Sea level changes and Neolithic hunter-fisher-gatherers in the centre of Tallinn,southern coast of the Gulf of Finland,Baltic Sea

Abstract

Relative sea level (RSL) changes and the palaeogeography of a Neolithic hunter-fisher-gatherer settlement site on the former shore of the Gulf of Finland in the city centre of Tallinn were reconstructed by implementing GIS in landscape modelling based on archaeological, sedimentary and shore displacement data. AMS radiocarbon dating of mammal bones from the cultural layer suggests the existence of the hunter-fisher-gatherer settlement around 5.1–4.8 cal. ka BP on a seaward inclining sandy beach of Tallinn palaeo-bay c. 100 m from the Litorina Sea shoreline and at about 2.4 m above the coeval sea level. The shoreline passed the study site at about 5.8 cal. ka BP and retreated towards northeast with an average speed of 13 m per century, while the RSL lowered by c. 2.5 mm annually. Combining radiocarbon dates of terrestrial and marine mammal bones from the Neolithic cultural layer, a marine reservoir effect of 350 ¹⁴C years for the brackish-water Baltic Sea was calculated. By using high-resolution archaeological data in combination with RSL and other geological proxies, we demonstrate new possibilities to reconstruct the palaeoenvironment of deeply buried coastal settlement sites and to predict a possible continuation of the cultural layer in heavily built-up areas.

Keywords

Baltic Sea landscape modelling Neolithic settlement palaeogeographic reconstruction relative sea level shoreline change

Introduction

In early-Holocene times, remarkable changes in the Baltic Sea ecosystem took place when the freshwater Ancylus Lake occupying the Baltic Sea basin drained into the ocean, allowing saline water to enter. The earliest evidence of saline water ingress is documented from the Blekinge archipelago (Berglund et al., 2005) and from the Bornholm basin (Andrén et al., 2000) in the southern Baltic Sea at about 9.8 cal. ka BP (Andrén et al., 2011). Around 8.5 cal. ka BP, the Baltic Sea became brackish and a marine phase called the Litorina Sea started with significantly increased primary production (Andrén et al., 2011), establishing favourable conditions for the exploitation of marine resources by people (Jöns, 2011; Kriiska and Lõugas, 2009). Transformation in the subsistence strategy of prehistoric humans was reflected in the pattern of settlement sites, which were then established on the seashores (e.g. Andersen, 1993a; Christensen, 1995; Kriiska, 2000; Larsson, 1997). These settlements are presently located at different elevations because of differential glacio-isostatic land uplift and eustatic sea level rise. In the southern Baltic Sea area, a number of Stone Age coastal settlement sites have been found lying in the present seabed because of Holocene sea level rise (Fischer, 2011; Lübke et al., 2011). In contrast, in the northern Baltic Sea areas, prehistoric coastal settlements have been uplifted because of post-glacial rebound: the older the site, the higher the altitude (Siiriäinen, 1982; Vaneeckhout, 2008). In transitional areas like the Gulf of Finland region, Mesolithic and Neolithic coastal settlements have also been uplifted and are presently located inland; however, Mesolithic sites are often buried under marine sediments because of mid-Holocene sea level rise, which temporarily exceeded the rate of the land uplift in this region (Gerasimov et al., 2010; Jussila and Kriiska, 2004; Rosentau et al., 2011; Veski et al., 2005). An important shift in subsistence strategy and a change in settlement pattern were related to the introduction of agriculture and animal husbandry in the Baltic Sea region around 6.0 cal. ka BP (e.g. Ahlfont et al., 1995; Andersen, 1993b). In many locations, settlements were moved to arable farmlands away from the coast (Jöns 2011). However, in the eastern Baltic Sea area, the process seems to be different as a large number of settlements were still established on the seashores. This is reflected in the finds of fishing and hunting tools and large numbers of bones of marine mammals together with Comb Ware pottery and the absence of farming indicators in Middle Neolithic cultural layers (Bērziņš, 2008; Gerasimov et al., 2010; Kriiska, 2003a; Rosentau et al., 2013; Veski et al., 2005). Our study site at Vabaduse Square in the centre of Tallinn on the southern coast of the Gulf of Finland shows a similar find assemblage (Kadakas et al., 2009, 2010; Figure 1).

Figure 1.

(a) Overview map with the location of the study area, comparable Stone Age coastal settlement sites in the eastern Baltic Sea region and present-day apparent land uplift isobases (Ekman, 1996); (b) distribution of Quaternary deposits in Tallinn area together with relative sea level isobases and the location of shoreline at c. 5.1 cal. ka BP (after Saarse et al. 2003 and Saarse and Vassiljev, 2010); and (c) location of Vabaduse Square excavation plot (red polygon) marked on the orthophoto (Estonian Land Board 2015).

During large-scale salvage excavations in 2008 and 2009, a large number of Middle and Late Neolithic artefacts and ecofacts including fishing hooks, harpoons, bones of marine and terrestrial fauna and Comb Ware pottery with a few Corded Ware sherds were unexpectedly found under 2.5-m-thick anthropogenic soil at Vabaduse Square. Archaeological (Kadakas, 2010; Kadakas et al., 2010) and osteological material (Lõugas and Tomek, 2013) from the settlement site has been studied in detail, suggesting the proximity of the seashore during the occupation. However, the palaeogeography, including relative sea level (RSL) and palaeoshoreline positions, has remained unclear in this infilled and heavily built-up urban area (Figure 1c).

This study aims to reconstruct palaeotopography and sea level changes of the Vabaduse Square Neolithic settlement site by implementing GIS in landscape modelling and using high resolution excavation data together with sedimentary and RSL proxies. We contribute to earlier knowledge by providing new radiocarbon dates and palaeogeographic and shore displacement reconstructions for the Tallinn area. The study sheds light on the use of coastal areas and the choice of settlements by specialized Neolithic hunter-fisher-gatherers with the distinctive maritime economy common to the eastern Baltic Sea region.

Archaeology of the Stone Age settlement at Vabaduse Square

The remains of the Neolithic settlement site were found at c. 15–16.5 m a.s.l. at Vabaduse Square in the centre of Tallinn, the capital of Estonia (59.4331°N, 24.7438°E). The Neolithic cultural layer of the site was mainly 20–30 cm and, in a few locations, up to 50 cm thick and was spread over an area of 2200 m² at least (Figure 1c). The Stone Age cultural layer was a former sandy land surface, which was covered by approximately 2- to 2.5-m-thick historical suburban soil. The locations of artefacts and faunal remains of the Neolithic settlement site indicate that the cultural layer was situated in its original place. From the excavated area, no certain remains of Neolithic built constructions were found. The only clear structures were five dark brown spots of sand with slightly rounded or conical bottoms that might be marks of pits dug by humans. The measures of the pits were generally between 1.05–1.2 m in width, 1.15–1.5 m in length, and the depth reached 25–35 cm. Whether they were fire places or depressions needed for household activities was not possible to determine (Kadakas et al., 2010).

The cultural layer was rather sparse in findings and the variability of the find collection was low. Out of the Stone Age assemblage of 2010 fragments, 875 pieces were pottery, 95% of which belonged to Late Comb Ware and 2% to Corded Ware ceramics (Kadakas, 2010; Kadakas et al., 2010). Rock and mineral (quartz) artefacts numbered 1099 altogether, the majority of which had been worked in the flaking technique using bipolar reduction (Kadakas, 2010). The scarce Comb Ware material had limited variation, suggesting that it was produced during rather a short period of time, for example, a few human generations (Kadakas, 2010). In total, 24 specimens of bone artefacts and bone fragments with working traces, mostly hunting and fishing tools, were found (Kadakas, 2010; Kadakas et al., 2010).

Osteological material from the Stone Age layer indicates the vicinity of the seashore. More than 3000 pieces of various faunal remains were collected. Most of the identifiable remains belonged to sea mammals: 16% to porpoise Phocoena phocoena and 44% to seals Phocidae (harp seal Phoca groenlandica 70% and ringed seal Phoca hispida 30% of seal bones). The number of terrestrial mammals and birds was also relatively large (Lõugas and Tomek, 2013). The osteological composition is similar to other mainland coastal sites dated from 6.0 to 4.5 cal. ka BP in the eastern Baltic Sea region, where seals with additions from the terrestrial environment like elk (Alces alces), wild boar (Sus scrofa) and beaver (Castor fiber) were the most hunted mammals. The finds of avian fauna were dominated by water fowls (e.g. Anatidae and Anseridae), and those of fishes by cod (Gadus morhua) from marine waters and pike (Esox lucius) and perch (Perca fluviatilis) from fresh/brackish water bodies (Lõugas and Tomek, 2013). The material from Vabaduse Square is distinguished from other Neolithic coastal sites in the region by its relatively high number of porpoise bones (see Figure 1a for sites with porpoise bone finds). According to the analysis of faunal remains, the area was most probably used as hunting and fishing camp in the marginal area of different biotypes (viz. sea, land and freshwater). The best time for hunting porpoise and harp seal was probably the summer/early autumn, when they came to feed in the bay (Lõugas, 1997, 1999; Lõugas and Tomek, 2013); however, the other animal and bird species do not allow us to delimit the use of the settlement to a particular period of the year.

Four or five more intense areas of use (finds of pottery, quartz, bone) were identified with rather empty areas in between. These concentration areas of finds might refer to camps that were set up in different years. Through analysing the distribution of quartz and pottery, it was possible to identify areas of different usage (e.g. quartz processing areas, cooking areas) in each of these supposed camp sites (Kadakas, 2010).

Material and methods

Sediment description

A sediment sequence was described in detail on site and sampled for loss-on-ignition (LOI), diatom and pollen analyses at the southwest part of the Vabaduse Square (Figures 1c and 5a). The distribution and thickness of the Neolithic cultural layer were mapped during the archaeological excavations. Altogether, four samples from the Neolithic cultural layer and three from the underlying deposit (units B and A in Figure 2) were analysed. The standard method of LOI according to Heiri et al. (2001) was used. Diatom samples were prepared using standard methods by Battarbee et al. (2001). The identification used was based on Krammer and Lange-Bertalot (1988, 1999–2004). The samples for pollen analysis were processed following the standard laboratory technique (Berglund and Ralska-Jasiewiczowa, 1986) with the additional use of a heavy liquid treatment (CdJ2 and KJ solution with the gravity of 2.2 g/cm³) to remove minerogenic material.

Figure 2.

Sediment stratigraphy from the Vabaduse Square excavation plot with five identified sedimentary units (A–E) and radiocarbon dates from the terrestrial mammal bones from the Neolithic cultural layer. The location of the sampling site is shown in Figure 5.

Radiocarbon chronology and shore displacement reconstruction

For the chronology, three AMS radiocarbon dates from published sources (Lõugas and Tomek, 2013) and two new dates were used to determine the most likely median age of the settlement site (Table 1). Three of the dates came from terrestrial animals (i.e. a wild boar and two elk bones) and two from marine mammals (harp seal). The bones had artificial scratches that demonstrated relations to humans.

Table 1.

AMS radiocarbon dates of the osteological material found from the Vabaduse Square Neolithic cultural layer. The radiocarbon ages of seal bones are calibrated to calendar years after subtraction of the calculated reservoir offset of 347 ± 73 ¹⁴C years.

Species	¹⁴C age	¹⁴C age corrected to R(t)	Cal. BP (95% intcal13)	Median	Sample code	Reference
Sus scrofa	4450 ± 40		5290–4880	5100	Poz-35401	Lõugas and Tomek (2013)
Alces alces	4240 ± 30		4860–4650	4830	Beta-409161	This study
Alces alces	4340 ± 30		4970–4850	4910	Beta-409162	This study
Phoca groenlandica	4630 ± 40	4283 ± 113	5280–4530	4860	Hela-1922	Lõugas and Tomek (2013)
Phoca groenlandica	4750 ± 40	4403 ± 113	5440–4660	5040	Hela-1923	Lõugas and Tomek (2013)

Based on the comparison of the ages of the marine mammals and the terrestrial mammals, an estimation of the reservoir effect was given as the difference between the mean ¹⁴C age of the marine mammals and that of the terrestrial mammals.

Both AMS and the conventional radiocarbon dates from regional geological sites (lakes and mires; Table 2) from approximately the same Litorina Sea isobase as the site itself and water level surfaces derived from Estonian coastal formation database (Saarse et al., 2003, 2007) were used for the construction of the shore displacement curve.

Table 2.

Radiocarbon dates and modelled isolation ages used for shore displacement reconstruction of Tallinn area. The location of the sites is shown in Figure 3b.

Site	Laboratory code	¹⁴C age (yr BP)	Calibrated ages (cal. yr BP)			Elevation (m a.s.l.)	Dated material/modelled event	Position to Ancylus Lake/sea level	References
Site	Laboratory code	¹⁴C age (yr BP)	95.4% probability range	Weighted average	Median	Elevation (m a.s.l.)	Dated material/modelled event	Position to Ancylus Lake/sea level	References
Aabla	Poz-33490	7280 ± 50	8190−8000	8090 ± 50	8100	18.75	Wood	Above	Saarse et al. (2009)
Jõelähtme	Tln-2349	9980 ± 120	11,970−11,200	11,530 ± 210	11,510	28.45 to 28.50	Peat	Above	Heinsalu (2000)
	Tln-2371	9640 ± 100	11,230−10,710	10,970 ± 150	10,970	28.40 to 28.45	Peat	Above	Heinsalu (2000)
	Tln-2370	9330 ± 90	10,750−10,260	10,530 ± 140	10,530	28.50 to 28.55	Peat	Above	Heinsalu (2000)
L. Maardu	Ua-2390	9490 ± 110	11,170−10,510	10,820 ± 190	10,810	25.75	Wood	Above	Veski, (1998)
L. Ülemiste	Tln-1859	9480 ± 95	11,130−10,510	10,810 ± 180	10,780	28.30	Peaty gyttja	Above	Saarse et al. (1997)
	Tln-1856	9300 ± 80	10,670−10,260	10,490 ± 120	10,490	28.40	Peaty gyttja	Ancylus Lake level	Saarse et al. (1997)
	Tln-1861	9145 ± 75	10,510−10,200	10,338 ± 90	10,330	28.90	Peaty gyttja	Below	Saarse et al. (1997)
Lepistiku	Tln-2504	3780 ± 50	4380−3990	4160 ± 90	4160	16.37	Peat	Above	Saarse et al. (2001)
Kroodi	Tln-2668	7730 ± 80	8700−8380	8520 ± 90	8510	19.50	Peat	Above	Saarse et al. (2003)
L. Harku	Poz-51453	1190 ± 30	1180−1010	1110 ± 40	1110	−1.55 to −1.60	Plant remains	Below	Grudzinska et al. (2014)
	Poz-51454	1270 ± 30	1290−1150	1220 ± 30	1220	−1.75 to −1.80	Plant remains	Below	Grudzinska et al. (2014)
	Poz-49185	1900 ± 40	1920−1730	1830 ± 50	1840	−3.04	Wood	Below	Grudzinska et al. (2014)
	Modelled		1110−970	1040 ± 70	1040	2.50	Isolation	Sea level	Grudzinska et al. (2014)
L. Klooga	Poz-42168	3760 ± 40	4240−3990	4120 ± 60	4120	8.40 to 8.45	Plant remains	Below	Grudzinska et al. (2013)
	Poz-42169	3840 ± 50	4420−4140	4260 ± 80	4260	7.85 to 7.90	Plant remains	Below	Grudzinska et al. (2013)
	Modelled		4290−4070	4180 ± 50	4180	11.80	Isolation	Sea level	Grudzinska et al. (2013)
L. Lohja	Poz-42171	2280 ± 30	2360−2160	2270 ± 60	2310	1.80 to 1.85	Pinus bark	Below	Grudzinska et al. (2013)
	Poz-42172	2490 ± 40	2730−2440	2580 ± 80	2580	1.55	Tilia wood	Below	Grudzinska et al. (2013)
	Modelled		2330−2110	2230 ± 70	2240	6.00	Isolation	Sea level	Grudzinska et al. (2013)
L. Käsmu	Poz-42177	1830 ± 30	1860−1630	1770 ± 40	1770	−0.26	Pinus bark	Below	Grudzinska et al. (2013)
	Poz-42166	1910 ± 30	1930−1780	1860 ± 30	1860	−0.43	Pinus wood	Below	Grudzinska et al. (2013)
	Modelled		1910−1780	1840 ± 30	1840	4.50	Isolation	Sea level	Grudzinska et al., 2013
L. Tänavjärv	Tln-3306	4490 ± 70	5320−4910	5130 ± 100	5130	15.19 to 15.24	Gyttja	Below	Grudzinska et al. (2013)
	Tln-3305	4600 ± 100	5460−5050	5260 ± 110	5280	15.16 to 15.19	Silty gyttja	Below	Grudzinska et al. (2013)
	Poz-42173	4930 ± 40	5730−5590	5660 ± 40	5650	14.93 to 14.98	Woody pieces	Below	Grudzinska et al. (2013)
	Modelled		565−5160	5410 ± 130	5420	17.40	Isolation	Sea level	Grudzinska et al. (2013)
Vääna	Tln-3036	5260 ± 60	6110−5900	5970 ± 60	5970	18.25 to 18.29	Peat	Above	Saarse et al. (2009)
	Tln-3037	5210 ± 60	6180−5960	6080 ± 70	6090	18.2 to 18.25	Peat	Above	Saarse et al. (2009)
	Poz-24245	6240 ± 40	7270−7040	7190 ± 60	7210	18.10	Plant remains	Below	Saarse et al. (2009)
	Poz-24267	7420 ± 40	8340−8170	8250 ± 50	8240	17.91 to 17.92	Plant remains	Below	Saarse et al. (2009)
	Poz-24246	8040 ± 50	9080−8730	8910 ± 90	8910	17.76 to 17.77	Gyttja	Below	Saarse et al. (2009)
	Modelled		9200−8780	8990 ± 110	8990	18.20	Lagoon onset	Sea level	Saarse et al. (2009)
	Modelled		7020−6600	6810 ± 110	6810	18.20	Isolation	Sea level	Saarse et al. (2009)
	Modelled		6300−6000	6150 ± 80	6150	18.20	Peat accumulation	Above	Saarse et al. (2009)

All radiocarbon dates were calibrated using the OxCal program (Ramsey, 2009) with the IntCal13 calibration curve (Reimer et al., 2013).

Palaeogeographic reconstructions

To enable an overall view of the formation of the Vabaduse Square area and its surroundings, the palaeogeography was reconstructed based on the shore displacement curve and a digital elevation model (DEM) of Tallinn centre with the present absolute elevation of the base of the medieval cultural layer. For creating the DEM, the data of pre-medieval surface elevations from Zobel (2009) and the data measured during the Vabaduse Square excavations were used. The resolution of the DEM was 2 m. RSL elevation of relevant time slices of the Stone Age period was taken from the shore displacement curve considering a linear regression between the isolation dates of Vääna lagoon and Lake Klooga (Figure 3). Palaeo-land surfaces were modelled by subtracting the palaeo-sea level elevation value from the DEM of the pre-medieval surface. Due to the small extent of the modelled area, the RSL differences resulting from the locational differences in glacio-isostatic land uplift were insignificant and were therefore neglected.

Figure 3.

(a) Holocene shore displacement curve with relative shore level indicators for Tallinn area in relation to radiocarbon-dated Vabaduse Square Neolithic cultural layer. A full list of radiocarbon dates and modelled isolation ages is given in Table 2. Shore mark elevations according to Saarse et al. (2003). Baltic Sea stage boundaries according to Andrén et al. (2011). (b) Overview map with the location of the sampling sites and isobases of the highest shoreline of the Litorina Sea.

Additionally, a detailed DEM of the Stone Age excavation plot with a resolution of 0.2 m was generated. The elevation data originated from tacheometry and levelling measurements during the excavations (Kadakas, 2010). The locations of the finds and disturbed parts of Stone Age cultural layer were marked. Based on the shore displacement curve, the elevations of the detailed DEM were recalculated to correspond to 5.1 and 4.8 cal. ka BP (contemporary to the median ages of the bones from the site). This enabled a detailed description of the microtopography of the settlement site. Distances from the findings to the shoreline and the relative height of the settlement area were determined.

Results

Sediment stratigraphy and chronology of the settlement

The stratigraphy of the Vabaduse Square site comprises five sedimentary units (A–E) presented in Figure 2.

The landward-dipping (16° SW) sandy deposits below the Neolithic cultural layer form the lowermost unit A. This sandy deposit contains ripples and interbeds of moderately to well-rounded gravel and pebbles and is interpreted as a coastal deposit formed in an active wave regime around mean sea level (Cooper, 2007; Orford et al., 1991).

The following up to 50-cm-thick (typically 20–30 cm) Neolithic cultural layer (unit B) consists of reworked fine-to-medium grained sand containing fragments of potsherds, quartz tools and production residue, bones and charcoal (a description of the finds is presented in the section ‘Archaeology of the Stone Age settlement at Vabaduse Square’). The elevation of the upper surface of the Neolithic cultural layer decreases continuously in the seaward direction, being 16.36 m a.s.l. in the southwest part of the study area and 15.05 m a.s.l. in the northeast part. The sediments of unit B are relatively poor in organic matter (<3%) and in some places contain secondary carbonates (tubes), most probably associated with roots. The lower section of the Neolithic cultural layer contains relatively poorly rounded gravel and pebbles (unit B1), while in the upper part (typically 5–10 cm) of the unit (B2), this coarse-grained material is missing. In a few places in the excavation plot, horizontal bedding structures are visible within the sediments of unit B1. However, most of the sediment of unit B is reworked by roots and by the activities of the Neolithic/Medieval humans. The contact with the underlying undisturbed sandy deposits of unit A is undulating and probably marks the reworking depth. According to sedimentary evidence, the Neolithic cultural layer (unit B) is interpreted as a reworked coastal deposit.

The Neolithic cultural layer and underlying deposits (units A and B) are extremely poor in diatoms and pollen, probably due to bad preservation conditions in a sandy environment. Only one cell of Hantzschia amphioxys and a few unidentified diatom fragments were encountered from the sediments of unit B. H. amphioxys is usually found in or on the ground and some populations in fresh or coastal waters. No diatoms characteristic of the Litorina Sea were found. Pollen samples from the Neolithic cultural layer (unit B) contain a few pollen grains of Chenopodiaceae, Asteraceae, Poaceae and a single Pinus. A high concentration of microscopic charcoal particles and several specimens of Acritarch occur in all analysed samples. Due to the lack of representative data, it was not possible to draw any specific conclusions on the basis of diatoms and pollen.

The median ages of the two elk and the one wild boar samples from unit B suggest the accumulation of animal bones and therefore habitation of the settlement around 5.1–4.8 cal. ka BP (Table 1). The clear offset between the terrestrial and marine samples indicates the marine reservoir effect of the seal bones. The calculated reservoir effect, considering the difference between the mean ¹⁴C age of the marine mammals and the mean ¹⁴C age of the terrestrial mammals, is about 350 radiocarbon years (347 ± 73 ¹⁴C years; see Table 1). The radiocarbon ages of the seal bones were calibrated to calendar years after subtraction of the estimated reservoir effect rate. The median values of 5.0 and 4.9 cal. ka BP were received.

The Neolithic cultural layer is covered by a 70-cm-thick organic-rich sandy medieval cultural layer (unit C) followed by two landfill layers (units D and E). The contact between the Neolithic and the medieval cultural layer is sharp and undulating.

Shore displacement curve for Tallinn area

A curve showing shore displacement in the Vabaduse Square area is presented in Figure 3. According to the water level surfaces derived from the Estonian coastal formation database (Saarse et al., 2003, 2007), the water level was 76–78 m a.s.l. during the Baltic Ice Lake (c. 11.7 cal. ka BP; Andren et al., 2001), 34–36 m a.s.l. during the Ancylus Lake transgression maximum (c. 10.5–9.8 cal. ka BP; Table 2) and 20–22 m a.s.l. during the Litorina Sea transgression maximum (c. 8.1–7.2 cal. ka BP; Table 2) in the study area. The water level before the Ancylus Lake transgression is not well known; however, according to the wood remains from the bottom of Lake Maardu and peat/peaty gyttja accumulation at Jõelähtme (Veski, 1998) and Lake Ülemiste (Saarse et al., 1997), it must have been lower than 25 m a.s.l. (Figure 3; Table 2). The timing and the changes in the water level during the Litorina transgression were determined according to bio- and chronostratigraphic records from Vääna lagoon (Saarse et al., 2009), which lies at a similar Litorina Sea isobase as Vabaduse Square in Tallinn (Figure 3b). The first appearance of the diatom species Mastogloia baltica, indicating the start of the marine phase at Vääna, is dated to c. 9.0 cal. ka BP, suggesting that the sea level was near the threshold level of the lagoon (i.e. 18.2 m a.s.l.) at that time. The RSL rise was slow and accelerated at about 8.1 cal. ka BP. The Kroodi peat (Saarse and Vassiljev, 2010) and the Aabla coastal lake (Saarse et al., 2010) suggest that the transgression maximum occurred after 8.1 cal. ka BP (Figure 3) and the shore marks (Saarse et al., 2003, 2007) indicate that the maximum sea level reached 21–22 m a.s.l. The exact timing of the maximum Litorina Sea level is not well defined, but bio- and chronostratigraphic evidence suggests that the sea level dropped below the threshold of Vääna lagoon at about 6.8 cal. ka BP, and continuous peat accumulation started there at c. 6.2 cal. ka BP. Recent studies of lake isolation in northwest Estonia (lakes Harku, Käsmu, Lohja, Klooga, Tänavjärv; Figure 3) indicate that the RSL decreased quasi-linearly after the Litorina Sea transgression maximum (Grudzinska et al., 2013, 2014).

Palaeogeography of the settlement

Palaeogeographic reconstructions of central Tallinn around Vabaduse Square are presented in Figure 4 for five time slices. These time slices were chosen in order to visualize shoreline changes since the formation of the transgression maximum shoreline of the Litorina Sea. During the formation of the highest shoreline of the Litorina Sea (according to the Vääna sequence), our study area (presently at 15.5–16.5 m a.s.l.) was submerged to a depth of about 5.5–6.5 m b.s.l. At that time, a narrow peninsula connecting Toompea bedrock hillock to the mainland was formed (Figure 1b), and our study area was located c. 180 m offshore (Figure 4a). From the start of the isolation of the Vääna lagoon at about 6.8 cal. ka BP until the isolation of Lake Klooga at about 4.2 cal. ka BP, the RSL lowering was modelled as a linear decrease of 2.5 mm per year. The modelling results show that the shoreline passed the settlement site area around 5.8 cal. ka BP (Figure 4b) and was at a distance of around 100 m at the dates from the cultural layer (5.1–4.8 cal. ka BP; Figure 4c and d). At that time, the settlement site was located on the shore of the open bay in the wind and wave shadow by the up to 28 m high Toompea bedrock hillock. Later, the shoreline continued retreating, and the Middle Neolithic hunter-fisher-gatherers probably abandoned the settlement. At about 4.2 cal. ka BP (Figure 4e), the settlement site area was already c. 230 m away from the seashore. The shoreline continued to retreat towards the north and northeast by about 13 m in 100 years.

Figure 4.

Palaeogeographic reconstructions of the centre of Tallinn. (a) The highest shoreline of the Litorina Sea: the most of central Tallinn is inundated, only the Toompea bedrock hillock is above the water. The shoreline retreats towards the north and northeast. (b) 5.8 cal. ka BP: the shoreline of the Litorina Sea crosses the location of the Stone Age site. (c) Possible earliest time of habitation c. 5.1 cal. ka BP: the site was located close to the seashore in the beach zone, protected from northerly winds by the Toompea bedrock hillock and a developing peninsula in the north. (d) 4.8 cal. ka BP: the shoreline retreats from the site. (e) 4.2 cal. ka BP: the location of the Stone Age site is out of the beach zone (c. 230 m away from the sea), the shoreline continues to retreat towards the north and northeast.

The detailed reconstruction of the Stone Age excavation area emphasizes that during its oldest phase of existence, the settlement was located 1.5–3.5 m above coeval average sea level (Figure 5). In the excavation area, the Neolithic land surface dips, and the concentration of findings decreases continuously towards the past seashore. The findings are located most densely in the flattest (gradient of <1°) section right inland from the steepest (gradient of 1.5°−2.0°) section of the seaward-dipping beach. The concentration areas form a shore-parallel row at a semi-constant elevation of 2 m above coeval sea level at 5.1 cal. ka BP and 2.7 m at 4.8 cal. ka BP located 80 and 120 m from the modelled shoreline positions, respectively. The dated bones were found from the same section.

Figure 5.

(a) Detailed palaeotopographic reconstruction of the Vabaduse Square Neolithic settlement site at c. 5.1 cal. ka BP, (b) 3D image of the surroundings of settlement site at c. 4.8 cal. ka BP with present street network of Tallinn centre, (c) profile illustrating the elevation and distance of the settlement from its coeval shoreline, (d) fragments of Comb Ware (AI 6917:527; left) and Corded Ware (AI 6917:376) ceramics and (e) bone artefacts (AI 6917:821, AI 6917:782, AI 6917:401) and (f) quartz tools (AI 6917:40, AI 6917:862, AI 6917:251) found from the Neolithic cultural layer. Photographs and locations of finds, disturbed parts of Stone Age layer and man-made pits according to Kadakas (2010).

Discussion

In the eastern Baltic Sea area, the beginning of maritime lifestyle and the selection of settlement sites on the coast started during the second half of the Mesolithic period at about 9.8–8.5 cal. ka BP (Gerasimov et al., 2010; Jöns, 2011; Kriiska, 2001) and continued in many areas also during Neolithic (Comb Ware) times. Preferred settlement locations were associated with former river mouths (Nuñez and Okkonen, 1999; Rosentau et al., 2011), major palaeo-lagoonal systems (Rosentau et al., 2013), sheltered bays and archipelago systems (Kriiska, 2003a), coastal lakes (Bērziņš, 2008; Loze, 2006) and only on a few occasions with shores of open bays (Carpelan et al., 2008; Kriiska, 2003a, 2007; Figure 1a) as in the case of the Vabaduse Square site (Figure 1b). Our palaeogeographic reconstructions suggest the existence of the hunter-fisher-gatherer site at around 5.1–4.8 cal. ka BP on a seaward-inclining sandy beach of the large Tallinn palaeo-bay at about 100 m from its coeval shoreline and c. 2.4 m above its coeval sea level. The settlement was well protected against the prevailing westerly storms and wind-induced waves by Toompea bedrock hillock. Modelling shows that the average RSL fall rate was at that time 2.5 mm per year and that the shoreline shifted continuously towards the northeast at an average speed of c. 13 m per century (Figures 3 and 4). The settlement was probably established higher than wave run-up. In the sheltered Tallinn bay with negligible tides, the sea level has been recorded to rise maximum +1.5 m during storm surges (Post and Kõuts, 2014), enabling the coastal settlement to be only a few 10s m inland from its coeval shoreline, even for a gentle beach slope. As the finds appear to be in situ and no Neolithic artefacts are found closer than 50 m from the 5.1 cal. ka BP shoreline (Figure 5), it can be presumed that mostly the wave run-up did not reach farther than 50 m from the mean shoreline and that people founded the seasonal settlement directly on the beach as close to the shoreline as possible. Compared with other Stone Age sites in the eastern Baltic Sea region, this kind of setting is common in all types of coastal settlements (Jussila and Kriiska, 2004) and is reasoned by maritime economics.

This settlement pattern probably reflects a specific socio-economic strategy of Neolithic hunter-fisher-gatherers common to the region. In the open bay sites, unlike those associated with lagoons, no dwelling depressions and in some cases (Karhusuo and Johanneksen asema sites, Carpelan et al., 2008; Figure 1a) even no ceramics have been found. Similarly, in the Vabaduse Square site, the number of finds is relatively small and there are no visible constructions of dwellings despite the large settlement area. This, together with the finds of Comb Ware pottery, lets us conclude that the Vabaduse Square site was used as seasonally revisited coastal hunting and fishing camp in Middle Neolithic times. However, the few Corded Ware pottery fragments suggest that it was also re-exploited to some extent by Late Neolithic farmers, when the seashore had retreated far from the site. Based on palaeogeographic reconstructions, the site was located at c. 4.2 m a.s.l. and c. 230 m from the seashore during the Late Neolithic period at about 4.2 cal. ka BP (Figure 4e). A similar pattern of re-habitation is also recognized in other settlement sites in the eastern Baltic Sea region (Kriiska 2003b; Kriiska et al., 2014).

As indicated by the faunal remains and the reconstructed palaeogeography, the settlement site presented a good location for seal and porpoise hunting trips into the bay and open sea. The harp seal, which lives in the North Atlantic and the Arctic Ocean today, used to come to feed and probably formed an independent population in the Baltic Sea at that time (Ukkonen, 2001) and was an important game in several locations, but mostly on islands. The number of Neolithic sites with porpoise – which still inhabits the Baltic in low numbers today – bones is limited on the eastern coast of the Baltic Sea (Figure 1a). Two sites with porpoise bones have been found in Finland (Tuorsniemi and Jokiniemi Sandliden), eight in Estonia (Naakamäe and Loona in Saaremaa Island, Kõpu XI in Hiiumaa Island and Riigiküla I, III, IV, Neolithic context of Kunda Lammasmägi and Vabaduse Square in NE Estonia), one in Latvia (Siliņupe) and two (Šventoji 2B and 3) in Lithuania (Daugnora and Girininkas, 2004; Leskinen and Pesonen, 2008; Lõugas, 1999; Seitsonen, 2008; Ukkonen, 2001; Zagorska, 2000; Figure 1a). Among the faunal remains of the Vabaduse Square site, terrestrial and freshwater game species, like elk and beaver, occur besides marine species, reflecting the fact that the site was located at the border of different biotopes: sea, land and freshwater (Lõugas and Tomek, 2013). During the excavations, it was not possible to identify any buried valleys or other indicators of former fluvial activity in the study area; however, these could have existed outside of the excavation plot. The nearest known freshwater body was the formerly existing Härjapea River at about 2 km east of Vabaduse Square (Figure 1b), which flowed out of Lake Ülemiste and broke through the limestone plateau down to the sea.

The good selection of terrestrial and marine mammal bones in the Vabaduse Square Neolithic cultural layer enabled the calculation of marine reservoir effect of 350 ¹⁴C years (Table 1). This is in agreement with available reservoir effect estimations from mid-Holocene subfossil seal bones from the Baltic Sea. Lindqvist and Possnert (1997) estimated 300 ¹⁴C years of reservoir effect for the mid-Holocene, Olsson (1980) suggested values of c. 270–350 ¹⁴C years for 19th- and 20th-century seal bones, while in Piličiauskas and Heron (2015), the reservoir effect values yielded only 190 ¹⁴C years. Ukkonen et al. (2014), summarizing previous studies, considered average reservoir ages of 300–400 ¹⁴C years as well justified for seal bones from the Baltic Sea throughout the Holocene.

The use of the high-resolution archaeological data in combination with shore level and other proxies in landscape modelling provided the possibility to predict the continuation of the cultural layer in the heavily built-up area in the centre of Tallinn (Figure 1c). Although the Vabaduse Square site is one of the largest Stone Age excavation plots studied in the eastern Baltic Sea region in the last few decades, the settlement site itself was probably larger. As seen from Figure 5a, the Stone Age cultural layer has been disturbed or removed by later building and road construction in many sections. Two of the concentration areas of finds are bordered by areas with a disturbed cultural layer, which means that the concentration areas were probably originally larger. In mid-south part, dense findings continue until the border of the excavation area, and probably continue farther in that direction (not excavated because of the presence of a road). As the concentration of finds decreases towards the former shoreline until the most seaward finds at c. 50 m from the shoreline of 5.1 cal. ka BP, it is most likely that the settlement continues to the southeast, parallel to the modelled shoreline, and does not spread farther seaward.

Conclusion

We presented palaeogeographic reconstructions for one of the largest Neolithic hunter-fisher-gatherer settlement sites studied in eastern Baltic Sea Region in the last few decades. We conclude that during its habitation at around 5.1–4.8 cal. ka BP, the site was located in the beach zone of a large open bay of the past Baltic Sea. By combining RSL, sedimentary and archaeological excavation data and with the implementation of GIS, we were able to reveal palaeotopography beneath an up to 2.5-m-thick layer of younger anthropogenic sediments of the densely built-up city centre, and to reconstruct the Neolithic seaward-dipping sandy palaeo-land surface of the site. The modelling results show that the study area was isostatically uplifted from the Litorina Sea at about 5.8 cal. ka BP, and the settlement site was located at c. 2.4 m above coeval sea level and at c. 100 m from the shoreline at about 5.1 cal. ka BP. The modelling results helped us to characterize prehistoric people’s selection of seasonal hunting and fishing camps on the beach of the open bay and to predict a possible continuation of the cultural layer. The knowledge acquired from this study provides a better prediction of the locations and discoveries of similar Neolithic open seashore sites, which is a rare type of settlement according to what is currently known about the Stone Age in the eastern Baltic Sea region. The results of this study verify the good potential of the application of detailed archaeological excavation data in GIS-based palaeogeographic reconstructions.

Footnotes

Acknowledgements

We thank two anonymous reviewers for valuable remarks and suggestions that enabled us to improve the manuscript.

Funding

This work was supported by the Estonian Science Foundation (grant number ETF9011) and by the Estonian Research Council (grant numbers IUT20-7 and PUT456).

References

Ahlfont

Guinard

Gustafson

. (1995) Patterns of Neolithic farming in Sweden. TOR 27(1): 133–184.

Andersen

(1993a) Mesolithic coastal settlement. In: Hvass

Storgaard

(eds) Digging into the Past: 25 Years of Archaeology in Denmark. Aarhus: Aarhus University Press, pp. 65–69.

Andersen

(1993b) Early agriculture. In: Hvass

Storgaard

(eds) Digging into the Past: 25 Years of Archaeology in Denmark. Aarhus: Aarhus University Press, pp. 88–91.

Andrén

Sohlenius

(2000) The Holocene history of the southwestern Baltic Sea as reflected in a sediment core from the Bornholm Basin. Boreas 29: 233–250.

Andrén

Björck

Andrén

. (2011) The development of the Baltic Sea Basin during the last 130 ka. In: Harff

Björck

Hoth

(eds) The Baltic Sea Basin. Berlin: Springer, pp. 75–97.

Battarbee

Jones

Flower

. (2001) Diatoms. In: Smol

Birks

HJB

Last

(eds) Tracking Environmental Change Using Lake Sediments, Volume 3: Terrestrial, Algal, and Siliceous Indicator. Dordrecht: Kluwer Academic Publishers, pp. 155–202.

Berglund

Ralska-Jasiewiczowa

(1986) Pollen analysis and pollendiagrams. In: Berglund

(ed.) Handbook of Holocene Palaeoecology and Palaeohydrology. Chichester: John Wiley & Sons, pp. 455–484.

Berglund

Sandgren

Barnekow

. (2005) Early Holocene history of the Baltic Sea, as reflected in coastal sediments in Blekinge, southeastern Sweden. Quaternary International 130: 111–139.

Bērziņš

(2008) Sārnate: Living by a coastal lake during the East Baltic Neolithic. Acta Universitatis Ouluensis B Humaniora 86. PhD Thesis, Oulu University.

10.

Carpelan

Uino

Gerasimov

(2008) Archaeology in the former municipality of Johannes. In: Lavento

Nordqvist

(eds) Karelian Isthmus: Stone Age Studies in 1998–2003. ISKOS 16. Helsinki: The Finnish Antiquarian Society, pp. 185–211.

11.

Christensen

(1995) The Littorina transgressions in Denmark. In: Fischer

(ed.) Man and Sea in the Mesolithic. Oxford: Oxbow Books, pp. 15–22.

12.

Cooper

(2007) High energy coasts sedimentary indicators. Encyclopedia of Quaternary Science 1: 2983–2993.

13.

Daugnora

Girininkas

(2004) Rytų Pabaltijo bendruomenių gyvensena XI–II tūkst. pr. Kr. Kaunas: Lietuvos veterinarijos akademijos leidykla.

14.

Ekman

(1996) A consistent map of the postglacial uplift of Fennoscandia. Terra Nova 8: 158–165.

15.

Fischer

(2011) Stone Age on the continental shelf. An eroding resource. In: Benjamin

Bonsall

Fischer

. (eds) Submerged Prehistory. Oxford: Oxbow Books, pp. 289–310.

16.

Gerasimov

Kriiska

Lisitsyn

(2010) Colonization of the Gulf of Finland (the Baltic Sea) coastal zone in the Stone Age. In: Golavnev

(ed.) III Northern Archaeological Congress. Papers. November 8–13, 2010 Khanty-Mansisk. Ekaterinburg: Academy of Sciences, pp. 28–53.

17.

Grudzinska

Saarse

Vassiljev

. (2013) Mid- and late-Holocene shoreline changes along the southern coast of the Gulf of Finland. Bulletin of the Geological Society of Finland 85: 19–34.

18.

Grudzinska

Saarse

Vassiljev

. (2014) Lake Harku, Limnea Sea lagoon in North Estonia: Biostratigraphy and shoreline changes. Baltica 27(1): 15–24.

19.

Heinsalu

(2000) Diatom stratigraphy and palaeoenvironment of the Yoldia Sea in northern Estonia. Proceedings of the Estonian Academy of Sciences: Geology 49(3): 218–243.

20.

Heiri

Lotter

Lemcke

(2001) Loss-on-ignition as a method for estimating organic and carbonate content in sediments: Reproducibility and comparability of results. Journal of Paleolimnology 25: 101–110.

21.

Jöns

(2011) Settlement development in the shadow of coastal changes – Case studies from the Baltic Rim. In: Harff

Björck

Hoth

(eds) The Baltic Sea Basin. Berlin: Springer, pp. 301–336.

22.

Jussila

Kriiska

(2004) Shore displacement chronology of the Estonian stone age. Estonian Journal of Archaeology 8(1): 3–32.

23.

Kadakas

(2010) Tallinna Vabaduse väljaku neoliitiline asulakoht Eesti samaaegsete rannikuasulate kontekstis. Master’s Thesis, University of Tartu.

24.

Kadakas

Lõugas

. (2009) Vabaduse väljak. Eesti mahukaim arheoloogiline uurimisobjekt. Horisont 5: 8–13.

25.

Kadakas

Vedru

Lõugas

. (2010) Rescue excavations of the Neolithic settlement site in Vabaduse Square, Tallinn. In: Oras

Russow

(eds) Archaeological Fieldwork in Estonia 2009. Tallinn: Muinsuskaitseamet, pp. 27–46.

26.

Krammer

Lange-Bertalot

(1988) Bacillariophyaceae 2. Teil Bacillariaceae, Epithemiaceae, Surirellaceae. In: Ettl

Gerloff

Heying

. (eds) Süsswasserflora von Mitteleuropa 2. Stuttgart: Gustav Fischer Verlag, 596 pp.

27.

Krammer

Lange-Bertalot

(1999–2004) Bacillariophyceae. In: Ettl

Gerloff

Heying

(eds) Süsswasserflora von Mitteleuropa (vol. 2). Heidelberg, Berlin: Spektrum Akademischer Verlag, pp. 3–4.

28.

Kriiska

(2000) Settlements of coastal Estonia and maritime hunter-gatherer economy. Lietuvos archeologija 19: 153–166.

29.

Kriiska

(2001) Stone Age settlement and economic processes in the Estonian coastal area and islands. PhD Thesis, University of Helsinki.

30.

Kriiska

(2003a) Colonisation of the west Estonian archipelago. In: Larsson

Kindgren

Knutson

. (eds) Mesolithic on the Move: Papers Presented at the Sixth International Conference on the Mesolithic in Europe. Oxford: Oxbow Books, pp. 20–28.

31.

Kriiska

(2003b) From hunter-gatherer to farmer. Changes in the Neolithic economy and settlement on Estonian territory. Archaeologia Lituana 4: 11−26.

32.

Kriiska

(2007) Saaremaa kiviaeg. In: Jänes-Kapp

Randma

Soosaar

(eds) Saaremaa: Ajalugu. Majandus. Kultuur. Tallinn: Koolibri, pp. 9–36.

33.

Kriiska

Lõugas

(2009) Stone Age settlement sites on an environmentally sensitive coastal area along the lower reaches of the river Pärnu (south-western Estonia), as indicators of changing settlement patterns, technologies and economies. In: McCartan

Woodman

Schulting

. (eds) Mesolithic Horizons: Papers Presented at the Seventh International Conference on the Mesolithic in Europe. Oxford: Oxbow Books, pp. 167–175.

34.

Kriiska

Nordqvist

Gerasimov

. (2014) Preliminary results of the research at Corded Ware sites in the Narva–Luga interfluve, Estonian–Russian border area in 2008–2014. In: Haak

Russow

(eds) Archaeological Fieldwork in Estonia 2014. Tallinn: Muinsuskaitseamet, pp. 39−50.

35.

Larsson

(1997) Coastal settlement during the Mesolithic and Neolithic periods in the southernmost part of Sweden. In: Król

(ed) The Built Environment of Coast Areas during the Stone Age. Gdańsk: Archaeological Museum, pp. 12–22.

36.

Leskinen

Pesonen

(2008) Vantaan esihistoria. Keuruu: Otavan Kirjapaino Oy.

37.

Lindqvist

Possnert

(1997) The subsistence economy and diet at Jakob/Ajvide, Eksta Parish and other prehistoric swelling and burial sites on Gotland in long-term perspective. In: Burenhult

(ed.) Remote Sensing: Applied Techniques for the Study of Cultural Resources and the Localization, Identification and Documentation of Sub-Surface Prehistoric Remains in Swedish Archaeology. Stockholm: Stockholm University, pp. 29–90.

38.

Loze

(2006) Neolīta apmetnes Ziemeļkurzemes kāpās. Rīga: Latvijas vēstures institūta apgāds.

39.

Lõugas

(1997) Post-Glacial development of vertebrate fauna in Estonian water bodies: A palaeozoological study. Dissertationes Biologicae Universitatis Tartuensis, 32. PhD Thesis, Tartu University.

40.

Lõugas

(1999) Postglacial development of fish and seal faunas in the Eastern Baltic water system. In: Benecke

(ed.) The Holocene History of the European Vertebrate Fauna: Modern Aspects of Research. Rahden: VML Verlag Marie Leidorf, pp. 185–200.

41.

Lõugas

Tomek

(2013) Marginal effect at the coastal area of Tallinn Bay: The marine, terrestrial and avial fauna as a source of subsistence during the Late Neolithic. Muinasaja Teadus 19: 463–485.

42.

Lübke

Schmölcke

Tauber

(2011) Mesolithic hunter-fishers in a changing world: A case study of submerged sites on the Jäckelberg, Wismar Bay, northeastern Germany. In: Benjamin

Bonsall

Pickard

. (eds) Submerged Prehistory. Oxford: Oxbow Books, pp. 21–37.

43.

Nuñez

Okkonen

(1999) Environmental background for the rise and fall of villages and megastructures in North Ostrobothnia 4000–2000 cal BC. In: Huurre

(ed.) Dig it all: Papers Dedicated to Ari Siiriäinen. Helsinki: The Archaeological Society of Finland, The Finnish Antiquarian Society, pp. 105–115.

44.

Olsson

(1980) Content of ¹⁴C in marine mammals from Northern Europe. Radiocarbon 22: 515–544.

45.

Orford

Carter

RWG

Jennings

(1991) Coarse clastic barrier environments: Evolution and implications for quaternary sea level interpretations. Quaternary International 9: 87–104.

46.

Piličiauskas

Heron

(2015) Aquatic radiocarbon reservoir offsets in the southeastern Baltic. Radiocarbon 57(4): 539–556.

47.

Post

Kõuts

(2014) Characteristics of cyclones causing extreme sea levels in the northern Baltic Sea. Oceanologia 56(2): 241–258.

48.

Ramsey

(2009) Bayesian analysis of radiocarbon dates. Radiocarbon 51: 337–360.

49.

Reimer

Bard

Bayliss

. (2013) IntCal13 and Marine13 radiocarbon age calibration curves 0–50,000 years cal. BP. Radiocarbon 55(4): 1869–1887.

50.

Rosentau

Muru

Kriiska

. (2013) Stone age settlement and Holocene shore displacement in the Narva-Luga Klint Bay area, eastern Gulf of Finland. Boreas 42(4): 912–931.

51.

Rosentau

Veski

Kriiska

. (2011) Palaeogeographic model for the SW Estonian coastal zone of the Baltic Sea. In: Harff

Björck

Hoth

(eds) The Baltic Sea Basin. Berlin: Springer, pp. 165–188.

52.

Saarse

Vassiljev

(2010) Holocene shore displacement in the surroundings of Tallinn, North Estonia. Estonian Journal of Earth Sciences 59(3): 207–215.

53.

Saarse

Heinsalu

Veski

(2009) Litorina Sea sediments of ancient Vääna Lagoon, northwestern Estonia. Estonian Journal of Earth Sciences 58(1): 85–93.

54.

Saarse

Vassiljev

Heinsalu

(2010) Reconstruction of the land-sea changes on the Juminda Peninsula, North Estonia, during the last 10 300 years. Baltica 23(2): 117–126.

55.

Saarse

Heinsalu

Poska

. (1997) Early Holocene shore displacement of the Baltic Sea east of Tallinn (N Estonia). Baltica 10: 13–24.

56.

Saarse

Vassiljev

Miidel

(2003) Simulation of the Baltic Sea shorelines in Estonia and neighbouring areas. Journal of Coastal Research 19(2): 261–268.

57.

Saarse

Veski

Miidel

(2001) Lepistiku buried peat. Proceedings of the Estonian Academy of Sciences: Geology 50(2): 128−134.

58.

Saarse

Vassiljev

Rosentau

. (2007) Reconstructed late glacial shore displacement in Estonia. Baltica 20(1–2): 35–45.

59.

Seitsonen

(2008) Osteological material from the Stone Age and Early Metal Period sites in Karelian Isthmus and Ladoga Karelia. In: Lavento

Nordqvist

(eds) Karelian Isthmus: Stone Age Studies in 1998–2003. ISKOS 16. Helsinki: The Finnish Antiquarian Society, pp. 266–283.

60.

Siiriäinen

(1982) Shore displacement and archaeology in Finland. Annales Academiae Scientiarum Fennicae A III 134: 173–184.

61.

Ukkonen

(2001) Shaped by the Ice Age: Reconstructing the History of Mammals in Finland during the Late Pleistocene and Early Holocene. PhD Thesis, University of Helsinki.

62.

Ukkonen

Aaris-Sørensen

Arppe

. (2014) An Arctic seal in temperate waters: History of the ringed seal (Pusa hispida) in the Baltic Sea and its adaptation to the changing environment. Holocene 24(12): 1694–1706.

63.

Vaneeckhout

(2008) Sedentism on the Finnish northwest coast: shoreline reduction and reduced mobility. Fennoscandia Archaeologica 25: 61–72.

64.

Veski

(1998) Vegetation history, human impact and palaeogeography of West Estonia. Pollen analytical studies of lake and bog sediment. STRIAE 38: 1–119.

65.

Veski

Heinsalu

Klassen

. (2005) Early Holocene coastal settlements and palaeoenvironment on the shore of the Baltic Sea at Pärnu, southwestern Estonia. Quaternary International 130(1): 75–85.

66.

Zagorska

(2000) Sea mammal hunting strategy in Eastern Baltic. Lietuvos archeologija 19: 275–285.

67.

Zobel

(2009) Tallinn (Reval) keskajal. Linnaehitus 13.-14. sajandil. Tallinn: Eesti Kunstiakadeemia.