Salty Matters

The Blog is written by me, John Warren. Once every three or four weeks or so I will post an article or two on an evaporite topic that has piqued my interest. On the Saltwork Publications webpage (under "the Works") there is a growing library of pdfs and epubs based on these blogs. These articles on the website have much higher resolution extractable graphics in than in the blog. There is also a link to this set of pdfs and epubs on the home page (

Life in modern Deepsea Hypersaline Lakes and Basins - DHALs and DHABs

John Warren - Sunday, September 30, 2018



Exuded salt karst brine on the deep ocean floor has a much higher density that the overlying seawater and so if there is an ongoing supply it tends to pond in seafloor lows (Figure 1a). The longterm character (hydrological stability over hundreds to thousands of years) of such density-stratified brine lakes, which form the centrepieces in deepsea hypersaline anoxic basins (DHAB), facilitate longterm ecologic niche sthe tability. The upper surface of a brine lake is marked by a halocline, which typically defines one or more nutrient, thermal and salinity interfaces (Figure 1b). There a light-independent chemosynthetic seep and lake biota can grow and flourish (Figure 1a). Escaping subsurface brines can entrain both hydrocarbons (mostly methane) and H2S, which are nutrients in the base of the chemosynthetic food chain. The salinity layering created by the halocline can be positioned as ; 1) a pelagic biotal interface, or 2) a brine lake edge (or shore) interface or 3) out in the lake the brine column base (i.e. a hypersaline-sediment interface) (Figure 1b).

In other places on a deep seafloor, the escaping salt-karst brines, with entrained methane and H2S, can form diffuse outflow or seep areas, without ever developing into a free-standing brine lake (position 4 in Figure 1a). Highly specialised chemosynthetic communities tend to dolonise the resulting density and salinity-stratified interfaces. And so, some chemosynthetic communities occupy a halocline interface in a pelagic position atop an open brine lake, while others inhabit a benthic position where the halocline intersects the deep seafloor (Figure 1). Anoxic hypersaline brine can also pond on the shallow seafloor in high latitude regions where the formation of sea ice create cryogenic brines (Kvitek et al, 1998). But this style of cryogenic seaflooor brine lake is more ephemeral and is not tied to major evaporite deposits, so is not considered further.

Two groups of megafauna with symbiotic methanotrophic or thiotrophic bacteria dominate chemoosynthetic communities in the salt-floored Gulf of Mexico: 1) bivalves, including bathymodiolin mussels and multiple families of clams and 2) vestimentiferan tubeworms in the polychaete family Siboglinidae. Both the vestimentiferan siboglinids and clams harbour microbial endosymbionts that utilise sulphide as an energy source, whereas different species of bathymodiolin mussels harbour either methanotrophic, thiotrophic, or both, types of symbionts (Figure 2).

Along the brine pool edge in the Gulf of Mexico

Hence, the mussel-tubeworm dominated brine-lake edge and seep biostromes in the Gulf of Mexico are dependent on chemosynthesising microbes as a food source. This community is the cold-water counterpart to warm-water chemosynthetic hydrothermal communities flourishing in high temperature waters the vicinity of black smoker vents (MacDonald, 1992; MacDonald et al., 2003). In both settings, it is methane and sulphide, not light, that provides the than DHALs energy source for the bacteria and archaea that make up the base of the chemosynthetic food chain.

Methanotrophic bacteria live symbiotically on a seep mussel’s gills, taking in methane and converting it to nutrients that nourish the mussels. The seep mussels (Bathymodiolus childressi and Calyptogena ponderosa continually waft methane-rich water through their gills to help their chemo-autotrophic bacterial symbionts grow and periodically harvest some of the excess growth. Their lifestyle means that seep mussels need to live near a supply of dissolved gas, so they can inhabit isolated seep outflows on the deep seafloor where gas is bubbling out, including the edges of mud volcano pools, but do best about the more stable and relatively quiescent edges of methane-saturated brine pools and lakes.

There they grow as a fringe to the brine pool, and exist about the pool rim, wherever they can keep their syphons above the halocline Figure 2a-d). They tend to construct a biogenic edge (biostrome) to the brine pool atop with sediment piles generally cemented by methanogenic calcite. Such rims typically extend some 5-10 metres behind the pool edge (Figure 2a; Smith et al., 2000). The inner edge of the mussel biostrome is elevated only a few centimetres from the surface of the pool and is distinguishable by an abundance of smaller individuals, present in high densities (Figure 2b). At the outer edge of the mussel biostrome, there is a high frequency of disarticulated shells and low densities of still living larger individuals.

Also living atop seafloor seeps and about some brine pools are knots and clusters of chemosynthetic polychaete tubeworms (Figures 2c, 3; Lamellibrachia luymesi and Seepiophila jonesi). Individual tubeworms (aka seep beard-worms) in a colony can be up to 2.5 m long with a microbe-dependent metabolism evolved to exploit the abundant H2S and methane seeping through the seafloor. Tubeworm colonies grow as rims and clumps atop H2S seeps, as at Bush Hill on the floor of the Gulf of Mexico (Figure 3a; Reilly et al., 1996; Dattagupta et al., 2006; McMullin et al., 2010). Tubeworm “bushes” in cold seep regions of the Gulf of Mexico are typically rooted in the H2S-rich muds (Figure 3b). Growing individual tubes actively extend down into the H2S-rich mud as well as up into the O2-rich water column giving the cluster a morphology similar to a tree or shrub. Their “roots” extend into the earth, while “branches” extend above. Continuing the plant analogy, it seems that tubeworm shrubs absorb H2S through their “roots” and O2 through their “branches” (Freytag et al., 2001; Bergquist et al., 2003). As a group, seep tubeworms are related to the giant rift tubeworm (Riftia pachptila), which inhabits active hydrothermal seeps in active seafloor rifts.

Via a specialised haemoglobin molecule, vestimentiferan tubeworms in the Gulf of Mexico provide H2S and O2 as nutrients to sulphur-oxidising bacteria living symbiotically in trophosome structures, which extend for up to 75% of the length of each tubeworm. Unlike hydrothermal tubeworms such as Riftia pachptila that grow to lengths of more than 2 metres in less than two years, Lamellibrachia luymesi grow very slowly for most of their lives. It takes from 170 to 250 years to grow to 2 meters in length, making them perhaps the longest living known invertebrate species (Bergquist et al., 2000). With five or six species currently known to flourish there, the brine-fed cold seeps of the Gulf of Mexico host the highest biodiversity of vestimentiferan siboglinid tubeworms worldwide.

There is a time-based evolution in the biotal make-up of chemosynthetic communities in the Gulf of Mexico (Glover et al., 2010 and references therein). The earliest stage of a cold seep is characterised by a high seepage rate and the release of large amounts of biogenic and thermogenic methane, H2S and oil (Sassen et al., 1994). As authigenic carbonates with specific negative δ13C values precipitate as a metabolic byproduct of microbial methanogenesis, they provide a necessary stable substrate for the settlement of larval vestimentiferans and seep mussels. These seep communities begin with mussel (Bathymodiolus childressi) beds containing high biomass communities of low diversity and high endemicity. Individual mussels live for 100–150 years, whereas mussel beds may persist for even longer periods, with growth rates of mussels primarily controlled by methane concentrations (Nix et al., 1995).

The next successional stage consists of vestimentiferan tubeworm aggregations dominated by Lamellibrachia luymesi and Seepiophila jonesi. Young tubeworm aggregations often overlap in time with, and usually persist past the stage of mussel beds. These tubeworm aggregations and their associated faunas go through a series of successional stages over a period of hundreds of years. Declines in seepage rates result from ongoing carbonate precipitation occluding pores and so forming aquitards, as well as the influence of L. luymesi on the local biogeochemistry as it extracts ever-larger volumes of H2S. In older tubeworm aggregations, biomass, density, and number of species per square metre decline in response to reduced sulphide concentrations.

Once seep habitat space becomes available, more of the non-endemic background species, such as amphipods, chitons, and limpets, can colonise the mussel and tubeworm aggregations. Due to the lowering concentrations of sulphide and methane, the free-living microbial primary productivity is reduced. The number of associated taxa is positively correlated with the size of the tubeworm-generated habitat, so diversity in this stage remains relatively high although the proportion of endemic species is smaller in the older aggregations. This final stage may last for centuries, as individual vestimentiferan tubeworms can live for over 400 years (Cordes et al., 2009).

Even as seepage of hydrocarbons declines in a particular seep site, the authigenic carbonate layers of relict seeps can still provide a stable seafloor substrate for marine filter feeders, such as cold-water corals. The scleractinians Lophelia pertusa and Madrepora oculata, several gorgonian, anthipatharian, and bamboo coral species form extensive reef structures atop now inactive seeps on the upper slope of the Gulf of Mexico (Schroeder et al., 2005). The corals obtain their food supply form the water column and are not dependent on chemosynthetic microbes. The coral communities also harbour distinct associated assemblages, consisting mainly of the general background marine fauna, but also contain a few species exclusively associated with the corals and a few species that are common to both coral and seep habitats

Although individual tubeworms and molluscs in chemosynthetic brine pool communities may live for more than 300-400 years, vagaries in the rate of brine and nutrient supply to the seafloor mean many mussel and tubeworm colonies are overwhelmed by a rising halocline and so die in a shorter space of time. Their partially decomposed remains can spread out as part of the organic-rich debris atop the halocline, along with bacterial, algal and faecal residues, where it is acted upon by a rich community of aerobic and anaerobic decomposers. If the organic matter is mineralised or attaches to other interface precipitates such as pyrite, it sinks to the anoxic brine pool bottom, where it is largely preserved and protected from further biodegradation.

The inherently unstable nature of the seafloor in the vicinity of active salt allochthons and brine lakes means it is subject to slumping, especially in the vicinity of brine fed mud volcanoes. In such settings, parts of the carbonate-rich biostrome rim are periodically killed “en masse” as sediment about a brine pool edge collapses, slumps and slides into anoxic pool waters, carrying with it the chemosynthetic community. As well as further elevating levels of preserved organics in the brine pool bottom sediments, this process also creates potential fossil lagerstaette. Death of seep communities, even if survives such catastrophic events, ultimately comes when the supply of seep gases and liquid hydrocarbons is cut off to any single seep.

Hardgrounds, seafloor stability & stable isotopes

Associated with the brine-pool communities, and helping form an initial stable seafloor substrate for the colonising seep invertebrates, are calcite-cemented biogenic crusts. These cemented hardgrounds precipitate as a microbial byproduct wherever methane and H2S are bubbling up in and around brine pool edges, and gases are being metabolised by chemosynthetic archaea and bacteria (Canet et al., 2006; Fu Chen et al., 2007; Feng et al., 2009). The resulting biogenic calcite crusts have δ13CPDB values ranging to as low as -53‰, which is characteristic of methanogenic carbon (Figure 4a). Seep sediments retain a group of unsaturated 2,6,10,15,19-pentamethylicosane (PMID) compounds, also produced by methane-oxidising archaea, with δ13CPDB values ranging from -107.2 to -115.5‰. In combination, the isotope values, textures and biomarkers indicate a combination of bacterially catalysed methane oxidation and sulphate reduction plexi in the crusts.

Fabrics of the two flat sides of methanogenic calcite crusts crust are texturally distinct. The “top” side is composed entirely of microcrystalline calcite, while the bottom is composed entirely of “wormy” carbonate cement that is interpreted as a random, low fidelity replacement of bacteria. (Figure 4b) “Wormy” carbonate cement coats microcrystalline calcite in the interior of the thick crust and dispersed pyrite framboids appear to be indicators of collaborating colonies of methane-oxidising archaea and sulphate-reducing bacteria. Fu Chen et al., (2007) propose that the “wormy” carbonate texture, particularly with microcrystalline calcite and pyrite framboids present, is a likely indicator of biologically controlled fabrics produced during methane oxidation and sulphate reduction.

Hypersaline brines and entrained gases escaping and pooling on the Gulf of Mexico seafloor do so either into quiescent brine lakes and pools or as mud chimneys and volcanoes (Figure 5; Joye et al., 2009). Both environments are anoxic and hypersaline, brine pools are typified by low fluid-flow rates and waters free of suspended sediment, while flow rates in mud volcano chimneys are more vigorous and the waters tend to be more turbulent and carry more suspended load. The sharp salinity transition between hypersaline brine and seawater typifies the water column in both settings, and a higher suspended particle load underscores the more rapid fluid-flow regime of the mud volcano (Figure 5a, f). Brines in both are mildly sulphidic; concentrations of dissolved inorganic carbon are elevated relative to seawater. Microbial abundance is 100 times higher in brines than in the overlying seawater (Figure 5a, f), showing that brine-derived substrates produce high microbial biomass. The brines are gas charged; the dominant dissolved alkane is methane (94-99.9%) with a stable carbon isotopic composition, 13C, of -62‰.

The feeder brines to the chemosynthetic communities in much of the Gulf of Mexico form via halite dissolution and so contain little to no sulphate. Seawater sulphate diffuses into the brine, and concentrations decrease with depth, reflecting a combination of microbial consumption through sulphate reduction (both sites) and upward advection of sulphate-free brine in a mud volcano (Figure5b, g). The hydrogen profile in the mud volcano brine is relatively uniform (hundreds of nanomolar), reflecting the potential importance of autotrophic acetogenesis and/or hydrogenotrophic methanogenesis. In the brine pool, however, hydrogen concentration increases to micromolar levels between depths ≈25 and 100 cm and remains high (≈µ6 M) to 180 cm, promoting acetogenesis. Such high hydrogen concentrations indicate active fermentation and substantial inputs of labile organic matter. Concentrations of dissolved organic carbon (DOC) increases with depth (Figure 5b, g), suggesting a deep-subsurface DOC source (thermogenic?). In the brine pool, extra labile DOC, probably coming from the surrounding chemosynthetic community can further stimulate fermentation (Joye et al., 2009)

Rates of acetate production and levels of sulphate reduction are much higher in brine pools, whereas the mud volcano supports much higher rates of methane production (Figure 5d, i). Joye et al. (2009) found no evidence of anaerobic oxidation of methane (AOM), despite high methane fluxes in both settings. It suggests both these systems are leaking methane into the overlying water column. Joye et al. conclude that the different halo-adapted microbial community compositions and metabolisms are linked to differences in dissolved-organic-matter input from the deep subsurface and different fluid advection rates between the two settings.

Clathrates and methane seeps in the Gulf of Mexico

Across the slope and rise in the Gulf of Mexico, where sea bottom temperatures are suitably low, methane hydrates (clathrates) form atop focused outflow zones and oil seeps are common at the sea surface above vent clathrates (Dalthorp and Naehr, 2011). Gas hydrate or clathrate is an ice-like crystalline mineral in which hydrocarbon and non-hydrocarbon gases are frozen within rigid molecular cages of water. They can be thought of gaseous permafrost. Their occurrence is not just tied to the cold temperature portion of the deep seafloor; clathrates are the dominant seals to large gas reservoirs in the permafrost regions of Siberia. Methane hydrates are common associations where methane, which can be thermogenically or biogenically sourced, occurs just below the deep cold seafloor. In much of world, it accumulates in seafloor regions independent of any underlying evaporite occurrence (Thakur and Rajput, 2011). Evaporite edges just tend to focus the outflow zones (Figure 6).

Clathrate formation on the seafloor requires bottom temperatures not encountered until the seafloor bottom lies beneath a water column 450-500 m deep. Beneath the clathrate-covered seafloor, temperature increases with depth and this limits the depth at which gas hydrates will occur, so below most clathrate layer is an accumulation of free gas is likely. Clathrates seeps in the vicinity off brine pools are not unique to, but are often very obvious about, salt allochthon edges where salt flow induces extensional faulting and funnels a focused rise of methane, degraded oil and H2S to the cold seafloor (Chapter 6). Hence, breaks in the lateral extent of the various salt sheets act as a focusing mechanism for escaping thermogenic and biogenic methane and other gases and fluids (Figures 3, 6; Fisher et al., 2000; MacDonald et al., 2003). Rapid burial of organic-entraining sediments in supra-allochthon minibasins encourages the creation of biogenic methane that sources much of the gas escaping to the seafloor away from salt-edge focused seeps. Hence, in the salt allochthon province of the northern Gulf of Mexico, there is a definite association between brine pool chemosynthetic communities, thicker gas hydrates and the edges of minibasins (Figure 6; Reilly et al., 1996; Milkov and Sassen, 2001).

In all these setting clathrates are a food source for various methanogenic microbes, and so there are different multi-cellular lifeforms dependent on these microbes. One obvious dependency is seen in the eco-niche occupied by a small 2-4 cm-long highly specialised polychaete called Hesiocaeca methanicola (Figure 7). It was discovered in 1997 flourishing in regions of methane hydrate atop the deep seafloor in the Gulf of Mexico (Fisher et al., 2000). These “ice worms” inhabit indentations (“burrows”) in blocks and layers of methane clathrate and glean or harvest biofilms of the methanotrophic bacteria that are metabolising methane on the block surface. In turn, the ice worm supplies oxygen to the methanotrophs and via its movement appears to contribute to the dissolution of hydrates. Mature ice worms can survive in an anoxic environment for up to 96 hours. The experiments oof Fisher et al., (2000) also showed that the larvae were dispersed by currents, and died after 20 days if they did not find a place to feed.

Brine lake biota in the Mediterranean Ridges

Eight brine lakes, L’Atalante, Bannock, Discovery, Kryos, Medee, Thetis, Tyro and Urania, have been discovered and studied in the Mediterranean Ridge region of the deep eastern Mediterranean over the last 20 years (Figure 8a; see part 1). The surfaces of these brine lakes lie between 3.0 and 3.5 km below sea level, and the salinity of their brines ranges from five to 15 times higher than that of seawater. In the Bannock Basin, the various brine-filled depressions or sub-basins create a closed outer moat around a central seafloor mound that is 10 km across (Figure 8b). The chemical composition of the Tyro Basin bottom brine is related to the dissolution of the underlying halite-dominated evaporites, while the chemical composition of the Bannock Basin (Libeccio Basin in the Bannock area) implies derivation from dissolving bittern salts (de Lange et al., 1990). In the “anoxic lakes region”, sodium chloride is predominantly sourced in the L’Atalante and Urania lakes, but L’Atalante is much richer in potassium chloride than the other lakes. The Discovery basin brine is almost exclusively the product of dissolution of magnesium chloride (bischofite) salts. It has a density of 1330 kg/m3, which makes it the densest naturally occurring brine yet discovered in the marine environment (Wallmann et al., 2002). Its concentration profile in sediment beneath the brine lake shows the age of this lake is between 700 and 2000 yr. The high concentration of magnesium chloride drives the dissolution of biogenic calcium carbonate but facilitates excellent preservation of siliceous microfossils and organic matter. In basin bottom muds there are large euhedral crystals of gypsum, up to 10cm across, precipitating from these brines (Cita 2006).

Of the Mediterranean brine lakes, Lake Medee is the largest, and fills a narrow depression at the Eastern edge of the abrupt cliffs of the small evaporite ridge located 70 nautical miles SW of Crete (Figure 8a). The lake depression is approximately 50 km in length with a surface area of about 110 km2 and a volume of nearly 9 km3, which places Lake Medee among the largest of the known DHALs in the deep-sea environment. Although all the Mediterranean DHALs lie geographically close to each other, their hydrochemical diversity suggests that dissolving salt mineralogies were different. Salinity levels are much higher in some dues to the presence off nearby bittern layers. For example, Discovery Lake and Lake Kryos have salinities and MgCl2 proportions indicative of bischofite dissolution. Even so, it seems like, mostly sulphate-reducers can still metabolise in the extremely saline MgCl2 waters of Lake Kryos (Steinle et al., 2018).

In contrast to the brine lakes and seeps in salt-allochthon terrane of the Gulf of Mexico, seep megafauna is so far absent in the various documented modern brine lakes along the Mediterranean Ridges (Figure 8d). The brine lakeshore edge communities are mostly microbial, as are the lifeforms that make up the pelagic biota off the halocline. Biological studies on the anoxic basins of the Eastern Mediterranean started after the discovery of gelatinous matter of organic origin in the brine lake sediments (Figure 8c; Brusa et al., 1997). The laminar gelatinous matter was observed within the cores containing anoxic sediments obtained during oceanographic expeditions for geological study of the Mediterranean Ridge. Microbiological and ultrastructural investigations were carried out on core sediment samples and on the overlying water. Various authors demonstrated the organic nature of the mucilaginous pellicles found in the cores and their relation with numerous microbic forms present in all the samples. Viable microorganisms, prevalently Gram-negative and aerobic as well as facultative anaerobes, were found in the halocline water samples. Different microbic forms were isolated in pure culture: a vibrio (Nitrosovibrio spp.), a coccus (Staphylococcus sp.) and some rods of the family Pseudomonadaceae. In addition, laminar formations were observed in a growth medium of mixed cultures that could be interpreted as the first stages of the mucilaginous pellicles seen in the cores. Earlier studies described the geological and physiochemical characteristics of such habitats (Erba et al. 1987; Cita et al. 1985). Subsequent work using metagenomic techniques have documented a prosperous microbial community inhabiting the halocline of most of the Mediterranean brine lakes.

DHAL interfaces in the Mediterranean Sea deeps act as hot spots of deep-sea microbial activity that significantly contribute to de novo organic matter production. Metabolically active prokaryotes are sharply stratified across the halocline interfaces in the various brine lakes and likely provide organic carbon and energy that sustain the microbial communities of the underlying salt-saturated brines. Since metagenomic analysis of DHALs is still in its infancy, the metabolic patterns prevailing in the organisms residing in the interior of DHALs remains mostly unknown. What is known is that the redox boundary at the brine/seawater interface provides energy to various types of chemolithic and heterotrophic communities. Aerobic oxidations of reduced manganese and iron, sulphide and intermediate sulphur species, diffusing from anaerobic brine lake interior to the oxygenated upper layers of the haloclines are highly exergonic processes capable of supporting an elevated biomass at DHAL interfaces (Yakimov et al., 2013). Depending on availability of oxygen and other electron acceptors bacterial autotrophic communities belonging to Alpha-, Gamma- and Epsilon-proteobacteria fix CO2 mainly via the Calvin-Benson-Bassham and the reductive tricarboxylic acid (rTCA) cycles, respectively.

Biomarker associations of the organics accumulating in the brine lakes define two depositional styles: typical marine and hypersaline (Burkova et al., 2000). For example, algal and bacterial biomarkers typical of saline environments were found in layers 0.60 to 0.75 m below the sediment surface in the Tyro Lake Basin, as well as standard marine indicators derived from pelagic fallout (“rain from heaven”). Saline indicators include; regular C-25 isoprenoids, squalane, lycopane, isolycopane, tetraterpenoid and tetrapyrrolic pigments, monoalkylcyclohexanes, tricyclic diterpanes, steranes, hopanes, bio- and geohopanes. According to Burkova et al. (2000), the saline organic signatures come from microbial mat layers, redeposited from a Messinian source into the sapropels of the modern depression. Alternatively, they may indicate the activities of a chemoautotrophic community, which flourishes at the halocline or around active brine vents. As in the Orca Basin, the organic content of the bottom sediments of the Mediterranean brine pools is much higher than is found in typical deep seafloor sediment (Figure 9a).

Anoxic hypersaline brines in Mediterranean brine lakes are highly sulphidic and among the most sulphidic bodies of water in the marine realm, with H2S concentrations consistently greater than 2-3 mmol (Table 1; Henneke et al., 1997). The brine body below the Urania chemocline is more than 100 m thick and contains up to 11 mM hydrogen sulphide, making it the most sulphidic water body in the known marine realm. In combination with the sulphide are very high levels of methane both in and below the halocline (≈5.56 mM; Borin et al., 2009). In contrast, there is little to no H2S in the anoxic bottom brine of the Orca Basin (Table 1). There the iron concentration is 2 ppm, a value more than 1000 times higher than in the overlying Gulf of Mexico seawater. Such high levels of reducible iron in the Orca Basin are thought to explain the lack of H2S in the bottom brine and a preponderance of framboidal pyrite and extractable iron in the bottom sediments (Sheu, 1987). Both the Orca Basin and the brine pools on the floor of the Mediterranean, show sulphate levels that can be more than twice that of the overlying seawater.

So, organic debris first formed at the halocline can then accumulated as pellicle layers within the pyritic bottom muds (laminites). Pellicular debris is also carried to the bottom during the emplacement of turbidites when the halocline is disturbed by turbid overflow (Figure 10; Erba, 1991). Hence, pellicular layers are typically aligned parallel to lamination, or are folded parallel to the sandy bases of the turbidite flows, or line up parallel to deformed layers within slumped sediment layers. Individual pellicle layers are 0.5 to 3 mm thick and dark greenish-grey in colour. Similar pellicular layers cover the surface of, or are locked within, recent gypsum crystals recovered from bottom sediments of the Bannock area. This gypsum is growing today on the bottom of the Bannock Basin, atop regions about the brine pool margin that are directly underlain by dissolving Miocene evaporites (Corselli and Aghib, 1987; Cita 2006). Other than the Dead Sea, it is one of the few modern examples of a deepwater evaporite, but its seepage-fed genesis means it is a poor analogue for deepwater basinwide salt units.

The community of bacteria and archaea flourishing at the halocline in sulphidic marine brine pools on the deep Mediterranean floor is quite diverse, mostly independent of primary production in the euphotic zone, with the number of identified unique halobacteria and haloarchea species expanding every year (Albuquerque et al., 2012). Bottom brine in the Urania brine lake has a salinity of 162‰, and the chemocline of the brine lake is some 3490m below the ocean surface, so only a minimal amount of phytoplanktonic organic carbon ever reaches the 20m thick chemocline. Yet the oxic waters of the upper part of the chemocline support a rich bacterial and archaeal assemblage in and below the interface between the hypersaline brine and the overlying seawater, much like the chemosynthetic bacterial community associated with the halocline in Lake Mahoney (Sass et al., 2001; Borin et al., 2009).

Sulphide concentration in the Urania Basin increases from 0 to 10 mM within a vertical interval of 5 m across the interface (Figure 11a). Within the halocline, the total bacterial cell counts and the exoenzyme activities are elevated and biogenic activity continues below the halocline. Bacterial sulphate reduction rates measured in this layer are ≈ 14 nmol SO4 cm-3 d-1 and are among the highest in the marine realm. They correspond to the zone of maximum bacterial activity in the chemocline (Figure 11b). Particulate organic content is 15 times greater than that in the overlying normal marine waters. A similar focus of microbial occurrence (bacterial and archaeal) is seen at the halocline in l’Atalante Basin and is probably typical of all chemocline layers in the various Bannock brine lakes (Yakimov et al., 2007)

Employing 11 cultivation methods, Sass et al. 2001 isolated a total of 70 bacterial strains from the chemocline in the Urania Basin (Figure 11a). These strains were identified as the flavobacteria, Alteromonas macleodii, and Halomonas aquamarina. All 70 strains could grow chemo-organoheterotrophically under oxic conditions. Twenty-one of the isolates could grow both chemo-organotrophically and chemo-lithotrophically (decomposers and fermenters). While the most probable numbers in most cases ranged between 0.006 and 4.3% of the total cell counts, an unusually high value of 54% was determined above the chemocline with media containing amino acids as the carbon and energy source.

Subsequent detailed work focused on the various layers that make up the Urania halocline showed the high sulphide levels in and below the halocline, make it a mecca for bacterial sulphate reducers, as do high levels of methane for the methanogens (Figure 11b; Borin et al., 2009). Microbial abundance showed a rapid increase by two orders of magnitude from 3.9 x 104 cells mL-1 in the deep oxic seawater immediately above the basin, up to 4.3 x 106 cells mL-1 in the first half of interface 1. Although less pronounced than in the first chemocline, a second increase in microbial counts occurred in interface 2. Deceleration of falling particulate organic matter from the highly productive interface 1, is probably responsible for stimulating microbial growth and hence cell numbers in interface 2. That is, compared to the overlying seawater column, bacterial cell numbers increased up to a hundred-fold in interface 1 and up to ten-fold in interface 2. This is a consequence of elevated nutrient availability, with higher numbers in the upper interface where the redox gradient was steeper. Bacterial and archaeal communities, analysed by DNA fingerprinting, 16S rRNA gene libraries, activity measurements, and cultivation, were highly stratified within the various layers of the chemocline and metabolically more active along the various chemocline layers, compared with normal seawater above, or the uniformly hypersaline brines below.

Detailed metagenome analysis of 16S rRNA gene sequences revealed that in both chemocline interfaces the e- and d-Proteobacteria were abundant, predominantly as sulphate reducers and sulphur oxidisers, respectively (Figure 11b). The only archaea in the first 50 cm of interface 1 were Crenarchaeota, which consist of organisms having sulphur-based metabolism, and hence could play a role in sulphur cycling in the upper interface. In the deepest layers of the basin below the halocline, MSBL1, putatively responsible for methanogenesis, dominated among archaea (Figure 11b). The work of Borin et al. (2009) illustrate that a well adapted and complex microbial community is thriving in the Urania basin’s extreme chemistry, The elevated biomass centred on the halocline is driven mainly by sulphur cycling and methanogenesis.

Similarly detailed studies of interface-controlled chemosynthetic communities in other Mediterranean DHALs have been documented in Lake Thetis (Ferrer et al., 2012; Oliveri et al., 2013) and Lake Medee (Yakimov et al., 2013). Medee Lake is the largest known DHAL on the Mediterranean seafloor and has two unique features: a complex geobiochemical stratification and an absence of chemolithoautotrophic Epsilonproteobacteria, which usually play the primary role in dark bicarbonate assimilation in DHALs interfaces worldwide. Presumably, because of these features, Medee is less productive and exhibits a reduced diversity of autochthonous prokaryotes in its interior brine layers. Indeed, the brine community almost exclusively consists of the members of euryarchaeal and bacterial KB1 candidate divisions which a ubiquitous in the DHAL biota worldwide. In Medee, as elsewhere, they are thriving on small organic molecules produced by a combination of degraded marine plankton and moderate halophiles living in the overlying stratified brine column.

Outside off the microbial makeup of DHAL communities, one of the more exciting discoveries in the brine lakes of the Mediterranean ridges is the likely discovery of multicellular life of the Phylum Loricifera (“Beard shells) capable of living and reproducing in the absence of oxygen. Loricifera (from Latin, lorica, corselet (armour) + ferre, to bear) is a phylum made up of very small to microscopic marine cycloneuralian sediment-dwelling animals with 37 described species. Their size ranges from 100 µm to ca. 1 mm and individuals are characterised by a protective outer case called a lorica and by their habitat, which is in the spaces between marine sediment particles. The phylum was first discovered in tidal sediments in 1983 and is among the most recently discovered groups of Metazoans. Individuals attach themselves quite firmly to the sediment substrate, and hence the phylum remained undiscovered for so long. In 2010, viable specimens of Spinoloricus cinziae, along with two other newly discovered species, Rugiloricus nov. sp. and Pliciloricus nov. sp., were found in the sediment core from below the anoxic L'Atalante basin of the Mediterranean Sea (Danovaro et al., 2010, 2016). The species cellular innards appear to be adapted for a zero-oxygen life as their mitochondria appear to act as hydrogenosomes, organelles which already provide energy in some anaerobic single-celled creatures known. Before their discovery, living and reproducing exclusively in an oxygen-free setting was thought to be a lifestyle open only to viruses and single-celled microorganisms. The ability of these anoxic brine-dwelling creatures to live solely in an oxygen-free environment is questioned still by other workers (Bernhard et al., 2015).

Neither Tyro nor Bannock Basin bottom sediments show a significant correlation between pyritic sulphur and the organic carbon in the bottom sediments, suggesting predominantly syngenetic pyrite evolution in bottom sediments of these brine lakes (Henneke et al., 1997). That is, both pyritic and humic sulphur preserved in the bottom sediments formed either in the lower water column or at the sediment-brine interface, not in the sediment itself. Ongoing diagenetic processes within the bottom sediments only form an additional 5% of the total pyrite. Van der Sloot et al. (1990) clearly showed that metal sulphides, as well as organics and other minerals, precipitate at the brine-seawater interface in the Tyro Basin, as they do in the Orca Basin. They found extremely high concentrations of Co (0.015%), Cu (1.35%) and Zn (0.28%) in suspended matter at the halocline. These high particulate Co, Cu and Zn concentrations correspond to sharp increases in dissolved sulphide across the interface (a redox front), and indicate precipitation of metal sulphides at the interface. Humic sulphur in the bottom sediments correlates with the pyritic sulphur distribution and is related to the amount of gelatinous pellicle derived from bacterial mats growing at the halocline between oxic seawater and bottom brine (Erba, 1991, Henneke et al., 1997).

Additionally, the degree of pyritisation in the sediments (DOP ≈ 0.62) indicates that present-day pyrite formation is limited by the reactivity of Fe in the Bannock and Tyro basins and not by the availability of organic matter, the latter being the process that limits pyrite formation in most normal marine settings (Figure 9b). The degree of pyritisation (DOP) is defined as [(pyritic iron)/(pyritic iron + reactive iron)]. Raiswell et al. (1988) showed that DOP in ancient sediments can distinguish anoxic from normal marine sediments. Anoxic sediments show DOP values between 0.55 and 0.93, while normal marine sediments have DOP values less than 0.42. The DOP levels in the Bannock and Tyro basins confirm observations made in ancient anoxic sediments. Thus, although the Tyro and Bannock basin brines differ in their major element chemistry, reflecting a different salt source, their reduced sulphur species chemistry appears to be similar, but is significantly different from standard marine systems and capable of precipitating metal sulphides above the sediment surface.

Life in the Red Sea brine deeps

The Atlantis II Deep marks the northern-most end of the Atlantis II Shagara- Erba Trough section, hosting numerous sub-deeps like the Discovery and Aswad Deep (Figure 12). In general, the Atlantis II Deep area has a smoother bathymetric character than the Thetis-Hadarba-Hatiba and Shagara-Aswad-Erba Troughs, due to massive inflow of salt and sediments from nearly all sides into the deep. In the Atlantis II deep, Siam et al. (2012) identified metagenomic archaeal groups in high relative abundance at the bottom of a sediment core from the Atlantis II Deep, which, as in the Kebrit Deep, are another case of the dominance of Archaea. Their results showed that the dominant archaeal inhabitants in the bottom layer (3.5 m depth to the seafloor) included Marine Benthic Group E, and the archaeal ANME-1 ( anaerobic methane consumers metagenome. The presence of the latter was also confirmed in a study of a barite mound in the Atlantis II Deep (Wang et al., 2015), but the former was not detected in this later study.

In metagenomic studies of the Atlantis II sediments, Cupriavidus (Betaproteobacteria) and Acinetobacter (Gammaproteobacteria) are the most abundant species in the surface layer (12 cm) and the bottom layer (222 cm) of a sediment core obtained in 2008. Both bacterial species were not the dominant inhabitants in the ABS core analysed in the present study. Due to tremendous differences between brine water and sediment chemistry in the Deep, their microbial communities differ remarkably. The lower convective layers of the Atlantis II and Discovery brine pools are dominated by Gammaproteobacteria, while Alphaproteobacteria and Betaproteobacteria are the major bacterial groups in the upper layers of Atlantis II sediment (Bougouffa et al., 2013). All the above discrepancies in composition of microbial communities in the two Deeps were probably caused by 1) primer selection for amplification of rRNA genes; 2) different microenvironments in the sampling sites; 3) taxonomic assignment criteria employed by different studies; 4) different experimental procedures, and 5) sampling bias due to low biomass in sampling sites. Except for these potential problems, this study demonstrates the profound changes in microbial communities in deep-sea hydrothermal sediment under the influence of extensive mineralisation process. Many of the groups detected in the S-rich Atlantis II section are likely to play a dominant role in the cycling of methane and sulphur due to their phylogenetic affiliations with bacteria and archaea involved in anaerobic methane oxidation and sulphate reduction.

In the Kebrit Deep on the deep floor of the Red Sea, an assemblage of halophilic archaea and bacteria similar to that of the DHALs of the Mediterranean Deeps flourish in hypersaline waters below the chemocline (Figure 13). Kebrit Deep (24°44’N, 36°17’E) measures 1 by 2.5 km, with a maximum depth of 1549 m and is one of the smallest salt allochthon-associated brine-pools of the Red Sea. It is located around 300 km nothwest the well-known metalliferous Atlantis II deep (see previous article). The Kebrit Deep is filled by an 84 m thick, anaerobic, slightly acidic brine lake (pH approximately 5.5) with a salinity of 260‰ and a temperature of 23.3°C (Antunes et al., 2011). The brine has a high gas content that is made up mainly of CO2, H2S, small amounts of N2, methane and ethane, with remarkably high quantities of H2S (12–14 mg S l-1; Hartmann et al., 1998). The presence of sulphur is self-evident by the strong, characteristic odour present in brine samples, and hence the name of the basin (Kebrit is the Arabic word for sulphur). Like the Atlantis II deep there are impregnated massive sulphides accumulations on the floor of Kebrit Deep. Kebrit samples are porous and fragile, and consist mainly of pyrite and sphalerite. Prior to gene sequencing studies, sulphur isotope values provided substantial evidence for biogenic sulphate reduction being involved in sulphide-forming processes in Kebrit Deep. They are linked to bacterial methane oxidation and sulphate reduction centred on the brine-seawater interface (see Chapter 15 in Warren 2016 for metallogenic details).

Most of the archaeal metagenomic sequences in Kebrit Deep cluster within the Thermoplasmatales (Marine group II, Marine Benthic group D, and the KTK-4A cluster) among the Euryarchaeota, while the remaining sequences do not show high similarity to any of the known phylogenetic groups (Figure 13). One of these sequences was shown to cluster with the later-described SA2 group, while another (accession number AJ133624) clusters together with two gene sequences from L’Atalante Basin waters, defining a novel deeply-branching phylogenetic lineage within the Crenarchaeota.

Gene sequencing studies on water samples from the brine-seawater interface in the Kebrit deep retrieved sequences from the KB1 group, as well as Clostridiales (mostly Halanaerobium), Spirochetes (ST12-K34/MSBL2 cluster), Epsilonproteobacteria and Actinobacteria, but no archaeal sequences were detected in these interface samples (Antunes et al.,2011). Under strictly anaerobic culture conditions, novel halophiles were isolated from samples of these waters and belong to the halophilic genus Halanaerobium. They are the first representatives of the genus obtained from deep-sea, anaerobic brine pools (Eder et al., 2001). Within the genus Halanaerobium, they represent new species that grow chemo-organotrophically at NaCl concentrations ranging from 5 to 34%. They contribute significantly to the anaerobic degradation of organic matter, which formed at the brine-seawater interface and is slowly settling into the bottom brine.

Similarities in the makeup of the Archaeal population, tied to similar metabolic process sets at the brine interface across various deep seafloor brine lakes in the Gulf of Mexico, the Mediterranean and the Red Sea. Compared with other hydrothermal sediments around the world, the Atlantis II hydrothermal field is unique in that sulphur and nitrogen oxides are low in the pore water of the sediments. This probably leads to lack of ANME . It seems, different geochemical conditions of hydrothermal marine and cool seep sediments across the deepsea sub-seafloor resulted in various niche-specific microbial communities.

Life in the Dead Sea

As defined in the salty matters article previous to this, the Dead Sea can be considered a continental counterpart of a marine DHAL where there is no overlying body of marine water. Instead, the Dead Sea brine mass is in direct contact with the atmosphere.

The Dead Sea provides one of nature’s supreme tests of survival of life. The negative-water balance in the Dead Sea hydrology over recent decades resulted in ever-rising salinity and divalent-cation ratios, cumulating in the current highly drawdown situation (See Warren 2016, Chapter 4 for a summary of the relevant hydrological evolution. Today the brines have reached a salinity level more than 348 /l total dissolved salts, with a high ratio of (Ca + Mg) to Na. Water activity (Aw, a measure based on the partial pressure of water vapour in a substance, and correlated with the ability to support microorganisms) of the Dead Sea is extremely low (Aw ≈ 0.669), even lower than that of saturated-NaCl solution (Aw ≈ 0.753±0.004), and is thus unbearable for most life forms (Kis-Papo et al., 2014).

Nevertheless, a number of halobacteria (Archaea), one green algal species (Dunaliella parva), and several fungal taxa withstand these extreme conditions(Kis-Papo et al., 2014). Most organisms in the Dead Sea survive in fresher-water spring refugia or in their dormant stages or and only revive when salinity is temporarily reduced during rare massive flooding events (Ionescu et al., 2012.

Effects of occasional freshening on biomass in stratified brine columns that are supersaline, not mesohaline, is clearly seen in the present “feast or famine” productivity cycle of the Dead Sea (Warren, 2011; Oren and Gurevich, 1995; Oren et al., 1995; Oren 2005). Dunaliella sp, a unicellular green alga variously described in the past as Dunaliella parva or Dunaliella viridis, is the sole primary producer in the Dead Sea waters. Then there are several types of halophilic archaea of the family Halobacteriaceae (prokaryotes) which consume organic compounds produced by the algae.

Two distinct periods of organic productivity (feast) have been documented in the upper lake water mass since the Dead Sea became holomictic in 1979 (Oren, 1993, 1999). The first mass developments of Dunaliella sp. (up to 8,800 cells/ml) began in the summer of 1980 following dilution of the saline upper water layers by the heavy winter rains of 1979-1980 Figure 14a, b). The rains drove a rapid rise of 1.5 metres in lake level and an increase in the level of phosphates in the lake’s surface waters (Figure 14c). This bloom was quickly followed by a blossoming in the numbers of red halophilic archaea (2 x 107 cells/ml), Dunaliella numbers then declined rapidly following the complete remixing of the water column and the associated increase in salinity of the upper water mass. By the end of 1982, Dunaliella had disappeared from the main surface water mass. Archaeal numbers underwent a slower decline.

During the period 1983-1991 the lake was holomictic, halite-saturated and no Dunaliella blooms were observed. Viable halophilic and halotolerant archaea were probably present in refugia about the lake edge during this period but in meagre numbers. Then heavy rains and floods of the winter of 1991-1992 raised the lake level by 2 metres and drove a new episode of meromictic stratification as the upper five metres of the water column was diluted to 70% of its normal surface salinity (Figure 14d). High densities of Dunaliella reappeared in this upper less saline water layer (up to 3 x 104 cells/ml) at the beginning of May 1992, rapidly declining to less than 40 cells/ml at the end of July 1992 (Figure 15). An associated bloom of heterotrophic haloarchaea (3 x 107 cells/ml) continued past July and continued to impart a reddish colour to the surface and nearsurface waters.

Much of the archaeal community was still present at the end of 1993, but the amount of carotenoid pigment per cell had decreased two- to three-fold between June 1992 and August 1993 (Oren and Gurevich, 1995). A remnant of the 1992 Dunaliella bloom maintained itself at the lower end of the pycnocline at depths between 7 and 13 m (September 1992- August 1993), perhaps chasing nutrients rather than light. Its photosynthetic activity was low, and very little stimulation of archaeal growth and activity was associated with this algal community (Figure 15). It seems that once stratification ends and the new holomictic period begins, the remaining Archaeal community, which was primarily restricted to the upper water layers above the halocline, spreads out more evenly over the entire upper water column until it too dies out. No substantial algal and archaeal blooms have developed in the Dead Sea since the winter floods of 1992-1993 until today

Underwater freshwater to brackish springs are likely refugia to much of the life in the Dead Sea and are inhabited by interesting microbial communities including chemolithotrophs, phototrophs, sulphate reducers, nitrifiers, iron oxidisers, iron reducers, and others. The springs also host numerous cyanobacterial and diatomatous mats with sulfate-reducers near the base of the foood chain (Oren et al., 2008; Ionescu et al., 2012). Sequences matching the 16S rRNA gene of known sulphate-reducing bacteria (SRB) and sulphur oxidising bacteria (SOB) were detexcted in all microbial mats centered on freshwater springs as well as in the Dead Sea water column (Häusler et al., 2014). Generally, sequence abundance of SRB and SOB was higher in the microbial mats than in the Dead Sea, indicating that the conditions for both groups are more favorable in the spring environments.

The springs also supply nitrogen, phosphorus and organic matter to the Dead Sea microbial communities. Due to frequent fluctuations in the freshwater flow volumes in the springs and local salinity, microorganisms that inhabit these springs must be capable of withstanding large and rapid salinity fluctuations and the population proportions vary according to the Spring chemistry (Ionescu et al., 2012).

Salt dissolution, seafloor salinity and halophilic extremophile populations

In most DHALs, the rate of vertical mixing across the extreme density gradients between brine and overlying seawater is extremely slow (Steinle et al., 2018). Hydrochemically, depending on the nature of the dissolving salt supply, seawater and DHAL brines can differ sharply in their solute composition, in particular, in the concentrations of the critical electron donors and acceptors so crucial to the functioning of life. In that a narrow (1– 3 m) chemocline (halocline) forms a transition zone between the two quite-different hydrologies that define a DHAL water column, microbial ecologies have evolved to inhabit particular portions of the halocline as well as the brine lake and the normal marine deepwater columns (Figure 16).

In contrast to the overlying seawater, the bottom brines are anoxic but contain electron acceptors other than oxygen most importantly sulphide and methane. Hence, hotspots of chemosynthetic (not photosynthetic) activity have evolved that flourish at these brine-seawater interfaces, where the principal reactions at the base of the food chain are anoxic and encompass sulphate reduction, methanogenesis, and microbial heterotrophy. Highly-adapted microbial life continues to function even in the most extreme hypersaline conditions found in some DHALs, such as in Lake Kryos where MgCl2-rich chemistries dominate, or in the Atlantis II Deep where there is a combination of extreme temperatures and salinities.

In the Gulf of Mexico, an endosymbiotic megafauna constructs methanogenically-cemented carbonate biostromes as lake fringe mussel-dominated communities or polychaete forests atop cool water H2S seeps. Both the microbial population and the megafauna that exploits this chemosynthetic base to the food chain flourish best in seafloor regions defined by the long-term focused escape of methane or H2S (Figure 16). Cool-seep brine lakes were first discovered in the Gulf of Mexico in the early 1980s, but similar hydrocarbon-dependent cool-seep communities with their own megafauna accumulations are now documented in other parts of the world characterised by the naturally-focused escape of hydrocarbons to the seafloor (for example, atop cool-water brine seeps along the slope and rise of the east and west coasts of North America and in the Black Sea.

The relative long-term stability of cool-seep ecology, tied to the chemical stability of the niche, is seen when lifespans of hydrothermal endosymbiotic communities living chemosynthetically about thermal vents along mid-oceanic ridges are compared to Gulf of Mexico communities. Endosymbiotic polychaete and clam species in the brine lakes and seeps of the Gulf of Mexico can live for a hundred or more years, while lifespans in similar endosymbiotic polychaete and clam species in hydrothermal ridges communities are less than 30-50 years.

Moving onshore, into the partial analogue offered by the salt-karst fed Dead Sea depression, we see Dead Sea biomass is subject to much shorter-term changes in the salinity and nutrient content of its uppermost water mass (Feast and Famine cycles as documented in Warren, 2011, 2016 Chapter 9). The freshening water mass above a lake halocline his ephemeral in the current longterm holomictic hydrology of the Dead Sea (see Warren 2016 chapter 4 for details). The changes in surface water salinity are tied to the periodic influx of a freshened upper water mass. These climatically-driven fluctuation to the the extent and activity of the halotolerant and halophilic community in the upper water mass, and the Feast or Famine responses of the Dead Sea biota, are different to the longterm niche stability created by the presence of a perennial oceanic water mass over a salt-karst induced halocline and brine lake in a DHAL sump on the deep seafloor. The latter is continually resupplied brine and chemosynthetic nutrients via the dissolution and focusing effect of the underlying salt sheet. The hydrology of a DHAL system only shuts down when all the mother salt is dissolved or cut off.

Accordingly, rather than the hundreds of years of longterm growth (albeit at relatively slow metabolic rates) that we see in a DHAL, in the Dead Sea we see that freshening facilitates a rapid spread of a halotolerant alga (Dunaliella sp.) and associated halophilic microbes and viruses. The propagation and persistence of a large biomass pulse in the Dead Sea is measured in timeframes of months. The halotolerant photo-synthesisers can only spread out from long-term refugia communities once the surface salinities fall to levels that allow the photosynthesising base too the Lake food chain inhabit fresher water springs regions about the lake margins. Comparison to the DHAL and Dead Sea communities underlines how life will evolve into any neighbourhood, even if conditions are extremely challenging


Albuquerque, L., M. Taborda, V. La Cono, M. Yakimov, and M. S. da Costa, 2012, Natrinema salaciae sp. nov., a halophilic archaeon isolated from the deep, hypersaline anoxic Lake Medee in the Eastern Mediterranean Sea: Systematic and Applied Microbiology, v. 35, p. 368-373.

Antunes, A., D. K. Ngugi, and U. Stingl, 2011, Microbiology of the Red Sea (and other) deep-sea anoxic brine lakes: Environmental Microbiology Reports, v. 3, p. 416-433.

Augustin, N., F. M. van der Zwan, C. W. Devey, M. Ligi, T. Kwasnitschka, P. Feldens, R. A. Bantan, and A. S. Basaham, 2016, Geomorphology of the central Red Sea Rift: Determining spreading processes: Geomorphology, v. 274, p. 162-179.

Bergquist, D. C., T. Ward, E. E. Cordes, T. McNelis, S. Howlett, R. Kosoff, S. Hourdez, R. Carney, and C. R. Fisher, 2003, Community structure of vestimentiferan-generated habitat islands from Gulf of Mexico cold seeps: Journal of Experimental Marine Biology and Ecology, v. 289, p. 197-222.

Bergquist, D. C., F. M. Williams, and C. R. Fisher, 2000, Longevity record for deep-sea invertebrate: Nature, v. 403, p. 499-500.

Bernhard, J. M., C. R. Morrison, E. Pape, D. J. Beaudoin, M. A. Todaro, M. G. Pachiadaki, K. A. Kormas, and V. P. Edgcomb, 2015, Metazoans of redoxcline sediments in Mediterranean deep-sea hypersaline anoxic basins: BMC Biology, v. 13, p. 105.

Borin, S., L. Brusetti, F. Mapelli, G. D'Auria, T. Brusa, M. Marzorati, A. Rizzi, M. Yakimov, D. Marty, G. J. De Lange, P. Van der Wielen, H. Bolhuis, T. J. McGenity, P. N. Polymenakou, E. Malinverno, L. Giuliano, C. Corselli, and D. Daffonchio, 2009, Sulfur cycling and methanogenesis primarily drive microbial colonization of the highly sulfidic Urania deep hypersaline basin: Proceedings of the National Academy of Sciences, v. 106, p. 9151-9156.

Bougouffa, S., J. K. Yang, O. O. Lee, Y. Wang, Y. Batang, A. Al-Suwailem, and G. Qian, 2013, Distinctive Microbial Community Structure in Highly Stratified Deep-Sea Brine Water Columns: Appl. Environ. Microbiol., v. 79, p. 3425-3437.

Brusa, T., E. Del Puppo, A. Ferrari, G. Rodondi, C. Andreis, and S. Pellegrini, 1997, Microbes in deep-sea anoxic basins: Microbiol.Res., v. 152, p. 45-56.

Burkova, V. N., E. A. Kurakolova, N. S. Vorob'eva, M. L. Kondakova, and O. K. Bazhenova, 2000, Hydrocarbons of the hypersaline environment of the Tyro deep-sea depression (eastern Mediterranean): Geochemistry International, v. 38, p. 883-894.

Camerlenghi, A., 1990, Anoxic Basins of the eastern Mediterranean: geological framework: Marine Chemistry, v. 31, p. 1-19.

Canet, C., R. M. Prol-Ledesma, E. Escobar-Briones, C. Mortera-Gutierrez, R. Lozano-Santa Cruz, C. Linares, E. Cienfuegos, and P. Morales-Puente, 2006, Mineralogical and geochemical characterization of hydrocarbon seep sediments from the Gulf of Mexico: Marine and Petroleum Geology, v. 23, p. 605-619.

Cita, M. B., 2006, Exhumation of Messinian evaporites in the deep-sea and creation of deep anoxic brine-filled collapsed basins: Sedimentary Geology, v. 188-189, p. 357-378.

Cita, M. B., K. A. Kastens, F. W. McCoy, F. Aghib, A. Cambi, A. Camerlenghi, C. Corselli, E. Erba, M. Giambastiani, T. Herbert, C. Leoni, P. Malinverno, A. Nosetto, and E. Parisi, 1985, Gypsum precipitation from cold brines in an anoxic basin in the eastern Mediterranean: Nature (London), v. 314, p. 152-154.

Cordes, E. E., D. C. Bergquist, and C. R. Fisher, 2009, Macro-Ecology of Gulf of Mexico Cold Seeps, Annual Review of Marine Science: 1, p. 143-168.

Corselli, C., and F. S. Aghib, 1987, Brine formation and gypsum precipitation in the Bannock Basin (eastern Mediterranean): Marine Geology, v. 75, p. 185-199.

Dalthorp, M., and T. H. Naehr, 2011, Structural and Stratigraphic Relationships of Hydrocarbon Seeps in the Northern Gulf of Mexico and Geological Factors Contributing to Migration Variations: Gulf Coast Association of Geological Societies Transactions, v. 61, p. 105-122.

Danovaro, R., A. Dell'Anno, A. Pusceddu, C. Gambi, I. Heiner, and R. Møbjerg Kristensen, 2010, The first metazoa living in permanently anoxic conditions: BMC Biology, v. 8, p. 30.

Danovaro, R., C. Gambi, A. Dell’Anno, C. Corinaldesi, A. Pusceddu, R. C. Neves, and R. M. Kristensen, 2016, The challenge of proving the existence of metazoan life in permanently anoxic deep-sea sediments: BMC Biology, v. 14, p. 43.

Dattagupta, S., L. L. Miles, M. S. Barnabei, and C. R. Fisher, 2006, The hydrocarbon seep tubeworm Lamellibrachia luymesi; primarily eliminates sulfate and hydrogen ions across its roots to conserve energy and ensure sulfide supply: Journal of Experimental Biology, v. 209, p. 3795.

de Lange, G. J., J. J. Middleburg, C. H. van der Weijden, G. Catalano, G. W. Luther, III, D. J. Hydes, J. R. W. Woittiez, and G. P. Klinkhammer, 1990, Composition of anoxic hypersaline brines in the Tyro and Bannock Basins, eastern Mediterranean: Marine Chemistry, v. 31, p. 63-88.

Edgcomb, P. V., and M. J. Bernhard, 2013, Heterotrophic Protists in Hypersaline Microbial Mats and Deep Hypersaline Basin Water Columns: Life, v. 3.

Erba, E., 1991, Deep mid-water bacterial mats from anoxic basins of the eastern Mediterranean: Marine Geology, v. 100, p. 83-101.

Erba, E., G. Rodondi, E. Parisi, H. L. Ten Haven, M. Nip, and J. W. De Leeuw, 1987, Gelatinous pellicles in deep anoxic hypersaline basins from the Eastern Mediterranean: Marine Geology, v. 75, p. 165-183.

Feng, D., H. H. Roberts, P. Di, and D. Chen, 2009, Characteristics of hydrocarbon seep-related rocks from the deep Gulf of Mexico: Gulf Coast Association of Geological Societies Transactions, v. 59, p. 271-275.

Ferrer, M., J. Werner, T. N. Chernikova, R. Bargiela, L. Fernández, V. La Cono, J. Waldmann, H. Teeling, O. V. Golyshina, F. O. Glöckner, M. M. Yakimov, P. N. Golyshin, and M. S. C. The, 2012, Unveiling microbial life in the new deep-sea hypersaline Lake Thetis. Part II: a metagenomic study: Environmental Microbiology, v. 14, p. 268-281.

Fisher, C. R., I. R. MacDonald, R. Sassen, C. M. Young, S. A. Macko, S. Hourdez, R. S. Carney, S. Joye, and E. McMullin, 2000, Methane Ice Worms: Hesiocaeca methanicola Colonizing Fossil Fuel Reserves: Naturwissenschaften, v. 87, p. 184-187.

Freytag, J. K., P. R. Girguis, D. C. Bergquist, J. P. Andras, J. J. Childress, and C. R. Fisher, 2001, A paradox resolved: Sulfide acquisition by roots of seep tubeworms sustains net chemoautotrophy: Proceedings of the National Academy of Sciences of the United States of America, v. 98, p. 13408-13413.

Fu Chen, D., Q. Liu, Z. Zhang, L. M. Cathles Iii, and H. H. Roberts, 2007, Biogenic fabrics in seep carbonates from an active gas vent site in Green Canyon Block 238, Gulf of Mexico: Marine and Petroleum Geology, v. 24, p. 313-320.

Glover, A. G., A. J. Gooday, D. M. Bailey, D. S. M. Billett, P. Chevaldonne, A. Colaco, J. Copley, D. Cuvelier, D. Desbruyeres, V. Kalogeropoulou, M. Klages, N. Lampadariou, C. Lejeusne, N. Mestre, G. L. J. Paterson, T. Perez, H. Ruhl, J. Sarrazin, T. Soltwedel, E. H. Soto, S. Thatje, A. Tselepides, S. Van Gaever, and A. Vanreusel, 2010, Temporal change in deep-sea benthic ecosystems: a review of the evidence from recent time-series studies: Advances In Marine Biology, v. 58, p. 1-95.

Hartmann, M., J. C. Scholten, P. Stoffers, and K. F. Wehner, 1998, Hydrographic structure of the brine-filled deeps in the Red Sea - New results from the Shaban, Kebrit, Atlantis II, and Discovery deeps: Marine Geology, v. 144, p. 311-330.

Häusler, S., M. Weber, C. Siebert, M. Holtappels, B. E. Noriega-Ortega, D. De Beer, and D. Ionescu, 2014, Sulfate reduction and sulfide oxidation in extremely steep salinity gradients formed by freshwater springs emerging into the Dead Sea: FEMS Microbiol Ecol., v. 90.

Henneke, E., G. W. Luther, G. J. Delange, and J. Hoefs, 1997, Sulphur speciation in anoxic hypersaline sediments from the Eastern Mediterranean Sea: Geochimica et Cosmochimica Acta, v. 61, p. 307-321.

Ionescu, D., C. Siebert, L. Polerecky, Y. Y. Munwes, C. Lott, S. Häusler, M. B. Ionescu, J. Peplies, F. O. Glöckner, A. Ramette, T. Rödiger, T. Dittmar, A. Oren, S. Geyer, H.-J. Stärk, M. Sauter, T. Licha, J. B. Laronne, and D. de Beer, 2012, Microbial and Chemical Characterization of Underwater Fresh Water Springs in the Dead Sea. PLoS ONE 7(6): e38319. doi:10.1371/journal.pone.0038319: PlosOne, v. 7, p. e38319.

Joye, S. B., V. A. Samarkin, B. N. Orcutt, I. R. MacDonald, K.-U. Hinrichs, M. Elvert, A. P. Teske, K. G. Lloyd, M. A. Lever, J. P. Montoya, and C. D. Meile, 2009, Metabolic variability in seafloor brines revealed by carbon and sulphur dynamics: Nature Geoscience, v. 2, p. 349-354.

Kis-Papo, T., A. R. Weig, R. Riley, D. Peršoh, A. Salamov, H. Sun, A. Lipzen, S. P. Wasser, G. Rambold, I. V. Grigoriev, and E. Nevo, 2014, Genomic adaptations of the halophilic Dead Sea filamentous fungus Eurotium rubrum: Nature Communications, v. 5, p. 3745.

Kvitek, R. G., K. E. Coonlan, and P. J. Iampietro, 1998, Black pools of death: hypoxic, brine-filled ice gouge depressions become lethal traps for benthic organisms in a shallow Arctic embayment: Marine Ecology Progress Series, v. 162, p. 1-10.

MacDonald, I. R., 1992, Sea-floor brine pools affect behavior, mortality, and preservation of fishes in the Gulf of Mexico: lagerstatten in the making?: Palaios, v. 7, p. 383-387.

MacDonald, I. R., W. W. Sager, and M. B. Peccini, 2003, Gas hydrate and chemosynthetic biota in mounded bathymetry at mid-slope hydrocarbon seeps: Northern Gulf of Mexico: Marine Geology, v. 198, p. 133-158.

McMullin, E. R., K. Nelson, C. R. Fisher, and S. W. Schaeffer, 2010, Population structure of two deep sea tubeworms, Lamellibrachia luymesi and Seepiophila jonesi, from the hydrocarbon seeps of the Gulf of Mexico: Deep Sea Research Part I: Oceanographic Research Papers, v. 57, p. 1499-1509.

Milkov, A. V., and R. Sassen, 2001, Economic Geology of the Gulf of Mexico and the Blake Ridge Gas Hydrate Provinces: Gulf Coast Association of Geological Societies Transactions, v. 51, p. 219-228.

Nix, E. R., C. R. Fisher, J. Vodenichar, and K. M. Scott, 1995, Physiological ecology of a mussel with methanotrophic endosymbionts at three hydrocarbon seep sites in the Gulf of Mexico: Marine Biology, v. 122, p. 605-617.

Oren, A., 1993, The Dead Sea - Alive again: Experientia, v. 49, p. 518-522.

Oren, A., 1999, Bioenergetic aspects of halophilism: Microbiology & Molecular Biology Reviews, v. 63, p. 334-348.

Oren, A., 2005, A century of Dunaliella research: 1905-2005, in N. Gunde-Cimerman, A. Oren, and A. Plemenitaš, eds., Adaptation to Life at High Salt Concentrations in Archaea, Bacteria, and Eukarya: Dordrecht, Netherlands, Springer, p. 491-502.

Oren, A., 2015, Halophilic microbial communities and their environments: Current Opinion in Biotechnology, v. 33, p. 119-124.

Oren, A., D. Ionescu, M. Y. Hindiyeh, and H. I. Malkawi, 2008, Microalgae and cyanobacteria of the Dead Sea and its surrounding springs: Israel Journal of Plant Sciences, v. 56, p. 1-13.

Oren, A., and N. Ben Yosef, 1997, Development and spatial distribution of an algal bloom in the Dead Sea: A remote sensing study: Aquatic Microbial Ecology, v. 13, p. 219-223.

Oren, A., G. Bratbak, and M. Heldal, 1997, Occurrence of virus-like particles in the Dead Sea: Extremophiles, v. 1, p. 143-149.

Oren, A., and P. Gurevich, 1995, Dynamics of a bloom of halophilic Archaea in the Dead Sea: Hydrobiologia, v. 315, p. 149-158.

Oren, A., P. Gurevich, D. A. Anati, E. Barkan, and B. Luz, 1995, A bloom of Dunaliella parva in the Dead Sea in 1992: biological and biogeochemical aspects: Hydrobiologia, v. 297, p. 173-185.

Raiswell, R., and D. E. Canfield, 1998, Sources of iron for pyrite formation in marine sediments: American Journal of Science, v. 298, p. 219-245.

Reilly, J. F., I. R. MacDonald, E. K. Biegert, and J. M. Brooks, 1996, Geologic controls on the distribution of chemosynthetic communities in the Gulf of Mexico, in D. Schumacher, and M. A. Abrams, eds., Hydrocarbon Migration and its Near-Surface Expression, American Association of Petroleum Geologists Memoir 66, p. 39-62.

Sass, A. M., H. Sass, M. J. L. Coolen, H. Cypionka, and J. Overmann, 2001, Microbial communities in the chemocline of a hypersaline deep- sea basin (Urania basin, Mediterranean Sea): Applied and Environmental Microbiology, v. 67, p. 5392-5402.

Sassen, R., I. R. MacDonald, A. G. Requejo, N. L. Guinasso, Jr., M. C. Kennicutt, II, S. T. Sweet, and J. M. Brooks, 1994, Organic geochemistry of sediments from chemosynthetic communities, Gulf of Mexico slope: Geo-Marine Letters, v. 14, p. 110-119.

Schroeder, W., S. Brooke, J. Olson, B. Phaneuf, J. McDonough, III, and P. Etnoyer, 2005, Occurrence of deep-water Lophelia pertusa and Madrepora oculata in the Gulf of Mexico, in A. Freiwald, and J. M. Roberts, eds., Cold-Water Corals and Ecosystems: Erlangen Earth Conference Series, Springer Berlin Heidelberg, p. 297-307.

Sheu, D. D., 1987, Sulfur and organic carbon contents in sediment cores from the Tyro and Orca basins: Marine Geology, v. 75, p. 157-164.

Siam, R., G. A. Mustafa, H. Sharaf, A. Moustafa, A. R. Ramadan, A. Antunes, V. B. Bajic, U. Stingl, N. G. R. Marsis, M. J. L. Coolen, S. Mitchell, A. J. S. Ferreira, and H. El Dorry, 2012, Unique Prokaryotic Consortia in Geochemically Distinct Sediments from Red Sea Atlantis II and Discovery Deep Brine Pools: PLoS ONE, v. 7, p. e42872.

Smith, E. B., K. M. Scott, E. R. Nix, C. Korte, and C. R. Fisher, 2000, Growth and Condition of Seep Mussels (Bathymodiolus childressi) at a Gulf of Mexico Brine Pool: Ecology, v. 81, p. 2392-2403.

Steinle, L., K. Knittel, N. Felber, C. Casalino, G. de Lange, C. Tessarolo, A. Stadnitskaia, J. S. Sinninghe Damsté, J. Zopfi, M. F. Lehmann, T. Treude, and H. Niemann, 2018, Life on the edge: active microbial communities in the Kryos MgCl2-brine basin at very low water activity: The ISME Journal.

Thakur, N. K., and S. Rajput, 2011, Exploration of Gas Hydrates - Geophysical Techniques: Berlin, Springer, 281 p.

Wallmann, K., F. S. Aghi, D. Castradori, M. B. Cita, E. Suess, J. Greinert, and D. Rickert, 2002, Sedimentation and formation of secondary minerals in the hypersaline Discovery Basin, eastern Mediterranean: Marine Geology, v. 186, p. 9-28.

Wang, Y., J. T. Li, L. S. He, B. Yang, Z. M. Gao, H. L. Cao, Z. B. Batang, A. Al-Suwailem, and P.-Y. Qian, 2015, Zonation of Microbial Communities by a Hydrothermal Mound in the Atlantis II Deep (the Red Sea): PLoS ONE, v. 10, p. e0140766.

Warren, J. K., 2011, Evaporitic source rocks: mesohaline responses to cycles of “famine or feast” in layered brines, Doug Shearman Memorial Volume, (Wiley-Blackwell) IAS Special Publication Number 43, p. 315-392.

Warren, J. K., 2016, Evaporites: A compendium (ISBN 978-3-319-13511-3): Berlin, Springer, 1854 p.

Yakimov, M. M., V. La Cono, R. Denaro, G. D'Auria, F. Decembrini, K. N. Timmis, P. N. Golyshin, and L. Giuliano, 2007, Primary producing prokaryotic communities of brine, interface and seawater above the halocline of deep anoxic lake L'Atalante, Eastern Mediterranean Sea: ISME J, v. 1, p. 743-755.

Yakimov, M. M., V. La Cono, V. Z. Slepak, G. La Spada, E. Arcadi, E. Messina, B. Mireno, L. S. Monticelli, D. Rojo, C. Barbas, O. V. Golyshina, M. Ferrer, P. N. Golyshin, and L. Giuliano, 2013, Microbial life in the Lake Medee, the largest deep-sea salt-saturated formation. : Scientific reports, 3;

Ziegenbalg, S. B., D. Birgel, L. Hoffmann-Sell, C. Pierre, J. M. Rouchy, and J. Peckmann, 2012, Anaerobic oxidation of methane in hypersaline Messinian environments revealed by 13C-depleted molecular fossils: Chemical Geology, v. 292-293, p. 140-148.


Salt Dissolution (2 of 5); Salt caves

John Warren - Saturday, September 30, 2017


Properties of evaporite karst

At first sight, medium-scale evaporite karst surface landforms, such as dolines, polje-like depressions, subsidence bowls and collapse dolines, appear near-identical to those found in and on carbonate karst (Part 1). Likewise, all the smaller-scale intracavern features, both dissolutional (pipes and cavities) and constructive calcite speleothems), can be preserved as karst zones in and above an evaporite host, even when the formative hydrology becomes inactive and the system is buried (palaeokarst). But, the much higher rates of dissolution of halite and gypsum compared to limestone or dolomite means there are significant differences in rates of formation and cave geometries compared to carbonate caves in nonevaporitic hosts. One cannot simply take models of caves developed by studies of carbonate karst and use them to unreservedly interpret evaporite caves.

Differences reflect the inherently higher solubility and flowability of the host salts at earth surface conditions and mean most evaporite-hosted caves show peculiarities related to higher dissolution rates compared to carbonate karst. For example, one most significant in a vadose halite cave is that evaporation of a near-saturated brine is aided by airflow and means many halite stalactites tend to curve into the direction of air flow (anemolites) rather than the subvertical dripstone mechanism that controls the shape of most calcite stalactites and stalagmites. Forti (2017) is an excellent summary of evaporite speleothems and formative mechanisms in both gypsum and halite caves.

The high solubility of subsurface evaporite beds atop deeply-circulating pressurised and jointed aquifers (confined hydrologies) means deep-phreatic chimneying (vertical shafts) and deep dissolution with upward-stoping are a common deep cavity-forming mechanism in and atop buried salt beds. This, in turn, means high volumes of overburden sediment can be swallowed quickly, especially in suprasalt regions with strong roof spans covered by loose sediment (Figure 1). In such settings, large at-surface natural sinkholes and circular structures (breccia chimneys) can daylight in days or even hours and so constitute significant geohazards (see Parts 4 and 5). This type of rapid point-sourced vertical stoping atop evaporites doesn’t just happen in continental settings. Similar circular structures atop breccia-filled vertical shafts or fault-bound chimneys form beneath deep waters of the Eastern Mediterranean via substratal breaching of the Messinian evaporite, followed by upward stoping through the overlying sediments (Bertoni and Cartwright, 2005).

Worldwide, much of the dissolution action in uplifting salt beds and masses begins well below the watertable. It takes place hundreds of metres or more below the surface at bathyphreatic depths where; 1) halite is leaching, 2) anhydrite is reconverting to gypsum and 3) where caprocks and other dissolution residues are forming atop diapiric salt features. Early dissolution is greatest at contacts between the salt edges and joints or fractures in adjacent aquifers. Thus patterns of jointing or fracturing control the extent and style of caves in a salt host and typically creates high-density maze caves. With halite, the only setting where a NaCl mass makes it to the surface is as an active or recently active diapir crest or namakier. Before its dissolution, the salt matrix hosting a cave-hosting diapiric halite is largely impervious, flowing and re-annealing. Hence, most of the karst action accessible for study in halite caves is vadose and tied to perched water tables; this is most obvious near the margins of namakier salt sheets and located well away from locations of active salt fountains.

If phreatic caves ever do form near to rising salt fountain domes, these early cavities are quickly closed by the pressurised salt flow needed to bring salt to surface. Rather than being telogenetic features, many of the metre-scale blebs of finely-layered gravitationally-aligned laminar halite seen within diapirs intersected in deep salt mines, and floating in a matrix of flow foliated coarse-crystalline halite, are the result of mesogenetic salt precipitating in gas-filled open cavities within the diapir mass (see Salty Matters October 31, 2016). These laminar cavity filling growth-aligned halite textures are not telogenetic, neither are the entrained salt clasts preserving primary (relict) depositional texture of the mother salt. Instead, they indicate the existence of former or present-day gas-filled cavities (N2, CH4 or CO2), which are a hazard occasionally encountered during salt and potash mining, for example in the diapirs of NE Germany (Hedlund, 2012). Their presence and their ability to blow out mine walls show that pressurised gas pockets and associated cavern fill textures are not unusual mesogenetic features in a flowing salt mass. Away from deep mine intersections of mesogenetic gas-filled cavities, most of the shallow cavities and fill textures of a flowing salt body and studied by speleologists are telogenetic and mostly vadose.

Cave-forming processes in a telogenetic salt cave in bedded salt bodies are either vadose or phreatic (both shallow and deep), with many gypsum/anhydrite caves showing textural evidence indicating transit from one realm to the other. At any one time, the transition from phreatic to vadose landforms is tied to depth to watertable. This is seen in the gypsiferous cap karst to diapirs and allochthons of Triassic salt in the Betic Cordillera of Spain (Calaforra and Pulido-Bosch, 1999). The crests of the salt structures are dominated by collapse dolines (vadose), this passes radially out into the belt of solution dolines occupied by seasonal saline lagoons and further out into rim of saline springs that form wherever the watertable intersects the landsurface (phreatic). Some caves pass from phreatic to vadose a number of times in their histories in response to watertable fluctuations tied to varying climate and tectonics (Columbu et al., 2015.

Vadose processes characterise the uppermost part of a karst aquifer and have air in pores above a water surface or perched watertable. Water drainage is free-flowing under gravity; cave passages drain downslope and show strong gravitational orientations (numerous sub-vertical features). Because they form above the watertable, vadose cave walls are subject to surface seepage, evaporation and drying. Speleothems decorating vadose cave walls indicate varying combinations of gravity and airflow and include; stalactites, stalagmites, cave popcorn, helicites and flowstones.

Caves can form in a salt bed on its way down (eogenesis; syndepositional to early burial phreatic), on its way up (evaporite telogenesis, generally begins at bathyphreatic depths) and at the bottom of the bed’s burial history (mesogenetic; e.g., gas-driven bathykarst).

Gypsum caves

Active caves beneath gypsum karst landforms first formed as deep phreatic maze caves and are characterised by dense passage networks with numerous contemporaneously closed loops, typically in bedded evaporites. Vadose cross sections of the same system can be quite large due to the high solubility and relative homogeneity of the host, especially if formed in a thick capstone. Heimkehle cave in the Zechstein anhydrites in the Harz Mountains of Germany has an overall passage length of more than 2 km, with large rooms up to 22 metres high and 65 metres wide. It was large enough to be used in World War II to house a factory manufacturing parts for JU88 aeroplanes (Knolle et al., 2013).

Other gypsum caves, some more than 150-200 km long, are well documented in Miocene gypsum hosts in west Ukraine (Table 1; Figure 2; Klimchouk, 2000; Klimchouk and Andrejchuk, 2003; Andrejchuk and Klimchouk, 2004; Klimchouk, 2007) and the Gypsum Plain of West Texas and New Mexico (Stafford et al., 2008, 2009, 2017) and much of our current understanding of the gypsum karst process comes from these regions.

Gypsum cave walls in the vadose realm are relatively undecorated compared to carbonate caves. But calcitic speleothems can form in a gypsum cave and are well documented, as are alabastrine flowstones and selenite rinds on cave walls, for example; in northwest Texas (McGregor et al., 1963), in Alabaster Cave and others in the Blaine Gypsum in western Oklahoma (Johnson, 1996; Bozeman and Bozeman 2002), and in quarries intersecting the gypsum karst system in the Kirschberg Evaporite member near Fredericksburg, Texas (Warren et al., 1990).

Phreatic gypsum caves

Phreatic caves in bedded evaporites atop artesian aquifers typically grow as upward stoping and branching blind flow loop caverns (phreatic cupolas) and chimneys, which can begin to grow deep below the watertable as bathyphreatic or hypogenic karst. For example, deep artesian systems drive cupola karst, and blind chimney stopes in the Black Hills of Dakota, the Elk Point Basin of Canada and the Optymistychna Cave, western Ukraine (Figures 2, 4a). Deeper in a basin, where bathyphreatic caves are bathed by centripetal mesogenetic crossflows, water flow is slow and phreatic karst is influenced by the escape of H2S- and CO2-rich basinal waters, not necessarily by the confined meteoric head. Fluid flow at these greater depths is driven by pore water gradients that reflect potentiometric variations in temperature, organic maturation, pressure and salinity of basinal waters.


Passages in mesogenetic and telogenetic evaporite-hosted caves tend to first develop near fractures and joints in adjacent aquifers and then expand into maze cave networks. Once growing in the main evaporite mass, some phreatic maze passages can show internal upward-directed switchback gradients independent of jointing in adjacent aquifers. Cave orientations within the evaporite bed and away from the aquifer inflow level are tied to internal inhomogeneities in the host bed such as less soluble or more soluble intrasalt beds and changes in mineral proportions.

Phreatic tubes are the most common passage shape in smooth-walled blind-dissolution pockets and cupolas in bathyphreatic gypsum caves. Tube or channelway shapes range in cross section from near-circular (isotopic dissolution of a soluble host) to elliptical tubes to canyon-shaped keyholes (Figures 3, 4a, b; Klimchouk, 1992, 1996). Ornamentation is minimal where undersaturated crossflow drives the rapid dissolutional breakdown of the cave wall and the resulting passages are smooth, with local scalloped dissolution irregularities. Dense intersecting maze cave networks first form in the early stages of exhumation at the contact between a tight but soluble gypsum/anhydrite unit and a less soluble carbonate or siliciclastic aquifer (Figure 2, 3). This less soluble bed is the supply conduit for groundwater crossflows that dissolve the edges of the initially impervious anhydrite/gypsum bed. The greatest rate of water supply to the dissolving gypsum contact is along joints and fractures in the adjacent aquifer bed (especially with limestone aquifers). Accordingly, the meshwork of caves penetrates the gypsum bed and tends to follow joint and fracture patterns of the adjacent aquifer.

More stagnant phreatic portions of telogenetic CaSO4 cave systems can be saturated and so precipitate isopachous crusts and crystal rinds. Gypsum is the commonplace isopachous precipitate in voids in this setting, while anhydrite tends to dominate in cavities in the deeper mesogenetic realm (Garcia-Guinea et al., 2002). By the time phreatic maze caves are exhumed and enter a vadose (epigene) setting, where they are accessible for speleological study, much of the earlier phreatic ornamentation has already been dissolved by the increasing undersaturated throughflow and cavern interconnection. This is associated with uplift into the more hydrologically-active phreatic realm, which always precedes entry into the vadose realm. (Warren et al., 1990).

Some of the longest and most complex phreatic maze cave systems in the world are found in Miocene gypsum in west Ukraine. Optymistychna Cave, with more than 214 km of surveyed passages, is the world’s longest gypsum cave and the second or third longest cave of any type (Table 1; Figure 2; Klimchouk, 2007). The world’s longest cave, at 550 km, is the carbonate-hosted Mammoth Cave of Kentucky. The West Ukraine region contains the five longest known gypsum caves in the world, accounting for well over half of the total known length of known gypsum caves on Earth. By area and volume, the world’s largest gypsum caves are: Ozernaja (330,000 m2 and 665,000 m3; with 122 km of documented passages it is also the world’s 10th longest), Zoloushka (305,000 m2 and 712,000 m3), followed by Optimisticheskaja Cave (260,000 m2 and 520,000 m3). They are all complex joint-controlled maze caves, formed under confined aquifer conditions that existed from the Pliocene to the Early Pleistocene (that is karstification on the way up - telogenesis or gypsum exhumation, initially focused on the underside of the bed). Their growth patterns indicate upward-transverse phreatic groundwater circulation, with ultimate cavern fusion across the gypsum bed. All these west Ukrainian caves were fed by artesian crossflow in sub-gypsum and supra-gypsum aquifers that were sourced in the Carpathian Mountains.

High rates of dissolution in phreatic gypsum caves, relative to rates of water crossflow, are indicated by bevelled, faceted and “keyhole” cross sections in the Ukrainian caves along with a lack of vadose wall ornamentation (Figures 3a 4; Klimchouk, 1996; Pfeiffer and Hahn, 1976). Keyholes indicate density stratification and convectional circulation in cave-forming waters, with shapes sometimes complicated by lithological discontinuities in the gypsum bed (Figure 3b; Kempe, 1972). Convection in caves is most pronounced where sluggish artesian flow and low flow velocities dominate.

This was the case in the Pliocene to early Pleistocene history of the maze caves of the western Ukraine (Klimchouk and Andrejchuk, 2003). At that time the deeply buried gypsum dissolved via upward growing but blind, phreatic cavities, with a reflux of somewhat denser “spent” waters sinking toward the base of the cave. Spent water was replaced by less dense inflow waters supplied from the lower aquifer (Figure 3a). This set up a natural density-stratified convection, which maintained fresher (less dense) waters near the phreatic cave roof. Upward growing blind caves tend to expand more at their tops driving the transition from subcircular to keyhole caverns along the cave conduit (Figure 4b). After dissolving gypsum and increasing in density, a portion of the “spent” cave water sank all the way back into the underlying aquifer where it once again joined the regional throughflow in the lower aquifer. Once a stoping cave breached the top aquifer water flow direction in the cave was controlled by temperature, pressure and density contrasts between aquifers on either side of the gypsum bed. It seems that post-breach most of the cave water continued to rise through the cave system into the overlying aquifer (Figures 3a, 4) Similar keyhole and cupola morphologies are developed in low flow rate bathyphreatic sulphuric acid caves in carbonate hosts (e.g., early stages in the formation of the Carlsbad and Lechuguilla caverns).

Formation of keyhole passages is not an exclusively phreatic phenomenon in density-stratified gypsum caves. Keyholes in the vadose portions of many telogenetic carbonate caves indicate the transition of the cave passage from phreatic, with circular cross sections, to vadose with deepening drainage slots at the base of the passages (e.g. Calaforra and Pulido-Bosch, 2003). Phreatic stoping, followed by vadose cavern enlargement, probably explains the close correlation between caprock sinkhole distribution and position of underlying vadose passages in Permian gypsum subcrop in the Kungur Cave region in the Russian Urals, where the caprock thickness is a little as 25 m (Figure 5).

Vadose gypsum caves

A lowering of the watertable, either by uplift or climatic change tied to increasing local aridity, converts a former hypogene to meteoric phreatic cave into a vadose cave. In the latter, the walls become ornamented with gypsum and calcite speleothems. This is the recent history of the accessible portions of the gypsum maze caves worldwide, including documented examples in New Mexico, the western Ukraine and Saudi Arabia where the passage into the middle-upper Pleistocene marks the transition from phreatic to vadose in most of the accessible caves (Stafford et al., 2008). It is characteristic also of the very recent history of some natural phreatic caves where water tables were artificially lowered to allow quarrying of gypsum (Warren et al., 1990; Klimchouk, 2012).

Climate shifts or watertable fluctuations at the early end of the burial cycle (salt on the way down) can create alternating vadose-phreatic conditions in evaporite beds in the early stages of burial and so create an early watertable-associated karst level in the accumulating evaporites. That is, a gypsum bed hosting a vadose hydrology on its way down into burial, may pass through the watertable a number of times before its final burial and passage into the mesogenetic realm. This is the case today in the captured recharge playas in central Australia and about the edges of some halite-filled salars in the Andes and many Canadian salt lakes (Last, 1993; Warren, 2016), where fluctuating hydrologic conditions alternate between vadose and phreatic. Similarly, Quaternary climate shifts have variably karstified the gypsiferous sediments of many Sinic playas. For example, Yaoru and Cooper (1997) document Pleistocene lake basins in north-west China, such as in the Chaidamu Basin, where exposed gypsum beds evidence karst overprints that include: corroded flutes, fissures, small caves and associated collapse breccias and roof falls, followed by phreatic evaporite cement overprints. Watertable fluctuation is a hydrological overprint that is preserved as alternating ornamented surfaces in gypsum caves, disconformities and cave fills in many ancient lacustrine gypsum units.

Halite Caves

Because of its high solubility, halite does not make it into outcrop or shallow subcrop as easily as gypsum/anhydrite. Where halite is at the surface, it tends to be in regions of Pleistocene halite deposition (salt on its way down) or in zones of active diapirism (salt coming up very quickly). Chabert and Courbon (1997) noted caves in ancient rock salt in several regions: Algeria (in diapirs, mostly as vertical to sub-vertical shafts and short caves up to 28shafts0 m long), Chile (diapirs, with caves 250-500 m long), Israel (in Mt Sedom diapir as vadose caves and tube caves several hundred metres in length and subvertical vadose shafts that are metres across), Romania (in diapirs, with caves up to several hundred metres long), Spain (in diapirs, with caves up to 650 m long), Tajikistan (several caves 300 to 2,500 m long and up to 120 m deep), and in the namakiers of Iran and the offshore island relicts where cavern lengths range from several hundred metres to kilometres. The maintenance of landscape elevation, which can be hundreds of metres above the surrounding plains, facilitates the creation of vadose caves in the diapir crest (Table 2).

Halite in an active namakier rises and spreads rapidly, so any karst in an active salt diapir tends to be a feature associated with the immediate underside of a caprock. Karst processes cannot deeply penetrate while salt is flowing, even when the plug rises more than 300 metres above the surrounds. So, more extensive halite hosted caves are best developed about the margins of a namakier where halite’s susceptibility to rapid dissolution means the length of a cave developed below the caprock can be substantial. There is a report of a single salt halite cave (Cave 3N) on Qeshm Island, Iran, with a passage length of more than 9 km (Bruthans et al., 2002). Once the rate of diapir rise has slowed or ceased, the positive topography of the now inactive diapir controls the depth of development of doline collapse inn the diapir itself. That is, deeper collapse dolines can now form in the more central topographically higher portions of a salt structure, once the rate of rise has slowed (e.g. Calaforra and Pulido-Bosch, 1999).

The high solubility of halite means a halite cave system can form in a few hundred years rather than the thousands to tens of thousands of years needed to form carbonate karst (Bruthans et al., 2010). Not only is rate of cavern formation swift (dissolution/downcutting of 20 mm/year), these rapidly forming karst features are hosted in and atop a flowing rock mass. This means some of our notions of karst process and cave stability, related to the rate and density of cavern expansion, need to be modified when dealing with halite karst. For example, unlike gypsum and carbonate karst, jointing is not significant in controlling cavern orientation in namakier karst.


Dead Sea karst

Halite caves occur in the Mt. Sedom diapir, where halokinetic Miocene salt is sporadically exposed beneath a weathering and fractured gypsiferous caprock (Figure 6a,b; Frumkin, 1996; Frumkin and Ford, 1995; Frumkin 2009). Water enters the various caves in Sedom diapir through breaches in the caprock (Figure 7b). Most of the caves in the higher parts of Mt. Sedom salt are vadose inlet caves; these are meandering steeply inclined tubes and canyon slots located within or immediately below the caprock. They form where salt solution quickly carves out near vertical slots and shafts (typically < 2m broad and much deeper) that lead down from the surface, sometimes along pre-existing fractures and shears in the salt. Inlet caves in the central portions of the mountain can only be accessed through their sinks and appear to have no distinct outlet (Figure 6c). All terminate several tens of metres above the regional watertable (e.g. Karbolot Cave; Figure 6c, d; Frumkin, 1994a, b, 1996).

The lower parts of inlet caves often contain steep silt and clay banks with surge marks that indicate occurrence of low energy water ponding, with variable residence times. Silt and clay sediments settling at the bottoms of inlet caves impede infiltration, extending the residence time of pond water (Frumkin, 1994a, 1996). Three of the studied caves in northern Mount Sedom had perennial ponds throughout the period 1984-1995. The ponds are perched, without any lithologic control, tens of metres above the nearest potential outlet at the foot of the mountain (Figure 6c). The water level in each pond differs from the others by tens of metres. All pond waters are highly concentrated, up to 324 g/l, with solutes consisting mainly of sodium and chlorine. Fresh inflow waters reach halite saturation within a few hours of reaching the pond. Both dissolution and precipitation features form the pond edges, and their equivalents can be seen on cave walls wherever ponds have dried out. Dissolution is indicated by horizontal notches, which connote density stratification in the ponds when aggressive fresh flood waters are temporarily diluting the upper parts of the pond. Subsequent saturation of holomictic pond waters is indicated by the growth of cm-scale halite crystals on the bottom and sides of the ponds.

Towards the periphery of Mt Sedom, the inlet caves lead down to laterally expanding vadose cave levels that drain onto the Dead Sea plain (Figure 6c,d, 7c). Sedom Cave, is the longest laterally expanding diapir cave, with an aggregate length between two subparallel conduits of 1.8 km. Malham Cave, another large perched and laterally expanding cave, lies a few hundred metres south of Sedom Cave (Figure 6a, c, d; Frumkin, 1996). It has an aggregate passage length of more than 5.5 km and reaches to some 194 m below the landsurface. There is an upper tier of mostly inactive passages and a lower active channel level. 14C dates on fossil wood in the upper cave level shows meteoric waters began to sculpt the upper cave more than 5,500 years ago. Ongoing uplift of Mt Sedom salt means the active channel level in Malham Cave is now downcut some 10-12 metres lower than when it began. Malham Cave passages quickly developed an open outlet through which floodwater escaped directly to the Dead Sea floor, proving that during this period some 4000 years ago rock salt had already risen above region hydrological base level at the Malham outlet point within the eastern escarpment (Figure 6c). Lashelshet Cave, an inlet cave on the highest point on the diapir cross section, has an even older age of more than 7,000 years since cave initiation. Caves in the northern part of Mt Sedom did not begin to form until some 3,000 years later (Frumkin, 1996).

Based on their study of the caves of Mt Sedom, Frumkin and Ford (1995) concluded cave passages develop in two main stages: (1) an early stage characterized by inlet caves with high downcutting rates into the rock salt bed, and steep passage gradients; (2) a mature laterally expanding stage characterized by lower downcutting rates and the establishment of a wider subhorizontal perched stream bed armoured with alluvial detritus. This style of cave tends to develop toward the periphery of the diapir mound. In the mature expanding stage downcutting rates are controlled by the uplift rate of the diapir and changes of the level of the Dead Sea.

Passages may aggrade to create wide flat bevelled passages and slots with thick sediment armoured bases (Figure 7c; Frumkin, 1998). A lack of a consistent phreatic level in the blind bottoms of perched water levels and the presence of the horizontal slots in the lower levels of Sedom Cave means dissolution in both types of caves is largely restricted to times of flooding and perched or backed up freshwater in the vadose zone. This explains the tapering passages of inlet caves and the widespread alternation of armouring and bevelling as well as formation of narrow horizontal meandering slots toward parts of the top of the meandering channel that is now Sedom Cave.

Mass balance calculations in the halite caves of Mt Sedom yield downcutting rates of 0.2 mm s-1 during peak flood conditions, this is about eight orders of magnitude higher than reported rates in any limestone cave stream (Frumkin and Ford, 1995). However, floods have a low recurrence interval in the arid climate of Mount Sedom so that long-term mean downcutting rates are lower: an average rate of 8.8 mm/ year was measured for the period 1986-1991, while Frumkin (2000) estimate the average regional vadose downcutting rate in the Mt Sedom karst region to be 20 mm/year. This is still at least three orders of magnitude higher than rates established for limestone caves and more than able to cope with the rate of supply of the diapiric salt.

The highly impervious nature of halite and its resupply in actively growing diapirs means that, unlike carbonate and gypsum caves, there is no real watertable level to define maximum cave development in a rising salt stem. Rather, the inlet caves are simple dissolution tubes where rainwater has accessed halite and sank until it was saturated and then dissolution stopped until the next flood. Toward the edge of the rising stem, these inlet caves breached the edge of the salt mound and vented their perched groundwaters to the surrounding plain (Figure 6; Sedom and Malham caves). This creates a downcutting and laterally expanding cave system, which is still some metres to tens of metres above the base level of the regional watertable in the surrounding plain. The expanding cave level is dominated by mostly horizontal growth, often with a sediment-armoured floor. It has numerous benches in the walls that probably reflect changes in the hydrological base level.

Dense anastomose cave networks that characterise gypsum caves are not found in the halite caves of Mt Sedom. This reflects the ability of diapiric halite to re-anneal and the fact that all exhumed halite that makes it to the surface is diapiric, not bedded. At-surface halite is not sandwiched between jointed aquifers above and below the dissolving layer. Rather it is a growing mound subject to dissolution at its top and sides.

Away from Mt Sedom, there are active collapse sinkholes and caverns forming in the alluvial fans and clastic aprons that overlie the bedded Quaternary lacustrine halite of the Dead Sea (Figure 8a, b; salt on its way down in the burial cycle). In the sediments around the lakeshore, the pace of karst collapse has accelerated in the last 60 years due to a drastic lowering of the circum-lake watertable and the associated lakeward migration of the saline-fresh water interface (Figure 9). For example, a series of collapse dolines 2-15 m diameter and up to 7 m deep, appeared in 1990 in the New Zohar area. In January 2001 a large sinkhole, some 20 m deep and 30 m wide, cut through the asphalt surface of the main road along the western shore of the Dead Sea. It was opened by the passage of a busload of tourists on their way from Ein Gedi to the Mineral Beach solarium. Existing tourist facilities, such as the Ein Gedi beachside parking, were shut down after the road was damaged and several buildings have since collapsed into sinkholes. Sinkholes have since developed in other areas about the Dead Sea Margin including Qalia, Ein Samar and Ein Gedi. The process began in the southern part of the Dead Sea coast and slowly spread northward along the Israeli coast. Collapse is more localised in the northern and southern regions on the Jordanian side, and across the region, continues to increase in frequency as the sea level falls (Ezersky and Frumkin, 2013).

Three main types of sinkhole or doline fill have been recognized atop the dissolving Holocene salt beds of Mt Sedom; 1) Gravel holes in alluvial fans, 2) mud holes in the intervening bays of laminated clay deposits between fans, and 3) a combination of both types at the front of young alluvial fans where they overlap mud flats. Fossil, relict sinkholes have been observed in the wadi channels cutting into some old alluvial fans, showing this is a natural and ongoing process. While lake levels continue to fall (Figure 9b,c), the potential for subsidence hazards related to karst collapse is ongoing.

Sinkholes and related subsidence have been the focus of much geological study of the halite caves, but Closson et al. (2010) pointed out that an even more significant and ultimately damaging environmental effect of the ongoing water level lowering is the hectometre and larger scale landslides along the retreating shorezone. In the 1990s, international builders created major tourist resorts and industrial plants along the Jordanian and Israeli shore while, during the same period, geological hazards triggered by the level lowering spread out. From the beginning of the year 2000, sinkholes, subsidence, landslides, and river erosion damaged infrastructures more and more frequently: dykes, bridges, roads, houses, factories, pipes, crops, etc. all suffered as a result.

There is evidence of an older set of widespread ground collapses, sinkholes and caves that are tied to an earlier substantial fall in the Dead Sea water level some 4 ka. It may even be that the events described in Genesis 14 in the Christian Bible took place at the time of a substantially lowered sea level. The described battle, which occurred prior to the fall of Sodom and Gomorrah, perhaps took place on the subaerially exposed flats of the Southern Basin of the Dead Sea. The “pits of slime” described in the fall perhaps were solution collapse sinkholes activated by the 4000 ka fall in the Dead Sea water level (Frumkin and Elitzur, 2002).

Halite karst in diapiric Hormuz salt, Middle East

Namakier outcrops in and about the Arabian Gulf range from structures actively extruding salt to those in ruins where salt has not flowed for tens of thousands of years (Figure 9). Likewise the halite caves developed in the namakiers of Iran, or their offshore island counterparts, show a broad range of ages and styles of salt cave development tied to the time since cessation of salt flow (Filippi et al., 2011; Bruthans et al., 2010; Talbot et al., 2009).

Surfaces of actively flowing namakiers on the Iranian mainland are characterised by karren flutes and pinnacles, with numerous small-medium dolines, collapse structures, swallow holes and small caves at their base. As at Mt Sedom, caves tend to be sediment-armoured meandering tube caves or subvertical canyon slots that are centred on joints in the salt beneath a thin suffusion mantle. In contrast, the halite caves in the diapiric cores of the many islands in the Arabian Gulf have a more mature bevelled meandering style with thicker sediment armouring on the cavern floor. Many of these caves breach the retreating edges of former namakiers and salt fountains.

Salt movement in the various diapiric cores of these islands is inactive, or is greatly reduced, compared to the Miocene when these structures were active namakiers. For example, Dragon Breath Cave on Hormuz Island is a linear meander tube cave fed by an ephemeral stream in a shallow valley filling with alluvium (Figure 11; Bosak et al., 1999; Filippi et al., 2011). The surrounding landscape is classic salt karst with numerous depressions, blind valleys, ponors and subrosion sinks. Together they form a highly pockmarked centripetally-ringed topography, which outlines those central parts of the island underlain by shallow subcropping Hormuz salt. The cave itself is one of a number of tube caves exiting about the edge of the zone of diapiric salt. Hosted in steeply dipping diapiric salt, it is around 100 m long with its main passage created by a minor ephemeral stream. Its near flat roof, with an average inclination of 5.4%, is a notable feature and reflects either joint-related dissolutional spalling of the roof or an earlier watertable slot related to backup of a freshened water body.

The current cave passage has cut down a metre or more into earlier cave floor sediments (sediment armour), which contain clasts up to 50 cm in diameter. The cave formed by initial ingress along a linear joint, which was then widened by salt dissolution, so allowing meandering of the stream trace within the salt. It is a cave system very similar to the mature stages of laterally expanding caves in Mt Sedom. The base level of the cave correlates with the surface of a widespread marine terrace, which is now uplifted some 20 metres above sealevel and defines much of the periphery of Hormuz Island. The raising of the marine terrace is related to the ongoing raising of the island via salt flow.

Bruthans et al. (2000, 2010) show that the style of karst landform developed in dissolving diapiric salt in the Arabian Gulf Islands reflects the thickness of the carapace that caps the dissolving salt core (Figure 10). They distinguished four classes of diapir cap, each with a particular association of superficial and underground karst forms, namely: 1) outcropping salt, 2) thin capping (0.5-2 m), 3) capping with moderate thickness (5-30 m), 4) capping with greater thickness (more than 30 m). Cap thickness controls or reflects: 1) the density of recharge points, with high densities of recharge points in the thinner caps; 2) the amount of concentrated recharge which occurs at each recharge point, with suffusion karst characterising thinner caps; 3) the rate of lowering the ground surface atop the salt, with the faster rates of lowering occurring beneath thinner caps, and 4) the amount of load transported by underground flood-streams into cave systems. The volume of sediment load tends to be locally higher and focused beneath the thicker caps, particularly when inflow streams abut the edges of a dissolving salt dome. The thickness of caps atop expanding halite caves does not appear to influence the shape or style of the cave developed within the salt mass; more important seems to be the thickness of cap in the recharge area of the cave and the type of recharge into the salt environment. That is, how much water is passing into the salt and is its flow ongoing or ephemeral?

Halite caves in the relatively mature salt stems of the various islands of the Arabian Gulf, unlike carbonate systems, can swallow and store huge volumes of clastic sediment, volumes that would clog the entrance to a carbonate system. The extreme solubility of halite enables the pace of dissolution/corrosion enlargement in a salt cave to keep pace with large amounts of sediment carried into the cave by external inflows (Figure 11b). Stream sediments arriving at the cave entrance, including boulders, move inside and are trapped within the salt itself. Sediment is not dumped outside the cave entrance, which is the typical situation in blind valley river mouths at carbonate caves (Bruthans et al., 2003). For example, coarse-grained sediment fractions are carried hundreds of metres into the cave by two large intermittent streams entering the upper part of the Ponor Cave (Hormuz Island). The clasts in the resulting intra-cave alluvial fan conglomerates range from cobble of several centimetres up to 1 m diameter boulders; only sand-sized particles make it to the lower part of the same cave.

Caves capable of storing such coarse alluvium within the cavern itself are halite-specific with no equivalent in a carbonate karst terrain. There the boulder-size fraction in the cave itself is the result of roof fall, and almost all stream-borne coarse alluvium is deposited outside the cave.


Evaporite Speleothems

Cave walls in zones of less intense dissolution and stream crossflow are decorated with halite, gypsum and anhydrite speleothems (Figure 12). Halite has a much higher potential to form macro- or mono-crystalline speleothems than calcite and gypsum (Forti 2017). Therefore, in most of the studied halite caves around the world, relatively large euhedral or hopper halite crystals have been observed as in the Iranian and Atacama caves (Forti, 2017; De Waele et al., 2009). The preferred location for these crystals are the pools in the cave entrances, where evaporation is sufficiently low to allow the development of euhedral crystals up to 10 cm in size (Figure 12, 13; Fillipi et al., 2011).

In the Iranian caves halite macrocrystals normally form also along streams, whereas in the Atacama Desert they are completely lacking. This is because they need time to develop and therefore the stream must remain active for at least a few days (Figure 12a; Filippi et al., 2011).

Monocrystalline stalactites are widespread in the Iranian caves; the most common of which are the skeletal forms (Figure 12a; Filippi et al., 2011). An idealised skeletal stalactite normally consists of a central rounded “stalactite” from which, at different heights, three smaller and shorter rounded branches develops, being equally spaced at an angle of 120° (Figure 12b). Moreover, each branch and the central stalactite form an angle of ~70°. These values show that the directions of the central columns and the side twigs correspond to that of the four cube diagonals. Finally, at the end of each twig, there is a small halite crystal with one of its diagonals perfectly coincident with the twig (Figure 12b).

It is therefore evident that the entire structure of these peculiar stalactites consists of a single crystal lattice, albeit with a fractal appearance, and this fact is also confirmed by the presence along the rounded column of evident crystal facets oriented in the same direction (Figure 12b). Air currents and other local perturbing factors may cause a deflection from the theoretical direction of both the main column and of the side twigs (anemolites; Figure 12a). Finally, the rounded structure of the central column and of the external twigs is normally covered by glazy halite suggesting that cycles of deposition and dissolution alternate, while no inner feeding tube is present within the central column.

By these observations, the genesis of these skeletal monocrystalline stalactites is induced as driven by solutions mainly coming from brines and sprays, that then flow via gravity and capillarity only on the external surface of the stalactites. The amount and the composition of these solutions must change in time, becoming sometimes slightly undersaturated, probably during rain falls. The location of these speleothems close to waterfalls, where sprays are easily formed support that interpretation (Figure 13; Filippi et al., 2011; Forti 2017).

Compared to the growth rates of calcite speleothems in carbonate caves the growth rates of evaporite speleothems are phenomenal. Halite stalactites, several metres long and curving into the direction of airflow, formed in the mouth of Dragon Breath cave in a few years rather than millennia needed for carbonate counterparts (Bokacs et al., 1999). In August 1997, a network of numerous halite stalagmites and stalactites blocked the entrance to Dragon Breath Cave. In March 1998 there were no remains of the speleothem meshwork, while in February 1999, the stalagmites had reappeared (Bosak et al., 1999). Similar halite structures occur in the caverns of Mt Sedom and on the wet roofs of some salt mines.

Telogenetic halite deposits forming in namakiers encompass a range of mechanisms and speleothem textures (Figure 13; Filippi et al., 2011): i) via crystallization in/on streams and pools, ii) from dripping, splashing and aerosol water, iii) from evaporation of seepage and capillary water, and iv) other types of evaporative deposits. The following examples of halite textures are distinguished in each of the above-mentioned groups: i) euhedral crystals, floating rafts (raft cones), thin brine surface crusts and films; ii) straw stalactites, macrocrystalline skeletal and hyaline deposits, aerosol deposits; iii) microcrystalline forms (crusts, stalactites and stalagmites, helictites); iv) macrocrystalline helictites, halite bottom fibres and spiders, crystals in fluvial sediments, euhedral halite crystals in rock salt, combined or transient forms and biologically induced deposits. The occurrence of particular forms depends strongly on the environment, in particular on the type of brine occurrence (pool, drip, splashing brine, microscopic capillary brine, etc.), flow rate and its variation, atmospheric humidity, evaporation rate and, in some cases, on the air flow direction. Combined or transitional secondary deposits can be observed if the conditions changed during the deposition. Euhedral halite crystals originate solely below the brine surface of supersaturated streams and lakes.

Macrocrystalline skeletal deposits occur at places with abundant irregular dripping and splashing (i.e., waterfalls, or places with strong dripping from the cave ceilings, etc.). Microcrystalline (fine-grained) deposits are generated by evaporation of capillary brine at places where brine is not present in a macroscopically visible form. Straw stalactites form at places where dripping is concentrated in small spots and is frequently sufficient to assure that the tip of the stalactite will not be overgrown by halite precipitates. If the tip is blocked by halite precipitates, the brine remaining in the straw will seep through the walls and helictites start to grow in some places.

Macrocrystalline skeletal halite deposits and straw stalactites usually grow after a major rain event when dripping is strong, while microcrystalline speleothems are formed continuously during much longer periods and ultimately (usually) overgrow the other types of speleothems during dry periods. The rate of secondary halite deposition is much faster compared to the carbonate karst. Some forms increase more than 0.5 m during the first year after a strong rain event; however, the age of speleothems is difficult to estimate, as they are often combinations of segments of various ages and growth periods alternate with long intervals of inactivity.

Anhydrite forms speleothems in preference to gypsum in those rare parts of a cave with very high salinity waters, But overall gypsum speleothems dominate Some of these gypsum speleothems can be quite large, up to a few metres long.

Unlike halite and gypsum caves, which are rich in halite and gypsum formations respectively, anhydrite caves do not host anhydrite speleothems at all. This is a direct consequence of the CaSO4 – CaSO4.2H2O solubility disequilibrium, which makes the hydrated mineral (gypsum) less soluble than the anhydrous one (anhydrite) at normal cave temperatures, thus totally hindering the development of secondary anhydrite formations. Most of the gypsum produced by hydration replaces anhydrite within the rock structure, and therefore anhydrite does not form any speleothems. Nonetheless, a minor part of this secondary gypsum may develop some small deposits. In the caves of the Upper Secchia Valley, small gypsum crusts and flowstones were observed where condensation water, after dissolving anhydrite, flows over the gallery roof or walls where air currents induce evaporation (Chiesi & Forti, 1988). In the same caves, when per ascensum capillary flow and evaporation are possible, euhedral aggregates of small gypsum crystals may develop on top of rock. The interested reader is referred to a comprehensive paper by Forti, 2017 dealing the great variety of halite and gypsum speleothems.

Anhydrite karst is known from several countries of the world, but in almost all cases it is located at depths that make direct exploration almost impossible (Ford & Williams, 2007). This is the reason why until present, only caves from two locations (South Harz in Germany and Upper Secchia Valley in Italy) were explored and their speleothems studied.

The Upper Secchia Valley anhydrite caves and their chemical deposits were already known when the first monograph on speleothems in gypsum caves was published (Forti, 1996). However, at that time these caves were incorrectly considered as formed in gypsum and therefore their deposits were described along with those hosted in classical gypsum karst.

The genesis and evolution of the German and Italian anhydrite caves are completely different; in fact, the first are hypogenic caves (Kempe, 2014) and lack any natural entrance, whereas the second ones are epigenic and often develop very close to the surface (Malavolti, 1949). Therefore, chemical deposits are different in the two locations and restricted to the peculiar environment that controlled the evolution of the caves.

Leaving aside the widespread secondary gypsum produced by the hydration of the host rock, anhydrite caves are extremely poor in chemical deposits. The lack of minerals in the hypogenic caves is because they were filled with near-stagnant water for most of the time during their development. In the epigenic caves, instead, the absence of cave minerals is mainly attributed to the strong increase in volume caused by hydration of anhydrite (that turns into gypsum), which makes the wall and the ceiling of these cavities extremely fractured. In this latter setting, the rather continuous breakdown normally inhibits the development of even small chemical deposits, which, in any case, are easily washed away by the frequent floods that characterise the Upper Secchia Valley. Despite all these restrictions, the anhydrite caves proved to be interesting not only from a mineralogical point of view, as they host one cave mineral (clinochlore, Chiesi & Forti, 1985) restricted to this environment, but also for the presence of a unique gypsum/anhydrite speleothem, i.e., the huge “leather like sheets” of Barbarossa Cave (Figure 14).



Part 1 and Part 2 of this set of articles dealing with evaporite dissolution emphasise the importance of rapid rates of volume loss in creating a unique set of karst landforms and speleothems. This rapidity creates cavities in a hydrological milieu of contrasting brine salinity and temperature interfaces and permeability contrasts.This inherent association of voids in a setting with abrupt chemical interfaces facilitates the enrichment levels of economic commodities (part 6) and drives rapid bed stoping and foundering that forms zones of significant natural and anthropogenically enhanced geohazards in the landscape (parts 4 and 5).


Alavi, M., 1996, Tectonostratigraphic synthesis and structural style of the Alborz mountain system in northern Iran: Journal of Geodynamics, v. 21, p. 1-33.

Andrejchuk, V. N., and A. B. Klimchouk, 2004, Mechanisms of karst breakdown formation in the gypsum karst of the fore-Ural region, Russia (from observations in the Kungurskaja Cave): Speleogenesis and Evolution of Karst Aquifers 2 (2), available at, (online journal): re-published: Implication of Speleological Studies for Karst Subsidence Hazard Assessment. International Journal of Speleology, Theme issue, N 31 (1-4), 89-114.

Bertoni, C., and J. A. Cartwright, 2005, 3D seismic analysis of circular evaporite dissolution structures, Eastern Mediterranean: Journal of the Geological Society, v. 162, p. 909-926.

Bosak, P., J. I. Bruthans, M. Fillippi, T. Svoboda, and J. Smid, 1999, Karst and salt caves in salt diapirs, SE Zagros Mountains (Iran): Acta Carsologica, v. 28, p. 41-75.

Bozeman, J., and S. Bozeman, 2002, Speleology of gypsum caves in Oklahoma: Carbonates and Evaporites, v. 17, p. 107-113.

Broughton, P. L., and D. Cotterill, 2017, Breccia pipe and sinkhole linked fluidized beds and debris flows in the Athabasca Oil Sands: dynamics of evaporite karst collapse-induced fault block collisions: Bulletin of Canadian Petroleum Geology, v. 65, p. 200.

Bruthans, J., M. Filippi, J. Smíd, and L. Palatinus, 2002, Trí Nahácu (Three Nudes) and Ghar-e Daneshyu caves - The world´s second and fifth longest salt caves: The International Caver 2002, Swindon, U.K., p. 27-36.

Bruthans, J., M. Filippi, M. Zare, Z. Churackova, N. Asadi, M. Fuchs, and J. Adamovic, 2010, Evolution of salt diapir and karst morphology during the last glacial cycle: Effects of sea-level oscillation, diapir and regional uplift, and erosion (Persian Gulf, Iran): Geomorphology, v. 121, p. 291-304.

Bruthans, J., and O. Zeman, 2003, Factors controlling exokarst morphology and sediment transport through caves: Comparison of carbonate and salt karst: Acta Carsologica, v. 32, p. 83-99.

Calaforra, J. M., and A. Pulido-Bosch, 1999, Gypsum karst features as evidence of diapiric processes in the Betic Cordillera, Southern Spain: Geomorphology, v. 29, p. 251-264.

Calaforra, J. M., and A. Pulido-Bosch, 2003, Evolution of the gypsum karst of Sorbas (SE Spain): Geomorphology, v. 50, p. 173-180.

Chabert, C., and P. Courbon, 1997, Atlas de cavités non calcaires du monde: Union Int. Spéléol. au prés Madame Carle, Paris, p. 1-120.

Chiesi, M., and P. Forti, 1985, Tre nuovi minerali per le grotte dell’Emilia-Romagna: Bollettino di Mineralogia e Paleontologia, v. 45, p. 14-18.

Closson, D., N. Abou Karaki, and F. Hallot, 2010, Landslides along the Jordanian Dead Sea coast triggered by the lake level lowering: Environmental Earth Sciences, v. 59, p. 1417-1430.

Columbu, A., J. De Waele, P. Forti, P. Montagna, V. Picotti, E. Pons-Branchu, J. Hellstrom, P. Bajo, and R. Drysdale, 2015, Gypsum caves as indicators of climate-driven river incision and aggradation in a rapidly uplifting region: Geology, v. 43, p. 539.

Davison, I., D. Bosence, G. I. Alsop, and M. H. Al-Aawah, 1996, Deformation and sedimentation around active Miocene salt diapirs on the Tihama Plain, northwest Yemen: Geological Society, London, Special Publications, v. 100, p. 23-39.

De Waele, J., P. Forti, V. Picotti, E. Galli, A. Rossi, G. A. Brook, L. Zini, and F. Cucchi, 2009, Cave deposits in Cordillera de la Sal (Atacama, Chile), v. 2, GeoActa, Special Publication, 97-111 p.

Demercian, S., P. Szatmari, and P. R. Cobbold, 1993, Style and pattern of salt diapirs due to thin-skinned gravitational gliding, Campos and Santos basins, offshore Brazil: Tectonophysics, v. 228, p. 393-423.

Ezersky, M., and A. Frumkin, 2013, Fault - Dissolution front relations and the Dead Sea sinkhole problem: Geomorphology, v. 201, p. 35-44.

Filippi, M., J. Bruthans, L. Palatinus, M. Zare, and N. Asadi, 2011, Secondary halite deposits in the Iranian salt karst: general description and origin: International Journal of Speleology, v. 40, p. 141-162.

Ford, D., and P. D. Williams, 2007, Karst hydrology and Geomorphology: New York, John Wiley and Sons, Ltd, 562 p.

Forti, P., 1996, Speleothems and cave minerals in gypsum caves: International Journal of Speleology, v. 25, p. 91-104.

Forti, P., 2017, Chemical deposits in evaporite caves: an overview: International Journal of Speleology, v. 46, p. 109-135.

Forti, P., and M. Chiesi, 1995, A proposito di un particolare tipo di calcite flottante osservato nella Grotta Grave Grubbo - CB 258 (Verzino, Calabria): Atti e Memorie della Commissione Grotte Boegan, v. 32, p. 43-53.

Frumkin, A., 1994a, Hydrology and denudation rates of halite karst: Journal of Hydrology, v. 162.

Frumkin, A., 1994b, Morphology and development of salt caves: National Speleological Society, Bulletin, v. 56, p. 82-95.

Frumkin, A., 1996, Determining the exposure age of a karst landscape: Quaternary Research, v. 46, p. 99-106.

Frumkin, A., 1998, Salt cave cross-sections and their paleoenvironmental implications: Geomorphology, v. 23, p. 183-191.

Frumkin, A., 2000, Speleogenesis in salt – the mount Sedom area, Israel, in A. V. Klimchouk, D. C. Ford, A. N. Palmer, and W. Dreybrodt, eds., Speleogenesis; Evolution of Karst Aquifers, Speleological Society of America, Huntsville, AL, p. 443-451.

Frumkin, A., and D. C. Ford, 1995, Rapid entrenchment of stream profiles in the salt caves of Mount Sedom, Israel: Earth Surface Processes & Landforms, v. 20, p. 139-152.

Garcia-Guinea, J., S. Morales, A. Delgado, C. Recio, and J. M. Calaforra, 2002, Formation of gigantic gypsum crystals: Journal of the Geological Society, v. 159, p. 347-350.

Hedlund, F. H., 2012, The extreme carbon dioxide outburst at the Menzengraben potash mine, 7 July 1953: Safety Science, v. 50, p. 537-553.

Jackson, M. P. A., D. G. Roberts, and S. Snelson, 1995, Salt tectonics, v. 65, American Society Petroleum Geologists Memoir, 454 p.

Johnson, K. J., 1996, Gypsum karst in the United States: International Journal of Speleology, v. 25, p. 183-193.

Kempe, S., 1972, Cave genesis in gypsum with particular reference to underwater conditions: Cave Science, v. 49, p. 1-6.

Kempe, S., 2014, How deep is hypogene? Gypsum caves in the South Harz: Karst Water Institute Special Publication, v. 18, p. 48-57.

Klimchouk, A., 2007, Hypogene Speleogenesis: Hydrological and morphogenetic perspective: Carlsbad, New Mexico, National Cave and Karst Research Institute.

Klimchouk, A., 2012, Ukraine Giant Gypsum Caves, in W. B. White, and D. C. Culver, eds., Encyclopedia of Caves (Second Edition): Amsterdam, Academic Press, p. 827-833.

Klimchouk, A. B., 1992, Large gypsum caves in the western Ukraine and their genesis: Cave Science, v. 19, p. 3-11.

Klimchouk, A. B., 1996, Gypsum karst of the world: Journal of Speleology, v. 25.

Klimchouk, A. B., 2000, Speleogenesis of great gypsum mazes in the Western Ukraine, in A. B. Klimchouk, D. Ford, A. N. Palmer, and W. Dreybrodt, eds., Speleogenesis: Evolution of karst aquifers: Huntsville, Natl. Speleol. Soc., p. 261-273.

Klimchouk, A. B., 2003, Conceptualisation of speleogenesis in multi-storey artesian systems: a model of transverse speleogenesis: Speleogenesis and Evolution of Karst Aquifers, The Virtual Scientific Journal,, 18 pages, v. 1.

Klimchouk, A. B., and V. N. Andrejchuk, 2003, Karst breakdown mechanisms from observations in the gypsum caves of the Western Ukraine: Implications for Subsidence Hazard Assessment, Speleogenesis and Evolution of Karst Aquifers, v. 1, The Virtual Scientific Journal:, p. 20 pages.

Knolle, F., S. Kempe, and L. E. P. Travassos, 2013, Nazi military use of German caves, Dr Benno Wolf and the world cave registry project: Conference: 16th International Congress of Speleology, Brno, July 21 -28, 2013, At Brno, Czech Republic, Volume: 1: 65-70.

Kulke, H., 1978, Tektonik und Petrographie einer Salinarformation am Beispiel der Trias des Atlassystems (NW-Afrika), in W. Zeil, ed., Geotektonische Forschungen, v. 55, p. 1-158.

Last, W. M., 1993, Salt dissolution features in saline lakes of the northern Great Plains, western Canada: Geomorphology, v. 8, p. 321-334.

Liro, L. M., 1992, Distribution of shallow salt structures, lower slope of the northern Gulf of Mexico, USA: Marine and Petroleum Geology, v. 9, p. 433-451.

Malavolti, F., 1949, Memoria del Comitato Scientifico Centrale del Club Alpino Italiano: Morfologia carsica del Trias gessoso-calcareo nell’Alta Valle del Secchia v. 1, p. 129-225.

McGregor, D. R., E. C. Pendery, and D. L. McGregor, 1963, Solution caves in gypsum, north central Texas: Journal of Geology, v. 71, p. 108-115.

Morley, C. K., B. Kongwung, A. A. Julapour, M. Abdolghafourian, M. Hajian, D. Waples, J. Warren, H. Otterdoom, K. Srisuriyon, and H. Kazemi, 2009, Structural development of a major late Cenozoic basin and transpressional belt in central Iran: The Central Basin in the Qom-Saveh area: Geosphere, v. 5, p. 325-362.

Palmer, A. N., and M. V. Palmer, 2000, Hydrochemical interpretation of cave patterns in the Guadalupe Mountains, New Mexico: Journal of Cave and Karst Studies, v. 62, p. 91-108.

Pfeiffer, D., and J. Hahn, 1976, Karst of Germany, in M. Herak, and V. T. Stringfield, eds., Karst: Important karst regions of the Northern Hemisphere: Amsterdam, Elsevier, p. 189-223.

Purser, B. H., 1973, Sedimentation around bathymetric highs in the southern Persian Gulf, in B. H. Purser, ed., The Persian Gulf: Holocene carbonate sedimentation and diagenesis in a shallow epicontinental sea: New York, Springer Verlag, p. 157-177.

Rezak, R., 1985, Local carbonate production on a terrigenous shelf: Gulf Coast Association of Geological Societies Transactions, v. 35, p. 477-483.

Salameh, E., and H. El-Naser, 2000, Changes in the Dead Sea Level and their Impacts on the Surrounding Groundwater Bodies: Acta hydrochimica et hydrobiologica, v. 28, p. 24-33.

Stafford, K. W., W. A. Brown, T. Ehrhart. Jon, A. F. Majzoub, and J. D. Woodard, 2017, Evaporite karst geohazards in the Delaware Basin, Texas: review of traditional karst studies coupled with geophysical and remote sensing characterization: International Journal of Speleology, v. 46, p. 169-180.

Stafford, K. W., L. Land, and G. Veni, 2009, Advances in hypogene karst studies National Cave and Karst Research Institute, Carlsbad, NM (NCKRI Symposium 1), Carlsbad, NM.

Stafford, K. W., R. Nance, L. Rosales-Lagarde, and P. J. Boston, 2008, Epigene and Hypogene Gypsum karst manifestations of the Castile formation: Eddy County, new Mexico and Culbesron County, Texas, USA: International Journal of Speleology, v. 37, p. 83-98.

Talbot, C. J., and V. Pohjola, 2009, Subaerial salt extrusions in Iran as analogues of ice sheets, streams and glaciers: Earth-Science Reviews, v. 97, p. 155-183.

Warren, J. K., 2016, Evaporites: A compendium (ISBN 978-3-319-13511-3): Berlin, Springer, 1854 p.

Warren, J. K., K. G. Havholm, M. R. Rosen, and M. J. Parsley, 1990, Evolution of gypsum karst in the Kirschberg Evaporite Member near Fredericksburg, Texas: Journal of Sedimentary Petrology, v. 60, p. 721-734.

Wu, S., A. W. Bally, and C. Cramez, 1990b, Allochthonous salt, structure and stratigraphy of the north- eastern Gulf of Mexico. Part II: structure: Marine and Petroleum Geology, v. 7, p. 334-370.

Wu, S., P. R. Vail, and C. Cramez, 1990a, Allochthonous salt, structure and stratigraphy of the north-eastern Gulf of Mexico. Part I: stratigraphy: Marine and Petroleum Geology, v. 7, p. 318-333.

Yaoru, L., and A. H. Cooper, 1997, Gypsum karst geohazards in China, in B. F. Beck, and J. B. Stephenson, eds., Engineering Geology and hydrogeology of Karst Terrains: Proceedings of the Sixth Multidisciplinary Conference on Sinkholes and the Engineering and Environmental Impacts of Karst Springfield, Missouri, 6-9 April 1997, Balkema, Rotterdam, p. 117-126.

Yechieli, Y., M. Abelson, A. Bein, O. Crouvi, and V. Shtivelman, 2006, Sinkhole "swarms" along the Dead Sea coast: Reflection of disturbance of lake and adjacent groundwater systems: Geological Society of America Bulletin, v. 118, p. 1075-1087.


Recent Posts


sepiolite anthropogenically enhanced salt dissolution ozone depletion Realmonte potash SedEx DHAB Bathymodiolus childressi cryogenic salt McArthur River Pb-Zn CO2: albedo sinkhole doline Prairie Evaporite causes of major extinction events Density log allo-suture trona Dallol saltpan Ripon Sumo Zabuye Lake North Pole Crescent potash NaSO4 salts gas outburst 13C halite DHAL kainitite Hadley cell: salt karst K2O from Gamma Log RHOB Lop Nur gem MOP perchlorate salt ablation breccia Five Island salt dome trend Magdalen's Road nuclear waste storage hydrogen organic matter Schoenite phreatomagmatic explosion Stebnyk potash 18O enrichment gypsum dune Calyptogena ponderosa base metal salt mine carbon cycle Warrawoona Group climate control on salt well log interpretation 18O eolian transport knistersalz African rift valley lakes anomalous salt zones dissolution collapse doline Kalush Potash Gamma log wireline log interpretation tachyhydrite Pilbara well blowout Red Sea Stebnik Potash vadose zone Phaneozoic climate meta-evaporite potash ore lithium carbonate geohazard supercontinent sulfate End-Triassic freefight lake collapse doline Belle Plain Member snake-skin chert Ethiopia intersalt natural geohazard saline giant chert salt leakage, dihedral angle, halite, halokinesis, salt flow, Hyperarid Seepiophila jonesi magadiite cauliflower chert Neutron Log epsomite Atlantis II Deep HYC Pb-Zn sulphate Hell Kettle flowing salt MVT deposit LIP salt tectonics Koeppen Climate Quaternary climate Kara bogaz gol lunette Archean evaporite-metal association anthropogenic potash saline clay seal capacity Hadley Cell Messinian Karabogazgol Patience Lake member mass die-back subsidence basin bischofite Lop Nor Dead Sea karst collapse circum-Atlantic Salt Basins dark salt GR log extrasalt recurring slope lines (RSL) Platform evaporite lithium battery brine lake edge High Magadi beds Ure Terrace Belle Isle salt mine SOP alkaline lake halocarbon Turkmenistan vanished evaporite Sulphate of potash Ingebright Lake Mesoproterozoic silica solubility namakier bedded potash Musley potash hydrothermal potash hectorite astrakanite Boulby Mine CaCl2 brine ancient climate Large Igneous Magmatic Province Neoproterozoic Oxygenation Event potash dihedral angle rockburst Lake Magadi nacholite Great Salt Lake 13C enrichment basinwide evaporite halophile deep meteoric potash stevensite mine stability carbon oxygen isotope cross plots jadarite vestimentiferan siboglinids MgSO4 depleted End-Cretaceous stable isotope black salt Deep seafloor hypersaline anoxic lake hydrohalite Catalayud Danakhil Depression, Afar solar concentrator pans mirabilite Noril'sk Nickel hydrological indicator oil gusher silicified anhydrite nodules salt periphery End-Permian Deep halite-hosted cave halogenated hydrocarbon Zaragoza causes of glaciation capillary zone endosymbiosis evaporite karst waste storage in salt cavity marine brine carnallitite Koppen climate lot's wife brine pan Precambrian evaporites lithium brine well logs in evaporites blowout crocodile skin chert potash ore price Mulhouse Basin CO2 mummifiction Weeks Island salt mine evaporite Jefferson Island salt mine hydrothermal karst sedimentary copper gas in salt brine evolution water on Mars salt seal venice water in modern-day Mars Badenian lapis lazuli SO2 Lomagundi Event methanogenesis intrasalt seawater evolution Lake Peigneur Muriate of potash Evaporite-source rock association auto-suture zeolite source rock Pangaea methane Proterozoic methanotrophic symbionts deep seafloor hypersaline anoxic basin sulfur Lamellibrachia luymesi halotolerant gassy salt York (Whitehall) Mine lazurite phreatic explosion MgSO4 enriched authigenic silica salt trade evaporite-hydrocarbon association Paleoproterozoic Oxygenation Event sulphur Salar de Atacama Clayton Valley playa: palygorskite Dead Sea saltworks antarcticite salt suture Dead Sea caves Mega-monsoon Thiotrphic symbionts evaporite dissolution Corocoro copper NPHI log sodium silicate solikamsk 2 sinjarite H2S Neoproterozoic halokinetic nitrogen