Abstract
The Baltic Sea has experienced marked changes in salinity over the Holocene due to a rising global sea level and the isostatic rebound of the area. Here we investigate the temporal transition of the Littorina Sea Transgression (~8000 cal BP) between the freshwater Ancylus Lake stage and the marine/brackish Littorina Sea stage using Osmium isotopes (i.e. 187Os/188Os values), supplemented by BIT index, TOC content and sedimentology. The 187Os/188Os values of surface sediments show a clear distinction between the North Sea, Kattegat Bay and Skagerrak Strait (~1) and that of the Baltic Sea Basin (~1.9 to 2.6), with more radiogenic 187Os/188Os compositions being associated with lower surface salinity values. A 187Os/188Os record determined from a 4.5 m-long core from the central Baltic (the anoxic Gotland Basin) reveals the transition from the freshwater Ancylus Lake stage into the Littorina Sea stage by a lowering of 187Os/188Os values coupled with the sharp decrease in BIT index values, and an increase in the TOC content, accompanied by the appearance of brackish water diatoms. Collectively we show that 187Os/188Os values can be applied to identify large-scale changes between freshwater and marine environments, as previously demonstrated for the Arctic Basin during the Eocene.
Introduction
The Baltic Sea Basin has had a long glaciological history being influenced by the continually shifting margins of the Fennoscandian Ice Sheet over the Quaternary (Andrén et al., 2015; Björck, 1995). At the Last Glacial Maximum, the Fennoscandian Ice Sheet covered the entire Baltic Basin (Rosentau et al., 2017). The retreat of the Fennoscandian Ice Sheet was not continuous, and over time differing rates of glacio-isostatic uplift and glacio-eustatic sea level rise have resulted in changes in basin palaeogeography that define a series of basin stages. Numerous cores from the Baltic Sea show records of these different basin stages, which include: the freshwater Baltic Ice Lake (~16–11.7 cal ka BP); the brackish Yoldia Sea (~11.7–10.7 cal ka BP); the freshwater Ancylus Lake and the brackish to marine Littorina Sea (present day conditions) separated by the Littorina Sea transgression (Andrén et al 2008, 2011; Bennike et al., 2021; Berglund et al., 2005; Björck, 1995; Jakobsson et al., 2007; Rosentau et al., 2017). The Littorina Sea transgression was time-transgressive, with marine waters being recorded earlier around Öresund (10.3 cal. ka BP) and the Great Belt (8.2 cal. ka BP) than further into the Baltic Sea Basin in Mecklenburg Bay (8.1 cal. ka BP) and the Arkona Basin (7.6 cal. ka BP; Bennike et al., 2021; Rößler et al., 2011).
During the transition from the Ancylus Lake to the Littorina Sea connections between the Baltic Basin and the palaeo-North Atlantic (via the North Sea) evolved to their present-day configurations with the main connections developing through the Danish Straits of Öresund and the Great Belt (Figure 1; Andrén et al., 2011; Jakobsson et al., 2007; Rosentau et al., 2017). Due to these altering connections, the basin changed from a freshwater lake (Ancylus Lake) to a more marine-brackish basin (Littorina Sea). The basin therefore experienced significant changes in salinity, freshwater versus marine input, temperature, productivity and sedimentation (Andrén et al., 2015; Björck, 1995; Rosentau et al., 2017), which were subsequently recorded in the geochemistry of the basin sediments. The Baltic Sea Basin generally experiences high sedimentation rates (~100–500 cm/1000 years; Andrén et al., 2015) and therefore can provide high resolution records, making it an excellent site for reconstructing longer and shorter timescale climatic oscillations and palaeoenvironments with high-temporal resolution (Andrén et al., 2015; Björck, 1995; Rosentau et al., 2017).
Map of the Baltic Sea Basin showing the locations of core P435/1-5 GC (and 303600-N – orange circle covers both core sites, exact locations found in Table 1), and surface sediment samples analysed in this study (pink circles). The pink circle labelled three covers surface sample EMB046/10-1, core top samples M86-1-07-1GC and M86-1-05-1GC (Ownsworth et al., 2024), and are labelled 3a, 3b and 3c in Table 1 and Figure 2. Colour map from ArcGIS, grey and white Europe map from Wikipedia (https://en.wikipedia.org/wiki/File:Europe_blank_map.png), and the basic geological data is adapted from Peucker-Ehrenbrink and Ravizza (1996) and Rosentau et al. (2017).
The palaeo-record of the Baltic Sea Basin has been well studied using numerous proxies (Andrén et al., 2015; Björck, 1995; Rosentau et al., 2017; Warden et al., 2017). This study applies the first use of osmium isotopes (187Os/188Os) to a sedimentary core record of the Gotland Basin in the Baltic Sea to investigate whether the 187Os/188Os composition of the sediment, which reflects the osmium isotope composition of the water column at the time of deposition, can distinguish between the freshwater Ancylus Lake stage and the brackish Littorina Sea stage. Hitherto, the utility of osmium isotopes has proved successful applied to restricted basins. For example, osmium isotopes were used to reconstruct the Messinian Salinity Crisis in the Mediterranean as it became restricted from input of waters from the Atlantic Ocean, resulting in a greater influence of the unradiogenic osmium (lower 187Os/188Os) derived from ultramafic units (ophiolites) present in the Mediterranean Sea catchment (Kuroda et al., 2016). Hence, the Messinian Salinity Crisis is recorded by a lowering of the 187Os/188Os value compared to 187Os/188Os of the open ocean at the time. Osmium isotopes have also been used to show the marine-lacustrine environment transitions of the Arctic Basin during the Eocene, with more radiogenic osmium (i.e. higher 187Os/188Os values) recorded in the Arctic Lake-stage due to the restriction from the open ocean values and hence increased influence of the more radiogenic old rocks around the Arctic Basin (Dickson et al., 2022; Poirier and Hillaire-Marcel, 2011). During the transition back to a marine setting of the Arctic Basin at 36 Ma the 187Os/188Os values decreased from ~1.3 to ~0.4–0.6 in line with the global open ocean values at that time. Osmium isotopes have similarly tracked basin evolution through relative sea-level change and glacial isostatic readjustment in northwest Scotland since the Last Glacial Maximum (Taylor et al., 2024).
Here, we show the present day 187Os/188Os measurements of surface sediments from across the Baltic Sea, Skagerrak and North Sea record the transition between marine waters in the North Sea, through brackish waters in the Danish Straits and the south of the Baltic Sea, through to more freshwater in the north and east of the Baltic Sea. Using this knowledge, we discuss the 187Os/188Os record, in conjunction with the BIT index record (Branched and Isoprenoid Tetraether index), fossilised diatoms, and core sedimentology, can provide a geochemical record of palaeoceanographic changes between the Ancylus Lake and the Littorina Sea stages in the Baltic Sea history.
Present day Baltic Sea Basin geography and geology
The Baltic Sea Basin (Figure 1) is one of the world’s largest brackish water bodies and is subdivided into several different basins and sub-basins (Andrén et al., 2015; Björck, 1995; Rosentau et al., 2017; Tuuling et al., 2011). The Baltic Sea is connected to the North Sea through the narrow and shallow straits between Sweden and Denmark (e.g. Öresund, the Great Belt Strait) and between Kattegat Bay and Skagerrak Strait. Together this system acts as an estuary (Björck, 1995; Dickens, 2013; Nordberg, 1991; Rosentau et al., 2017; Tuuling et al., 2011).
The surface salinity in the Baltic Basin is mainly controlled by varying freshwater inputs and a restricted input of marine waters from the North Sea (Andrén et al., 2011). Salinity varies between almost freshwater (0–5 PSU) in the Gulfs of Bothnia and Finland, to 6–8 PSU in the central Baltic and 10–15 PSU in the south Baltic and Danish Straits into the Kattegat (Figure 2; Björck, 1995; Ducrotoy and Elliott, 2008; Ehlin, 1981; Jiang et al., 1997; Melvasalo et al., 1981; Omstedt et al., 2004; Peucker-Ehrenbrink and Ravizza, 1996; Rosentau et al., 2017). The Baltic Sea drains a large area of 1.6 million km2 (Andrén et al., 2015; Björck, 1995), with half of the freshwater inputs originating from rivers draining south/southeast of the basin, and half draining the Precambrian Shield in the north (Peucker-Ehrenbrink and Ravizza, 1996).
(a) The 187Os/188Os data for surface samples in this study (pink circles), core top sample P435/1-5 GC, core top samples M86-1-07-1GC and M86-1-05-1GC (Ownsworth et al., 2024; inset (b)), and Mn nodules (green circles; Peucker-Ehrenbrink and Ravizza, 1996) across the Baltic Sea, Skagerrak Strait and the Kattegat. The 187Os/188Os values are also shown for continental runoff from Finland, south-Sweden, Denmark and northern Germany derived from lake ore and bog ore Fe-Mn oxidates (dark grey values; Peucker-Ehrenbrink and Ravizza, 1996). Estimated salinity across the study area is plotted in purple (PSU), which is visualised by sea surface salinity measurements at respective areas divided by dashed lines (Sjöqvist et al., 2015). These salinity gradients also agree with data from other studies and have not changed since measurements started in the early 1900s (e.g. Ehlin, 1981; Melvasalo et al., 1981; Omstedt et al., 2004). Map sourced from ArcGIS.
The north, northeast and west of the Baltic Sea Basin is characterised by crystalline basement or ‘shield’ (Precambrian: Late Proterozoic – Late Archean), whereas the south and east is dominated by the European sedimentary platform of Phanerozoic age, and a small part of south-western Baltic Sea lies on the west European sedimentary platform of Palaeozoic age, and comprises of mainly calcareous, clastic, sandstone and clay rocks and deposits (Figure 1; Björck, 1995; Peucker-Ehrenbrink and Ravizza, 1996; Rosentau et al., 2017).
Materials and methods
Sampling
Surface sediments were obtained from locations across the Baltic Sea, Kattegat, Skagerrak and North Sea (Figure 1; Table 1) by using a multi-corer. Samples were dried and shelly detritus and larger clasts >2 mm were removed by sieving before powdering.
Details of surface sediment samples and the P435/1-5 GC and 303600-N cores of the central Baltic.
Two sediment gravity cores were obtained from the anoxic Gotland Basin in the central Baltic. The 11 m-long core P435/1-5 GC (56°57.954°N, 19°22.210°E, 178 m depth) was collected on research cruise RV ‘Poseidon’ in 2012 (Figure 1; Table 1). The upper 4.5 m was analysed for Re-Os abundance and isotopic composition; analysis of the TOC content and BIT index was restricted to the 2 to 4.5 m interval of the core. Core 303600-N was obtained from a nearby location ~35 km away (56°55.01°N, 19°19.99°E, 170 m depth; Figure 1; Table 1), which had been previously utilised for TOC content and BIT index values (Warden et al., 2017, 2018). Here, both cores were used for the study of diatom assemblages, with the two cores correlated based on TOC and BIT index records (Supplemental Figure 1).
Determination of the TOC content
Total carbon (TC) and total inorganic carbon (TIC) for core P453/1-5 GC was determined at ~3 cm intervals using 100 mg of freeze-dried sediment that was diluted with 40% H3PO4 and incinerated at 1200°C on a Multi Elemental Analyser 4000 from Analytik Jena at the Leibniz Institute for Baltic Sea Research, Warnemünde. Total organic carbon (TOC) was calculated based on TOC (wt. %) = TC (wt. %) – TIC (wt. %).
Rhenium-osmium (Re-Os) isotope analysis
Twelve surface sediments (0–1 cm) and twenty 1-cm stratigraphic thick sediment horizons from core P435/1-5 GC were analysed for their Re-Os chemistry. The sediment samples were dried in air at 50°C overnight and then powdered using an agate pestle and mortar. The Re-Os analysis is based on established methods that utilises a chemical digestion technique which preferentially liberates the hydrogenous Re and Os from organic matter, thus avoiding incorporation of Re and/or Os possibly held in detrital silicate (Selby and Creaser, 2003). Briefly, ~1 g of powdered sample is sealed into a Carius tube with 8 ml CrO3 4N H2SO4 and tracer solution (enriched in 185Re and 190Os), which was then heated at 220°C for 48 h. The Carius tubes were opened, and a solvent extraction was undertaken using chloroform (CHCl3) to separate the Os for back extraction into HBr. Purification of the Re fraction is achieved using a 5N NaOH – acetone extraction, and HCl-HNO3 anion bead chromatography. The Os fraction was further purified by a H2SO4-CrO3-HBr microdistillation. The purified Os and Re fractions were loaded on Pt and Ni filaments, respectively (Selby and Creaser, 2003), with the addition of ~0.5 µl BaNO3 and BaOH activator solutions, respectively. The isotope compositions of the purified Re and Os fractions were measured using negative thermal ionisation mass spectrometry (NTIMS; Creaser et al., 1991; Völkening et al., 1991) using a ThermoScientific TRITON mass spectrometer via faraday cups for Re and electron multiplier (SEM) in peak hopping mode for Os. Osmium isotopic ratios were calculated relative to 188Os and corrected for mass fractionation using a 192Os/188Os value of 3.08261 (Nier, 1937). The oxide corrected 185Re/187Re was normalised using a 185Re/187Re value of 0.59738 (Gramlich et al., 1973).
Total procedural blanks for Re and Os during this study are 19.04 ± 4.67 pg and 0.08 ± 0.05 pg, respectively, with a 187Os/188Os of 0.27± 0.20 (1 S.D, N = 8). The percent blank correction ranges between 0.02 and 0.68 (Re) and 0.01 and 0.70 (Os). Uncertainties for 187Re/188Os and 187Os/188Os are determined through full propagation of uncertainties in Re and Os mass spectrometer measurements, blank abundances and isotopic compositions, spike calibrations and reproducibility of standard Re and Os isotopic values. Given the young age of the sediment samples analysed, the correction for any ingrowth of radiogenic 187Os to the measured present-day 187Os/188Os composition is less than that of the minimum 187Os/188Os uncertainty (~0.01; Table 2). As such, the measured present-day values are presented and discussed (Table 2). In-house standard solutions (DROsS 50 pg and Re Std 125 pg) are run repeatedly throughout each batch of samples to monitor mass spectrometer reproducibility. The Re standard yielded an average 185Re/187Re value of 0.59834 ± 0.00063 (1 S.D., N = 25). The DROsS (Durham Romil Osmium Standard) osmium solution yielded average 187Os/188Os values of 0.16080 ± 0.00025 (1 S.D., N = 25). The isotope compositions of these Re and Os solutions are consistent with uncertainty to those published by Nowell et al. (2008) and previous studies (e.g. Taylor et al., 2025).
Synopsis of the Re-Os data for core P435/1-5 GC and surface sediment samples. Uncertainty is ± 2SE.
Data of surface sediment samples M86-1-07-1GC and M86-1-05-1GC from Ownsworth et al. (2024).
Diatom assemblages
The diatom assemblage of cores P435/1-5 GC and 303600-N was studied at the Palaeo-oceanology Unit at the Faculty of Geosciences, University of Szczecin focusing on the identification of the freshwater Ancylus Lake to the marine-brackish Littorina Sea transition. In core P435/1-5 GC between 4.5 and 2 m, 1 cm-thick sediment horizons were analysed every 9–12 cm. The diatoms present were classified as either brackish or freshwater. Sample preparation follows previously published methods (Harff et al., 2011; Kotrys et al., 2014). To identify the depth position of the Ancylus Lake to Littorina Sea transition (cf e.g. Andrén et al., 2000) in 303600-N we use the percentages of the marine planktonic species Pseudosolenia calcar-avis (Schultze) Sundström to the freshwater taxa Aulacoseira spp. from the lowermost part of core 303600-N. The palaeo-ecological preferences of these two species groups are presented in previous publications (e.g. Andrén et al., 2000; Harff et al., 2011; Kaiser et al., 2016; Kotrys et al., 2014).
BIT index
The BIT index records the abundance of certain branched glycerol dialkyl glycerol tetraethers (bGDGTs) obtained from terrestrial organisms, compared to an isoprenoid GDGT called ‘crenarchaeol’ which marine Archaea produce (Kim et al., 2006; Sinninghe Damsté et al., 2002). This therefore allows basic changes between a comparatively more terrestrial (value of 1) versus more marine environment (value of 0) to be inferred.
Sediment samples from core P453/1-5 GC were freeze-dried, ground, homogenised and extracted using Dionex™ accelerated solvent extraction (ASE). Analysis of glycerol dialkyl glycerol tetraethers (GDGTs) and BIT (Branched and Isoprenoid Tetraether) index calculation was performed as described previously (Warden et al., 2018).
Results
To understand the 187Os/188Os geochemistry of the Baltic Basin and to apply this in the reconstruction of its palaeoceanographic evolution through to the Holocene, we investigated (i) a set of surface sediments from across the Baltic Sea, Kattegat, Skagerrak and North Sea, and (ii) a sediment core from the anoxic Gotland Basin covering the Ancylus Lake and Littorina Sea stages of the Baltic Sea.
The 187Os/188Os values of surface sediments
To characterise the present-day 187Os/188Os values along a marine to brackish to relatively freshwater transition, 12 surface sediments were analysed from across the Baltic Sea, Kattegat, Skagerrak and North Sea (Figure 1; Table 1). These sediments comprised a mix of sandy, gritty, muddy to clayey sediments with occasional shelly detritus and larger clasts > 2 mm. The Re and total Os abundances vary between 0.81 and 4.20 ppb, and 16 and 167 ppt, respectively (Table 2). The 187Os/188Os values range between 0.93 and 2.25 (Table 2; Figure 2). Collectively in 187Os/188Os versus 1/192Os space no appreciable trend is observed, even removing the single outlier with a high 1/192Os value (sample 242970-4; Figure 3a). The 187Os/188Os values in the surface sediments of the Skagerrak and off western Norway (samples 1–3) are ~1. Moving through the Kattegat into the southwest of the Baltic Sea values increase from ~1.25 to 1.67 (samples 4–9). In the mid Baltic Sea and northern Baltic Sea surface sediment values increase to 1.88 and 1.95 and then further to 2.10 and 2.24, respectively (samples P435/1-5 GC - 10 and 11–12, respectively).
(a) The 187Os/188Os values for the surface sediments analysed in this study (R2 = 0.76) and Mn nodules from the seafloor (R2 = 0.81; surface sediment and Mn nodules together R2 = 0.73; Peucker-Ehrenbrink and Ravizza, 1996) plotted against sea surface salinity (PSU; Sjöqvist et al., 2015). Also shown is a binary mixing line between salinity and 187Os/188Os based on salinity end members of 30 and 0.1 PSU and 187Os/188Os end members of 1 and 2.6 of this study and Peucker-Ehrenbrink and Ravizza (1996), and an average water osmium abundance of 10 pg/kg (Rooney et al., 2024 and references therein); (b) 187Os/188Os versus 1/192Os plot for present-day surface samples; (c) 187Os/188Os versus 1/192Os plot for core P435/1-5 GC samples. See text for discussion.
Holocene sedimentary record
To understand 187Os/188Os changes through the Holocene, core P435/1-5 GC was used to create a record over time of 187Os/188Os, TOC, diatom assemblages and BIT index. This is then further supplemented with previously published data on core 303600-N including the TOC content (Warden et al., 2017) and BIT index values (Warden et al., 2018).
Rhenium and osmium abundances and 187Os/188Os values
The Re and Os abundances vary between 0.4 and 12.9 ppb and 57 and 224 ppt, respectively within P435/1-5 GC (Figure 4; Table 2). Between 450 and 240 cm Re and Os abundance remain relatively similar with an average of ~1 ppb and 80 ppt, respectively. Rhenium and Os abundances rise to a peak at 200 cm of 12.5 ppb and 224 ppt, respectively, before decreasing back to ~110 ppt Os and 10 ppb Re up to the core top, staying more elevated compared to the deeper part of the core. The 187Os/188Os values range between 1.87 and 2.76. Between 450 and 240 cm the 187Os/188Os values are the highest, largely >2.5, except for a drop to 2.2 at 430 cm and to 2.4 at 386 cm. The 187Os/188Os values then decrease from 2.5 to 2.0 between 220 and 165 cm. This is followed by a more gradual decrease in 187Os/188Os values from 2.0 to 1.9 between 165 cm and the core top. As shown for the surface samples, collectively no appreciable trend is observed in 187Os/188Os versus 1/192Os space (Figure 3b).
Fresh and brackish diatoms, 187Os/188Os, BIT index, total Os (ppt), Re (ppb) and TOC (%) data for core P435/1-5 GC against depth. The lighter coloured lines for the records of the BIT index and TOC represent core 303600-N data from Warden et al. (2018). Diatom data for core 303600-N is also shown with freshwater diatoms (Aulacoseira spp %) in light blue and brackish diatoms (Pseudosolenia calcar-avis) in dark blue solid lines (ranging 182–198 cm depth). For total Os, Re and 187Os/188Os the symbol size is greater than the uncertainty of the data (see Table 2). Dashed lines show the division between the two lithofacies. Cores have been correlated together based on TOC and BIT index (Supplemental Figure 1).
TOC content
Between 450 and 240 cm the TOC values are low, between 0.4% and 1.5%. At 240 cm the TOC values begin to increase quickly to a peak of 15% at ~220 cm and then fluctuate between 4% and 12% to the core top as seen in the data from core 303600-N (Figure 4; Table 3).
Total carbon (TC), total inorganic carbon (TIC) and total organic carbon (TC) for core P453/1-5 GC.
Diatom assemblages
Core P435/1-5 GC displays limited occurrences of diatoms (blue ovals; Figure 4; Table 4). Freshwater diatoms are recorded at depths of 370.5 and 358.5 cm. No diatoms are found between the depths of 350 and 320 cm. Freshwater diatoms are recorded between 310 and 210 cm. The occurrence of brackish diatoms is then noted at 199.5 cm. The diatom assemblage of core 303600-N (solid blue lines) shows a decrease in the freshwater taxa Aulacoseira spp from 76% to 3% at a depth of ~210 cm and a sudden increase in the marine planktonic species Pseudosolenia calcar-avis from 1% to 37% at ~205 cm (Figure 4, Table 5).
Appearance of fresh and brackish diatoms in core P435/1-5 GC.
Percentages of freshwater and brackish diatoms in core 303600-N.
BIT index
The BIT index varied substantially in the 200–450 cm section of core P435/1-5 GC (Figure 4; Table 6). In general, values markedly increased from <0.1 at 200 cm to a peak value of 0.86 at 286 cm. At greater depth the BIT values decreased but remained high (average ca. 0.4) with a second maximum of 0.65 at 364 cm, and two dips to 0.14 and 0.22 at 391 and 316 cm, respectively. The general profile of the BIT index record is similar to that obtained previously for the adjacent 303600-N core (Warden et al., 2018; Figure 4).
BIT index values for core P435/1-5 GC.
Discussion
187Os/188Os composition characteristics of the present-day Baltic Sea basin
Understanding of the present-day environment and sediment geochemistry in the Baltic Sea is key to investigating the evolution of the Baltic Sea Basin through the Late Quaternary and Holocene. This study focusses on the use of osmium isotopes to infer the water 187Os/188Os composition (Cohen et al., 1999; Ravizza and Turekian, 1989) together with sedimentological, diatom analysis, BIT index and TOC abundance of the Baltic Basin during the transition of the Ancylus Lake into the Littorina Sea. Establishing the present-day 187Os/188Os signature of the Baltic and North Sea is critical to assess the potential of 187Os/188Os to track changes in water mass exchange through time. The application of 187Os/188Os chemostratigraphy allows the differentiation between long- and short-term changes in the environment due to the relatively short residence time of osmium in the oceans of 1–50 kyrs, with many studies noting a residence time of <10 kyrs (Levasseur et al., 1999; Ownsworth et al., 2023, 2024; Oxburgh, 1998, 2001; Peucker-Ehrenbrink and Ravizza, 2012; Rooney et al., 2016; Sharma et al., 1997; Taylor et al., 2024). The 187Os/188Os signature of present-day oxic open seawater has an average value of 1.060 ± 0.005 and is relatively similar across the global oceans (Burton et al., 1999; Cohen, 2004; Levasseur et al., 1998; Peucker-Ehrenbrink and Ravizza, 2000; Rooney et al., 2016; Sharma et al., 1997). Spatial and temporal variations in 187Os/188Os can exist that reflect local influences in the ocean chemistry due to variations in terrestrial sources, especially in more isolated/restricted seawater basins (Burton et al., 2010; Chen et al., 2009; Gannoun and Burton, 2014; Kuroda et al., 2016; Ownsworth et al., 2023, 2024; Paquay and Ravizza, 2012; Peucker-Ehrenbrink and Ravizza, 2012; Plint et al., 2025; Poirier and Hillaire-Marcel, 2011; Rooney et al., 2016; Taylor et al., 2024, 2025). The surrounding geology and variations in flux of eroded material, as well as marine influence is therefore important in understanding the 187Os/188Os of sediments through a core section, and thus by inference the 187Os/188Os of the water column at the time of deposition.
The Baltic Sea catchment is dominated by the Precambrian crystalline basement (Baltic Shield; Figure 1) comprising rocks and runoff with a highly radiogenic 187Os/188Os signature (~3.6 to 7.9; Figure 2; Pegram et al., 1992; Peucker-Ehrenbrink and Ravizza, 1996). Freshwater runoff from the Baltic Shield therefore delivers a radiogenic signal into the Baltic Basin, and potentially an even greater radiogenic signature due to preferential weathering of felsic matrices that possess a more radiogenic 187Os/188Os composition (Peucker-Ehrenbrink and Blum, 1998). This is demonstrated by 187Os/188Os values of surface sediments (this study) and of Mn nodules of the Baltic Basin (Pegram et al., 1992; Peucker-Ehrenbrink and Ravizza, 1996; 187Os/188Os = ⩾~1.5 to 2.6; Figure 2).
The transition from the Kattegat through the Baltic Sea possesses present-day 187Os/188Os compositions ranging from 1.3 in Kattegat Bay through to 2.6 in the Gulf of Bothnia (N = 4 Mn nodules; Pegram et al. (1992, Peucker-Ehrenbrink and Ravizza (1996, N = 10 surface sediments; this study; Figure 2). Manganese nodules have been extensively used to accurately record the 187Os/188Os composition of the water column from which they formed, as have organic-bearing sediments (Peucker-Ehrenbrink and Ravizza, 1996, 2012; Peucker-Ehrenbrink and Ravizza, 2000; Rooney et al., 2024 and references therein). Direct comparison of Mn nodule and sediment 187Os/188Os compositions and by inference the water column 187Os/188Os can only be evaluated for samples that are spatially associated and occur in water columns of similar salinity. In this case, location 7 and 8 of this study are nearby the Mn nodule sample of Mecklenburg Bight and possess similar 187Os/188Os values (Location 7 = 1.49; location 8 = 1.53; Mn nodule = 1.49; Figure 2). Although spatially distinct, the Mn nodule value of the Bothnian Sea is more radiogenic than the closest (~125 and 160 km) sediment surface samples (2.50 vs 1.95, 2.10, 2.24; Figure 2). Although all similarly radiogenic, the slightly lower surface sediment 187Os/188Os values may record local variations in the osmium isotope composition of the crustal source input and/or anthropogenic influence (Chen et al., 2009) in comparison to the Mn nodule which is more centrally positioned in Bothnian Sea and was archived before 1950 (Peucker-Ehrenbrink and Ravizza, 1996). Additionally, variation between Mn nodule and sediment 187Os/188Os values may relate to slow accumulation and diffusion rates of osmium associated with Mn nodules (Peucker-Ehrenbrink and Ravizza, 2000 and references therein).
The 187Os/188Os compositions of the surface sediment for the Baltic Sea in general are significantly more radiogenic than those from the North Sea and Skagerrak (~0.93–1.10, N = 5 surface sediments; this study and Ownsworth et al., 2024; Figure 2). The North Sea and Skagerrak surface samples record 187Os/188Os values (0.93–1.10; Figure 2) that are similar and slightly lower than the average open ocean value (~1.06). The slightly less radiogenic value of 0.93 could be due to several factors including influx of less radiogenic osmium from coastal riverine systems and potentially localised anthropogenic pollution, which commonly has an unradiogenic isotopic signature (e.g. Chen et al., 2009). For example, the Alcoa Lista Al and Eramet Porsgrunn Mn smelters occupy coastal locations nearby the sampling site (e.g. location 2 which exhibits the lowest 187Os/188Os value of 0.93; N = 2; Table 6). The more radiogenic 187Os/188Os composition moving from the Skagerrak through the Kattegat into the Baltic Sea from the narrow straits around Denmark to the south-central Baltic Sea coincides with the present day decreased marine influence from the North Sea. This is shown by the decrease in surface salinity from normal marine values in the North Sea (30–35 PSU) to 30–15 PSU through the Skagerrak and Kattegat, 10 PSU in the south Baltic Sea, 8–6 PSU in the Central Baltic and then almost fresh water at 0.5 PSU in the outer Gulf of Finland and Gulf of Bothnia (Figure 2; Björck, 1995; Dickens, 2013; Ducrotoy and Elliott, 2008; Ehlin, 1981; Jiang et al., 1997; Melvasalo et al., 1981; Peucker-Ehrenbrink and Ravizza, 1996; Rosentau et al., 2017). The decreasing salinity indicates an increased influence of freshwater terrestrial runoff, and hence a predominance in the supply of terrestrially sourced radiogenic material from the Precambrian Baltic Shield, which possesses 187Os/188Os between 3.6 and 7.9 (Pegram et al., 1992; Peucker-Ehrenbrink and Ravizza, 1996).
The 187Os/188Os values show a moderate increase between the North Sea and the Danish Straits. From the Danish Straits the 187Os/188Os values become progressively more radiogenic through the central and northern regions of the Baltic Sea, corresponding to the reduction in marine influence through the narrow Danish Straits (Figure 3c). There is no clear linear relationship in 187Os/188Os versus 1/192Os space (Figure 3a), that is, a mixing line between an open seawater and terrestrial derived osmium isotope signal. This might be, in part, because this data presentation does not take into consideration that the uptake of osmium from the water column (i.e. hydrogenous osmium i.e. associated with organic matter), is highly variable, being a function of the depositional environmental conditions and the organic matter type (Harris et al., 2013). Moreover, it could also be related to the transport of water through the Danish Sounds which takes place in the upper 10–15 m layer. During the outward phase, relatively freshwater masses move into the Belt Sea, where the water becomes mixed vertically with saline masses (Ehlin, 1981). During the next period of inflow, the water mass returns to the Baltic Sea. Based on the data of this study and Peucker-Ehrenbrink and Ravizza (1996) the 187Os/188Os composition inflowing into the Bornholm Basin from the Belt Sea would have a more radiogenic 187Os/188Os composition (~1.5; Figure 2) than that of the open ocean (~1.0; Figure 2). In addition, starting from the Bornholm Basin the basins buffer circulation, with the Gulfs of Bothnia and Riga being topographically isolated from the saline water below the halocline and only receiving surface water (Elken and Matthäus, 2008). The latter (i.e. non binary mixing) is further supported by the 187Os/188Os data not falling along a simple binary mixing model (Figure 3c) that predicts a linear relationship between salinity and osmium isotope composition using salinity end members of 30 and 0.1 PSU and 187Os/188Os end members of 1 and 2.6 of this study and Peucker-Ehrenbrink and Ravizza (1996), and an average water osmium abundance of 10 pg/kg (Rooney et al., 2024 and references therein).
Based on the modern-day sediment signature, 187Os/188Os compositions of Holocene sediments from the core within the southern Gotland Basin should identify periods of isolation from a marine signal (i.e. from the North Sea) with a relative increase in the flux of terrestrially eroded material into the basin and reduced dilution by marine incursion. This would be similar to the present-day 187Os/188Os composition of the northern and eastern extremities of the Baltic Sea, and thus starkly different to that of the Late Pleistocene and Holocene global ocean (ca. 1; Paquay and Ravizza, 2012; Rooney et al., 2016). Thus, the sediment archive of core P435/1-5 GC should record the palaeoceanographic changes in the Baltic Sea basin.
The transition of the Ancylus Lake to Littorina Sea
Based on the obtained data of this study, sedimentological features of core P435/1-5 GC and 303600-N together with published data (Warden et al., 2017, 2018), the sediments have been divided into two lithofacies (Figure 4). Lithofacies 1 encompasses the deepest part of the core from ~450 cm to a depth of ~220 cm. Here the sediments are a mixture of mainly clay with some silt and sand, and with some disturbed laminations of varying thickness (mm-cm scale) and some more massive sections. Here TOC, Re and Os abundances are low and stable (2–6%, 0.6–2 ppb and ~60–136 ppt, respectively), whilst 187Os/188Os and BIT index values are high (2.24–2.76, and 0.15–0.82, respectively), and freshwater diatoms are present. Lithofacies 2 encompasses the section of core between ~220 and 0 cm. Here the sediments are homogenous, with very fine and even laminations, TOC, Re and Os abundances are at their highest (3–15%, 7.7–12.5 ppb and 111–224 ppt, respectively), and 187Os/188Os and BIT index values have decreased to their lowest (1.87–2.25, and 0.03–0.06, respectively), and the first appearance of brackish diatoms are noted at the start of lithofacies 2. The point at which the freshwater diatoms disappear, and the brackish diatoms appear is utilised as the main factor in determining the boundary of lithofacies 1 and 2 at 205 cm. The below discussion focusses on the transition between the two lithofacies.
As with many other cores from the Baltic region, the P435/1-5 GC core used in this study lacks any datable fossil material in the pre-Littorina sediments (Hyttinen et al., 2014; Kögler and Larsen, 1979; Kortekaas, 2007; Moros et al., 2002). Thus, any discussion of the ages of sediments in core P435/1-5 GC below are poorly constrained and can only be inferred from correlating changes in sedimentology (which, basin-wide have been attributed to the different stages) and any accompanying geochemical changes (187Os/188Os, BIT index) to sediments of the Baltic Sea stages. The sedimentology of the Ancylus Lake stage is characterised by grey clays of lacustrine origin with more heterogenous sediments and occasional disturbed laminations of varying thickness (Kögler and Larsen, 1979; Rosentau et al., 2017). The Littorina Sea sediments are characteristically laminated organic-rich alternating muds and clay-gyttjas with homogenous laminations (Moros et al., 2002, 2020; Warden et al., 2017, 2018; Zillén et al., 2008). The appearance of brackish diatoms (and high TOC values) in Baltic Sea sediments marks the establishment of brackish marine conditions. Although no absolute age constraints can be placed on the sediments of core P435/1-5 GC, the transition to brackish conditions signified by diatoms and TOC has been dated in other cores close to 8 ka (Moros et al., 2020; Zillén et al., 2008).
Ancylus Lake
The transition from the marine Yoldia Sea stage to the freshwater Ancylus Lake stage was driven by glacio-isostatic uplift following deglaciation of the Fennoscandian Ice Sheet. Due to isostatic uplift, channels connecting the Baltic Basin to the Skagerrak and into the North Sea over south-central Sweden closed, allowing the build-up and development of the freshwater Ancylus Lake (Andrén et al., 2002, 2011; Berglund et al., 2005; Björck, 1995; Rosentau et al., 2017). Between ~400 and 220 cm the sedimentological record in core P435/1-5 GC (lithofacies 1) is dominated by heterogenous laminations of varying thicknesses with some more massive units. The 187Os/188Os values increase from 2.2 at 425 cm to a peak of 2.8 at 350 cm with values remaining between 2.6 and 2.7 from 425 to 240 cm (Figure 4). These values are more radiogenic than the northernmost present-day Baltic Sea sediment sample values (2.1–2.3), representing the most freshwater influenced values of the present Baltic Sea. The 187Os/188Os values of 2.6 to 2.8 are, therefore, likely to represent a much lower marine influence than the present-day, with greater influence of run-off and meltwater from the Fennoscandian Ice Sheet moving over the Baltic Shield. Coincident with the radiogenic 187Os/188Os values, the BIT index is higher at >0.3 with intervals reaching 0.6 and 0.8 at 364 and 286 cm, respectively. These BIT values indicate a greater flux of terrestrial/freshwater material and a more continental input (Kim et al., 2006; Wang et al., 2013; Warden et al., 2018). The TOC levels are consistently very low (<0.5%) from 400 to 280 cm (except for a minor peak at 380–390 cm), then gradually increase to 1.5% by 230 cm and then abruptly to ~15% at the boundary between lithofacies 1 and 2 (~205 cm). This indicates initially very low productivity in lithofacies 1, increasing slightly towards the top.
Freshwater diatoms are also recorded at 370 and 360 cm and at various intervals between 310 and 210 cm (Figure 4). However, in general diatoms are scarce between 450 and 300 cm indicating the low productivity of this lower section of the Ancylus Lake sequence.
The radiogenic nature of the 187Os/188Os record is interpreted to reflect freshwater conditions of the Baltic Basin during the deposition of lithofacies 1 between 450 and 205 cm. This is supported by other collected and published data including diatoms, TOC, BIT index and several organic geochemical records (Sinninghe Damsté et al., 2022; Sollai et al., 2017; Warden et al., 2017, 2018). With this information alongside the sedimentology of the core, the interpretation of this interval is that it represents deposition during the Ancylus Lake stage.
Littorina Sea
The combination of decreasing glacio-isostatic uplift along with glacial-eustatic changes led to a relative sea-level rise allowing areas through the Danish Straits such as Öresund and the Great Belt channels to connect the Baltic Basin and the North Sea. This deepening connection allows the Ancylus Lake to drain but also leads to the influx of marine waters into the Baltic Basin (the Littorina Sea Transgression), over time creating the marine/brackish Littorina Sea stage (Andrén et al., 2011; Bennike et al., 2021; Rosentau et al., 2017). The Littorina Sea stage is characterised by brackish water conditions, as marine waters were able to flow and mix into the Baltic Basin through the Danish Straits as the Fennoscandian Ice Sheet retreated (Andrén et al., 2011; Berglund et al., 2005; Björck, 1995; Rosentau et al., 2017). The transport of water through the Danish Straits was likely similar to present day whereby during the outward phase, relatively freshwater masses move into the Belt Sea and are mixed with saline masses. Then, during the next period of inflow, the mixed water mass returns to the Littorina Sea (Ehlin, 1981). This stage is represented in cores across the Baltic Basin by homogenous, finely laminated olive-green sediments with higher TOC values in comparison to Ancylus Lake sediments (Moros et al., 2002; Rosentau et al., 2017).
The sedimentology of core P435/1-5 GC from 205 cm to the core top (lithofacies 2) is characterised by fine, homogenous clay-dominated laminations. From ~240 to 170 cm the 187Os/188Os values decrease from 2.60 to 1.97 and then gradually to the present-day value 1.88 at the core top (Figure 4). This indicates a decrease in radiogenic, terrestrially sourced material from freshwater flux and/or an increased influence from open marine waters, lowering the 187Os/188Os values to 1.88 and in line with other surface samples taken from the present-day around the Baltic Sea (Figure 2). The BIT index values decline from 0.83 to their lowest at 0.03 between 280 and 200 cm and then remain at this low level to the core top. A value of 0 for the BIT index represents a marine end member (Kim et al., 2006; Wang et al., 2013; Warden et al., 2018). Thus, the decline and consistently low values in the BIT index in P435/1-5 GC and 303,600 -N represents a significant increase and continuous influence of marine waters. A sudden decrease in the BIT index between the Ancylus Lake and Littorina Sea stages has also been noted in sedimentary successions from the Arkona and Gotland Basins (Warden et al., 2018). Contemporaneously, TOC levels begin to increase significantly for the first time over the length of the core at 250 cm, and then much more rapidly at 230 cm rising from <2% to 13% and remain relatively high through lithofacies 2 (average ~6%) compared to lithofacies 1 (average ~ 0.5%; Warden et al., 2017). This is likely a result of increased productivity due to general climate amelioration, and a reduction in meltwater influx diluting the organic material produced from in situ productivity. Increasing TOC values are found from sediments of the Littorina Sea stage across many cores around the Baltic Sea (Kögler and Larsen, 1979; Moros et al., 2002). Total Os and Re values increase to their highest levels (224 ppt and 12 ppb, respectively) coincident with the rise in TOC and decrease in 187Os/188Os and BIT index (Fig4) this figure reference hasnt been altered and linked like all the others. Rhenium and osmium are organophilic and so are concentrated in the organic matter of organic-rich sediments during deposition from seawater. The increase in abundance of Re and Os at ~200 cm could relate to increased chelation by the initial marine water incursion, higher organic matter content and change in organic matter type related to the paleoceanographic change from the Ancylus Lake and Littorina Sea stages (Cohen, 2004; Cohen et al., 1999; Harris et al., 2013; Ravizza and Turekian, 1989; Selby et al., 2013; Selby and Creaser, 2003; Taylor et al., 2025). Additionally, at ~205 cm the first occurrence of brackish diatoms is recorded in both cores P435/1-5 GC and 303600 N, with an abrupt switch from freshwater to brackish diatoms. Diatom analysis was not undertaken from 180 cm to the surface, but abundant brackish to marine diatoms are found throughout the Baltic Sea in core sediments deposited during the Littorina Sea stage (Andrén et al., 2011; Moros et al., 2002).
The decline to less radiogenic 187Os/188Os values in lithofacies 2 between ~220 cm and core top therefore indicates an increased influence of less radiogenic marine waters and less influence of more radiogenic freshwater sources. Although, likely similar to present day, the 187Os/188Os mixing into the Bornholm Basin would have been more radiogenic than the open ocean (~1.06) and more likely ⩾1.5 at ~8 ka. This is in part due to the marine water entering Kattegat Bay from eastern Skagerrak being more radiogenic at ~8 ka (⩾1.1; Ownsworth et al., 2024), and due to the mixing of waters in the Belt Sea as discussed above. As such, this may explain the absence of a mixing line in 187Os/188Os versus 1/192Os space (Figure 3b) between an open seawater (~1) and terrestrial derived osmium isotope signal (~2).
The transition from the Ancylus Lake to Littorina Sea stages is further supported by other collected and published data including a low BIT index (Warden et al., 2018), higher TOC (Warden et al., 2017) and a sharp increase in the percentage of brackish diatoms. This alongside the sedimentology of the core indicates that the transition from lithofacies 1 to 2 represents the transition from the freshwater Ancylus Lake stage to the brackish Littorina Sea stage of the Baltic Sea Basin history.
Deeper sections
In the deeper section of the core ~390–386 cm there are several changes in our dataset that, combined, could provide evidence for a relatively restricted and short-lived marine influx (Figure 4). Between the depths of ~390–386 cm there is a dip in the BIT index to 0.14, close to the marine end member of 0 (Kim et al., 2006; Wang et al., 2013; Warden et al., 2018), which could indicate an increase in marine input. There is also a small lowering in the 187Os/188Os values from 2.6 to 2.4, which would indicate a less radiogenic source of Os. An increased marine influence would achieve this. A minor increase of ~1% in TOC also occurs at this same depth (Figure 4). This small TOC spike can potentially be correlated with similar spikes in cores from other areas in the Baltic Sea (e.g. Andrén et al., 2000; Moros et al., 2002; Warden et al., 2018). However, especially without means of dating the core, and with such small deviations, it is very difficult to make any definitive conclusions as to the cause of the TOC enrichment. Despite this, the data may become useful in later studies.
Implications and conclusions
Through combining modern day surface sediment data with past core records including BIT index, and TOC, this study demonstrates the utility of osmium isotopes for reconstructing and understanding the connection of the Baltic Sea Basin to the open ocean. This technique has the potential to be applied to other similarly restricted basins. A strong correlation between salinity and 187Os/188Os is recorded showing that the 187Os/188Os composition of sediments can be used to differentiate between a largely marine environment and a largely freshwater environment. It is shown that the use of 187Os/188Os is effective in distinguishing between freshwater and marine phases where there are large and longer-term drainage or inflow events, such as in this study with the Littorina Sea transgression between the freshwater Ancylus Lake stage and the brackish Littorina Sea stage of the Baltic Sea history. This also supports the findings of 187Os/188Os values tracking the Arctic Lake to Arctic Ocean transition (Dickson et al., 2022; Poirier and Hillaire-Marcel, 2011).
Supplemental Material
sj-pdf-1-hol-10.1177_09596836261432434 – Supplemental material for Tracking postglacial palaeoceanographic change of the Baltic Sea Basin using Osmium isotopes: Transition between the freshwater Ancylus Lake and the marine/brackish Littorina Sea
Supplemental material, sj-pdf-1-hol-10.1177_09596836261432434 for Tracking postglacial palaeoceanographic change of the Baltic Sea Basin using Osmium isotopes: Transition between the freshwater Ancylus Lake and the marine/brackish Littorina Sea by Emma Ownsworth, Matthias Moros, David Selby, Jeremy Lloyd, Yang Li, Jaap Sinninghe Damste and Slawomir Dobosz in The Holocene
Footnotes
Acknowledgements
We are grateful to laboratory assistance from Chris Ottley, Geoffrey Nowell and Emily Unsworth at Durham University.
Ethical considerations
Ethical approval not required.
Author Contributions
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was part funded by the NERC IAPETUS DTP, grant code NE/L002590/1. Funding contribution came also from a Yorkshire Geological Society (YGS) research fund grant.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data availability statement
The data that support the findings of this study are available in the article and supplementary material.
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
