong Kong show the sea level got to present level at about 6,000 BP. (Meacham 1978). The boring hole ND1 in North Vietnam shows evidence that the sea was at 12,000 BP to –45 m (Shigeko 2001). Radiocarbon date at Ha Lung cave (North Viet Nam) stated that before 6000 BP fresh water snails were collected in this region, where at present is - 10 - 20 m under water (Viet 1989). More recent research confirmed the middle Holocene highest sea level (+4m to +5m) in Vietnam occured during 6,000 – 5,000 BP (Viet 2002; Thuan 2002; Tiep 1998).

            The changes of sea level in the last 20,000 years led the markable reduction of living surface for the Hoabinhian (Bellwood 1985). However it happened in Vietnam only in some traditional Hoabinhian food exploiting areas, for example, in Ha Long Bay. The reduction of the Hoabinhian living surface occured only in the middle Holocene. Until now we do not have evidence for a migration wave, due to raising sea levels, from the “lower Hoabinhian“ into traditional Hoabinhian moutainous areas before 6000 BP.

 

 Mean Temperature during Hoabinhian

            The temperature at 20,000 - 15,000 BP was regarded as lowest during last 20,000 years. After CLIMAP (1981) the mean sea surface temperature (SST) in South China sea was ca 2 ^oC lower at present. Basing on pollen data in Leizhou peninsula (Southern China), Zheng et al. (1999) estimated, that the temperature at LGM in montane forest altitude 800m was ca. 4 ^o – 6 ^oC lower than today. The lapse rate here was 0,61 ^oC per 100 m. After Zhang (1984), we could estimate that the mean temperature in North Vietnam at LGM was 0 – 5^oC in January and 25 ^oC in July. Prell (1982) presented an isomap SST for the West Sunda sea, in which north Viet Nam and South China Sea belong to the region having SST of ca 24 ^oC in August (ca 4 ^oC lower than present). The pollen record from Taiwan suggested the mean temperature at 24,000 - 14,000 BP was 2 to 6 ^oC lower than at present(Tsukada 1966). In Japan, the pollen data confirmed that the mean temperature reduced 8 – 9 ^oC in northern Hokkaido, 7,7 - 8,7 ^oC in Northern most Honshu, 7,2 - 8,4 ^oC in Central mountains, 6,5 ^oC in Chugoku district and 5 – 6 ^oC in Kyushu (Tsukuda 1983). In the Yangtze Delta, pollen data showed the mean temperature was below ca 5 ^oC at LGM (Lin et al 1989).

            For the case of southern Sunda, Verstappen (1975) proposed it was 3 – 5 ^oC cooler while Peterson (1968) estimated the mean temperature at Niah cave region in Wurm II glacial 5 – 7 ^oC lower than at today, and rainfall reduced ca 30%. The pollen data taken by Stuijts (1984) in west Java (Indonesia) yielded the evidences for the decrease of mean temperature ca 5 – 6 ^oC lower than present during 16,800 - 10,870 BP.          

            According to Frenzel (1968) in the Wurm Glacial almost the surface of the northern Hoabinhian region was covered by broad leaf evergreen forest. After Aigner, this forest in China corresponds to a climate, which’s temperature was about 5 ^oC lower than today. In Vietnam, the pollen data recorded from Nguom rock shelter at ca 23,000 BP and from Xom Trai cave (18,500 - 16,600 BP) included frequently some species such as juglans spp., querous spp., salix spp., carpinus spp., lithocarpus spp., and minica spp.. These plant types  correspond to a cool forest vegetation (or montane forest II) which exist presently in altitudes of 1800 m to 2400m in Taiwan and central Sumatra (see Tsukada 1966 and Morley 1982). The presence of these species in the lower altitudes in Vietnam suggest a reducing of mean temperature ca 5 – 8 ^oC lower than present [2]. The palaeontological records supports this cooling. The fossils of Ailuropoda were found in Vietnam at Phung Quyen rock shelter at ca. 18,000 BP and Pongo bones were recovered at Nguom rock shelter at about 23,000 BP.

            The macro plant remains collected from some Hoabinhian caves show a relative clear change of the flora exploited by Hoabinhian inhabitants during 18,500 to 8,000 BP.

            In Xom Trai cave, 90% of total collected plant remains belong to charred shells of a probably walnut (juglans) or gam (gnetum sp), which can be seen today usualy in the high mountain forest. These charred fragments were found also in Lang Vanh, Mai Da Dieu and Dong Cang. They were used as ideal materials for radiocarbon dating. In Xom Trai, Lang Vanh and Mai Da Dieu radiocarbon dates measured from these charcoal show ages in range of 20,000 to 16,000 BP. In the later Hoabinhian sites, such plant remains were total absent. 

            In Taiwan, Tsukada (1966) mentions a particular event in the pollen history that : "near the top of zone M [3], Juglans cathayensis Dode suddenly appears, but decreases before the M/R zonal boundary (ca. 12 - 10,000 BP - NV)." and "Juglans cathayensis, associated with Cunninghamia [Konishii Hayata], may give a reliable estimate of the mean temperature (lower by 2 ^oC – 6 ^oC) between ca. 25,000 and ca. 14,000 BP" (Tsukada 1966:547). Juglans in Jihyuetans pollen history represented cool climate at LGM. Presently this genus grows common in the cool-temperature forest at altitudes from 1800 m to 2400 m in Taiwan.

            Since 15,000 years ago sea temperature began to increase corresponding to the rise of sea levels. From data of core V35-5 in South China Sea (in 7 ^o12'N – 112 ^o5'E) and core SONNEN 50-37KL, in 18 ^o55'N – 115 ^o45'E, which located not far from east margin of Sunda shelf (in 7 ^o12'N – 112 ^o5'E) Broecker discovered the abruft termination of the last glacial, in which distinct phases of sedimentation rate and productivity of foraminifera organisms  during 18,000 to 6000 BP were recognized. Phase I (from 18,000 - 13,000 BP) the sedimentation rate was rapid whilst the productivity of foraminifera organisms was slow. In the phase II (from 13,000 to 6000 BP) it is reverse, (Broecker et al 1990).

            All pollen data, whether from northern part (Taiwan, by Tsukada 1966) or from southern part (east Malaysia and Sumatra, by Morley 1982, Maloney 1986) or further southeast ward from New Guinea by Flenley (1979) demonstrated a climatic rapid amelioration during 14,000 to 12,000 BP. Morley, Flenley and Tsukada recorded apeak warming higher ca 2 ^oC than at presentat 10 to 8000 BP in Sumatra, at 8000 BP in New Guinea and New Zealand, and at about 6000 BP in Taiwan.

            The climate changed also in Vietnam during 12,500 to 8,000 BP (Nguyen Viet et al., 1988). At Con Moong, we recovered from cultural sediment 2 a different assemblage of plant remains.

            1- The earlier assemblage: From the lower layers, where radiocarbon dated from 12,500 till 10,500 BP, humans collected nuts, attested by the recovery of thousands of fragments of nut shells determined as probable Aleurites spp. or Castanopsis spp..Such trees grow today mainly in mountain regions of north Viet Nam.

             Castanopsis spp. according Yen (1977) grows presently in Thailand high mountain forest, whilst in a higher latitude (for example) in Taiwan, this genus is found in the warm temperature forest 500 to 1800 m in altitude from sea level (Tsukada 1966). In New Guinea  this genus is seen frequently in lower mountain forests  — montane forest I at 1500 - 1800 m altitude (Fleney 1985; see Morley 1982).

            This kind of nut fragments had been associated frequently with Celtis spp. stone cores in dried form. Celtis spp. cores existed relatively common in the Hoabinhian sites. On a block of cemented sediment brought by M.Colani from My Te cave[4] we have recognized 8 cores of Celtis, which were stucked on a broad mammalian bone. Celtis is recorded also at Nguom, Xom Trai, Con Moong. According  to Yen (1977), Celtis dried cores were found in Spirit cave, Thailand. Celtis is one of the characteristic species of the sub-montane forest in Java — common at altitute ca. 1000 - 1400 m (Morley 1982). In the some higher altitude, for example in Taiwan, Celtis grows common in subtropical rain forest below ca. 500 m altitute (Tsukada 1966). Celtis pollen was recorded frequently at Zengpiyan cave (northern Kwangzi) in the layers 9,000 - 7,000 BP (Wang, 1989).

            2- The later assemblage : In the upper level of Con Moong cave (from layers A4b to A2) occurred mostly fruit stones of Canarium spp. while Juglans spp. was absent. The radiocarbon dates measured associated with such plant remains showed a time depth of 10,500 BP to 8500 BP. Cannarium may be characterize for beginning of Holocene in Vietnam as well as in Southeast Asia. Some of this kind of plantd were determined by Yen as Canarium spp. or more confidently by Thin et al. (1987) as Canarium tonkinensis Engl.. Canarium was recorded in Dong Cang, Hang Muoi, Hang Doi and upper level of Mai Da Dieu. In the Southeast Asian Hoabinhian context charred fragments of Canarium were documented in Spirit cave, Tham Pa Chan, Banyan Valley cave (Yen 1977), and Ongbah cave (Soerensen 1988). It's notable that almost all Canarium remains found in Southeast Asia were younger than 11,000 BP [5] . This species is very common till now in Mainland Southeast Asia.

            The seemly sudden explosion of Canarium in the vegetable menu of Con Moong dwellers in the time span of 9,500 - 8,500 BP marked a warmering phasein Vietnam, which is supported by evidence in Taiwan (Tsukada 1966), in Sumatra (Morley 1982) and in  New Guinea (Fleney 1979, 1985). The COHMAP explained it that as result of decreasing distance of earth – sun, especially after 9000 BP. (COHMAP 1988: 1044, 1048).

            In Taiwan Tsukada (1966) estimated for this warmer phase ca. 2 – 3 ^oC higher than at present at the peak of 6000 BP. In Sumatra, at 10,000 - 8600 BP the mean temperature increased possibly by 2 ^oC (Morley, 1982).

 

Precipitation during Hoabinhian

 

            Ha van Tan (1985) suggested that the rockfall layer at Nguom rockshelter reflected a dry and cold climate before 23,000 BP. Hoang Du and Nguyen Duc Tung (1977) defined the pollen data in the lowest layer in Con Moong excavation 1975-1976 to represent a  dry climate. This layer have the 14C- age of 12,000 - 11,000 years ago. Two above opinions represent the generally accepted view by almost all quaternary researchers in Viet Nam. These views accepts that the late Pleistocene climate was generally dry and cool. This general conception was supported by climate models proposed by COHMAP (1988) for the case of the LGM climate. However, through our research, there were many evidences for the condition of a humid climate with strong rainfall during LGM in Vietnam.

           The relative frequent presences of mollusc remains in the cultural deposits during the time span from 23,000 to later18,600 BP in Nguom rock shelter (Long et al. 1986) as well as the consequent increase of mollusc remains during the Hoabinhian period tells of a long dry period covering in total the late Pleistocene. The stratigraphy of some sites in Southern China existing in the late upper Pleistocene (Q3) reflected a identical palaeontological event that some representative species of the Ailuropoda - Stegodon Fauna became rare. In contrast, the shells of mollusc increased visible.

            Based on mollusc data from Nguom and Tham Khuong, we can estimate that at about 30,000 BP the climate in the region, for example in Tham Khuong (1,200 m altitude, 21 ^o35N) was not too dry and not too cold. The presence of a lot of shells of the land snail Cyclophorus in the lowest layer of Tham Khuong cave supports this picture. From 30,000 BP (the upper part of Tham Khuong) to 23,000 BP. However, the climate became perhaps very dry and cool [6]

Researching the surfaces of shells of Melania spp. collected in Xom Trai, we recognized that they were destroyed always by being smashed using stream stones . It is evidence of the role of floods caused by strong rainfall. The fast running flood water caused the pushing of snail shells against stones, smashing them in the process. This phenomenon suggest strong rainfall periods in Muong Vang valley during 18,500 to16,000 BP. It is in contrary to the traditional theory, which estimates a dry period during the Last Glacial Maximum (LGM) in Vietnam.

            In the climatic amelioration rainfall increased in almost all regions of the World. COHMAP (1988) pointed out that rainfall was 20 to 30% more compared to present levels, particulary during 9000 - 6000 BP. The lake level data from China supported this opinion. In the northern Hoabinhian region, the pollen data from Zengpiyan exhibited a mild-humid to warm-humid climate during 9000 - 7000 BP.

            In Vietnam, we observed at Con Moong cave that at different archaeological layers, the nature of food remains differed. This observation also demonstrate two distinct moist phases corresponding to the increase of land snails Cyclophorus spp. and Ranguna spp.[7]at 12,000 - 10,000 BP (peak at about 10,500 BP) and 9500 - 9000 BP (peak at about 9,300 BP) (Viet et al 1988). These phases also correspond to the abundance of charred plant food remains Castanopsis spp. and Canarium spp..

        

Hoabinhian Food Remains

 

     With the excavations in the Mea Hong Son area (North west Thailand) lead by Chester Gorman in the 1960s our knowledge of the Hoabinhian menu became more detail. Gorman was also the first researcher to address this question. Before Gorman there was a common notion that prehistoric organic remains hardly survive under tropical conditions. But this is not correct anymore in the case of Hoabinhian sites, where human habitat were mainly in caves or under rock shelters. The plant remains as well as fossils of small animals, even of insects eggs and exoskeletons, excavated at Xom Trai, Con Moong contributed  to the change of thinking.

            Gorman (1970, 1971, 1977) and then Yen (1977) had recovered firstly believable data on Hoabinhian-like culture plant remains collected by them at the Thailand sites of Spirit cave, Tham Pachan and Banyan Valley cave. The excavations at Khao Talu , Men, Petch Kuha and Heap caves (West Thailand) during 1977-1979 supplied more plant remains which were botanical investigated by Pynamarn (1989). In the excavations at Ongbah cave (West Thailand) Sorensen (1988) recorded some Canarium fragments. In Viet Nam, Hoabinhian plant remains were found in many sites: Xom Trai, Con Moong, Dong Cang, Mai Da Dieu, Hang Doi, My Te, Sung Sam, Nguom, Lang Vanh, Hang Muoi. However systematic recovery were only carried out at the excavations of Xom Trai and Con Moong sites(Viet  2000a).

            The list of Hoabinhian animal bone remains was presented by Gorman (1971) consisting 32 species from 26 sites. Pookajorn (1988) presented a comparative list of animals found at Ban Kao Hoabinhian caves and Mea Hong Son Hoabinhian caves. In Viet Nam, Long (1984) presented a complete picture of animal species collected from almost all Hoabinhian excavations in Viet Nam. From the lists mentioned above we can recognize an abundance of animal and plant materials from Hoabinhian sites, but it is still not substantial enough for a significant palaeoeconomic study. Some excavations such as in Xom Trai in 1986, Con Moong in 1987, Dong Can in 1987 and Mai Da Dieu in 1989, where animal bone recovery were best documented, are not yet completely published. The Hoabinhian economic picture stays, therefore, only at the very general as sketched over 30 years ago by Gorman (1971, 1977), Dunn (1970), Glover (1977) and Higham (1977).  More recently Surin Pookajorn (1988) and Albrecht (2000) tried a more detailed study of this topic through an ethnoarchaeological approach. But the archaeological, specially the botanical and animal remains, were not  documented and analysed palaeoeconomically.

         Basing on the data documented and analysed by us and by our collaborators during last 20 years on Hoabinhian food,  I would discuss below some of our insights.

 

General Context of Hoabinhian Food Collecting in Vietnam

 

Hoabinhian living in Vietnam has two main characteristics: Mountainous valley food exploitation and limestone cave or rockshelter housing. There are four cases that have been carefully researched: Con Moong, Xom Trai –Lang Vanh, Dong Can and Sung Sam. At Con Moong, Sung Sam, the Hoabinhian people occupied high altitude caves ( about 30-50m over valley niveau). The valleys here were small and without stream. Humans must have had difficulty getting water during the dry season. Meal remains from Con Moong cultural sediments showed only products occurring during warm or rainy season ( Summer ) with advantage of land snail Cyclophorus spp., of rock crabs Ranguna spp. and charred stones of Canarium fruits. The limestone mountain here occurred more than 60% per gatherer’s food exploiting catchement ( 5 km radius from the cave ). The same situation is seen also in Sung Sam cave. Against it, man could get from Hoabinhian sediment in Dong Can cave mainly products of the Winter season. Here it occurred few limestone mountain (resource of Cyclophorus), few earthern ferlarit humus hills (resource of castanopis or canarium fruits) but had more water sources like big streams even in the dry season. Bones of big stream fishes and of tortoises dominated the food remains assemblage; with 90 % of measured fish vertebra evidenced they were killed during Winter. Very few shells of snailwere found here.

In Xom Trai cave and Lang Vanh rock-shelter of “ Adam Garden” Muong Vang valley remains of food from all seasons were found. The catch basin occurred here with only 0.03% of limestone, but 5% water surface, 40% valley surface and rest about 55% ferlarit humus earthern hills. The main food of Xom Trai and Lang Vanh inhabitants were represented by stream snails Melania spp.. Ca. 47.000 snail shells and 5 kg mammal bone remains corresponding to 78.5 kg edible meal from snail and 35 kg from mammal could be counted from each cubic meter [8]. That means, in the very good natural condition as Muong Vang valley, the Hoabinhian people need not move seasonally. The existence of three contemporary Hoabinhian sites (Xom Trai, Lang Vanh, Xom Tre) in the same hunting and gathering area could mean that they were rest places of the same Hoabinhian group. The moving cycle period was perhaps three or five years, when the natural food resource condition became bad as a result of over exploitation of resources. Depending on the food resources, the Hoabinhian hunters and gatherers were moving mainly seasonally. They stayed everywhere only for a short time. In very few cases the Hoabinhian could settle in one place for the entire year.

The general food resources of Hoabinhian were gathered from the following environmental conditions:

-                           Limestone rock-mountains (delivering land snail Cyclophorus and some small mammals),

-                           Different mountain water sources such as streams, small rivers, valley swamps, and lakes (providing snail melania, angulyagra, bivalve corbicula and fish),

-                           Valley earthen surfaces and felarit humus hill surfaces (delivering nuts and fruits, roots, fungis, vegetables, wild cereals and many wild mammals).

The traditional Hoabinhian food strategy oriented on collecting snails and other plant products. They hunted big mammals perhaps only occasionally.

 

The Hoabinhian strategy of food exploitation on the basic of a common broad spectrum food exploiting (hunting, fishing and collecting)[9].

 

         From the quantitative data of total analysable food remains excavated at Xom Trai cave (1982, 1986) and at Con Moong cave (1986) the determined orientation is not subsistence based on hunted preys, but rather on collected food. A quantitative research on food sources yielded from animal world at Xom Trai exhibited the notable domination of the Antimelania snails. Over 60% of the edible meat weight exploited from animals belong to stream snails (Viet 1990b). This number has a great significance, because from it we have an idea of how much time inhabitants of  Xom Trai spent daily for collecting and for preparing of snails. Our experimental study shows that with a relatively abundant source, a 30 years old woman had to spend nearly two hours to get ca. 5 kg stream snails (corresponding to ca. 2,000 calories) and ca. two hours for preparation and consuming these snails.

            The orientation to non-vertebrate animal resources, especially snails (gastropods)[10] is an important feature of the food exploitation of the Hoabinhian. This orientation differed in each distinct ecological niche or environmental condition. For example, in some sites, the shells of stream snails dominated (cases of Xom Trai, Hang Muoi, Lang Vanh) but in another one the land-snail shells dominated (cases of Phung Quyen, Sung Sam, Con Moong). The differences of aquatic resource and limestone rock surface determined the percentage of mollusc sources in the food composition in each Hoabinhian site.

            A food orientation can be found also in the plant food exploitation. Over 90% of the plant remains excavated at Xom Trai (1982) belong to Juglans nut. The same percentage of the plant remains collected at Con Moong (1986) belonged to Canarium (upper layers) and to Castanopsis (lower layers).

            The statistic calculations of the animal bone fragments showed the dominance of turtle in the hunting of Xom Trai cave dwellers : 26% of total bone fragments belong to this species. The same situation could be seen in Con Moong and Dong Cang caves.

            The quantitative research shows the local differences of Hoabinhian food collecting. For example, in Dong Can, the Hoabinhian people consumed mainly fish, tortoise and Canarium nuts. But people used in Con Moong mainly land snail Cyclophorus, stream snail Melania, rock-crabs Ranguna and nut Castanopsis, fruits Celtis in lower layers and fruit Canarium in upper layers.

            We may infer that snails, fruits or nuts were the source of daily nutrition needs of Hoabinhian inhabitants. This orientation was not in contrast with the broader spectrum of hunting and collecting strategy, but they do characterize specific  Hoabinhian food strategy.

 

Seasonality

         Chester Gorman had suggested the possibility of seasonal occupation in Northwest Thailand Hoabinhian (Gorman 1971:313).  Researching the case of Con Moong arises a question, whether the cave was occupied during winters, when presently at this season almost all natural streams in around a 2 km radius are nearly total dry. The major food remains demonstrated that they were collected mainly in summer, for example Canarium (July-October), land snail Cyclophorus (April-October), rock crabs Ranguna (April-October). The measures of growth lines on some shell samples in the layers B3a and B4a (ca. 12,000 BP), show that they were collected directly after the "hungry period" — about March or April (Viet 1990). In contrast, researching growth lines on fish bones — a main food of Dong Cang inhabitants  — evidenced the occupations happened only in Winter during No vember to February (Viet et al. 2001).

            In Xan Trai cave, however, the food remain data did not suggest seasonality. The food sources in this region permited the inhabitants to stay for long periods of time.  Pookajorn’s (1988) enthographic study of the current population of Phi Tong Luang living in the area describes a migration pattern. This gives us an insight on the possibility that the archaeological sites located near each other  (for example, case of Xom Trai, Lang Vanh and Xom Tre) belong to only one Hoabinhian group. This group migrated not seasonally but moved in the area at  three or five years intervals depending on the decreasing  availability of  food resources due to human exploitation.

 

The Changes in  Hoabinhian Food Strategy in the Da But Culture

The Da But culture existed in northern Vietnam during the Middle Holocene Sea transgression. radiocarbon dates of this culture suggested it’s beginning at 5^th-6^th millennium BCE and ended at about the 3^rd millennium BCE. As presented above, the Hoabinhian food tradition in the site was oriented towards the exploitation of  shellfish from mountainous environs,  and some nuts and fruits. There was a shift in the Dabutian diet with the exploitation of a new shellfish : bivalve Corbicula found in swamp and lake environments.  In some Hoabinhian caves, especially at the east border of the limestone rock massif where Thanh Hoa, Hoa Binh, Ninh Binh and Da Phuc site are located, we could see the increasing percentage of valley swamp shellfish (bivalve Corbicula, snail Angulyagra).

 During the period 10,000 to 8000 BP,  rainfall increased in the continent. Under influences of sea trangression many coastal deltas were formed.  This lead to the creation of many swamp and lakes, conducive to the growth of  mollusk populations. In the early Holocene, almost  all swarmp and lakes were fresh water near Dabut. Compared to Hoabinhian traditional catchments, the new forming coast deltas had more aquatic surfaces . The populations of swamp shellfish developed quicker than land and stream snails. This allowed the Dabutians to settle longer in a place. They lived mainly throughout  the year by collecting bivalvia Corbicular. However we noticed in the upper stratigraphic layers of Dabut  shell middens, marine shellfish remains increased, which may be seen as evidence for the nearing of the sea to the sites studied.

In another case of Dabutian site at Go Trung, the people lived mainly on fishing. At this site thousands of  fragments of marine fish bones and many stone net weights were excavated. In another site Con Co Ngua, the Dabutian collected aside from  Corbicular,  many other marine shellfish.  This may be seen as a cultural adaptation to the higher sea level condition.

 The change in food strategy lead to a change in the way of living of people in the area. The Dabutian needed to use more edge polishing axes to work with wood to build  open -air housing ( house or shelter building), more potteries for cooking possibly Corbicular, and the dead were interned in a burial field, buried in a squat/sitting position. The first evidence for domestication of dog, pig and water buffalo were suggested from bones excavated in Dabut sediments. However, the Dabutian lived mainly by a non-producing food subsistence strategy. They were basically gatherers, collectors and hunters. The introduction of Hoabinhian in the area  caused the change in food subsistence strategy. The Hoabinhians  became the most developed group in early and middle Holocene in Northern Middle Vietnam. In this culture the characteristics of a  Neolithic society are first established.

 

---