Formation of phreatic caves in an eogenetic karst aquifer by CO2 enrichment at lower water tables and subsequent flooding by sea level rise

Formation of extensive phreatic caves in eogenetic karst aquifers is widely believed to require mixing of fresh and saltwater. Extensive phreatic caves also occur, however, in eogenetic karst aquifers where fresh and saltwater do not mix, for example in the upper Floridan aquifer. These caves are thought to have formed in their modern settings by dissolution from sinking streams or by convergence of groundwater flow paths on springs. Alternatively, these caves have been hypothesized to have formed at lower water tables during sea level low‐stands. These hypotheses have not previously been tested against one another. Analyzing morphological data and water chemistry from caves in the Suwannee River Basin in north‐central Florida and water chemistry from wells in the central Florida carbonate platform indicates that phreatic caves within the Suwannee River Basin most likely formed at lower water tables during lower sea levels. Consideration of the hydrological and geochemical constraints posed by the upper Floridan aquifer leads to the conclusion that cave formation was most likely driven by dissolution of vadose CO₂ gas into the groundwater. Sea level rise and a wetter climate during the mid‐Holocene lifted the water table above the elevation of the caves and placed the caves tens of meters below the modern water table. When rising water tables reached the land surface, surface streams formed. Incision of surface streams breached the pre‐existing caves to form modern springs, which provide access to the phreatic caves. Phreatic caves in the Suwannee River Basin are thus relict and have no causal relationship with modern surficial drainage systems. Neither mixing dissolution nor sinking streams are necessary to form laterally extensive phreatic caves in eogenetic karst aquifers. Dissolution at water tables, potentially driven by vadose CO₂ gas, offers an underappreciated mechanism to form cavernous porosity in eogenetic carbonate rocks. Copyright © 2012 John Wiley & Sons, Ltd.     

Vadose CO2 gas drives dissolution at water tables in eogenetic karst aquifers more than mixing dissolution

Most models of cave formation in limestone that remains near its depositional environment and has not been deeply buried (i.e. eogenetic limestone) invoke dissolution from mixing of waters that have different ionic strengths or have equilibrated with calcite at different _pCO₂ values. In eogenetic karst aquifers lacking saline water, mixing of vadose and phreatic waters is thought to form caves. We show here calcite dissolution in a cave in eogenetic limestone occurred due to increases in vadose CO₂ gas concentrations and subsequent dissolution of CO₂ into groundwater, not by mixing dissolution. We collected high‐resolution time series measurements (1 year) of specific conductivity (SpC), temperature, meteorological data, and synoptic water chemical composition from a water table cave in central Florida (Briar Cave). We found SpC, _pCO₂ and calcite undersaturation increased through late summer, when Briar Cave experienced little ventilation by outside air, and decreased through winter, when increased ventilation lowered cave CO_2(g) concentrations. We hypothesize dissolution occurred when water flowed from aquifer regions with low _pCO₂ into the cave, which had elevated _pCO₂. Elevated _pCO₂ would be promoted by fractures connecting the soil to the water table. Simple geochemical models demonstrate that changes in _pCO₂ of less than 1% along flow paths are an order of magnitude more efficient at dissolving limestone than mixing of vadose and phreatic water. We conclude that spatially or temporally variable vadose CO_2(g) concentrations are responsible for cave formation because mixing is too slow to generate observed cave sizes in the time available for formation. While this study emphasized dissolution, gas exchange between the atmosphere and karst aquifer vadose zones that is facilitated by conduits likely exerts important controls on other geochemical processes in limestone critical zones by transporting oxygen deep into vadose zones, creating redox boundaries that would not exist in the absence of caves. Copyright © 2014 John Wiley & Sons, Ltd.     

Organic carbon inputs,common ions and degassing: rethinking mixing dissolution in coastal eogenetic carbonate aquifers

Caves deliver freshwater from coastal carbonate landscapes to estuaries but how these caves form and grow remains poorly understood. Models suggest fresh and salt water mixing drives dissolution in eogenetic limestone, but have rarely been validated through sampling of mixing waters. Here we assess controls on carbonate mineral saturation states using new and legacy geochemical data that were collected in vertical profiles through three cenotes and one borehole in the Yucatan Peninsula. Results suggest saturation states are primarily controlled by carbon fluxes rather than mixing. Undersaturation predicted by mixing models that rely on idealized end members is diminished or eliminated when end members are collected from above and below actual mixing zones. Undersaturation due to mixing is limited by CO₂ degassing from fresh water in karst windows, which results in calcite supersaturation. With respect to saline groundwater, controls on capacity for mixing dissolution were more varied. Oxidation of organic carbon increased pCO₂ of saline groundwater in caves (pCO₂ = 10^–2.06 to 10^–0.96 atm) relative to matrix porosity (10^–2.39 atm) and local seawater (10^–3.12 atm). The impact of increased pCO₂ on saturation state, however, depended on the geochemical composition of the saline water and the magnitude of organic carbon oxidation. Carbonate undersaturation due to mixing was limited where gypsum dissolution (Cenote Angelita) or sulfate reduction (Cenote Calica) increased concentrations of common ions (Ca²⁺ or HCO₃^?, respectively). Maximum undersaturation was found to occur in mixtures including saline water that had ion concentrations and ratios similar to seawater, but with moderately elevated pCO₂ (Cenote Eden). Undersaturation, however, was dominated by the initial undersaturation of the saline end member, mixing was irrelevant. Our results add to a growing body of literature that suggests oxidation of organic carbon, and not mixing dissolution, is the dominant control on cave formation and enlargement in coastal eogenetic karst aquifers. Copyright © 2016 John Wiley & Sons, Ltd.     

Seasonal δ13CDIC sourcing and geochemical flux in telogenetic epikarst of south-central Kentucky

Telogenetic epikarst carbon sourcing and transport processes and their associated hydrogeochemical responses are complex and dynamic. Carbon dioxide (CO₂) transport rates in the epikarst zone are often driven by hydrogeochemical responses, which influence carbonate dissolution and conduit formation. This study examines the influence of land use on carbon sourcing and carbonate dissolution kinetics through a comparative analysis of separate, but similar, epikarst systems in south-central Kentucky. The use of high-resolution hydrogeochemical data from multiple data loggers and isotope analysis from collected water samples reflects the processes within these epikarst aquifers, which are estimated to contribute significantly to bedrock dissolution. Results indicate that, in an agricultural setting, long-term variability and dissolution is governed by seasonal production of CO₂ . In a more urbanized, shallower epikarst system, land cover may affect CO₂ transport between the soil and underlying bedrock. This concentration of CO₂ potentially contributes to ongoing dissolution and conduit development, irrespective of seasonality. The observed responses in telogenetic epikarst systems seem to be more similar to eogenetic settings, which is suggested to be driven by CO₂ transport occurring independent of high matrix porosity. The results of this study indicate site-specific responses with respect to both geochemical and δ¹³C_DIC changes on a seasonal scale, despite regional geologic similarities. The results indicate that further comparative analyses between rural and urban landscapes in other karst settings is needed to delineate the impact of land use and seasonality on dissolution and carbon sourcing during karst formation processes. © 2019 John Wiley & Sons, Ltd.     

Springflow hydrographs: eogenetic vs. telogenetic karst

Matrix permeability in the range of 10(-11) to 10(-14) m(2) characterizes eogenetic karst, where limestones have not been deeply buried. In contrast, limestones of postburial, telogenetic karst have matrix permeabilities on the order of 10(-15) to 10(-20) m(2). Is this difference in matrix permeability paralleled by a difference in the behavior of springs draining eogenetic and telogenetic karst? Log Q/Q(min) flow duration curves from 11 eogenetic-karst springs in Florida and 12 telogenetic-karst springs in Missouri, Kentucky, and Switzerland, plot in different fields because of the disparate slopes of the curves. The substantially lower flow variability in eogenetic-karst springs, which results in the steeper slopes of their flow duration curves, also makes for a strong contrast in patterns (e.g., "flashiness") between the eogenetic-karst and telogenetic-karst spring hydrographs. With respect to both spring hydrographs and the flow duration curves derived from them, the eogenetic-karst springs of Florida are more like basalt springs of Idaho than the telogenetic-karst springs of the study. From time-series analyses on discharge records for 31 springs and published time-series results for 28 additional sites spanning 11 countries, we conclude that (1) the ratio of maximum to mean (Q(max)/Q(mean)) discharge is less in springs of eogenetic karst than springs of telogenetic karst; (2) aquifer inertia (system memory) is larger in eogenetic karst; (3) eogenetic-karst aquifers take longer to respond to input signals; and (4) high-frequency events affect discharge less in eogenetic karst. All four of these results are consistent with the hypothesis that accessible storage is larger in eogenetic-karst aquifers than in telogenetic-karst aquifers.     

The role of gypsum and/or dolomite dissolution in tufa precipitation: lessons from the hydrochemistry of a carbonate–sulphate karst system

The precipitation of freshwater carbonates (tufa) along karstic rivers is enhanced by degassing of carbon dioxide (CO₂) downstream of karstic springs. However, in most karstic springs CO₂ degassing is not enough to force the precipitation of tufa sediments. Little is known about the role of dissolution of gypsum or dolomite in the hydrochemistry of these systems and how this affects the formation of tufa deposits. Here we present a monitoring study conducted over a year in Trabaque River (Spain). The river has typical karst hydrological dynamics with water sinking upstream and re‐emerging downstream of the canyon. Mixing of calcium–magnesium bicarbonate and calcium sulphate waters downstream of the sink enhances the dissolution of carbonates and potentially plays a positive role in the formation of tufa sediments. However, due to the common‐ion effect, dissolution of dolomite and/or gypsum causes precipitation of underground calcite cements as part of the incongruent dissolution of dolomite/dedolomitization process, which limits the precipitation of tufa sediments. Current precipitation of tufa is scant compared to previous Holocene tufa deposits, which likely precipitated from solutions with higher saturation indexes of calcite (SIcc values) than nowadays. Limited incongruent dissolution of dolomite/dedolomitization favours higher SIcc values. This circumstance occurs when waters with relatively high supersaturation of dolomite and low SO₄^2? composition sink in the upper sector of the canyon. In such a scenario, the process of mixing waters enhances the exclusive dissolution of limestones, preventing the precipitation of calcite within the aquifer and favouring the increase of SIcc values downstream of the springs. Such conditions were recorded during periods of high water level of the aquifers and during floods. This research shows that the common‐ion effect caused by the dissolution of gypsum and/or dolomite rocks can limit [or favour] the precipitation of tufa sediments depending on the occurrence [or not] of incongruent dissolution of dolomite/dedolomitization. Copyright © 2016 John Wiley & Sons, Ltd.     

Carbon dioxide concentration in temperate climate caves and parent soils over an altitudinal gradient and its influence on speleothem growth and fabrics

Carbon dioxide (CO₂) concentrations in caves and parent soils in the Italian Alps have been studied along a 2100 m altitudinal range – corresponding to a mean annual temperature (MAT) range of 12°C – in order to investigate the relationship between MAT, soil pCO₂ and cave air pCO₂, and to test the influence of soil pCO₂ on speleothem growth and fabric to ultimately gain insight into their palaeoclimatic significance in temperate climate settings. Our findings indicate that soil CO₂ is linearly correlated to MAT and its mean annual concentration is described by the equation: soil CO₂ (ppmv) = 1112 + 460 MAT. Soil pCO₂ can also be exponentially correlated to actual evapotranspiration. The pCO₂ in the aquifer is linearly correlated to MAT at the infiltration site and is more influenced by summer soil pCO₂. Cave air CO₂ in the innermost part of the caves exhibits a similar seasonal pattern, and commonly reaches concentrations of about 15% to 35%, with respect to the corresponding soil values, and is exponentially correlated to the MAT at the infiltration site. The combination of these parameters (soil pCO₂, dripwater pCO₂ and cave air pCO₂) results in speleothem growth and controls their fabrics which are typical of four MAT/elevation belts broadly corresponding to the present‐day vegetation zones. In the lower montane zone [100–800 m above sea level (a.s.l.)] speleothems mostly consist of columnar fabric, in the upper montane zone (800–1600 m a.s.l.) both columnar and dendritic fabrics are common, the Subalpine zone (1600–2200 m a.s.l.) is characterized mostly by moonmilk deposits, whereas in the Alpine zone (above 2200 m a.s.l.) no speleothems are forming today. Therefore, fabric changes in fossil speleothem can potentially be used to reconstruct MAT changes in temperate climate karst areas. Copyright © 2015 John Wiley & Sons, Ltd.     

Dissolution in a variably confined carbonate platform: effects of allogenic runoff,hydraulic damming of groundwater inputs,and surface–groundwater exchange at the basin scale

In variably confined carbonate platforms, impermeable confining units collect rainfall over large areas and deliver runoff to rivers or conduits in unconfined portions of platforms. Runoff can increase river stage or conduit heads in unconfined portions of platforms faster than local infiltration of rainfall can increase groundwater heads, causing hydraulic gradients between rivers, conduits and the aquifer to reverse. Gradient reversals cause flood waters to flow from rivers and conduits into the aquifer where they can dissolve limestone. Previous work on impacts of gradient reversals on dissolution has primarily emphasized individual caves and little research has been conducted at basin scales. To address this gap in knowledge, we used legacy data to assess how a gradient of aquifer confinement across the Suwannee River Basin, north‐central Florida affected locations, magnitudes and processes of dissolution during 2005–2007, a period with extreme ranges of discharge. During intense rain events, runoff from the confining unit increased river stage above groundwater heads in unconfined portions of the platform, hydraulically damming inputs of groundwater along a 200 km reach of river. Hydraulic damming allowed allogenic runoff with SI_CAL < ?4 to fill the entire river channel and flow into the aquifer via reversing springs. Storage of runoff in the aquifer decreased peak river discharges downstream and contributed to dissolution within the aquifer. Temporary storage of allogenic runoff in karst aquifers represents hyporheic exchange at a scale that is larger than found in streams flowing over non‐karst aquifers because conduits in karst aquifers extend the area available for exchange beyond river beds deep into aquifers. Post‐depositional porosity in variably confined carbonate platforms should thus be enhanced along rivers that originate on confining units. This distribution should be considered in models of porosity distribution used to manage water and hydrocarbon resources in carbonate rocks. Copyright © 2013 John Wiley & Sons, Ltd.     

Temporal cycles of karst denudation in northwest Georgia,U.S.A.

Time patterns of karst denudation in northwest Georgia (U.S.A.) were investigated at three spring sites for 12 months and at five stream sites for 10 years. Rainfall was evenly distributed and showed no significant seasonality. At the springs, as well as the streams, water hardness was largely controlled by discharge. At the springs, soil pCO₂ and water pH were strongly correlated (r + -0·69 to -0·83). Solute transport in spring waters was highly seasonal, with two conduit flow springs removing more limestone in the winter, and the diffuse flow spring removing more during the growing season. At the stream sites, most denudation occurred during the winter and spring seasons, and least during the summer. Fourier analysis showed that variations in denudation occur on deterministic (long-wave) as well as stochastic (shortwave) time scales. As contributing variables, discharge varied in short-wave and long-wave cycles, whereas soil pCO₂ showed only a long-wave cycle. The 12 month deterministic cycles were the most important, with changes in discharge taking precedence over soil pCO₂. Time series regression explains up to 69 per cent of changes in denudation through rain and soil pCO₂. Time cycles in available water are the key controlling factor of denudation, and amounts of available soil CO₂ may not be as important in the temporal patterns of karst downwearing as has been believed previously.     

The flank margin model for dissolution cave development in carbonate platforms

The Bahama Islands contain many abandoned dissolution caves at elevations between two and seven metres above current sea level. The development of dissolution caves in tropical carbonate islands is dependent on the position and nature of the freshwater lens. Lens position is controlled by sea level, which in stable carbonate platforms like the Bahamas is a function of glacioeustatic sea level still stands. Caves in the Bahamas that are currently subaerial must have developed during past higher sea levels. During the Late Quaternary, sea levels higher than present have been relatively short-lived, and that limits the amount of time that a freshwater lens could be situated at the elevation required for the cave formation. The Bahama Islands are low-lying landforms where only aeolian ridges extend to elevations higher than six metres above current sea level. Past high sea level events greatly reduced the exposed land area of the Bahama Islands, thus also limiting both the catchment for and size of freshwater lenses. Caves must be younger than the rock in which they are developed; most subaerial Bahamian caves are found in limestones that are less than 150000 years old. Development of large dissolution caves under these limitations of time and lens size requires a powerful dissolutional mechanism. The mixing of discharging freshwater with tide-pulsed incoming marine water under the flanks of emergent dune ridges may have produced the conditions necessary. Bahamian caves formed by this process are phreatic chambers with complex interconnections and blind tubes. Their presence demonstrates that significant dissolution can occur rapidly as a result of the mixing of fresh and marine waters beneath small carbonate islands.     

Tidal pumping and biogeochemical processes: Dissolution within the tidal capillary fringe of eogenetic coastal carbonates

Growing evidence suggests microbial respiration of dissolved organic carbon (DOC) may be a principal driver of subsurface dissolution and cave formation in eogenetic carbonate rock. Analyses of samples of vadose zone gasses, and geochemical and hydrological data collected from shallow, uncased wells on San Salvador Island, Bahamas, suggest tidally varying water tables may help fuel microbial respiration and dissolution through oxygenation. Respiration of soil organic carbon transported to water tables generates dysaerobic to anaerobic groundwater, limiting aerobic microbial processes. Positive correlations of carbon dioxide (CO₂), radon-222 (²²²Rn) and water table elevation indicate, however, that tidal pumping of water tables pulls atmospheric air that is rich in oxygen, and low in CO₂ and ²²²Rn, into contact with the tidal capillary fringe during falling tides. Ratios of CO₂ and O₂ in vadose gas relative to the atmosphere indicate this atmospheric oxygen fuels respiration within newly-exposed, wetted bedrock. Deficits of expected CO₂ relative to O₂ concentrations indicate some respired CO₂ is likely removed by carbonate mineral dissolution. Tidal pumping also appears capable of transferring oxygen to the freshwater lens, where it could also contribute to respiration and dissolution; dissolved oxygen concentrations at the water table are at least 5% saturated and decline to anaerobic conditions 1–2 m below. Our results demonstrate how tidal pumping of air to vadose zones can drive mineral dissolution reactions that are focused near water tables and may contribute to the formation of laterally continuous vuggy horizons and potentially caves. © 2020 John Wiley & Sons, Ltd.     

Dissolution-filling mechanism of atmospheric precipitation controlled by both thermodynamics and kinetics

Affected by structural uplift,the Ordovician carbonate rockbed in the Tarim Basin,China,was exposed to dissolution and reformation of atmospheric precipitation many times,and formed a large quantity of karst caves serving as hydrocarbon reservoir.However,drilling in Tahe area showed that many large karst caves,small pores and fractures are filled by calcite,resulting in decrease in their reservoir ability.Calcite filled in the karst caves has very light oxygen isotopic composition and87Sr/86Sr ratio.Its 18OPDB ranges from 21.2‰to 13.3‰with the average of 16.3‰and its87Sr/86Sr ratio ranges from0.709561 to 0.710070 with the average of 0.709843.The isotope composition showed that calcite is related to atmospheric precipitation.Theoretic analyses indicated that the dissolving and filling actions of the precipitation on carbonate rocks are controlled by both thermodynamic and kinetic mechanisms.Among them,the thermodynamic factor determines that the precipitation during its flow from the earth surface downward plays important roles on carbonate rocks from dissolution to saturation,further sedimentation,and finally filling.In other words,the depth of the karstification development is not unrestricted,but limited by the precipitation beneath the earth surface.On the other hand,the kinetic factor controls the intensity,depth,and breadth of the karstification development,that is,the karstification is also affected by topographic,geomorphologic,climatic factors,the degree of fracture or fault,etc.Therefore,subject to their joint effects,the karstification of the precipitation on the Ordovician carbonate rocks occurs only within a certain depth(most about 200 m)under the unconformity surface,deeper than which carbonate minerals begin to sedimentate and fill the karst caves that were formed previously.     

From subsurface to surface: a multidisciplinary approach to decoding uplift histories in tectonically-active karst landscapes

We present an integrated study of subsurface and surficial karst landforms to unravel the uplift history of karst landscape in a tectonically-active area. To this end, we apply a multidisciplinary approach by combining cave geomorphology and Th/U dating of speleothems with remote sensing plus geophysical imaging of surface landforms. We use as an example Mt. Menikio in northern Greece where four caves share well-defined epiphreatic/shallow phreatic characteristics that are related to the distribution of surface and buried doline fields and provide evidence for three distinct water table stillstands (e.g. expressed as cave levels) now lying at ~130 m, ~800 m and ~1600 m a.m.s.l. Our dating constraints delimit the age of the lower water table stillstand prior to 77 ka ago and imply a maximum rate of relative base level drop of 0.45 mma^-1, which is consistent with relative tectonic uplift rate estimates along currently active normal faults. We interpret the elevation of the higher water table stillstands to reflect earlier phases of uplift related to the regional tectonic events associated with the development of the North Anatolian Fault and the Northern Aegean area. Our analysis shows that the combined study of epiphreatic/shallow phreatic caves and surficial karst landforms together, is a robust way to investigate the uplift history of a karst landscape in a tectonically-active setting. © 2019 John Wiley & Sons, Ltd.     

Water and air dynamics within a deep vadose zone of a karst massif: Observations from the Lukina jama–Trojama cave system (−1,431 m) in Dinaric karst (Croatia)

Extreme heterogeneity of karst systems makes them very challenging to study. Various processes within the system affect its global response, usually measured at karst springs. Research conducted in caves provides a unique opportunity for in situ analysis of separate processes in karst underground. The aim of the present study was to research the water and air dynamics within a deep karst system. Air and water basic physical parameters across the Lukina jama–Trojama cave system (?1,431 m) were continuously monitored during a 1‐year period. Recorded hydrograph of the siphon lake at the bottom of the cave was used to interpret the characteristics of an unexplored phreatic/epiphreatic conduit network. Water origin in the siphon was determined based on temperature and electrical conductivity. Air temperature and humidity monitoring revealed a strong inflow of air of sub‐zero temperature into the upper portion of the cave during winter. Cave passage morphology was interpreted as the main determinant of air dynamics, which caused ice to accumulate extensively in the upper portions of the cave and caused the temperature on the top of the homothermic zone to be significantly below the mean outside temperature. Air dynamics also lowered the temperature of water flowing through the cave vadose zone and feeding the phreatic zone of the massif. The pronounced temperature difference between the phreatic zone and the top of the homothermic zone probably contributed to the thermal gradient observed in the cave, which is steeper than in ice‐free caves in the area. Our results enabled the development of a conceptual model that describes coupling between air and water dynamics in the cave system and its surroundings.     

Effects of lithology on solution development in carbonate aquifers

Groundwater flow in karst aquifers is extremely difficult to predict because of the presence of solution-enlarged fractures and conduits in the bedrock. The location and volume of this solution development is the result of interrelated geologic, hydrologic and climatic factors. This paper examines lithologic properties which influence apparent solution rates in carbonate aquifers as indicated by weathered profiles in caverns. Rock samples collected from cavern entrances and walls in southwestern Missouri, U.S.A., were analyzed for grain size, sorting, percent quartz, percent magnesium carbonate, percent opaque minerals and percent acid-insoluble residue. Discriminant function analyses of these measurements show that grain size is the most important parameter in distinguishing the grouping of apparently resistant and non-resistant samples. Small grain size favors dissolution and strongly influences the relief of the weathered profile regardless of rock composition. Significant variability in magnesium carbonate content may also affect the apparent solution rate but to a lesser degree than grain size.     

Rates of dissolution of aluminosilicates in seawater

Dissolution of eight clay minerals, four zeolites, and quartz in seawater has been monitored for81/2 years. For most of the minerals, dissolution can be described as a first-order reaction in which dissolved silica approaches from undersaturation steady concentration values with time. Characteristic reaction rate constants (k₁) are of the order of 10^?7 sec^?1. One of the zeolites, clinoptilolite, shows a different dissolution behavior: SiO₂ concentration in solution reaches a high value within one year, followed by a decline to a lower value, suggestive of precipitation of another silicate phase (possibly sepiolite).A mathematical solution is given for a kinetic equation combining the parabolic-rate and first-order rate processes. It is shown that in a wide range of silicate dissolution reactions taking place over long periods of time, the presence of the parabolic-rate dissolution processes cannot be detected, thereby making its inclusion in the kinetic equations unnecessary. The experimental rates of dissolution are comparable to the SiO₂^? dissolution rates in oceanic sediments near the sediment/water interface. But deeper in the sediment, the calculated dissolution rates are significantly lower than the near-interface and experimental values.     

Late quaternary speleogenesis and landscape evolution in the northern Apennine evaporite areas

Gypsum beds host the majority of the caves in the north‐eastern flank of the Apennines, in the Emilia Romagna region (Italy). More than six hundred of these caves have been surveyed, including the longest known epigenic gypsum cave systems in the world (Spipola‐Acquafredda, ~11 km). Although this area has been intensively studied from a geological point of view, the age of the caves has never been investigated in detail. The rapid dissolution of gypsum and uplift history of the area have led to the long‐held view that speleogenesis commenced only during the last 130 000 years. Epigenic caves only form when the surface drainage system efficiently conveys water into the underground. In the study area, this was achieved after the dismantling of most of the impervious sediments covering the gypsum and the development of protovalleys and sinkholes. The time necessary for these processes can by constrained by understanding when caves were first formed. The minimum age of karst voids can be indirectly estimated by dating the infilling sediments. U–Th dating of carbonate speleothems growing in gypsum caves has been applied to 20 samples from 14 different caves from the Spipola‐Acquafredda, Monte Tondo‐Re Tiberio, Stella‐Rio Basino, Monte Mauro, and Castelnuovo systems. The results show that: (i) caves have been forming since at least ~600 kyr ago; (ii) the peak of speleogenesis was reached during relatively cold climate stages, when rivers formed terraces at the surface and aggradation caused paragenesis in the stable cave levels; (iii) ~200 000 years were necessary for the dismantling of most of the sediments covering the karstifiable gypsum and the development of a surface mature drainage network. Besides providing a significant contribution to the understanding of evaporite karst evolution in the Apennines, this study refines our knowledge on the timescale of geomorphological processes in a region affected by rapid uplifting. Copyright © 2016 John Wiley & Sons, Ltd.     

The role of antecedent groundwater heads in controlling transient aquifer storage and flood peak attenuation in karst watersheds

Transient storage of floodwaters in aquifers is known to attenuate peak flows in rivers and drive subsurface dissolution. Transient aquifer storage could be enhanced in watersheds overlying karst aquifers where caves facilitate surface and groundwater exchange. Few studies, however, have examined controls on, or magnitudes of, transient aquifer storage or flood peak attenuation in karstic watersheds. Here we evaluate flood peak attenuation with multiple linear regression analyses of 10 years of river and groundwater data from the Suwannee River, which flows over the karstic upper Floridan aquifer in north-central Florida and experiences frequent flooding. Regressions show antecedent river stage exerts the dominant control on magnitudes of transient aquifer storage, with recharge and time to peak having secondary controls. Specifically, low antecedent stages result in larger magnitudes of transient aquifer storage and thus greater flood attenuation than conditions of elevated antecedent stage. These findings suggest subsurface weathering, including cave formation and enlargement, caused by transient aquifer storage could occur on a more frequent basis in aquifers where groundwater table elevation is lowered due to anthropogenic or climatic influences. Our work also shows that measures of groundwater table elevation prior to an event could be used to improve predictive flood models. © 2018 John Wiley & Sons, Ltd.     

Hydrogeochemical characteristics of surface water and groundwater in the karst basin,southwest China

Groundwater is a very significant water source used for irrigation and drinking purposes in the karst region, and therefore understanding the hydrogeochemistry of karst water is extremely important. Surface water and groundwater were collected, and major chemical compositions and environmental isotopes in the water were measured in order to reveal the geochemical processes affecting water quality in the Gaoping karst basin, southwest China. Dominated by Ca²⁺, Mg²⁺, HCO₃^? and SO₄^2?, the groundwater is typically characterized by Ca? Mg? HCO₃ type in a shallow aquifer, and Ca? Mg? SO₄ type in a deeper aquifer. Dissolution of dolomite aquifer with gypsiferous rocks and dedolomitization in karst aquifers are important processes for chemical compositions of water in the study basin, and produce water with increased Mg²⁺, Ca²⁺ and SO₄^2? concentrations, and also increased TDS in surface water and groundwater. Mg²⁺/Ca²⁺ molar ratios in groundwater decrease slightly due to dedolomitization, while the mixing of discharge of groundwater with high Mg²⁺/Ca²⁺ ratios may be responsible for Mg²⁺/Ca²⁺ ratios obviously increasing in surface water, and Mg²⁺/Ca²⁺ ratios in both surface water and groundwater finally tending to a constant. In combination with environmental isotopic analyses, the major mechanism responsible for the water chemistry and its geochemical evolution in the study basin can be revealed as being mainly from the water–rock interaction in karst aquifers, the agricultural irrigation and its infiltration, the mixing of surface water and groundwater and the water movement along faults and joints in the karst basin. Copyright © 2009 John Wiley & Sons, Ltd.     

Speleogenesis of an exhumed hydrothermal sulphuric acid karst in Cambrian carbonates (Mount San Giovanni,Sardinia)

In the past few years the systematic study of caves intercepted by mine workings in southwest Sardinia has permitted us to observe morphologies due to rare speleogenetic and minerogenetic processes related to ancient hydrothermal activity. These relic morphologies are slowly being overprinted by recent speleogenetic processes that tend to obscure the hypogene origin of these caves. A combined geomorphological and mineralogical investigation has permitted a fairly detailed reconstruction of the various phases of evolution of these caves. Cave formation had already started in Cambrian times, but culminated in the Carboniferous, when most of the large voids still accessible today were formed. A key role in carbonate dissolution was played by sulphuric acid formed by the oxidation of the polymetallic ores present in the rocks since the Cambrian. During the Quaternary a variety of minerals formed inside the caves: calcite and aragonite, that yielded sequences of palaeo‐environmental interest, and also barite, phosgenite, hydrozincite, hemimorphite and many others. These minerals are in part due to a phreatic thermal hypogenic cave forming phase, and in part to later epigene overprinting in an oxidizing environment rich in polymetallic ores. Massive gypsum deposits, elsewhere typical of this kind of caves, are entirely absent due to dissolution during both the phreatic cave formation and the later epigenic stage. Copyright © 2013 John Wiley & Sons, Ltd.