Dating the Trollesgave Site and the Bromme Culture E Chronological ﬁx-Points for the Lateglacial Settlement of Southern Scandinavia

Journal of Archaeological Science 40 (2013) 4663e4674

Contents lists available at SciVerse ScienceDirect

Journal of Archaeological Science

journal homepage: http://www.elsevier.com/locate/jas

Dating the Trollesgave site and the Bromme culture e chronological ﬁx-points for the Lateglacial settlement of Southern Scandinavia

Anders Fischer a,b,*, Morten Fischer Mortensen b, Peter Steen Henriksen b, Dorte Rørbeck Mathiassen c, Jesper Olsen d a The Danish Agency for Culture, H.C. Andersens Boulevard 2, DK-1553 Copenhagen V, Denmark b The National Museum of Denmark, Danish Prehistory and Environmental Archaeology, Ny Vestergade 11, DK-1471 Copenhagen K, Denmark c Nordvænget 18, 2tv, DK-3520 Farum, Denmark d Department of Physics and Astronomy, Aarhus University, Ny Munkegade 120, DK-8000 Aarhus C, Denmark article info abstract

Article history: The Bromme culture belongs to the Lateglacial, the period when people settled in the recently degla- Received 24 March 2013 ciated Southern Scandinavia. Until now there have been only a few imprecise ﬁx-points relating to the Received in revised form chronological position of this archaeological culture. This situation can now be improved with the aid of 13 June 2013 research results from a Bromme culture settlement at Trollesgave in SE Denmark. Using pollen and plant Accepted 30 June 2013 macrofossil data, Lateglacial lacustrine deposits containing waste material from the settlement can be assigned to the end of the climatically mild Allerød period. A series of 14C dates establishes the age of the Keywords: settlement as c. 10 826 49 14C years BP (12 871e12 590 cal yr BP). By correlation with climate data from Late Palaeolithic Bromme culture the Greenland ice cores, the occupation can be assigned to the early part of the cold climatic zone GS-1, Allerød thus demonstrating that the global ice-core climate zones are not absolutely synchronous with the Younger Dryas regional division into biozones. Pollen analysis Ó 2013 Elsevier Ltd. All rights reserved. Climate change Southern Scandinavia

1. Introduction (Eriksen, 2002; Clausen, 2003; Pedersen, 2009; Petersen, 2009; Burdukiewicz, 2011; Riede and Edinborough, 2012). Whether it As the archaeological record stands today, the Bromme culture also extended over northern Poland and parts of England represents the northernmost human presence in Western Europe (Sørensen, 2010) remains unresolved (cf. Pedersen, 2009,19;Riede during the Lateglacial (Fischer et al., 2013). The culture was origi- et al. 2011). The culture is usually taken to be an expression of the nally defined on the basis of the flint inventory from the locality at first complete adaptation to, and year-round presence in, the Bromme, eastern Denmark (Fig. 1). This is characterised by only recently deglaciated landscape of Southern Scandinavia. An three types of tool with retouch: scrapers, burins and, not least, element in this adaptation was that the meticulous and economical robust tanged points (Westerby, 1946, 1987; Mathisassen, 1948; (in terms of raw material) flint technology that had been practised Fischer and Nielsen, 1987). Over time, as further finds assem- in the preceding and generally more southerly distributed blages have been discovered, it has become increasingly clear that Hamburg and Federmesser cultures was abandoned in favour of a the features common to the culture also include a characteristic significantly simpler, in technical terms, and markedly more prof- approach to the production of blades (Andersen, 1973; Fischer, ligate craft tradition. This change probably reflects the relative 1990; Madsen, 1992, 1996). abundance of large flint nodules of good quality to which the The territory covered by the Bromme culture comprises population gained access when they made their entry into the present-day Denmark, the southernmost part of Sweden, the Southern Scandinavian young-moraine landscape (Fischer, 1991; northern parts of the countries located south of the Baltic and the Petersen, 2009). adjacent, now sea-covered, parts of the Baltic and the North Sea Despite the fact that research has been carried out into the Bromme culture through several decades, it remains one of the most poorly-dated archaeological cultures of the North European low- * Corresponding author. The Danish Agency for Culture, H.C. Andersens Boule- fi vard 2, DK-1553 Copenhagen V, Denmark. lands. Signi cant cultural-historical questions therefore remain un- E-mail address: afi[email protected] (A. Fischer). resolved. For example: Did the culture emerge under the influence of

The present article presents a preliminary status with respect to the dating of the Bromme culture and attempts thereby to establish a foundation for resolution of some of the fundamental archaeological problems associated speciﬁcally with the Lateglacial in Southern Scandinavia, as well as with the Late Palaeolithic epoch generally in the North European lowlands. The Trollesgave settlement plays a critical role in this investigation in that a compre- hensive body of archaeological and environmental evidence from this locality offers special opportunities to establish a relatively solid chronological ﬁx-point. The material recovered from this locality has now been brought out for analysis almost 40 years after it was excavated because, in the interim, no better basis has emerged for dating Late Palaeolithic settlements in Denmark. The sampling and analysis procedures were state of the art in the 1970s, but by today’s standards there is an unfortunate lack of loss on ignition and magnetic susceptibility data.

2. Material and methods

2.1. The Trollesgave site

Soon after its discovery in the early 1970s the Trollesgave site was identified as a subject for further investigation, in part because it offered a hope of material being obtainable for scientific dating (Fischer and Mortensen, 1977; Fischer, 1990). The Late Palaeolithic settlements of the lowlands of western and central Europe are usually associated with sandy, naturally drained sediments. Due to biological decomposition and leaching, as well as the abrasive effect of freeze and thaw processes, pollen, charcoal, faunal remains etc. e materials that could form the basis for scientific dating e are hardly ever preserved (e.g. Taute, 1968,19e27; Tromnau, 1975a, 1975b; Sulgostowska, 1989; De Bie and Caspar, 2000; Johansson, 2003; Bogen et al. 2003; Fischer, 2013). At Trollesgave, however, deep, waterlogged lacustrine deposits existed in the vicinity of the settlement area. By continuing the excavation out into these wetland deposits, it proved possible to demonstrate the presence of cultural remains in an undisturbed layer which could then be dated with the aid of scientific methods. As will become apparent below, this exer- cise was associated with significant difficulties. The locality had e like most Late Palaeolithic lake-shore sites in NW Europe e been exposed to complex and extensive erosion and solifluction processes. In the following, a summary is given of the outcome of the archaeological-geological fieldwork carried out between 1973 and Fig. 1. The location of Holmegård Bog and the site of Trollesgave, relative to the 1975 on the basis of which the date presented below was estab- present-day geography and the extent of land, sea and ice cap in Southern Scandinavia lished. It included the excavation of the main SE trench (Fig. 2), during Middle to Late Allerød times (A). Other Bromme culture sites mentioned in the where intact lacustrine deposits were encountered containing a text are also marked on the map. Dots indicate sites that have been dated radiomet- small number of discarded flint artefacts from the Bromme culture. rically. The map of the Holmegård Bog area (B) gives the topographical location of the Trollesgave site and other well-defined Late Palaeolithic activity areas recorded during In the other excavation trenches in the wetland deposits, the intensive field-surveying along the shores of the former lake system. Geographical Lateglacial sediments were generally heavily disturbed or features drawn on the basis of M. Houmark-Nielsen 2012, Fig. 14e13 (A), and Fischer completely eroded away and none of them contained in situ Late 1985, Johansson 2003 (B). Palaeolithic material. the Allerød period’s mild climate and the rapidly improving ecological potential for a hunter-fisher-gatherer existence in South- 2.2. The stratigraphy ern Scandinavia (Fischer,1991; Eriksen, 2002)? Or did it first develop late in Allerød times in response to an ecological catastrophe which In principle, the stratigraphy of the lacustrine deposits at Trol- took place around 12 900 cal yr BP; i.e. when the Lacher See volcano lesgave represents a standard succession from lake to bog to carr, spewed out colossal ash layers which then settled as a suffocating with gyttja deposits at the base, then fen peat and uppermost blanket across enormous areas further to the south (Riede, 2008; wood (alder) peat (for definitions of the stratigraphic terms Riede and Edinborough, 2012;cf.Sørensen, 2010)? Did the employed see Troels-Smith, 1955). The natural accumulation of Bromme culture (and humans in general) disappear from Southern organic sediments is, however, interrupted in several places by Scandinavia as a consequence of the extreme cold during the major erosive deposits of minerogenic material, resulting in a far Younger Dryas (Petersen, 2009)? What was the nature of the cultural from simple stratigraphy. This is demonstrated here by two sec- and demographic substrate from which the culture arose? And did it tions cut respectively parallel with, and at right angles to, the enjoy a cultural and genetic afterlife in subsequent periods? original lake shore (Figs. 3e5). A. Fischer et al. / Journal of Archaeological Science 40 (2013) 4663e4674 4665

Fig. 4. Section through the lake deposits running parallel to the original lake shore, seen from the west. Layers 2ae6 represent the period from Allerød times to the Early Preboreal. Layer 2b was found to contain sporadic finds discarded from the nearby Fig. 2. Excavated parts of the Trollesgave site and the depth of the sediment in the Bromme culture settlement. The numbers running up the vertical edges of the section adjacent areas of the bog, mapped by geoelectric survey. The darker the hatching, the give the height in metres above Danish Ordnance Datum. The vertical arrow shows the deeper the lacustrine and bog deposits. The greatest depth of wetland sediments is c. point of contact between Figs. 4 and 5. 7 m. This article gives an account of depositional conditions in the SE excavation trench Sediment descriptions for Figs. 4 and 5. 1: Stratified sand and gravel, in places with a (white polygon). Geo-resistance features drawn on the basis of Mathiassen (1986). substantial content of white chalk lumps. 2a: Weakly stratified, coarse detritus gyttja of varying sand content. The abundance of wood fragments increased away from the shore. In the easternmost parts of the trench the layer consisted primarily of branches and trunks up to 8e9 cm in diameter e mostly of birch but also willow and juniper. The layer was also found to contain seeds of buckbean (Menyanthes trifoliata) and sedges (Carex sp.) and possibly mare's-tail (Hippuris vulgaris). 2b: Weakly stratified, very sandy gyttja containing twig fragments, charcoal and occasional stones. The content of wood fragments decreased gradually upwards e possibly representing a decrease in water depth. 3: Clay gyttja containing fragments of twigs and branches and occasional charcoal particles. Stripes of fine sand uppermost. 4: Stratified sand and gravel. 5: Humus-rich sand. 6: Sandy sediment with clay, gravel and stones and some charcoal. Increasing content away from the lake shore of gyttja lumps and pieces of wood eroded out of older lake sediments. 7: Gyttja or fen peat. 8: Gravel and sand, humus-rich in parts. 9: Peat or gyttja with fen peat. 10: Gravel and sand, humus-rich in parts. 11: Peat e gradually shifting from fen peat at its base to wood (alder) peat uppermost. 12: Sandy, very peat-rich plough soil and recent intrusions from this.

With respect to the dating of the Lateglacial settlement at the site, attention is focussed primarily on layer 2b and its content of cultural remains in primary context. This layer was found completely undisturbed at the NE end of the trench, where it was sealed by layer 3 (Figs. 3, 4 and 6). The minerogenic layers (6, 8 and 10) represent erosion of the littoral zone and the slope running up from the lake. They result from geological processes of markedly differing character. The lower horizon (layer 6) was, in the eastern part of the trench, formed by a series of overlapping flows of relatively fluid material, which incorporated large quantities of lake sediments e not least derived from the near-coastal parts of layer 2. The structure and sorting of the material are consistent with the layer being deposited under water. In the western part of the trench, the layer was completely dominated by minerogenic material which, to a very significant extent, derived from the slope that led up from the lake shore at that time (Fig. 7). An up to 7 m deep depression located east of the settlement resulted partly from the melting of residual glacial ice (see Fig. 2). Fig. 3. The main SE trench showing the positions of the surveyed and drawn sections. Subsidence of the deposits on top of the ice caused deformation The curly brackets show the location of the two sections reproduced in Figs. 4 and 5. The two arrows show the positions of Fig. 6A and B. The horizontal extent of the gyttja and folding of layers as observed in the long trench located containing artefacts (layer 2) is also marked, as well as the area where this layer was immediately to the north of the main trench dealt with here. This completely intact and sealed by clay gyttja (layer 3). The extent of layer 2 towards the deformation definitely took place in Allerød times and possibly also west (light shading) corresponds approximately to the extent of the lake at the time of subsequently (Mathiassen, 1986). The collapse of the lake basin occupation, although the solifluction lobes (layer 6) had disturbed the shore-near part floor created steep sides which destabilised the littoral zone and of layer 2 at the northern and southern ends of the trench. The star symbol at the fl western side of the trench marks an accumulation of flint debitage e a flint workshop presumably contributed to the soli uction that took place in Early 14 which was apparently located on the lake shore. Preboreal times (cf. C dates from layer 6). Fig. 5. Section cut at right angles to the original lake shoreline, seen from the south (sediment descriptions are listed below Fig. 4). At the base, melt-water deposits (layer 1). Above these are sediments laid down in shallow water close to the lake shore (layer 2). The lowermost part of this (2a, pollen-dated to the Allerød period) contained numerous twigs and branches, many of which had been gnawed by beavers (Castor fiber). The lower parts of the stratigraphy had been heavily deformed and eroded by a mud flow (layer 6, from the early part of the Preboreal biozone). In the western part of the section, postglacial wetland deposits (layers 7, 9 and 11) alternated with minerogenic erosion deposits originating from the slope above the lake shore (layers 8 and 10). The vertical arrow shows the point of contact between Figs. 4 and 5.

Fig. 6. Vertical sections seen from the south (A) and west (B) (cf. Fig. 3). On Fig. 6B, soliﬂuction deposits cut down through the sandy layers 5 and 4 as well as the dark clay gyttja (layer 3). The twig-rich sand layer with cultural remains (layer 2b) was intact here. The dark colour of layer 2a was due to its substantial content of organic material. The white sections of the surveying pole are 20 cm in length. Photos A. Fischer.

Fig. 7. A soliﬂuction lobe of pale minerogenic materials (layer 6) has forced its way down through the twig-rich Allerød gyttja (layer 2a). The same square metre seen from the south (left) and east (right). Photos A. Fischer. A. Fischer et al. / Journal of Archaeological Science 40 (2013) 4663e4674 4667

At the base of the western end of the trench there was a sand/ gravel deposit that was locally rich in flint debitage from the Bromme culture. The pieces of flint could be refitted to produce a small number of flint nodules. The fragments derived from the individual nodules formed small, partially overlapping concentrations which collectively represent the apparently virtually undisturbed remains of a workshop used by two flint knappers, located by the shore of the lake (Fig. 8; cf. Donahue and Fischer, 2013). A handful of flakes and a core from this workshop had been thrown out into the lake and were incorporated within layer 2b. Fig. 9 illustrates these important links between the activity areas on land and the datable lacustrine deposits. The age of some of the layers was established in approximate terms already during the actual excavation. Preliminary pollen analyses indicated that layer 2a dated from the Allerød period, layer 3 from the Younger Dryas, and layer 7 from advanced parts of postglacial times.

Fig. 9. Horizontal distribution of refitted flint fragments (tools and debitage), most of 2.3. Cultural remains in the stratigraphy which were found in the workshop shown in Fig. 8. One of the flakes had been thrown out into shallow water where it was found incorporated within lacustrine sediments, Worked flint was demonstrated in four horizons within the layer 2b from Late Allerød times (grey shading). wetland sediments encountered in the SE main trench: layers 2b, 6, The overall evaluation of where in the stratigraphy Late Palae- 8 and 10. If the genesis of these sediments is taken into consider- olithic cultural traces occur in situ therefore points, clearly and ation, then only the artefacts recovered from layer 2b can be unequivocally, to layer 2b. considered to represent definite activity traces in a primary context. These artefacts (e.g. Fig.10) stand out by being markedly larger than the components which otherwise characterise the deposit, i.e. sand 2.4. Pollen and plant macro-remains and small plant remains. This suggests that they represent waste thrown out from activity areas on and above the lake shore (cf. A pollen diagram was produced in 1985 on the basis of a series of Figs. 8 and 9). samples taken from P7 (Fig. 3) with the aim of establishing the Charcoal and charcoal dust are considered possible traces of boundary between the Allerød and Younger Dryas periods. A total of human occupation in the immediate vicinity, and were found in varying quantities in more or less all the deposits encountered in the trench. The greatest concentration of large, often only partially charred, pieces of wood was demonstrated in layers 2b and 6. Microscopic charcoal was also observed in limited quantities in layer 2a and, more frequently, in layers 2b and 3 (Fig. 11). The high concentration in layer 2b indicates local burning.

Fig. 8. Sandy deposit containing flintdebitagewhich could be refitted to produce virtually complete flint nodules. A few flakes and a core from these refitted items were found in Fig. 10. A burin of Bromme character found in layer 2b. Length 6 cm. Photo A. Mik- undisturbed gyttja (layer 2b) some metres out into the lake basin. Photo A. Fischer. kelsen, Danish National Museum. 4668 A. Fischer et al. / Journal of Archaeological Science 40 (2013) 4663e4674

Fig. 11. Summary of the pollen data from Trollesgave (upper) and from cavities of two faunal remains found at the Bromme site (lower). In the Trollesgave diagram the local pollen zone boundaries are marked by broken lines. Significant fluctuations in for example oak fern (Gymnocarpium) in the upper part of layer 2b and in crowberry (Empetrum) and heathers (Ericaceae) in the lower and middle parts of layer 3 indicate that the sediments do not incorporate any significant eroded material from older settlements. eight samples were analysed, covering layer 2b and the adjacent of pollen from dwarf birch and grasses shows, however, that the parts of layers 2a and 3 (Kolstrup, 1985). Each sample was prepared forest was light and open and perhaps had a more mosaic-like using standard techniques, mounted in glycerine and analysed structure. This situation is also seen in several other pollen at 400, 650 and 1000 magnification. Identification follows studies and is ascribed to the Allerød period (Jessen, 1920a,b; Fægri and Iversen (1975). The results are summarised in Fig.11, with Andersen, 1980; Kolstrup, 1982, 2002; Kolstrup and Buchardt, trees, shrubs, dwarf shrubs and herbs being included in the pollen 1982; Bennike et al., 2004). sum. The diameter of Betula pollen was measured in order to TG2: Upper part of layer 2b. Characterised by a marked reduc- separate approximately pollen of tree birch from that of dwarf birch tion in tree birch type from around 40%e12%, while the percentages (Betula nana). Grains of B. nana tend to be smaller, but an absolute of pine, grasses, willow and oak fern (Gymnocarpium) increase. separation is impossible since there is an overlap in grain size be- The sharp reduction in tree birch type is caused by a cooler tween B. nana, Betula pubescens and Betula verrucosa (e.g. Birks, climate. The pine maximum is an effect of the decrease in local 1968; Usinger, 1977; Andersen, 1980). In the present study a fixed pollen production, giving a relatively larger proportion of pollen threshold at 20 mm was used to distinguish B. nana type from tree transported over long distances. Increased values of grasses and birch types (e.g. Berglund and Digerfeldt, 1971; Usinger, 1975, 1978; oak fern are likely to represent a response to a more open forest Stockmarr, 1975; Usinger, 1978; Kolstrup, 1982). These data should environment (e.g. Krog, 1954). This change can be correlated bio- be seen as a supplement to the results of macrofossil analysis, which stratigraphically with a similar change observed during the final provides a more powerful tool for determining the local presence of phase of the Allerød period at a number of other localities in the different Betula species. The pollen diagram shown in Fig. 11 Denmark (e.g. Iversen, 1942; Krog, 1954; Mortensen et al. 2011). includes previously unpublished data from the analysis of two TG3: The lower and middle part of layer 3. Birch is strongly pollen samples taken from bones recovered from the Bromme locus dominated by dwarf birch (B. nana) type which has values of be- classicus. tween 31 and 46%. There is generally a more diverse pollen and The pollen diagram can be divided into three local pollen zones: spore spectrum with several light-demanding taxa such as saxi- TG1: Layer 2a and the lower and middle parts of layer 2b. frage (Saxifraga), small clubmoss (Selaginella selaginoides) and sun Dominated by dwarf birch (B. nana) type (25e40%) and common rose (Helianthemum). Values of crowberry (Empetrum)/heathers tree birch type (B. pubescens type) (20e35%) along with grasses (Ericaceae) increase in the second half of the pollen zone. (Poaceae)(5e15%). The pollen assemblage in this zone shows a complete regime In this zone the majority of the landscape was covered by open shift and collapse of the forest ecosystem between the two zones. woodland dominated by tree birch. The relatively high percentage Given its great diversity of open and cold habitat herbs, such as A. Fischer et al. / Journal of Archaeological Science 40 (2013) 4663e4674 4669

Fig. 11. (continued).

small clubmoss (S. selaginoides), purple saxifrage (Saxifraga oppo- fits well, from an ecological standpoint, into the general picture of sitifolia) type, mugwort (Artemisia) etc., in biostratigraphic terms the woodland of the late Allerød (e.g. Iversen, 1954; Mortensen TG3 belongs in the Younger Dryas. The increase in crowberry et al. 2011). Charcoal was observed in the upper samples which (Empetrum) and heathers (Ericaceae) is typical of the late phase of correspond to layer 2b. the Younger Dryas and the earliest part of the Preboreal (Iversen, The plant macrofossils found in layer 3 are typical of the Younger 1954; Usinger, 2002; de Klerk, 2008). It indicates that there is a Dryas. They are dominated by dwarf birch (B. nana), willow (Salix), major hiatus between layers 2b and 3 e an inference which is mountain avens (Dryas octopetala) and crowberry (Empetrum nig- supported by the sharply defined border between these two layers. rum). Remains of common birch (B. pubescens) were not observed. Similar hiatuses from this period are also seen at other sites in The changes seen in the last phase of the Allerød period (TG2) Northern Germany and Denmark (Usinger, 1981). were presumably prompted by a general fall in temperature which In order to gain more information about the local environment resulted in more open woodland. However woodland was still the of the settlement, a series of eight plant macrofossil samples were dominant habitat type in the area even though tree birch is reduced analysed (in 2009). Six samples were selected from layer 2, and two in the pollen assemblage. This is apparent from the plant macro- from layer 3 (from P4 and P6, respectively, cf. Fig. 3). The samples fossil data which show tree birch (B. pubescens) to be the dominant had been stored in ethanol for 35 years prior to the analyses. A sub- species, whereas dwarf birch (B. nana) is completely absent. Simi- sample of 50 ml was extracted from each sample and wet-sieved larly, the cold-tolerant pioneer plants that normally characterise through meshes with a minimum mesh size of 0.25 mm. The the Younger Dryas period in Denmark are also absent from TG2. preservation of organic material was relatively good and there were no indications of significant redeposition. The results of the ana- 2.5. 14C dates lyses are given in Fig. 12. Since the samples for the analysis of plant macro-remains and pollen derive from different sampling locations Two series of 14C analyses were carried out with the aim of it is not possible to present an absolute correlation between the two determining an absolute age for the Late Palaeolithic finds at the diagrams (Figs. 11 and 12). site. The first series (from the 1970s) was obtained by decay The plant macrofossils from layer 2 are characteristic of the counting, while the second (from 2012) employed the AMS Allerød period. They are dominated by common birch (B. pubescens) method. As the latter was able to operate with very small samples, and willow (Salix sp.), but also include an element of dwarf birch it was possible to use material for which the stratigraphic position (B. nana). Dewberry (Rubus caesius), which has edible blackberry- was recorded much more exactly than was generally the case in the like fruits, was also found in one of the samples. This species has former series. The AMS method also allowed exclusive use of ma- not previously been found in Lateglacial deposits in Denmark but it terial of either very modest (thin twigs of willow) or negligible 4670 A. Fischer et al. / Journal of Archaeological Science 40 (2013) 4663e4674

Fig. 12. Plant macrofossil data, listed in stratigraphical order. Layer 2 is dominated by species associated with the form of open woodland that characterised the Late Allerød period. Layer 3 is dominated by cold-tolerant species that are characteristic of the Younger Dryas. biological age (cone scales and seeds), thereby making it plausible end of the Allerød period but to an advanced point in time during that the dates were obtained for material coeval with the deposi- its final phase. tion of the sediments in question. In principle there was a risk of redeposited material being included in the sample series. In prac- 3. Results tice, however, there appear to be no problems of this kind, as a consistent relationship is observed between the position of the The collective result of the various dating methods indicates samples in the stratigraphy and their date (Table 1). that the Bromme culture activities at Trollesgave took place late in When the calibrated 14C dates are shown with the Greenland the final phase of the Allerød biozone, after 11 126 14CyrBP ice-core oxygen isotope data (d18O), it becomes apparent that the (Table 1). There are two dates for the sediment containing in situ lower half of layer 3 belongs to an advanced part of the cold climatic settlement remains from the Bromme culture of, respectively, period GS-1 (e.g. Björck et al.1998; Rasmussen et al. 2006; cf. 11 070 120 (K-2641) and 10 826 49 14C yr BP (AAR-16019). On Blockley et al. 2012)(Fig. 13). Compared with the pollen data out- calibration, the first of these dates falls in the interval 13 199e lined above, the shift from the Allerød period to the Younger Dryas 12 670 cal yr BP (95.4% confidence interval). When compared with must have taken place during the already mentioned hiatus be- the climate data from the Greenland ice cores (Fig. 13), this date tween layers 2b and 3. As the refuse thrown out from the settle- broadly brackets the boundary between the warm period GI-1 and ment appears in the middle of layer 2b, the date of the Bromme the subsequent cold period GS-1. This interval overlaps completely culture settlement does not, accordingly, correspond to the actual with the second date which, on calibration, falls within a much

Table 1 14C dates for lacustrine deposits in the main SE trench e listed in stratigraphical and chronological order. All the samples comprise material of limited biological age. The ages given below e with the exception of K-2256 e have been corrected for isotope fractionation. The correction of K-2256 would probably have altered the date by less than 20 years. The d13C values for the AMS (AAR numbers) samples are derived from the actual accelerator measurements, not from separate measurements by mass spectrometer. Calibration of 14C dates was carried out with OxCal v4.1.7 (Ramsey, 2009) using the atmospheric calibration curve, IntCal09 (Reimer et al., 2009).

Layer no. 14C years BP Calibrated age BP Nature of sample d13C-value (&) Lab. no. (95.4% conﬁdence interval)

6 9760 110 11 599e10 740 Partially charred branch fragment of juniper (Juniperus sp.) 22.8 K-2510 6 9980 150 12 076e11 161 Charcoal of willow (Salix sp.). Several small pieces found close 25.3 K-2511 together and therefore assumed to originally be from one piece 3, lower half 10 287 44 12 381e11 830 Cone scales and seeds of dwarf birch (Betula nana) 24 AAR-16020 2b, middle 10 826 49 12 872e12 592 Piece of willow (Salix sp.) charcoal 27 AAR-16019 2b 11 070 120 13 199e12 670 Three pieces of charcoal and partially charred wood of, respectively, 24.5 K-2641 willow (Salix sp.), birch (Betula sp.) and poplar (Populus sp.), originating from two square metres and at various depths in the layer 2 11 100 160 13 281e12 661 Partially charred branch fragment of willow (Salix sp.) 24.6 K-2509 Transition 2b/2a 11 126 44 13 155e12 799 Piece of willow (Salix sp.) charcoal 28 AAR-16021 2a 11 190 130 13 317e12 737 Branch or trunk of birch (Betula sp.), diameter 6.5 cm, well-preserved Not measured K-2256 bark. Contained numerous rootlets from later vegetation on the site. These were as far as possible removed prior to dating 2a, middle 11 237 42 13 288e12 958 Piece of willow (Salix sp.) charcoal 21 AAR-16018 A. Fischer et al. / Journal of Archaeological Science 40 (2013) 4663e4674 4671

Fig. 13. Correlation of climate data from the Greenland ice cap with 14C dates significant for the establishment of the chronological position of the Bromme culture. The individual dates are shown as probability curves with a 95.4% confidence interval. The cultural horizon in Trollesgave layer 2b belongs to the early part of the GS-1 period. Greenland ice-core climate data from Rasmussen et al. (2006). narrower interval of 12 872e12 592 cal yr BP, corresponding to the period. He was not able to exclude the possibility that it extended a early part of the cold period GS-1. Statistically there is no conflict little into the Younger Dryas, but he considered this unlikely. between the two dates for layer 2b and AAR-16019 can therefore be In the 1980s, the finds recovered from Bromme in the 1940s considered as the best estimate so far for the date of Bromme were subjected to a reappraisal. This established that the Bromme culture settlement at Trollesgave. culture settlement remains recovered from the site corresponded It should be pointed out that the ice-core climate dating of to at least seven horizontally distinct concentrations, each of which Trollesgave to the cold GS-1 period and the vegetation-history could very well originate from its own temporally distinct period of (pollen and plant macro-remains) date to the warm Allerød occupation. Concentration A was characterised by relatively large period are fully compatible, because the earliest part of GS-1 is tanged points with preserved platform remnants, while the adja- characterised by a relatively mild climate (Fig. 13). This example cent and apparently stratigraphically higher concentration B had simply underlines the oft-stated fact that ice-core climate zones GI- tanged points that were generally smaller and often had their 1 and GS-1 are not directly synonymous (synchronous) with, platform remnants removed by retouch (Fischer and Nielsen, 1987). respectively, biozones Bølling-Allerød and Younger Dryas (cf. Iversen’s date relates primarily to concentration A. Subsequently, a Mortensen et al. 2011; Fischer et al. 2013). bone fragment recovered from the Bromme culture layer (‘the dark cultural horizon’) during the excavations in the 1940s was 14C dated to 10 720 90 14C yr BP, 12 862e12 432 cal yr BP, 95.4% confidence 4. Discussion e other chronological fix-points from the interval (AAR-4539) (upper row on Fig. 13). This was a surprisingly Bromme culture late date for a sample expected to derive from the Allerød biozone (cf. Pedersen, 2009, 20). Iversen’s perception of the dating of the The Bromme locus classicus (Fig. 1)wasthesubjectofextensive settlement was, however, reinforced by two pollen analyses of investigations in 1945e46 (Mathisassen, 1948;cf.Westerby, 1946, sediment adhering to the above-mentioned bone and to a fragment 1987). As a consequence, Iversen (1948) was able to establish e pri- of reindeer (Rangifer tarandus) antler, also found in the Bromme marily via pollen analysis e that the settlement dated from the Allerød culture horizon. The pollen data revealed that these two faunal 4672 A. Fischer et al. / Journal of Archaeological Science 40 (2013) 4663e4674 remains became incorporated in the sediments at a time with Trollesgave appears, accordingly, to belong to a relative late part of relatively mild climate (Fig. 11). This impression was further the Bromme culture. In this, as yet, only provisionally defined underlined by the fact that the bone fragment is from an elk (Alces phase (Fischer, 1978; Johansson, 2003,87e89), tanged points are alces), i.e. the characteristic animal for the region during the Allerød dominated by examples which have had the platform remnant period (cf. Iversen, 1967; Aaris-Sørensen, 2009). The elk bone was removed either by breaking off or by retouch. Moreover, double- found in a wetland area located, respectively, c. 13 and 16 m blow burins dominate relative to burins on an existing surface perpendicularly down the slope from finds concentrations A and B. and burins on retouch. Finds concentration B at Bromme can also This excludes the possibility of linking the 14C date specifically with be assigned to this phase, whereas Bromme A appears, on typo- one or other of the presumably chronologically discrete activity logical and possibly also stratigraphic grounds, to be slightly older concentrations. (cf. Fischer and Nielsen, 1987). A more marked typological Most of the other Bromme culture sites that were excavated contrast, relative to Trollesgave and Bromme B, is seen in the in- following the ground-breaking investigations at Bromme itself ventories from Løvenholm and Hollendskær (Fig. 1). These are have only provided little and very uncertain information relevant to characterised by tanged points with a preserved platform remnant the dating of the culture. The most frequent observation relates to and are also remarkable in containing a large proportion of end large vertical spreads (several decimetres) of the apparently chro- scrapers with heavy lateral retouch. The latter presumably refer nologically homogeneous flint assemblages in sandy subsoil. This backwards in time to the Federmesser culture (Madsen, 1983; was the case at Segebro (Salomonsson, 1964), Trollesgave (Fischer Nilsson, 1989). There are, however, no scientific dates for these and Mortensen, 1977, 1978; Fischer, 1990), Langå (Madsen, 1983), probably early examples of the Bromme culture. A recent re- Stoksbjerg Vest (Johansson, 2003) and Stoksbjerg Bro (Pedersen, examination of the finds from Nørre Lyngby does, perhaps, indi- 2009). This is perceived as a phenomenon resulting from frost cate that a typologically early part of the culture existed during the movements (cryoturbation) in the soil during a period of extreme later part of the Greenland ice core period GI-1, perhaps within the cold, probably the middle of the Younger Dryas. At Trollesgave, interval 11 600e11 10 0 14C yr BP (c. 13 200 cal yr BP) (Fischer et al. Stoksbjerg Vest (Jacobsen, 1977) and Stoksbjerg Bro (Pedersen and 2013). Petersen, 2006) there was also evidence of what appear to be in- volutions and frost-sorted ground (cf. Wood and Johnson, 1978). 5. Conclusion These characteristically cold-period phenomena could, at all three localities, be shown to have taken place after the Bromme culture The Lateglacial settlements of the lowlands of Northern Europe artefacts had been deposited. A further indication that the Bromme only rarely offer opportunities for dating by scientific methods. culture apparently dates from the time before the coldest part of With a couple of exceptions (Iversen, 1948; Clausen, 1995; cf. GS-1 is provided by observations from the locality of Bonderup. Weber et al. 2010; Mortensen et al. 2013), the settlement areas Here, re-deposited flint artefacts of Bromme character were found demonstrated to date have been located in association with natu- in solifluction layers, which can be assigned on their pollen content rally drained mineral soil, where chemical and biological decom- to the Younger Dryas biozone. These lay stratigraphically above a position of organic material, in combination with bioturbation and gyttja layer from the Younger Dryas and below a gyttja layer from cryoturbation of the sediments, have erased any possibility of the earliest Preboreal (Fischer, 1982; Fredskild, 1982). pollen- and 14C dating. As for radiometric dates for the Bromme culture, apart from Trollesgave illustrates the naturally-determined dating prob- locus classicus itself, only two other sites are of use. One is the lems associated with Lateglacial settlements in the North European Fensmark locality, where there is a single 14C date, for a piece of lowlands. The reason that it was possible nevertheless to establish a charcoal from a solifluction layer containing flint artefacts from relatively precise chronological fix-point is, to a very great extent, the Bromme culture, of 10 810 120 14C yr BP (13 066e due to the fieldwork being carried out, in its entirety, with complete 12 545 cal yr BP, 95.4% confidence interval, OxA-3614) (Fischer, integration of methods of archaeology and natural sciences. Large 2013). The second is Trollesgave and, in addition to the 14C parts of the locality’s chronologically significant lacustrine deposits dates discussed above, there are three thermoluminescence dates remain intact and are now subject to legal protection (Fischer, for worked flint from a hearth unequivocally linked to the 2013). Consequently, possibilities exist for further investigations Bromme occupation (Fischer, 1990). Their mean values lie some- at the site with up- to-date methods. what later than the date for the settlement obtained on the basis The combined results obtained by the various scientific disci- of the 14C dates for the lacustrine deposits, although their wide plines involved in the investigations show that Trollesgave belongs dating range overlaps with the latter. to such a late part of the warm Allerød biozone that, in absolute If the perspective is broadened to include all four of the calendar years, it falls within the early part of the ice-core climatic archaeological cultures identified from the Lateglacial in Southern zone GS-1, when the temperature was generally falling. Accord- Scandinavia (Hamburg, Federmesser, Bromme and Ahrensburg), ingly, the observations from the site show that global ice-core cli- this situation does not change markedly. Only one further locality matic zones are not synchronous with the regional NW European is added: Slotseng e a water-deposited faunal and flint assem- division into biozones. blage, probably from the Hamburg culture, for which there are The other 14C dates for settlements of the Bromme culture, both pollen and 14C dates (Holm, 1996; Mortensen et al. 2008, Fensmark and Bromme locus classicus, similarly fall so late that they 2013). A significant improvement is obtained if the perspective must also be assigned to the early part of the ice-core cold period is widened further to include NW Germany (Riede et al. 2010; GS-1. Other data presented in this article indicate that the Bromme Weber et al. 2011), but the overall picture for the entire Late culture dates from before the coldest part of GS-1. Vegetation his- Palaeolithic in Southern Scandinavia remains poor. The Lateglacial tory studies suggest that the Allerød ecosystem prevailed a good remains an epoch where the formulation of over-arching culture- way into GS-1, perhaps until c. 12 650 cal yr BP (Andersen et al. historical syntheses is rendered difficult by a lack of chronological 2000; cf. Brauer et al. 2008; Mortensen et al. 2011). The data fix-points. presently available thereby provide a foundation for considering The presently available fix-points for the Bromme culture are whether the end of the Bromme culture was linked with the clearly not representative of the entire ‘lifetime’ of the culture, breakdown of the Allerød environment and the establishment of which apparently included several typologically different stages. the Younger Dryas cold regime. A. Fischer et al. / Journal of Archaeological Science 40 (2013) 4663e4674 4673

Acknowledgements Donahue, R., Fischer, A., 2013. The Late Glacial Trollesgave settlement, Denmark e evidence for a Family Hunting Unit. Journal of Archaeological Science (submitted for publication). The excavations were carried out under the auspices of Næstved Eriksen, B.V., 2002. Reconsidering the geochronological framework of Lateglacial Museum and the National Museum of Denmark. The site archive is hunteregatherer colonization of southern Scandinavia. In: Eriksen, B.V., stored at the National Museum under ref. nos. NM1 1041/75 and Bratlund, B. (Eds.), Recent Studies in the Final Palaeolithic of the European Plain, Jutland Archaeological Society Publications, vol. 39, pp. 25e41. NM8 A5807. Højbjerg. The manuscript was prepared under the auspices of the Danish Fischer, A., 1978. På sporet af overgangen mellem palæoliticum og mesoliticum i National Museum, research initiative Northern Worlds by way of a Sydskandinavien. (Summary: tracing the transition from the Palaeolithic to the Mesolithic in southern Scandinavia). Hikuin 4, 27e50, 150e152. grant from the Danish Council for Independent Research (FKK ref. Fischer, A., 1982. Bonderupebopladsen. Det manglende led mellem dansk palæo- no. 273-08-0424). litikum og mesolitikum? Antikvariske studier 5, 87e100. The investigations of the main SE trench took place with Fischer, A., 1985. Late Palaeolithic finds. In: Kristiansen, K. (Ed.), Archaeological fi Formation Processes. The Representativity of Archaeological Remains from inspiration and active eld- and laboratory assistance from Jørgen Danish Prehistory. Nationalmuseets Forlag, Copenhagen, pp. 81e88. Troels-Smithy and his staff of the Danish National Museum’s Fischer, A., 1990. A Late Palaeolithic ‘school’ of flint-knapping at Trollesgave, Department of Environmental Archaeology: Charlie Christensen, Denmark. Results from refitting. Acta Archaeologica 60, 33e49. Bent Fredskild, Claus Malmros and Kjeld Christensen. Fischer, A., 1991. Pioneers in deglaciated landscapes: the expansion and adaptation of Late Palaeolithic societies in Southern Scandinavia. In: Barton, N., Dr. Phil Else Kolstrup, Uppsala, and cand.mag. Ingrid Sørensen, Roberts, A.J., Roe, D.A. (Eds.), The Late Glacial in North-west Europe. Human Copenhagen, are thanked for making previously unpublished pol- Adaptation and Environmental Change at the End of the Pleistocene. Council for e len data available for this study. The owners of the field at the time British Archaeology, Oxford, pp. 100 121. CBA Report 77. Fischer, A., 2013. The Fensmark settlement and the almost invisible Late Palaeolithic of the investigations, the Nordsted-Jørgensen family, are warmly in Danish field archaeology. Danish Journal of Archaeology 1 (2), 123e141. thanked for accommodation and permission to work at the site. Fischer, A., Mortensen, B.N., 1977. Trollesgave-bopladsen. Et eksempel på anven- Thanks also to everyone who participated in the field activities. The delse af EDB inden for arkæologien. Nationalmuseets Arbejdsmark, Copenha- gen, pp. 90e95. Drawing Studio of Moesgaard Museum (Aarhus) is acknowledged Fischer, A., Mortensen, B.N., 1978. Report on the use of computers for description for producing the graphics. Sincere thanks are due to archaeologist, and analysis of Palaeolithic and Mesolithic occupation areas. In: Kristiansen, K. mag.art. Lisbeth Pedersen, Kalundborg, for critical reading and (Ed.), New Directions in Scandinavian Archaeology. The National Museum of Denmark, Copenhagen, pp. 7e22. constructive suggestions with respect to earlier versions of this Fischer, A., Nielsen, F.O.S., 1987. Senistidens bopladser ved Bromme e en nybear- manuscript. Dr. David Earle Robinson is thanked for swift and bejdning af Westerbys og Mathiassens fund (Summary in German: späteiszei- precise translation of the manuscript into English and for making tliche Wohnplätze bei Bromme. Eine Neubearbeitung des Fundes von Erik Westerby und Therkel Mathiassen). Aarbøger for nordisk oldkyndighed og useful suggestions for its content. Historie 1986, 5e69. Fischer, A., Clemmensen, L.B., Donahue, R., Heinemeier, J., Lykke-Andersen, H., Lysdahl, P., Mortensen, M.F., Olsen, J., Petersen, P.V., 2013. Late Palaeolithic References Nørre Lyngby e a Northern Outpost Close to the West Coast of Europe. Quartär 60 (in press). Aaris-Sørensen, K., 2009. Diversity and dynamics of the mammalian fauna in Fredskild, B., 1982. Nogle pollenanalyser fra prøvegravningen på stenalderbopladsen Denmark throughout the last glacial-interglacial cycle, 115e0 kyr BP. Fossils and Bonderup. Antikvariske studier 5, 100e103. Strata 57, 1e59. Fægri, K., Iversen, J., 1975. Textbook of Pollen Analysis, third ed. Scandinavian Andersen, S.H., 1973. Bro. En senglacial boplads på Fyn. (Summary in English: Bro. A University Books, Copenhagen. Late Glacial settlement of northern Funen). Kuml 1972, 6e60. Holm, J., 1996. The earliest settlement of Denmark. In: Larsson, L. (Ed.), The Earliest Andersen, S. Th, 1980. Early and Late Weichselian chronology and birch assem- Settlement of Scandinavia and its Relationship with Neighbouring Areas, Acta blages in Denmark. Boreas 9, 53e69. Archaeologica Lundensia, Series in 8o, No. 24. Almquist and Wiksell, Stockholm, Andersen, C.S., Björck, S., Bennike, O., Heinemeier, J., Kromer, B., 2000. What do 14C pp. 43e59. changes across the Gerzensee oscillation/GI-1b event imply for deglacial Houmark-Nielsen, M., 2012. Palaeogeography of Southern Scandinavia, Allerød oscillation? Journal of Quaternary Science 15 (3), 203e214. Interstadial. Naturen i Danmark, Geologien, second ed. Gyldendal, Copenhagen. Bennike, O., Jensen, J.B., Lemke, W., Kuijpers, A., Lomholt, S.T., 2004. Late- and Iversen, J., 1942. En pollenanalytisk Tidsfæstelse af Ferskvandslagene ved Nørre postglacial history of the Great Belt, Denmark. Boreas 33, 18e33. Lyngby. Medd. fra Dansk Geol. Forening. København bd 10, 130e151. Berglund, B.E., Digerfeldt, E., 1971. A palaeoecological study of the Late-Glacial lake Iversen, J., 1948. Geologisk datering af en senglacial boplads ved Bromme. Aarbøger Torreberga. Okios 21, 98e128. for Nordisk Oldkyndighed og Historie 1946, 198e231. Birks, H.J.B., 1968. The indentification of Betula nana pollen. New Phytologist 67, Iversen, J., 1954. The Late-glacial Flora of Denmark and Its Relation to Climate and 309e314. Soil. DGU. Række 2, Nr. 80, pp. 87e119. Björck, S., Walker, M.J.C., Cwynar, L.C., Johnsen, S., Knudsen, K.-L., Lowe, J.J., Iversen, J., 1967. Naturens udvikling siden sidste istid. In: Nørrevang, A., Meyer, T.J. Wohlfarth, B., INTIMATE Members, 1998. An event stratigraphy for the last (Eds.), Danmarks Natur. Bind 1. Landskabernes opståen. Politikens Forlag, termination in the North Atlantic region based on the Greenland ice-core record: a København, pp. 345e448. proposal by the INTIMATE group. Journal of Quaternary Science 13 (4), 283e292. Jacobsen, E.M., 1977. Periglaciale fænomener fra yngre Dryas i Danmark. In: Dansk Blockley, S.P.E., Lane, C.S., Hardiman, M., Rasmussen, S.O., Seierstad, I.K., Geologisk Forening, Årsskrift for 1977, pp. 9e11. Steffensen, J.P., Svensson, A., Lotter, A.F., Turney, C.S.M., Ramsey, C.B., 2012. Jessen, K., 1920a. Moseundersøgelser i det nordøstlige Sjælland. Med Bemærkninger Synchronisation of palaeoenvironmental records over the last 60,000 years, and om Træers og Buskes Indvandring og Vegetationens Historie. In: Danmarks an extended INTIMATE event stratigraphy to 48,000 b2k. Quaternary Science Geologiske Undersøgelser, II. række, 34. Reviews 36, 2e10. Jessen, K.,1920b. Moseundersøgelser i det nordøstlige Sjælland. DGU. Række II. Nr. 34. Bogen, C., Hilgers, A., Kaiser, K., Kühn, P., Lidke, G., 2003. Archäologie, Pedologie und Johansson, A.D., 2003. Stoksbjerg Vest. Et senpalæolitisk fundkompleks ved Pors- geochronologie spätpaläolithischer Fundplätze in der Ueckermünde Heide mose, Sydsjælland. Fra Bromme- til Ahrensburgkultur i Norden. In: Nordiske (Mecklenburg-Vorpommern). Archäologisches Korrespondenzblatt 33, 1e20. Fortidsminder Serie C, Bind 3. Det Kongelige Nordiske Oldskriftselskab, Brauer, A., Haug, G.H., Dulski, P., Sigman, D.M., Negendank, J.F.W., 2008. An abrupt Copenhagen. wind shift in western Europe at the onset of the Younger Dryas cold period. Kolstrup, E., 1982. Late-glacial pollen diagrams from Hjelm and Draved Mose Nature Geoscience 1, 520e523. (Denmark) with a suggestion of the possibility of drought during the earlier Burdukiewicz, J.M., 2011. Late Glacial hunteregatherer reactions to the Younger Dryas. Review of Palaeobotany and Palynology 36, 35e63. Dryas cooling event in the southern and eastern Baltic regions of Europe. Kolstrup, E., 1985. Pollenanalytisk undersøgelse af profil x:121,00; y:103,23, Trol- Quaternary International 242, 302e312. lesgave. Copenhagen (unpublished). Rapport. Clausen, I., 1995. Alt Duvenstedt, Kreis Rendsburg-Eckernförde, LA 121. Ein Kolstrup, E., 2002. Some classical methods used for reconstruction of Lateglacial Ajrensburger Kulturvorkommen in allerødzeitlichem Boden. Archäologische environments in the European plain: potentials and limitations. In: Nachrichten aus Schleswig-Holstein 6, 103e126. Eriksen, B.V., Bratlund, B. (Eds.), Recent Studies in the Final Palaeolithic of the Clausen, I., 2003. Das allerødzeitlische Rengeweihbeil aus Klappholz LA 63, Kreis European Plain, Proceedings of a U.I.S.P.P. Symposium, Stockholm1999. Jutland Schleswig-Flensburg. Offa-Zeitschrift 59/60, 15e39. Archaeological Society Publications 39, pp. 11e23. De Bie, M., Caspar, J.-P., 2000. Rekem. A Federmesser Camp on the Meuse River Kolstrup, E., Buchardt, B., 1982. A pollen analytical investigation supported by an 18 Bank. Leuven University Press, Leuven. O-record of a Late Glacial lake deposit at Graenge (Denmark). Review of de Klerk, P., 2008. Patterns in vegetation and sedimentation during the Weichselian Palaeobotany and Palynology 36, 205e230. 14 Late-glacial in north-eastern Germany. Journal of Biogeography 35 (7), 1308e Krog, H., 1954. Pollen Analytical Investigation of a C-dated Alleröd Section from 1322. Ruds Vedby. DGU. II Række, Nr.80, pp. 120e140. 4674 A. Fischer et al. / Journal of Archaeological Science 40 (2013) 4663e4674

Madsen, B., 1983. New evidence of late Palaeolithic settlement in east Jutland. Riede, F., Grimm, S.B., Weber, M.-J., Fahlke, J.M., 2010. Neue daten für alte grabungen Journal of Danish Archaeology 2, 12e31. e ein Beitrag zur spätglazialen Archäologie und Faunengeschichte Nord- Madsen, B., 1992. Hamburgkulturens flintteknologi i Jels. (Summary in English: the deutschlands. Archäologisches Korrespondenzblatt 40, 297e316. Hamburgian flint technology at Jels). In: Holm, J., Rieck, F. (Eds.), Istidsjægere Riede, F., Laursen, S.T., Hertz, E., 2011. Federmesserkulturen i Danmark. Kuml 2011, ved Jelssøerne. Skrifter fra Museumsrådet for Sønderjyllands Amt 5, Haderslev, 9e38. pp. 93e131. Salomonsson, B., 1964. Decouverte d’une habitation du tardi-glaciaire a Segebro, Madsen, B., 1996. Late Palaeolithic cultures of South Scandinavia e tools, traditions Scanie, Suede. Acta Archaeologica 35 (1), 1e28. and technology. In: Larsson, L. (Ed.), The Earliest Settlement of Scandinavia and Stockmarr, J., 1975. Biostratigraphic studies in Late Weichselian sediments near Its Relationship with Neighbouring Areas, Acta Archaeologica Lundensia, Series Bøllingsø. DGU. Aarbog 1974, 71e89. in 8o, No. 24. Almquist and Wiksell, Stockholm, pp. 61e73. Sulgostowska, Z., 1989. Prahistoria miedzyrzecza Wisly, Niema i Dniestru u schylku Mathiassen, D.R., 1986. En sedimentologisk og økologisk undersøgelse af et lacus- plejstocenu. Panstwowe Muzeum Archeologiczne w Warszawie, Warszawa. trint aflejringsbassin fra Ældre Dryas til Præboral Tid i Holmegårds Mose, Sørensen, L., 2010. The Lacher See volcanic eruption e challenging the idea of Sjælland. Bd. 1 & 2. Cand. scient. thesis. Copenhagen University, Copenhagen. cultural disruption. Acta Archaeologica 81, 276e287. Mathisassen, T., 1948. En senglacial boplads ved Bromme. Aarbøger for nordisk Taute, W., 1968. Die Stielspitzen-gruppen im nördlichen Mitteleuropa. In: Ein Bei- Oldkyndighed og Historie 1946, 121e197. trag zur Kenntnis der späten Altsteinzeit. Fundamenta Reihe A, Bd. 5. Köln. Mortensen, M.F., Holm, J., Christensen, C., Aaris-Sørensen, K., 2008. De første Troels-Smith, J., 1955. Karakterisering af løse jordarter. Characterization of Uncon- mennesker i Danmark. Nationalmuseets Arbejdsmark, Copenhagen, pp. 69e82. solidated Sediments. In: Danmarks Geologiske Undersøgelse, IV. Række, Bd. 3, Mortensen, M.F., Birks, H.H., Christensen, C., Holm, J., Noe-Nygaard, N., No. 10. Copenhagen. Odgaard, B.V., Olsen, J., Rasmussen, K.L., 2011. Late-glacial vegetation develop- Tromnau, G., 1975a. Neue Ausgrabungen im Ahrensburger Tunneltal. Wachholtz, ment in Denmark e new evidence based on macrofossils and pollen from Neumünster. Slotseng, a small-scale site in southern Jutland. Quaternary Sciences Reviews Tromnau, G., 1975b. Die Fundplätze der Hamburgerkultur von Heber und Deimern, 30, 2534e3550. Kreis Soltau. In: Materialhefte zur Ur- und Frühgeschichte Niedersachsens. Mortensen, M.F., Olsen, J., Holm, J., Christensen, C., Submitted, 2013. Right Time, Verlag August Lax, Hildesheim. Right Place e Dating the Havelte Phase in Slotseng, Denmark. In: British Usinger, H., 1975. Pollenanalytische und stratigraphische Untersuchungen an zwei Archaeological Reports. Archaeopress. Spätglazial-Vorkommen in Schleswig-Holstein. Mitteilungen der Arbeitsge- Nilsson, T., 1989. Senglacial bosættelse i Vendsyssel. Kuml 1987, 47e75. meinschaft Geobotanik in Schleswig-Holstein und Hamburg 25, 1e183. Pedersen, K.B., 2009. Istidsjægere omkring Knudshoved Odde. Stederne og men- Usinger, H., 1977. Bölling-Interstadial and Laacher Bimstuff in einem neuen Spät- neskene. Museerne.dk. Vordingborg. glazial-Profil aus dem Vallensgaard Mose/Bornholm. Mit pollen-grössen- Pedersen, K.B., Petersen, E.B., 2006. Variability in the Late Palaeolithic: the example statistischer Trennung der Birken. DGU Aarbog 1977, 5e29. of Stoksbjerg Bro, Zealand, Denmark. In: Hansen, K.M., Pedersen, K.B. (Eds.), Usinger, H., 1978. Pollen- und Großrestanalysen Untersuchungen zur Frage des Across the Western Baltic. Vordingborg. Sydsjællands Museums Publikationer, Bölling-Interstadials und der Spätglacialen Baumbirken-Einwanderung in pp. 75e92. Schleswig-Holstein. Schr. Naturw. Ver. Schlesw.-Holst 48, 41e61. Petersen, E.B., 2009. The human settlement of southern Scandinavia 12500e Usinger, H.,1981.Einweit verbreiteter Hiatus in spätglazialen Seesedimenten: Mögliche 8700 cal. BC. In: Street, M., Barton, N., Terberger, T. (Eds.), Humans, Environment Ursache für Fehlinterpretation von Pollendiagrammen und Hinwies auf klimatisch and Chronology of the Late Glacial of the North European Plain. Römisch- verursachte Seespiegelbewegungen. Eiszeitalter u. Gegenwart 31, 91e107. Germanischen Zentralmuseum, Mainz, pp. 89e129. Usinger, H., 2002. Vegetation and climate of the lowlands of northern Central Ramsey, C., 2009. Bayesian analysis of radiocarbon dates. Radiocarbon 51 (1), Europe and adjacent areas around the Younger Dryas e Preboreal transition e 337e360. with special emphasis on the Preboreal oscillation. In: Terberger, T., Rasmussen, S.O., Andersen, K.K., Svensson, A., Steffensen, J.P., Vinther, B.M., Eriksen, B.V. (Eds.), Hunters in a changing world. Environment and Archaeology Clausen, H.B., Andersen, M.L.S., Johnsen, S.J., Larsen, L.B., Dahl-Jensen, D., of the Pleistocene-Holocene Transition (ca. 11000-9000 B.C.) in Northern Cen- Bigler, M., Röthlisberger, R., Fischer, H., Goto-Azuma, K., Hansson, M.E., Ruth, U., tral Europe. Workshop of the U.I.S.P.P.-Commission XXXII at Greifswald 2002. 2006. A new Greenland ice core chronology for the last glacial termination. Inter-nationale Archöologie, Arbeitsgemeinschaft, Symposium, Tagung, Kon- Journal of Geophysical Research D: Atmospheres D6, 1e16. gress 5 (Rahden/Westfalen 2004), pp. 1e26. Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Ramsey, C.B., Weber, M.-J., Clausen, I., Houseley, A.R., Miller, C.E., Riede, F., Usinger, H., 2010. New Buck, C.E., Burr, G.S., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P., information on the Havelte group site Ahrenshöft LA 58 D (Nordfriesland, Hajdas, I., Heaton, T.J., Hogg, A.G., Hughen, K.A., Kaiser, K.F., Kromer, B., Germany) e preliminary results of the 2008 fieldwork. Quartär 57, 7e24. McCormac, F.G., Manning, S.W., Reimer, R.W., Richards, D.A., Southon, J.R., Weber, M.-J., Grimm, S.B., Baales, M., 2011. Between warm and cold: Impact of the Talamo, S., Turney, C.S.M., van der Plicht, J., Weyhenmeye, C.E., 2009. Intcal09 and Younger Dryas on human behavior in Central Europe. Quaternary International Marine09 radiocarbon age calibration curves, 0e50,000 years Cal BP. Radio- 242, 277e301. carbon 51, 1111e1150. Westerby, E., 1946. Da Danmarks ældste Stenalderboplads blev fundet. Berlingske Riede, F., 2008. The Laacher see-eruption (12,920 BP) and material culture change at Aftenavis, 22e23. oktober 1946. Copenhagen. the end of the Allerød in northern Europe. Journal of Archaeological Science 35, Westerby, E., 1987. Da Danmarks ældste Stenalderboplads blev fundet. Aarbøger for 591e599. Nordisk Oldkyndighed og Historie 1986, 43e69. Riede, F., Edinborough, K., 2012. Bayesian radiocarbon models for the cultural Wood, W.R., Johnson, D.L., 1978. A survey of disturbance processes in archaeological transition during the Allerød in southern Scandinavia. Journal of Archaeological site formation. In: Schiffer, M.B. (Ed.), Advances in Archaeological Method and Science 39, 744e756. Theory 1. Academic Press, pp. 315e381.