Title: Influence of transboundary transport of trace elements on mountain peat geochemistry (Sudetes, Central Europe)

Mountain ombrotrophic peatlands in Central Europe are an important stock of transboundary contamination both of natural and anthropogenic origin. The Snie _ zka Mountain (West Sudetes) forms a significant orographic barrier and receives aerosols from broadly-recognized anthropogenic sources (production and use of stainless steel, processing of uranium, coal combustion, nuclear weapon tests, and Chernobyl accident). The main objective of the study was to assess the pattern of distribution and origin of trace elements and to distinguish the long-range transport vs. local signals in two 210Pb and 14C e dated peat cores from the highest summit of the Karkonosze (West Sudetes) spanning the last 280 years. Maximum values and accumulations of almost all investigated elements (Pb, Zn, Cu, Ni, Cr, Ti, Al, U, Sc, and REE) were identified around the 1970s. The analysis of peat using scanning electron microscopy (SEM) confirmed the occurrence of spheroidal aluminosilicate fly ash particles (SAP) in the topmost 40 cm (from AD 1938) together with a maximum of mullite (3Al2O3$2SiO2), an anthropogenic marker originating from coal-based power plants. The overall 206Pb/207Pb signature ranges from 1.160 to 1.173, indicating a predominant contribution of anthropogenic Pb. Human activities promote the release of mobile 234U, due to the weaker bonds to mineral structure, and cause the radiogenic disequilibrium between 238U and its daughter 234U. © 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).


Introduction
Industrial activities caused dramatic changes in ecosystems, releasing uncontrolled amounts of trace elements influencing both biotic and abiotic components (Rose, 2015;Waters et al., 2016Waters et al., , 2018. Atmospheric pollutants are transported by wind and dispersed at different geographical scales. The range of dispersion is dependent on the size of the particle and its chemical behaviour in the air (Samson, 1988). Larger particles are generally deposited close to the emitter (Samson, 1988) while regionally-recognized lighter particles can be transported from hundreds to thousands of kilometres, passing through cities, states, and countries, and strengthen the local pollution (Munn, 1972;Erel et al., 2002Erel et al., , 2006Bergin et al., 2005). It results in a more global, uniform pollutant signal over continents for certain pollutants (i.e. contaminants carried by submicronic aerosols: artificial radionuclides, lead derived from the combustion of leaded gasoline).
Mountain ecosystems such as lakes and peatlands are sensitive archives of long-range transport (Arellano et al., 2011;Catalan et al., 2013;Le Roux et al., 2016;Okamoto and Tanimoto, 2016). Notwithstanding, the received input of dust depends on height above sea level (a.s.l.). The altitudes between 0 and~2000 m a.s.l. are mainly subjected to vertical transport of Potentially Harmful Trace Elements (PHTE) (Le Roux et al., 2016). The free tropospheric zone, located above 2000 m a.s.l. is characterised by more lateral transport of PHTE (Catalan et al., 2013;Le Roux et al., 2016) and ecosystems located there receive dust mostly from long-range transport. The mutual effect of different climate factors like wind trajectories and precipitation play the most important role in mountain pollution (Catalan et al., 2013;Le Roux et al., 2016).
Peat bogs, exclusively fed by dry and wet atmospheric deposition, are significant biological filters and traps for PHTE and other elements. They give essential information about past and present changes in the environment, especially when applying a multiproxy approach (Lamentowicz et al., 2013;Gałka et al., 2019). Mountain peat bogs, located on plateaus, where the slope effect is minimal, are great archives of geochemical records of various origin (Le Roux et al., 2016).
The calculated accumulation of peat and, therefore, accumulations of elements enable to assess the past trends in pollution as well as to assess the sources of observed enrichments (e.g. Shotyk et al., 1998Shotyk et al., , 2002. REE elements are believed to be good indicators of lithogenic activity and natural sources of dust (e.g. Le Pratte et al., 2017;Vanneste et al., 2016), but they can also be indicators of anthropogenic activity (e.g. Fiałkiewicz-Kozieł et al., 2016).
The Sudetes (located at the border of Poland, the Czech Republic, and Germany) are an important region with long-time traditions of mining and ore processing (i.e. Fe, Cu, Pb, U, and Au) that may have had a great effect on PHTE deposited onto mountains. This mountain range is additionally surrounded by economically significant industrial centres and lignite-fired power plants, causing dramatic deterioration of environment in the 1970s and 1980s in Poland, the Czech Republic, and Germany. Due to dramatic environmental impact this area (e.g. Mazurski, 1986), named "Black Triangle" (BT), was qualified by the United Nations Environment Programme (UNEP) as an "ecological disaster zone" (Grübler, 2002;Kol a r et al., 2015). Sparse investigations about mineralogy, pollen analysis, sulfur and carbon isotopes exist (i.e., Popowski, 2005;Skrzypek et al., 2009;Kajukało et al., 2016), but information about past PHTE contamination in peat bogs from the Sudetes is scarce (i.e. Strzyszcz and Magiera, 2001;Fiałkiewicz-Kozieł et al., 2015a). The most detailed anthropogenic signal in central Europe was estimated in the Czech Republic (i.e., Novak et al., , 2008Ettler et al., 2004;Ettler et al., 2006;Bohd alkov a et al., 2014). Only a few multiproxy studies including advanced geochemical analyses were carried out on peatlands from Poland (e.g. De Vleeschouwer et al., 2009;Fiałkiewicz-Kozieł et al., 2018).
Here we present a detailed multiproxy geochemical record from two cores of the Na R owni pod Snie _ zką (NRS) peatland, located at the feet of the Snie _ zka summit (1400 m a.s.l.), the highest mountain of the Sudetes. We hypothesised that long-range transport mainly contributes to the pollution of NRS. Our aims are: (1) to assess the pattern of distribution and origin of trace elements and REE (2) to distinguish the most significant signals of pollution in areas impacted by transboundary anthropogenic pollution and (3) to explain the fate of strategic elements like uranium after anthropogenically-driven release to the atmosphere.

Study site
The Sudetes are surrounded by industrial centres located in Poland (Lower Silesia) as well as in the Czech Republic (Northern Bohemia) and Germany (Saxony and Brandenburg). These three areas constitute the so-called Black Triangle (Fig. 1). In the 1970s and 1980s, the "Black Triangle" was one of the most polluted areas in Europe due to the uncontrolled density of factories and power plants, which, working without filters, released enormous amounts of alkali dust and toxic gases (mainly SO 2 ) in the atmosphere (Ję drysek et al., 2002;Szynkiewicz et al., 2008).
The Na R owni pod Snie _ zką (NRS) peatland (50 44 0 18,31 00 N, 15 41 0 59,72 00 E) is located on a plateau (1350e1450 m a.s.l.) in the Karkonosze range, close to the highest summit of Mt Snie _ zka (1602 m a.s.l.). The Karkonosze area is~60 km long and up to 20 km wide granitic massif in the West Sudetes at the northern periphery of the Bohemian Massif, straddling the Czech/Polish border and forming the eastern extremity of the European Variscan belt (Mazur et al., 2006). The Karkonosze unit was lifted during Alpine orogeny, forming currently a steep rocky mountain range with numerous glacial landforms. Consequently, it is a significant orographic barrier for different types of air masses. During the 64% of the year, the area is directly influenced by the zonal circulation of oceanic air masses from the North Atlantic, flowing over the lowlands of Western Europe. 30% of the year is characterised by the influence of polar continental air masses, flowing from the east, 4%by arctic air from the north and 2% -by tropical air masses, coming from the south (Sobik et al., 2014). The mean annual air temperature at Mt.

Coring and subsampling
Two cores of 63 and 67 cm long (Sn1 and Sn2, respectively), were retrieved in the spring of 2012 using a stainless steel 10 Â 10 Â 100 cm Wardenaar corer (Wardenaar, 1987). The distance between them was 20 m. Monoliths were wrapped in plastic bag, transported to the laboratory in Pozna n (UAM), and stored in the laboratory cooler. Fresh cores were cut into 1-cm thick slices (except the top 6 cm, which were cut on 2 cm slices) using a carbon steel knife. Fresh cutting was necessary for detailed mineralogical analysis as freezing the peat core (as it is frequently done before slicing) would have caused a transformation of minerals by changing the temperature and water conditions of the catotelm. Protocols from Givelet et al. (2004) and De Vleeschouwer et al. (2010) were used to minimise contamination. Several geological samples such as the local rocks (granite, hornfels from Snie _ zka), lignite coal (mine Turosz ow), fly ash from Power Plant Tur ow, aircraft fuel (Laboratory Warter fuels, Płock, Poland) were also collected.

Ash content, bulk density, plant macrofossils
Dried bulk density was determined using fresh material collected with a 5-cm 3 beaker. Ash content (AC), used to quantify the relative proportion of mineral fraction in the peat, was determined by burning the dry samples at 550 C overnight. Highresolution (1-cm peat slices) plant macrofossil analysis was used to reconstruct local ecological conditions and peat-forming plants in contiguous samples of approximately 20 cm 3 in the two profiles (comp. Suppl.1). The samples were washed and sieved under a warm water current over 0.2-mm mesh sieves. Vascular plants and brown mosses composition were determined on the basis of carpological remains and vegetative fragments (leaves, rootlets, epidermis) using the available identification keys (e.g. Smith, 2004;Mauquoy and van Geel, 2007). The identification of Sphagnum to species level was carried out separately on stem leaves using specific keys (H€ olzer, 2010; Laine et al., 2011).

Dating
The Sn1 profile was analyzed by gamma ( 137 Cs) and alpha ( 210 Po, 238 Pu, 239þ240 Pu, 234,238 U) spectrometry (comp. Table 1, Fig. 2). Separated slices were dried at 105 C, homogenised, and taken for gamma spectrometric measurement. For 137 Cs analyses, samples were packed into 100 ml polypropylene cylindrical containers and were measured using high-resolution gamma spectrometry with a planar HPGe (high-purity germanium) detector with a composite foil window made of carbon fiber which support a capon foil covered with ultra-thin aluminum foil (homemade by Institute of Nuclear Physics PAS Krakow and electronics by Silena S.p.A.). A small amount of sample material was deposited on the bottom of the container. Activities of 137 Cs were determined via the 137m Ba emission peak at 662 keV. Spectra were collected for 12e72 h, depending on the activity of the samples. The activities of 210 Po, 238 Pu, 239þ240 Pu, and 234,238 U were determined for about 1 g of dried samples. Samples were digested with radioactive tracers ( 208 Po, 242 Pu, 232 U) and a concentrated mixture of HNO 3, HCl, H 3 BO 3, and H 2 O 2 and slowly dried. Details of this procedure were described in previous articles (Mietelski et al., 2008;Łokas et al., 2013;Mr oz et al., 2017). The activity of total 210 Pb was determined indirectly by measuring its decay product, 210 Po, using alpha spectrometry. 210 Po was chemically extracted from the material. Po isotopes were deposited on an Ag disc. Solution after polonium separation was used for Pu separation. Plutonium alpha sources were prepared by the NdF 3 microcoprecipitation method (Sill, 1987;Rao and Cooper, 1995). The effluent solution (8 M HNO 3 ) after Pu separation was used for U determination. Uranium was coprecipitated directly from this elution using NdF 3 and Mohr's Salt (ammonium iron sulfate) to obtain a thin spectrometric source (Łokas et al., 2010;Mietelski et al., 2016). Activity concentrations for polonium, plutonium, and uranium isotopes were determined using alpha spectrometers (Silena AlphaQuattro, Ortec Alpha Duo or Canberra 7401; all equipped with Canberra or Ortec ionimplanted silicon detectors). For this radiochemical procedure blank samples were made of reagents and were analyzed after each thirty sample. Alongside blanks and samples, the reference materials (IAEA Moss-Soil 447 and IAEA Sediment 385) were analyzed to ensure the quality of measurements. The obtained results: 447 ( 234 U-20.8 ± 1.8; 238 U -21.4 ± 2.2; 238 Pu -0.14 ± 0.01; 239þ240 Pu -4.96 ± 0.32, 210 Po -415 ± 10; 137 Cs -421 ± 22); 385 (234U-27.9 ± 1.9; 238U -29.4 ± 2.2; 238Pu -0.40 ± 0.06; 239 þ 240Pu -2.70 ± 0.23) fall within 93e99% of certified values.
Seven samples from Sn1 and five samples from Sn2 were subjected to 14 C measurements in Radiocarbon Laboratory in Pozna n. Hand-picked plant macrofossils, stored in MilliQ water, were selected for dating (Table 2). An absolute chronology is based on (i) the age-depth models calculated on 14 C dates in the OxCal v. 4.3 software (Bronk Ramsey, 1995) applying the IntCal13 (Reimer et al., 2013) and BOMB13NH1 (Hua et al., 2013) atmospheric curves as the calibration set and (ii) 210 Pb dates obtained for the section of 0e52 cm in the Sn-1 profile (Fig. 3). Additionally, patterns of SAP and mullite were used as chronomarkers and included in age-depth model for Sn2 (comp. Fiałkiewicz-Kozieł et al., 2016). For the calculation of the models based on 14 C dates, the P_Sequence command (Bronk Ramsey, 2008) with k parameters equal to 1 cm À1 and log 10 (k/k 0 ) ¼ 1 was applied. In the case of the Sn1 profile, the section between 0 and 52 cm, being dated using 210 Pb method (CRS model) (Table 1 and Fig. 2), was validated by 14 C dates. The CRS model assumes a constant rate of supply of unsupported 210 Pb to the peat surface despite variable sedimentation rates (Appleby and  (Sobik et al., 2014). The localisation of Wismut company (Wolkersdorfer, 1995), localisation of Czech uranium mines and mills (Cech ak and Kluson, 2006), spotted line e the distance between the source of emission and destination, pink spot e investigated Polish peatlands, described in the text. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) Oldfield, 1978). Unsupported activity concentrations of 210 Pb, which were determined for each layer by subtracting the supported activities from the total 210 Pb activities. The supported level of 210 Pb was calculated by using the mean activity of the bottom layers of the peat profile (14 ± 2 Bq kg À1 ).
The lower section of the profile (52e64 cm) was based on the Bayesian age-depth model constructed using the mixed 210 Pb (date from a depth of 52.5 cm introduced using C_Date command) and 14 C dates. In the case of the Sn2 model, the calendar date AD 1938 ± 2 was introduced (using C_Date command) at a depth of 26.5 cm as an extrapolated date of SAP presence in Sn1 (see Fig. 3). Moreover, we used one boundary (Boundary command) at a depth of 44 cm due to the rapid change in bulk density (Suppl.2), which might be a trace of two sections of different peat accumulation rates (see Fiałkiewicz-Kozieł et al., 2014, Fiałkiewicz-Kozieł et al., 2015b. For the better readability of the modelled dates, they were expressed as m values (cal. AD) in the following sections of text.
The accumulation rate (AR) was calculated using concentration (mg g À1 ) of element, bulk density, and peat accumulation rate using the following equation: AR (mg m À2 yr À1 ) ¼ concentration* BD (g cm À3 )*PAR (cm y À1 ) *10.
The dust flux was calculated using the sum of REE concentrations (mg g À1 ) in the bulk peat using the following equation : Wedepohl, 1995), PAR is the peat accumulation rate (cm yr À1 ) calculated according to formula (h2-h1)/(y1-y2), where h-depth, y-age, and BD is the bulk density (g cm À3 ).

Pb isotopes
Ten to 500 mg of dry peat and rocks powder was taken in order to obtain 2000 ng of Pb in the final solution. The sample powders were ashed (550 C, 4 h) prior to digestion in 6 ml HF suprapur (Merck) þ 1 ml HNO 3 65% cc. sub. in Teflon beakers (120 C, 48 h) in class 1000 clean room (UAM, Pozna n). Dried residues were dissolved in HBr 6% cc. sub. prior to chromatographic separation. Lead was separated using anionic exchange micro-columns (Weis et al., 2005) and sub-boiled distilled acids. The measurements were conducted on TIMS Finnigan MAT-261 special (UAM). The instrumental drift was controlled by standard bracketing using NBS981 standard data (Galer and Abouchami, 1998 (Taylor et al., 2015).

Mineralogy
The shape, size, morphology, and chemical composition of dust particles were determined in peat samples using the backscattered electron detector of a scanning electron microscope (SEM) equipped with an energy dispersive system (Philips XL30 ESEM/EDS). The accelerating voltage was 15 kV and 10 mm the working distance. Air-dried peat samples were gently homogenised using a corundum mortar and pestle. A thin layer of each homogenised sample was fixed to a double-sided 9 mm carbon tab, placed on an aluminium stub, and carbon-coated prior to analysis.
Mullite content was determined in Sn1 profile using x-ray diffraction (XRD). For the XRD analysis, peat samples were ashed at 550 C and treated with 1 M HCl for 15 min to remove the acidsoluble ash fraction (Sapkota, 2006). The residue was dried, ground in an agate mortar, and analyzed using a Panalytical X'Pert PRO -PW 3040/60 X-ray diffractometer. The instrument was equipped with a Ni-filtered Cu Ka source radiation (l ¼ 1.540598 Å) and an X'Celerator strip detector. Samples were scanned within a 2Q interval of 2.5e65 , with a step size of 0.01 2Q and counting time of 300 s. Identification and quantification of mineral phases was done by means of the X'Pert HighScore Plus Software using the newest ICSD database. The detection limit of the XRD method was 0.5e2%, depending on the sample mineral composition. Analytical precision and accuracy were ±3%.

Chronological control ( 210 Pb, 14 C)
The 210 Pb activities (Table 1) gradually declined with increasing depth and became constant at depths ranging from 52.5 to 54.5 cm, which was consistent with the equilibrium depth of total 210 Pb and supported 210 Pb (14 ± 2 Bq kg À1 ). Fig. 2 shows the CRS age-depth relationships of Sn1 peat profile. The peat section 0 -49 cm accumulated in 149 years, corresponding to an average accumulation rate of 0.53 ± 0.11 cm yr À1 .
The data obtained from 14 C activity (Table 2) suggests a hiatus in the Sn2 profile. Therefore, for further interpretation, we opted only to use the top 44 cm, which spans a similar time interval comparing to Sn1.
The A model for the Sn1 model revealed 17%, whereas for the Sn2 it was 64% (Fig. 3). The minimum advised as the critical value for the age-depth model robustness is 60% (Bronk Ramsey, 2008

137 Cs, 239þ240 Pu
Activity concentrations and inventories of anthropogenic radionuclides ( 137 Cs, 238 Pu, 239þ240 Pu) are presented in Table 1 and in Fig. 2. The radionuclide inventory is understood here as the activity concentration of a given radionuclide contained in the profile (Bq kg À1 ) per unit surface area (kg m À2 ). The activity concentrations of 137 Cs range from 81 ± 45 to 2255 ± 104 Bq kg À1 . Most of the total Cs activity concentration is retained in the upper four layers (0e9 cm). Two distinct peaks of 137 Cs activity can be found at 23.5 cm (1011 ± 48 Bq kg À1 ) and 26.5 cm depth (699 ± 58 Bq kg À1 ), and can be attributed to Chernobyl (1986) and Nuclear weapon test (1963) signal (Table 1, Fig. 2).
Activity concentrations of 239þ240 Pu range between 0.06 ± 0.04 Bq kg À1 to 22.86 ± 1.99 Bq kg À1 (dry weight) with a maximum value at 33.5 cm depth. Activity concentrations for 238 Pu in this profile are much lower than for 239þ240 Pu. The minimum value is < 0.03 Bq kg À1 , the maximum equals 0.69 ± 0.14 Bq kg À1 at the same depth as 239þ240 Pu (Table 1).
Several shifts in concentration and accumulation of investigated elements can be distinguished.
In Sn1 the layer between 50.5 and 62.5 cm (AD 1739e1872) was characterised by the decreased Pb concentration from 218 mg kg À1 at the bottom of the profile to twice lower value of 111 mg kg À1 at the depth 58.5 cm (AD 1780), which slightly fluctuated till 50.5 cm (AD 1872). The calculated Pb accumulations are the highest at the bottom (62.5 cm) e 20 mg m À2 y À1 (Fig. 4). The increase in Al and Ti concentration from 57.5 cm to 50.5 cm; (AD 1789e1872) with maximum at 55.5e53.5 cm (AD 1806e1828) is less visible in accumulation rates (comp. Fig. 4). Other elements displayed only minor fluctuations during that time.
In Sn2 the described period was represented by three samples within a depth interval of 45.5 cme43.5 cm (AD 1760e1883). This layer was characterised by the highest density and low accumulation rates.
The depth from 50.5 cm to 35.5 cm (AD 1870e1954) in Sn1 was characterised by only small variations in both concentration and accumulation of most elements (Suppl.2, Fig. 4). In Sn2, the depth from 43.5 cm to 20.5 cm (AD 1883e1952) reflected a more complex, but comparable trend to Sn1 (Supl.2, Fig. 4). In both profiles, an anomaly in Cr and Ni is observed within the period. In Sn1, a distinct peak in Cr (33 mg kg À1 ) and Ni (15 mg kg À1 ) concentrations is seen at 46.5 cm (AD 1904;Fig. 5). The similar, but broaden peak was also observed in Sn 2 at a depth of 33.5 cm (AD 1912) for Al, Ti, Cr, Ni, and Cu and at 31.5 cm (AD 1922) for Cr and Ni (comp. Suppl.2,Figs. 4,Fig.5). Pb also revealed a small increase in concentration as well as accumulation in both profiles.
The peat layer 22.5 cme35.5 cm (AD 1954e1988) in Sn1 and 10.5 cme19.5 cm (AD 1955e1978) in Sn2 was characterised by the most pronounced changes in concentrations and accumulations of all elements.
Rare Earth Element (REE) concentrations assessed for Sn1 are given in the supplementary data (Suppl.2) and presented to show the significance of industrial activity on the cycle of lithogenic elements. The total sum of P REE varied from 1.2 to 40 mg kg À1 and was used to assess the dust flux (Fig. 4). From the bottom to the 31.5 cm both concentrations and accumulations presented almost straight line with only slight fluctuations. The highest dust accumulation rate was observed at the depth of 30.5 cm (AD 1970) (Fig. 4) and corresponded to other elements in Sn1. The accumulation rate of La as a representative of REE displayed a similar pattern to dust flux reaching maximum values in 1970 (Fig. 4).

Tool for distinguishing sources of deposited elements (Pb isotopes)
Lead isotopic signatures revealed similar age and depth patterns in both peat profiles during the last 100 years. 208 Fig. 4). Generally, the ratio 206 Pb/ 207 Pb decreased to less radiogenic values from the bottom to the top of the profiles and revealed similarity to the anthropogenic sources (Figs. 4 and 6).

Mineralogy
The distinguishing of mineral particles in the peat profiles gave an opportunity to further confirm the sources. Spheroidal aluminosilicate fly ash particles (SAP) were the dominant technogenic dust particles detected in the samples using SEM. In the Sn1 profile, SAP were found within a depth range of 0e40 cm and in Sn2 at depths between 0 and 27 cm. This result was used to calculate chronology (comp. 2.4 and Fig. 3). The majority of the SAP particles in the Sn1 profile were within a size range of <1e9.5 mm, independent of the depth, which indicates the long-range transport. The mean value was 2.4 and the median -2.0 mm (n ¼ 180). A small number of much larger (up to 50 mm), less regular, and highly porous aluminosilicates were found within a depth range of 25e35 cm in Sn1.

Steel factories signal
An increase in Cr and Ni accumulation was described in peatlands from the United States (Cole et al., 1990), Poland (De Vleeschouwer et al., 2009), Belgium (Allan et al., 2013), the Czech Republic (Bohd alkov a et al., 2018) and Germany (Gałka et al., 2019). A distinguishable signal around 1904 ± 15 in Sn1 and from  1922 ± 7e1928 (±6) in Sn2 in the CreNi accumulation profiles is also observed in Snie _ zka, with only slight fluctuations in trace and REE accumulation rates (Fig. 5). Ni and Cr are suggested to be immobile in peat bogs (Krachler et al., 2003;Allan et al., 2013), but research indicating mobility also occurred (Nieminen et al., 2002;Ukonmaanaho et al., 2004). Despite the existence of these contradicting conclusions, the repeatable signal in both profiles (Sn1, Sn2) gives us confidence in the possible immobility of those elements. A sharp increase in Cr and Ni in Cowles bog (USA, Northern Indiana), observed in 1928e1942, was explained as booming stainless-steel industry (Cole et al., 1990). The invention of stainless steel in 1912 was a milestone and caused the boosting of industrial development in many countries (Corker, 2016). In Europe, the leading producer of CreNi steel was Krupp company, the owner of many steel factories, located in Germany (Ruhr area, Kiel) and Poland (presently), for example in Szklary (Ger. Glasendorf) (see Fig. 1). The Szklary Ni-Steel factory was located about 120 km east of Snie _ zka, and operated from 1901 to 1920 on both local and imported Ni ores (Furmankiewicz and Krzy _ zanowski, 2008). In Belgium, the CreNi increase was attributed to industrial activities from the Ruhr region, also belonging to the Krupp company, 300 km from the investigated peatland (Allan et al., 2013). The Ruhr stainless steel industries and their use of NieCr steel alloys promoted the emission of these two metals (Corker, 2016). In the Słowi nskie Błota bog (Northern Poland), 7 mg kg À1 was detected in peat layer accumulated in AD 1920, similarly to the NRS bog, but the accumulation rate in Słowi nskie Błota was much lower (De Vleeschouwer et al., 2009). The Słowi nskie Błota bog is located about 400 km to Kiel and could also be affected by nearby Krupp steel factories. While the exact number of steel factories operating during that time is unknown, the use of primitive technology without particulate-emission control systems has certainly contributed to the elevated Ni and Cr accumulations found in the NRS peat profiles. This is also supported by slag particles found in the Sn1 peat profile at the depth of the highest CreNi concentration and accumulation (Fig. 5). Such particles were not found in the topmost layers. The elevated peak in Cr and Ni in the 1970s and 1980s, together with other investigated elements, is attributed to extensive brown coal combustion (comp. Figs. 4 and 5).

Uranium processing signal
The Sudetes lie in the northern part of the Bohemian Massif, the most important uranium ore district in Europe with many deposits of various sizes, both in the Czech Republic and bordering eastern Germany. The total historical uranium production is estimated at 350 Mt for the region (OECD-IAEA, 2003). The total production of uranium ores in Germany (Wismut company) from 1946 to 2012 was c. 220 Mt. In Czechoslovakia (presently as the Czech Republic and Slovakia), the total uranium production from 1945 to 2017 was c. 112 Mt (Diehl, 2011). Before becoming a strategic material, uranium was also mined as a by-product during the exploitation of copper, iron, or arsenic ores and used for colouration in glass and porcelain industries (Proch azka et al., 2009). Natural uranium consists of three alpha radioactive isotopes: 99.2745% of 238 U, 0.7200% of 235 U, and 0.0054% of 234 U (Boryło and Skwarzec, 2014). The specific activity of 238 U is much lower (1.24 Â 104 Bq g À1 ) compared to 234 U (2.30 Â 108 Bq g À1 ) (Browne and Firestone, 1986). A 234 U/ 238 U ratio close to 1 is typically found in natural samples including rock samples and sediments (Boryło, 2013). The moss Pleurozium schreberi from northern Poland revealed the isotopic ratio also around 1 (Boryło et al., 2017). A recognizable regional signature is seen in the U record of Snie _ zka peat profiles (Fig. 7). In Sn1 and Sn2 profiles, two periods in uranium deposition can be distinguished. The obtained values are two orders of magnitude higher than values reported in the mountainous Etang de la Gru ere bog in Switzerland   (Boryło and Skwarzec, 2014). The geochemical explanation of radioactivity imbalance is the looser bonds of 234 U atoms in mineral structures, making them easier to leach during physicochemical erosion (Fleischer and Raabe, 1978). Industrial activities, fossil fuel combustion, phosphate fertilizers in agriculture, and domestic and industrial sewage trigger the increase of uranium concentration and disequilibrium of 234 U/ 238 U ratio (Boryło, 2013;Boryło and Skwarzec, 2014).
The elevated values, together with an increased 234 U activity coincide with the discovery of uranium and increasing mining and processing of uranium as well as coal combustion. The U peak around 1840e1890 is interpreted here as originating from the glass/porcelain industry. The glassworks used local rock material and charcoal, both enriched in uranium, as well as pure uranium for colouration. This glass industry gained its highest popularity around 1880e1920 (Rene, 2008). Several glassworks existed in the Sudetes and in the South of Bohemia as well as in Germany, Austria, England, and France during that time (Rene, 2008;Proch azka et al., 2009).
The increase in uranium accumulation rate starting from the year 1938 and reaching a maximum in the 1970se1980s, reflects the pollution from fossil fuels' burning, in line with other proxies. Uranium was classified as a coalphile element by Ketris and Yudovich (2009). The brown coals from the nearby Tur ow coal field have twice higher U content (4.4 mg kg À1 ) than the average for Polish brown (2.2 mg kg À1 ) and hard (2.0 mg kg À1 ) coals (Bojakowska et al., 2008). The uranium concentration of Czech brown coal varies from 0.60 to 4.03 mg kg À1 (Bouska and Pesek, 1999). The mining of uranium ores constitutes an additional source of uranium supply to the atmosphere. Taking into account the fact that industrial activities promote the release of 234 U to the environment, we propose here that ombrotrophic peatlands can record the influence of humans on the biogeochemical cycle of uranium.

Chernobyl signal
The Chernobyl accident caused a dramatic release of many radionuclides, which were dispersed through Europe. The highest content of radionuclides reached the Belarus boundary, then split up towards Scandinavia and the southwestern part of Europe. It appeared in Poland two days after the accident. The most polluted area was the Opolskie (Southern Poland) and the part of Lower Silesian voivodeship, including Karpacz town, located at the feet of Snie _ zka (Łukaszek eChmielewska et al., 2018). The heterogeneous pattern of radionuclide fallout was caused by diverse weather conditions.
There is a strong increase of 137 Cs in the upper 4 cm of Sn1, probably due to plant retention (See Fig. 2 and Table 1). Similarly to K, 137 Cs can be uptaken by roots (Gerdol et al., 1994) and accumulated in living plants. This phenomenon was also described elsewhere in peatlands (Mr oz et al., 2017;Fiałkiewicz-Kozieł et al., 2014;Rosen et al., 2009). We suspect that the 137 Cs from Chernobyl may be present in the top peat layers. The total inventory of 137 Cs in Sn1 is 8970 ± 1300 Bq m À2 (Table 1). The estimated weapons Cs deposition for Poland is 982 Bq m À2 , but the mean total (including Chernobyl) value for 137 Cs inventory for Poland is 3770 Bq m À2 (Stach, 1996). In the Upper Odra valley (SW Poland), the 137 Cs inventory equaled to 5230 Bq m À2 (Porę ba and Bluszcz, 2007).
The mean inventory value for the other regions of southern Poland (e.g. Opole and Katowice provinces) equals to 11,240 Bq m À2 and 6800 Bq m À2 respectively (Stach, 1996), but the contribution of 137 Cs of Chernobyl fallout in this region is about 80% of the total 137 Cs fallout. The mean value for 137 Cs inventory for the Tatra National Park (Mietelski et al., 2008) in southern Poland is 7800 Bq m À2 , and this value is comparable with Sn1. In the Czech Republic, the area from Snie _ zka toward Prague was characterised by 3000e10,000 Bq m À2 , but on NW from Snie _ zka only 1000e3000 Bq m À2 was detected (Hanslík et al., 2018). The results are in good agreement with previous studies, conducted in higher elevation mountains such as the Tatra Mts.

Nuclear weapon test signal
There are two main sources of plutonium isotopes: global radioactive atmospheric fallout due to nuclear weapon tests and the Chernobyl accident. Plutonium from the global fallout was spread worldwide, but Pu from the Chernobyl accident was only dispersed in the form of fuel particles and was deposited mostly in the north-eastern and eastern parts of Poland (Mietelski et al., 2016). Non-volatile elements like Pu can be transported on larger aerosol "hot particles" (Cuddihy et al., 1989;Devell et al., 1986). The average deposition of global fallout of 239þ240 Pu for Poland was 58 Bq m À2 (UNSCEAR, 1982), and about 4% of the activity of 239þ240 Pu is due to the average 238 Pu deposition from weapon tests and fallout after the SNAP-9A satellite accidental burn up over the Madagascar. Therefore, we observed lower values for 238 Pu than for 239þ240 Pu. The observed maximum of 239þ240 Pu activity concentrations characterize the global fallout peak of 1963 and is commonly used to identify specific horizons in lake sediments, peat, and other deposits, as well as to estimate their accretion rates. The total inventory of 239þ240 Pu and 238 Pu is 76.6 ± 7.8 Bq m À2 and 2.20 ± 0.71 Bq m À2 , respectively. The value of 239þ240 Pu deposition is slightly higher to the Polish average of 58 Bq m À2 predicted by the UNSCEAR (1982). It is, however, in line with values found by Mietelski et al. (2008) in the Ko scieliska Valley area, Tatra Mts. The 238 Pu/ 239þ240 Pu activity ratios were calculated for layers with the maximum activity of these isotopes. In the Sn1, the values: 0.018 ± 0.005 to 0.051 ± 0.019 with a mean value of 0.029 ± 0.010 correspond to global fallout, including the SNAP-9A satellite crash (0.03e0.05) and there is no presence of Chernobyl-origin plutonium. The 238 Pu/ 239þ240 Pu activity ratio for Chernobyl fallout is 0.50 (Mietelski and Wą s, 1995) detected only in northern Poland (Mietelski et al., 2016). In the Czech Republic the deposition of plutonium isotopes over former Czechoslovakia (H€ olgye and Filgas, 1995) varies from 10.2 to 108.8 Bq m À2 for 239þ240 Pu and from <0.5 to 6.2 Bq m À2 for cumulative 238 Pu (H€ olgye and Filgas, 1995;H€ olgye and Malý, 2000).

Coal combustion central Europe signal
In Europe, lead in the atmosphere originated from ore processing, coal combustion, and gasoline usage (Pacyna et al., 2007). The isotopic signature of galena and coal in Central European records are very similar in areas where coal is the main source of lead (Vile et al., 2000;Cimova et al., 2016). Leaded gasoline, which was an important source of lead in western European countries, is hidden here by coal signatures (Fig. 6). The 206 Pb/ 207 Pb signature of aircraft fuel (1.172) taken presently from the main producer of fuel (Laboratory Warter fuels, Płock, Poland) (Table 3), similar to Polish leaded gasoline used in the 1980s (1.174, Yao et al., 2015), is placed within the range of coal signature.
SAP and mullite which form during industrial coal combustion have been deposited on the Snie _ zka plateau since 1938 (Figs. 3 and 4), which is later than the electrification of Black Triangle (Krajniak, 2017) and later than the occurrence of fly ash particles in peatlands located close/within coal-based regions (Yang et al., 2001;Smieja-Kr ol et al., 2019). Lower abundance and technological restrictions (e.g., lower chimneys resulting in proximal dissemination only) of the earlier times of industrialisation (before~1950) often give a local/regional signal (Waters et al., 2018)  The acceleration of coal usage is seen in all proxies shortly after 1970 when the total production of lignite was 260 Gt in Germany ( € Oko-Institut, 2017), 80 Gt in Czech Republic (Vrablik et al., 2017) and 39 Gt in Poland (Kasztelewicz, 2018). The 1970s are characterised by the highest Pb accumulation rate in both profiles: Sn1 (77 mg m À2 yr À1 ) and Sn2 (78 mg m À2 yr À1 ) (Fig. 4) comparable to other peatlands from the Black Triangle region (60e65 mg m À2 yr À1 ; Novak et al., 2008) or other heavy-industry impacted regions (100 mg m À2 yr À1 ; Allan et al., 2013). The highest sum of REE and other elements regarded as lithogenic was also noted for that time and attributed to lignite combustion (comp. with dust flux in Fig. 4., which is calculated based on P REE and La AR as REE representative). It is known that coals contain REE, Th, U, Ti, Zr, and Sc (Bouska and Pesek, 1999;Ketris and Yudovich, 2009;Dolnickova et al., 2012), and that these elements are further concentrated in fly ash (e.g., Vassilev et al., 2001;Jones et al., 2012) emitted in significant amounts during coal combustion. The peak of the booming industry in the 1970s is also recorded by less radiogenic 206 Pb/ 207 Pb ratios (1.161e1.162), distinguished in both profiles (Figs. 4 and 6, Table 3). This is comparable to isotopic signature of the Czech industry (Novak et al., , 2008Zuna et al., 2011;Bohd alkov a et al., 2014) and soils in Colditz (eastern Germany within the Black Triangle) (1.161e1.167) (Haack et al., 2003). The isotopic signature of Colditz soil (see Fig. 1) was significantly distinct from the Pb isotopic signature of other locations in western or southern Germany.
The Pb signature is also distinct from the rest of the Polish industry (Cu exploitation, Tur ow power plant activity) located in the Lower Silesian area (comp. Fig. 6, Table 3, Tyszka et al., 2012).
Despite its location in the Black Triangle, the Snie _ zka peatland reveals similarities in its Pb isotopic composition with north Poland (De Vleeschouwer et al., 2009), the Czech Republic , Switzerland (Weiss et al., 1999) and Belgium . It, however, displays discrepancies with Pu scizna Mała -SE Poland , where distribution patterns are more local due to their foothill character (comp. Fig. 1). The Tatra mountains act here as an effective barrier for long-range transported pollutants. The closeness of the eastern border with various precipitation regimes and wind directions, as well the burning of peat and coal influenced significantly the isotopic ratio of Pb in peatlands from SE Poland, while the Snie _ zka peatland displays less radiogenic profiles, suggesting that they receive more long-range transported pollutants, mainly from the west because of the dominant westerlies (comp. Figs. 1  and 6).
A second peak is visible around 1982e1986 and can also be attributed to lignite burning. This period is characterised by the highest production of coal in Poland but also by the highest dust emissions in Saxony (Zimmermann and Bothmer, 2000).
From 1987 onwards, a substantial decrease in all element concentration is observed as a result of reinforced pollution management in Europe. The least radiogenic 206 Pb/ 207 Pb isotopic ratio̴ 1.150 (Figs. 4 and 6, Table 3) is comparable to the values of fly ash originating from waste incinerators (1.14e1.16) from different parts of Europe (Komarek et al., 2008) as well as of unleaded gasoline (1.157) (Haack et al., 2002) and modern aerosols (Bollh€ ofer and Rosman, 2001), as observed elsewhere in Polish peatlands (De Vleeschouwer et al., 2009).

Conclusions
Both peat profiles, sampled in the ombrotrophic mountain peatland, record relatively similar geochemical changes linked to historical industrial activities, despite their different peat accumulation rates. The use of a wide spectrum of geochemical proxies allows distinguishing several broadly recognized sources of pollutants.
1. The Black Triangle industrial activities (the Czech Republic, Germany and Poland) as well as the broad Central Europe coal use caused the most significant changes in the geochemistry of Snie _ zka peatland and were the dominant source of pollution and atmospheric emission of most elements such as PHTE (Pb, Zn, Cu, Ni, Cr). All the elements classified as "lithogenic" on a geological viewpoint (Ti, Al, Sc, REE), also increased during 1970'as a result of the intensive use of coal, pointing out that one should be cautious when using them as reference elements to calculate enrichment factors, as part of these lithogenic elements though are issued from anthropogenic activities. 2. The leaded gasoline signal is hidden by the coal isotopic ratio, which has been abundantly used in central Europe. While the local signal linked to Polish industry is weakened by the elevated altitude, long-range transport is distinguishable, despite the severe pollution of the region. It is confirmed by the size of SAP, which varies from <1e9.5 mm and indicates distal particles. 3. Lead isotopes have indicated a significant contribution of the neighbouring industry to the contamination of Snie _ zka. 4. Cr and Ni depositions are strongly influenced by Ni ore smelting and production of stainless steel, also attesting of long-range transport, probably from Germany. 5. The 239þ240 Pu activity concentrations fingerprint the global fallout peak of nuclear weapon tests, while Chernobyl signal is confirmed only by 137 Cs activity. 6. The uranium activity concentration profile displays a complex pattern, and the section corresponding to AD 1828e1938 is clearly dissociated from the topmost part of the profile. The observed disequilibrium in the 234 U /238 U ratio, exceeding 1.0, indicates an anthropogenic influence on the release of more mobile 234 U. 7. Overall, the Sudetes behave like a typical mountain critical zone despite the localisation on lower altitude, and therefore receive more regional, long-range sources of aerosols.