Preliminary Results and Research Perspectives on the Submerged Stone Age Sites in Storstrømmen, Denmark

3.1 Storstrømmen, Lokalitet A and B

The sites Lokalitet A and Lokalitet B are located at water depths of 10–12 m on both sides of a former lake/freshwater system (Figure 2). Within the basin, organic sediments were preserved from approximately 13 m below the sea level. One hundred fifty-four test pits were dug with a mechanical excavator in the area, revealing three concentrations of worked flint. Eight hundred six pieces of worked flint along with 207 fragments of organic material (fragmented bone, charcoal, burned hazelnut shells, and fishbone) were found during the test excavations. Based on morphological and technological traits of the flint material, the northern-most concentration (Lokalitet B) is dated to the Maglemose culture (8900–6550 BC). The two other concentrations (Lokalitet A) date to the Maglemose (8900–6550 BC) and/or Kongemose culture (6550–5400 BC). All flint materials were found in a transgressional gravel layer, which was most likely redeposited during the transgression of the sites. The sites could be bypassed during the construction work and no excavation was conducted in the area (Johansen, 2017).

3.2 Vordingborg Vesthavn

The site Vordingborg Vesthavn was found during test excavations in the shipping lane of Vordingborg Harbour (Bohr & Hyttel, 2016) (Figure 2).

An area of 27 m² was excavated at the site (Figure 3). The excavated area is estimated to make up approximately 1.4% of the site’s total area based on the distribution of worked flint recorded during test excavations. The site is located around 8–9 m below the present sea level, on the edge of a depression in the postglacial terrain. Gyttja deposits within the basin indicate that the site, during its time of use, was located on the edge of a freshwater basin or stream, where stones and flint blanks were eroded out of the moraine surface. A tightly packed layer of boulders and stones, including flint, was present at the site. The excavated worked flint at the site numbered 8,636 pieces and consists mostly of flakes and test-knapped nodules. Formal tools are almost completely absent in the material (Table 2). The site is interpreted as a gathering and test knapping site. It has not been possible to date the site precisely. However, the site is overgrown by alder swamp peat formation. Radiocarbon dates, from the top of the peat layer (beta-531294: 6242 ± 131 cal. BC) and top of the underlying layer (beta-534946: 6359 ± 93 cal. BC), indicate that the alder swamp formed over the site during the early Kongemose culture and gives a terminus ante quem for the site (Fiedler, 2019).

Figure 3

Vordingborg Vesthavn, site plan and profiles. Contour lines represent the depth of the seabed.

3.3 Orehoved Sejlrende

As a consequence of the bridge construction, the entry channel to the nearby harbour of Orehoved had to be relocated. During test excavations in the area of the new channel, a large amount of flint, dating to the Kongemose culture (6550–5400 BC), emerged in the bucket of the excavator. The material contained large regular blades, cores, core axes, oblique arrowheads, and a large amount of flakes and charred flint. In addition, organic materials such as animal bones and charred hazelnut shells were found. The material was derived from two sites located on either side of the mouth of a former small inlet, running north – south, and opening into Storstrømmen. The inlet only appears faintly on the interpreted postglacial terrain (Figure 2), but stood out clearly during test pitting in the area (Gutehall & Dencker, 2015). The mouth of the inlet was quite narrow, approximately 45 m, depending on the stage of marine transgression in the area. At both sites, the archaeological material was located in two find horizons, consisting of sand/gravel, located 4–7.3 m below the present sea level. Both sites were, based on the morphology of the arrowheads (Petersen, 2008, p. 88; Sørensen, 2017, pp. 39–43), dated to the Kongemose culture’s Villingebæk phase (6150–5800 BC) with a smaller number of the Vedbæk phase (5800–5400 BC). The recorded find distribution of the two sites spanned an area over 26,140 m², exceeding the development area toward both the east and west.

As it was not feasible to excavate the entire area, supplemental test excavations were conducted with a mechanical excavator and with divers to identify the areas with best preservation conditions for organic material and the most likely areas where parts of the original settlement surfaces might have survived the erosional effects of the transgression. Furthermore, a microscopic analysis of the lithic material was conducted to determine to what extent the artefacts from the various test pits were modified by natural processes and thus redeposited and out of the primary context (Donahue, & Burroni, 2015; Donahue, Fischer, Burroni, Malm, & Johansen, 2019). Finally, an excavation with divers was undertaken at the western side of the inlet. The aim was to reach a clearer understanding of stratigraphy, looking for artefact-rich deposits of apparent undisturbed nature and searching for near-shore gyttja deposits containing organic cultural materials. A total of 94 m² were excavated at water depths of 4.7–5.7 m (Johansen & Ravn, 2018). Within the excavated area (Figure 4) were found 25,178 pieces of worked flint and 6,552 animal bones/antlers. The stratigraphy of the site consisted from the top of (1) a layer of marine sand, (2) a coarse-grain marine gyttja with relatively few Mesolithic artefacts, and (3) a sand/gravel layer (interpreted as a beach- or transgressional layer, with hundreds of artefacts per square meter, all of which appear to have been washed out from the settlement surface and redeposited when the site was inundated), and (4) a terrestrial dirt/humus layer with many, very well-preserved artefacts of organic material and flint. Layer 4 most likely represents the sediment of the original settlement surface, but the micro-ware analysis of flint material from the layer (Donahue et al., 2019), as well as the archaeobotanical analysis of the layer (Hald & Jessen, 2018), indicates that the sediment was washed out of primary context. Radiocarbon-dated hazelnut shells from layer 3 (beta-479655: 6112 ± 96 cal. BC) and layer 4 (beta-479654: 5969 ± 64 cal. BC and beta-479656: 6030 ± 40 cal. BC) confirm the dating of the site to the Kongemosian Villingebæk phase.

Figure 4

Orehoved Sejlrende, site plan and profile. Contour lines represent the depth of the seabed.

Even though the micro-wear analysis suggests that flint from all layers has been redeposited to some degree, the state of preservation of flint artefacts, in particular form the lowest organic layer, was extraordinary, the majority being sharp and unpatinated. The lithic material illustrates a settlement site where the whole range of lithic debitage is represented (Table 2). The majority of the tools are blade tools (Table 2 and Figure 5a), produced on large regular blades made by indirect technique with antler punch, as is typical for the Kongemose culture (Karsten & Knarrström, 2003, pp. 38–49; Sørensen, 2017, p. 38). A large amount of core axes was also excavated (Figure 5b). Three hundred two oblique arrowheads were found at the site. One hundred eighty-five of these are, based on morphological traits, dated to the Villingebæk phase (6150–5800 BC), 44 to the Vedbæk phase (5800–5400 BC), and the remaining were too fragmented to be securely ascribed to type. Furthermore, 280 microburins, deriving from the production of arrowheads were excavated. Fragments of two large pointed flint weapons were also found (Figure 5c). Noteworthy also are the two hammer stones with ground edge facets, which were recovered (Figure 5d and e). Hammer stones with ground edge facets have been found in other Danish Kongemose contexts (Sørensen, 2017, pp. 56–57), and a possible interpretation of use for these stones is retouching, for example, for notches when micro-burin technique was employed for arrowhead making (Sørensen, 2017, pp. 56–57). One intact and three fragmentary pecked stone axes were found at the site. One butt fragment has an initiated hourglass-shaped perforation. The axe is broken through the unfinished hole (Figure 5f and g). A decorated piece of amber with a blind-drilled hole was also found (Figure 5h).

Figure 5

Artefacts from Orehoved Sejrende, (a) blade knives, (b) core axe made on blade core, (c) fragment of large pointed weapon, (d–e) faceted hammer stones, (f–g) butt of round-butted pecked stone axe with initiated hourglass shaped perforation (h) decorated piece of amber with striations and blind drill hole (i) fragment of wooden shaft. Photos: (a–b) Irina Baez-Westerberg, (c–i) Morten Johansen.

Likewise, layer 4 had exceptional preservation conditions for organic material. Organic tools found at the site include a 15 cm long fragment of a broken wooden shaft made of acer (Acer sp.) with a worked knob at the base (Figure 5i), three antler punches, two antler axes, and seven bone bodkins. Species determination of the 3,960 mammal bones (Brodersen, 2019) suggest that red deer (Cervus cervus), wild boar (Sus scrofa), and roe deer (Capreolus capreolus) were the most common game animals. Also present in the assemblage were: beaver (Castor fiber), wolf (Canis lupes), dog (Canis familiaris), fox (Vulpes vulpes), hedgehog (Erinaceus eropaeus), wild cat (Felis silvetris), water vole (Arvicola terrestris), otter (Lutra lutra), and seal (Phoca sp.) (Table 3). The 127 bird bones represent at least 17 taxa with the highest number of fragments from swans (Cygnus sp.) (Table 3). The majority of the identified bird species are freshwater birds (ibid.). Likewise, the fishbone material (3,155 bones and bone fragments) consisted mainly (95%) of freshwater fish (Table 3) (Magnussen, 2019b). The high freshwater indications at the site might at first glance seem unexpected, with considerations of the high salinity otherwise demonstrated in Danish waters during the Atlantic period (Aaris-Sørensen, 1988); moreover, dated sea-level indication points show a fixed link between the Great Belt and Storstrømmen from the onset of the Atlantic period (Astrup, 2018; Binder, 2019). However, the fishbone material from the contemporary Kongemose site Argusgrunden (Figure 1), located 13 km west of Orehoved Sejlrende, is also comprised mainly of freshwater fish (Møhl, 1987), indicating that during the Villingebæk phase, Storstrømmen was still a freshwater system with little or no marine influence.

Two out of the three human teeth found at Orehoved Sejlrende are currently being analysed for of δ¹³C and δ¹⁵N (Anders Fisher, pers. com. March 10, 2021). The stable isotope values may contribute an insight to the composition of the human diet concerning, terrestrial, marine, and freshwater sources.

3.4 Kalverev, Lokalitet X

Another site from the Kongemose culture was located, at the same depth and directly across the sound, 1.7 km north of Orehoved Sejlrende. The site was originally located on an island or peninsula, and the remains of which today is the island of Masnedø Kalv (Figure 2). Gyttja layers at the Kalverev Lokalitet X site from around 5.5 m below the present sea level indicate that the site was located at the very edge of the water during its occupation. The topography of the postglacial terrain suggests that the coastline at the site was relatively straight. Today, the area is overlain by up to 4 m of marine sand and silt, so before excavating with divers, extensive amounts of sediment had to be removed with mechanical excavator. During test excavations in the area, the extent of the site was estimated to be around 8,600 m². Only 21 m² were finally excavated with divers, giving a relatively modest insight into the site (Figure 6). The lithic material excavated at the site consisted of 7,654 pieces of worked or charred flint. Among these were 58 oblique arrowheads, 39 of which could morphologically be ascribed to the Villingebæk phase (6150–5800 BC) and eight to the Vedbæk phase (5800–5400 BC). The lithic assemblage had a close resemblance to the material from Orehoved Sejlrende (Table 2). Likewise, the stratigraphy at the site, with a transgressional gravel layer at a depth of approximately 5–6 m, mirrors the situation at Orehoved Sejlrende. A radiocarbon date of a nutshell and charcoal from the transgressional layer (beta-499650: 5994 ± 68 cal. BC and beta-479651: 5958 ± 59 cal. BC) and a bone fragment from the underlying gyttja layer (beta-479652: 6225 ± 132 cal. BC) place the site in the Villingebæk phase (6150–5800 BC). The 779 mammal bones and bone fragments excavated at the site have not yet undergone species determination. The 474 fish bones found at the site predominantly consist of freshwater species (98%). The dominating species are cyprinids (Cyprinidae) 45%, northern pike (Esox lucius) 27%, and European perch (Perca fluviatilis) 23% (Magnussen, 2019a).

Figure 6

Kalverev, Lokalitet X, site plan and profile. Contour lines represent the depth of the seabed. As several meters of sediment were removed with mechanical excavator prior to diver excavations, the profile only shows the one meter of sediment above moraine till that was left by the excavator.

3.5 Orenæs, Kabel 8

During test excavations, a site from the Ertebølle culture (5400–3900 BC) was found approximately 200 m north of the present coast of Falster, where the small stream Lygemose Å today has its outlet (Figure 2). The site was located only 1 m below the present sea level. The sediments in the area consisted of a complex stratigraphy of organic sediment and several layers of washed out flint material from the settlement site. Stratigraphic and archaeobotanical analyses of the sediment sequence indicate that the area was once a brackish/freshwater lake or lagoon protected from the coast by a barrier. During transgression of the area, the lake/lagoon was gradually more affected by marine intrusion, possibly by periodic flooding of the barrier (Jessen & Hald, 2018). Radiocarbon-dated oak (Quercus sp.) buds from top (beta-509394: 5110 ± 101 cal. BC) and bottom (beta-509395: 5057 ± 148 cal. BC) of the lowest organic layer that shows marine ingression indicate that the area was first transgressed by marine waters between 5100 and 5000 BC.

Because the excavation was development led, it was only possible to excavate within the narrow trench of the cable corridor. The washed-out archaeological material could be followed over a 12 m stretch along the cable route (Figure 7). No settlement surface was preserved. Apart from the material found in the lowermost layer of freshwater gyttja, most likely representing the shallow-water, near-shore refuse layer of the settlement, all archaeological material is relocated and washed out into the area from higher ground during the transgressions of the area. A total of 5,755 pieces of worked or charred flint were found within the 12 excavated square meters, and among these 38 transverse arrowheads, the majority of the Stationsvej type. Trylleskov and Ålekistebro types were also present. Organic material consisted of several sharpened stakes, some driven vertically into the seabed, a leister prong, six bone bodkins, one antler punch, the tip of a slotted bone point with residues of resins preserved within one of the grooves, 1,511 fragments of animal bones, as well as charcoal, burned hazelnut shells, and different seeds. The composition of the fishbone material is of special interest: 492 fish bones were collected during sieving, the majority from the lower, and fresh/brackish deposited, gyttja layer. Despite the deposition environment, all bones derive from saltwater species. The dominating species is cod (Gadidae) 91%, but plaice/flounder/dab (P. platessa/P. flesus/L. limanda) and turbot/brill (Psetta maxima/Scophthalmus rhombus), garfish (Belone belone), and herring (Clupea harengus) were also present (Magnussen, 2018).

Figure 7

Orenæs, Kabel 8, site plan and profile.