δ34S values in gypsum are used to evaluate the fate of sulfur in the hypersaline Lake Lisan, the late Pleistocene precursor of the Dead Sea (70-14 ka BP), and applied as a paleo-limnological tracer. The Ca-chloride Lake Lisan evolved through meromictic periods characterized by precipitation of authigenic aragonite and holomictic episodes characterized by enhanced gypsum precipitation. The lake deposited two major gypsum units: the "Lower Gypsum unit" (deposited at ∼56 ka) showing δ34S values of 18-20‰, and the "Upper Gypsum unit" (deposited at 17 ka) displaying significantly higher δ34S values of 26-28‰. Laminated and disseminated gypsum, residing within the aragonite, exhibit δ34S values in the range of -26‰ to 1‰. The isotopic composition of the gypsum was dictated by freshwater sulfate input that replenished the upper layer of the lake (the mixolimnion), bacterial sulfate reduction (BSR) that occurred under the anoxic conditions of the lower brine (the monimolimnion), and mixing between these two layers. During meromictic periods, the sulfate reservoir in the lower brine was replenished by precipitation of gypsum from the upper layer, and its subsequent dissolution due to sulfate deficiency induced by BSR activity. This process describes a "sulfur pump" mechanism and its effect on δ34S in the water can be modeled by a modified Rayleigh distillation equation. Steady state δ34S values (∼40‰) were reached in the lower brine after long meromictic periods. Following overturn episodes, induced by diminishing freshwater input and lake level decline, large quantities of δ34S enriched gypsum precipitated. The negative δ34S values in laminated and disseminated gypsum provide evidence for BSR activity in the lower brine that removed isotopically depleted sulfides from the water column, causing significant isotopic enrichment of remaining sulfate. Following the lake desiccation, the sediments were exposed and the latter sulfides oxidized and re-crystallized as gypsum.
Bibliographical noteFunding Information:
We thank Yeshu Kolodny who participated in the first stage of this project before being kidnapped to the dean's office, Amitai Katz for illuminating discussions and Ruth Binstock for laboratory work. We wish to thank T. Lyons and two other anonymous reviewers for constructive and helpful reviews. The study was supported by BSF Grant # 2000271 to I.G. and M.S.
- Bacterial sulfate reduction
- Lake Lisan
- Rayleigh distillation
- Sulfur isotopes