PII S0016-7037(00)00540-8 Dynamics of the carbon dioxide system in the Dead Sea EUGENI BARKAN,BOAZ LUZ, and BOAZ LAZAR* The Institute of Earth Sciences, The Hebrew University, Jerusalem 91904, Israel (Received January 26, 2000; accepted in revised form August 24, 2000) Abstract—–Because the Dead Sea is a Ca-chloride hypersaline lake and is virtually lifeless, it is an excellent system to investigate the physical and chemical paths of the carbon cycle in terrestrial aquatic environments that are generally overwhelmed by biologic carbon fluxes. The Dead Sea is known to form massive aragonite deposits in the form of varves at present and during the Lisan period (late Pleistocene). The commonly accepted season for the main aragonite precipitation was summer, and the trigger for precipitation was attributed to evaporation and the warming of surface water (“whitening” events). To determine the main carbon fluxes in the Dead Sea, we followed changes in the carbonate system of the mixed layer, from February 1993 to December 1994, after stratification formed due to heavy flooding during the winter of 1992. The stratification isolated the mixed layer, a relatively small reservoir from the main brine body, causing an amplification of the chemical signals. The data show that partial pressure of CO 2 in the lake was very high (2000 atm). Total alkalinity and total carbon of the mixed layer decreased with time, whereas  13 C increased. The high P CO 2 originates from precipitation of aragonite and implies that in many aquatic systems it may originate from an inorganic process and not only from degradation of organic matter. Thermodynamic calculation estimated the degree of aragonite saturation to be 15 when 10% of Dead Sea brine with a high Ca content mixed with 90% runoff freshwater with high-bicarbonate content. Therefore, mixing during winter flooding triggers massive aragonite deposition in the Dead Sea. The general conclusion is that inorganic carbonate precipitation by mixing of two solutions, one supplying Ca 2+ and the other HCO 3 - should be considered in the evaluation of the carbonate system in a wide range of aquatic environments. A mass-balance model for total alkalinity, total carbon, and carbon isotopes reveals two main carbon sinks and one carbon source. The sinks are a chemical deposition of aragonite (1.4 mol  m -2  y -1 ), and CO 2 escape to the atmosphere (4 mol  m -2  y -1 ), and the source is bicarbonate input by floods (2.1 mol  m -2  y -1 ). The precipitation rate of the present Dead Sea is approximately sixfold lower than the average precipitation rate during the Lisan period, implying a wetter climate during that period. The present CO 2 escape rate from the mixed layer is twice the bicarbonate input and threefold the aragonite precipitation rate, indicating a net CO 2 loss. We suggest that such a scenario is possible if the Dead Sea was meromictic (stratified) for a very long period of time. Copyright © 2001 Elsevier Science Ltd 1. INTRODUCTION The carbon cycle of most of the lakes worldwide is con- trolled mainly by biologic degradation processes which have high partial pressure of CO 2 (Cole et al., 1994). The Dead Sea, being a Ca-chloride hypersaline lake (Neev and Emery, 1967) with total dissolved salts  278 g  kg -1 , is nearly barren (Oren, 1997). Therefore, it is an ideal “natural laboratory” to investigate the impact of physical and chemical processes on the carbon cycle of continental water masses. The Dead Sea is a terminal lake situated in the arid Dead Sea/Red Sea rift valley. Its brine precipitates aragonite inorganically (e.g., Neev and Emery, 1967) despite its low pH value of 6 (Amit and Bentor, 1971). Several tens of thousands of aragonite varves (compris- ing the Lisan Formation) were deposited on the bottom of its historic predecessor, Lake Lisan during the Late Pleistocene. This finding suggests that chemical precipitation of CaCO 3 has been an important path of the carbon cycle in the area for a long time (e.g., Niemi, 1997). The first study of the carbonate system in the Dead Sea was conducted in the pioneering research of Neev and Emery (1967). They showed that two summer min- eral precipitation events (brine “whitening”), resulted in a gyp- sum to aragonite weight ratio  3.14. It clearly indicates that aragonite varves comprised of almost 100% CaCO 3 did not originate from “whitening” events and that an alternative ex- planation is needed. Paleolimnologic studies conducted to date (e.g., Stein et al., 1997) attributed the source of CO 3 2- in the aragonite to HCO 3 - input with freshwater runoff, but the trigger and timing of this precipitation remains unclear. In the present study, we measured the carbonate system parameters in the mixed layer of the modern Dead Sea, and by using carbon balances and thermodynamic calculations we show that most of the aragonite precipitates in response to winter flooding. The first systematic study of the isotopic composition of the carbonate system was conducted by Nissenbaum and Kaplan (1976). They reported significant depletion of 13 C of the dis- solved inorganic carbon and attributed this depletion to an incomplete isotopic equilibration of inorganic carbon intro- duced by the influx of water from the Jordan River and winter floods. However, their reported  13 C values are much smaller than those of Neev and Emery (1967) and of Luz et al. (1997). Additional studies of the carbonate system were conducted by Stiller in 1984 to 1985 and are reported in Luz et al. (1997). *Author to whom correspondence should be addressed (boazl @vms.huji.ac.il). Pergamon Geochimica et Cosmochimica Acta, Vol. 65, No. 3, pp. 355–368, 2001 Copyright © 2001 Elsevier Science Ltd Printed in the USA. All rights reserved 0016-7037/01 $20.00 + .00 355