Insight Into Hartoušov Mofette, Czech Republic: Tales by the Fluids

The Cheb Basin (Czech Republic) is characterized by emanations of magma-derived gases and repeated occurrences of mid-crustal earthquake swarms with small to intermediate magnitudes (ML < 4.5). Associated intense mantle degassing occurs at the Hartoušov Mofette, a representative site for the Cheb Basin. Here, we performed 14 sampling campaigns between June 2019 and March 2020. Gas samples of fluids ascending in two boreholes (F1, ∼28 m depth and F2, ∼108 m depth) and from a nearby natural mofette were analyzed for their chemical (CO2, N2, O2, Ar, He, CH4, and H2) and isotope compositions (noble gases and CO2). CO2 concentrations were above 99.1% in most samples, while O2 and N2 were below 0.6%. He ranged from 19 to 34 μmol/mol and CH4 was mostly below 12 μmol/mol. Isotope compositions of helium and carbon in CO2 ranged from 5.39 to 5.86 RA and from −2.4 to −1.3 ‰ versus VPDB, respectively. Solubility differences of the investigated gases resulted in fluctuations of their chemical compositions. These differences were accompanied by observed changes of gas fluxes in the field and at the monitoring station for F1. Variations in solubilities and fluxes also impacted the chemical concentration of the gases and the δ13C values that were also likely influenced by Fischer-Tropsch type reactions. The combination of (a) the Bernard ratio, (b) CH4/3He distributions, (c) P-T conditions, (d) heat flow, and (e) the sedimentary regime led to the hypothesis that CH4 may be of mixed biogenic and volcanic/geothermal origin with a noticeable atmospheric contribution. The drilling of a third borehole (F3) with a depth of ∼238 m in August 2019 has been crucial for providing insights into the complex system of Hartoušov Mofette.


INTRODUCTION
Gases from various tectonic regions can differ in their geochemical characteristics (Lupton, 1983). Therefore, detailed studies of geothermal gases often contribute to an improved understanding of tectonic and geological settings and their corresponding fluids (Giggenbach, 1992(Giggenbach, , 1996Lee et al., 2005). Furthermore, geodynamic settings are often influenced by seismic, volcanic, and geothermal activities and that also plays an essential role in earth degassing (Irwin and Barnes, 1980). Hydrothermal fluids transport volatiles from the deep crust or mantle to the surface, while their circulation in the crust can also enhance geodynamic processes. The generation and transportation of these fluids are linked to a plethora of faulting processes (including nucleation, propagation, arrest, and recurrence of earthquake ruptures, fault creep or slow earthquakes, and the long-term structural and compositional evolution of fault zones - Hickman et al., 1995). Their study therefore offers a tool for monitoring and better understanding mantle and crustal processes.
Geogenic CO 2 discharges are widespread throughout central Europe (Pearce et al., 2004). Although they occur in diverse geological and geodynamic settings their distribution is principally controlled by the Cenozoic rift systems and associated Tertiary volcanism. The Eger Rift (Czech Republic, Figure 1A) is an intraplate region without active volcanism (youngest volcanic activity took place 0.29 Ma ago- Mrlina et al., 2009). However, emanations of magma-derived gases take place in the western part of the Cheb Basin (Weinlich et al., 1999;Geissler et al., 2005). This area associates with earthquake swarms that are likely induced by the ascent of the magmatic fluids (Parotidis et al., 2003). Variations in the gas flow together with chemical and isotope compositions were noticed during periods of seismic activity (Bräuer et al., 2018). For instance, increase in the gas flow was observed at the Hartoušov mofette field (HMF) after a series of earthquakes in 2014 (Fischer et al., 2017). During the same period, the gas flow in the mofette field of Dolní Cástkov (Cheb Basin) decreased drastically. Post-seismic shifts in δ 13 C CO2 and 3 He/ 4 He after the small swarm on the 4th and 5th of December 1994 were documented by Weise et al. (2001). They estimated the fluid transport velocities in the upper crust at 400 m/d for the Bublák mofette field. Spatial and temporal increase of mantle-derived helium contributions in the eastern Cheb Basin suggested that fluid injection channels reach down to the lithospheric mantle (Bräuer et al., 2005(Bräuer et al., , 2009. It is worth noting that the subcontinental lithospheric mantle (SCLM) contribution at HMF increased from 38% in 1993 to 89% in 2016 (Bräuer et al., 2018), while the isotope ratio of He at the Bublák mofette reached 6.3 R A before the earthquake swarms in 2000 and 2008. This increase indicated that ascending magma from the SCLM into the crust might have triggered the swarm.
These observations highlight the complexity and uniqueness of the Cheb Basin, and illustrate the need to deepen our knowledge on the system. Especially the relationship between the fluids and the seismic activity demands further research. Toward this goal, this investigation focuses on monitoring fluids encountered at different depths of the HMF ( Figure 1B). Gases emerging in the mofette (surface expression), as well as in two boreholes (F1 and F2 with depths ∼28 m ∼108 m, respectively) have been sampled regularly over 1 year. Our study aims to identify key processes affecting the gases in the "micro"system of HMF. We also aimed to show first results of gases during the drilling period of a third borehole. This was drilled during the study period, to a depth of ∼238 m. Moreover, our study takes a first step toward the determination of the origin of CH 4 . This work is part of the Mofette Research (MoRe) project, which combines geochemical and geophysical methods to document how fluids can actively act in a non-volcanic rift setting of intraplate seismicity.

STUDY AREA
The Cheb Basin is an asymmetric intracontinental basin that lies within the western part of the Eger Rift close to the Nový Kostel focal zone (NK, Fischer and Horálek, 2003; Figure 1). In its eastern part, the basin is defined by the NNW-SSE Mariánské Lázně Fault Zone (MLFZ). Its petrological regime includes rock sequences of Upper Cambrian to Ordovician age and areas characterized by Late Variscan intrusions that are dominated by granites. Its crystalline basement consists of muscovite granites of the Smrčiny/Fichtelgebirge Pluton (Hecht et al., 1997). These combine with crystalline schists of the Saxothuringian Zone of the Variscan Orogen (Fiala and Vejnar, 2004). The basin was formed during the Late Tertiary and Quaternary by the re-activation of Hercynian faults and separated microplates present within the basement (Bankwitz et al., 2003;Babuška and Plomerová, 2008) or either of them. The Cheb Basin was filled with 300 m-thick fluvial and lacustrine deposits (Špičáková et al., 2000;Nickschick et al., 2015). Volcanism in the western part of the Eger Rift was weak and of Quaternary age, and the youngest volcanic activity in the area took place 0.29 Ma ago (Mrlina et al., 2009).
Magma-derived volatiles present in the Cheb Basin consist of CO 2 -rich waters, wells, and dry gas vents (Weinlich et al., 1999). The dominant gas species is CO 2 (>98%) with minor amounts of N 2 , O 2 , Ar, He, and CH 4 (Weinlich et al., 1999;Kämpf et al., 2013;Bräuer et al., 2018). Isotope compositions for both helium and carbon in CO 2 indicate a mantle origin of the fluids (Bräuer et al., 2018). Along with the emanation of magma-derived gases, the recurrence of swarms with small to intermediate magnitudes (M L < 4.5) is characteristic for the area. Parotidis et al. (2003) hypothesized that the ascent of magmatic fluids triggers most of these anomalous earthquake activities. However, Fischer et al. (2014) suggested that fluids are probably not the only triggering factor of the swarms. It is worth noting that the lack of CO 2 emanations in the epicentral area of the Nový Kostel (NK) focal zone likely results from an impermeable rock formation above the fault zone (Bräuer et al., 2009). Nonetheless, it has produced almost 80% of the earthquake swarms during the past 30 years (Fischer and Michálek, 2008).
In order to investigate relationships between geodynamic processes, CO 2 degassing, and the continental "deep biosphere, " three boreholes were drilled in the intensively degassing HMF (Dahm et al., 2013, 26 t/d of diffuse soil degassing; Kämpf et al., 2019). F1 [corresponding to 1H-031 in Bräuer et al. (2018) andVP8303 in Fischer et al. (2020)] was drilled in 2007 down to ∼28 m below ground and taps into a CO 2saturated, confined aquifer. The F2 borehole [HJB-1 in Bräuer et al. (2018)] has a depth of ∼108 m, and was drilled in spring 2016 to investigate whether the increased fluid and substrate flow can accelerate microbial life in active fault zones and CO 2 conduits (Bussert et al., 2017). In August 2019, a third borehole (F3) was drilled to a depth of ∼238 m. Its principal aim is to investigate the relation between geogenic degassing and earthquake activity by combination of geochemical and geophysical techniques. Further details about the borehole configuration can be found in Woith et al. (2020).

MATERIALS AND METHODS
From June 2019 to March 2020, 14 sampling campaigns took place in the HMF, and 40 gas samples were collected. Samples were obtained from the free gas phase after passage through water. They were collected in Exetainer vials and glass vessels with two vacuum stopcocks. The containers were first filled with water. Subsequently, gas was collected with tubing that was directly connected to the borehole head. Its upward flow replaced the water in the vessel. For mofette samples, the tubing was connected to an inverted funnel, which was immersed in the water. Three samples were collected per sampling site on each sampling day; one of these samples was used to determine the gas composition and the other two for noble gas and carbon isotope analyses of CO 2 , respectively.
For analyses of their chemical composition, gases were sampled in 1,000 cm 3 glass vessels. The CO 2 contents were determined volumetrically after absorption in a KOH solution with a precision of 0.1 cm 3 . The remaining CO 2free aliquot was analyzed at a commercial laboratory in the Czech Republic (Labor Union) using gas chromatography. N 2 and O 2 had a relative precision of ± 3%, while the minor components (Ar, He, H 2 , CH 4 , and light hydrocarbons) were determined at relative precisions of ±10-40%. The detection limit for H 2 and He was <100 µmol/mol of CO 2 -free gas, while for CH 4 and other light hydrocarbons it was <1 µmol/mol. This corresponded to 0.5 µmol/mol and 0.005 µmol/mol, respectively, in gas with CO 2 contents of 99.5%. Further details of the method are also available in Weinlich et al. (1998).
δ 13 C analyses were performed from the Exetainer vials (12 cm 3 ) at GFZ Potsdam with a gas chromatograph (GC 6890N, Agilent Technologies, equipped with Plot column) coupled via a combustion device (GC-C/TC III, Thermo Fisher Scientific) to an isotope ratio mass spectrometer (MAT253, Thermo Fisher Scientific). The δ 13 C values are reported in versus Vienna Pee Dee Belemnite (VPDB) standard with a standard deviation of ±0.3 .
Samples for noble gas analyses were collected in 250 cm 3 glass vessels with two vacuum stopcocks. The isotopic composition of noble gases was determined for selected samples at GFZ Potsdam. For this purpose, aliquots were analyzed for noble gas concentrations and isotopic compositions with a VG 5400 noble gas mass spectrometer after removing the active gas components in a gas purification line. Details of the analytical procedure can be found in Niedermann et al. (1997). The precision of He isotope measurements generally ranges from ±1.7% to ± 2.4%, with only four samples [(7), (18), (31), and (33)] being above this range (± 4.7% in average). The precision of 4 He/ 20 Ne generally ranges from ± 5.5% to ± 16% at 2σ (95% confidence) level. An exception to the latter range is sample (25) (± 84%), because of an extremely low Ne concentration. The measured 3 He/ 4 He ratios were corrected for atmospheric He contributions by assuming that contributions of 20 Ne are entirely of atmospheric origin, and by using the 4 He/ 20 Ne ratio of air (0.319) according to Craig et al. (1978): where 3/4 and 4/20 are the 3 He/ 4 He and 4 He/ 20 Ne ratios. Corrections were <3%, except for sample (1) (∼5%).  Bräuer et al. (2018) were considered as background values, and used for comparison. It is worth mentioning that literature data from Weinlich et al. (1999) and Kämpf et al. (2013) were not used in our study, because they were influenced by seismic activity, and may result in misleading conclusions. CO 2 concentrations exceeded 99 % in most samples, while O 2 and N 2 were mostly below 0.3 and 0.6%, respectively ( Table 1). Six samples presented lower CO 2 contents (down to 92.5%), most likely due to air contamination during sampling that was manifested in elevated O 2 and N 2 concentrations (up to 1.6 and 5.9%, respectively). These samples are included and presented in the current study, however, they are given in gray color and are signed with "_AC, " which stands for Air Contamination.
He in the uncontaminated samples ranged from 18 to 34 µmol/mol, whereas lower concentrations were found in samples with enhanced air component. Ar showed a wide range of concentrations (24 to 835 µmol/mol). Concentrations of CH 4 were generally lower than 12 µmol/mol, while enrichments up to 23 µmol/mol were found in F1 before drilling and during drilling at a core depth of ∼125 m. The lowest CH 4 value (0.5 µmol/mol) was found at F2 in the sample collected after the perforation of the F3 steel casing after the drilling period. H 2 was always below detection limit, apart from one sample collected from F2 when the drilling reached its final depth. C 2 H 4 and C 3 H 8 were almost always below detection limit apart from three borehole samples [(5) and (6) for C 2 H 4 and (5) and (27) for C 3 H 8 ]. C 2 H 6 was present in concentrations lower than 0.24 µmol/mol.
Noble gases, as well as carbon isotopes, were measured only in selected samples. The isotope composition of He showed values between 5.37 and 5.86 R A (where R A is the atmospheric 3 He/ 4 He ratio of 1.39 × 10 −6 ). Corresponding 4 He/ 20 Ne ratios covered a wide range of values (5.2 to ∼19,000). The samples analyzed for δ 13 C CO2 were in the range of −2.4 to −1.3 vs. VPDB.
Water temperature and gas flow in F1, the wellhead and bottom-hole pressure in F2, and meteorological data were obtained from the monitoring station located at F1 in the HMF ( Figure 1B). These data were obtained from the CO 2 network (CarbonNet) of Charles University of Prague. Details are provided in the Section of Data Availability Statement.

DISCUSSION
Processes Affecting the Gases CO 2 is the dominant gas component emerging from the HMF, with He isotopes and δ 13 C CO2 showing a mantle origin (Kämpf et al., 2013;Bräuer et al., 2018). In Figure 2A, the atmospheric gas component (here represented by N 2 ) is plotted together with CO 2 and CH 4 , which are characteristic of hydrothermal-type components (Giggenbach et al., 1993). This ternary plot aims to identify secondary processes, such as dissolution (Italiano et al., 2014a). As expected, N 2 concentrations are strongly enhanced in the air contaminated samples (gray-colored symbols). Borehole samples (13) and (26), collected after the perforation, as well as samples from F1 collected when F3 arrived at its final depth [(8)] and after the drilling process [(10), (11), and (14)] are also enriched in N 2 . This suggests a CO 2 loss that was probably caused by dissolution processes. An increase in CH 4 content is found for two F1 samples [(1) and (6) - Figure 2A]. In the He-N 2 -Ar ternary diagram ( Figure 2B -after Giggenbach et al., 1983), the vast majority of the samples are distributed inside the triangle delimited by N 2 /Ar ratios of air and air-saturated water (ASW), and the He apex. This distribution shows a two-component mixing relationship between mantle and atmospheric sources. Furthermore, most of the N 2 /Ar ratios approach the value of 50 (Table 1), the typical value of airsaturated waters (Heaton and Vogel, 1981;Fischer et al., 1998). This suggests that N 2 originates from a mixture of shallow airsaturated fluids and deep circulating groundwaters. Samples (17) and (30), collected when the drilling of F3 arrived at ∼2 m and ∼41 m depth, respectively, revealed an excess of Ar for so far unknown reasons, while samples affected by atmospheric contamination plot close to the air and ASW points.
Enrichments of the geogenic components are evident in the binary plot of He/Ar versus N 2 /O 2 (Figure 3A), with F2 presenting the most extreme values ( Table 1). In most cases, samples of HMF have stable CH 4 and He contents ( Figure 3B). As expected, the variability of CH 4 is somewhat higher in samples not affected by air contamination (Figure 3B). The higher He concentration in the F2 samples collected at the end [(22)] and after the F3 drilling [(25)] indicates a CO 2 loss. This loss may have been caused by the strong solubility contrast of these gases in aquatic environments (Reid et al., 1987). This may have been enhanced by the low and non-thermal temperatures of the HMF (up to 15 • C for F1 and F2 and up to 20 • C for the Mofette during the summer period). D' Alessandro et al. (2014) suggested that ( 3 He/ 4 He) c is the air-corrected 3 He/ 4 He ratio (Craig et al., 1978). 4 He concentrations in this table were obtained by relating the He partial pressure determined by noble mass spectrometry to the total pressure in the gas vessel used for the noble gas analyses, assuming that the latter was equal to atmospheric pressure. 2σ uncertainties of these values are estimated at ∼10-15%.
such contrasts may result in strong enrichments in less soluble gases, when a gas mixture rises through unsaturated waters at depleted gas/water ratios. Enrichment in He content was also noticed in the aircontaminated sample (28) ( Figure 3B and Table 1). However, a different approach in calculating the He concentration yields a much lower value (Table 3). This suggests that this difference may result from an analytical problem during gas chromatography. Samples from F1 collected during the 1st campaign [(1)] and when F3 arrived at ∼125 m depth [(6)] exhibit enrichments in CH 4 (Figure 3B). This was likely caused by biogenic generation of the gas, a hypothesis promoted by the sedimentary formations (Daskalopoulou et al., 2018), and also supported by the depleted values of the atmospheric components (Goff and Janik, 2002;Easley et al., 2011). It should be noted that CH 4 content changes are common, as the gas can be involved in many production and consumption processes (Rolston et al., 1993). Another possible scenario that may apply to sample (1), collected in June 2019, is that CH 4 was accumulated at the wellhead for the period that the borehole was sealed and due to its low density, it was among the first gases to escape. Consumption by microbial or inorganic oxidation of the gas may explain the depletion in CH 4 that is noticed in the borehole gases collected after the perforation [ (13) and (26)]. Indeed, N 2 /O 2 ratios of these samples and also of samples (17) and (30) are similar to air. This indicates that the atmospheric component of meteoric water has been modified by redox reactions that took place either in the subsoil or in the aquifer (D'Alessandro et al., 2010).

Relation Between Fluid Parameters and CO 2
Mean bubble fraction and water temperature at F1 are presented in Figure 4, along with the wellhead pressure at F2 and the meteorological conditions in the HMF over the drilling period. The mean bubble fraction was calculated according to Fischer et al. (2020), who showed that in boreholes and narrow tubelike mofettes, gas bubble contents in a water-gas mixture can be quantified from the pressure difference over a fixed depth interval. The high gas/water ratio is expressed as an increase in the gas flow and the high bubble fraction. This may also FIGURE 2 | (A) N 2 -CO 2 -CH 4 ternary diagram after Giggenbach et al. (1993) and (B) N 2 -He-Ar ternary diagram after Giggenbach et al. (1983) of the gas emitted in the HMF. The abbreviation _AC stands for air contamination, while data of samples collected during 2016 are from Bräuer et al. (2018). cause higher δ 13 C CO2 values as recognized in samples (4) and (6) (Figure 5). It might also be responsible for the depletion of the less soluble gases observed in samples (2) and (10) (Figures 4A,D).
Elevated gas flow and bubble fraction values in F1 correspond to depleted atmospheric components in the borehole and vice versa (Figures 4A,D). Provided that the concentration of atmospheric gases in the groundwater is constant, this observation indicates that mixing with a component of geogenic origin modifies the ratio. Moreover, low gas/water ratios (i.e., less intense bubbling) favor the dissolution, and hence the loss of CO 2 to the water (Reid et al., 1987). At the same time the limited ascent of geogenic gas can be contaminated more easily. The lack of gas flow and bubble fraction data at F2 and the mofette does not allow us to arrive at a similar conclusion for these locations.
Perturbations on the bubble fraction and the gas flow (F1) and on the wellhead pressure (F2) were observed when F3 was at a depth of ∼38 m ( Figure 4D) and ∼110 m (Figure 4B), respectively. This indicates a hydraulic connection between the boreholes. This hypothesis was also supported by Woith et al. (2020), who documented a decrease in the gas flow and bubble fraction ( Figure 4D) on the 30th of August. At the same time, the Rn concentration of the F3 drill mud reached its maximum. It is worth mentioning that a similar observation to the hydraulic connectivity, though on a larger scale, was made by Kämpf et al. (2013). They identified two connected conduit systems (Bublák and Hartoušov) that were interpreted as highly permeable substructures inside the Počatky-Plesná Fault Zone (PPZ). This observation was further studied by Nickschick et al. (2015), who hypothesized the existence of pull-apart basin-like structures inside the PPZ, and it was tested by Kämpf et al. (2019), who instead identified en-echelon faults, which act as fluid channels to depth.
Barometric pressures (note inverted scale in Figure 4E) are negatively correlated to the gas flow and bubble fraction at F1 (Woith et al., 2020). Fischer et al. (2020) explained this by proposing that elevated barometric pressures contribute to the dissolution of CO 2 that in turn hampers degassing. Even though the water temperature seems to be mostly constant, its perturbations are negatively correlated to the gas flow and bubble fraction (note inverted scale in Figure 4C). Despite its minor importance, it should be mentioned that the fractionation of C isotopes increases with decreasing temperature (Bottinga, 1968).
The relation between gases of different solubilities and δ 13 C CO2 observed during the sampling period is presented in Figure 5. Gases collected when F3 reached a depth of ∼41 m [F1 (4)] and ∼125 m [F1 (6)] show a shift toward 13 C-enriched isotope values that may indicate fractionation due to preferential 12 C CO2 loss caused by phase separation. These phenomena have been observed for dissolved gases from the Apennines (Italy; Chiodini et al., 2000Chiodini et al., , 2013, the Southern Volcanic Zone of Chile (Ray et al., 2009), the Amik Basin (Turkey; Yuce et al., 2014), eastern Australia (Italiano et al., 2014b;Ring et al., 2016), and the Eastern Carpathians-Transylvanian Basin (Romania; Italiano et al., 2017). Moreover, the vigorous bubbling visible in the field and enhanced flow at F1  cf. Figure 4), suggest extensive degassing. This process promotes intense gas separation that subsequently causes isotope fractionation related to water-gas interactions.
Another possible explanation for elevated CH 4 contents and associated δ 13 C CO2 values are Fischer-Tropsch type (FTT) reactions, during which gaseous CO 2 becomes reduced to produce abiogenic CH 4 (Berndt et al., 1996;Horita and Berndt, 1999;Foustoukos and Seyfried, 2004). Details on FTT reactions and abiogenic CH 4 are provided in the Section "Potential Origins of CH 4 ."  Sample (35) presents a shift toward more negative δ 13 C values that may be related to a decrease in flow observed in the field. It should be also taken into consideration that the dissolution of CO 2 in the water results in CO 2 loss and lower δ 13 C CO2 values. Moreover, if the hydraulic connection is real, then the mixing of gases ascending from different depths may have had an impact on the δ 13 C and chemical composition. Overall the atmospheric components should decrease with increasing depth, while the geogenic components should increase. However, due to the limited amount of data and the standard deviation of the δ 13 C CO2 values, no conclusions regarding the isotopic changes can be reached in the present work.

Potential Origins of CH 4
In geothermal and volcanic systems, the CH 4 / 3 He ratio may differentiate possible methane sources (e.g., Welhan and Craig, 1983;Poreda et al., 1988); 3 He is mantle-derived, and the FIGURE 6 | Air-corrected ( 3 He/ 4 He) c vs. CH 4 / 3 He diagram (cf. Poreda et al., 1988). The abbreviation _AC stands for air contamination, while data of samples collected during 2016 are from Bräuer et al. (2018). CH 4 / 3 He ratio may reflect the proportions of crustal and mantle-derived methane (Poreda et al., 1992). In Figure 6, the ( 3 He/ 4 He) c ratios are plotted against the CH 4 / 3 He ratios of the samples collected in the HMF. The values of CH 4 should be considered as minimum estimates, because this gas can be involved in various post-genetic processes, and can also be consumed by methanotrophic bacteria (Murrell and Jetten, 2009;Kip et al., 2012;Gagliano et al., 2020). The CH 4 / 3 He values range from 1.6 × 10 4 to 8.9 × 10 4 , with two samples [(1), (33)] that reach up to 1.4 × 10 5 , and one sample [(12)] with only 9.3 × 10 3 . It should be noted that the highest [sample (1)] and the lowest [sample (12)] CH 4 / 3 He ratios are observed for samples with elevated and depleted CH 4 contents. Likewise, 3 He/ 4 He ratios were relatively constant apart from one sample collected in the mofette during the drilling period [(33), F3 was at ∼197 m depth]. This sample presented a lower isotope ratio, which was most probably due to a higher contribution of crustal He. Except for this single occurrence, no correlation between the 3 He/ 4 He ratios and the drilling depth of F3 was evident.
According to Poreda et al. (1988) and Sakata (1991), low CH 4 contents correspond to low CH 4 / 3 He ratios in CO 2 -rich magmatic gases. This suggests the common origin of CH 4 and He. Mantle-derived fluids typically have CH 4 / 3 He ratios between 10 5 and 10 8 , while those of crust-derived fluids are much higher, and range from 10 8 to 10 11 (Welhan and Craig, 1979;Giggenbach et al., 1993;Dai et al., 2005). The CH 4 / 3 He ratios for most of the gases collected in the HMF are lower than 10 5 (Figure 6). This indicates that CH 4 may be of volcanic/geothermal origin (Etiope, 2015). If this is the case, CH 4 may have been generated from mantle-and limestone-derived CO 2 or from either of them, according to chemical reactions like Fischer-Tropsch type (FTT) reactions: nCO + (2n + 1)H 2 = C n H 2n+2 + nH 2 O (2) (Craig, 1953;Berndt et al., 1996;Horita and Berndt, 1999;McCollom and Seewald, 2001), and in the presence of Fe-bearing phase catalysts: Here FeO and FeO 1.5 represent iron in the two oxidation states that are generally present in minerals of crustal rocks (Giggenbach, 1987). Both the FTT reactions and the presence of Fe-bearing phase catalysts are fundamental conditions for the chemical and isotope exchange between CO 2 and CH 4 and the establishment of an equilibrium.
In an attempt to determine the P-T conditions under which the gases were formed, the CO 2 /Ar and CH 4 /Ar ratios were used as geoindicators (Figure 7). Figure 7 was constructed assuming fugacities of H 2 O and CO 2 as described in Chiodini et al. (2007). Redox conditions were defined according to D'Amore and Panichi (1980). Excluding the samples affected by air contamination, most samples cluster at T from ∼100 to ∼120 o C and P CO2 from ∼30 to ∼70 bar. These values can be considered as medium to low P-T conditions that promote the generation of thermogenic CH 4 (Hunt, 1996). Some samples [10), (13), (17), (26), (30), (54)] are characterized by even lower P-T conditions, indicating that the impact of the atmospheric end-member is not negligible in this case. The elevated heat flow values (70-80 mW/m 2 , locally as high as 90 mW/m 2 ; Cermák, 1994) recognized in the area may have contributed to the depleted light hydrocarbon content. It is worth noting that FIGURE 7 | Plot of log(CH 4 /Ar) vs. log(CO 2 /Ar) after Chiodini et al. (2007). The theoretical P CO2 -T grids assume that the redox conditions are fixed by the D'Amore and Panichi (1980) buffer. The abbreviation _AC stands for air contamination, while data of samples collected during 2016 are from Bräuer et al. (2018). the majority of the gases have values of CH 4 /(C 2 H 6 +C 3 H 8 ) (Bernard ratio;Bernard et al., 1978) in the range of 9 to 93. According to Bernard et al. (1978), these ratios characterize gases of thermogenic origin that are produced by decay of organic matter. If this applies to the gases of the HMF, samples (1), (6), (18), (22), (23), (24), (25), (38), and (39) with Bernard ratios between 100 and 1,000 may reflect secondary processes that stem from inorganic or organic oxidation of CH 4 or mixing of microbial and thermogenic CH 4 . Methanogenic archaea detected in various mofette waters in the Cheb Basin (Krauze et al., 2017) may support a biogenic origin of at least parts of the CH 4 present. Moreover, Liu et al. (2018) suggested that the anaerobic and acidophilic taxa identified in soils and sediments collected from two 3-m cores in the HMF may perform methanogenesis under reducing conditions. Both scenarios seem plausible, however, we cannot confirm or dismiss either of them because the rather low CH 4 concentrations ( Table 1) did not allow the determination of δ 13 C CH4 and δ 2 H CH4 values. Thus, any conclusion on the origin of the gas may be misleading. However, the extraordinary low CH 4 / 3 He ratios may point to a limited presence of organic matter in the overall process of CH 4 formation. Together with the medium to low P-T conditions, it can be hypothesized that CH 4 in the HMF originates from mixing of volcanic/geothermal and biogenic CH 4 with a non-negligible atmospheric contribution. This hypothesis seems plausible based on the sedimentary formations that characterize the HMF, together with its elevated heat flow values (Čermák, 1994), and the mantle-derived He and CO 2 (Weinlich et al., 1999;Kämpf et al., 2013;Bräuer et al., 2018; this study Tables 2, 3). Procesi et al. (2019) summarized that these are some of the criteria that describe sediment-hosted geothermal systems. Such hybrid geological systems exhibit both volcanohydrothermal and sedimentary features, in which geothermal and sedimentary domains interact and yield mixtures of inorganic and biogenic gases.

CONCLUSION
The present study investigated processes that affected gases of the HMF in a period when a third borehole was drilled, and no seismic event took place. Changes in CO 2 concentrations are mainly attributed to different solubilities of the gases. This was combined with dilution caused by an enhanced content of atmospheric components. Perturbations on the gas flows, dissolution, and likely FTT reactions had effects on the isotope compositions of CO 2 . Moreover, data from the wellhead pressure of F2 as well as from the gas flow of F1 support the assumption that the boreholes are hydraulically connected.
Variations in CH 4 contents indicated generation and consumption of this gas. Despite the lack of CH 4 isotope data, we suggest that CH 4 may be of mixed biogenic and volcanic/geothermal origin. In any case, it derives from sediments characterized by medium to low P-T conditions. It is unclear if it is solely derived from the mantle or also influenced by limestone-derived CO 2 under conditions in which FTT reactions and Fe catalysts are present. The drilling of the third borehole highlighted the complexity of the HMF in a seismically quiescent period. It opens the opportunity for further geochemical investigations in the area.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Materials, further inquiries can be directed to the corresponding author.

AUTHOR CONTRIBUTIONS
HW, MZ, SN, and JB developed the conception and design of the MoRe project. KD was responsible for the gas sampling, while all the authors assisted in the field campaigns. AV-H was responsible for the δ 13 C CO2 analyses, while SN was responsible for the noble gas analyses. JV was responsible for the gas flow, pressure, and temperature measurements. KD wrote the first draft of the manuscript. All authors contributed to manuscript revision, read and approved the submitted version.