The First Seal Hunter Families on Gotland On the Mesolithic Occupation in the Stora Förvar Cave

The article presents some results of a joint interdisciplinary research project, The Stora Förvar Cave and Gotlands peopling, faunal history and subsistence economyldiet development from the Boreal to the Subatlantic, initiated by Christian Lindqvist in 1991. Its objectives include investigations of a number of crucial issues in a long-term perspective, such as the initial settlement, the early faunal history, the early subsistence economy and diet, but also the character of the Mesolithic-Neolithic shift on Gotland, by means of human and zooosteological, carbon isotope and ancient DNA analyses. The article presents and discusses artefact, osteological, and "C and "C data and interpretations concerning the duration and character of the Mesolithic occupation —temporary kill/butchering site, seasonal hunting station, semi-sedentary base camp or burial cave —as well as osteobiographical data on the identified human individuals and their burial customs.


INTRODUCTION
The Stora Förvar Cave is situated 21.3 m. a.s.l. (at the cave mouth) on the northern part of the island of Stora Karlsö, ca. 8 km to the west of Gotland in the Central Baltic Sea (figs. 1-2). The first hunters, according to AMSdates on bone collagen, arrived on the island of Stora Karlsö about 7,200 cal. BC at the latest. It is probable that they came along the south-east Swedish coast via the Baltic island of Öland and across the Ancylus Lake to Stora Karlsö, where the distance was shortest, although there are of course other possibilities. The distance from the Swedish mainland to Gotland is a good 80 km. The distance from the east Baltic coast is ca. 140-150 km and from the south Baltic coast ca. 225-230 km. Since the early Mesolithic material culture on Gotlandperhaps due to the importance of marine huntingespecially concerning the flint and bone/antler artefacts, is fairly unsophisticated compared to its mainland counterparts, it is not possible to connect it with any particular source area on the mainland, where the material culture was adapted to terrestrial large game hunting, for example of aurochs, elk, red deer, roe deer and wild boar.
The Stora Förvar Cave was excavated by Lars Kolmodin and Hjalmar Stolpe in 1888-93. In spite of its importance (the very wellpreserved faunal and human remains, the long layer sequence, etc. ) the material has to a large extent not yet been analysed and published. During the last seven years, one of the authors (CL) has analysed the faunal and human remains from one of the excavation areas (parcelle G, with a 3.3 m deep deposit excavated in 30 cm spits, see figs. 3-4, tab. I), together with faunal remains from ten other Mesolithic dwelling sites as well as some Neolithic materials from main Gotland.
The cultural layer in the inner part of the cave seems largely to have been composed of ashes, probably due to the continuous existence ofa large hearth, which is clearly visible in a photo taken over a century ago by one of the excavators, Hjalmar Stolpe ( fig. 4).
Although fire-damaged bones are generally uncommon finds, fairly numerous burnt bones are found in parcelle G, level 8 (G.8), that is, just below the Neolithic cultural layer beginning in G.7. Among these burnt bones are also found charcoal and some charred hazelnut shells. A hazelnut shell from G.8 has been AMS-dated to 7795 +105 BP (Ua-2937), 6,696 -6,465 cal. BC (I a). Since there are cut-marks on several ofthe seal bones and since numerous flints were found in the ashy cultural layer, the skins, meat and fat were probably processed beside a hearth in the cave. This activity was most intensive during a few hundred years, ca. 7,300-7,000 cal. BC, and interrupted less than a millennium later, about 6,200 cal. BC, perhaps due to the flooding of shore terrasses up to 18-20.5 m. a.s.l. by the first Litorina transgressions. The same thing seems to have occurred at Visborgs Kungsladugård in Visby country parish (Lindqvist 1997c:102,fig. 6). On this occasion three known shore-bound dwelling sites on main Gotland, Strå kalkbrott in  Other caves in the klint and the rai sed beaches ofthe Ancylus Lake and the Litorina Sea are mari ed, as is a fi eshwater fen (''Myren ') and three other freshwater wells. The stratigraphy does not seem to be disturbed by burials. In the foreground, finds in the base level ofthe Farea seem to be marked with sticks. Large rooffall rocks made the excavation difficult. Photo: Hjalmar Stolpe, l89/-92. Hildebrandt & Jones 1992:388-389;Jones & Hildebrandt 1995:78-90;Lyman 1995:46-47;Lindqvist 1997c:101-102,fig. 6).
It is significant that so few terrestrial species are represented in the faunal remains from the Mesolithic dwelling sites on Gotland, that the meals seem to have been composed to 99% of marine foodstuffs. The Mesolithic (and Neolithic) subsistence economy, seasonality and faunal history of Gotland have recently been summarized elsewhere (Lindqvist & Possnert 1997c:35, 39-44, 51-64, 73-76, 79-81;Lindqvist & Storå 1997:22-23). The grey seal cubs were apparently mainly hunted during the winter or spring, but the ringed seals were hunted during the late summer and autumn. The salmons were caught during the summer half of the year. Migrating aquatic birds and hares were mainly caught during the autumn/winter and spring. Initially mountain hare (Lepus timidus) was present on the islands of Stora Karlsö and Gotland, and later red fox (Vulpes vulpes) appeared. These species may have crossed the sea on fairly thin winter ice. The presence of dog, hedgehog or field mouse has not been proved during this earlier Mesolithic period on Gotland, but they appear later during the final Mesolithic or Neolithic periods. This is due to the fact that Gotland is a zoogeographically isolated island in the central Baltic Sea, and quadruped terrestrial animals may only have crossed the Baltic Sea on winter ice, on floating tree trunks, or were unintentionally transported by man in canoes, rafts, etc. The access to boats or large canoes, which are depicted on the hunters' Stone Age rock engravings along the North Fennoscandian coasts (Lindqvist 1983(Lindqvist :3-6, 1984(Lindqvist :24-27, 1994, was naturally a prerequisite for the human settlement of Gotland, as well as for a successful sea mammal exploitation and fishing during a period with a milder climate than today. During the earliest period of occupation, according to the animal bone material from levels G. 11-8 in the Stora Förvar Cave, there apparently was a breeding colony ofhundreds ofgrey seals (Halichoerus grypus) on the island of Stora Karlsö. The climate during the postglacial climatic optimum late Boreal and early Atlantic periods was warm, and the mild winter ice conditions seem to have favoured the grey seal, which fairly recently may have immigrated from the north Atlantic Ocean. It may have immigrated into the Ioldia Sea across the Närke Strait in Middle Sweden, and then established a population that became isolated in the Ancylus Lake. The warm climate is also indicated by the rare occurrence of the ringed seal (Phoca (Pusa) hispida), which demands firm ice during the breeding season (Lindqvist & Possnert 1997c:39-44, 51-64). The early occurrence of water chestnut (Trapa natans) Current Swedish Archaeolagy, Vat 7, /999 In spite of the access to the large seal resource, it is a little mysterious that people chose to settle on Gotland, which lacked other terrestrial game than hare. They probably were quite surprised when they realized that this large islandin spite of a lush vegetation and plenty of fresh waterentirely lacked large terrestrial game. The almost total lack on Gotland (except possibly in the Stora Bjärs grave in Stenkyrka parish) of formal geometric microlithic tools and microblade cores, which are typical forms in the material culture on the mainland, is probably due to this lack of large terrestrial game on Gotland and the island of Stora Karlsö. The two so-called bird arrowheads, that is slotted bone points, a typical artefact in the Maglemosian technocomplex during the late Boreal and early Atlantic zone on the mainland (Lidén 1942;Welinder 1971), from G.7-6 in the Stora Förvar Cave (Rydh 1931:28,fig. 9;Schnittger & Rydh 1940:65, Pl. l:1) and one from the Stora Bjärs grave (Arwidsson 1979:21fig. 4 (Boule & Vallois 1957:196, 198-199;Gore 1994:8-9;Stringer & Gamble 1994:13-14).
The crania and extremity bones showed remarkable traces of work (cut-marks and splitting), that were interpreted as signs of (butchering and) marrow-extraction and hence cannibalism. Although Gustav Retzius, in a lecture addressed to the Swedish Society for Anthropology and Geography in November 21 1890, noted that the skulls were not of the same primitive type as the Neanderthals, he interpreted the marks as signs of cannibalism (Retzius 1890;Rydh 1931:23,41-42;Schnittger &, Rydh 1940:38-39, 42, 46-48 Västergötland. This was, as we shall see, a misinterpretation. In fact, the earliest Stora Förvar Cave humans are not only earlier in date than the megalithic tombs, but also earlier than the Bäckaskog burial in Scania Current Swedish Archaeotogy, Vot 7, t999 found in 1939 (Hanson 1941;Althin 1950;Munthe 1954;Gejvall 1970;Welinder 1971), the two or three Kambs burials found in 1939 and 1947 (Arwidsson 1949;Gejvall 1949;Larsson 1982) and the Stora Bjars burial found in 1953 (Munthe 1954;Arwidsson 1979;Gejvall 1979;Larsson 1982)     Additionally there are a number of Vertebrae and Costae, apparently belonging to various individuals, some of them mature.
Three sacral vertebrae probably belong to individual no. 3.
Hence, there are at least four hitherto dated individuals from the ca. 8,200 BP (c. 7,500-7,000 cal BC) interval, and three dated individuals later than 7,900 BP in the Mesolithic levels from the cave and one from G.7. Including the undated fire-damaged cranial bones, and the fire-damaged clavicle, totally 9 or 10 individuals are represented.
Of the cranial bones only a couple are robust enough to stem from adult men; most of them seem to stem from young individuals.
ON THE STORA FÖRVAR CHRONO-LOGY AND STRATIGRAPHY Several earlier scholars have assigned the aceramic lower levels in the Stora Förvar Cave to the Vrå/Säter II period, which was apprehended as corresponding to the period of the so-called axe dwelling sites, thought to be contemporaneous with the dalmen period, approximately 5,000/4, 500 BP (Nihlén 1927:65;Rydh 1931:11,19;Schnittger & Rydh 1940:65, 75, 78-79;Stenberger 1964:96, 102;Clark 1976:115,fig. I;Knape & Ericson 1983:169, 172, 1988Ericson 1989:192-193,198;Ericson &, Knape 1991:198, 202-203). This is, according to our artefact and osteological studies and dating results, obviously a misinterpretation. However, Henric Munthe (1940Munthe ( :197, 225, 1954claimed that the two so-called bird arrowheads, that is slotted bone points of Lidén's type A (Lidén 1942:28-30)from G.7-6, which are typical artefacts for the Maglemosian techno-complex during the late Boreal and early Atlantic zones and the final Ancylus and early Mastogloia stages (Welinder 1971), were associated with the first occupation in the Stora Förvar Cave and that it occurred already during the Ancylus Lake stage. Also Mats Malmer (1962:940)assumed the presence of Mesolithic layers in the cave. In a couple of lectures given at the Dept. of Archaeology, Stockholm University, in 1982 and 1984 one of the authors (CL) claimed that these early levels (G. 11-8) must be Mesolithic. This conclusion was based on the almost total lack of pottery finds in the lower levels; that the flints in this preceramic cultural layer were mainly composed of narrow and thin blades apparently pressed from several conical platform cores, whereas there are relatively more wide and thick flakes in the ceramic cultural layer; and that the fauna also differ markedly from the fauna in the ceramic cultural layer. A calculation, based on the assumption that the sedimentation rate had been constant throughout the stratigraphy, and according to which the base level could be dated approximately to 6, 000-6,500 BP, was also presented.
Furthermore, in a lecture given by us (CL and GP) at the Department of Archaeology in Stockholm in 1993 (as well as at international conferences, for instance in Esbjerg in 1993 and in Copenhagen 1998), we presented a new chronological interpretation based on a series of AMS "C dates on human and animal bone collagen from the cave, implying that the levels G. 11-8 in fact comprised the earliest known Mesolithic cultural layer on Gotland, and that there was a chronological and stratigraphical hiatus between ca. 7,440 and 5,500 BP (See tab. 2 and figs. 17-18).
Since datings on bone collagen of marineadapted humans arein this case due to the consumption of grey seal and salmon, and later harp seal, porpoise, cod and herringmore or less affected by the reservoir effect, and as such show an apparently older age than dates conducted on contemporaneous terrestrial material, and since the degree of the reservoir effect on bone material from various Baltic stages in prehistoric time was unknown, we also made datings on mountain hare (Lepus timidus) bones found in the same layers as the human bones described above.   Since the hare as a terrestrial species is unaffected by the reservoir effect, dates on hare bone are more reliable than on bone collagen of seal and marine-adapted man; the latter is apparently influenced by a reservoir effect of about 100 years in contrast to the hare bones during the early Mesolithic on Gotland, and seem to give somewhat earlier datings ( fig. 17).
Taken together, this indicates that resources were available and indeed utilised more or less all the year round. Earlier interpretations include the extreme view that the cave was merely a dump for skinned seal bodies, and that the cave was at most a temporary seal-hunting station for seal cub fur extraction (Knape & Ericson 1988;Ericson & Knape 1991).The find of skeletal remains of a suckling baby, a girl and two women in the thick Mesolithic cultural layer in the cave, however, may instead imply that the occupation was a semi-permanent base camp. Or should we interpret it the other way around: that young girls and women with suckling babies joined the men at the temporary hunting stations?
There is, however, still another hypothetical alternative: that the human remains in the cave at least partly derive from persons who had died elsewhere, and subsequently were transported to the island of Stora Karlsö.
The fact that human bones, including firedamaged cranial bones, crushed cranial bones and a frontal bone with cut-marks (flg. 14, 16), from a number of individuals of various living ages and sex have been identified among seal bones, flints, stone axes and ashes in the Stora Förvar Cave, at the same time as human bones are missing at the other Mesolithic dwelling sites on Gotlandexcept at Svalings, where a 4.5~3 cm cranial fragment was identified by Elias Dahr, although the find circumstances are unclear and the bone has not been possible to retrieve (Munthe, Hede & Lundqvist 1928:87; Munthe 1954:700)may imply that separate, burial ceremonies diverging from the formalised burial customs ('hocker' graves) at Stora Bjärs in Stenkyrka parish and Kambs in Lummelunda parish on northwest Gotland (Althin 1950;Arwidsson 1949Arwidsson , 1979Munthe 1954;Larsson 1982) (see fig. 1 and fig. 19ab) occurred in the Stora Förvar Cave on Stora Karlsö. Perhaps they took place within the framework of an ancestral cult that included some peculiar activities such as skeleting, scalping and ritual cannibalism, followed by a quite different type of disposal of human bones in the meal-scrap midden inside the cave.
The frontal bone with cut-marks and the unburnt, fire-damaged or burnt cranial bone fragments, showing signs of crushing (flg. 14, 16) are not unique finds, since Palaeolithic, Mesolithic and Neolithic human skull and postcranial bones with cut-marks etc. also occur at for example Vindija and Krapina in Croatia (Gore 1996:8),Gough's Cave, Ched- Fig. 19.a-b. The male burial from Stora Bjärs (left) and the male burial no. Ifrom Kambs (right). The Stora Bjärs grave is AMS "C-dated on bone collagen to 6, 992-6, 595 cal. BC (å"C: -17.8%o vs PDB), and a female skeleton (no. II) from Kambs is conventionally "C-dated on bane to 7, 007-6, 622 cal. BC (å"C. -18.0%o vs PDB) (Arwidsson 1949, Larsson 1982. There are also a couple ofinj uries: a hole in the right parietal with signs of healing and a fresh woundin the left side ofthe mandible. Furthermore a slotted bone point was found in the pelvis. This male had apparently been involved in some physical conflicts resulting in inj uries that ultimatelv caused his death. This is remarkable, since the population on Gotland dut ing the final Boreal/ early Atlantic zone shift is apprehended as fairly small. dar, in England (Currant, Jacobi & Stringer 1989:135fig. 5-6), Dyrholmen  Sweden, Zvejnieki in Latvia, and Jushnij Olenij Ostrov in East Carelia (Russia). Although evidently not found in formal burials, the two skulls first found in the Stora Förvar Cave seem to have been found together with postcranial bones surrounded by rocks (a rooffall accident?), and in 1892 Hjalmar Stolpe found ". . . parts of a human skeleton, (some ribs and vertebrae, half a pelvis, both femora and the lower legs and fragments of an arm) brought together in a pile .. . " on the floor of the cave, probably in parcelle F or G. Stolpe interpreted these finds in terms of cannibalism, but it could perhaps have represented a skeleton 'tied together' in the same way as at Kambs, albeit disturbed. Stolpe also mentions numerous human bones, including skull bones showing signs of blows and burnt human bones, from F. 1 1 (Schnittger & Rydh 1940:47-48). The latter human bones, which have not been possible to study and which may have been lost, could in fact belong to the same individuals as those analysed by us in the adjacent parcelle G.
It remains to be investigated whether such informal disposals of numerous dead individuals have been conducted also in other caves on the islands of Stora and Lilla Karlsö. Provided approximately the same number of individuals were disposed of in each cave (there may be more individuals than these nine if also the Mesolithic levels in the other excavation areas in the Stora Förvar Cave are included), it would mean that perhaps as many as a hundred individuals are buried on these islands, which may have comprised an analogy to such burial islands as Jushni Olenij Ostrov in East Carelia or Téviec and Hoedic in France. Viewed