Geochemical Evolution of Great Salt Lake, Utah, USA
- First Online:
- Cite this article as:
- Jones, B.F., Naftz, D.L., Spencer, R.J. et al. Aquat Geochem (2009) 15: 95. doi:10.1007/s10498-008-9047-y
- 192 Views
The Great Salt Lake (GSL) of Utah, USA, is the largest saline lake in North America, and its brines are some of the most concentrated anywhere in the world. The lake occupies a closed basin system whose chemistry reflects solute inputs from the weathering of a diverse suite of rocks in its drainage basin. GSL is the remnant of a much larger lacustrine body, Lake Bonneville, and it has a long history of carbonate deposition. Inflow to the lake is from three major rivers that drain mountain ranges to the east and empty into the southern arm of the lake, from precipitation directly on the lake, and from minor groundwater inflow. Outflow is by evaporation. The greatest solute inputs are from calcium bicarbonate river waters mixed with sodium chloride-type springs and groundwaters. Prior to 1930 the lake concentration inversely tracked lake volume, which reflected climatic variation in the drainage, but since then salt precipitation and re-solution, primarily halite and mirabilite, have periodically modified lake-brine chemistry through density stratification and compositional differentiation. In addition, construction of a railway causeway has restricted circulation, nearly isolating the northern from the southern part of the lake, leading to halite precipitation in the north. These and other conditions have created brine differentiation, mixing, and fractional precipitation of salts as major factors in solute evolution. Pore fluids and diagenetic reactions have been identified as important sources and especially sinks for CaCO3, Mg, and K in the lake, depending on the concentration gradient and clays.
KeywordsLakesSalineEvaporationMixingCalcium bicarbonateSodium chloridePrecipitationRe-solutionHaliteMirabiliteCalciteDolomiteMg-silicateDiffusionPore-fluidsDiagenesisClimate
Dimensional properties of Great Salt Lake
4196 ft; 1279 m
1200 mi2; 31,000 ha
1.5 × 107 ac-ft; 1.9 × 1010 m3
River inflow (1931–1976)
Water temperature range
Air temperature range
2.9 × 106 ac-ft; 3.6 × 1010 m3
1.9 × 106 ac-ft; 2.3 × 109 m3
0.9 × 106 ac-ft; 1.1 × 109 m3
<O to +35°C
−6 to +33°C
28% North arm; 12% south arm
~28% North arm; ~27% south arm
~28% North arm; ~6% south arm
In this paper we present a review of the hydrologic history of the lake basin in terms of lake level changes that occurred in response to changes in the balance between inflow and outflow. The hydrologic history is based on a review of previous work, primarily on ancient shorelines. We then present information on the chemistry of the various inflow components to the system. This is followed by a review of the concepts employed in interpreting the geochemical evolution of brines. The geochemical evolution of the lake water chemistry within the basin is discussed in the context of the changes in the hydrologic history and changes in relative inputs of inflow components with different chemical compositions in response to the hydrologic changes employing the principles of brine evolution. The discussion of the geochemical evolution of the lake above provides a context for the discussion of the modern lake that follows.
2 Hydrologic History of the Basin
The lake basin now occupied by GSL began forming in response to tectonic extension in the eastern Basin and Range Province in the middle Tertiary (Miller 1991). Lake sediments of Miocene age have been identified in the basin, both in deep cores from the basin center and in outcrops of faulted offshore sediments on the basin margins (Miller 1991). Most previous work on the lacustrine history of the basin has focused on the late Pleistocene (Gilbert 1890; Scott et al. 1983; Spencer et al. 1984; Oviatt et al. 1992), but older Pleistocene lakes have also been investigated (Kowalewska and Cohen 1998; Balch et al. 2005).
A summary of the late-Pleistocene history of Lake Bonneville is of particular interest here as the precursor to hypersaline GSL. Prior to about 28,000 14C year B.P. the lake level was apparently similar to Holocene GSL, although no shorelines of this age have been identified and dated. Pre-Bonneville sediments in cores from the floor of GSL are composed of laminated mud and fine sand, and both saline-tolerant ostracodes and brine-shrimp fecal pellets are present in different samples, but the sediments are not dominated by brine-shrimp fecal-pellet mud as are Holocene GSL sediments (Spencer et al. 1984; Oviatt and Thompson 2008, unpublished data).
Lake Bonneville briefly overflowed into the Snake River drainage at its highest level and formed the Bonneville shoreline. Lake Bonneville catastrophically dropped about 100 m in altitude as the overflow threshold washed out in an event known as the Bonneville flood (Gilbert 1890; Oviatt et al. 1992; O’Connor 1993). The lake continued to discharge across a bedrock-floored threshold into the Snake River drainage for approximately 2500 years, during which time the Provo shoreline formed throughout the basin (Gilbert 1890; Godsey et al. 2005). The rapid closed-basin regressive phase began about 12,000 14C year B.P. and dropped the lake to altitudes comparable to modern GSL by about 11,000 14C year B.P. (Oviatt et al. 2005). A brief rise to the Gilbert shoreline at roughly 10,000 14C year B.P. was followed by Holocene fluctuations of GSL within an altitude range of 6 m (1280–1286 masl), and a volume range of about 40–300 km3 (Fig. 1).
During the transgressive phase of Lake Bonneville, lake level and lake volume progressively increased, and the water became more dilute. In sediment cores from GSL that cover the transgressive phase, laminations containing large percentages of endogenic aragonite are interbedded with the dominantly calcite mud (Spencer et al. 1984; Oviatt and Thompson 2008 unpublished data). Aragonite laminations may indicate times when lake volume declined and the Mg to Ca ratio increased, possibly in response to an influx of water from the Sevier arm of the lake, where the river inflow has a relatively high Mg/Ca ratio (Oviatt et al. 1994; Pedone 2004).
Aragonite replaces calcite as the dominant mineral in post-Provo, regressive-phase sediments. This was caused by lake-volume decrease and progressive removal of Ca from the water as carbonate minerals precipitated, leading to an increase in the Mg/Ca ratio, and a shift from calcite to aragonite precipitation (Spencer et al. 1984). Aragonite dominates the carbonate mineral assemblage throughout the Holocene sedimentary column of GSL (Spencer et al. 1984; Oviatt and Thompson 2008 unpublished data). The end of the regressive phase of Lake Bonneville, prior to the development of the Gilbert shoreline (Fig. 1), marks the beginning of the GSL in something close to its modern configuration.
3 Geochemistry of Inflowing Waters
Representative recent analyses of the three major inflows to Great Salt Lake, the Bear, Weber, and Jordan Rivers
Chemical analyses of major inflows
Chemical analyses of Great Salt Lake (Gilbert Bay)
Normative salt assemblage
SNORM calculation results for Great Salt Lake (site N1022 at 1 m depth)
KMgCl3 · 6H2O
MgSO4 · H2O
MgCl2 · 6H2O
Li2SO4 · H2O
Mg2B6O11 · 15H2O
SNORM calculation results for seawater
MgCl2 · 6H2O
KMgCl3 · 6H2O
MgSO4 · H2O
Mg2B6O11 · 15H2O
Li2SO4 · H2O
Mg(NO3)2 · 6H2O
4 Geochemical Evolution of Brines—Concepts
The early precipitation of alkaline-earth carbonates is a crucial step in understanding brine evolution. There are several alkaline-earth carbonates present in many closed basin sediments. Among these are calcite, Mg-bearing calcite (less than about 12 mole percent Mg), and aragonite. All of these phases have relatively low solubility. Carbonates with higher magnesium contents such as dolomite and magnesite also have low solubility, but are not common as primary precipitates in closed basin sediments, and are rare or absent as primary precipitates during the early stages of evaporative concentration. Calcium is preferentially removed from solution relative to magnesium during the precipitation of calcite, Mg-bearing calcite and aragonite, resulting in an increase in the ratio of Mg to Ca in solution during precipitation of these minerals. Experimental work by Fuchtbauer and Hardie (1976) shows a precipitation sequence from calcite to Mg-bearing calcite to aragonite with increasing Mg to Ca in solution. This is expected to occur in closed basin systems as inflow waters with relatively low Mg to Ca in solution are concentrated and alkaline earth carbonates precipitate.
The first mineral to precipitate from a concentrating water is determined by its position with respect to the primary precipitation field for minerals on the ternary phase diagram in Fig. 5. The body of this phase diagram is the primary precipitation field for calcite. Therefore, the first mineral to precipitate from waters that plot within the body of the triangle is calcite (or possibly Mg-bearing calcite or aragonite). During mineral precipitation the composition of the water moves directly away from the composition of the mineral. For instance, waters at points 1 and 2 on Fig. 5 will precipitate calcite and move directly away from the calcite compositional point. Continued evaporative concentration and calcite precipitation results in migration of the water composition to the join between HCO3 plus CO3 and SO4 as indicated by the arrows in Fig. 5. These waters become depleted in Ca during the precipitation of calcite; a portion of the HCO3 and CO3 and SO4 remain in solution after calcite precipitation. These waters eventually may precipitate a number of more soluble salts, including sodium carbonate and/or sodium sulfate salts, but not calcium sulfate (gypsum). Waters near composition 1 precipitate sodium carbonates prior to sodium sulfate and waters near composition 2 precipitate sodium sulfate prior to sodium carbonate. All waters may precipitate halite (and other salts) at some point (see Spencer and Hardie 1990, for more details).
There are three distinct compositional fields separated by two chemical divides as a result of the precipitation of relatively insoluble minerals in the system displayed in Fig. 5. The join between the calcite compositional point and the SO4 apex represents the divide between “Alkaline” and “Neutral” waters, and the join from calcite to gypsum represents the divide between “Neutral” and “Calcium Chloride” waters. Neutral waters, such as at composition 3 on Fig. 5, precipitate calcite and move directly away from the calcite composition. Continued evaporative concentration and calcite precipitation brings these waters to the gypsum stability field located along the join between Ca and SO4 in Fig. 5. Waters are depleted in HCO3 and CO3 along this join and gypsum is the dominant precipitate. Waters move along the phase boundary away from the gypsum compositional point, toward the SO4 corner of the diagram. Waters are depleted in Ca and further concentration leads to the precipitation of halite and a variety of other salts including K and Mg sulfates and chlorides.
Calcium chloride waters such as at composition 4 on Fig. 5, precipitate calcite and move directly away from the calcite composition. Continued evaporative concentration and calcite precipitation brings these waters to the gypsum stability field located along the join between Ca and SO4 in Fig. 5. Waters are depleted in HCO3 and CO3 along this join and gypsum is the dominant precipitate. Waters move along the phase boundary away from the gypsum compositional point, toward the Ca corner of the diagram. Waters are depleted in SO4 and further concentration leads to the precipitation of halite as well as K, Mg, and Ca chlorides.
5 Geochemical Evolution of GSL
Salt budget over 35,000 years (as calculated by Spencer et al. 1985b)
Percentage of input accounted for
Sediment from mid-lake cores that cover the transgressive phase (Fig. 1) varies from laminated to mottled to massive in appearance. The mineralogy of the sediment is dominated by detrital quartz (~50%) and detrital clay minerals (~30%) and lesser amounts of endogenic low Mg calcite (<20%) (Spencer et al. 1984). The sediments contain fresh to brackish water ostracodes and well-preserved algal filaments. Overall the sediment is consistent with a rising lake receiving dilute, high discharge waters with a composition dominated by sodium bicarbonate-type waters as a result of the weathering of silicates (i.e., Bear and Weber Rivers), carrying significant detrital sediments load. The lake does not appear to have increased systematically in concentration during the transgressive stage, but rather to have become more dilute. However, there are short intervals of light colored sediment that contain small amounts of aragonite (10–15%) along with calcite (5–15%) and variable amounts of quartz (~10–50%) that punctuate this interval (Spencer et al. 1984). The aragonitic intervals are bracketed by ages of 21,050 14C year B.P. and 19,540 14C year B.P. (Thompson et al. 1990) and may represent fluctuations near the Stansbury shoreline (see Fig. 1). Oviatt and Thompson (2008 unpublished data) also report sediment within the transgressive interval containing large percentages of endogenic aragonite interbedded with the dominantly calcite mud. Aragonite laminations may indicate times when lake volume declined and the Mg to Ca ratio increased, although there is no indication of a progressive shift in the lake chemistry as seen in the major drawdown discussed below. These aragonitic intervals may be in response to an influx of water from the Sevier arm of the lake, where the river inflow has a relatively high Mg/Ca ratio (Oviatt et al. 1994; Pedone 2004).
During the development of the Bonneville shoreline, the Bonneville flood, and the Provo shoreline (Fig. 1) the lake was overflowing into the Snake River drainage system. Therefore, from the perspective of the deep lake sediment and geochemical evolution of waters, this interval is grouped with the transgressive phase and was characterized by a relatively fresh lake receiving high discharge waters.
The regressive phase of the lake is clearly recorded in the mid-lake cores. The most dramatic changes in the mid-lake cores occur over a 25-cm interval of sediment. At the base of this interval the sediment mineralogy consists of detrital quartz (~55%), detrital clay minerals (~30%) and low Mg calcite (~15%), typical of the transgressive sediments described above. These sediments also contain ostracodes indicative of the freshest water conditions in the entire core record. Over an 8 cm interval the quartz content of the sediment drops systematically from 55% to 20%, while calcite content increases continuously from 15% to 55% and the Mg content of the calcite increases steadily form near zero to 11 mole percent. Immediately above this is a 2 cm interval containing about 50% aragonite, 5% calcite, 15% quartz, and 30% clay minerals. Above this is an interval of what appears to be reworked sediment with mud coated and abraded ostracodes. Brine shrimp cysts first appear in muds just below the coated ostracodes and initially even with some uncoated ones, but they are much more abundant above the carbonate coated ostracodes. This indicates that the lake had become saline and was reaching relatively low levels. The interval covering the regressive phase described above is bracketed by ages of 12,400 14C year B.P. and 11,970 14C year B.P. (Thompson et al. 1990).1
The regressive interval described above covers the precipitation of alkaline-earth carbonates during the rapid drawdown of the lake as a result of evaporative concentration. The sequence from calcite, through Mg-bearing calcite with increasing Mg-content to aragonite is consistent with the experimental work by Fuchtbauer and Hardie (1976) in response to an increase in the Mg to Ca ratio in solution as calcium is preferentially removed from waters to form the carbonates. The beginning of the evaporative phase chemistry is signaled by the increase in the amount of calcite and the Mg-content of the calcite. The shift from Mg-calcite to aragonite is simply a point in the continuous sequence. The lake continued to become more concentrated beyond this point as indicated by the presence of brine shrimp.
The chemistry of the lake has changed through time. Modern GSL brines are within the “Neutral” field as indicated on Fig. 7b. Lake levels appear to have fluctuated near present day levels during the past 9000 years. Sediments are dominated by brine shrimp pellets composed of aragonitic mud. Disrupted, chemoturbated sediments indicate the growth and dissolution of mirabilite throughout the interval. However, gypsum is present as discoidal crystals, likely of diagenetic origin, within the upper 1.5 m of sediment, above the Mazama tephra with an age of 6730 14C year B.P. (7627 cal year B.P.) (Zdanowicz et al. 1999), and as a primary precipitate in the upper 10 cm of sediment. The presence of gypsum indicates a “Neutral” composition of the brines. Thus, there is a difference in mineral precipitation sequence which can be explained in terms of the composition of low versus high inflow discharge. The major inputs to the lake under low discharge conditions are the “Alkaline” waters of the Bear, Weber, and Jordan rivers, plus a contribution from springs, such as the Malad Springs, along the Wasatch front. The spring waters are of the “Calcium Chloride” type, see Fig. 7b. There is a much higher relative contribution of the “Calcium Chloride” solutes during low-discharge than at high-discharge conditions. The continued input of the spring water component during the 10,000–11,000 year low stand has resulted in the migration of the lake water chemistry into the neutral field, to a point where it follows the evaporation path indicated by the dotted line in Fig. 7b.
6 The Modern Lake System
Salt (NaCl) precipitation occurred during the drought of the 1930s (1935–1945) and again in the 1960s. In the South Arm the salt was apparently re-dissolved by 1972, but stayed saturated in the north arm until the large lake-level rise in the 1980s. From 1965 until 2002, there was a slight decline in SO4, Mg, K, and Ca, whereas Na and Cl increased. For the period 1966–1981 there was an increase in the relation of total dissolved solids (TDS) versus stage in the south arm and constancy in the north arm, until the heavy rains and South Arm inflow increased dramatically. Salt precipitation began again in the summer of 1992 and continued to the present (Gwynn 2002). Prominent minor elements (Br, B, Li) increased relative to NaCl. Exchange across the railroad causeway has kept the concentration trends similar. Bidirectional flow related to a brine interface existed across the causeway from pre-1966 until mid 1991 because of the lack of return flow thru the causeway and vertical mixing. Since 1992 the north arm has been at halite saturation, so there was no inverse relation to lake level. Since 1993 there has been only south to north flow or none at all. The lake-level rise since 1995 has increased the south to north surface flow, but compaction of fine-grained sediments in the causeway fill has shut off the underflow return. Movement southward of south arm bottom brine over the mid-lake ridge from the north to the south was noted by Hahl and Handy (1969).
Farmington Bay has apparently often been isolated from main GSL, but under average conditions, has a salinity about half of the main lake. The proportion of major ions is about the same as in the lake. The bay acts as a biological treatment lagoon for nutrients, N and P, from the Jordan River and sewage-treatment plants. Six mineral-extraction industries operate on the lake. These produce potash fertilizer, Mg metal, and Cl gas, as well as a variety of salts.
7 Geochemical Dynamics
As in all natural waters, the composition of any saline lake is dominated by less than ten major solutes, Na, K, Ca, Mg, Cl, SO4, dissolved inorganic carbon as HCO3 + CO3 and sometimes SiO2. Of these, Na is the most common cation, and often is nearly the only cation in solution. The short-term processes noted above mainly affect Na, SO4, and Cl, whereas Mg and K remain mostly solute conservative (lost only to long-term pore-fluid diffusion, diagenetic mineral formation in sediment, aerosols, or the mineral extraction industries). In the Great Basin, overall compositional trends can be described almost entirely in terms of the anions in saline waters (Jones 1966). However, Bodine and Jones (1986) found it more useful to classify saline waters according to normative analyses of their principal anion-cation associations based on relative mineral solubility. Further, in the latter stages of chemical evolution the ratios with respect to the alkalies and sulfo-chloride become more important, and it is advantageous to consider associations within the entire solute matrix. For this purpose, the computer program SNORM calculates the equilibrium salt assemblage expected for any water taken to dryness at 25°C and atmospheric pCO2. Thus, regardless of the initial composition, saline waters can be compared in terms of the total salt assemblage to be expected on complete evaporation, including the presence or absence of characteristic double salts. Using recalculation of the assemblage to “simple salts,” multiphase mineral assemblages can be reduced to simple characteristic cation-anion associations.
A significant example of the importance of redox processes to minor elements at GSL is the case of mercury. The high concentrations of Cl and Br in GSL may enhance the atmospheric deposition of Hg to the lake surface. Previous studies (Mason and Gill 2005) have shown that the presence of halogens such as Br and Cl in the marine boundary layer (MBL) can act as oxidizing agents that will transform the relatively inert Hg0 in the atmosphere to reactive gaseous Hg (RGHg). Once formed, RGHg has high deposition velocities and has been observed to be rapidly removed from the atmosphere in polar locations relative to Hg0 (Schroeder et al. 1998). The abundance of Br and Cl in the MBL above GSL may favor the formation of RGHg and enhance Hg deposition rates. Once inorganic forms of Hg enter GSL, the physical and chemical conditions may be ideal for Hg methylation. Previous work has shown that marine sediment rich in organic matter and dissolved sulfide have rapid CH3Hg production rates in conjunction with rapid rates of SO4 reduction (King et al. 2000). Low dissolved O2 saturation in hypersaline systems such as GSL can support high rates of SO4 reduction. The SO4 reduction rates measured in GSL were higher than 6000 nmol/cm3/day, one of the highest rates reported in a natural environment (Ingvorsen and Brandt 2002). In laboratory experiments, King et al. (2000) determined that SO4 reducing bacteria capable of acetate utilization in their metabolic pathways are the most efficient at methylating Hg. Acetate-utilizing bacteria (Desulfobacter halotolerans and Desulfocella halophila) capable of high methyl-Hg production have been isolated from sediment in the south arm of GSL (Ingvorsen and Brandt 2002).
Whole water samples were collected from GSL in 2003, 2005, and 2007 and analyzed for total Hg and CH3Hg. The highest concentrations of total Hg were found in water samples collected from the DBL, ranging from 7 to >100 ng/l (Naftz et al. 2008), with a median concentration of 42 ng/l. A significant proportion (31–60%) of the total Hg in water samples from the DBL was composed of CH3Hg. Concentrations of CH3Hg in the DBL ranged from 0.84 to >30 ng/l. The concentration of CH3Hg measured in GSL is among the highest measured in surface water by the USGS Mercury Research Laboratory. For comparison, CH3Hg in wholewater samples collected from Maryland reservoirs ranged from 0.007 to 0.493 ng/l (Mason and Sveinsdottir 2003).
Besides mixing and associated smaller scale processes (degassing, temperature change), evaporative concentration is the dominant mechanism effecting mineral formation and the solute evolution accompanying sequential mineral saturation and precipitation. The sequence is readily illustrated with a flow chart such as modified from Eugster and Hardie (1978) (Fig. 4) and based on relative solubility. Thus the first minerals to precipitate from GSL waters increasing in salinity are the alkaline-earth carbonates. The equilibrium precipitation of calcite helps illustrate a major control on the solute evolution of saline lake waters: the “chemical divide” (first defined by Hardie and Eugster 1970).
As discussed in the section on “concepts,” for the precipitation of pure mineral, cation and anion both contribute to the precipitate and are lost from the solution in equal proportion. At the same time, the solubility product of cation and anion must be constant, so as the mineral precipitates, unequal amounts remain in solution and the constituent of greater proportion will become dominant in solution on further concentration, i.e., increasing Ca will decrease the level of CO3 or vice versa. Thus, early precipitation of calcite determines whether the remaining solution will become carbonate rich or poor.
As noted earlier, an effective technique for applying the “chemical divide” concept is the prediction of solute evolution through use of the “Spencer Triangle” (Fig. 5). To summarize the explanation given in the “concepts” section, in a triangular diagram of the system Ca-SO4(CO3 + HCO3) a line connecting a point representing a particular solute distribution and a particular precipitate composition will define the direction of the chemical evolution of the water. The join between CaCO3 and the SO4 apex represents the calcite divide, and the join from CaCO3 to CaSO4 represents the gypsum divide. Solutions will move away from these divides as the respective minerals precipitate. NaCl is considered ubiquitous. Each triangular field is named according to the principle solutes left in the final fluid.
In historic time the lake seems to have passed through the CaCO3 divide repeatedly. As a result, carbonate precipitation processes have left the residual brines strongly enriched in Mg, depleted in Ca, and impoverished in bicarbonate. Mirabilite still occurs intermittently through brine chilling. According to Spencer et al. (1985b) the sequence expected from modern brines at 25°C is aragonite, small amounts of gypsum, a little glauberite, and then a large mass of halite. The variations in mineral precipitate sequences, the fossil mirabilite found in the sediment, and the halite dominance in salts precipitated at present are a testimony to the changing influence of input sources and sinks through time (Spencer et al. 1985b).
8 Summary and Conclusions
The GSL is the largest saline lake of North America, and its brines are some of the most concentrated anywhere in the world. The lake occupies a closed basin system whose chemistry reflects solute inputs from the weathering of rocks ranging in age from Precambrian and Paleozoic in the high mountains on the east to Tertiary and Quaternary sediments in the desert basins to the west. Lithologies cover a complete range including crystalline rocks (both intrusive and metamorphic), clastic sediments of all types, large sections of carbonates, and small sections of both marine and lacustrine evaporites. GSL is the remnant of a much larger lacustrine body, Lake Bonneville, and its geochemical evolution can be considered to have begun with the drop in the level of Bonneville at the end of the Pleistocene. Evidence of the effect of evaporative concentration (and CaCO3 precipitation) on high stand waters can be readily seen in the tufa deposits on older shorelines.
The water balance of GSL depends primarily on the inflow from three major rivers draining the ranges to the east and precipitation directly on the lake. The greatest solute inputs are from calcium bicarbonate river waters mixed with sodium chloride type springs and groundwaters. Prior to 1930, the lake concentration inversely tracked lake volume, which reflected climatic variation in the drainage. However, since that time, salt precipitation, primarily halite and mirabilite, and re-solution have periodically modified lake brine chemistry through density stratification and formation of brine pockets and compositional differentiation. In addition, construction of a railway causeway has restricted circulation, nearly isolating the northern from the southern part of the lake, which receives 90% of the inflow, leading to halite precipitation in the north. Widespread halite precipitation occurred prior to railroad-causeway construction, especially in the south, as a result of severe droughts. The presence of a mid-lake subaqueous ridge, probably had the same effect. These conditions have emphasized brine differentiation, mixing, and fractional precipitation of salts as major factors in solute evolution, especially in relatively shallow systems.
The work at GSL has also highlighted the role of pore fluids as sources and sinks of solutes in the lake, depending on the concentration gradient. Also, diagenetic reactions in the sediments cause gypsum to precipitate or dissolve irregularly because calcium levels are so low after extensive carbonate deposition. Alkalinity and low calcium concentrations are usually near constancy at aragonite rather than calcite saturation, because of increasing Mg/Ca ratio. In addition to carbonate, significant amounts of dissolved Mg and K are lost to clays through diagenetic pore-fluid reactions. Accounting for pore-fluid diffusion has permitted a reasonable balance to be made between solute input over time and the quantities now found in solution and sediments. Excess amounts of calcium, carbonate, and silica can be attributed to detrital input.
Comparison of analyzed and computer determinations of a halite-saturated pond sequence from GSL indicated similar points of crystallization for minor salt phases. Compositional plots of major ions versus bromide suggested bitter salt (kainite) uptake of Br began at 700 ppm (as compared to 70 ppm for halite), and computer calculations produced solubility values for phases subsidiary to halite in the GLS evaporative pond sequence.
In some cores the aragonite peak is replaced by a “dolomite” peak (primary XRD peak near 31 degrees 2-theta), and although aragonite does increase in this interval, “dolomite” is the dominant carbonate mineral. Could the “dolomite” be a secondary replacement of aragonite?