Journal of Quaternary Science Page 2 of 57

1 2 3 1 Geology, stratigraphy and palaeoenvironmental evolution of the Stephanorhinus 4 5 2 6 kirchbergensis-bearing Quaternary palaeolake(s) of Gorzów Wielkopolski (NW Poland, 7 8 3 Central Europe) 9 10 4 11 12 5 Artur Sobczyk a*, Ryszard K. Borówka b, Janusz Badura c, Renata Stachowicz-Rybka d, 13 14 b e b b 15 6 Julita Tomkowiak , Anna Hrynowiecka , Karolina Bloom , Joanna Sławińska , 16 17 7 Kamila Mianowicz b, Michał Tomczak f, Mateusz Pitura a, Bartosz Kotrys g, Mariusz 18 19 8 Lamentowicz h, Piotr Kołaczek h, Niska Monika i, Dariusz Tarnawski j, Marcin Kadej j, 20 21 k l m d 22 9 Jan Kotusz , Piotr Moska , Marek Krąpiec , Krzysztof Stachowicz , Bartosz Bieniek 23 24 10 b, Krzysztof Siedlik b, Małgorzata Bąk f, Jolanta Piątek d, Joanna Lenarczyk d, Marcin 25 26 11 Olkowicz c, J. van der Made n, Adam Kotowski o, Krzysztof Stefaniak o 27 28 29 12 a Department of Structural Geology and Geological Mapping, Institute of Geological 30 13 Sciences, University of Wroclaw, pl. M. Borna 9, 50-204 Wrocław, Poland; 31 32 14 b Geology and Palaeogeography Unit, Faculty of Geosciences, University of Szczecin, 33 15 Mickiewicza 18, 70-383, Szczecin, Poland 34 c 35 16 Polish Geological Institute – National Research Institute, Jaworowa alley 19, 53-122 36 17 Wrocław, Poland 37 18 d W. Szafer Institute of Botany, Polish Academy of Sciences, Lubicz 46, 31-512 Kraków, 38 39 19 Poland 40 20 e Polish Geological Institute - National Research Institute, Marine Geology Branch, 41 21 Kościerska 5, 80-328 Gdańsk, Poland 42 43 22 f Palaeoceanography Unit, Faculty of Geosciences, University of Szczecin, Mickiewicza 18, 44 23 70-383, Szczecin, Poland 45 46 24 g Polish Geological Institute - National Research Institute, Wieniawskiego 20, 71-130 47 25 Szczecin, Poland 48 h 49 26 Department of Biogeography and Palaeoecology, Institute of Geoecology and 50 27 Geoinformation, Adam Mickiewicz University in Poznań, Krygowskiego 10, 61-880 Poznań, 51 28 Poland 52 i 53 29 Institute of Geography, Pomeranian University in Słupsk, Partyzantów 27a., 76-200 Słupsk, 54 30 Poland 55 31 j Department of Invertrabete Biology, Evolution and Conservation, Institute of Environmental 56 32 57 Biology, University of Wroclaw, Sienkiewicza 21, 50-335 Wrocław, Poland 58 33 k Museum of Natural History, University of Wroclaw, Sienkiewicza 21, 50-335 Wrocław, 59 34 Poland 60

http://mc.manuscriptcentral.com/jqs Page 3 of 57 Journal of Quaternary Science

1 2 3 35 l Division of Radioisotopes, Institute of Physics, Centre for Science and Education, Silesian 4 36 University of Technology, 22B Konarskiego st., 44-100 Gliwice, Poland 5 6 37 m AGH University of Science and Technology, Faculty of Geology, Geophysics and 7 38 Environment Protection, Chair of General Geology and Geotourism, Mickiewicza alley 30, 8 39 30-059 Kraków, Poland 9 10 40 n Museo Nacional de Ciencias Naturales, Consejo Superior de Investigaciones Científicas. C. 11 41 José Gutiérrez Abascal 2, 28006 Madrid, Spain 12 o 13 42 Department of Palaeozoology, Institute of Environmental Biology, University of Wroclaw, 14 43 Sienkiewicza 21, 50-335 Wrocław, Poland 15 16 44 17 45 *corresponding author: [email protected] 18 19 20 46 21 22 47 Keywords: Quaternary stratigraphy, Eemian interglacial, Palynology and macrofossils 23 24 25 48 analysis, OSL dating, radiocarbon dating 26 27 49 28 29 50 Abstract 30 31 51 32 The sedimentary succession exposed in the Gorzów Wielkopolski area includes Eemian 33 34 52 Interglacial (MIS 5e) or Early Weichselian (MIS 5d-e) deposits. The sedimentary sequence has 35 36 53 been the object of intense interdisciplinary study, which resulted in the identification of at least 37 38 54 two palaeolake horizons. Both yielded fossil remains of large mammals, alongside pollen and 39 40 41 55 plant macrofossils. All these proxies have been used to reconstruct the environmental 42 43 56 conditions prevailing at the time of deposition, as well as to define the geological context and 44 45 57 the biochronologic position of the fauna. OSL dating of the glaciofluvial layers of the GS3 46 47 48 58 succession to 141,4±3,0 (below the lower palaeolake) and 79,0±1,5 ka (above the upper 49 50 59 palaeolake) indicate that the site formed during the Middle-Late Pleistocene (MIS 6-MIS 5). 51 52 60 Radiocarbon dating of the lacustrine organic matter revealed a tight cluster of Middle 53 54 61 55 Pleniglacial Period (MIS 3) ages in the range of ~41-32 ka cal BP (Hengelo – Denekamp 56 57 62 Interstadials). Holocene organic layers have also been found, with 14C ages within a range of 58 59 60

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 4 of 57

1 2 3 63 4,330 – 4,280 cal. BP (Neolithic). Pollen and plant macrofossil records, together with 4 5 64 6 sedimentological and geochemical data, confirm the dating to the Eemian Interglacial. 7 8 65 9 10 66 Introduction 11 12 67 A well-preserved, nearly complete skeleton of a tandem-horned rhinoceros Stephanorhinus 13 14 15 68 kirchbergensis (Jäger, 1839) has recently been found at an Eemian Interglacial site near Gorzów 16 17 69 Wielkopolski, in NW Poland. The remains have been discovered in 2016 during works along 18 19 70 S3 Polish national road. They were located at the southern end of an escarpment along the road. 20 21 22 71 The skeleton was preserved in a succession that had accumulated in a small, shallow kettle 23 24 72 palaeolake formed during the Scandinavian ice-sheet retreat. The Gorzów deposits belong to at 25 26 73 least two limnic cycles, an older one, dating to the Eemian Interglacial (MIS 5e), and an 27 28 74 29 overlying one of Mid-Weichselian (MIS 3) age; the rhinoceros remains were preserved in a 30 31 75 gyttja layer of the former unit. Besides the skeletal remains of Stephanorhinus kirchbergensis, 32 33 76 the site also yielded an isolated left metacarpal bone of fallow deer, Dama dama from the basal 34 35 77 part of a peat layer that lies over the rhinoceros-bearing gyttja. The rhinoceros skeleton includes 36 37 38 78 the cranium, with full dentition, and almost 120 postcranial bones; the lumbar and caudal part 39 40 79 of the spinal column, as well as the left hind limb and a few minor bones were not preserved. 41 42 80 All the bones are exceptionally preserved; the skull of the rhinoceros suffered diagenetic 43 44 45 81 crushing, and damages caused by the roadworks. 46 47 82 In the European Lowland area, Eemian Interglacial and Early Weichselian deposits are usually 48 49 83 found south of the limit of maximum extension of the Weichselian (Vistulian) glaciation (Marks 50 51 84 52 et al., 2016), whereas within the post-Weichselian areas they are reported only from boreholes 53 54 85 and pits (Fig. 1). Up until now, only a few Eemian Interglacial sites were known in NW Poland, 55 56 86 and palynological and malacological analysis had been performed in only five of them, i.e. 57 58 87 Radówek (Urbański and Winter, 2005), Rzecino (Mirosław-Grabowska, 2008; Winter et al., 59 60

http://mc.manuscriptcentral.com/jqs Page 5 of 57 Journal of Quaternary Science

1 2 3 88 2008; Niska and Mirosław-Grabowska, 2015), Rusinowo (Stark et al., 1932), Słubice 4 5 89 6 (Skompski, 1980; Alexandrowicz and Alexandrowicz, 2010) and Nakło (Noryśkiewicz, 1978). 7 8 90 Further to the west, Eemian deposits have been documented around the municipality of Poznań 9 10 91 and the southern region of Wielkopolska (Straszewska and Stupnicka, 1979; Kuszell and 11 12 92 Malkiewicz, 1999; Bruj and Roman, 2007). From the German Brandenburg and Mecklenburg 13 14 15 93 regions, sediments of Eemian age have been reported by Eissmann (2002), Brose et al. (2006), 16 17 94 Hermsdorf and Strahl (2008) and Rother et al. (2017). Middle Pleniglacial deposits of the same 18 19 95 age as the upper peats horizon from section GS3 (Fig. 1) of Gorzów Wielkopolski occur in 20 21 22 96 brown-coal open-cast mines in the region of Lausitz (Eastern Germany) (see Mol, 1997; Bos et 23 24 97 al., 2001). Kozarski et al. (1980) found Brørup strata in a brickyard nearby Drezdenko (Stary 25 26 98 Kurow). Only a few Eemian and Early Weichselian mollusc faunas from these sites were 27 28 99 29 described (Skompski, 1996; Alexandrowicz and Alexandrowicz, 2010), whereas no vertebrate 30 31 100 remains had ever been found preserved in situ until now. Other fossil-bearing localities with 32 33 101 remains of large mammals are Imbramowice, near Wrocław (Marciszak et al., 2017), Warsaw 34 35 102 (Czyżewska, 1962; Borsuk-Białynicka and Jakubowski, 1972), and Konin (Lorek, 1988), and 36 37 38 103 still other similar sites are known in Eastern Germany (Kolfschoten, 2000; Eismann, 2002; 39 40 104 Gaudzinski-Windheuser et al., 2014). The nearly complete skeleton of Stephanorhinus 41 42 105 kirchbergensis (Jäger, 1839) from Gorzów Wielkopolski is therefore key reference for the 43 44 45 106 analysis of the Central European Eemian/Early Weichselian. This study presents the 46 47 107 preliminary results of an interdisciplinary research project aimed at performing an in-depth 48 49 108 analysis of the rhinoceros from Gorzów Wielkopolski. The paper focuses in particular on the 50 51 109 52 environmental conditions under which the animal lived, as well as on later environmental 53 54 110 changes. 55 56 111 Regional setting - study area 57 58 59 60

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 6 of 57

1 2 3 112 The studied area is located in NW Poland (Fig. 2), ca. 6 km to the W of Gorzów Wielkopolski 4 5 113 6 municipality (51°43' N; 15°10' E), along a deep constructional excavation along the S3 national 7 8 114 road. The road now cuts through the valley escarpment separating Gorzów Plain, to the north, 9 10 115 from the ice-marginal valley of Toruń-Eberswalde to the south. Gorzów Plain is a vast ground 11 12 116 morainic plateau intercalated with outwash plain deposits, formed during the maximum 13 14 15 117 extension of the Weichselian glaciation in Western Poland (Germany – Brandenburg phase; 16 17 118 Poland – Leszno phase) when, during its climax, the glacial sheets reached the Leszno area, ca. 18 19 119 90 km south of the studied area (Fig. 2A). The ground moraine area is drained to the south by 20 21 22 120 a system of syn- to post-glacial channels, which form a dense network of small erosional, 23 24 121 mostly dry, deeply-incised valleys that cut through the morphological crevasse. Gorzów Plain 25 26 122 is between 70 to 90 m high a.s.l., whereas the base of the Toruń-Eberswalde valley reaches 18- 27 28 123 29 20 m a.s.l. The fossil bones were found at site GS3 (Fig. 2C); they were preserved in a narrow 30 31 124 melt-out depression filled with Eemian lacustrine calcareous gyttja (Fig. 3). Two limnic 32 33 125 horizons occur in the central GS3 section (Fig. 4A): the lower one is a lacustrine unit of Eemian 34 35 126 age, the upper one is of Weichselian (Vistulian) age. The lacustrine succession is capped by a 36 37 38 127 peat bog deposit (Fig. 4-6), whereas the upper horizon by glaciofluvial and glacial beds. Section 39 40 128 GS1 includes about 11 m of lacustrine deposits that end with fine and medium sands 41 42 129 interbedded with silty sands. Section GS2 includes 13.5 m of uniform, unsorted, glaciofluvial 43 44 45 130 sands and gravels. The entire GS2 profile is perhaps an older, presumably Saalian glaciofluvial 46 47 131 sequence, possibly correlated with the basal parts of the GS1 and GS3 successions (Fig. 3, 4). 48 49 132 Methods 50 51 133 52 Fieldwork and sampling strategy 53 54 134 Geological and geomorphological maps of the Quaternary strata exposed along a road cut 55 56 135 nearby Gorzów Wielkopolski were developed based on a geodetic survey that served to 57 58 136 establish a series of cross-sections through which the spatial extent of the extinct lake deposits 59 60

http://mc.manuscriptcentral.com/jqs Page 7 of 57 Journal of Quaternary Science

1 2 3 137 was defined (Fig. 6). A Leica SR530 dual-frequency GPS RTK receiver supplied with a VRS 4 5 138 6 net correction device was used for the geodetic survey. Geochemical, sedimentological and 7 8 139 environmental reconstructions were based on evidence collected from sections of ca. 11 m of 9 10 140 the GS1 (Tab. 1, 2) and GS3 (Tab. 3, 4) successions (Fig. 2C, 3). Moreover, radiocarbon and 11 12 141 optically stimulated luminescence dating were performed to define the chronological extent of 13 14 15 142 the exposed sediments (Tab. 5, 6). The GS1, GS2 and GS3 sequences were located on the west 16 17 143 side of the S3 road, and section GS4 on the east (Fig. 5, 6). After angle correction, the four 18 19 144 sections GS1-GS4 reached the total thicknesses of 21.5, 13.5, 22 and 23 metres, respectively 20 21 22 145 (Fig. 3, 6). A comparative Racław profile (Fig. 3) was built based on a sketchy description of 23 24 146 the section drilled by the cartographic borehole no. 62 (Piotrowski et al., 2010). Detailed 25 26 147 sedimentological, geochemical, palynological and plant macrofossil analyses have been 27 28 148 29 performed of the GS1, and GS3 sections (Figs. 7-10) aimed to define the characters of the 30 31 149 palaeolake deposits and the environmental context in which they accumulated. 32 33 150 34 35 151 Geochemical and sedimentological analyses 36 37 38 152 A total of 171 and 181 sediment samples were analysed from the sections GS1 and GS3, 39 40 153 respectively (Fig. 7, 8); a constant sampling pace of 6 cm per ca. 11 metres was maintained for 41 42 154 the lacustrine sediments of the GS1 sequence as well as for the bottom part of the GS3 section. 43 44 45 155 All samples were freeze-dried in a Beta 1-8 LD plus Lyophiliser (Christ). Following 46 47 156 homogenisation, some samples were analysed for grain size distribution, using a Mastersizer 48 49 157 3000 laser grain size analyser (Malvern Instruments Ltd.). This permitted to determine the 50 51 158 52 relative percentages of the different grain size fractions, as well as to calculate graphically, 53 54 159 according to Folk and Ward (1957), their major properties: mean grain diameter (MD), standard 55 56 160 deviation (SD), skewness (Sk), and kurtosis (KG). In addition, loss on ignition at 550ºC was 57 58 161 calculated on about 2 g sediment samples dried at 105ºC to estimate the percentages of organic 59 60

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 8 of 57

1 2 3 162 and mineral matter content. The organic matter-deprived ash remaining after sediment 4 5 163 6 combustion was treated with 2 ml 10% hydrochloric acid and 2 ml hydrogen peroxide in a 7 8 164 Speedwave microwave mineraliser (Berghof). Atomic absorption spectrometry assays of the 9 10 165 solutions obtained, aimed at revealing their content of Na, K, Ca, Mg, Fe, Mn, Cu, and Zn, were 11 12 166 conducted using an AAS SOLAAR 969 spectrometer (Unicam). The air-dry sediment samples 13 14 15 167 were also used to determine per cent contributions of total carbon, total nitrogen, and total 16 17 168 sulphur, using a VARIOMAX CNS analyser (Elementar). All the lithological and geochemical 18 19 169 analyses were carried out at the laboratory of the Geology and Palaeography Unit of the 20 21 22 170 University of Szczecin. The data were subjected to statistical treatment, primarily with 23 24 171 Microsoft Office Excel software as well as with PAST Palaeontological Statistics (Hammer et 25 26 172 al., 2001) to identify geochemical horizons. The grain size data were stored and processed using 27 28 173 29 GRADISTAT software (Blott and Pye, 2001), and C2 programme (Juggins, 2007) was used to 30 31 174 visualise the results of laboratory analyses. 32 33 175 Bone remains 34 35 176 The teeth of the rhinoceros were measured according to Van der Made (2010) and Shpansky 36 37 38 177 (2016), and the postcranial bones following Gromova (1935, 1960), Guérin (1980) and Van der 39 40 178 Made (2010); Von den Driesch (1976), Di Stefano (1996), Lister (1996) were used for the 41 42 179 measurement and identification of the metacarpal bone of fallow deer. 43 44 45 180 Pollen and plant macroremains 46 47 181 Thirty-five samples for pollen analysis were collected from a depth of 10.26 to 1.38 m (from 48 49 182 bottom to top) of section GS3 (Fig. 9). The samples were acetolyzed according to Erdtman's 50 51 183 3 52 method (1960). Before acetolysis, a tablet containing Lycopodium spores per 1 cm was added 53 54 184 to each sample as a quantitative indicator (Stockmar, 1971; Berglund and Ralska-Jasiewiczowa, 55 56 185 1986). Pollen spectra were counted on two slides from a surface area of 20x20 mm routinely 57 58 186 up to a total of ca. 500 grains. Thirty-eight samples, each of 150 ml of sediment, were collected 59 60

http://mc.manuscriptcentral.com/jqs Page 9 of 57 Journal of Quaternary Science

1 2 3 187 for plant macroremains analysis from a depth of 10.26-2.10 m of section GS3, at an average 4 5 188 6 pace of 10-40 cm (the same adopted for the pollen sampling). The samples were macerated with 7 8 189 10% KOH, boiled to a pulp, and wet-sieved using a φ 0.2 mm mesh sieve. The concentrated 9 10 190 sieved residue was then sorted under a stereoscopic microscope. Plant remains that could be 11 12 191 identified were selected and placed in a mixture of glycerine, water and ethyl alcohol (1:1:1) 13 14 15 192 with the addition of thymol. Macrofossils were identified with the use of keys, atlases (e.g. 16 17 193 Cappers et al., 2006; Velichkevich and Zastawniak, 2006, 2008), literature data, as well as using 18 19 194 a reference collection of present-day seeds and fruits, as well as collections of fossil floras 20 21 22 195 housed in the Palaeobotanical Museum of the W. Szafer Institute of Botany of the Polish 23 24 196 Academy of Sciences, Kraków. Qualitative and quantitative results of the plant macrofossil 25 26 197 analysis are presented in a diagram (Fig. 10) plotted with POLPAL software for Windows 27 28 198 29 (Nalepka and Walanus, 2003). Local Macrofossil Assemblage Zones (L MAZ) were 30 31 199 distinguished in the diagram of macroscopic plant remains from section GS3; they were 32 33 200 numbered and labelled, from base to top, GS3-M1 to GS3-M8. 34 35 201 Geochronological data 36 37 38 202 Three optically stimulated luminescence datings (OSL; Fig. 11, Tab. 6) and five radiocarbon 39 40 203 14C datings (Tab. 5) were performed for representative stratigraphic horizons of sections GS3 41 42 204 and GS4 (Fig. 3-5). All OSL measurements were made on individual quartz grains by the 43 44 45 205 Department of Radioisotopes of the Silesian University of Technology, according to standard 46 47 206 procedures (for details see Aitken, 1998; Galbraith et al., 1999; Murray and Wintle, 2000; 48 49 207 Guérin et al., 2011). OSL datings were obtained from samples of glaciofluvial sands collected 50 51 208 52 from three different levels of section GS3. The samples were collected with steel tubes (1) 21.8 53 54 209 m b.s.l. at the bottom of the lower palaeolake section; (2) 16.5 m b.s.l., from a ca. 2 m-thick 55 56 210 interlayer between the two lacustrine successions and (3) 10.5 m b.s.l., ca. 0.5 m at the top of 57 58 211 the upper (second) palaeolake section. Dose rates were calculated using the conversion factors 59 60

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 10 of 57

1 2 3 212 of Guerin et al. (2011). For beta dose rate the cosmic ray dose-rate to the site was determined 4 5 213 6 as described by Prescott and Stephan (1982). Sand-sized grains of quartz (90–200 µm in 7 8 214 diameter) were extracted from the sample using standard purification procedures (e.g., Aitken, 9 10 215 1998). All OSL measurements were made using an automated Risø TL/OSL DA-20 reader for 11 12 216 multi-grain aliquots, each of ca. 1 mg. The stimulation light source was a blue (470 ± 30 nm) 13 14 15 217 light emitting diode array (LED) delivering 50 mW/cm2 at the sample (Bøtter-Jensen et al. 16 17 218 2000). Detection was through 7.5 mm of Hoya U-340 filter. Equivalent doses have been 18 19 219 determined using the single-aliquot regenerative-dose (SAR) protocol (Murray and Wintle, 20 21 22 220 2000). Final OSL ages have been calculated using the Central Age Model (Galbraith et al. 1999) 23 24 221 and presented on radial plots. Plant macrofossils and peat of the upper palaeolake succession 25 26 222 of GS3 and two peat layers of section GS4 have been radiocarbon dated: the upper layer of GS4 27 28 223 29 is located in a hollow on the top of glacial till and a lower one just below the base of glacial 30 31 224 succession (Fig. 3-5, Tab. 5). 32 33 225 34 35 226 Results 36 37 38 227 Lithology and geochemistry 39 40 228 Section GS1 41 42 229 Lithological and geochemical analyses of ca. 11.5 m of limnic deposits at the bottom of section 43 44 45 230 GS1 (Fig. 4B, 7; Tab. 1, 2) revealed the clear dominance of silt fractions in the range from 46 47 231 about 60 to 90%. The highest admixture of sand fractions was found in the upper part of the 48 49 232 succession, from 0.0 to about 3.5 m (Fig. 7A). In the middle of the stratified sequence, at 3.5- 50 51 233 52 9.5 m, features a minor share of fine sand. It is only below 9.5 m that the relative amounts of 53 54 234 both medium and fine sand markedly increase. The mean grain size (Mz) of the limnic deposits 55 56 235 ranges between 5 and 7 φ units. Sorting (SD) varies but gradually increases moving downward 57 58 236 along the sequence. Most grain size distributions are negatively skewed (Sk); symmetric and 59 60

http://mc.manuscriptcentral.com/jqs Page 11 of 57 Journal of Quaternary Science

1 2 3 237 positively skewed distributions occur only at 2-3 m of the GS1 section. Kurtosis (KG) ranges 4 5 238 6 from values higher than 1.1, in the bottom part of the succession, to 0.7-0.8, in the upper part. 7 8 239 Organic matter (Fig. 7B) is distributed almost homogenously throughout the section, and varies 9 10 240 from about 4 to 6%; only between 10.5-11.1 m it reaches values of 8-11%. The organic matter 11 12 241 content correlates significantly with the concentrations of total sulphur and iron (Tab. 2). Total 13 14 15 242 carbon content in the limnic deposits tends to decrease gradually moving downwards along the 16 17 243 section and is most strongly and positively correlated with calcium, magnesium, and total 18 19 244 nitrogen. On the other hand, negative correlations were obtained with zinc, potassium, and 20 21 22 245 sodium. The highest total nitrogen concentrations were recorded between about 0.75-4.0 m and 23 24 246 7.0-7.3 m. The total sulphur concentration remains virtually stable at about 0.015% in the upper 25 26 247 and middle part of the sequence, while in the lower half it rises from 0.015 to 0.045 %. Total 27 28 248 29 sulphur concentrations increase steadily, peaking in the organic-matter-rich layer at 8-11 m. 30 31 249 Noteworthy is also a strong, positive correlation between total C, total N and Ca concentrations. 32 33 250 Na/K and Fe/Mn concentrations remain stable in all but the lower part of the section, where 34 35 251 higher values are reached. No significant variations were observed in the concentrations of other 36 37 38 252 chemical elements. 39 40 253 Section GS3 41 42 254 The ca. 11 m of section GS3 analysed for this study includes gyttja, limnic chalk, silty sand and 43 44 45 255 peat layers (Fig. 4A, 8; Tab. 3). There is a dominance of silts, mixed with minor amounts of 46 47 256 fine and very fine sand. Sands characterise in particular the middle part of the section, at around 48 49 257 4.5–6.0 m; they lay over 60 cm of black peat (Fig. 4A). Mean grain sizes (Mz) range around 50 51 258 52 5–6 φ units; sorting (SD) constantly reaches 1.5 φ (Fig. 8A). Skewness (Sk) changes remarkably 53 54 259 only at the bottom and in the central part of the section, ranging from -0.2 to 0.5 phi. Kurtosis 55 56 260 (KG) is persistently one phi all but at the very bottom (10.5 m) of the sequence. Geochemical 57 58 261 analyses revealed significantly high Ca content (200–320 mg/g) in the limnic facies and 59 60

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 12 of 57

1 2 3 262 elevated levels of organic matter at 0–2 m and 6.5–7.0 m (Tab. 4). The highest values of the 4 5 263 6 total carbon, nitrogen and sulphur indexes occur at a transition from the organic mud to peats. 7 8 264 Ca/Mg and Na/K indexes reach the highest values in the gyttja/limnic chalk horizons, whereas 9 10 265 Fe/Mn increase significantly for the peat horizon. Na+K+Mg/Ca ratio reaches maximum values 11 12 266 at 0–2 m and 4.5–6.5 m, and is negligible elsewhere. Six geochemical zones have been 13 14 15 267 identified in section GS3. 16 17 268 Palaeobotanical analysis 18 19 269 The frequency of pollen was from high to very high. The results of the pollen analysis of section 20 21 22 270 GS3 are presented graphically in a percentage pollen diagram (Fig. 9) plotted using POLPAL 23 24 271 software for Windows (Nalepka and Walanus, 2003); 12 Local Pollen Assemblage Zones (L 25 26 272 PAZ) could be distinguished, based on the samples GS3-1 to GS3-12. 27 28 273 29 Cool climatic conditions, typical of final glacial phases, prevailed when organic sediments 30 31 274 started accumulating in the lake (GS3-1 Betula-NAP L PAZ, Fig. 9). The Late Saalian 32 33 275 (Wartanian) glaciation was characterised by very deforested landscapes, rarely overgrown with 34 35 276 Betula, Juniperus and with a very few trees of Pinus. On the other hand, there were numerous 36 37 38 277 steppe communities with Poaceae and Artemisia as well as tundras with Betula nana and 39 40 278 Cyperaceae (Fig. 9). During this period, at the level of GS3-M1 L MAZ (Fig. 10) the local 41 42 279 vegetation was initially dominated by Characeae, which appeared in the lake. At the bottom of 43 44 45 280 the reservoir, they occurred in communities that preferred clear water rich in calcium carbonate, 46 47 281 which encrusted their thalli. As a result, the sediment was supplemented with calcium 48 49 282 carbonate. 50 51 283 52 The beginning of the Eemian Interglacial was marked by the development of bright and open 53 54 284 birch-pine forests (GS3-2 Betula-Pinus), with an admixture of Ulmus in more humid places 55 56 285 (GS3-3 Pinus-Betula-Ulmus; Fig. 9). Herbaceous vegetation was becoming increasingly rare at 57 58 286 the time, which indicates a significant increase of tree crowns density and that climatic 59 60

http://mc.manuscriptcentral.com/jqs Page 13 of 57 Journal of Quaternary Science

1 2 3 287 conditions had improved. Then forests evolved into the Quercetum mixtum communities (GS3- 4 5 288 6 4 Quercus-Fraxinus-Ulmus), in which Quercus became dominant, and Pinus was abundant. 7 8 289 There was a characteristic small increase in the participation of herbaceous plants, which 9 10 290 indicate that these communities have become brighter again. The first four L PAZs are included 11 12 291 in a thin layer of sediment (36 cm), which indicate a slow rate of sedimentation or compaction 13 14 15 292 of sediment. In the level of GS3-M2 L MAZ (Fig. 10), the local presence of several tree species 16 17 293 growing close to the share of the lake is confirmed by the fruits and fruit scales of Betula sect. 18 19 294 Albae, fruits Alnus glutinosa as well as by the remains of Pinus sylvestris. At that time there 20 21 22 295 were already communities similar to contemporary reed bed with Typha sp. in the lake. Najas 23 24 296 marina and Nuphar lutea appeared at this time, which indicates that the water in the reservoir 25 26 297 had grown progressively warmer. 27 28 298 29 The appearance of Corylus in association with Quercus (GS3-5 Corylus-Quercus-Alnus) and, 30 31 299 successively, also with Tilia cordata (GS3-6 Corylus-Tilia-Alnus) indicates a further 32 33 300 improvement of climatic conditions. Alnus was also important at this time; it was accompanied 34 35 301 by Ulmus and Fraxinus, which indicates the development of riparian communities. In addition 36 37 38 302 to the macroremains of birch, pine and alder, the level of GS3-M3 L MAZ provided also 39 40 303 remains of Tilia platyphyllos, which indicates the local presence of this other species of lime in 41 42 304 the vicinity of the reservoir. The communities of largely rooted macrohydrophytes were 43 44 45 305 dominated by floating species, among which Najas marina, Nuphar lutea, Brasenia sp., and 46 47 306 Trapa natans. 48 49 307 The climatic optimum of the Eemian Interglacial was characterised by the development of oak- 50 51 308 52 hornbeam forests, including Carpinus, Corylus, Quercus and Tilia cordata, as well as of 53 54 309 riparian forests, mainly represented by Alnus (GS3-7 Carpinus-Corylus-Alnus; Fig. 9). The 55 56 310 upper part of the level of GS3-M3 L MAZ (Fig. 10) provided fruits of Carpinus betulus and 57 58 311 nuts of Corylus avellana, while in the lake, alongside other species characteristic of the climate 59 60

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 14 of 57

1 2 3 312 optimum, such as Brasenia sp., and Trapa natans, Najas marina was ever more present. The 4 5 313 6 level of GS3-M4 L MAZ, which is also included in time period of the climatic optimum, was 7 8 314 characterised by an increased presence of rush vegetation, including Cladium mariscus, Typha 9 10 315 sp. and Schoenoplectus lacustris. The very significant increase in the proportion of peat 11 12 316 vegetation such as Eleocharis palustris and Carax pseudocyperus, versus the preceding zones 13 14 15 317 most likely results from gradual shallowing of the basin and intensive overgrowth of its shores. 16 17 318 The disappearance of Carpinus, which was gradually replaced by pinewoods including Picea, 18 19 319 Abies and Pinus sylvestris, indicates a climatic and soil deterioration (GS3-8 Picea-Abies- 20 21 22 320 Carpinus). As a consequence, Pinus completely dominated the surrounding landscape (GS3-9 23 24 321 Pinus; Fig. 9), which is tantamount to the approaching end of the Eemian Interglacial. The 25 26 322 results of the analysis of macroremains from level GS3-M5 L MAZ (Fig. 10) indicates that the 27 28 323 29 reservoir was progressively infilling with sediments. The diaspores of aquatic plants have 30 31 324 disappeared, and peat vegetation, with numerous mosses and sedges, begins to dominate. A 32 33 325 dwarf birch appeared again in the environment, and this indicates the deterioration of climatic 34 35 326 conditions. 36 37 38 327 The cooling of the climate at the beginning of Weichselian Glaciation is confirmed by 39 40 328 the dwindling and significant withdrawal of woody communities. A characteristic element of 41 42 329 this time was Calluna vulgaris, together with numerous herbaceous plants (GS3-10 Calluna 43 44 45 330 vulgaris-NAP-Pinus); Juniperus and Pinus also appeared in this period (Fig. 9). The reservoir 46 47 331 became shallow and was dominated by Sphagnum, which created peat bog. Humid areas were 48 49 332 inhabited by tundra communities, typified by the presence of Betula nana. The depletion of 50 51 333 52 species, the drop in frequency of plant macroremains in all the ecological groups present in the 53 54 334 GS3-M6 L MAZ level and the highest amounts of sclerotia of Cenococcum geophilum in the 55 56 335 entire sequence indicate that strong processes of resedimentation were underway along the 57 58 59 60

http://mc.manuscriptcentral.com/jqs Page 15 of 57 Journal of Quaternary Science

1 2 3 336 banks of the reservoir, which explains the numerous occurrence of thermophilic trees pollen 4 5 337 6 (Fig. 10). 7 8 338 Further cooling led to the withdrawal of Pinus, as well as to the spread of a pioneering 9 10 339 and heliotropic birch-tree in a landscape dominated by herbaceous communities (GS3-11 11 12 340 Betula-NAP). Relatively stable, but cool climate, led to the development of rare pine and birch 13 14 15 341 forests mixed with herbaceous vegetation (lower part of GS3-12 Pinus-Betula, Fig. 9). The 16 17 342 numerous macroremains of dwarf birch (Betula nana) at the level GS3-M7 L MAZ indicate an 18 19 343 increase in the amount of shrub-tundra-type vegetation surrounded by a lake. The aquatic 20 21 22 344 vegetation of Potamogeton filiformis is also indicative of cold climatic conditions. The results 23 24 345 of the plant macroremains analysis in the last determined level GS3-M8 L MAZ correlate with 25 26 346 those of the palynological analysis performed in the upper part of the level GS3-12 L PAZ, 27 28 347 29 where variable values of Pinus and Betula indicate fluctuations in unstable climatic conditions, 30 31 348 but with a slight warming trend. The transition of from Najas marina and Najas flexilis aquatic 32 33 349 plant assemblages at this level may indicate a slight short-term improvement of climatic 34 35 350 conditions, while Characeae oospores denote a new deepening of the palaeolake (Fig. 10). 36 37 38 351 Palaeobotanical analyzes of other samples from section GS3 are in progress. 39 40 352 Bone remains 41 42 353 Some bones of S. kirchbergensis were found still preserved in anatomical connection; others 43 44 45 354 were disarticulated and displaced (Fig. 4F). The bones surfaces are virtually unabraded and 46 47 355 unaffected by weathering. They show no evidence of hunting or butchering by humans, nor of 48 49 356 carnivore ravaging. The teeth are heavily worn, which indicates that the individual reached a 50 51 357 52 relatively old age. Based on the size of its long bones, the individual stood about 182 cm at the 53 54 358 withers, which makes it one of the largest known representatives of the species. Its size probably 55 56 359 discouraged predators from attacking it. 57 58 59 60

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 16 of 57

1 2 3 360 The fallow deer from Gorzów Wielkopolski is the earliest known representative of D. dama in 4 5 361 6 the Polish fossil record. Up to this moment, the only known fossil remains of fallow deer were 7 8 362 known from the Lower Pleistocene (MIS 58-40) deposits of Żabia Cave, which are assigned to 9 10 363 Dama cf. farnetensis (Stefaniak, 2015). Other finds of fallow deer, mentioned in earlier 11 12 364 literature (Kowalski, 1959; Czyżewska, 1989) are reported from historical sites and are now 13 14 15 365 lost to record. The extant D. dama was introduced in Poland in the early Middle Ages as a park 16 17 366 animal (Kowalski, 1959; Czyżewska, 1989). This discovery adds to the list of Polish 18 19 367 Interglacial faunas. 20 21 22 368 Geochronological data 23 24 369 Combined optically stimulated luminescence and radiocarbon dating of the GS3 sequence (Fig. 25 26 370 3) revealed a time period ranging from the Late Saalian (ca. 140 ka) to the Mid-Holocene (ca. 27 28 371 29 5 ka). OSL results were obtained on quartz grains collected from basal, central and upper 30 31 372 inorganic horizons under, between and over, respectively, the lower and upper palaeolake 32 33 373 deposits. All OSL dated samples failed statistical P(χ2) homogeneity test, therefore they could 34 35 374 be interpreted in the context of mixture model as representing different age populations with 36 37 38 375 age dispersion for individual sample bracketing between 6.4 to 16% (Fig. 11). However, in our 39 40 376 interpretation we do refer more likely to the central age model when discussing the OSL results, 41 42 377 as the statistical approximation of each dataset, bearing in mind the OSL method limitations. 43 44 45 378 The OSL central ages are in the range between 141,4±3,0 and 79,0±1,5 ka (Tab. 6, Fig. 11) and 46 47 379 regularly grow younger toward the top of the sequence. The upper palaeolake, macrofossil- 48 49 380 bearing, organic-rich, black peat horizon laying just above the OSL-dated GdTL-2831 sample 50 51 381 52 (79,0±1,5 ka) was radiocarbon dated. This sample gave an age of 40,661 cal median years BP 53 54 382 (Tab. 5). Radiocarbon datings of four samples collected from middle and upper peat horizons 55 56 383 (Fig. 5B) of section GS4 yielded two groups of ages: the oldest (36,172 – 33,236 cal med yr 57 58 384 BP) were recorded in the middle peat and can be referred to the Mid-Pleniglacial (MIS3) phase; 59 60

http://mc.manuscriptcentral.com/jqs Page 17 of 57 Journal of Quaternary Science

1 2 3 385 the youngest Mid-Holocene ages (4,940 – 4,852 cal med yr BP) were recorded in the peat 4 5 386 6 located on top of the glacial till. The OSL datings defined the minimum age of glaciofluvial 7 8 387 sequence that deposited before formation of first, Eemian palaeolake, whereas the radiocarbon 9 10 388 ages place clearer chronological limits to the history of the evolution of the younger palaeolake. 11 12 389 Discussion 13 14 15 390 The results of lithological and geochemical studies of the sediments filling the analysed 16 17 391 sedimentary basin, as well as the radiocarbon and OSL ages and the results of the 18 19 392 palaeobotanical analysis, delineate distinct stages of evolution of the palaeolakes (see Table 7 20 21 22 393 for summary). The local stratigraphic succession starts to deposit at ca. 170-180 ka during the 23 24 394 Late Saalian (Wartanian) glaciation in a glacial/glaciofluvial environment, as indicated by the 25 26 395 oldest OSL dated quartz grains, some time before the formation of the lower palaeolake (Fig. 27 28 396 29 12). The lake began to develop in the Early Eemian coinciding with climate warming; carbonate 30 31 397 gyttja accumulated, together with a fairly significant amount of allochthonous minerals. In the 32 33 398 coastal zone, sandy material reached the basin; a quite significant proportion of it was found in 34 35 399 the lower part of section GS1 (Fig. 4B). During this phase, the lake was supplied not only by 36 37 38 400 calcium carbonate-rich groundwater but also, to a large extent, by surface waters, which brought 39 40 401 in allochthonous materials (Fig.8). The high proportion of manganese and iron oxides and 41 42 402 hydroxides in the sediments of this phase indicates that oxidising conditions prevailed both in 43 44 45 403 the coastal zone and off the lake shore. High concentrations of manganese have repeatedly been 46 47 404 reported in the sediments deposited during the early stages of development of some lakes of the 48 49 405 temperate climatic zone (see Łącka et al., 1998; Borówka and Tomkowiak, 2010). According 50 51 406 52 to Crerara et al. (1971/72), this is a characteristic phenomenon observed in shallow, well- 53 54 407 oxygenated, oligotrophic reservoirs where conditions favoured the precipitation of manganese 55 56 408 hydroxides in the form of Mn(OH)4. The quite high content of Ca in the form of CaCO3 may 57 58 409 be associated with the inflow of ground waters very rich in -HCO and ++Ca ions, derived from 59 3 60

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 18 of 57

1 2 3 410 the decalcification of glaciofluvial sediments at the bottom of the sedimentation basin. In the 4 5 411 6 initial phase of the reservoir, during the Late Saalian glaciation (MIS 6, GS3-1 L PAZ; Fig. 9), 7 8 412 these sediments deposited under cold, dry climatic conditions, when the landscapes were 9 10 413 dominated by steppe communities of Poaceae and Artemisia and tundras of Betula nana and 11 12 414 Cyperaceae. They continued accumulating for almost all the Eemian Interglacial (MIS 5e) from 13 14 15 415 E1 (GS3-2 L PAZ) to E5 (GS3-7 L PAZ; Fig. 89), under improving climatic conditions that 16 17 416 peaked in the climatic optimum, when forests were dominated by varied oak-hornbeam 18 19 417 communities. The communities of macrophytes spread in the palaeo-reservoir (GS3-M1 to 20 21 22 418 GS3-M4 L MAZs) when the early lake was still relatively deep, reaching up to approximately 23 24 419 10 m or more, and had transparent waters rich in calcium carbonate. Over time, the lake turned 25 26 420 shallow and eutrophied (Fig.10); five cycles of limnic chalk deposited along its margins, and 27 28 421 29 two in its central part. At the current stage of research, it can only be assumed that this cyclicity 30 31 422 is driven by climate. The Ca/Mg, and Na/K curves (Fig. 7, 8) show a clear predominance of 32 33 423 chemical over mechanical denudation during the accumulation of the lower limnic chalk, 34 35 424 although the upper part of both sequences shows a marked decrease in the Na/K ratio. Curves 36 37 38 425 illustrating the variability of the so-called erosion index (Na+K+Mg/Ca) for both exposures 39 40 426 show that only during the initial phase these were important processes of sedimentation, and 41 42 427 became negligible later on. This advanced stage of sedimentation correlates with the 43 44 45 428 palaeobotanical samples GS3-8 and GS3-9 L PAZs (E6 and E7, Fig. 9), which correspond to 46 47 429 the declining Eemian Interglacial when climatic conditions gradually deteriorated, and boreal 48 49 430 pine forests became dominant. The latter is the time that corresponds to level GS3-M5 L MAZ 50 51 431 52 (Fig. 10), which in fact documents a progressive progradation of the lake margins and 53 54 432 shallowing and shrinking of the palaeo-reservoir. The remains of Stephanorhinus lay in this 55 56 433 lowermost part of the section (Fig. 4F). The last phase of the Eemian lake ended up with the 57 58 434 accumulation of the ca 0.6 m of D. dama-bearing peat. In this phase in section GS 1, there was 59 60

http://mc.manuscriptcentral.com/jqs Page 19 of 57 Journal of Quaternary Science

1 2 3 435 the accumulation of ~ 5 cm of sandy-silts containing a significant amount of organic matter. It 4 5 436 6 is worth noting that in both sections GS1 and GS3 the transitional portion between the lower 7 8 437 chalk level and the layer of organic sediments is not visible (Fig. 4A), which indicates a fairly 9 10 438 rapid change in the water level, as well as in the conditions of sedimentation in the basin. The 11 12 439 pollen spectra (GS3-10 L PAZ; Fig. 9) tells us that significant climatic cooling and drying, and 13 14 15 440 thus deforestation, occurred at this time. The area surrounding the lake was then overgrown 16 17 441 mainly by heathland, steppe and tundra communities. The absence of a dense plant cover led to 18 19 442 an increase in surface flow and redeposition of sediment. The lake shrank, became shallow, and 20 21 22 443 transformed into a peat bog, consistently with the results of the macroremains analysis (GS3- 23 24 444 M6 L MAZ, Fig. 10). The lower palaeolake deposits were overlain by silt-sand sediments; at 25 26 445 the same time sands deposited in the upper part of section GS1. Based on lithology and 27 28 446 29 structural analysis, this over 10 m thick part of the sequence GS1 is interpreted as a delta 30 31 447 deposit, or an inflow cone accumulated from the southern side of the sedimentation basin. OSL 32 33 448 dating sandy inlayer sample of section GS3 gave a central age of 109,9±2,5 ka (MIS5d), which 34 35 449 could be correlated with the onset of the Weichselian glaciation (Fig. 12). This result agrees 36 37 38 450 with the palaeobotanical observations which attest to a cold climate and a landscape dominated 39 40 451 by specific herbaceous vegetation communities and with the variable spread of woody Pinus 41 42 452 and Betula communities. This horizon, therefore, documents shifting climatic conditions 43 44 45 453 towards colder contexts. This observation correlates with GS3-11 L PAZ, the lower part of 46 47 454 GS3-12 L PAZ (Fig. 9), and with GS3-M7 L MAZ (Fig. 10), which also confirms cold climate 48 49 455 with tundra vegetation. The upper part of GS3-12 L PAZ (Fig. 9) and level GS3-8m L MAZ 50 51 456 52 (Fig. 10) indicate improved climatic conditions and an increased level of the lake water. 53 54 457 The upper deposits undoubtedly indicate a return of lake sedimentation, as shown by the 55 56 458 increase in Ca content (Fig. 8A) to 250-300 mg / g (60-75% CaCO3 calculated). In contrast to 57 58 459 the lower limnic chalk, Ca in this layer of carbonate sediments is positively correlated with 59 60

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 20 of 57

1 2 3 460 general sulphur, and negatively with the content of organic matter (LOI), TOC, K, Fe, Cu and 4 5 461 6 Zn (Tab. 4). This proves that there was a pure chemical precipitation of carbonates, and no 7 8 462 contribution by aquatic vegetation. This was most probably associated with the shallowness of 9 10 463 the lake, where well-oxygenated waters favoured the precipitation of manganese hydroxides, 11 12 464 which was particularly marked at the boundary between the subzones GZ3-4a and 4b. The 13 14 15 465 presence of oxidising conditions is also confirmed by the low Fe/Mn ratio, which persists 16 17 466 throughout the period of accumulation of the upper limnic chalk layer (Fig. 8). Moving upwards 18 19 467 in sequence GS3 we meet a new set of bog sediments, which are radiocarbon-dated to 40,661 20 21 22 468 cal med yr BP. A thin layer of peat lies directly on the upper limnic chalk, and is covered by a 23 24 469 series of glacial sediments; this peat was also dated in section GS4 to about 37,7 - 32,4 ka cal 25 26 470 BP (Tab. 5). On this basis, it can be assumed that the peat and organic-mineral sediments that 27 28 471 29 rest on the upper limnic chalk accumulated in rather dry and cold climatic conditions short 30 31 472 before the last ice sheet covered the area of the Polish Lowlands (see Marks et al., 2016). The 32 33 473 chemical composition of mineral-organic sediments indicates a significant amount of erosion 34 35 474 processes and mechanical denudation during the filling of the basin. However, chemical 36 37 38 475 denudation processes were not negligible, as evidenced by the reduced, but notable amount of 39 40 476 Ca, as well as by the pattern of the curve of the Ca/Mg ratio (Fig. 8). The sedimentary basin 41 42 477 had become a periodically flooded swamp. There is a chronological anomaly between the 43 44 45 478 radiocarbon age (41,385-39,928 ka cal BP) of the upper lake sediments and the OSL-dating to 46 47 479 79,0±3,4 ka of the overlying glaciofluvial unit. This apparent age inversion between the 14C 48 49 480 and OSL datings might be explained supposing the reworking and re-deposition of the quartz 50 51 481 52 grains used for the OSL analysis. In fact, Piotrowski and Sochan (2008, 2009) and Piotrowski 53 54 482 et al. (2010) reported a radiocarbon age of 34,360 ±1,500 ka cal BP (site RAC62 in fig. 2B) for 55 56 483 the peat horizon lying over the gyttja layer that were drilled by the cartographic borehole no. 57 58 484 62 (Racław), ca. 5 km NW of Gorzów area (fig. 2B), (Fig. 3). 59 60

http://mc.manuscriptcentral.com/jqs Page 21 of 57 Journal of Quaternary Science

1 2 3 485 East of the RAC site (Fig. 2B) peats yielded a radiocarbon age > 35,0 ka BP (uncalibrated, 4 5 486 6 open). Finally, glaciofluvial sediments from Łupowo (LUP in fig.2B) have TL ages ranging 7 8 487 from 84,0 ± 12,7 to 79,2 ± 11,9 ka BP (Piotrowski et al., 2010). In the light of the radiometric 9 10 488 ages obtanied for this study, it can be speculated that a substantial amount of pre-Weichselian 11 12 489 quartz grains in the glaciofluvial deposit of section GS3 was reworked during the middle 13 14 15 490 Pleniglacial. 16 17 491 During the last stage of evolution of the study area there was the transgression and withdrawal 18 19 492 of the Weichselian ice cover towards the area of Pomerania, in the course of a time period of 20 21 22 493 ranging from about 24,500 to 17,500 years (Rotnicki and Borówka 1990, 1995a, b; Marks 2002, 23 24 494 2012; Stroeven et al., 2015). Despite the lithology of the sediments that fill the analysed 25 26 495 sedimentary basin could suggest that the rock elements were highly susceptible to shear under 27 28 496 29 the stresses imposed by the vertical pressure of ice and by the transgression of an ice sheet that 30 31 497 moved relatively fast in this area, no glaciotectonic disturbances were noted. This occurs when 32 33 498 the edge of the ice sheet pauses for a long time in a place. After some time, the pressure squeezes 34 35 499 the underlying deposits out in front of the margin of the ice sheet (Rotnicki 1976, Jaroszewski 36 37 38 500 1991). Similarly, glacier recession from the Poznań (to the south from the research area) to the 39 40 501 Pomeranian (to the north from the research area) phase occurred fairly rapidly in this area. This 41 42 502 is consistent with the recent reconstructions of the speed of deglaciation of the European 43 44 45 503 Lowland (Stroeven et al., 2015). The process of glaciation in the study area caused the 46 47 504 accumulation of a relatively thick glacial series, which in the GS3 and Racław successions (Fig. 48 49 505 3) exceed the overall thickness of 10 metres. These observations suggest that Gorzów 50 51 506 52 palaeolakes have to survived Scandinavian ice sheet advance during the Last Glacial Maximum, 53 54 507 at the same time preserving Eemian large mammal remains. Following this observation, if we 55 56 508 would assume a normal post-mortem position of rhinoceros skeleton preceding burial 57 58 59 60

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 22 of 57

1 2 3 509 processes, ice-sheet overriding fossilised lacustrine sequence could be at least partly 4 5 510 6 responsible for on-site observed Stephanorhinus kirchbergensis skeleton disarticulation. 7 8 511 9 10 512 Conclusions 11 12 513 Remains of rhinoceroses of the genus Stephanorhinus are known from many European sites, 13 14 15 514 but they generally consist of only a few bones or teeth (Billa and Zervanová, 2015). For this 16 17 515 reason, the specimen from Gorzów Wielkopolski is a significant find. Preliminary analyses of 18 19 516 this rhinoceros-bearing site indicate that the local sedimentary succession extends throughout a 20 21 22 517 period running from the Eemian interglacial and Vistulian glaciation up to the middle 23 24 518 Plenivistulian (MIS 5-3). The Gorzów Wielkopolski stratigraphic sequence was the result of a 25 26 519 virtually uninterrupted accumulation of siliciclastic and organic sediments from the decline of 27 28 520 29 the Saalian glaciation (MIS 6), through the Eemian interglacial (MIS 5e) to the Hengelo and 30 31 521 Denekamp interstadials (MIS 3). Sites with similar periods are Hengelo and Moershoofd, in the 32 33 522 eastern Netherlands, and Oerel, in Germany (Behre and Lade, 1986, Vandenberghe and Plicht, 34 35 523 2016). Remains of S. kirchbergensis had been found in other 10 Polish sites, but most are of 36 37 38 524 unknown age. Polish large mammal fauna is fairly imperfectly known. Most mammal-bearing 39 40 525 sites are caves in the Kraków-Częstochowa Upland: Nietoperzowa Cave, Ciemna Cave, Biśnik 41 42 526 Cave, Cave in Dziadowa Skała, Deszczowa Cave. The faunal assemblages from these sites are 43 44 45 527 largely dominated by cervids, but lists include the brown bear, the cave bear, the cave lion, the 46 47 528 grey wolf, the horse Equus ferus, the woolly rhinoceros, the woolly mammoth, the aurochs, and 48 49 529 the steppe bison. Polish mammal-bearing open-air sites include Imbramowice, Jóźwin near 50 51 530 52 Konin and Warsaw-Szczęśliwice, which yielded remains of four large mammal taxa, i.e., the 53 54 531 straight-tusked elephant Palaeoloxodon antiquus, the horse E. ferus, the woolly rhinoceros, and 55 56 532 S. kirchbergensis. In total there are 17 large mammal species (Kowalski, 1959; Czyżewska, 57 58 533 1962, 1989; Jakubowski et al., 1968; Borsuk-Białynicka and Jakubowski, 1972; Jakubowski, 59 60

http://mc.manuscriptcentral.com/jqs Page 23 of 57 Journal of Quaternary Science

1 2 3 534 1988, 1996; Kubiak, 1989a, b; Wojtal, 2007; Stefaniak et al. 2009; Made et al., 2014; Valde- 4 5 535 6 Nowak and Nadachowski, 2014; Stefaniak, 2015; Marciszak et al., 2017). Several species, 7 8 536 whose remains were found in western European interglacial sites (Koenigswald, 1991; Kahlke, 9 10 537 1999; Kolfschoten, 2000; Eissmann 2002; Pushkina, 2007; Musil, 2010), are missing from the 11 12 538 Polish fossil record (e.g., Equus hydruntinus, Stephanorhinus hemitoechus, Hippopotamus 13 14 15 539 amphibius, Bubalus murrensis). The new site of Gorzów Wielkopolski contributes new 16 17 540 knowledge about the interglacial fauna of Poland and Central Europe. 18 19 541 20 21 22 542 Acknowledgements 23 24 543 This research was supported by grant no. 0201/2048/18 “Life and death of extinct rhino 25 26 544 (Stephanorhinus sp.) from Western Poland: a multiproxy palaeoenvironmental approach” 27 28 545 29 financed by the National Science Centre, Poland. LiDAR DTM data presented in this study 30 31 546 were used under an academic licence no. DIO.DFT.DSI.7211.1619.2015_PL_N and 32 33 547 DIO.DFT.7211.9874.2015_PL_N awarded to the Faculty of Earth Sciences and the 34 35 548 Environmental Management University of Wrocław, according to the Polish legal regulations 36 37 38 549 of the administration of Head Office of Land Surveying and Cartography. We are also 39 40 550 immensely grateful to Paul Mazza for his constructive comments and edits to the early version 41 42 551 of this manuscript, which helped us thoroughly improve the text. 43 44 45 552 46 47 553 References 48 49 554 Aitken, M.J., 1998. An introduction to optical dating. The dating of Quaternary sediments by 50 51 555 52 the use of photon-stimulated luminescence. Oxford University Press, Oxford. 53 54 556 Alexandrowicz, S.W., Alexandrowicz, W.P., 2010. Molluscs of the Eemian Interglacial in 55 56 557 Poland. Annales Societatis Geologorum Poloniae 80, 69–87. 57 58 59 60

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 24 of 57

1 2 3 558 Berglund, B.E., Ralska-Jasiewiczowa, M., 1986. Pollen analysis and pollen diagrams. In: 4 5 559 6 Berglund, B.E., Ralska-Jasiewiczowa, M. (Eds.), Handbook of Holocene Palaeoecology and 7 8 560 Palaeohydrology. John Wiley and Sons Ltd., Chichester – New York, pp. 455–484. 9 10 561 Billa, E.M.E., Zervanová, J., 2015. New Stephanorhinus kirchbergensis (Jäger, 1839) 11 12 562 (Mammalia, Rhinocerotidae) in Eurasia – an account. Addenda to previous work. Gortania, 13 14 15 563 Geologia, Paleontologia, Paletnologia 36, 55–68. 16 17 564 Blott, S.J., Pye, K., 2001. Gradistat: a grain size distribution and statistics package for the 18 19 565 analysis of unconsolidated sediments. Earth Surf. Process. Landforms 26, 1237–1248. 20 21 22 566 Borówka, R.K., Tomkowiak, J., 2010. Skład chemiczny osadów z profilu torfowiska Żabieniec. 23 24 567 In: Twardy, J., Żurek, S., Forysiak, J. (Eds.), Torfowisko Żabieniec. Warunki naturalne, rozwój 25 26 568 i zapis zmian paleoekologicznych w jego osadach. Bogucki Prace Muzeum Ziemi, Poznań, pp. 27 28 569 29 163–172. 30 31 570 Borsuk-Białynicka, M., Jakubowski, G., 1972. The skull of Dicerorhinus mercki (Jäger) from 32 33 571 Warsaw. [In Polish] Prace Muzeum Ziemi 20, 187–199. 34 35 572 Bos, A.A.J., Bohncke, S.J.P., Kasse, C., Vandenberghe, J., 2001. Vegetation and climate during 36 37 38 573 the Weichselian Early Glacial and Pleniglacial in the Niederlausitz, eastern Germany – 39 40 574 macrofossil and pollen evidence. Journal of Quaternary Science 16, 269–289. 41 42 575 Bøtter-Jensen, L., Bulur, E., Duller, G.A.T., Murray, A.S., 2000. Advances in luminescence 43 44 45 576 instrument systems. Radiation Measurements 32, 523–528. 46 47 577 Brose, F., Luckert, J., Müller, H., Schulz, R., Strahl, J., Thieke, H.U., 2006. Das Eem von 48 49 578 Vevais – ein bedeutendes Geotop in Ostbrandenburg [The Eemian of Vevais – an important 50 51 579 52 geotope of the Eastern Brandenburg area]. Brandenburgische geowissenschaftliche Beiträge 13 53 54 580 (1/2), 155–164. 55 56 581 Bruj, M., Roman, M., 2007. The Eemian lakeland extent in Poland versus stratigraphical 57 58 582 position of the Middle Polish Glaciations [In Polish]. Biuletyn PIG-PIB 425, 27–34.Cappers, 59 60

http://mc.manuscriptcentral.com/jqs Page 25 of 57 Journal of Quaternary Science

1 2 3 583 R.T.J., Bekker, R.M., Jans, J.E.A., 2006. Digital seed atlas of the Netherlands. Groningen, 4 5 584 6 Barkhuis/Groningen University Library. 7 8 585 Crerar, D.A., Cormick, R.K., Barnes, H.L., 1971/72. Organic control on the organic 9 10 586 geochemistry of manganese. Acta Mineralogica-Petrographica 20, 217–226. 11 12 587 Czyżewska, T., 1962. Upper dentition of Dicerorhinus mercki (Jäger) from Szczęśliwice near 13 14 15 588 Warszawa, Poland. [In Polish] Acta Palaeontologica Polonica 7, (1-2), 223–234. 16 17 589 Czyżewska, T., 1989. Parzystokopytne – Artiodactyla. In: Kowalski, K. (Ed.), Historia i 18 19 590 ewolucja lądowej fauny Polski. [In Polish] Folia Quaternaria, 59-60, 209–217. 20 21 22 591 Di Stefano, G., 1996. Identification of fallow deer remains on the basis of its skeletal features: 23 24 592 taxonomical considerations. Bollettino della Società Paleontologica Italiana 34, 323–331. 25 26 593 Driesch, A., von den., 1976. A guide to the measurement of animal bones from archaeological 27 28 594 29 sites. Peabody Museum Bulletin 1, 1–136. 30 31 595 Eissmann, L., 2002. Quaternary geology of eastern Germany (Saxony, Saxon–Anhalt, South 32 33 596 Brandenburg, Thüringia), type area of the Elsterian and Saalian Stages in Europe. Quaternary 34 35 597 Science Reviews 21, 1275–1346. 36 37 38 598 Erdtman, G., 1960. The acetolysis method. Svensk Botanisk Tidskrift 54, 561–564. 39 40 599 Folk, R., Ward, W., 1957. Brazos River bar: a study in significance of grain size parameters. 41 42 600 Journal of Sedimentary Petrology 27, (1), 3–26. 43 44 45 601 Galbraith, R.F., 1990. The radial plot: graphical assessment of spread in ages. Nucl. Tracks 46 47 602 Radiat. Meas. 17, 207–214. 48 49 603 Galbraith, R.F., Roberts, R.G., Laslett, G.M., Yoshida, H., Olley, J.M., 1999. Optical dating of 50 51 604 52 single and multiple grains of quartz from Jinminum Rock Shelter, Northern 12 Australia. Part 53 54 605 I, experimental design and statistical models. Archaeometry 41, 1835–1857 55 56 57 58 59 60

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 26 of 57

1 2 3 606 Gaudzinski-Windheuser, S., Kindler, L., Pop, E., Roebroeks, W., Smith, G., 2014. The Eemian 4 5 607 6 Interglacial lake-landscape at Neumark-Nord (Germany) and its potential for our knowledge of 7 8 608 hominin subsistence strategies. Quaternary International 331, 3–38. 9 10 609 Gromova, V., 1935. On the remnants of rhinoceros Merck (Rhinoceros mercki Jaeg.) from the 11 12 610 Lower Volga. Trudy PIN RAN 4, 91–136. 13 14 15 611 Gromova, V., 1960. Determination key to mammals of USSR based on postracianl bones. [In 16 17 612 Russian.] Izdatel’stvo Akademii Nauk SSSR, Moscow. 18 19 613 Guérin, C. 1980. Les Rhinoceros (Mammalia, Perissodactyla) du Miocene terminal au 20 21 22 614 Pleistocene superieur en Europe Occidentale. Comparison avec le especes actuelles. Documents 23 24 615 du Laboratoire de Geologie de la Faculte des Sciences de Lyon 79, 423–783. 25 26 616 Guérin, G., Mercier, N., Adamiec, G., 2011. Dose-rate conversion factors: update. Ancient TL 27 28 617 29 29, 5–8. 30 31 618 Hammer, Ø., Harper, D.A.T., Ryan, P.D., 2001. Past: Paleontological Statistics Software 32 33 619 Package for Education and Data Analysis. Palaeontologia Electronica 4, (1), 1–9. 34 35 620 Hemming, S.R., 2004. Heinrich events: massive late Pleistocene detritus layers of the North 36 37 38 621 Atlantic and their global climate imprint. Review of Geophysics 42 (1), 235–273. 39 40 622 Hermsdorf, N., Strahl, J., 2008. Karte der Eem–Vorkommen des Landes Brandenburg. 41 42 623 Brandenburgische Geowissenschaftliche Beiträge 15 (1/2), 23–55. 43 44 45 624 Jakubowski, G., 1988. Finding of forest elephant – Palaeoloxodon antiquus (Falconer & 46 47 625 Cautley, 1847) in Upper Pleistocene deposits of outcrop, Jóźwin of Brown Coal Mine “Konin”. 48 49 626 [In Polish] Zeszyty Muzealne 2, 13–87. 50 51 627 52 Jakubowski, G., 1996. Forest elephant Palaeoloxodon antiquus (Falconer and Cautley, 1847) 53 54 628 from Poland. [In Polish] Prace Muzeum Ziemi 43, 85–109. 55 56 629 Jaroszewski, W., 1991. Rozważania geologiczno-strukturalne nad genezą deformacji 57 58 630 glacitektonicznych. [In Polish] Annales Sosietatis Geologorum Poloniae 61, 153–206. 59 60

http://mc.manuscriptcentral.com/jqs Page 27 of 57 Journal of Quaternary Science

1 2 3 631 Juggins, S., 2007. C2: Software for ecological and palaeoecological data analysis and 4 5 632 6 visualisation. User guide Version 1.5. www.staff.ncl.ac.uk/stephen.juggins/C2.pdf. 7 8 633 Kahlke, R. D., 1999. The history of the origin, evolution and dispersal of the late Pleistocene 9 10 634 Mammuthus – Coleodonta faunal complex in Eurasia (large mammals). Rapid Fenske 11 12 635 Companies, Rapid City, SD. 53–55, 1–219. 13 14 15 636 Koenigswald, von., W., 1991. Exoten in der Großsäuger-Fauna des letzten Interglacials von 16 17 637 Mitteleuropa. Eiszeitalter und Gegenwart 41, 70–84. 18 19 638 Kolfschoten, van. Th., 2000. The Eemian mammal fauna of central Europe. Geologie en 20 21 22 639 Mijnbouw. Netherlands Journal of Geosciences 79 (2/3), 269–81. 23 24 640 Kowalski, K., 1959. A Catalogue of the Pleistocene Mammals of Poland. [In Polish] PWN. 25 26 641 Warszawa-Wrocław. 27 28 642 29 Kozarski, S., Nowaczyk, B., Tobolski, K., 1980. Results of studies of deposits from Stary 30 31 643 Kurów near Drezdenko assigned to the Brørup Interstadial. [In Polish] Przegląd Geologiczny 32 33 644 28 (4), 210–214. 34 35 645 Kubiak, H., 1989a. Perissodactyla. In: Kowalski, K. (Ed.), History and evolution of the terrestial 36 37 38 646 fauna of Poland. [In Polish] Folia Quaternaria 59-60, 197–201. 39 40 647 Kubiak, H., 1989b. Trąbowce – Proboscidea. [In Polish.]In: Kowalski, K. (Ed.), History and 41 42 648 evolution of the terrestial fauna of Poland. [In Polish] Folia Quaternaria 59-60, 203–208. 43 44 45 649 Kuszell, T., Malkiewicz, M., 1999. Palynological profiles of the Eemian and Early Vistulian in 46 47 650 south–western Poland. Acta Palaeobotanica, Supplement 2, 487–490. 48 49 651 Lister, A. M., 1996. The morphological distinction between bones and teeth of Fallow Deer 50 51 652 52 (Dama dama) and Red Deer (Cervus elaphus). International Journal of Osteoarchaeology 6, 53 54 653 119–143. 55 56 57 58 59 60

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 28 of 57

1 2 3 654 Litt, T., Junge, F., Böttger, T., 1996. Climate during the Eemian in north-central Europe – a 4 5 655 6 critical review of the palaeobotanical and stable isotope data from central Germany. Veget. Hist. 7 8 656 Archeobot. 5: 247–257. 9 10 657 Lorek, I., 1988. The Dicerorhinus kierchbergensis (Jäger, 1839) remains from Eemian 11 12 658 sediments of Coalmine Konin (Strip Mine Jóźwin). [In Polish] Zeszyty Muzealne 2, 105–112. 13 14 15 659 Łącka, B., Starnawska, E., Kuźniarski, M., 1988. Mineralogy and geochemistry of the Younger 16 17 660 Dryas sediments from Lake Gościąż. In: Ralska-Jasiewiczowa M., Goslar T., Madeyska T., 18 19 661 Starkel L. (Eds.), Lake Gościąż, Central Poland. A monographic study. Kraków, pp. 124–128. 20 21 22 662 Made, J. van der., 2010. The rhinos from the Middle Pleistocene of Neumark-Nord (Saxony- 23 24 663 Anhalt). In: Mania D. (Ed.), Neumark Nord - Ein interglaziales Ökosystem des 25 26 664 mittelpaläolithischen Menschen (Veröffentlichungen des Landesamtes für Archäologie - 27 28 665 29 Landesmuseum für Vorgeschichte Sachsen-Anhalt), pp. 333–500. 30 31 666 Made, J. van der, Stefaniak, K., Marciszak, A., 2014. The Polish fossil record of the wolf Canis 32 33 667 and deer Alces, Capreolus, Megaloceros, Dama and Cervus in an evolutionary perspective. 34 35 668 Quaternary International 326-327, 406–430. 36 37 38 669 Marciszak, A., Kotowski, A., Przybylski, B., Badura, J., Wiśniewski, A., Stefaniak, K., 2017. 39 40 670 Large mammals from historical collections of open-air sites of Silesia (southern Poland) with 41 42 671 special reference to carnivores and rhinoceros. Historical Biology, doi: 43 44 45 672 10.1080/08912963.2017.1388377. 46 47 673 Marks, L., 2002. Last Glacial Maximum in Poland. Quaternary Science Reviews 21, 103–110. 48 49 674 Marks, L., 2012. Timing of the Late Vistulian (Weichselian) glacial phases in Poland. 50 51 675 52 Quaternary Science Reviews 44, 81–88. 53 54 676 Marks, L., Dzierżek, J., Janiszewski, R., Kaczorowski, J., Lindner, L., Majecka, A., Makos, M., 55 56 677 Szymanek, M., Tołoczko-Pasek, A., Woronko, B., 2016. Quaternary stratigraphy and 57 58 678 palaeogeography of Poland. Acta Geologica Polonica 66 (3), 403–427. 59 60

http://mc.manuscriptcentral.com/jqs Page 29 of 57 Journal of Quaternary Science

1 2 3 679 Mirosław-Grabowska, J., 2008. Reconstruction of lake evolution at Rzecino (NW Poland) 4 5 680 6 during the Eemian Interglacial and Early Vistulian on the basis of stable isotope analysis. 7 8 681 Annual Report of the Polish Academy of Science, 90–92. 9 10 682 Mol, J., 1997. Fluvial response to Weichselian climate changes in the Niederlausitz (Germany). 11 12 683 Journal of Quaternary Science 12 (1), 43–60. 13 14 15 684 Murray, A.S., Wintle, A.G., 2000. Luminescence dating of quartz using an improved single 16 17 685 aliquot regenerative-dose protocol. Radiation Measurements 32, 57–73. 18 19 686 Musil, R., 2010. The environment of the Middle Palaeolithic sites in Central and Eastern 20 21 22 687 Europe. In: Burdukiewicz, J.M., Wiśniewski, A. (Eds.), Middle Palaeolithic Human Activity 23 24 688 and Palaeoecology: New Discoveries and Ideas. Acta Universitatis Wratislaviensis 3207, 25 26 689 Studia Archeologiczne 41, 121–179. 27 28 690 29 Nalepka, D., Walanus, A., 2003. Data processing in pollen analysis. Acta Palaeobotanica 43 30 31 691 (1), 125–134. 32 33 692 Niska, M., Mirosław–Grabowska, J., 2015. Eemian environmental changes recorded in lake 34 35 693 deposits from Rzecino (NW Poland): Cladocera, isotopic and selected geochemical data. 36 37 38 694 Journal of Paleolimnology 53, 89–105. 39 40 695 Noryśkiewicz, B. 1978. The Eemian Interglacial at Nakło on the River Noteć (N Poland). Acta 41 42 696 Palaeobotanica 19 (1), 67–112. 43 44 45 697 Piotrowski, A., Sochan, A., 2008. Objaśnienia do Szczegółowej Mapy Geologicznej Polski w 46 47 698 skali 1:50 000, ark. Gorzów Wielkopolski. PIG Warszawa. 48 49 699 Piotrowski, A., Sochan, A., 2009. Szczegółowa Mapa Geologiczna Polski w skali 1:50 000, 50 51 700 52 ark. Gorzów Wielkopolski. PIG Warszawa. 53 54 701 Piotrowski, A., Krupiński, K., Krzymińska, J., 2010. Organic deposits of Eemian Interglacial 55 56 702 in Racław in Gorzów Wielkopolski district. [In Polish] In: Karnkowski, P., Piotrowski, A. 57 58 703 (Eds.), Budowa geologiczna, geologia naftowa, wody geotermalne i ochrona środowiska bloku 59 60

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 30 of 57

1 2 3 704 Gorzowa – Pojezierza Myśliborskiego. 80 Zjazd Naukowy Polskiego Towarzystwa 4 5 705 6 Geologicznego, Szczecin 11–14 września 2010 r, pp. 37–41. 7 8 706 Prescott, J.R., Stephan, L.G., 1982. The contribution of cosmic radiation to the environmental 9 10 707 dose for thermoluminescence dating. Latitude, altitude and depth dependencies. TLS II-1, 16– 11 12 708 25.Pushkina, D., 2007. The Pleistocene easternmost distribution in Eurasia of the species 13 14 15 709 associated with the Eemian Palaeoloxodon antiquus assemblage. Mammalian Review 37(3), 16 17 710 224-245. 18 19 711 Reimer, P.J., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Ramsey, C.B., Buck, C.E., 20 21 22 712 Cheng, H., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P., Haflidason, H., 23 24 713 Hajdas, I., Hatté, C., Heaton, T.J., Hoffmann, D.L., Hogg, A.G., Hughen, K.A., Kaiser, K.F., 25 26 714 Kromer, B., Manning, S.W., Niu, M., Reimer, R.W., Richards, D.A., Scott, E.M., Southon, 27 28 715 29 J.R., Staff, R.A., Turney, C.S.M., van der Plicht, J., 2013. IntCal13 and Marine13 radiocarbon 30 31 716 age calibration curves 0–50,000 years cal BP. Radiocarbon 55 (4), 1869–1887. 32 33 717 Rother, H., Lorenz, S., Börner, A., Kenzler, M., Siermann, N., Fülling, A., Hrynowiecka, A., 34 35 718 Forler, D., Kuznetsov, V., Maksimov, F., Starikova, A., 2017. The terrestial Eemian to late 36 37 38 719 Weichselian sediment record at Beckentin (NE Germany): First results from lithostratigraphic, 39 40 720 palynological and geochronological analyses. Quaternary International (in press) 41 42 721 http://dx.doi.org/10.1016/j.quaint.2017.08.009 43 44 45 722 Rotnicki, K., 1976. The theoretical basis for a model of the origin of glaciotectonic 46 47 723 deformations. Quaestiones Geographicae 3, 103–139. 48 49 724 Rotnicki, K., Borówka, R.K., 1990. New data on the age of the maximum advance of the 50 51 725 52 Vistulian ice sheet during the Leszno Phase. Quaternary Studies in Poland 9, 73–83. 53 54 726 Rotnicki, K., Borówka, R.K., 1995a. The last cold period in the Gardno-Łeba coastal Plain. 55 56 727 Journal of Coastal Research, Special Issue 22, 225–229. 57 58 59 60

http://mc.manuscriptcentral.com/jqs Page 31 of 57 Journal of Quaternary Science

1 2 3 728 Rotnicki, K., Borówka, R.K., 1995b. Dating of the upper pleni-Vistulian Scandinavian ice sheet 4 5 729 6 in the Polish Baltic middle coast. Prace PIG 149, 84–89. 7 8 730 Shpansky, A.V., 2016. New finds of Merck rhinoceros (Stephanorhinus kirchbergensis Jäger 9 10 731 1839) (Rhinocerotidae, Mammalia) in Ob area Tomsk region. Geosphere Research 1, 24–39. 11 12 732 Skompski, S. 1980. Nowe stanowiska mięczaków z osadów interglacjalnych w zachodniej 13 14 15 733 Polsce. [In Polish] Biuletyn Instytutu Geologicznego 322, 5–29. 16 17 734 Stark, P., Firbas, F., Overbeck, F. 1932. Die Vegetationsentwicklung des Interglazials von 18 19 735 Rinnersdorf in der ostlichen Mark Brandenburg. Abhandlungen des Naturwissenschaftlichen 20 21 22 736 Vereins zu Bremen, 28, 105–130. 23 24 737 Stefaniak, K., 2015. Neogene and Quaternary Cervidae from Poland. Institute of Systematics 25 26 738 and Evolution of Animals, Polish Academy of Sciences. Kraków. 27 28 739 29 Stefaniak, K., Socha, P., Nadachowski, A., Tomek, T., 2009. Palaeontological studies in the 30 31 740 Częstochowa Upland. In: Stefaniak, K., Tyc, A., Socha P. (Eds.), Karst of the Częstochowa 32 33 741 Upland and of the Eastern Sudetes: palaeoenvironments and protection. Studies of the Faculty 34 35 742 of Earth Sciences, University of Silesia, No. 56, Sosnowiec-Wrocław, pp. 85–144. 36 37 38 743 Stockmar, J., 1971. Tablets with spores used in absolute pollen analysis. Pollen et Spores 13, 39 40 744 615–621. 41 42 745 Straszewska, K., Stupnicka. E. 1979. Sites of the Quaternary lacustrine and peaty deposits in 43 44 45 746 Poland. Bulletin of the Polish Academy of Sciences 27 (3-4), 169–177. 46 47 747 Stroeven, A.P., Hättestrand, C., Kleman, J., Heyman, J., Fabel, D., Fredin, O., Goodfellow, 48 49 748 B,W., Harbor, J.M., Jansen, J.D., Olsen, L., Caffee, M.W., Fink, D., Lundqvist, J., Rosqvist, 50 51 749 52 G.C., Strömberg, B., Jansson, K.N., 2015. Deglaciation of Fennoscandia. Quaternary Science 53 54 750 Reviews 147, 91–121. 55 56 57 58 59 60

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 32 of 57

1 2 3 751 Urbański, K., Winter, H., 2005. The Eemian Interglacial in the section Radówek (Łagów 4 5 752 6 lakeland, western Poland) and its implication for till lithostratigraphy.[In Polish with English 7 8 753 summary]. Przegląd Geologiczny 53, (5), 418–424. 9 10 754 Valde-Nowak, P., Nadachowski, A., 2014. Micoquian assemblage and environmental 11 12 755 conditions for the Neanderthals in Obłazowa Cave, Western Carpathians, Poland. Quaternary 13 14 15 756 International 326-327, 146–156. 16 17 757 Vandenberghe, J., van der Plicht, J., 2016. The age of the Hengelo interstadial revisited. 18 19 758 Quaternary Geochronology 32, 21–28. 20 21 22 759 Velichkevich, F. YU., Zastawniak, E., 2006. Atlas of Pleistocene vascular plant macroremains 23 24 760 of Central and Eastern Europe, Part I – Pteridophytes and monocotyledons. Szafer Institute of 25 26 761 Botany, Polish Academy of Sciences, Cracow. 27 28 762 29 Velichkevich, F. YU., Zastawniak, E., 2008. Atlas of vascular plant macroremains from the 30 31 763 Pleistocene of central and eastern Europe, Part II – Herbaceous dicotyledons. Szafer Institute 32 33 764 of Botany, Polish Academy of Sciences, Cracow. 34 35 765 Vermeesch, P., 2009. RadialPlotter: a Java application for fission track, luminescence and other 36 37 38 766 radial plots. Radiation Measurements 44 (4), 409–410. 39 40 767 Winograd, I.J., Landwehr, J.M., Ludwig, K.R., Coplen, T.B., Riggs, A.C., 1997. Duration and 41 42 768 structure of the past four interglaciations. Quaternary Research 48, 141–154. 43 44 45 769 Winter, H., Dobracka, E., Ciszek, D., 2008. Multiproxy studies of Eemian and early Vistulian 46 47 770 sediments at Rzecino site (Łobez Upland, Western Pomerania Lakeland). Biuletyn 48 49 771 Państwowego Instytutu Geologicznego 428, 93–110. 50 51 772 52 Wojtal, P., 2007. Zooarchaeological studies of the Late Pleistocene sites in Poland. Institute of 53 54 773 Systematics and Evolution of Animals, Polish Academy of Science, Kraków. 55 56 774 57 58 775 59 60

http://mc.manuscriptcentral.com/jqs Page 33 of 57 Journal of Quaternary Science

1 2 3 776 TABLES 4 5 777 Table 1. Lithology of the GS1 profile (lower part). 6 7 778 8 Depth Height Lithology 9 (m) (m a.s.l.) 10 11 0.00 - 0.45 51.20 - 50.75 Light grey-beige silty sand 12 0.45 - 2.30 50.75 - 48.90 Light grey limnic chalk 13 14 2.30 - 3.50 48.90 - 47.70 Light brown limnic chalk with admixture of sand 15 3.50 - 9.50 47.70 - 41.70 Brown limnic chalk 16 9.50 - 11.40 41.70 - 39.80 Brown calcareous gyttja with admixture of sand 17 18 11.40 - 11.60 39.80 - 39.60 Light brown fine and medium sand 19 20 779 21 780 22 781 23 782 24 783 25 784 26 785 27 28 786 29 787 30 788 31 789 32 790 33 791 34 35 792 36 793 37 794 38 795 39 796 40 797 41 798 42 43 799 44 800 45 801 46 802 47 803 48 804 49 805 50 51 806 52 807 53 808 54 809 55 810 56 811 57 58 812 59 813 60 814

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 34 of 57

1 2 3 815 Table 2. Matrix of correlation (r) between the contents of organic matter and selected 4 816 geochemical elements in the limnic deposits of the Gorzów GS2 profile. 5 817 6 818 7 LOI Total Total Total TOC Na K Ca Mg Fe Mn Cu Zn 8 9 C N S 10 LOI 1 11 12 Total C -0,36 1 13 Total N -0,15 0,65 1 14 15 Total S 0,75 -0,30 -0,25 1 16 TOC 0,02 0,68 0,82 -0,02 1 17 18 Na 0,02 -0,54 -0,60 0,09 -0,49 1 19 K -0,38 0,59 -0,21 -0,36 -0,41 0,41 1 20 21 Ca -0,13 0,78 0,43 -0,15 0,44 -0,42 -0,65 1 22 Mg -0,20 0,43 -0,01 -0,16 0,28 0,21 -0,32 0,32 1 23 24 Fe 0,73 -0,61 -0,42 0,58 -0,34 0,28 -0,01 -0,39 -0,26 1 25 Mn -0,07 -0,46 -0,39 -0,12 -0,40 0,39 0,57 -0,42 -0,23 0,42 1 26 27 Cu -0,10 -0,43 -0,36 -0,11 -0,37 0,56 0,51 -0,34 -0,03 0,17 0,56 1 28 Zn -0,06 -0,73 -0,35 -0,07 -0,47 0,50 0,82 -0,72 -0,43 0,31 0,67 0,51 1 29 30 819 31 820 32 821 33 822 34 823 35 36 824 37 825 38 826 39 827 40 828 41 829 42 830 43 44 831 45 832 46 833 47 834 48 835 49 836 50 837 51 52 838 53 839 54 840 55 841 56 842 57 843 58 59 844 60 845

http://mc.manuscriptcentral.com/jqs Page 35 of 57 Journal of Quaternary Science

1 2 3 846 Table 3. Lithology of the GS3 profile (lower part). 4 847 5 Depth Height Lithology 6 7 (m) (m a.s.l.) 8 0.00 - 1.80 59.61 - 57.81 Dark grey sandy and silty peat with intercalations of organic silts 9 10 1.80 - 1.93 57.81 - 57.68 Black peat 11 1.93 - 4.18 57.68 - 55.43 Light grey limnic chalk 12 13 4.18 - 6.30 55.43 - 53.31 Light brown silts and sandy-silts 14 6.30 - 6.94 53.31 - 52.67 Black peat 15 16 6.94 - 10.93 52.67 - 48.68 Grey limnic chalk 17 848 18 849 19 20 850 21 851 22 852 23 853 24 854 25 855 26 856 27 28 857 29 858 30 859 31 860 32 861 33 862 34 35 863 36 864 37 865 38 866 39 867 40 868 41 869 42 43 870 44 871 45 872 46 873 47 874 48 875 49 876 50 51 877 52 878 53 879 54 880 55 881 56 882 57 58 883 59 884 60 885

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 36 of 57

1 2 3 886 Table 4. Matrix of correlation (r) between the contents of organic matter and selected 4 887 geochemical elements for the limnic chalk from the Gorzów GS3-3 profile (bold - lower chalk; 5 888 6 regular - upper chalk). 7 889 8 LOI Total Total Total TOC Na K Ca Mg Fe Mn Cu Zn 9 C N S 10 11 LOI 1 0,74 0,81 0,81 0,88 0,06 0,30 -0,74 -0,34 0,63 0,45 0,60 0,68 12 Total C 0,48 1 0,81 0,42 0,84 0,14 0,07 -0,36 -0,11 0,22 0,09 0,34 0,37 13 14 Total N 0,68 0,79 1 0,66 0,80 -0,03 0,20 -0,44 -0,08 0,61 0,56 0,47 0,55 15 Total S 0,63 0,45 0,80 1 0,76 0,01 0,44 0,83 -0,45 0,90 0,62 0,74 0,85 16 TOC 0,71 0,82 0,97 0,78 1 0,27 0,46 -0,68 -0,21 0,60 -0,28 0,66 0,74 17 18 Na -0,12 -0,59 -0,65 -0,41 -0,64 1 0,66 -0,18 0,19 0,08 -0,39 0,30 0,27 19 K -0,29 -0,62 -0,48 -0,28 -0,52 0,28 1 -0,53 0,22 0,56 0,08 0,63 0,68 20 21 Ca 0,24 0,87 0,54 0,22 0,57 -0,52 -0,51 1 0,34 -0,69 -0,27 -0,68 -0,75 22 Mg 0,05 0,54 0,16 -0,01 0,23 -0,19 -0,61 0,53 1 -0,24 -0,28 -0,31 -0,36 23 24 Fe -0,20 -0,81 -0,50 -0,14 -0,53 0,46 0,64 -0,83 -0,66 1 0,69 0,71 0,80 25 Mn -0,29 -0,49 -0,41 -0,24 -0,40 0,33 0,07 -0,39 -0,18 0,19 1 0,36 0,39 26 27 Cu 0,07 0,11 0,03 0,05 0,08 -0,19 0,32 0,25 0,02 -0,11 -0,10 1 0,79 28 Zn -0,27 -0,85 -0,58 -0,26 -0,60 0,40 0,73 -0,78 -0,66 0,85 0,25 0,04 1 29 30 890 31 891 32 892 33 893 34 35 894 36 895 37 896 38 897 39 898 40 899 41 900 42 43 901 44 902 45 903 46 904 47 905 48 906 49 50 907 51 908 52 909 53 910 54 911 55 912 56 913 57 58 914 59 915 60

http://mc.manuscriptcentral.com/jqs Page 37 of 57 Journal of Quaternary Science

1 2 3 916 Table 5. Results of radiocarbon dating, ages calibrated using OxCal INTCAL13 approach 4 917 (Reimer et al., 2013). Explanations for abbreviations: m a.s.l. (metres above sea level); m b.s.l. 5 918 6 (metres below surface level). 7 919 8 Sample Laboratory Sampling Dated material Conventional Cal 1 sigma years BP Cal 9 number code altitude and geological age (95%) median 10 11 (m a.s.l.) / context years BP years 12 depth BP 13 (m b.s.l.) 14 15 GS3-1 Beta-93676 58.7 / Macrofossils 36,010 ± 330 41,385 – 39,928 40,661 16 12.0 17 Upper 18 19 palaeolake peat, 20 just below OSL 21 22 sample 23 24 GS4-2 Poz-97161 58.1 / 9.9 Upper 32,200 ± 500 37,710 – 35,032 36,172 25 palaeolake peat 26 27 GS4-3 Poz-97162 58.0 / Upper 29,060 ± 310 33,906 – 32,402 33,236 28 10.0 palaeolake peat 29 30 GS4-4 MKL-3527 59.9 / 8.1 Peat on the 4,280 ± 110 5,280 – 5,163 (7,1%) 4,852 31 32 glacial till top 5,135 – 5,105 (1,5%) 33 5,077 – 4,526 (86,9%) 34 35 GS4-5 MKL-3526 60.0 / 8.0 Peat on the 4,330 ± 120 5,303 – 4,780 (79,6%) 4,940 36 37 glacial till top 4,770 – 4,580 (15,8%) 38 39 920 40 921 41 922 42 923 43 924 44 45 925 46 926 47 927 48 928 49 929 50 930 51 52 931 53 932 54 933 55 934 56 935 57 936 58 937 59 60 938

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 38 of 57

1 2 3 939 Table 6. Laboratory codes of investigated GS3 profile, samples depth (m b.s.l. = meters below 4 940 surface level), specific activities of natural radionuclides, dose rates, calculated mean ages for 5 941 6 all investigated fractions using CAM model. 7 942 8 9 Lab. Sampling U Th K Dose rate OSL Number of 10 Code depth (Bq/kg) (Bq/kg) (Bq/kg) (Gy/ka) central measured 11 12 (m b.s.l.) age aliquots 13 (ka) 14 GdTL-2831 10.5 4.4±0.1 6.0±0.2 192±6 0.77±0.03 79.0±1.5 19 15 16 17 GdTL-2830 16.5 4.8±0.1 4.7±0.1 206±6 0.74±0.03 109.9±2.5 58 18 GdTL-2829 21.8 8.4±0.2 6.7±0.2 257±7 0.90±0.04 141.4±3.0 47 19 20 943 21 944 22 23 945 24 25 946 26 947 27 28 948 29 30 949 31 950 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

http://mc.manuscriptcentral.com/jqs Page 39 of 57 Journal of Quaternary Science

1 2 3 951 Tabel 7. Distinct phases of the sedimentary basin development at the Gorzów Wielkopolski palaeolakes research area. 4 5 Phases of the L MAZ 6 sedimentary Lithology Geochemistry L PAZ Bone remains Hydroclimatic conditions 7 8 basin 9 development Section GS1 Section GS3 Section GS1 Section GS3 Section GS3 Section GS3 10 Phase I LLZ-2 LLZ-1a GLZ-2 GLZ-1a L PAZ-1 M1 Climate typical for park 11 (40.0-41.7 m a.s.l.) (48.6-49.2 m a.s.l.) (40.0-40.25 m a.s.l.) (48.6-49.2 m a.s.l.) Betula-NAP Characeae, Typha tundra; vast free open space 12 Calcareous gyttja Calcareous gyttja - low organic matter - increased content of K, (Betula nana, sp., Ceratophyllum resulting in intensification 13 - increased content of - sand fraction recognized content and TOC Fe, Mn, Zn Juniperus, Artemisia, demersum of mineral matter delivery 14 15 medium- to fine grained - Mz value in the range - increased content - upward decrease of Poaceae) to the lake basin; onset of 16 sand (ca. 10-20%) between 2 and 6 phi of lithophilic Fe/Mn and erosion index permafrost degradation and 17 - increase of Mz from 5.5 elements (Na, K, Ca-rich groundwater influx 18 at the bottom to 6.5 phi Mg) as well as Fe to the lake. Increase of lake 19 in top and Mg water level and growth of 20 GLZ-3 (40.25-41.0 m redox conditions in 21 a.s.l.) southernmost (deeper) part 22 23 - increased content of the lake. 24 of organic matter, 25 S and Fe 26 - decrease content Late Saalian 27 of Mn Glaciation/Early Eemian 28 (MIS 6 / MIS 5e) 29 30 Phase II LLZ-3 - LLZ-6 LLZ-1b & 1c GLZ-4 GLZ-1b & 1c L PAZ-2 - LPAZ-7 M2 - M4 Succession of plants typical 31 (41.0-50.6 m a.s.l.) (49.2-52,6 m a.s.l.) (41.0-50.9 m a.s.l.) (49.2-52.6 m a.s.l.) Betula-Pinus Najas marina for the Eemian Interglacial. 32 Calcareous gyttja Calcareous gyttja /limnic - high content of Ca, Mg, - high content of Ca, Mg, Pinus-Betula-Ulmus Typha sp. Nuphar Intensive Ca-rich 33 - in the lower part (LLZ-3 chalk Mn and total carbon Mn, TC and progressive Quercus-Fraxinus- lutea, Nyphaea groundwater inflow to the 34 & 4) initially increase of - slightly increase of mid- (TC), decrease of Fe content Ulmus alba, Brasenia sp., lake; stepwise shallowing 35 clay and decrease of silt to coarse-grained mud - decrease of Fe and S - low content of organic Corylus-Quarcus- Trapa natans, of the lake coinciding with 36 fraction content content (GLZ-4a & 4b) matter, TOC and sulphur Alnus Betula, Pinus, calcareous sediments 37 38 - above (LLZ-5 & 6) - continuous decrease of Corylus-Tilia-Alnus Alnus, Tilia, deposition, with upward 39 increase of sand content Mg (GLZ-4c, 4d & 4e) 40 41 42 43 http://mc.manuscriptcentral.com/jqs 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Journal of Quaternary Science Page 40 of 57

1 2 3 and decrease of clay and - increase of potassium Carpinus-Corylus- Carpinus, Corylus, increasing allochthonous 4 silt fraction content (GLZ-4d & 4e) Alnus Juniperus, Picea mineral deposits content. 5 - decrease of Mz and 6 increase of SD values Rhinoceros 7 8 bone remains 9 10 Deer bone 11 (uppermost part 12 of lacustrine Eemian Interglacial 13 deposits) (MIS 5e) 14 15 16 Phase III LZZ-7 LLZ-2 GLZ-5 GLZ-2 LPAZ-8 M5 Development of a valley 17 (50.6-51.3 m a.s.l.) (52.6-53.2 m a.s.l.) (50.9-51.3 m a.s.l.) (52.6-53.2 m a.s.l.) Picea-Abies- Betula, Pinus, bog evolved on the 18 Silty sand with organic matter, Peat with sand at the bottom - at the bottom (~50.9 m - high content of organic Carpinus Alnus, lacustrine sediments under 19 especially at 50.9 m a.sl. a.s.l.) higher content of matter, C, N, S and TOC Betula nana cold climatic conditions, 20 organic matter, TOC, N, - initially high content of Rapid increase of Carex completed with species 21 S and Fe Fe, afterwards pine pollen typical for tundra 22 23 - residual content of Ca decreasing gradually percentage, with at vegetation. Probable rise of 24 and increase of Na, K, - low content of Ca, Mg, the same time permafrost in the ground 25 Cu and Zn concentration Mn decreasing and interruption of 26 importance of oak, intensive feeding of the 27 hazel and hornbeam basin by a groundwater. 28 29 End of Eemian/Early 30 31 Weichselian (MIS 5d-5a) 32 Phase IV Upper part of the section (51.3 LLZ-3 GLZ-3 LPAZ-9 i 10 M6 Accumulation of mineral 33 - 60.0 m a.s.l.) (53.2 - 55.4 m a.s.l.) (53.2- 55.4 m a.s.l.) Pinus Small amount of deposits, sandy silts and 34 Fine- and medium-grained Sands with a changing grain- - - raised content of C.v. - NAP-Pinus plant macrofossils, silts in low flow conditions, 35 sands, deltaic size, intercalated with silts, lithophyte elements – - significant content Betula nana formation of delta at the 36 horizontally laminated Na, K, Mg, Cu and Zn; of NAP (30-50%) at the upper part southern margin of the 37 38 visible correlation - Co-occurrence of basin. Low groundwater 39 between their content tundra species basin recharge and high 40 41 42 43 http://mc.manuscriptcentral.com/jqs 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 41 of 57 Journal of Quaternary Science

1 2 3 and a percentage of silt- (Betula nana) and surface outwash may 4 clay fraction in the plants with higher indicate existence of 5 deposit thermal requirements permafrost in the ground. 6 - insignificant content of (i.a. Tilia) reflects 7 8 Ca that they accumulated 9 as a secondary Early Weichselian 10 deposit (MIS 4) 11 12 Phase V LLZ-4 (55.4-57.6 m a.s.l.) GLZ-4 (55.4-57.6 m LPAZ-11 i 12 M7 & M8 Reactivation of lacustrine 13 - Calcareous gyttja /limnic - a.s.l.) Betula - NAP At the beginning, basin intensively fed with 14 chalk, small amount of sandy - high content of Ca, Mg, Pinus-Betula numerous Betula Ca-rich groundwater, 15 16 silts Fe and Mn, especially at Gradually decreasing nana and Typha climatic conditions shift 17 the initial stage (GLZ- content of NAP sp. macrofossils toward optimum 18 4a) (M7), next 19 replaced with 20 freshwater plants 21 macroremains - 22 23 Characeae, Najas 24 Marina, Najas 25 flexilis (M8) 26 27 Early MIS 3 28 Phase VI LLZ-5 & 6 (57.6-59. m a.s.l.) GLZ-5 i 6 (57.6-59.6 m Deterioration of shallow 29 - Highly disintegrated peat with - a.s.l.) - - mesotrophic lake and 30 31 20 to 80% of organic matter - high content of organic valley-type periodically 32 content. Radiocarbon age matter, C, N, S and TOC, flooded peat grow 33 ~40,661 cal. BP for plants especially at GLZ-5 34 macrofossils (60 m a.s.l.) level and between 58,0 35 and 59,1 m a.s.l. Late MIS 3 36 - high and gradually (Hengelo - Denekamp ?) 37 38 increasing Na, K, Mg, 39 Cu, and Zn content, 40 41 42 43 http://mc.manuscriptcentral.com/jqs 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Journal of Quaternary Science Page 42 of 57

1 2 3 which correlates well 4 with increasing values of 5 erosion index 6 - small but important 7 8 content of Ca and Mn 9 952 10 11 953 12 13 14 954 15 955 16 956 17 18 19 957 20 21 958 22 23 959 24 25 26 960 27 28 961 29 30 962 31 32 963 33 34 35 964 36 37 965 38 39 40 41 42 43 http://mc.manuscriptcentral.com/jqs 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 43 of 57 Journal of Quaternary Science

1 2 3 966 4 5 967 6 FIGURE CAPTIONS 7 8 968 Fig.1 Location of the Eemian interglacial deposits in Poland and Eastern Germany with 9 10 969 indicated Pleistocene Scandinavian Ice Sheet maximum extents. Map explanations: black dots 11 12 970 – Eemian interglacial sites reported in the literature; red dots – sites with fossil bones findings; 13 14 15 971 green dots – sites with palynological and malacological analyses. 16 17 972 18 19 973 Fig.2 A – Location of the research area in Poland at the peak of the Last Glacial Maximum 20 21 22 974 (LGM) during the Late Weichselian (Vistulian) Scandinavian ice-sheet advance; B – LiDAR- 23 24 975 based digital elevation model showing the morphological escarpment separating the Gorzów 25 26 976 Lowland (to the north) from W-E-oriented, Toruń-Eberswalde ice-marginal valley. Please note 27 28 977 29 the network of small post-LGM dendritic basins with valleys roughly oriented to the south. 30 31 978 Sites of interest mentioned in the text: RAC62 – Borehole Racław 62, which drilled a peat 32 33 979 horizon radiocarbon-dated to 34,360 ±1,500 ka cal. BP (Lod-820); RAC –borehole with a peat 34 35 980 horizon dated to > 35,000 BP (Lod-872), LUP – Łupowo site, TL ages from 84.0 ± 12.7 to 79.2 36 37 38 981 ± 11.9 ka BP; all sites locations and data after Piotrowski et al. (2010). C – Detailed LiDAR 39 40 982 representation of the Gorzów Wielkopolski (GS1-GS4) sequences; GS3 is where the bones of 41 42 983 rhinoceros, the subject of the present study, were found. 43 44 45 984 46 47 985 Fig.3 Stratigraphic exposures (in metres) in the Gorzów Wielkopolski area as well as in the 48 49 986 Racław 62 borehole (the latter is based on data from Piotrowski et al., 2010). Legend: 1 – 50 51 987 52 Weichselian till with mixed sands and gravels; 2 – fine and medium grain sand; 3 – unsorted 53 54 988 sand mixed with gravel; 4 – silty sand; 5 – limnic chalk; 6 – gyttja; 7 – sandy silt; 8 – silt; 9 – 55 56 989 peat; 10 – OSL age ka BP; 11 – calibrated median radiocarbon age (see table 5 for details); 12 57 58 990 – bones. 59 60

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 44 of 57

1 2 3 991 Fig.4 Photographs of sections exposed at the Gorzów Wielkopolski research area. A – section 4 5 992 6 GS3: 1 – Weichselian till with mixed sands and gravels; 2 – fine and medium grain sand (OSL 7 8 993 sample GdTL-2831); 3 – upper Pleni-Weichselian peat; 4 – limnic chalk and gyttja; 5 – sands 9 10 994 and silty sands; 6 – lower peat; 7 – limnic chalk and gyttja; 8 – brown gyttja; 9 – Saalian 11 12 995 glaciofluvial sands and gravels. B – section GS1: 1 – fluvial sands (fan delta); 2 – limnic chalk 13 14 15 996 and gyttja; 3 – Saalian glaciofluvial sands. C – section GS4; 1 – organic sands and silt, locally 16 17 997 with peat; 2 – mixed glaciofluvial sands and gravelly diamicton; 3 – lacustrine sands, silt and 18 19 998 calcareous gyttja; 4 – peats; 5 – lacustrine calcareous gyttja; 6 – fine and medium grain sand; 20 21 22 999 7 – unsorted glaciofluvial sands and gravels. D – basal part of sequence GS3: 1 – grey limnic 23 24 1000 chalk and gyttja; 2 – brown gyttja; 3 – Saalian glaciofluvial sands and gravels. E – rhinoceros 25 26 1001 and deer bone-bearing section: 1 – Weichselian till with mixed sands and gravels; 2 – limnic 27 28 1002 29 chalk and gyttja; 3 – sands and silty sands; 4 – peat; 5 – grey limnic chalk and gyttja, showing 30 31 1003 the site where the bone remains were retrieved; 6 – brown gyttja; 7 – Saalian glaciofluvial sands. 32 33 1004 F – section GS3; the nearly complete skull of Stephanorhinus kirchbergensis; a malacological 34 35 1005 horizon can be recognised just above the skull. 36 37 38 1006 39 40 1007 Fig.5 Eastern (A) and western (B) geological cross-sections of the site, projected on N-S plane. 41 42 1008 Section GS3 showing where bones were found, as well as section GS4; arrows indicate 43 44 14 45 1009 sampling spots and relative C calibrated median radiocarbon ages (see tab.5). Legend: 1 – 46 47 1010 mixed glaciofluvial sands and gravelly diamicton; 2 – lacustrine lower unit (calcareous gyttja); 48 49 1011 3 – lacustrine upper unit (sand, silt and calcareous gyttja); 4 – fine and medium grain sand; 5 – 50 51 1012 52 limnic chalk; 6 – peat; 7 – unsorted glaciofluvial sands and gravels; 8 – organic sands and silts; 53 54 1013 9 – peats: (a) lower, (b) middle and (c) upper horizon. 55 56 1014 57 58 59 60

http://mc.manuscriptcentral.com/jqs Page 45 of 57 Journal of Quaternary Science

1 2 3 1015 Fig.6 (A) Map showing the location of detailed geological cross-sections (B) based on a 4 5 1016 6 geodetic survey conducted in the Gorzów Wielkopolski area, along the S3 national road 7 8 1017 excavation. Location of the GS1-GS4 sections and the calibrated median radiocarbon dated 9 10 1018 levels. For the legend of the lithostratigraphic units see Figure 5. 11 12 1019 13 14 15 1020 Fig.7 Results of the sedimentological and geochemical analyses of section GS1. See text for 16 17 1021 explanations. 18 19 1022 20 21 22 1023 Fig.8 Results of the sedimentological and geochemical analyses of section GS3 profile. See 23 24 1024 text for explanations. 25 26 1025 27 28 1026 29 Fig.9 Simplified pollen diagram of section GS3. Acronyms GS3 L PAZ are explained in the 30 31 1027 text. Legend: R PAZ – Regional Pollen Assemblage Zone, BZ – Biostratigraphic Zone; L.G. – 32 33 1028 Late Glacial; L.W. – Late Wartanian; S – Saalian. 34 35 1029 36 37 38 1030 Fig.10 Summary diagram of macrofossils from section GS3. Legend of abbreviations: f - fruit, 39 40 1031 fsc - fruit scales, sc - scales, s - seed, en - endocarp, n - needle, co - cone, sk - sclerocia, , l – 41 42 1032 leaf. The scale below diagram shows the quantitative results of the plant macroremains. Legend: 43 44 45 1033 L MAZ – Local Macrofossils Assemblage Zone, L.G. – Late Glacial; L.W. – Late Wartanian 46 47 1034 (Late-Saalian); S – Saalian. 48 49 1035 50 51 1036 52 Fig.11 Radial plots (Galbraith, 1990) of single grain quartz OSL ages for all investigated 53 54 1037 samples from section GS3 with the estimated representation of peak age for separate grains 55 56 1038 populations revealed with mixture algorithm. All scatterplots have been prepared with a 57 58 1039 RadialPlotter software (Vermeesch, 2009). 59 60

http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 46 of 57

1 2 3 1040 Fig.12 Middle- to Late Pleistocene climatic changes based on SPECMAP foraminiferal δ18O 4 5 1041 6 time series (from Winograd et al., 1997) and range of distribution of Gorzów Wielkopolski 7 8 1042 palaeolakes deposits based on 14C and OSL data, with an indication of Heinrich (H1-H6) 9 10 1043 glacial-depositional events (from Hemming, 2004). 11 12 1044 13 14 15 1045 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

http://mc.manuscriptcentral.com/jqs Page 47 of 57 Journal of Quaternary Science

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 Fig.1 Location of the Eemian interglacial deposits in Poland and Eastern Germany with indicated Pleistocene 37 Scandinavian Ice Sheet maximum extents. Map explanations: black dots – Eemian interglacial sites reported 38 in the literature; red dots – sites with fossil bones findings; green dots – sites with palynological and malacological analyses. 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 48 of 57

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 Fig.2 A – Location of the research area in Poland at the peak of the Last Glacial Maximum (LGM) during the 27 Late Weichselian (Vistulian) Scandinavian ice-sheet advance; B – LiDAR-based digital elevation model showing the morphological escarpment separating the Gorzów Lowland (to the north) from W-E-oriented, 28 Toruń-Eberswalde ice-marginal valley. Please note the network of small post-LGM dendritic basins with 29 valleys roughly oriented to the south. Sites of interest mentioned in the text: RAC62 – Borehole Racław 62, 30 which drilled a peat horizon radiocarbon-dated to 34,360 ±1,500 ka cal. BP (Lod-820); RAC –borehole with 31 a peat horizon dated to > 35,000 BP (Lod-872), LUP – Łupowo site, TL ages from 84.0 ± 12.7 to 79.2 ± 32 11.9 ka BP; all sites locations and data after Piotrowski et al. (2010). C – Detailed LiDAR representation of 33 the Gorzów Wielkopolski (GS1-GS4) sequences; GS3 is where the bones of rhinoceros, the subject of the present study, were found. 34 35 165x97mm (300 x 300 DPI) 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/jqs Page 49 of 57 Journal of Quaternary Science

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Fig.3 Stratigraphic exposures (in metres) in the Gorzów Wielkopolski area as well as in the Racław 62 46 borehole (the latter is based on data from Piotrowski et al., 2010). Legend: 1 – Weichselian till with mixed 47 sands and gravels; 2 – fine and medium grain sand; 3 – unsorted sand mixed with gravel; 4 – silty sand; 5 – limnic chalk; 6 – gyttja; 7 – sandy silt; 8 – silt; 9 – peat; 10 – OSL age ka BP; 11 – calibrated median 48 radiocarbon age (see table 5 for details); 12 – bones. 49 50 106x176mm (300 x 300 DPI) 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 50 of 57

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Fig.4 Photographs of sections exposed at the Gorzów Wielkopolski research area. A – section GS3: 1 – 46 Weichselian till with mixed sands and gravels; 2 – fine and medium grain sand (OSL sample GdTL-2831); 3 47 – upper Pleni-Weichselian peat; 4 – limnic chalk and gyttja; 5 – sands and silty sands; 6 – lower peat; 7 – limnic chalk and gyttja; 8 – brown gyttja; 9 – Saalian glaciofluvial sands and gravels. B – section GS1: 1 – 48 fluvial sands (fan delta); 2 – limnic chalk and gyttja; 3 – Saalian glaciofluvial sands. C – section GS4; 1 – 49 organic sands and silt, locally with peat; 2 – mixed glaciofluvial sands and gravelly diamicton; 3 – lacustrine 50 sands, silt and calcareous gyttja; 4 – peats; 5 – lacustrine calcareous gyttja; 6 – fine and medium grain 51 sand; 7 – unsorted glaciofluvial sands and gravels. D – basal part of sequence GS3: 1 – grey limnic chalk 52 and gyttja; 2 – brown gyttja; 3 – Saalian glaciofluvial sands and gravels. E – rhinoceros and deer bone- 53 bearing section: 1 – Weichselian till with mixed sands and gravels; 2 – limnic chalk and gyttja; 3 – sands and silty sands; 4 – peat; 5 – grey limnic chalk and gyttja, showing the site where the bone remains were 54 retrieved; 6 – brown gyttja; 7 – Saalian glaciofluvial sands. F – section GS3; the nearly complete skull of 55 Stephanorhinus kirchbergensis; a malacological horizon can be recognised just above the skull. 56 57 58 59 60 http://mc.manuscriptcentral.com/jqs Page 51 of 57 Journal of Quaternary Science

1 2 3 209x296mm (300 x 300 DPI) 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 52 of 57

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Fig.5 Eastern (A) and western (B) geological cross-sections of the site, projected on N-S plane. Section GS3 31 showing where bones were found, as well as section GS4; arrows indicate sampling spots and relative 14C 32 calibrated median radiocarbon ages (see tab.5). Legend: 1 – mixed glaciofluvial sands and gravelly diamicton; 2 – lacustrine lower unit (calcareous gyttja); 3 – lacustrine upper unit (sand, silt and calcareous 33 gyttja); 4 – fine and medium grain sand; 5 – limnic chalk; 6 – peat; 7 – unsorted glaciofluvial sands and 34 gravels; 8 – organic sands and silts; 9 – peats: (a) lower, (b) middle and (c) upper horizon. 35 36 179x129mm (300 x 300 DPI) 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/jqs Page 53 of 57 Journal of Quaternary Science

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Fig.6 (A) Map showing the location of detailed geological cross-sections (B) based on a geodetic survey 46 conducted in the Gorzów Wielkopolski area, along the S3 national road excavation. Location of the GS1-GS4 47 sections and the calibrated median radiocarbon dated levels. For the legend of the lithostratigraphic units see Figure 5. 48 49 209x285mm (300 x 300 DPI) 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 54 of 57

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 Fig.7 Results of the sedimentological and geochemical analyses of section GS1. See text for explanations. 38 39 198x190mm (300 x 300 DPI) 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/jqs Page 55 of 57 Journal of Quaternary Science

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 Fig.8 Results of the sedimentological and geochemical analyses of section GS3 profile. See text for 37 explanations. 38 39 199x186mm (300 x 300 DPI) 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 56 of 57

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Fig.9 Simplified pollen diagram of section GS3. Acronyms GS3 L PAZ are explained in the text. Legend: R 34 PAZ – Regional Pollen Assemblage Zone, BZ – Biostratigraphic Zone; L.G. – Late Glacial; L.W. – Late Wartanian; S – Saalian. 35 36 168x136mm (300 x 300 DPI) 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/jqs Page 57 of 57 Journal of Quaternary Science

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 Fig.10 Summary diagram of macrofossils from section GS3. Legend of abbreviations: f - fruit, fsc - fruit 32 scales, sc - scales, s - seed, en - endocarp, n - needle, co - cone, sk - sclerocia, , l – leaf. The scale below 33 diagram shows the quantitative results of the plant macroremains. Legend: L MAZ – Local Macrofossils Assemblage Zone, L.G. – Late Glacial; L.W. – Late Wartanian (Late-Saalian); S – Saalian. 34 35 186x139mm (300 x 300 DPI) 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/jqs Journal of Quaternary Science Page 58 of 57

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Fig.11 Radial plots (Galbraith, 1990) of single grain quartz OSL ages for all investigated samples from 18 section GS3 with the estimated representation of peak age for separate grains populations revealed with 19 mixture algorithm. All scatterplots have been prepared with a RadialPlotter software (Vermeesch, 2009). 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/jqs Page 59 of 57 Journal of Quaternary Science

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Fig.12 Middle- to Late Pleistocene climatic changes based on SPECMAP foraminiferal δ18O time series (from 29 Winograd et al., 1997) and range of distribution of Gorzów Wielkopolski palaeolakes deposits based on 14C 30 and OSL data, with an indication of Heinrich (H1-H6) glacial-depositional events (from Hemming, 2004). 31 32 115x77mm (300 x 300 DPI) 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/jqs