Geomorphology 228 (2015) 526–535

Contents lists available at ScienceDirect

Geomorphology journal homepage: www.elsevier.com/locate/geomorph

Genesis of folia in a non-thermal epigenic cave (Matanzas, Cuba) Ilenia Maria D'Angeli a,⁎, Jo De Waele a,b, Osmany Ceballo Melendres c, Nicola Tisato d,e,1,2, Francesco Sauro a,b, Esteban Ruben Grau Gonzales e,f,2, Stefano M. Bernasconi g, Stefano Torriani h, Tomaso R.R. Bontognali g a

Department of Biological, Geological and Environmental Sciences, Bologna University, Via Zamboni 67, 40126 Bologna, Italy Associazione di Esplorazioni Geografiche la Venta, Via Priamo Tron 35/F, 31030 Treviso, Italy Meteorological Center Sancti Spíritus, Comandante Fajardo Final, s/n, Sancti Spíritus, Cuba d University of Toronto, Civil Engineering Department, 35 St. George Street, M5S1A4 Toronto, Canada e La Salle 3D f Comité Espeleológico de Matanzas, Sociedad Espeleologica de Cuba, Cuba g ETH Zurich, Geological Institute, Soneggstrasse 5, 8092 Zurich, Switzerland h ETH Zurich, Institut fur Integrative Biologie, Universitätstrasse 2, 8092 Zurich, Switzerland b c

a r t i c l e

i n f o

Article history: Received 20 April 2014 Received in revised form 4 September 2014 Accepted 8 September 2014 Available online 16 September 2014 Keywords: Cave pool speleothems Folia CO2 degassing Evaporation Tropical cave Chemical cave deposits

a b s t r a c t Folia are an unusual speleothem type resembling inverted cups or bracket fungi. The mechanism of folia formation is not fully understood and is the subject of an ongoing debate. This study focuses on an occurrence of folia present in Santa Catalina Cave, a non-thermal epigenic cave located close to Matanzas (Cuba). The sedimentology, morphology, petrology, permeability and geochemistry of these folia have been studied to gain new insight on the processes leading to their development. It is concluded that folia in Santa Catalina Cave formed at the top of a fluctuating water body, through CO2-degassing or evaporation, which may have been enhanced by the proximity to cave entrances. Two observations strongly support our conclusions. (1) When compared to other subaqueous speleothems (e.g. cave clouds) present in the same rooms, folia occur exclusively within a limited vertical interval that likely represents an ancient water level. Folia occur together with calcite rafts and tower cones that developed, respectively, on top of and below the water level. This suggests that a fluctuating interface is required for folia formation. (2) The measured permeability of the folia is too high to trap gas bubbles. Thus, in contrast to what has been proposed in other studies, trapped bubbles of CO2 cannot be invoked as the key factor determining the genesis and morphology of folia in this subaqueous environment. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Folia are relatively rare speleothems resembling inverted rimstone dams (also described as inverted cups, bells, or bracket fungi) growing mostly on overhanging ledges and walls, often covering the bedrock completely (Hill and Forti, 1997). In scientific papers folia have been reported from 31 caves globally, but the number of these locations is likely to increase. To the list of 25 caves reported in Audra et al. (2009), we have added La Baume Cave in France (reported in Davis, 2012), Odelsteinhöhle in Austria (Plan and De Waele, 2011), Cave of the Winds, Colorado, and Fort Stanton Cave, New Mexico, both reported in Davis (2012), Cueva de Villa Luz in Mexico (Hose, 2009), the Sima de la Higuera near Murcia in Spain (Gázquez and Calaforra, 2013), the ⁎ Corresponding author. Tel.: +39 3285879656. E-mail addresses: [email protected] (I.M. D'Angeli), [email protected] (J. De Waele), [email protected] (O.C. Melendres), [email protected] (N. Tisato), [email protected] (F. Sauro), [email protected] (E.R.G. Gonzales), [email protected] (S.M. Bernasconi), [email protected] (S. Torriani), [email protected] (T.R.R. Bontognali). 1 Previously at: ETH Zurich, Geological Institute, Soneggstrasse 5, 8092 Zurich, Switzerland. 2 http://www.lasalle3d.com/.

http://dx.doi.org/10.1016/j.geomorph.2014.09.006 0169-555X/© 2014 Elsevier B.V. All rights reserved.

cavelets of Glenwood Canyon, Colorado (Polyak et al., 2013), and the Santa Catalina Cave near Matanzas in Cuba, subject of this paper. Most of the caves in which folia have been reported are of thermal and/or hypogenic origin, such as Sima de la Higuera (Gázquez and Calaforra, 2013), Devil's Hole in Nevada (Kolesar and Riggs, 2004), Giusti cave in Tuscany, Italy (Piccini, 2000), and the Buda Hills caves (Takacsné Bolner, 1993; Leel-Ossy et al., 2011). However, this is not the rule, as folia have also been described from epigenic-cold caves, such as Hurricane Crawl Cave in California (Davis, 2012) and Odelsteinhöhle in Styria, Austria (Plan and De Waele, 2011). This new occurrence of folia in Santa Catalina Cave appears to be another example of a non-thermal epigenic cave. Many of the reported folia are found in association with subaqueous or water-surface type speleothems forming in supersaturated lakes or pools, such as rafts and cones, and cave mammillaries, and folia are sometimes found associated with typical degassing corrosion morphologies such as bubble trails (Audra et al., 2009). The exact mechanism and the key conditions required for the development of folia are not fully clear, and are the subject of an ongoing debate. Existing hypotheses can be divided into three fundamentally different models: (1) the fluctuating-interface particle-accretion theory of Davis (1997, 2012);

I.M. D'Angeli et al. / Geomorphology 228 (2015) 526–535

(2) the thermal phreatic degassing theory of Green (1991, 1997) or its slightly modified version of hypogenic degassing just below the water level theory (Audra et al., 2009); and, ultimately, (3) the brine-mixing theory proposed by Queen (2009). According to Davis (2012), most carbonate folia form by accretion from adherent particles at a fluctuating water level, and would thus be very reliable water level indicators. For example, folia growing during the rise and fall of the water level (marking this level with a precision of only a few centimeters). Green (1991) and Audra et al. (2009) believe folia form in a subaqueous environment, by CO2-degassing close to but below the water table. The water level is precisely recorded, according to these authors, by the top of the folia zone. Queen (2009) proposes a possible mechanism of folia formation at indefinite depth in the phreatic zone by mixing of fresh water with brine water. Folia would thus indicate the location of the brine/freshwater interface and, as a consequence, would be unrelated to the position of the water table or water level. Each of these genetic theories requires particular conditions and are valid for specific folia occurrences, thus there is not yet a clear and unified theory. This study presents new data on folia that formed in an epigenic nonthermal cave, Santa Catalina Cave, in Cuba. Detailed sedimentological, morphological, petrographical and geochemical analyses, permeability measurements, and SEM examinations were performed on folia in this tropical near-to-the-coast cave. Finally, the possible genesis of these speleothems in this new setting is discussed. 2. The Santa Catalina Cave The Santa Catalina Cave is located approximately 20 km east of Matanzas, on the northern coast of Cuba, in an area known for its beautiful and well-decorated caves (Fig. 1). In 1996 Santa Catalina Cave was

527

declared a National Monument by the Cuban Government for its exceptional speleothems and its historical value. The area is characterized by the outcropping of a series of marine terraces composed of eogenetic limestones dated from Pliocene to present (Ducloz, 1963). The entrances of the cave open on the Yucayo terrace (Lower Pleistocene) at around 20 m asl and less than 1 km from the present coastline. The upper level of the cave is composed of a maze of passages connecting irregular wide and narrow rooms. Based on the general morphology, it can be classified as a flank margin cave, which formed in the mixing zone between fresh and brackish water (Mylroie and Carew, 1990, 2000). Regional uplift and a lowering of sea level have moved this upper sector of the cave out of the mixing zone such that mixing speleogenesis is no longer active. The upper part of the cave hosts speleothems such as stalactites, stalagmites, columns, gours, flowstones, and cave pearls. Besides these conventional morphologies, widespread in many Cuban and tropical caves, Santa Catalina Cave also hosts more exotic speleothems, including cave clouds, tower cones, significant deposits of calcite rafts, folia, and the very rare composite speleothems called “mushrooms” (Fig. 2). The floor of the studied part of the cave is: i) flat, except for some local collapses or flowstones, and the roof is never higher than 6 m; and ii) nearly entirely covered with whitish calcite rafts, which are composed of flakes up to 1 cm wide and less than 1 mm thick (Fig. 2A). These observations indicate that in the past the cave hosted pools with water and that evaporation (or CO2-degassing) at the pool surface was high enough to precipitate calcite. Calcite raft deposits can reach a thickness of N 2 m, especially along the sides of the passages. In the center of the galleries, below dripping points, calcite rafts sank to the bottom of the pools forming typical tower cones, entirely comprised of packed calcite flakes (Fig. 2B). Both calcite rafts and tower cones appear to be eroded away by a water flow along the central parts of the

Fig. 1. Location of Santa Catalina Cave, east of the small village of Carbonera and only 1 km from the northern coast of Cuba. From Google Maps.

528

I.M. D'Angeli et al. / Geomorphology 228 (2015) 526–535

Fig. 2. The exceptional speleothems of Santa Catalina Cave: A. Heap of calcite rafts (white) partly eroded by running water. Note the thin yellowish flowstone with some stalagmites covering and protecting the rafts. B. One of the tower cones cut in half, entirely composed of piled up white calcite rafts (hammer for scale). C. The balconies protruding from the cave walls. D. Detail of C, in which the ribbons on the lower side of the balconies are clearly visible. Also note the cave clouds at the base of the balcony, and the overlapping thickened calcite rafts. E. Folia. Note the highest folia growing directly on the cave clouds, while the lower ones are bigger and piled up one on the other. F. Cave clouds, some covered with subaerial popcorn, other showing signs of condensation-corrosion and some small bubble trails (behind the person). G. Almost 2 m tall mushrooms. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

passages. Locally this erosion removed more than 1 m of calcite rafts, and carved the lowest part of the mushroom stipes and the cave walls (often covered with mammillaries) most exposed to the water flow. The original thickness of the raft deposits is preserved in the most sheltered corners of the passages, farthest away from the erosional flow (Fig. 2C–D).

3. Methods Fieldwork in Santa Catalina Cave was conducted in December 2012, with two explorations of the upper level of this 10-km long cave system. The location and distribution of folia were carefully mapped, especially

I.M. D'Angeli et al. / Geomorphology 228 (2015) 526–535

regarding their elevation range and their position with respect to the entrances of the cave and the overlying topography. One sample of folia, with a diameter of 9.6 cm and a height of 6 cm, was collected from a marginal spot (for esthetic reasons), not visible from the normal pathway in the cave (central isolated folia in Fig. 3C, see also Fig. 4). The sample was cut in half along the vertical axis passing through its center (Fig. 4D). From one of these parts a 10.5 × 6.5 cm thin section was prepared for petrographic analyses. The microscopes used were a Zeiss Axioplan equipped with a Deltapix DP200 camera at the University of Bologna, and an optical system constituted by optiphot and jenoptik instruments at the ETH Zurich. Five small sub-samples (named from 3.1.1 to 3.1.5) were taken from the other half for SEM analyses (see Fig. 5A for location). We used a SEM Jeol JSM-5400 electron microscope, digitized with an iXRF 550i video card and equipped with a Si-drift detector for Energy Dispersive X-ray Spectroscopy at the BIGEA Department at the University of Bologna. Nineteen sub-samples of 100–200 μg were taken with a dental drill from various locations along the folia (see Fig. 5B) for stable isotope analyses. The isotopic composition of carbonate was measured according to the method described in detail in Breitenbach and Bernasconi (2011). Briefly, 100–200 μg aliquots of powder were filled in 12 ml Exetainers

529

(Labco, High Wycombe, UK) and flushed with pure Helium. The samples were reacted with 3–5 drops of 100% phosphoric acid at 70 °C with a ThermoFisher GasBench device connected to a ThermoFisher Delta V mass spectrometer. The average long-term reproducibility of the measurements based on replicated standards was ± 0.05‰ for δ13C and ±0.06‰ for δ18O. The instrument was calibrated with the international standards NBS19 (δ13C = 1.95 and δ18O = −2.2‰) and NBS18 (δ13C = −5.01 and δ18O = −23.01‰). The isotope values are reported here in the conventional delta notation with respect to VPDB (Vienna Pee Dee Belemnite). Four 12-mm diameter cores were drilled from the folia sample; two samples were vertically oriented, one was horizontal and the last one was cored obliquely, forming an angle of ~ 45° to the vertical axis. Sample end-faces were not ground (to keep the boundary permeabilities intact) and as a consequence were irregular. Permeability measurements on the four cores were performed at the Rock Deformation Laboratory of the ETH Zurich by means of a falling head permeameter (Wilson et al., 2000). The device is composed of a high precision graduated burette equipped with a valve hydraulically connected to the sample through a silicon hose, and an electronic chronometer able to evaluate the time elapsed before a pre-defined amount of water has flowed out of the

Fig. 3. Examples of folia. A. One of the niches with abundant folia, some covered with subaerial popcorn, growing directly on cave clouds. B. The center of these cave clouds can sometimes still be seen. C. Folia with its central lower part covered with popcorn like calcite overgrowth: the central isolated folia has been sampled (see Fig. 4). D. A large group of stacked folia developing over a vertical range of 80–100 cm.

530

I.M. D'Angeli et al. / Geomorphology 228 (2015) 526–535

Fig. 4. The sampled folia (S.C. 3.1). A. Oblique view: note the calcite bubble on its flank. B. Bottom view: the cup-shaped lower part of the folia has a continuous overgrowth of popcorn-like calcite. C. Detail of the calcite bubble on the side of the folia. D. The longitudinal cut of the folia through its center. Note the porous texture of the inner part of the speleothem, and the sheet-like calcite along its upper flanks (top right and left). The folia was connected to the roof (cave cloud) at their upper center through a 1 cm diameter stem.

burette through the sample (Fig. 6). The samples were enclosed in heat shrink tubing (i.e. jacket), which is impermeable and ensures a proper coupling between the curved surface of the sample and the jacket itself. Previously, the permeameter was calibrated employing standard samples: three samples with different lengths (19, 31 and 60 mm) composed of alumina oxide (Al2O3) spheres of 2 mm diameter, and a cylindrical alumina oxide sample (which can be considered waterproof). The latter calibration was necessary to evaluate the permeability of the interface between the curved surface of the sample and the internal wall of the jacket. Such permeability reflects the minimum measurable permeability, which, in fact, was much lower (i.e. ~10−10 cm2) than that measured for the folia samples (i.e. 10−6–10−7). Folia samples were prepared by a stack of the following: i) Al2O3 spheres (AlOS), ii) one folia core (~20 mm in length), and iii) again a layer made of AlOS. Such spheres were used to create planar sample end-faces without significantly decreasing the permeability of the stack (K) (Fig. 6). According to the calibration we calculated the permeability of the AlOS (Ka). Moreover,

the lengths of the AlOS (La) and of the folia cores (Ls) were computed accordingly to the total length of the stack (L = La + Ls), the weight of the folia sample and of the stack, and the density of the AlOS. Finally, according to the Darcy law, the folia permeability (Ks) was calculated as: Ks ¼ Ls=ððL=K Þ−ðLa=KaÞÞ:

ð1Þ

4. Results 4.1. Occurrence of folia within Santa Catalina Cave and association with other speleothems Folia are found only in parts of Santa Catalina Cave that are close to the entrances, on overhanging walls or along the roof of apparently blind passages or niches (Fig. 2E). They occur in a well-defined vertical range forming an 80–100 cm thick belt, probably due to tidal variation (Cuba has a microtidal regime with a daily tidal range of b2 m), with

Fig. 5. A. Petrographic zonation of the folia sample: cc = columnar calcite, c + a = calcite and aragonite, cr = calcite rafts, C = macrocrystalline calcite, mc = microcrystalline calcite, pc = popcorn. The stars indicate the samples taken for SEM analyses. B. Location of the samples analyzed for stable isotopes.

I.M. D'Angeli et al. / Geomorphology 228 (2015) 526–535

531

areas near the roof, these mammillaries are corroded, showing a pitted surface, and locally they also show small centimeter-wide and decimeter-long sub-vertical channels, probably produced by CO 2 bubbles creating channels of enhanced corrosion (i.e. bubble trails). Such features were probably produced by condensation corrosion (Fig. 2F). In the lower parts of the galleries, mammillaries display a typical mosaic pattern, clear evidence of an erosion episode. The folia are accompanied by shelfstone-like deposits that contour large rooms. These shelfstone-resembling speleothems can protrude from the walls for over a meter, forming an 80–100 cm thick hanging bench (Fig. 2C–D). Their surface is rough, having a bubbly appearance (i.e. similar to popcorn) and often drapery-like ribbons on their underside (Fig. 2D). The heads of the mushroom-shaped speleothem (Fig. 2G) in Santa Catalina Cave are also located at the same altitude of these balconies (Nuñez Jimenez, 1973).

4.2. Macroscopic morphology

Fig. 6. Scheme of the falling head permeameter and the folia sample (right).

the uppermost folia growing directly on the mammillaries, and the lowermost folia stacked one upon the other (Fig. 7). A nearly horizontal level defines the lower limit below which folia are absent. In contrast, the occurrence of mammillary calcite is not limited to this precise vertical horizon, and extends from the roof right down to the floor, reaching thicknesses of several centimeters. Locally and especially in the higher

The folia present in Santa Catalina Cave show the typical inverted shallow-cup morphology, with a more or less horizontal rim reaching diameters of 20 cm (Fig. 3). In the center of some folia deposits, mammillaries are visible (Fig. 3B) but often their underside is covered with popcorn-like calcite deposits (Fig. 3C–D). The folia sampled for laboratory analysis includes a secondary 2 cm “subfolia”, which developed on its outer (upper) sidewall. This subfolia contained a calcite bubble-like feature on its inside, similar to the ones documented by Audra et al. (2009) (Fig. 4A, C). As described above, the lower concave part of the bell is composed of a knobbly calcite structure resembling popcorn (Fig. 4A–B). Visual inspection of the cross-section of the sample reveals that the central inner part is highly porous, with massive parts of calcite lining millimeter-sized voids with small calcite (and/or aragonite) crystals growing inside (Fig. 4D). This more massive portion of the folia forms a cone-shaped part of the speleothem with its tip pointing upward. On the side of this cone, calcite rafts form an outer shell. The outermost layer of calcite of the folia is more massive and slightly layered. The lower part of the folia, composing the inner surface of the shallow bell, is globular, with compact calcite knobs of up to 2 mm in diameter. This outer calcite is finely layered and massive.

Fig. 7. Overview of one of the central rooms in Santa Catalina Cave. Note the 80–100 cm broad folia band and the presence of mammillaries both above and below the folia. Photo provided by Antonio Danieli, La Salle 3D.

532

I.M. D'Angeli et al. / Geomorphology 228 (2015) 526–535

4.3. Microscopic morphology Under the microscope, four domains presenting different textures can be distinguished in the folia (Fig. 5A). The inner core, making up the oldest part of the folia, is composed of a mixture of relatively well-developed, large aragonite and calcite crystals with less well crystallized areas composed of smaller calcite crystals in between (Fig. 8B). Voids are occupied by newly formed fine, needle-like crystals of aragonite (Fig. 8F). The inner core of the folia is surrounded by a mass of calcite and aragonite rafts, composed of tabular and flaky crystals (Fig. 8C). The outermost 2–3 mm of calcite is macrocrystalline and layered (Fig. 8D). The outer part of the lower area, constituting the knobs, consists of microcrystalline calcite crystals and is likely the youngest layer of the folia (Fig. 8E).

The sampled folia grew directly on mammillary calcite, which was sampled through a drill core after the folia were removed. This mammillary calcite is made out of layered columnar calcite (Figs. 5A, 8A). Electron microscope analyses confirm the observations made with the petrographic microscope (Fig. 9). The lowermost younger calcite deposits (popcorn, pc in Fig. 5A) are composed of macrocrystalline calcite crystals showing evident signs of corrosion (Fig. 9A). In these outer layers of the examined folia, very small needle-shaped crystals of halite occur on the surface (Fig. 9B). The inner part of the folia appears to be composed of a mixture of rhombohedric calcite (Fig. 9C) and prismatic aragonite crystals. A similar association occurs in the calcite raft layers of the speleothem (Fig. 9D–F), where aragonite predominates and calcite crystals are smaller. No microfossils of fossilized extracellular

Fig. 8. Thin sections of folia, all in parallel nicols unless stated otherwise. A. Columnar calcite of the cave cloud on which the folia grows; note the layering. B. Well developed and big aragonite and calcite crystals of the inner, older core. C. Tabular and flaky crystals of cave rafts along the external side of the folia. D. The 2–3-mm thick layered macrocrystalline calcite forming the lateral outer layer of the folia. E. Microcrystalline calcite crystals on the lower side of the folia (crossed nicols). F. Needle-like aragonite crystals in a small vug in the inner core.

I.M. D'Angeli et al. / Geomorphology 228 (2015) 526–535

533

Fig. 9. Scanning electron microscope images of the folia sample. For location of samples 3.1.1–3.1.5 see Fig. 5A. A. Macrocrystalline corroded calcite crystals of the youngest and outermost lower layer (popcorn). B. Tiny salt (NaCl) needle found in the outer and lowermost part of the folia rim. C. Mixture of well-developed calcite crystals, microcrystalline calcite and minor presence of aragonite needles in the inner part of the folia. D. Prismatic aragonite crystals together with microcrystalline calcite in the cave raft zone of the folia. E. Small prismatic aragonite crystals. F. Aragonite crystals grown over macrocrystalline calcite.

polymeric substances suggesting that a biofilm was present at the time the folia formed have been identified within the studied samples. 4.4. Stable isotopes Powders drilled from the various domains of the folia sample have been analyzed for δ13C and δ18O (Fig. 5B and Table 1). Since folia grow from the center to the rim, samples taken near the rim and the uppermost central samples are the youngest and the oldest, respectively. The youngest samples (5C, 5E, 5F, 5G, 55A, 55B, 55G, 55H, and 55I) have values ranging between − 4.01 and −4.52‰ in δ18O and −8.70 and − 10.74‰ in δ13 C, while the oldest ones (5A, 5D, 5H, 55D, and 55F) ranging between − 3.51 and − 3.96‰ in δ18O and − 6.63 and − 8.17‰ in δ13 C. The samples show a clear decrease in both δ18O and δ13C values from the oldest (top) samples toward the younger ones with a strong positive correlation (r2 = 0.92).

4.5. Permeability In order to test the possible mechanism of CO2 trapping in folia, the permeability on four subsamples of the same folia was measured. These subsamples were obtained by drilling at various angles (horizontal, vertical, and at 45° through the central part of the folia). The folia permeability values measured with the falling head permeameter yielded very high values, close to the instrumental limit of the device. Permeability was found to range between 10−6 and 10−7 cm2, similar to coarse sand or fine gravel.

5. Discussion and conclusions Folia in Santa Catalina Cave occur in association with mammillaries, cave rafts, tower cones, balconies, and mushroom-shaped complex

534

I.M. D'Angeli et al. / Geomorphology 228 (2015) 526–535

Table 1 Stable isotope composition of folia.

Youngest calcite

Calcite rafts-aragonite

Oldest calcite

Samples

δ13C

δ18O

Description

55I 5F 55H 5E 55B 55A 55G 5G 5C 55L 55M 5B 55C 55E 5H 5D 55F 55D 5A

−10.74 −10.56 −9.56 −10.16 −10.38 −8.7 −9.63 −10.26 −10.24 −6.11 −7.52 −6.54 −6.03 −7.22 −6.63 −7.98 −8.17 −7.34 −7.21

−4.52 −4.41 −4.25 −4.19 −4.37 −4.01 −4.5 −4.25 −4.27 −3.57 −3.65 −3.39 −3.21 −3.55 −3.51 −3.94 −3.96 −3.77 −3.64

Calcite layers bottom Calcite layers bottom Calcite layers bottom Calcite layers bottom Calcite layers bottom Calcite raft outside top Calcite layers bottom Outside lateral border Outside lateral border Aragonite in internal void Aragonite in internal void Top lateral void Top lateral void Center lateral void Center Center Center Top inside Top

speleothems. All these speleothems are interpreted as mineral concretions that form below or at the surface of supersaturated water pools. Mammillaries are subaqueous speleothems that form close to the water table (Hill and Forti, 1997), and are actively growing in Devil's Hole, Nevada (Szabo et al., 1994; Kolesar and Riggs, 2004). At Devil's Hole, they occur associated with nicely developed folia. The association of mammillary calcite with folia, calcite rafts and small bubble trails is further evidence that they form close to and below the water table. This characteristic has been used to reconstruct past water table position in the Grand Canyon (Polyak et al., 2008). In a cave intercepted by a mine in Sardinia, mammillaries are accompanied by bubble trails and very rare oxidation vents that are corrosional conical holes due to oxidation of sulfides in the core of the mammillaries. Both corrosional morphologies are clear evidence of CO2 degassing close to the water table (De Waele and Forti, 2006). The position of Santa Catalina Cave relative to sea level dictates that the mammillaries grew in shallow phreatic conditions. Mammillaries in Santa Catalina Cave are the earliest phreatic speleothem forming, covering a greater elevation range than folia, tower cones and calcite rafts and often forming the substrate of growing folia (Fig. 7). Most likely, mammillaries formed when the cave was largely underwater, probably during a period of higher sea level, and thus higher water table level. Calcite rafts are formed at the surface of water bodies by strong CO2 degassing or evaporation (Jones, 1989; Taylor et al., 2004). Tower cones in Santa Catalina Cave, formed by piles of rafts sunk by drip water, are evidence for efficient raft formation. Tower cones have been reported in many caves, such as Adaouste cave in Provence (Audra et al., 2002), Giusti cave in Tuscany, Italy (Piccini, 2000), many caves in the Buda Hills in Hungary (Leel-Ossy et al., 2011), Sima de la Higuera near Murcia, Spain (Gázquez and Calaforra, 2013), and of course Lechuguilla cave in New Mexico (Davis, 2000). These towers of calcite rafts can have sides as steep as 80° because the single flakes are cemented by calcite overgrowths. Although they are often found in thermal and/or hypogenic caves, they are not unique to these, having been reported also for example in alpine caves (Wildberger, 1987) or in an epigenic cold water cave in Sardinia (Caddeo et al., 2008). We postulate that the rafts and tower cones in Santa Catalina Cave formed in a period characterized by a lower water table level compared to when the mammillary calcite formed. A drop in water level, related to a transition toward drier conditions or to a drop in sea (and thus water table) level, might have caused transition from phreatic to vadose conditions. This could have increased the water–air surface in the caves and the in-cave airflow, and consequently enhanced evaporation and CO2 degassing. A lower water level in the caves would likely also decrease the relative humidity in the air, leading to more evaporation and CO2 degassing. This would have caused

the formation of these large amounts of calcite rafts in the Santa Catalina Cave. Folia develop directly upon mammillary calcite on overhanging walls or roofs, in an altitudinal range always higher or equal to that of the occurrences of both calcite rafts and tower cones (Fig. 7). This observation suggests that folia formed in the area of vertical oscillation of the water–air interface. The fact that folia develop along an 80–100 cm thick belt on the cave walls, with the uppermost folia smaller and the lowest reaching the biggest sizes, suggests that the water level lowered slowly, followed by an abrupt emptying of the pools or abrupt changes in water chemistry. Santa Catalina folia, rafts, tower cones, and mammillaries all suggest shallow phreatic conditions in an environment conducive to rapid degassing of CO2 and/or evaporation. This interpretation is also consistent with the stable isotope geochemistry of the studied folia. Indeed, the strong co-variation between δ13C and δ18O (Fig. 10) indicates that the isotopic composition of the carbonate is mainly determined by kinetic effects, and that evaporation or rapid degassing of CO2 were important factors for the formation of the folia (Mickler et al., 2006; Lachniet, 2009). It is reasonable to think that the extremely high permeability and the significant porosity of the folia cannot trap CO2. Therefore, we rule out the hypothesis that Santa Catalina's folia formed through the mechanism proposed by Green (1991) and Audra et al. (2009). Nevertheless, CO2 degassing is fundamental for creating the typical supersaturated conditions required for calcite precipitation, but does not allow for the trapping of stable air bubbles underneath the growing folia. The speleothem assemblage and stable isotope values favor rapid CO2 degassing and/or evaporation as a driving mechanism for origin of these speleothems, including the folia. Together, our observations are consistent with the model of Davis (2012) and more specifically with the mechanism proposed by Kolesar and Riggs (2004), suggesting that folia form in the intertidal zone of the cave pool, where water films above the water–air interface, and the uppermost water layer degas CO2 more rapidly, or evaporate, creating the ideal conditions for the

Fig. 10. δ13C and δ18O variation in 18 subsamples from the folia. The samples are visibly ranked in younger and older ones, or calcite raft-secondary aragonite.

I.M. D'Angeli et al. / Geomorphology 228 (2015) 526–535

rapid formation of calcite crystals. Such a process can also induce the incorporation of clay particles in between the calcite flakes, becoming part of the folia structure. Greater evaporation and CO2 degassing on the outer ledges of the folia could be caused, for instance, by the fact that the inner part is shielded from the air flow and thus less prone to evaporation/degassing. Considering such a scenario, precipitation would be faster along the folia rims, increasing the speleothem size outward and conferring the typical cup shape. The presence of multiple entrances and the related high air-flow, the relatively high air temperatures, and the vicinity of the external surface are conditions enhancing the evaporation of the water from the cave pools, the CO2 degassing, and the consequent oversaturation with respect CaCO3 of the solution. Given the short distance of Santa Catalina Cave from the sea and the altitude of the folia, it is reasonable to believe their formation is related to fluctuations of the water level driven by tidal movements during a period in which sea level was probably higher than today. Dating of these speleothems (mammillaries and folia) will be crucial for evaluating this hypothesis. Acknowledgments The speleological expedition was made possible thanks to the experience and organization of La Salle 3D. The SEM analyses were carried out with the help of Prof. Giorgio Gasparotto of BIGEA, Bologna University. Dr. Crisogono Vasconcelos from the Geomicrobiology Laboratory of ETH Zurich is thanked for financial and infrastructural support, while Prof. Ercilio Vento Canosa helped to obtain the necessary permissions for sampling. Scurion GmbH provided the special cave lamps that made life easier underground, especially for cave photographers. All cave pictures are courtesy of La Salle 3D. Special thanks to Antonio Danieli (La Salle 3D) who provided the photograph in Fig. 7. We thank the editor and two anonymous reviewers for their useful comments and corrections that greatly improved the manuscript. References Audra, P., Bigot, J.-Y., Mocochain, L., 2002. Hypogenic caves in Provence (France). Specific features and sediments. Acta Carsologica 31, 33–50. Audra, P., Mocochain, L., Bigot, J.Y., Nobecourt, J.C., 2009. The association between bubble trails and folia: a morphological and sedimentary indicator of hypogenic speleogenesis by degassing, example from Adaouste Cave (Provence, France). Int. J. Speleol. 38, 93–102. Breitenbach, S.F.M., Bernasconi, S.M., 2011. Carbon and oxygen isotope analysis of small carbonate samples (20 to 100 μg) with a GasBench II preparation device. Rapid Commun. Mass Spectrom. 25, 1910–1914. Caddeo, G.A., Caredda, A.M., De Waele, J., Frau, F., 2008. Il ricco patrimonio speleotemico della Grotta di Is Zuddas (Santadi, Sardegna sud-occidentale). XX Congresso Nazionale di Speleologia, Memorie dell'Istituto Italiano di Speleologia, s.II vol. XXI. Federazione Speleologica Sarda, Iglesias, pp. 296–307. Davis, D.G., 1997. Folia in Hurricane Crawl Cave and Crystal Sequoia Cave. San Francisco Bay Chapter Newsletter. Natl. Speleol. Soc. 40 (5).

535

Davis, D.G., 2000. Extraordinary features of Lechuguilla Cave, Guadalupe Mountains. J. Cave Karst Stud. 62, 147–157. Davis, D.G., 2012. In defense of a fluctuating-interface, particle-accretion origin of folia. Int. J. Speleol. 41, 189–198. De Waele, J., Forti, P., 2006. A new hypogean karst form: the oxidation vent. Z. Geomorphol. Suppl.bd 147, 107–127. Ducloz, C., 1963. Etude géomorphologiue de la région de Matanzas, Cuba. Arch. Sci. Génève 16 (2), 351–422. Gázquez, F., Calaforra, J.M., 2013. Origin of double-tower raft cones in hypogenic caves. Earth Surf. Process. Landf. 38, 1655–1661. Green, D.J., 1991. On the origin of the folia and rims. Salt Lake Grotto Technical Note. 88, pp. 182–196. Green, D.J., 1997. The origin of folia. Salt Lake Grotto Technical Note. 96, pp. 51–60. Hill, C.A., Forti, P., 1997. Cave Minerals of the World. National Speleological Society, Huntsville, Alabama, USA (463 pp.). Hose, L., 2009. Recent observations in a remarkably dynamic, sulfide-rich, hypogenic cave in southern Mexico. In: White, W.B. (Ed.), 15th International Congress of Speleology. National Speleological Society, Kerrville, Texas, USA, pp. 1525–1530. Jones, B., 1989. Calcite rafts, peloids, and micrite in cave deposits from Cayman Brac, British West Indies. Can. J. Earth Sci. 26, 654–664. Kolesar, P.T., Riggs, A.C., 2004. Influence of depositional environment on Devil's Hole calcite morphology and petrology. In: Sasowsky, I.D., Mylroie, J.E. (Eds.), Studies of Cave Sediments: Physical and Chemical Records of Paleoclimate. Kluwer/Plenum Academic Press, New York, pp. 227–241. Lachniet, M., 2009. Climatic and environmental controls on speleothem oxygen-isotope values. Quat. Sci. Rev. 28, 412–432. Leel-Ossy, S., Szanyi, G., Suranyi, G., 2011. Minerals and speleothems of the Jozsef-hegy Cave (Budapest, Hungary). Int. J. Speleol. 40, 191–203. Mickler, P.J., Stern, L.A., Banner, J.L., 2006. Large kinetic isotope effects in modern speleothems. Geol. Soc. Am. Bull. 118 (1–2), 65–81. Mylroie, J.E., Carew, J.L., 1990. The flank margin model for dissolution cave development in carbonate platforms. Earth Surf. Process. Landf. 15, 413–424. Mylroie, J.E., Carew, J.L., 2000. Speleogenesis in coastal and oceanic settings. In: Klimchouk, A.B., Ford, D.C., Palmer, A.N., Dreybrodt, W. (Eds.), Speleogenesis: Evolution of Karst Aquifers. National Speleological Society, Huntsville, pp. 226–233. Nuñez Jimenez, A., 1973. Las formaciones fungiformes y su importancia para conocer las fluctuaciones del mar. VIth International Congress on Speleology. Czech Speleological Society, Olomouc, pp. 519–527. Piccini, L., 2000. Il carsismo di origine idrotermale del Colle di Monsummano (Pistoia Toscana). Le Grotte d'Italia V (1), 33–43. Plan, L., De Waele, J., 2011. Folia in der Odelsteinhöhle, Steiermark. Erstnachweis im deutschsprachigen Raum. Die Höhle 62, 54–57. Polyak, V.J., Hill, C.A., Asmerom, Y., 2008. Age and evolution of the Grand Canyon revealed by U–Pb dating of water table-type speleothems. Science 319, 1377–1380. Polyak, V.J., DuChene, H.R., Davis, D.G., Palmer, A.N., Palmer, M.V., Asmerom, Y., 2013. Incision history of Glenwood Canyon, Colorado, USA, from the uranium-series analyses of water-table speleothems. Int. J. Speleol. 42 (3), 193–202. Queen, M., 2009. The lost city: hot springs, mixing and a possible model for folia development. In: White, W.B. (Ed.), 15th International Congress on Speleology. National Speleological Society, Kerrville, Texas, USA, pp. 1650–1656. Szabo, B.J., Kolesar, P.T., Riggs, A.C., Winograd, I.J., Ludwig, K.R., 1994. Paleoclimatic inferences from a 120,000-yr calcite record of water-table fluctuation in Browns Room of Devils Hole, Nevada. Quat. Res. 41 (1), 59–69. Takacsné Bolner, K., 1993. Rare types of carbonate speleothems. Karszt és Barlang 1–2, 29–36. Taylor, M.P., Drysdale, R.N., Carthew, K.D., 2004. The formation and environmental significance of calcite rafts in tropical tufa-depositing rivers of northern Australia. Sedimentology 51, 1089–1101. Wildberger, A., 1987. Höhlenkegel und damit verknüpfte Sinterformationen im Hölloch (Zentralschweiz) sowie daraus abgeleitete Hinweise zur Höhlenentwicklung. 8° Congress National of the Speleological Society, Switzerland, pp. 137–142. Wilson, M.A., Hoff, W.D., Brown, R.J.E., Carter, M.A., 2000. A falling head permeameter for the measurement of the hydraulic conductivity of granular solids. Rev. Sci. Instrum. 71, 3942–3946.