Speleothem oxygen record-thermal or moisture changes proxy ? A case study of multiproxy record from MIS 5 / MIS 6 age speleothems from Demänová Cave System

The speleothems are an important source of paleoclimatic information in the land environment. The basic advantages of speleothems are the high potential of preservation; the possibility of precise dating by the U-series method; many different 10 proxies like stable isotopes, trace elements, and microfabric which can be interpreted in the term of paleoclimate. The JS9 stalagmite was collected in Demänová Cave System (Slovakia). Presently this region of Europe is under influence of transitional and continental climate. However, in the past, it could be under stronger influence of the continental climate during cold glacial episodes and under wetter transitional climate during interglacial. The multiproxy record of the JS9 stalagmite represents ca. 60 ka period (143 – 83 ka). The multiproxy interpretation of the JS9 record shows that long time tendencies of 15 δ18O have thermal nature while the short time δO signal reflects changes in humidity. In opposition to the records from the Alps and the northern Tatra mountains, the δO record of JS9 has instant decrease episodes during Termination II.

log construction based on methodology proposed by Frisia (2015). The microscopic analyses were performed in the Institute of Geological Sciences at the Polish Academy of Sciences (Warsaw, Poland).

U-series and age-depth model
Ten calcite samples (0.1 -0.5g) were collected by drilling of the JS9 speleothem through its growing axes. The samples were drilled as thick as possible with average thickness of 2.5±0.2 mm. The chemical preparation oof the samples was made 95 at the U-series Laboratory of the Institute of Geological Sciences, Polish Academy of Sciences (Warsaw, Poland). At the beginning of chemical procedure the spike ( 233 U, 236 U and 229 Th) was added to control its efficiency. At first step, the samples were heated up for the decomposition of potential organic matter. After that the samples were soluted in nitric acid.
Finally the uranium, and the thorium were separated from the solution by chromatographic method using TRU Resin (Hellstrom, 2003). Apart from the regular samples the internal standards and blank samples were processed by the same 100 procedure. The measurements of U and Th isotopic compositions of all the samples and standards were done at the Institute of Geology of the CAS, v. v. i. (Prague, Czech Republic) by a double-focusing sector-field ICP mass analyser (Element 2, Thermo Finnigan MAT). The spectrometer settings was at a low mass resolution (m/Δm ≥ 300).
We assume the initial contamination of the samples by 230 Th, 234 U and 238 U isotopes. For the correction of obtained ages, a changed version of the Hellstrom algorithm was applied (Hellstrom, 2006). The used algorithm searches for the lowest values 110 of initial contamination by 230 Th, 234 U and 238 U isotopes from detrital sources, which were able to correct series of ages in stratigraphic order. The age-depth model was calculated by the MOD-AGE algorithm (Hercman and Pawlak, 2012).

Stable isotopes
The samples for measurement of stable isotopic composition were drilled by the Micro-Mill with a drill bit diameter of 0.1 115 mm. The final number of obtained samples was 290. At the first stage, JS9 sample was sampled along with its growth axe with a resolution of one sample/mm. To minimalize the difference in resolution between the lower and upper part of the studied record caused by the sedimentation rate, which is slower for the lower part. The lower part of a stalagmite from 0 to 40 mm was additionally sampled with a resolution of one sample/0.3 mm. The isotope composition of O and C were measured by a Thermo Kiel IV carbonate device connected to a Finnigan Delta Plus IRMS spectrometer in dual inlet mode. The results were 120 normalized to three international standards, NBS 19, NBS 18, and IAEA CO 8, and were reported relative to the V-PDB international standard. The analytical precision (1σ) was better than 0.03 ‰ and 0.08 ‰ for δ 13 C and δ 18 O, respectively. The https://doi.org/10.5194/cp-2020-125 Preprint. Discussion started: 26 October 2020 c Author(s) 2020. CC BY 4.0 License. reproducibility was checked by measurement of two internal standards after every 12 samples (for δ13C: ±0.03 ‰; for δ 18 O: ±0.08 ‰). The analyses were performed in the Stable Isotope Laboratory (Institute of Geological Sciences, Polish Academy of Sciences) in Warsaw. 125

Trace elements
The trace elements content was analysed from thin sections by an Analyte Excite Excimer Laser Ablation System with a wavelength of 193 nm connected with an Element 2 inductively coupled plasma mass spectrometer (Thermo Finnigan), using a laser output of 50% with 10-Hz pulses, we achieved a fluence of 2.44 J/cm 2 . The width of each line was 50 μm, and the laser speed during each scan was 5 μm/s. Additional details of the LA-ICP-MS analytical procedure were described by Eggins et 130 al. (1997). The measurements of near-surface trace elements content, namely: Mg, Sr, Na, Ba, P Si, Fe, Mn were performed at medium resolution. The obtained raw data were normalized to Ca. Finally the data were smoothed by the adjected averaging method using 10 nearby data points.

Petrography 135
The results of petrographic studies are presented on (Fig. 2). JS9 sample is a 155 mm long columnar stalagmite with 80 mm diameter. Macroscopically the JS9 stalagmite is built from laminated calcite (Fig. 2 A). The colour of the lamina's changes from light crème to dark brown ( Fig. 2 A). The part of the stalagmite between 75 and 85 mm has a grey colour. The light crème laminas between 40 and 75 mm have a zone of macroscopically visible porosity in the axial part of the stalagmite. A 140 microscopic analysis of the calcite crystal appearances and the identification of the texture features of the studied material shows that most of the observed stalagmite is composed of columnar polycrystals with length to width ratio usually > 10:1.
The "fibre-like" calcite individuals compose each polycrystal. The overall appearance of this layer is like spherulite consisted of bundles of elongated crystals bending outward (Fig. 2 B, C). The polycrystals show brush extinction converging away from the substrate when the rotating table is turned clockwise. The characteristics mentioned above are like those described in the 145 work of Frisia (2015), which indicates that Columnar radiaxial fibrous (Crf) is a dominant fabric in JS9 stalagmite (Fig. 2 B).
Parts of stalagmite built from Crf are separated by usually thin layers dark in Cross Polarized Light consisted of small calcite crystals and detrital material (Fig. 2 C). The appearance of these thin layers indicating on Micrite fabric (M) described in the work of Frisia (2015). The micrite fabric is most common in the middle part and in the youngest layers of the JS9 stalagmite ( Fig. 2 C).
The results of 10 U-series dates are presented in Table 1. The reported errors are 2σ, they vary from 0.8 to 2.8%. The analyzed samples did not show any visible detrital contamination at the dissolution stage. However, 4 from measured samples have the 155 230 Th/232 Th ratio lower than 300. In the case of measurement by mass spectrometry, those samples should be considered as the samples contaminated by the detrital thorium (Hellstrom, 2006). Therefore, the whole profile was corrected by using a modified procedure proposed by Hellstrom (2006). Used procedure considers the possibility of contamination not only by 230Th like in original Hellstrom's procedure but also by 234U and 238U (Błaszczyk et al., 2020). The result of correction shows that the corrected ages are within the error range of the un-corrected ages (Table 1). 160 Based on the U-series dating results, the age-depth models for JS9 stalagmite (Fig. 3) were created. According to the obtained age-depth model, the deposition of JS9 stalagmite started at 142±4 ka and ended at 84±3 ka. The JS9 stalagmite growth rate is not uniform. From 142 to 110 ka its growth rate is relative slow 1.4 mm/ka, after 112 ka it has the episode of fast grow 11.5 mm/ka, that episode ends at 108.5 ka and the grows rate slow down to 1.9 mm/ka, last intensive change of grows rate is after 94.5 ka and it increases to 4.2 mm/ka. 165

Stable Isotopes
The obtained isotopic records ( Fig. 4 A A). The δ 13 C record express changes of its values from -1‰ to -9.8‰ ( Fig. 4 B). The average amplitude of the δ 13 C value for short time changes is close to 1‰. In opposition to δ 18 O record, the δ 13 C curve is dominated by episodes of lower and higher values. They are divided by large-scale shifts ( Fig. 4 B). From 143 to 139 the δ 13 C value rise to -1. Since 139 to 130 ka (9ka 175 long) the δ 13 C value drop from -2 to -7‰ (5‰) and it has low ca. -8 to 110ka. Since 110 to 107 ka δ 13 C value growths from -9.3 to -2.6‰ (6.7‰) and decrease from -1 to -8.5 ‰ (7 ‰) at 101 ka. Since 100 to 85 ka the value of δ 13 C oscillates around -8.2 ‰. After 85 ka the δ 13 C value growth to -5.6‰ (Fig. 4 B).

Trace elements 180
The results of trace elements content measurements are presented on Records of Na, P, Fe, and Mn content repeats a similar pattern (Fig. 4 F, G, H, I), they have three intervals of increased values: before 138 ka; from 106 to 98 ka, and after 92.5 ka. From 98 to 92 ka the records of Na, P, Fe, and Mn content have the interval of lower values, this interval is also visible for Ba and Sr content, only the record of Mg content behaves in a different way 190 here. The record of Fe content has the biggest number of peaks (Fig 4 H), several from them are repeated by Na, P, and Mn content (Fig 4 F, G, I). In comparison to those four records, the record of P content has the lowest amplitude of its peaks (Fig   4 G).
The record of Si content shows a few different patterns. The most visible is the short maximum at 102 ka which goes into 195 minimum almost immediately at 101ka, except that the amplitude of Si content changes is rather low. However, from 122 to 102 ka it has an increasing trend, a similar trend can be observed for Na and P content (Fig 4 F, G, J). Similarly, to records of Fe, Mn, P, Na, and Ba content the record of Si content has increased values before 138 ka and after 98 ka.

Discussion 200
The main factor shaping the δ 18 O composition of western and central European speleothems is temperature (Moseley et al., 2015;Kern et al. 2019;Comas-Bru et al., 2020). Exemplary, stalagmite form Cobra cave, located on the northern coast of Spain (Fig. 5), reflect the changes of oceanic moisture isotopic composition which is dependent from the temperature (Rossi  (Demény 2017). The δ 18 O record from studied JS9 stalagmite has general positive trend since 143 to 130 ka. According to data from other speleothems, those long-time tendency clearly relates to improvement of thermal conditions after the MIS-6 glaciation maximum and before MIS-6/MIS -5e transition Moseley et al. 2015;Meyer et al., 2008;Holzkamper et 210 al., 2004).
In opposition, the short time signal observed in JS9 δ 18 O record may not have only thermal nature. The short time δ 18 O signal should be interpreted together with the other proxies (Fig. 4). The other important proxy is δ 13 C, from 143 to 137 the short time δ 13 C signal of JS9 stalagmite has the same trend as its δ18O short time signal. From 137 to 130 ka the short-time signal of δ 13 C record becomes opposite to the short time δ 18 O trend (Fig. 4). The δ 13 C value of speleothems calcite depends on the 215 proportion of CO2 from a soil source and from a host rocks source. The CO2 from a soil source is enriched in 12C. The level of soil development depends on climatic conditions like temperature and humidity. During warm and wet conditions, the soil is well developed. The well-developed soil cover results in lower δ 13 C value while the high value of δ 13 C and δ 18 O proxies can be the result of dryer conditions. In opposition, low δ13C and δ 18 O may be interpreted as a sign of wetter and colder climate.
The high value of δ 13 C and low value of δ 18 O relate to the cold climate and the opposite situation can be interpreted as warmer 220 interglacial conditions (Gascoyne, 1992;Genty et al., 2006, Couchoud et al., 2009. During the periods when the δ 18 O and https://doi.org/10.5194/cp-2020-125 Preprint. Discussion started: 26 October 2020 c Author(s) 2020. CC BY 4.0 License. δ 13 C have opposite trends, the short time isotopic signal reflects more the thermal conditions (137 -130 ka), while the same trends of both proxies can be caused by changes in precipitation. In the case of a period between 143 -137ka the elevated values of δ 18 O and δ 13 C can be interpreted as a period of dry and cold continental climate (Fig. 4 A, B).  (Fig. 6). The difference between Low Tatra Mountains and located on the northern slopes of Tatra Mountains caves (ca. 39 km towards to NE) shows that, Tatra Mountains were important 240 climatic barrier for the moisture at that time. In case of JS9 stalagmite, the δ 13 C and trace elements content records did not show any signal which could be equivalent to the rapid change on δ 18 O record at 130 ka (Fig 4). Therefore, recorded 1.2‰ negative shift must be caused by factors which affect only the δ 18 O proxy. Generally, in the region of Central Europe, the beginning of MIS-5e is related to change from continental to more transient climate (Demény et al. 2017;Moseley et al. 2015). The MIS-5e in JS9 stalagmite can be divided into two parts. First part (127 -122 ka) has lower values of δ 18 O and δ 13 C which 255 suggest well developed soil cover and wetter climate (Fig 4 A, B). Additionally, the Mg content during this period is lower than its average value with local minimum at 124 ka, which support the interpretation about wetter climate (Fig 4 C). Therefore, the climate during this period was more transitional than continental.
The second part (122 -115 ka) has elevated values of δ 18 O and δ 13 C, which can be interpreted as a long time change to a more continental climate with dryer conditions and less-developed soil cover. The records of Sr, Fe, Mn, P contents have picks of 260 high values (Fig 4 G, H (Demény et al. 2017;Pawlak et al., 2020). 265 Record from Magurska cave located in Tatra Mts. (Fig. 6) seems to be more similar to JS9 record during the period of 122 -115 ka than during older 127 -122 ka period. Other European records show high sensitivity for changes in the amount of precipitation during MIS -5e. Exemplary, the δ 18 O record of a stalagmite from Bourgeois-Delaunay cave (Couchoud et al. 2009) shows millennial changes with amplitude lower than 1 ‰. Those changes are repeated by changes in the δ 13 C record (Fig. 6). However, the δ 13 C changes are not synchronous, and they are shifted several hundred years towards younger ages. 270 The authors' interpretation here considers the changes in the amount of precipitation as the main driver of δ 18 O changes and slower vegetation response.
During the 107 -102ka interval the δ 18 O and δ 13 C values of JS9 stalagmite were elevated and their values are the highest in comparison to the whole recorded period (Fig 4). The elevated values of stable isotopes relate to elevated values of Fe, Mn, P, and Na content. Elements, such as Fe, Mn, and Si may be transported as detrital particles or submicron-size colloids (Fairchild 275 and Treble, 2009). Additionally, all elements that can be incorporated into the calcite structure can be transported as the absorbed ions on the clay mineral structure. During dryer periods under higher aeolian supply conditions, particles can be transported into the cave environment without water transportation (Hu et al. 2005). In the case of JS9 stalagmite, the interpretation of dry (107 -102ka) interval is probable and is in accordance with δ 18 O, δ 13 C, and Mg proxies. The elevates values of geochemical proxies relate to micrite microfabric (Fig 4, K). It appears plausible that the presence of micrite fabric 280 (M) is indicative of bio-influenced processes as micrite layers may be associated with shifts to more positive values in the C isotope ratios (Kaźmierczak et al., 1996). In the Nullarbor sample the δ 13 C values shifts from -10.5‰ to -4.0‰ in stromatoliticlike micrite (M) layers. This phenomenon was interpreted as a possible result of microbial colonization of the speleothem surface during a dry period (Frisia et al., 2012). According to all those informations, the 107 -102ka interval in the record from JS9 stalagmite can be interpreted as a stable period of a dry continental climate. However, the record from the other 285 regions shows the important climate changes at that time. Exemplary, the Greenland ice cores records at the time of 107 -102 ka (Fig.6)  Magurska Cave (Fig. 6) and by the records from the northern rim of the Alps (Boch et al., 2011). Therefore, the JS9 record expresses specific local conditions during the 107 -102 ka period. 290 The fast change of δ 18 O record of JS9 stalagmite towards to the lower values at 101 ka is repeated by similar behaviour of δ13C, Mg, Sr, Na, P, Mn, Si proxies (Fig 4 A, B, C, D, F, G, H, J) and it is related to end of microfabric from micrite (M) and beginning of Columnar radiaxial fibrous (Crf). That reflects the beginning of more humid and colder conditions. This episode happens synchronically to the cessation of growing the stalagmite from Magurska cave (Tatra Mts.). The next short episode of low δ18O values of JS9 record at 96 -94ka is also expressed on the NALPS δ 18 O records (Fig 6)  After the 92 ka the Mn, Fe, P, Na, Ba and Sr content in JS9 record start to grow (Fig. 4). It can be related to the increased weathering processes due to poor vegetation conditions and in consequence the lower level of soil development. After 85 ka 300 it is expressed by elevated value of δ 13 C proxy and finally the cessation of JS9 stalagmite growth at 83 ka.

315
• The negative shift during termination II is not observed in the records of the same age located on the west in Alps and on the northern slopes of Tatra Mountains. It shows that Carpathian Belt was important climatic barrier at that time.
• During the MIS-5e the δ 18 O record of JS9 stalagmite has stable mean value ca. -7.6 ‰. The observed ca. 1‰ short time oscillations relate to changes in amount of atmospheric precipitation. This interpretation is supported by the δ 13 C, Mg, Sr and Ba proxies.

Competing interests 330
The author declare that he has no conflict of interest 8. Acknowledgements.