Description of subsurface features at COM

Of the 60 basaltic flows that compose COM, the young Blue Dragon flow (∼2.1 ka) contains the majority of accessible lava tubes and caves. This high abundance of lava tubes and caves is because of the young age of the flow, as it has had less time to undergo gravitational collapse. As a result, the Blue Dragon flow hosts a majority of the secondary sulphate deposits found at COM; this is probably the result of its morphological characteristics rather than any chemical differences from adjacent COM flows (Stout et al. Reference Stout, Nicholls and Kuntz1994; Richardson et al. Reference Richardson, Hinman, McHenry, Kotler, Knipe and Scott2012).

Three locations within the Blue Dragon flow (Wilderness Caves, Arco Tunnel and Spatter Cones) were chosen for mineral sampling based on mineral occurrence, mineral abundance, cave accessibility and lack of public access. The extent of the Blue Dragon flow and cave locations used in this study is shown in Fig. 1(a). A representative photograph (Fig. 1(b)) shows the absence of surface vegetation at the Blue Dragon flow. The Wilderness Caves area is a collection of roughly six small lava tubes and caves covering an area ∼0.13 km². Approximately 2 km to the west of the Wilderness Caves area is the branching lava tunnel of Arco Tunnel. Arco Tunnel is composed of a series of connected lava tunnels that extend for over 1 km. The final sampling location comes from the two hollow magma chambers located within adjacent spatter cones (Crystal Pit and Snow Cone Pit). These spatter cones are approximately two and half kilometres west of Arco Tunnel on the western boundary of the Blue Dragon flow. These magma chambers are only accessible through narrow <20 m vertical throats (Peck Reference Peck1974). The hollow magma chambers of Crystal and Snow Cone Pit are only two of three such formations accessible in North America known to the authors because similar magma chambers tend to be inaccessible due to lava infill, weathering and/or ice plugs.

Description of secondary deposits

The secondary Na-sulphate minerals were found intermittently dispersed as localized, efflorescent, white and powdery deposits in small cavities on the ceilings and walls as well as mounds on the floors and are described in detail in Richardson et al. (Reference Richardson, Hinman, McHenry, Kotler, Knipe and Scott2012); thus, only a brief summary is provided here. The secondary deposits appear to be seasonal because the size and amount of each mineral deposit varied on a semi-annual basis. The depths of floor deposits ranged between 1 and 10 cm (Fig. 2) and are not considered a direct weathering product of the host basalts because the authigenic mineral-cave floor interface was quite sharp with no mineralogical or gradient into the underlying basalt. Secondary deposits were also found infilling small cavities and cracks on cave ceilings and walls. No correlation was observed between the ceiling and floor deposits.

Fig. 2. Photograph of thenardite deposit in a wall cavity inside Arco Tunnel.

Figure 3(a) shows an SEM image of a secondary deposit. The deposits are primarily composed of small crystals with a few larger, needle-like crystals. Given that organic film fragments have been detected on other evaporite minerals (Galeev et al. Reference Galeev, Vinokurov, Mouraviev and Osin2009), SEM images were acquired of the COM thernadite samples; however, no discernible features suggestive of organic film were observed. The EDX spectrum of the elements present is provided in Fig. 3(b) and shows the presence of Na, S, O, Ca and C. XRD analysis of the sample showed that it is predominantly thenardite when compared with published ICDD data (Fig. 3(c)). The carbon present in the EDX spectrum is probably due to the presence of Na-carbonate species such as burkeite (Richardson et al. Reference Richardson, Hinman, McHenry, Kotler, Knipe and Scott2012).

Fig. 3. (a) SEM image of efflorescent secondary mineral deposit. (b) EDX spectrum showing elements associated with deposit. (c) Background corrected XRD pattern of secondary mineral demonstrating that it is primarily thenardite.

Sulphate formation

The secondary sulphate deposits probably form in part from the primary weathering of sulphidic minerals from accessory pyrite and possibly from basaltic glass from the overlying host basalt. Sulphur concentration in the Blue Dragon flow was reported to be ∼0.06 wt% (Richardson et al. Reference Richardson, Hinman, McHenry, Kotler, Knipe and Scott2012). The specific factors controlling the kinetics and mechanisms of sulphide oxidation to sulphate are still poorly understood. Oxidation of pyrite may be oxidized by either molecular oxygen or ferric iron and is typically very slow under abiotic conditions. However, micro-organisms can accelerate this process by at least six orders of magnitude (Sampson et al. Reference Sampson, Phillips and Ball2000; Pisapia et al. Reference Pisapia, Chaussidon, Mustin and Humbert2007). Pyrite oxidation is complicated and involves many intermediate sulphur species with oxidation states between −I and +VI (McGuire et al. Reference McGuire, Edwards, Banfield and Hamers2001; Heidel & Tichomirowa Reference Heidel and Tichomirowa2011). The equations below summarize the general oxidation mechanism of pyrite without intermediate reactions (Singer & Stumm Reference Singer and Stumm1970):

(1)$$\eqalign{{\rm FeS}_{2({\rm s})} + 7/2{\rm O}_{2({\rm g})} + {\rm H}_2 {\rm O}_{({\rm l})} \to {\rm Fe}^{2 +} _{({\rm aq})} + 2{\rm SO}_{4({\rm aq})}^{2-} + 2{\rm H}^ + _{({\rm aq})} $$

(2)$$\eqalign{{\rm FeS}_{{\rm 2}({\rm s})} + {\rm 14Fe}^{{\rm 3} +} _{({\rm aq})} + {\rm 8H}_{\rm 2} {\rm O}_{({\rm l})} \to {\rm 15Fe}^{{\rm 2} +} _{({\rm aq})}+ {\rm 2SO}_{4({\rm aq})}^{{\rm 2} -}+ {\rm 16H}^ + _{({\rm aq})}$$

Sulphur isotope analysis was performed to further explore the transformation of pyrite and help determine potential involvement of microbes in formation of the deposits by identifying preferential fractionation of sulphur isotopes.

Sulphur fractionation

Sulphur isotopic systems have been suggested as a potential means for searching for signs of life on Mars by Parnell et al. (Reference Parnell, Boyce, Osinski, Izawa, Banerjee, Flemming and Lee2012), especially sulphates that are prolific on the Martian surface. The oxidation of sulphidic minerals to sulphate through a series of intermediate species represents an important energy-yielding pathway for endolithic micro-organisms. Iron and sulphur species are two of the most prevalent redox active elements, acting as an important metabolic energy source for sulphur oxidizing organisms (Balci et al. Reference Balci, Shanks, Mayer and Mandernack2007; Philippot et al. Reference Philippot, Van Zuilen, Lepot, Thomazo, Farquhar and Van Kranendonk2007; Wacey et al. Reference Wacey, Saunders, Brasier and Kilburn2011). As FeS₂ is one of the most abundant metal sulphide minerals in basaltic rocks, it is considered the main source of reduced sulphur in COM basalts. Although sulphur isotopes have been used to track bio-oxidation in many settings (Peterson Reference Peterson1999; Bottrell & Newton Reference Bottrell and Newton2006; Peters et al. Reference Peters, Strauss and Farquhar2009, Reference Peters, Strauss, Farquhar, Ockert, Eickmann and Jost2010; Zerkle et al. Reference Zerkle, Farquhar, Johnston, Cox and Canfield2009; Takai & Nakamura Reference Takai and Nakamura2011), little is known about the direct oxidation pathways and mechanism of bio-oxidation of sulphidic minerals in basalts. Analysis of ∂³⁴S fractionation values between the host basalt and the secondary deposits values was performed in hopes it would give an insight into the oxidation pathways and the role of micro-organisms in formation of the deposits.

Sulphur fractionation values between the secondary sulphate deposits and the host basalt (∆³⁴S_SO4-b) are shown in Table 1. The largest ∆³⁴S_SO4-b depletion was −4.3‰, with an average depletion of −2.6‰ (n=12, 1σ). No correlation with the cave location or within individual caves was observed regarding ∂³⁴S values. This lack of spatial correlation may imply that similar bio-oxidation pathways occur throughout the sampling area. These small but significant differences may imply biological oxidation as abiotic sulphide oxidation fractionation values typically are unidirectional and are generally indistinguishable from their parent mineral (Rye et al. Reference Rye, Bethke and Wasserman1992; Taylor & Wheeler Reference Taylor, Wheeler, Alpers and Blowes1994; Lefticariu et al. Reference Lefticariu, Pratt and Ripley2006; Balci et al. Reference Balci, Shanks, Mayer and Mandernack2007; Zerkle et al. Reference Zerkle, Farquhar, Johnston, Cox and Canfield2009) Oxidation is likely to proceed by biotic mechanisms because the fractionation observed in the COM samples (Table 1) exceeds reported abiotic values (Table 2). Previous biotic oxidation experiments of sulphidic minerals have yielded slightly more depleted ∂³⁴S fractionation values than the abiotic values and reflect the ∆³⁴S_SO4-b values observed from the COM secondary deposits in a better manner.

Table 1. Sulphur isotopic values of secondary deposits and host basalts

^a ±0.2‰ error.

^b δ³⁴S=[(³⁴S/³²S)_sample/(³⁴S/³²S)_VCDT]×1000.

^c Average of basalt δ³⁴S values used.

^d Standard deviation=1.0.

Table 2. Literature compilation of biotic and abiotic sulphur fractionation for oxidation of sulphidic minerals

^a Non-stoichiometric pathway.

^b Stoichiometric pathway.

Bio-oxidation of pyrite involves several intermediate species, such as polythionates (S_xO_y²⁻) and elemental sulphur. During bio-oxidation, disproportionation micro-organisms utilize these intermediate compounds as both electron donors and acceptors and show no preferential uptake between the sulphur isotopes. This results in largely negligible ∂³⁴S values between sulphate and sulphide (Habicht et al. Reference Habicht, Canfield and Rethmeier1998; Smock et al. Reference Smock, Bottcher and Cypionka1998; Bottcher et al. Reference Bottcher, Thamdrup and Vennemann2001; Finster Reference Finster2008). The presence of disproportionating bacteria may cause ∂³⁴S values from the secondary deposits to remain near the host basaltic values. Depletion in ∆³⁴S_SO4-b may imply biological activity, however, it is possible that these fractionation values may be due to stoichiometric (breaking of ionic S-FeS₂ bonds) abiotic oxidation. More extensive experiments would be required to elucidate the exact role of microbial activity and detailed bio-oxidation pathways involved in the formation of secondary sulphate deposits.

Presence of bio/organic compounds

Biological activity associated with mineral deposits can be inferred through various methods, including chemical or isotopic signatures, which can be preserved in the minerals (Boston et al. Reference Boston2001). The secondary deposits at COM were investigated for chemical and isotopic biological signatures, which could suggest direct or indirect biological involvement in mineralization of the Na-sulphate deposits. The presence of bio/organic compounds in the secondary minerals was determined using geomatrix-assisted laser desorption/ionization (GALDI) (Yan et al. Reference Yan, Stoner, Kotler, Hinman and Scott2007a). This technique uses a mineral matrix to assist in desorption and ionization of bio/organic compounds with little to no sample preparation. The ability of minerals to facilitate in the desorption and ionization of bio/organic compounds is a primary focus in previous studies of GALDI-FTICR-MS as a viable technique for bio/organic compound detection (Yan et al. Reference Yan, Stoner, Kotler, Hinman and Scott2007a, Reference Yan, Stoner and Scottb; Kotler et al. Reference Kotler, Hinman, Yan, Stoner and Scott2008; Richardson et al. Reference Richardson, Hinman, McJunkin, Kotler and Scott2008, Reference Richardson, Hinman and Scott2009). Previous studies using synthetic and natural thenardite have demonstrated the ability of thenardite to assist in the ionization and detection of bio/organic compounds down to detection levels of 3 ppt (Richardson et al. Reference Richardson, Hinman, McJunkin, Kotler and Scott2008, Reference Richardson, Hinman and Scott2009).

To assist in identification of any associated bio/organic peaks in the mass spectra from the COM secondary deposits, the spectra were compared with a suite of FTICR-MS standard spectra of inorganic thenardite, Na-carbonate (trona, natron), and physical combinations between these Na-sulphate and Na-carbonate minerals (Richardson et al. Reference Richardson, Hinman, McJunkin, Kotler and Scott2008, Reference Richardson, Hinman and Scott2009, Reference Richardson, Hinman, McHenry, Kotler, Knipe and Scott2012). All spectra of the standard Na-sulphate and Na-carbonate minerals had peaks with mass defects (i.e. the non-integer value after the decimal point) suggestive of inorganic constituents. Mass defects must be a linear sum of the mass defects of individual elemental components. Common non-hydrogen elements associated with bio/organic compounds have mass defects near 0.00 u (e.g. ¹²C at 12.000 u, ¹⁶O at 15.995 u). Hydrogen (at 1.008 u) tends to dominate the mass defects of bio/organic compounds because there are usually twice as many hydrogen atoms as other elements. Such distinction between an inorganic and organic ion based on the peak's mass defect is easily achieved when the ion has a mass-to-charge (m/z) ratio less than 400, because inorganic ions tend to have mass defects that are near to or would round up to next nominal integer mass, whereas bio/organic compounds tend to have mass defects that would round down to next integer mass. As the m/z rises above ∼400, the trends in mass defects for bio/organic and inorganic ions tend to shift (e.g. organic ions accumulate enough hydrogen atoms causing the mass defect to appear to indicate inorganic elements as the m/z value approaches the next integer).

Figure 4(a) shows a two-dimensional LD-FTICR-MS map where each circle represents a spectrum obtained from a single laser shot. The spectra represented by the open circles in the map only had peaks indicative of inorganic species. The coloured or shaded circles indicate the occurrence of peaks suggestive of bio/organic compounds. The heterogeneous distribution of inorganic and organic peaks in the map is common to all COM secondary deposits analysed. A representative negative mode spectrum from a secondary deposit collected from a wall cavity in Arco Tunnel is shown in Fig. 4(b). The inorganic peaks observed in the spectrum have been reported in previous studies of Na-sulphate spectra (Poels et al. Reference Poels, Van Vaeck and Gijbels1998; Van Vaeck et al. Reference Van Vaeck, Adriaens and Adams1998; Richardson et al. Reference Richardson, Hinman, McJunkin, Kotler and Scott2008, Reference Richardson, Hinman, McHenry, Kotler, Knipe and Scott2012). Peaks at m/z 183.081 and 339.040 (Figs. 4(c) and (d), respectively) have mass defects suggestive of bio/organic compounds. These peaks are related to bio/organic compounds because of the (1) absence of the peaks in synthetic inorganic Na-sulphate and Na-carbonate standards, (2) mass defects suggestive of bio/organic elements, and (3) isotopic distributions that correspond to the theoretical isotopic distribution of suggestive bio/organic formulas. The bio/organic related peaks are an example of complex cluster ions, similar to that reported for glycine with jarosite (Kotler et al. Reference Kotler, Hinman, Yan, Stoner and Scott2008) and stearic acid with thenardite (Richardson et al. Reference Richardson, Hinman, McJunkin, Kotler and Scott2008). Using a systematic procedure based on the mass defects and isotopic distributions outlined by Kotler et al. (Reference Kotler, Hinman, Yan, Stoner and Scott2008), the most probably composition for the peaks at m/z 183 and 339 are C₈H₁₆SONa⁻ and C₁₀H₂₀S₃O₅Na⁻, respectively. The occurrence of these complex cluster ions is not unusual, as cluster ions often form because of complex reactions in the laser desorption plume or in the gas phase (Karas et al. Reference Karas, Bachmann and Hillenkamp1985; Knochenmuss et al. Reference Knochenmuss, Dubois, Dale and Zenobi1996; Budimir et al. Reference Budimir, Blais, Fournier and Tabet2007). These reactions lead to ions larger than the expected molecular ion due to formation of adducts from the addition of matrix components and/or analyte species (Knochenmuss et al. Reference Knochenmuss, Dubois, Dale and Zenobi1996; Karas & Kruger Reference Karas and Kruger2000; Ham et al. Reference Ham, Durham and Scott2003; Budimir et al. Reference Budimir, Blais, Fournier and Tabet2007). The exact identification of the original bio/organic compound was unobtainable as it is quite difficult to ascertain without undertaking a systematic experiment using a variety of bio/organic compound/thenardite combinations. Such a systematic experiment was conducted to identify that the peak at m/z 390.260 observed in evaporitic thenardite samples from Searles Lake, CA originated from the presence of stearic acid (Richardson et al. Reference Richardson, Hinman, McJunkin, Kotler and Scott2008), which was also observed in the COM samples (Table 3).

Fig. 4. (a) LD-FTICR-MS chemical pseudo-image showing heterogeneity of chemical composition within the sample. Colored circles indicate areas with potential organic compounds: green for m/z 183, orange for m/z 339 and black for a mix of m/z 183 and 339. Blank circles are locations with peaks indicative of inorganic only compounds. (b) Negative mode laser desorption ion FTICR-MS spectrum of secondary sulphate deposits. Peaks at (c) m/z 183 and (d) m/z 339 have mass defects and isotopic distributions suggestive of bio/organic cluster ions.

Table 3. Suggestive chemical formulae for bio/organic peaks associated with the secondary sulphate deposits observed in the LD-FTICR-MS spectra

AT, Arco Tunnel; WC, Wilderness Caves; SC, Snow Cone Pit; CP, Crystal Pit.

A representative positive mode spectrum (Fig. 5) shows several high mass peaks, most of which are easily identified as inorganic based on their mass defects (values were generally >0.5 u), their occurrence in the inorganic standard spectra and inorganic peaks (such as m/z 165 and 181) that have been previously reported in Na₂SO₄ spectra (Poels et al. Reference Poels, Van Vaeck and Gijbels1998; Van Vaeck et al. Reference Van Vaeck, Adriaens and Adams1998; Ignatova et al. Reference Ignatova, Van Vaeck, Gijbels and Adams2003; Richardson et al. Reference Richardson, Hinman and Scott2009). Close inspection revealed that the peaks at m/z 141.052 and 170.044 have mass defects suggestive of bio/organic compounds and are absent in the spectra of the inorganic standards. Peak identification in the positive spectra was more difficult because the minor isotopic peaks were less distinct because of poor signal-to-noise (i.e. low abundance). However, peak identification was obtained by comparing the observed mass defects with theoretical mass defects. Using this method, the peak at m/z 141.052 was found to have only one likely composition with a theoretical weight of 141.053 u, corresponding to a formula of C₅H₁₀O₃Na⁺. Under the FTICR-MS parameters used, the mass accuracy is ∼0.003 u with a resolution of 10 000. Precise identification of the peak at m/z 170.044 was unobtainable because several likely formulae could account for the mass defect. However, this peak is considered to be from a bio/organic compound because of its mass defect and its absence in the inorganic Na-sulphate and Na-carbonate spectra. The formula C₅H₁₀O₃Na⁺ could correspond to a small molecular or fragment ion similar in structure to diethyl carbonate, ethyl lactate, or hydroxyvaleric acid that have a C=O group as a potential site for cation attachment of Na⁺.

Fig. 5. Positive mode LD-FTICR-MS spectrum of a secondary sulphate deposit from a floor deposit within a cave in the Wilderness Caves area.

A number of complex bio/organic or organic cluster ions are seen in the positive and negative spectra of the secondary deposits. A comprehensive list of these positive and negative mode peaks and their suggestive chemical formulae is shown in Table 3. Approximately 30% of these suggestive formulas had C : H : O ratios that resemble ratios observed in lipids, sugars, and/or amino acids. Unfortunately, many of the suggestive formulas indicate that the ions may result from gas-phase reactions in the laser desorption or ablation plume. These gas-phase reactions can lead to unusual chemical formulae. Previous FTICR-MS studies have yielded similar gas-phase reactions with glycine mixed with jarosite (Kotler et al. Reference Kotler, Hinman, Yan, Stoner and Scott2008) and stearic acid with thenardite (Richardson et al. Reference Richardson, Hinman, McJunkin, Kotler and Scott2008).

Additional evidence for associated bio/organic compounds in the secondary deposits was gathered using FTIR techniques because Bullen et al. (Reference Bullen, Oehrle, Bennett, Taylor and Barton2008) have shown that FTIR can detect low levels of metabolic compounds on minerals associated with caves. Even though FTIR does not typically have the sensitivity (∼0.1 wt%) (Materials Characterization Lab 2012) that is possible with LD-FTICR-MS (i.e. 3 ppt) (Richardson et al. Reference Richardson, Hinman, McJunkin, Kotler and Scott2008), several bands were observed that support the presence of organic compounds. Representative FTIR spectra are shown in Fig. 6. The inorganic bands in Fig. 6(a) are representative of thenardite; a more detailed description of these bands can be found in Richardson et al. (Reference Richardson, Hinman, McHenry, Kotler, Knipe and Scott2012). Evidence of organic-related bands can be seen at 559, 545 and 536 cm⁻¹, which are the characteristics of stretching and rocking of singly bonded carbon compounds (Smith Reference Smith1999). In Fig. 6(b), the prominent bands result from a combination of sulphate and carbonate oxyanions (Richardson et al. Reference Richardson, Hinman, McHenry, Kotler, Knipe and Scott2012). Similar to Fig. 4(a), several bands could represent organic compounds. The bands at 536 and 580 cm⁻¹ are representative of singly bonded organic compounds (Smith Reference Smith1999). In addition, the band at 785 cm⁻¹ is reminiscent of a C–H stretch associated with aromatic organic compounds (Coates Reference Coates and Meyers2000).

Fig. 6. Representative FTIR spectra of secondary deposits. (a) FTIR spectrum dominated by sulphate oxyanion, whereas FTIR spectrum (b) shows additional bands corresponding to carbonate oxyanion. Both spectra show bands suggestive of organic compounds (indicated by *).

Potential sources of bio/organic compounds

There are several potential sources that might account for the presence of bio/organic compounds in the secondary mineral deposits, including accidental emplacement of compounds and microbial activity. Emplacement of extraneous bio/organic compounds into the precipitates may be due to (1) plant detritus moving downward through the basalts via cracks and fissures with subsequent deposition with the secondary deposits, (2) organic aerosols carried through the cave openings and (3) post-depositional interaction with cave animals. Plant detritus is not likely because surface vegetation is extremely sparse in the Blue Dragon flow (Fig. 1(b)). Transportation of organic aerosols in the caves and their subsequent emplacement on the ceiling and floor deposits seems highly unlikely, especially considering the extent and complexity of some of the subsurface features (i.e. Arco Tunnel). Post-depositional interaction with cave animals is also unlikely to produce the detected organic compounds, especially for ceiling and wall deposits. To limit the possibility post-depositional interaction with cave animals, mineral samples were retrieved from the interior of the deposits rather than from the surfaces. Furthermore, no evidence of animal interaction was found near the secondary deposits. Microbial activity occurring in or around the basalt is another potential source of bio/organic compounds.

Proposed mechanisms for secondary deposits

A plausible explanation for the presence of secondary Na-sulphate deposits is the abiotic and biotic physiochemical weathering of the overlying basalt (Fig. 7). Downward migrating meteoric water interacts with primary alkali-rich plagioclase minerals (which constitutes a majority of the groundmass phase) leaching sodium ions. The Na-rich deposits are consistent with the elevated alkali concentrations reported in the Blue Dragon flow (Stout et al. Reference Stout, Nicholls and Kuntz1994; Richardson et al. Reference Richardson, Hinman, McHenry, Kotler, Knipe and Scott2012). Na-leaching has been observed as a source for Na-rich secondary minerals (halides) in other basaltic cave settings (Forti Reference Forti2005). Isolated regions of basaltic glass could be another source of sulphur, especially since the Blue Dragon flow has a high concentration (∼25 wt% by volume) of glass (Stout et al. Reference Stout, Nicholls and Kuntz1994). Varying oxidation states of sulphur (S⁶⁺ and S²⁻), along with Fe-sulphide globules, have been reported in basaltic glasses (Métrich et al. Reference Métrich, Berry, O'Neill and Susini2009). Thus, bio-oxidation of Fe-sulphides and S²⁻ in basaltic glass could account for the sulphate oxyanion observed in the secondary deposits. Sulphur-oxidizing endolithic organisms, such as certain Thiobacillus species, have been reported in basaltic cracks and glass (Thorseth et al. Reference Thorseth, Furnes and Tumyr1995; McKinley et al. Reference McKinley, Stevens and Westall2000; Edwards et al. Reference Edwards, McCollom, Konishi and Buseck2003; Cousins et al. Reference Cousins, Smellie, Jones and Crawford2009; Izawa et al. Reference Izawa, Banerjee, Flemming, Bridge and Schultz2010). Abiotic and biotic-mediated oxidation of these reduced sulphur species could produce and subsequently introduce the sulphate oxyanion into migrating water. Bio/organic compounds (such as lipids, sugars, proteins, amino acids and their degradation products) could be incorporated into migrating water along with sodium and sulphate ions. Upon evaporation in the ceiling cracks and cave floors, a fraction of the bio/organic compounds would be incorporated in the secondary deposits (Fig. 7).

Fig. 7. Schematic illustrating biotic and abiotic weathering of the overlying basalts at COM and subsequent mineralization of secondary Na-sulphate minerals on the cave floors.

The need for sulphur oxidation, in conjunction with FTICR-MS and FTIR spectra, demonstrate that bio/organic compounds are associated with the secondary deposits, and thus, some degree of biogenic activity is probably involved in the formation of secondary sulphate deposits. While plausible, evidence for the proposed scheme is not definitive. To the authors’ knowledge, no investigation of biogenic activity related to Na-sulphate mineralization in volcanic settings has been conducted. Therefore, in the future, an exhaustive microbial study would need to be conducted to identify and confirm microbial species associated with secondary mineralization.