Elsevier

Quaternary Science Reviews

Volume 221, 1 October 2019, 105871
Quaternary Science Reviews

Mount Sedom salt diapir - Source for sulfate replenishment and gypsum supersaturation in the last glacial Dead Sea (Lake Lisan)

https://doi.org/10.1016/j.quascirev.2019.105871Get rights and content

Highlights

  • Last Glacial Dead Sea hypolimnion compositions were investigated.

  • Dissolution of the Sedom salt diapir replenished Cl, Na+ and SO42− reservoirs.

  • Cool and diluted paleo-epilimnic waters facilitated dissolution.

  • Solute laden solution penetrated the chemocline and mixed with the hypolimnion.

  • Process facilitated the transition from a halite to gypsum precipitating lake.

Abstract

During the late Quaternary several hypersaline lakes occupied the tectonic depression of the Dead Sea Basin, depositing sequences of primary-evaporitic mineral phases: aragonite (CaCO3), gypsum (CaSO4·2H2O) and halite (NaCl). Aragonite and gypsum were the dominant primary mineral phases during the glacial periods and their formation required significant import of bicarbonate (HCO3) and sulfate (SO42−) ions to the lake. While bicarbonate was likely derived from dissolution of calcite in the watershed, the sources of sulfate remained elusive. Here we investigate and quantify the long-term sulfate reservoir changes in the deep waters (hypolimnion) of Lake Lisan (the last glacial Dead Sea) using concentrations and stable isotopes of sulfur in pore-fluids from the cores that were drilled at the lake floor (2010–11) by ICDP (International Continental Drilling Program). From ca. 117ka, pore-fluid sulfate concentrations increased and the brine attained supersaturation with respect to gypsum, peaking during the last glacial maximum (LGM; ca. 20ka). Stable isotopes of pore-fluid sulfate (δ34S and δ18O) are similar to the values found in bulk sulfate minerals from the nearby Mount Sedom salt diapir. We suggest that relatively diluted and cool paleo-epilimnion water facilitated dissolution of halite and anhydrite (CaSO4) of the Mt. Sedom salt diapir, resulting in a localized increase in solution density. Subsequently, this solution sank and mixed with saline hypolimnion water, simultaneously replenishing chloride, sodium and sulfate reservoirs, while diluting it with respect to other solutes. The mixing of the calcium-rich gypsum saturated hypolimnion and the sulfate-rich sinking brine from above resulted in gypsum supersaturation.

Introduction

Lake Lisan, the precursor of the modern Dead Sea during the last glacial period ca. 70-14 thousand years ago (ka), filled the tectonic depressions of the Dead Sea Basin and the Jordan Valley. At its highest stand, ca. 27ka to 23ka, with a lake level over 270 m (m) higher than today, it expanded over the area from Hazeva in the South to the Sea of Galilee in the North (Bartov et al., 2002; Begin et al., 1974, Neev and Emery, 1967; Haase-Schramm et al., 2004, Stein and Goldstein, 2017) (Fig. 1a). Its sheer size and lake level was determined by input of freshwater runoff from its 40 000 km2 drainage basin (Stein et al., 1997), which was controlled mostly by the East Mediterranean–Levant hydroclimate (e.g. Stein, 2001, 2014; Torfstein et al., 2013). As a result of increasing freshwater runoff, a stable stratified (meromictic) configuration developed whereby a dense saline hypolimnion layer was capped by a less dense and saline epilimnion layer (Stein et al., 1997; Torfstein et al., 2005; Lazar et al., 2014).

The sedimentary record of Lake Lisan, termed the Lisan Formation (Fm.), is abundant with aragonite and gypsum, which are the primary evaporitic mineral phases, as well as silty detritus material (Katz et al., 1977; Stein et al., 1997; Haliva-Cohen et al., 2012). Similar to the Holocene Dead Sea, Lake Lisan was composed of calcium-chloride (Ca-Cl) brine characterized by excess calcium relative to sulfate and dissolved inorganic carbon species combined (mole Ca2+ > SO42− + HCO3 + CO32−), and chloride in excess of sodium (mole Cl > Na+). The Ca-Cl brine's deficit in bicarbonate and sulfate required a supply of these ions to the lake from external sources to allow the deposition of aragonite and gypsum layers dominating the Lisan Fm. (Stein et al., 1997). It was suggested that dissolution of carbonates from Lake Lisan's watershed, including settled desert dust, by freshwater runoff, may have been the primary source of bicarbonate during that time (Belmaker et al., 2014). However, the origin of sulfate ions that allowed the precipitation of abundant gypsum layers in Lake Lisan and other lakes filling the Dead Sea Basin is less obvious (Torfstein et al., 2008; Torfstein and Turchyn, 2017).

Significant contributors of sulfate to the modern day Dead Sea are freshwater and saline springs (Starinsky, 1974; Stein et al., 1997; Torfstein et al., 2008). The upper Jordan River, which originates from the freshwater springs discharging at Mt. Hermon, as well as the Yarmouk River, Wadi Mujib, and Wadi Hasa could have provided sulfate to Lake Lisan (Torfstein et al., 2005). Additionally, freshwater runoff may also have derived sulfate from the dissolution of gypsic soils in the Lake Lisan's watershed (e.g. Palchan et al., 2019), though the extent of this sulfate source has yet to be addressed. A recent study by Weber et al. (2018) showed that the saline springs currently discharging to the Dead Sea most likely originate from Lake Lisan water. This finding raises a question whether the observed modern thermal saline springs found in and around the Dead Sea were active during the last glacial period. Indeed, in an earlier study, it was suggested that these types of Ca-Cl brines minimally contributed to the chemical budget of Lake Lisan (Stein et al., 1997).

Given the above sources of sulfate in freshwater inflow and assuming that saline springs were active during the last glacial period, it was estimated that sulfate could have accumulated in the epilimnion to the point of gypsum saturation (Torfstein et al., 2008). In order for thick (>10 cm) primary gypsum layers to have deposited, the hypolimnion must have also been saturated with respect to gypsum. To complicate matters Torfstein et al. (2005) showed that the Lisan Fm. gypsum carries an isotope signature (δ34S and δ18O) which is characteristic of sulfate that underwent microbial sulfate reduction. During microbial sulfate reduction, sulfate concentrations are decreased which would result in a decrease of the degree of saturation with respect to gypsum, and thus the saturated solution would become undersaturated. It was suggested that microbial sulfate reduction occurred in the oxygen-depleted hypolimnion, a process that enriched the isotope composition of the residual sulfate in 34S and 18O, respectively. To compensate for decreasing sulfate concentrations a ‘sulfur pump’ mechanism was suggested; gypsum that precipitated from the epilimnion sank and dissolved in the hypolimnion, thereby replenishing its sulfate reservoir and keeping it at saturation with respect to gypsum. Periodic overturn and mixing between hypolimnion and epilimnion waters then produced the isotope enriched gypsum (Torfstein et al., 2005). This suggested ‘sulfur pump’ assumed that hypolimnion salinity and composition was similar to the present day Dead Sea, and given a high calcium concentration it could not have maintained high concentration of sulfate. While this model can explain the formation of isotope enriched gypsum in the Lisan Fm., it lacks direct evidence of the chemical composition Lake Lisan's hypolimnion.

The International Continental Drilling Program (ICDP) Dead Sea Deep Drilling Project (DSDDP) (2010–11) provides for the first time a window into the chemical composition of the hypolimnion (Lazar et al., 2014; Levy et al., 2017, 2018). Pore-fluids were extracted from the DSDDP 5017-1-A core that was drilled at the center of the lake at a water depth of 300 m where over 450 m of lake sediment spanning back ca. 220ka was recovered (Neugebauer et al., 2014; Torfstein et al., 2015, Kitagawa et al., 2017). The pore-fluids show that Lake Lisan's hypolimnion was significantly diluted relative to pore-fluids from the previous interglacial, and relative to present day Dead Sea. This is best evident by the drop in the concentration of magnesium (Mg2+) and bromide (Br), which are considered to be conservative elements in the system (i.e. the concentration changes in pore-fluids are equal to one another and, thus, do not reflect chemical processes that may have occurred in the lake or during early diagenesis). Indirect evidence for this dilution is by the concentration decrease of the non-conservative chloride (Cl) (Lazar et al., 2014), but this decrease is moderated, relative to Mg2+ and Br, due to halite dissolution (NaCl) (Levy et al., 2017). Based on Mg2+ and Br alone it was suggested that Lake Lisan reached a dilution factor of ∼4 at its peak, in comparison to the modern Dead Sea (Levy et al., 2017). The increasing sodium chloride ratio (Na/Cl) in the pore-fluids in the Lisan Fm. suggests that dissolution of halite took place throughout most of the extended last glacial period (i.e. starting from ca. 117ka or marine isotope age [MIS] 5e). It was suggested that dissolution of rock salt at the Mt. Sedom salt diapir, found to the South-West of the modern Dead Sea (Fig. 1a), significantly replenished hypolimnion sodium and chloride inventories (Levy et al., 2018).

Using pore-fluid compositions we further investigate how Lake Lisan's hypolimnion composition changed. Given the suggestion that dissolution of the Sedom salt diapir played a significant role on sodium and chloride replenishment in Lake Lisan, we can also address the question whether dissolution of the salt diapir contributed to the lake sulfate budget. For this, we use the concentrations and stable isotopes of pore-fluid sulfate from the long ICDP cores, both at the deep and marginal sites of the Dead Sea, along with sulfate isotopes in minerals from Mt. Sedom. Based on the results we propose a conceptual model of solute fluxes from Mt. Sedom to Lake Lisan's hypolimnion, which can also explain the dilution of the hypolimnion under meromictic stratified conditions (Levy et al., 2017) and account for the thickness of the gypsum layers that were deposited intermittently in Lake Lisan.

Both microbial sulfate reduction and gypsum precipitation were observed in the modern Dead Sea. Before the lake overturn in 1979 (Steinhorn, 1985), microbial sulfate reduction occurred in the anoxic hypolimnion layer of the lake (Nissenbaum and Kaplan, 1976). Massive gypsum precipitation occurred at the end of the 1950's and early 1960's as water level started to recede due to lowered freshwater input into the lake (Neev and Emery, 1967). Since then, both the gypsum precipitation potential and the sulfate to calcium ratio (SO42−/Ca2+) continuously decreased (Reznik et al., 2009). The small amount of gypsum that currently precipitates in the Dead Sea is masked by massive halite (NaCl) precipitation (Herut et al., 1998; Stiller et al., 1997; Gavrieli, 1997). Nevertheless, the saturation state of gypsum continued to rise and reached Ω = 1.4 in 2009. The slow kinetics of gypsum nucleation which allows for this supersaturation may be explained by the low SO42−/Ca2+ ratio in the brine (Reznik et al., 2009).

The Mt. Sedom salt diapir (the purple area in Fig. 1b), located at the south-west side of the Dead Sea basin, comprises a sedimentary section of around 2000m mostly comprising rock-salt (halite with some anhydrite; Fig. 2a). These sediments, termed the Sedom Fm. (Zak, 1967), originally deposited in the late Miocene to Pliocene Sedom lagoon which were then buried. During the Pleistocene, the rock-salt uplifted and breached through the overlying lacustrine sediments. Interaction with shallow ground water resulted in rock-salt dissolution (Vroman, 1950). This dissolution was suggested to have occurred by a Ca-Cl brine, saturated (or attaining saturation in the process) with respect to anhydrite/gypsum, which removed the halite components while leaving behind the less soluble anhydrite. Indeed it was suggested to have occurred at times when the lake transgressed over the western margins of the basin reaching the diapir (Zak, 1967). The widespread dissolution of the rock-salt formed a near horizontal salt mirror (green line in Fig. 2a), and a caprock of ∼40 m thickness above, composed mostly of anhydrite and other insoluble sedimentary material (Zak, 1967, Fig. 2a). The roof of the caprock is an erosional surface on top of which are dotted remnants of parts of the Lisan Formation, implying that the diapir was submerged in the Lake Lisan water column.

Section snippets

Sampling

Details of the lithology of the DSDDP cores are reported in Neugebauer et al. (2014). Core 5017-1-A was drilled close to the middle of the modern lake (Fig. 1) at a water depth of ∼300 m (∼720 m below mean sea level = m bmsl). Core 5017-3-C was drilled off the shore of Ein Gedi Spa (Fig. 1) at water depth of ∼2.5 m. The chronology of the 5017-1-A is given by Torfstein et al. (2015) and Kitagawa et al. (2017). The chronology of core 5017-3-C was constructed by lithological correlation to core

Results

Chloride (Cl), magnesium (Mg2+) and calcium (Ca2+) concentrations from pore-fluids extracted from the DSDDP 5017-1-A core show abundant variability over the entire core depth (Fig. 3a–c). Bromide (Br) shows similar trends to magnesium (Levy et al., 2017) which is evident by the uniform bromide to magnesium (Br/Mg2+) ratio (pink circles in Fig. 3d). Due to this, both bromide and magnesium are regarded as conservative ions in the pore-fluids (Levy et al., 2017). Both ions are also considered

General

Given that the pore-fluid magnesium concentrations are representative of hypolimnion concentrations, the long term decrease in sediments corresponding to ca. 117ka to 12.5ka was suggested to be the result of increasing freshwater runoff into the lake along with mixing of these diluted waters with hypolimnion brine (Levy et al., 2017; Lazar et al., 2014). Chloride concentrations (Fig. 3a and i) also decrease within this depth interval, however the magnitude of dilution of chloride is much less

Summary and conclusions

This study presents an investigation of the long-term sulfate reservoir changes in the hypolimnion of Lake Lisan. We analyzed concentrations and stable isotopes in pore-fluids from cores that were drilled at the lake floor by ICDP (International Continental Drilling Programs). The main results and conclusions of the study are:

  • Significant amounts of chloride, sodium and sulfate ions were replenished in the lake due to dissolution of halite and anhydrite from the Mt. Sedom salt diapir.

  • The

Acknowledgements

We would like to thank both A. Starinsky and an anonymous reviewer for their fruifull and insighful reviews. We thank all who participated in the drilling operations, opening and descriptions of the drilled cores. We sincerely thank I. Swaed from GSI, E. Eliani-Russak, M. Adler, I. Bar-Or, N. Avrahamov, and A. Russak from BGU for taking part in sampling and measuring during the first pore-fluid extration campaign. D. Stiber, G. Sharabi, O. Berlin and other members of the GSI geochemical

References (54)

  • B. Herut et al.

    Coprecipitation of trace and minor elements in modern authigenic halites from the hypersaline Dead Sea brine

    Geochem. Cosmochim. Acta

    (1998)
  • A. Katz et al.

    The geochemical evolution of the Pleistocene Lake Lisan-Dead Sea system

    Geochem. Cosmochim. Acta

    (1977)
  • Y. Kiro et al.

    Relationships between lake-level changes and water and salt budgets in the Dead Sea during extreme aridities in the Eastern Mediterranean

    Earth Planet. Sci. Lett.

    (2017)
  • B. Lazar et al.

    Long-term freshening of the Dead Sea brine revealed by porewater Cl‾ and δ18O in ICDP Dead Sea deep-drill

    Earth Planet. Sci. Lett.

    (2014)
  • E.J. Levy et al.

    Salt precipitation and dissolution in the late Quaternary Dead Sea: evidence from chemical and δ 37 Cl composition of pore-fluids and halites

    Earth Planet. Sci. Lett.

    (2018)
  • G.W. Luther et al.

    Reduced sulfur in the hypersaline anoxic basins of the Mediterranean Sea

    Mar. Chem.

    (1990)
  • I. Neugebauer et al.

    Lithology of the long sediment record recovered by the ICDP Dead Sea deep drilling Project (DSDDP)

    Quat. Sci. Rev.

    (2014)
  • D. Palchan et al.

    Mobilization of fine detritus to the Dead Sea Basin during the late glacial and early Holocene

    Quat. Sci. Rev.

    (2019)
  • I.J. Reznik et al.

    Kinetics of gypsum nucleation and crystal growth from Dead Sea brine

    Geochem. Cosmochim. Acta

    (2009)
  • M. Stein et al.

    Strontium isotopic, chemical, and sedimentological evidence for the evolution of Lake Lisan and the Dead Sea

    Geochem. Cosmochim. Acta

    (1997)
  • A. Torfstein et al.

    The sources and evolution of sulfur in the hypersaline Lake Lisan (paleo-Dead Sea)

    Earth Planet. Sci. Lett.

    (2005)
  • A. Torfstein et al.

    Gypsum as a monitor of the paleo-limnological–hydrological conditions in lake Lisan and the Dead Sea

    Geochem. Cosmochim. Acta

    (2008)
  • A. Torfstein et al.

    Impacts of abrupt climate changes in the levant from last glacial Dead Sea levels

    Quat. Sci. Rev.

    (2013)
  • A. Torfstein et al.

    Dead Sea drawdown and monsoonal impacts in the Levant during the last interglacial

    Earth Planet. Sci. Lett.

    (2015)
  • A. Vengosh et al.

    Relics of evaporated sea water in deep basins of the eastern Mediterranean

    Mar. Geol.

    (1993)
  • N. Weber et al.

    The circulation of the Dead Sea brine in the regional aquifer

    Earth Planet. Sci. Lett.

    (2018)
  • R. Weinberger et al.

    Late-Pleistocene rise of the Sedom diapir on the backdrop of water-level fluctuations of Lake Lisan, Dead Sea basin

    Quat. Int.

    (2007)
  • Cited by (11)

    • Boron geochemistry reveals the evolution of Dead Sea brines

      2023, Earth and Planetary Science Letters
    • Hydrological and thermodynamic controls on late Holocene gypsum formation by mixing saline groundwater and Dead Sea brine

      2022, Geochimica et Cosmochimica Acta
      Citation Excerpt :

      This long-term process would have led to a significant depletion of the Dead Sea brine with respect to SO42−. However, the brine has maintained supersaturation with respect to gypsum over long periods (Torfstein et al., 2008; Reznik et al., 2009b; Levy et al., 2019). Consequently, any addition of SO42− that reaches the lake could potentially cause gypsum precipitation.

    • The paleohydrological implications of aragonite precipitation under contrasting climates in the endorheic Dead Sea and its precursors revealed by experimental investigations

      2021, Chemical Geology
      Citation Excerpt :

      Two experimental tracks were carried out in parallel; one using contemporary Dead Sea brine and another using a reconstructed Lake Lisan brine, prepared in the laboratory. The composition of Lake Lisan was reconstructed based on porewater composition from MIS2 sediments recovered from the deep Dead Sea cores (ICDP-DSDDP; Table 1; Levy et al., 2019; Levy et al., 2017; Levy et al., 2018). Because aragonite precipitates in the epilimnion of the lake (Belmaker et al., 2019; Neev and Emery, 1967; Stein et al., 1997), the concentrations of the reconstructed Lake Lisan brine were adjusted to 100 TDS [gr ∙ L−1] in accordance with salinity reconstructions for MIS2 (Begin et al., 2004).

    View all citing articles on Scopus
    View full text