Intensified microbial sulfate reduction in the deep Dead Sea during the early Holocene Mediterranean sapropel 1 deposition

Abstract The hypersaline Dead Sea and its sediments are natural laboratories for studying extremophile microorganism habitat response to environmental change. In modern times, increased freshwater runoff to the lake surface waters resulted in stratification and dilution of the upper water column followed by microbial blooms. However, whether these events facilitated a microbial response in the deep lake and sediments is obscure. Here we investigate archived evidence of microbial processes and changing regional hydroclimate conditions by reconstructing deep Dead Sea chemical compositions from pore fluid major ion concentration and stable S, O, and C isotopes, together with lipid biomarkers preserved in the hypersaline deep Dead Sea ICDP‐drilled core sediments dating to the early Holocene (ca. 10,000 years BP). Following a significant negative lake water balance resulting in salt layer deposits at the start of the Holocene, there was a general period of positive net water balance at 9500–8300 years BP. The pore fluid isotopic composition of sulfate exhibit evidence of intensified microbial sulfate reduction, where both δ34S and δ18O of sulfate show a sharp increase from estimated base values of 15.0‰ and 13.9‰ to 40.2‰ and 20.4‰, respectively, and a δ34S vs. δ18O slope of 0.26. The presence of the n‐C17 alkane biomarker in the sediments suggests an increase of cyanobacteria or phytoplankton contribution to the bulk organic matter that reached the deepest parts of the Dead Sea. Although hydrologically disconnected, both the Mediterranean Sea and the Dead Sea microbial ecosystems responded to increased freshwater runoff during the early Holocene, with the former depositing the organic‐rich sapropel 1 layer due to anoxic water column conditions. In the Dead Sea prolonged positive net water balance facilitated primary production and algal blooms in the upper waters and intensified microbial sulfate reduction in the hypolimnion and/or at the sediment–brine interface.


| INTRODUC TI ON
The Dead Sea is a terminal hypersaline lake (maximum water depth of ~300 m) located within a tectonically formed depression in an arid part of the Eastern Mediterranean ( Figure 1). Before human intervention its water level reflected the water balance between freshwater influx, originating mostly from Mediterranean-derived winter rainfall within the northern and central region of its 40,000-km 2 watershed, and evaporation (Enzel et al., 2003). Being highly concentrated (total dissolved solids: ~350 g L −1 ; density: ~1.24 Kg L −1 ), the Dead Sea is an extreme natural habitat, although specific populations of prokaryotes such as halophilic archaea and bacteria are adapted to this environment (Nissenbaum, 1975;Oren, 2010). The high brine density of the Dead Sea makes it extremely susceptible to stratification following heavy rainfall-derived freshwater runoff events.
Generally two conditions limit primary production in the Dead Sea: (1) high salinity and (2) phosphate nutrient availability (Oren et al., 2004;Oren & Shilo, 1985), both factors, which are impacted by freshwater runoff. Over the past few decades, spontaneous blooms of the unicellular primary producer green alga Dunaliella parva directly followed by a dense, red-pigmented halophilic archaea bloom have appeared twice, in the summer of 1980 and the spring of 1992, following particularly rainy winters with increased freshwater runoff that led to stratification of the Dead Sea water column and the formation of less saline surface waters (Oren, 1993;Oren et al., 1995).
Prior to 1979 overturn of the Dead Sea water column, the deep layer of the lake (hypolimnion) was anoxic and contained hydrogen sulfide (H 2 S) (Nissenbaum, 1975;Nissenbaum & Kaplan, 1976). The sulfur isotope values of the H 2 S (δ 34 S H 2 S = −19.6 to −21.7‰) were lower than that of the dissolved SO 4 2− (δ 34 S SO 4 = +15.9‰), and the H 2 S was suggested to have formed via microbial sulfate reduction (MSR) in the hypolimnion and/or in the underlying sediment interstitial water. In its generalized form, organoclastic-type MSR involves the oxidation of an organic substrate coupled to the reduction of SO 4 2− (Equation 1): where CH 2 O represents the mean of a variety of organic compounds in a subsurface marine setting (Arndt et al., 2013), and H 2 S the final reduced S product.
Whether modern deep-dwelling MSR communities in the hypolimnion and/or deep sediments respond to transient periods of increased freshwater runoff is unknown, however, the sediment record may be used to shed light on whether such a connection occurred in the past. There have been a number of studies on MSR in the paleo-Dead Sea using the distribution and stable isotope composition of S-bearing minerals, such as gypsum (CaSO 4 ·2H 2 O), pyrite (FeS 2 ), elemental sulfur nodules (S 0 ), greigite (Fe 3 O 4 ), and mackinawite (FeS) (Bishop et al., 2013;Thomas et al., 2016;Torfstein et al., 2005Torfstein et al., , 2008Torfstein & Turchyn, 2017).
Investigation of MSR in paleolimnological studies has typically focused on the distribution and isotope composition of authigenic S-bearing minerals. In modern environmental studies a more direct approach, such as measuring dissolved SO 4 2− concentrations and its stable isotopic composition, is typically used (Avrahamov et al., 2014;Gavrieli et al., 2001).
During late 2010 and early 2011 the Dead Sea Deep Drilling Project (DSDDP) was undertaken as part of a joint international scientific effort and the International Continental Scientific Drilling Program (ICDP). Core 5017-1-A drilled from the deep Dead Sea (water column of ~300 m) comprises layered evaporites and clastic deposits spanning around 450-m length and provides a nearcontinuous depositional record covering >200 ka (Neugebauer et al., 2014;Torfstein et al., 2015). The core sediments allows for an investigation of the deep Dead Sea subsurface biosphere, an extreme habitat characterized by high salinity, toxic concentrations of divalent cations, and a lack of labile organic matter (OM) and nutrients (Bodaker et al., 2010;Ionescu et al., 2012;Oren, 1999).
The analyses of biosignatures associated with this deep biosphere through the study of DNA and organominerals have been linked to changes in microbial diversity, microbial activity, and past paleoenvironmental and paleolimnological conditions (Thomas et al., 2014;Thomas et al., 2016).
(1) 2CH 2 O + SO 2− 4 → H 2 S + 2HCO − 3 F I G U R E 1 Map of the study region with the location of Dead Sea DSDDP core 5017-1-A (red star) The pore fluid samples extracted from the deep Dead Sea sediments in core 5017-1-A were shown to have been derived from deep lake brine trapped during sediment deposition . Subsurface advection and diffusion of pore fluid dissolved constituents were significantly hampered due to the abundant impermeable halite layers, high dynamic viscosity of hypersaline fluids, absence of bioturbation, and overall fast sedimentation rate . This is particularly true for the upper ~90m of core sediments corresponding to the Holocene. Pore fluid magnesium (Mg 2+ ) and bromide (Br − ) concentrations are conservative in core 5017-1-A (i.e., they did not significantly participate in chemical reactions and remained soluble; Supplementary Figure S1), and these ions provide a unique multimillennial scale record of dilution and concentration of the deep Dead Sea . The relatively high resolution of pore fluid compositional changes in the halite-rich early Holocene sediments together with evidence of significant hydroclimate variability that occurred during this time makes it a suitable interval for reconstructing both microbial and hydroclimate processes, and for determining possible connections between them. During the early Holocene, there was a period of prolonged and relatively wet regional hydroclimate conditions which resulted in depositional changes in the Dead Sea (Neugebauer et al., 2017). In this study, we investigate the 5017-1-A core sediment distribution and pore fluid concentrations, combined with pore fluid stable S, O, and C isotope ratios, lipid biomarkers, and organic C isotope compositions to reconstruct regional hydroclimate changes, microbial processes, and microbial sulfate reduction activity in the early Holocene Dead Sea.

| DSDDP 5017-1-A Holocene chronology
The 456-meter sedimentary core 5017-1-A was extracted from the deepest part of the lake at 300-m water column depth (~720 m below mean sea level) at 31°30′28.98″N, 35°28′15.60″E. The chronology of the investigated sediment sequence for Holocene sediments ( Figure 2a) is based on a linear interpolation of calibrated radiocarbon ages derived from terrestrial plant remains from site 5017-1 (Kitagawa et al., 2017;Neugebauer et al., 2014;Figure 2b).
The calibration was done using the IntCal13 calibration dataset (Reimer et al., 2013) within the OxCal 4.3 software.

| Binary net lake water balance curve
A binary net lake water balance curve for the Holocene was drawn based on the appearance of indicative sediment facies in core 5017-1-A (Neugebauer et al., 2014;Torfstein et al., 2015;Figure 2c)

| Pore fluid extraction, handling, and analyses
A description of sediment handling for core catcher material, taken during the drilling campaign of winter 2010-2011 and core section sediments, taken during July 2012, including pore fluid extraction methods, post-extraction treatment, and measurement of major cation and anion concentrations are given by Levy et al. (2017Levy et al. ( , 2019 approach at 25°C (Parkhurst & Appelo, 1999;Pitzer, 1973;Reznik et al., 2009). (2) where I. A. P. and K CaSO 4 •2H 2 O denote ion activity product and the gypsum solubility product, respectively. and is the fractionation factor during MSR. An isotope fractionation ( ) range between 22‰ and 67‰ was used (Deusner et al., 2014).

| Quantification of MSR
Following the calculation of f, the pre-MSR SO 4 2− concentrations and degree of saturation with respect to gypsum were estimated.
Additionally, we use a stoichiometric ratio of organic C:SO 4 of 2:1 based on the generalized pathway of MSR (i.e., Equation 1) to obtain a first-order approximation for the amount of organic carbon oxidized (weight %; relative to sediment) by the calculated  (Neugebauer et al., 2014 andKitagawa et al., 2017). Green triangles indicate depths where lipid biomarkers were analyzed for this study (samples S17, S18, S19, S21; Figure

| Calculating δ 13 C of added DIC
The observed changes in DIC concentrations and respective 13 C DIC in measured pore fluid compositions were used to calculate the addition of DIC derived from OM oxidation (between two corresponding depths: initial (i) to final (f)) using the following mass balance equation for a closed system (Equation 5):

| Lipid biomarker extraction and analysis
Four sediment samples between 65 and 91 meters below the lake floor (mblf) were analyzed for their lipid composition (S17, S18, S19, and S21; Figure 3). Samples were freeze-dried, ground, and extracted through sonication (methanol (MeOH) twice, MeOH/dichloromethane (DCM) (1:1) twice, and DCM three times). Sulfur was removed by activated copper. Precipitates were filtered out and lipids separated into five fractions by chromatography over a deactivated column of silica gel. Fraction F1 was eluted with hexane/DCM (9:1), F2 with hexane/DCM (1:1), F3 with DCM, F4 with ethyl acetate, and F5 with MeOH. Fraction F4 was silylated with pyridine/bis(trimethylsilyl) trifluoroacetamide (BSTFA) 1:1 (v/v). Only fractions F1 and F4 are presented in this article. Gas chromatography-mass spectrometry analyses (GC-MS) were performed using an HP 6890 Series Plus gas chromatograph equipped with a cool on-column injector and coupled to an Agilent 5975C (VL MSD) mass spectrometer. The GC was equipped with an HP5 column (30 m × 0.25 mm × 0.25 µm, RESTEK). Samples were injected at 60°C (held for 30 s) before oven temperature was increased to 130°C at 20°C min −1 , then to 250°C (5°C min −1 ) and 300°C (3°C min −1 , held for 45 min). Compoundspecific carbon isotope (δ 13 C) analyses were done using an HP7890B GC coupled to an Isoprime visION isotope ratio mass spectrometer via a GC-5 combustion interface operating at 870°C. The GC was equipped with a BPX5 column (30 m × 0.25 mm × 0.10 µm, SGE Analytical Science) and a cool on-column injector, and the oven temperature was programmed for GC-MS analyses. The B4 standard mixture (Arndt Schimmelmann, Indiana University, USA) was used to externally calibrate compound-specific 13 C values, and 13 C values of alcohols were corrected from the BSTFA-derivatizing agent (Thomas, Grossi, et al., 2019).

| Dead Sea early Holocene sediment interval
The Holocene interval of 5017-1-A is 88 m. The early Holocene "interval of interest" (a term we will use hereafter) is a depth interval located between 70 to 61 mblf and dates to ca. 9500- (4) Halite (NaCl) evaporite layers dating to the start of the Holocene deposited as a result of a sequence of lake level drops (negative water balance), which began at the end of the last glacial period (Torfstein et al., 2015).   Ar-OH F1 Ar S17 S18 S19 S21 S17 S18 S19 S21  Figure S1) together suggest that halite dissolution likely occurred coeval with lake dilution as emphasized by the Mg 2+ (Kiro et al., 2017;Levy et al., 2017Levy et al., , 2018. (i.e., "sulfur pump model"; Gavrieli et al., 2001;Torfstein et al., 2005).

| Sulfate
Saline springs on the western side of the Dead Sea, which have concentrations in excess of modern-day brine (~10 mm), may also contribute sulfate to the Dead Sea (Gavrieli et al., 2001;Torfstein et al., 2005;Weber et al., 2022). Springs such as the present-day Ein Qedem were activated following the regional aridity and associated lake level drops at the end of the late Pleistocene (Weber et al., 2018) and may have been active at the start of the Holocene (Weber et al., 2022). An additional SO 4 2− source and enrichment mechanism for the hypolimnion is the dissolution of calcium sulfate minerals in the upper waters (i.e., at the Mt. Sedom salt diapir) and transfer via gravity-driven brine flows internally in the Dead Sea, a mechanism that was suggested for the last glacial period (Levy et al., 2019).
Other potential sources include SO 4 2− enrichment occurring via in situ gypsum dissolution, in the subsurface or early Holocene water column, or reduced mineral bound S oxidation. Gypsum dissolution may occur directly following MSR in the hypolimnion or the subsurface sediments provided that SO 4 2− concentrations decrease to the extent that the solution becomes undersaturated with respect to gypsum (Gavrieli et al., 2001;Torfstein et al., 2005).
Unlike isotope fractionation associated with both organoclastic sulfate reduction and sulfate reduction coupled to methane oxidation (AOM).
Assuming closed system and Rayleigh-type distillation isotope fractionation (Equation 3), we estimate that between ~66 and 30% of the SO 4 2− had thus been reduced (i.e., f = 0.33 and 0.70). Pore fluid SO 4 2− concentrations in the interval of interest remain at or slightly above saturation with respect to gypsum (Figure 2e). This would suggest that one of the following mechanisms may have occurred: (a) prior to MSR, pore fluid was initially at a higher degree of supersaturation with respect to gypsum (mechanism A), (b) the MSR was accompanied by gypsum dissolution (mechanism B), or (c) a combination of both mechanisms occurred (i.e., A followed by or alongside B would have been reduced (Gavrieli et al., 2001;Torfstein et al., 2005) and facilitated by more OM oxidation. , S 2+ ) and there is an exchange of oxygen isotopes with water before the final sulfide product is released to the surrounding aqueous anoxic environment (Rees, 1973).
As some of these intermediates are re-oxidized back to SO 4 Viewed as the slope in a δ 18 O SO 4 vs. δ 34 S SO 4 plot, relatively steep slopes have been shown to characterize relatively slow net MSR rates (as low as 10 −12 mol cm −3 year −1 ; representing higher re-oxidation) while moderate slopes characterize relatively fast net MSR rates (maximum of 5 × 10 −4 mol cm −3 year −1 ; limited re-oxidation) (Antler et al., 2013. Other processes are known to have an effect on the δ 18 O SO 4 vs. δ 34 S SO 4 plot such as the sulfide and pyrite oxidation (Balci et al., 2007) and disproportionation (Böttcher et al., 2001(Böttcher et al., , 2005; however, their effect should only decrease the relative estimated calculated rates of MSR. Additionally, it was shown that when the sulfate reduction rates are high (i.e., the δ 18 O SO 4 vs. δ 34 S SO 4 slope is low), the effect of transport such as diffusion and sedimentation is negligible on the measured slope (Fotherby et al., 2020).  δ 18 O SO 4 slopes than all the modern regional sites (~1.1 to 2; Figure 4) and were suggested to have recorded a very slow net MSR rate in the stratified (paleo) water column. Comparatively, gypsum from the preceding interglacial Amora Fm. has a slope of ~0.55 (Torfstein & Turchyn, 2017).
The slope value of 0.26 from the interval of interest is comparable to the slope from areas of intensified MSR such as the East Anglian salt marshes (black filled triangles; the slope of 0.24) (Antler et al., 2019). It may also be comparable to modern hypersaline groundwater around the shores of the western Dead Sea banks (Avrahamov et al., 2014). It is reasonable to assume that rates of MSR may have been high; in the Dead Sea, microbial mats from underwater emerging springs close to the western shores have large spatial heterogeneity in sulfate reduction rates, with values up to 10 nmol cm −3 day −1 (4 × 10 −6 mol cm −3 year −1 ) detected in saline springs (Häusler et al., 2014).
In this model for interpretation of the evolution of δ 34 S SO 4 vs. there is an increase of DIC from the mean value of 0.6 mmol Kg −1 to between 2.1 and 1.9 mmol Kg −1 , coupled with a drop in δ 13 C DIC from mean value of −5.4‰ to between −16‰ and −12‰ (Figure 2g).
These δ 13 C DIC values would be indicative of extensive AOM-SR had they been more negative. Partial organic carbon oxidation into a relatively large and isotopically 13 C-enriched DIC pool may have occurred. An isotope mass balance calculation can be used to determine the δ 13 C of OM oxidized. Using averages of pre-peak δ 13 C DIC & DIC and peak δ 13 C DIC & DIC values for this mass balance calculation must also account for lake dilution (as expressed by Mg 2+ ) and an unknown DIC enrichment, which adds uncertainty. Thus, a mass balance calculation that calculates the δ 13 C of OM oxidized within the peak δ 13 C DIC depth interval based on discrete measurements is used (see methods; Equation 5), providing a δ 13 C OM value of −47‰. This estimate is lower than the range of primary derived OM in the Dead Sea ( Figure 5) and closer to that of CH 4 (−40‰) found in the Dead Sea hypersaline groundwater along the western margins (Avrahamov et al., 2014). However, this calculation is somewhat speculative as it is based only on two discrete measurements; had there been more measurements in this interval the calculated δ 13 C OM would give more confidence. Furthermore, it should be noted that the depth of maximum δ 18 O SO 4 and δ 34 S SO 4 values overlap only partly with the depths of the lowest δ 13 C DIC peak (Figure 2g). This is due to the low vertical resolution of pore fluid sampling and limitation in volumes of extracted pore fluid that did not allow measuring δ 18 O SO 4 and δ 34 S SO 4 at the most depleted δ 13 C DIC anomaly.
Lipid biomarkers from sediments can be used to detect evidence of methanogenesis, the presence of anaerobic methanotrophic archaea (ANME), and syntrophic sulfate-reducing bacteria (SRB) involved in AOM (Blumenberg et al., 2004). Nonisoprenoid macrocyclic glycerol diethers (a-g in Figure 3d; S17, S18, S19) support the existence of bacterial communities potentially involved in the S-cycle and have been detected in hypersaline sulfidic sediments (Baudrand et al., 2010) or environments with AOM (van Dongen et al., 2007). Macrocyclic archaeol (Figure 3d; S17) has been identified in the methanogenic archaea Methanococcus janaschii isolated from a deep-sea hydrothermal vent (Comita & Gagosian, 1983). Phytane detected in S17 and S18 (Figure 3c) can constitute a marker of methanogenic archaeal communities (Risatti et al., 1984;Schouten et al., 1997;Tornabene et al., 1979), although this compound has also been attributed to halophilic archaea in hypersaline environments and, in particular, the Dead Sea realm (Anderson et al., 1977). It is noteworthy that crocetane and pentamethylicosane (PMI), which can be strong indicators of AOM  were not detected in the investigated S17, S18, S19) (Hinrichs et al., 2000). The combination of archaeol and hydroxyarchaeol has been proposed as a biomarker signature specific to some CH 4 seeps (Blumenberg et al., 2004;Knittel & Boetius, 2009), where anaerobic CH 4 oxidizers and methanogenic archaea can co-exist and produce a similar set of biomarkers (Schouten et al., 1997). Unfortunately, the low amount of hydroxyarchaeol in the investigated samples did not allow for the measurement of its δ 13 C values that could have informed us on the occurrence of AOM. On the other hand, the range of δ 13 C values measured for Dead Sea archaeol (between −30 and −25‰) indicates an origin of this compound different from ANME, and its abundance along with that of extended archaeol (Figure 3d) rather supports halophilic archaea of the Haloarchaea class as a major biological source of archaeol (Vandier et al., 2021). These organisms classically dominate in the Dead Sea modern and ancient waters (Bodaker et al., 2010;Thomas & Ariztegui, 2019), and sediments (Grice et al., 1998;Thomas et al., 2015, Thomas, Grossi, et al., 2019. The set of biomarkers and their carbon stable isotope composition present in this interval therefore support a hypersaline environment dominated by halophilic archaea. The slightly depleted δ 13 C values measured for the total lipid extracts (ranging between −38‰ and −36‰) compared to that of archaeol still potentially argue for the presence of methane in this environment, without indicating AOM-SR.

| Autochthonous vs. allochthonous sources of OM
The availability of OM and its lability are important factors determining the rate of MSR in saline environments (marine : Schubert et al. (2000); lacustrine: Glombitza et al., 2013). In the modern Dead Sea, there is a low OM influx from the water column to the sediment (low benthic oxygen uptake of 0.46 mmol m −2 day −1 ), a result of the general lack of primary production and low input of terrestrial organic carbon to the lake (Häusler et al., 2014). It has previously been suggested that the slow rate of MSR in sediments of the Dead Sea is due to the lack of availability and/or OM quality (Häusler et al., 2014;Thomas et al., 2016). Lipid biomarkers can be used to estimate the presence of allochthonous and autochthonous derived OM. The presence of n-alkanes with an odd-over-even carbon number predominance centered at C 29 and C 31 in all sediment samples investigated (S17, S18, S19, S21; Figure 3c) infers a proportion of allochthonous derived OM, probably from plant waxes (Meyers, & Ishiwatari, 1993). This confirms findings by Oldenburg et al. (2000) defining terrestrial (allochthonous) OM as an important source to the Dead Sea sediment.
However, in the sediment interval of interest (S17 and S18), the presence of n-C 17 alkane points to an additional contribution from cyanobacteria (Sachse et al., 2006) or phytoplankton to the bulk OM, that appears together with evidence of increased freshwater influx and intensified MSR. This lipid biomarker is not found in the S19 and S20 samples from the halite-rich interval below. The presence of phytane (as a degradation product of the chlorophyll side chain in anoxic environments) may also be indicative of cya- Archaeal specific compounds in halite (core 5017-1-A) 1 Archaeal specific compounds in ld and aad (core 5017-1-A) -This study Lake phytoplankton specific compounds (late Holocene sediments) 2 Total lipid extracts (core 5017-1-A) -This study -40 δ 13 C (‰ ; vs. VPDB) organoclastic type MSR (Ebert et al., 2020;Thomas et al., 2016) or alternatively led to enhanced methane production that potentially fueled AOM-SR.

| Regional climate driver of MSR in the Dead Sea
Intensified MSR and changes in lipid biomarker distribution were coeval with positive net water balance in the Dead Sea, ultimately resulting from wet regional hydroclimate conditions. Global climate change on a multi-millennial scale inferred from decreasing and increasing CO 2 concentrations in Antarctic ice-core records emphasize the natural oscillation between glacial and interglacial periods, respectively (Figure 6a; Monnin et al., 2001;Pépin et al., 2001;Petit et al., 1999). Illustrating the control of global climate on the net water balance of the terminal Dead Sea, similar trends are found in the deep Dead Sea composition as shown by decreasing (positive water balance) and increasing (negative water balance) conservative Mg 2+ concentrations from pore fluids from 5017-1-A (Figure 6b; Levy et al., 2017). Perturbing these long-term global climate patterns in the Mg 2+ record is a dilution F I G U R E 6 Paleoclimate and Dead Sea time-series records. From top to bottom (a) Compiled atmospheric CO 2 record from Antarctic ice cores (Monnin et al., 2001;Pépin et al., 2001;Petit et al., 1999); (b) Dead Sea pore fluid Mg 2+ record from 5017-1-A (correlation after Levy et al., 2017). Focusing on the last 15 kyr's: Soreq speleothem records of (c) δ 18 O (blue) and (d) δ 13 C (red) (Bar-Matthews et al., 2003;Grant et al., 2012); (e) Dead Sea Mg 2+ concentrations  Increased rainfall in the Eastern Mediterranean during the early Holocene is evident in speleothem records from Jeita cave (western Lebanon) (Verheyden et al., 2008), Soreq cave (central Israel) (Bar-Matthews et al., 2003), and other regional paleoclimate records (Cheng et al., 2015;Robinson et al., 2006). Calcium carbonate speleothem stable oxygen (δ 18 O) and carbon (δ 13 C) isotope ratios from the Soreq cave provide high-resolution paleoclimate records of local rainfall conditions in relative proximity to the Dead Sea (found ~50 km east; Figure 1) and suggest that the early Holocene was marked by increased rainfall relative to the preceding drier onset of Holocene/post-Younger Dryas (YD) and following mid and late Holocene (Bar-Matthews et al., 2003). Speleothem The Dead Sea received most of its freshwater runoff from the Jordan River via the northern catchment during the early Holocene (Palchan et al., 2019). Westwards beyond the Dead Sea and Soreq cave in the adjacent Mediterranean Sea and central Mediterranean region, microbial ecology responses to hydroclimate changes occurred (Ariztegui et al., 2000). In both the Mediterranean Sea and the Dead Sea freshwater-derived and less saline buoyant surface water layers formed. Increased freshwater influx to the Mediterranean Sea predominantly from the River Nile and derived from African monsoon rainfall, led to the formation of organic-rich sapropel layer S1, spanning from 10,500 to 6100 years BP (Grant et al., 2016). Stratification in parts of the Mediterranean Sea resulted in eutrophication, oxygen-poor lower water, and benthic azoic conditions, which culminated in the deposition of the sulfide and C organic rich, sapropel layer 1 (S1) (e.g., Almogi-Labin et al., 2009;Rohling et al., 2015). In the epilimnion of the stratified Dead Sea primary productivity occurred as evident by the presence of n-C 17 alkane, while in the deep Dead Sea (hypolimnion) intensified MSR occurred at/below the sediment-water interface (Figure 6f). Collectively, the marine and lacustrine evidence suggest that the microbial ecology in both the surface and deep Mediterranean Sea and Dead Sea were independently responding to the increased influx of freshwater during the early Holocene.

| SUMMARY
The present study provides insight into the microbial response in the Dead Sea to positive net water balance caused by enhanced regional hydroclimate activity between ca. 9500-8300 years BP. This was done by measuring the Dead Sea drilled ICDP core sediments pore fluid concentrations and stable S, O, and C isotopes, combined with sediment lipid biomarkers. Positive lake net water balance was accompanied by depositional changes and surface water dilution, which facilitated enhanced microbial processes in both the surface and deep Dead Sea. Analogous to modern-day spontaneous blooms of the primary producer Dunaliella parva and halophilic archaea following increased freshwater runoff, lipid biomarkers archived from the deep sediments suggest the onset of upper water column productivity following positive lake net water balance. In the anoxic hypolimnion and/or its bottom sediments there was a microbial response manifested as intensified microbial sulfate reduction.

ACK N OWLED G M ENTS
We wish to thank the editors and anonymous reviewers for their reviews that considerably improved the paper. We sincerely thank our long-term collaborators Y. Yechieli, M. Stein, and B. Lazar along with scientific members of the DSDDP drilling project in Israel and beyond. We also thank the people involved in pore fluid sampling and analysis from the GSI: I. Swaed, G. Sharabi, D. Stiber, O. Berlin; and BGU: E. Elliani-Russak, N. Avrahamov, J. Ganor, and A. Reis. We would like to also thank I. Antheaume from Université de Lyon for lipid biomarker and isotope measurements. Lastly, We deeply thank and pay our gratitude to our colleague and pioneer of Dead Sea geochemical research Dr. Mariana Stiller who passed away in July 2020.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.