Degassing Rhythms and Fluctuations of Geogenic Gases in A Red Wood-Ant Nest and in Soil in The Neuwied Basin (East Eifel Volcanic Field, Germany)

Geochemical tracers of crustal fluids (CO2, He, Rn) provide a useful tool for the identification of buried fault structures. We acquired geochemical data during 7-months of continual sampling to identify causal processes underlying correlations between ambient air and degassing patterns of three gases (CO2, He, Rn) in a nest of red wood ants (Formica polyctena; “RWA”) and the soil at Goloring in the Neuwied Basin, a part of the East Eifel Volcanic Field (EEVF). We explored whether temporal relations and degassing rhythms in soil and nest gas concentrations could be indicators of hidden faults through which the gases migrate to the surface from depth. In nest gas, the coupled system of CO2-He and He concentrations exceeding atmospheric standards 2-3 fold suggested that RWA nests may be biological indicators of hidden degassing faults and fractures at small scales. Equivalently periodic degassing infradian rhythms in the RWA nest, soil, and three nearby minerals springs suggested NW-SE and NE-SW tectonic linkages. Because volcanic activity in the EEVF is dormant, more detailed information on the EEVF’s tectonic, magmatic, and degassing systems and its active tectonic fault zones are needed. Such data could provide additional insights into earthquake processes that are related to magmatic processes at the lower crust.


Introduction
The seismically active East Eifel Volcanic Field (EEVF) and its adjoining Neuwied basin have been the focus of many vulcanological, geochemical, petrochemical, and tectonic investigations. These have focused on dormant but not extinct volcanic activity [1][2][3]; the present-day NW-SE-directed compressional stress field and its related seismic activity [4,5]; gas composition and chemical tracers of mineral waters [6][7][8]; and mofettes along the Laacher See or at Obermendig [9]. Data collection and monitoring only has been annual, for example, [8] or short-term (4 days; [9]).
Geochemical tracers of crustal fluid for example, carbon dioxide (CO 2 ), helium (He), and radon (Rn) can identify buried fault structures in bedrocks [10,11]. Changes in soil gas concentrations reflect heterogeneities linked to soils or tectonic structures, for example, [9,12]. Faults and fracture networks from macro-to micro-scale are preferential pathways of lateral and vertical degassing, for example, [10,[13][14][15]. Important mechanisms driving fluid flow and keeping fractures open are compressive stress, volume changes of pore fluid or the rock matrix, and fluid movement or buoyancy [16]. Variation in structure along and between faults is common; damage intensity and thickness of the damaged zone vary laterally towards the core zone. Depending on positions of SG probes within the damaged zone, fluid flow and degassing may be limited (modified after [36]).
We found that NG appears to be associated with SG indicating fault-related micro-seepage of geogases; the degassing rhythm between the soil and nest is associated with degassing rhythms of three nearby mineral springs; and that degassing patterns are independent of earth tides and meteorological conditions. This study is part III of the research project "GeoBio-Interactions" in which we also monitored geochemistry of three mineral waters ≈6 km from the Goloring site [35] and the association of RWA nests and fault-related CH 4 degassing at Goloring [28].

Study Area
The Goloring site (Figure 2a), with its Iron age henge sanctuary (Figure 2b), is located southeast of the Laacher See volcano, and close to the Ochtendung Fault Zone in the seismically active Neuwied Basin, which is part of the Quaternary East Eifel Volcanic field (EEVF) in western Germany (Figure 2a). During the last 700 ka, intensive intra-continental Quaternary volcanism took place in the EEVF, with its youngest event being a phreato-plinian eruption of the Laacher See volcano ≈12,900 years ago. Today, the volcanic activity is dormant but not extinct [1,3]. Complex major tectonic and magmatic processes, such as plume-related thermal expansion of the mantle-lithosphere [2] and reactivation of Variscan thrust faults due to the present-day compressional stress field oriented in NW-SE direction affect the study area [5]. Weak to moderate earthquakes, which occur mostly in a shallow crustal depth (≤ 15 km) with local magnitudes (M L , Richter scale) rarely exceeding 4.0, are concentrated in the seismically active Ochtendunger Fault Zone (OFZ; Figure 2a; [4]). Berberich et al. [28] provide a more complete geological, tectonic, and volcanological description of the Goloring study site. Our Goloring study site is situated in the center of a triangle-shaped study area formed by the three previously-investigated mineral springs (Flöcksmühle in the Nette river near Ochtendung [hereafter: 'Nette'], Waldmühle in Mülheim-Kärlich [hereafter: 'Kärlich'] and 'Kobern' in Kobern-Gondorf (Figure 2a; [35]). No fault zones had been reported and identified previously from the Goloring study site; local earthquakes magnitude never exceeded M L = 2, and focal depths of earthquakes near it never exceeded 28 km during our sampling campaign [37].

Gas Sampling and Geochemical Analyses
We measured gases in ambient air (AA), soil (SG; Figure 2d), and the RWA nest (NG) at the Iron age henge sanctuary at the Goloring site ( Figure 2b) biweekly from 1 March-30 September 2016 (7-M; 16 times) and every eight hours between 12 July and 11 August 2016 (4-W; 83 times), yielding a total of 2673 gas samples. Gas sampling followed procedures described by Berberich [24] using analytical equipment and sampling methods described by Berberich et al. [28,35]. Briefly, AA was sampled at 1 m height, 1 m away from the RWA nest. The stainless-steel RWA nest-gas probe (Figure 2b; inner diameter 0.6 cm), equipped with a flexible tip attached to a pushable rod and a sealable outlet for docking sampling equipment, was inserted 1 m into the RWA nest. It remained there, unmoved, during the entire 7-M (including 4-W) sampling campaign. Before the start of the sampling in March, the probe was evacuated twice by pushing the rod using a 20-mL syringe. After this, the outlet was closed to prevent atmospheric influence. Thereafter, the outlet was opened only after docking the sampling unit to it. Seven permanent soil gas probes (Figure 2c), in locations chosen on information from previous investigations, were installed to 1 m depth, either ≈ 2 m (SG1), 30 m (SG 2 and 7), or 60 m (SG 3-6) from the RWA nest (Figure 1c; Hinkle 1994). Occurrences of maximum helium anomalies (> 11 ppm) in SG3, SG4, and SG6 ≈ 60 m away from the nest could be attributed clearly to operator error during analyses. The map (a) shows tectonic structures (black lines) and probability density of the earthquake events from 1977-2016 related to the Ochtendunger Fault Zone (OFZ; rainbow contours showing the hot spots (red color) of earthquake events within the OFZ rarely exceeding local magnitude of 4.0; modified after [28]). The inset shows the location of the study site within Germany. CO 2 , He, and Rn were sampled at the Goloring site with its Iron-aged henge sanctuary from a RWA nest, soil, and the ambient air; no fault zones had been reported and identified previously from the Goloring study site (b). Photographs show (c) the permanent nest gas probe (white arrow) and the RWA nest, (d) example of a permanent soil gas (SG1) probe with marked flag, and (e) the meteorological station (all photographs: G. M. Berberich).

External Factors
We used data on earthquakes, Earth tides, and meteorological conditions (Figure 2e) (published by Berberich et al. [28].

Data Analysis
All analyses were done using R version 3.3.2 (R Core Team 2016, www.R-project.org) or MATLAB R2017a (www.mathworks.com).
We used the "median + 2MAD" method [38] to separate true peaks in gas concentrations from background or naturally-elevated concentrations: any observation greater than the overall median + 2MAD was considered a peak concentration [38]. For interpreting the significance of the correlation coefficient, we followed Hinkle et al. [39].
Analysis of fluctuations followed Berberich et al. [35]; cross-correlation analyses were used to investigate temporal relations between degassing patterns of sampled springs and carrier-tracer gas relations.
Because meteorological variables were strongly correlated, we used principal component analysis (R function prcomp) on centered and scaled data to create composite "weather" variables (i.e., principal axes) that were used in subsequent analyses.
We used modified Fourier analysis (sampling rate = 8; Matlab 2017a) of the 4-W gas (NG, SG) and Earth-tide data [40] to test for temporal rhythms. Because the observation interval corresponds to an infinite signal multiplied by a rectangular window, a Blackman window [41] was applied to suppress the side lobes of the rectangular window. Because average degassing produced a large peak in the origin of the amplitude spectrum, this peak was removed to reveal any low-frequency components due to the main lobe of the window function [28].

Temporal Variations of Concentrations and Carrier-Trace Gas Couples in SG and NG
Cross-correlations were weakly positive between CO 2 in NG and SG mean (SG1-7) with a time lag of ≈1 day, and moderately negative between He in NG and SG mean with a time lag of ≈8 h ( Figure 4a). Cross-correlations between CO 2 and He in NG and AA were strong with a time lag of ≈4 days (Figure 4b). No cross-correlations were found for Rn among the samples.

External Factors
Stable meteorological conditions persisted during the campaign [28]. The degassing processes from NG and SG did not appear to be associated with meteorological conditions (Appendix A, Figures A1 and A2).

Gases in Ambient Air, Nest and Soil
This is the first time that AA, NG and SG samples have been monitored in parallel in a long-term survey in the Neuwied Basin. The continual bi-weekly and 8-hour sampling intervals generated a robust geochemical data set for the Goloring site. Prior geochemical analyses in the EEVF were based only on annual [8] and short-term surveys of soil gases (4 days, [9]; Appendix G).

CO 2
Ohashi et al. [32] and Risch et al. [30] found that RWA nests are point sources of CO 2 by measuring CO 2 fluxes from RWA nests at 10-cm maximum depth. Ohashi et al. [32] suggested that surficial CO 2 emissions from RWA nests originate from: (1) respiration processes of RWAs and other invertebrates within the nest; (2) root respiration by vascular plants within or beneath the nest; and (3) microbial decomposition of nest material. According to Hinkle [45], surficial gas samples from 0.0-0.8 m are influenced by the atmosphere. We took gas measurements at 1-m depth without atmospheric influence and the maximum NG concentrations (10.8 Vol. %; 2016) were comparable to others we took in June 2010 (≈15.0 Vol. %; Appendix G). CO 2 degassing (measured at 1-m depth) from 2-100 Vol. % can derive from deep fault zones and may be related to recent or post-volcanic metamorphic processes in carbonate rocks [46]. Our findings are comparable to ones of the Arabia Fault (8.2 Vol. %), and Terme S. Giovanni (18.1 Vol. %), a main thermal spring at the Rapolano Fault within the Neogene Siena-Radicofani Basin (Central Italy; [47]). We conclude that the NG results imply that the RWA nest is located above the fault core zone and indicate a degassing vent at the study site ( Figure 1).
Most of the SG concentrations are slightly elevated, but the very high levels in SG5 and SG1 indicate CO 2 anomalies (following [42]). Although higher ones have been reported (e.g., [9]), these are comparable to those recorded from the actively degassing Rapolano Fault [47] and therefore may be associated with an actively degassing but unknown fault on the Goloring site. Median concentrations were comparable to findings of actively degassing vents by Gal et al. [9] for the Laacher See pasture and Obermendig site, and to random samplings on the Goloring site (Appendix G).

He
He concentrations lower than the atmospheric standard of 5.22 ppm ( [18]; class I in Table 2) are considered to represent undisturbed background levels [48]; all other anomaly classes (II-V; Table 2) indicate tectonic influences.
He concentrations exceeding the atmospheric standard of 5.22 ppm ( [18]; Table 2) were twice as high in NG than in SG in the 4-W samples, lending further support for the RWA nest being located above the core fault zone (Figure 1). The high NG concentrations support the notion that RWA nests are useful biological indicators for degassing faults at small, local scales [24,25,28]. This conclusion is further supported by the temporal analyses that indicate a coupled system of CO 2 -He in NG. Tectonically active zones are known for high He fluxes through permeable fractures. Compressive stress and seismic activity maintain permeabilities and lead to gas anomalies at the surface [10,16,49].
Differences between NG and SG1 concentrations may be attributed to different soil characteristics, different basement geology [9], variation in structure, damage intensity, and thickness along and between faults [36] or even to an unknown fault separating both locations. SG1, SG5, and SG7 probably are located on a less developed fault segment, for example, in the damaged zone, so that there is less permeability for fluid flow and degassing (Figure 1). Faults exhibiting minor gas emissions are often synthetic faults that root into main faults [50].
It was not possible to verify the validity of He concentrations below the atmospheric standard or its concentrations in AA. Such a validation of these low values would require a follow-up study to assess sampling or analytical errors.

Rn
Although Rn concentrations at the Goloring site in AA and NG were within the background value (< 20 BqL −1 ; [43]), SG concentrations were up to 8-fold higher. Median SG concentrations differed up to 23-fold among sampling locations, confirming non-stationary variation in soil or bedrock at small scales [51]. In total, maximum Rn concentrations in SG at the Goloring site were up to 5-fold (7-M samples) or 2-fold higher (4-W samples) than those reported by Gal et al. [9] (Appendix G), indicating an increased Rn potential with a locally high potential (> 100 BqL −1 ) for the Goloring site [44]. Local high Rn anomalies (> 100 BqL −1 ) are associated with tectonic fault zones and clefts caused by advective gas transport along faults between the interbedding layers of Lower Devonian clay and siltstone bedrocks and the Cenozoic sediment basin fillings [52,53]. These concentrations also are comparable to hazardous sites along the Rapolano Fault [47]. The small-scale variability in Rn degassing at Goloring can be attributed to a linear fault-linked anomaly [10,13], suggesting a degassing in the NE-SW direction (Variscan direction) and the NW-SE direction (corresponding to the present-day main stress direction; [5]). Both accord with the local "Radon-Potential Map" [44]. Our data augment this "Radon-Potential Map", complement knowledge of Rn anomalies in the Neuwied Basin, and extend information on geogenic radon potential for this area.

Time Series
Bi-weekly NG and SG samples were more variable than 4-W samples in the three environments. This supports our conclusion that results derived from samples taken at long intervals may lead to erroneous conclusions [28]. For example, Griesshaber [54] and Clauser et al. [7] concluded from annual samples that CO 2 is the primary carrier in the Eifel fluid-rock system. Results of our higher-frequency samples cannot confirm this; we could only identify a CO 2 -He coupled system in NG and SG6 and a CO 2 -Rn coupled system in SG6 and SG7. Elsewhere, we hypothesized that geogenic gases in this part of the Neuwied Basin might be transported by another carrier gas, such as N 2 ( [28]; see also Bräuer et al. [8]). Bräuer et al. [8] found N 2 to be a carrier gas at the periphery along the Rhine. Future investigations throughout the Neuwied Basin should investigate the N 2 -He carrier-trace gas couple also in soil gas samples.

Meteorological Conditions
The degassing processes from NG and SG did not appear to be influenced by meteorological conditions (cf. [9,45,55,56]). Additionally, our results do not support the hypothesis that temperature is a dominant controller of CO 2 production (cf. [57]). The reason for these differences could be related to the frequency of measurements. We monitored them continuously on site where as others used daily values recorded at distant meteorological stations [9,57]

Earthquakes
We observed no effects of earthquakes on gas concentrations in the biweekly (7-M) samples. This is attributable to either: (a) the small size of the earthquakes; (b) their large distance (>10 km) from the Goloring site; or that (c) biweekly sampling intervals missed the influence of the small earthquakes. Alternatively, evidence of seismic influence on fluctuation patterns were observed after the earthquake that occurred on 3 August 2016 at Nickenich (≈11 km) while we were sampling every 8 h. The decline of CO 2 , He, and Rn concentrations in NG and SG observed ≈1 day before the earthquake can be explained by: (1) an increase in compressive stress; (2) volume changes of the pore fluid or rock matrix; or (3) a permeability change of conduits at the nest and soil sample locations [16,49]. As there was only one such nearby event, however, we cannot assert that there is a general relationship between earthquakes and nest or soil gas concentrations. The recent occurrences of deep earthquakes that are related to magmatic processes at the lower crust suggest continuous monitoring in this youngest volcanic field in Germany. Such data also would help assess relationships between gas flux dynamics and earthquake events in the Neuwied Basin.

Earth tides
Deformations of the Earth's crust by Earth tides are associated with cyclic variations in water-table levels within the rock strata and have been suggested to influence gas concentrations (e.g., Rn; [23,58]). Though all NG and SG probes were ≤ 60 m from one another, we only observed an effect of Earth tides on the fluctuation patterns in 25% of the probes for the 7-M samples. This result could be explained by: (1) the 8-hour sampling interval being too long to capture effects of semi-diurnal earth tides; (2) the study area being too far away from coastlines, ameliorating influences of Earth tides; or (3) in the case of Rn, the emanating layer being located too deeply so that any pumped Rn is diluted during migration [23].

Comparison with Mineral Springs Nette, Kärlich and Kobern
A comparison of NG and SG concentrations with previously-investigated gas concentrations in nearby mineral springs [35] showed similar median He concentrations in SG, NG, and Kobern (Appendix D). We observed highly or very highly (CO 2 ) to moderate (He) correlations in SG mean and concentrations at Nette, Kärlich, and Kobern (Appendix B, Figures A3 and A4). Similar median Rn concentrations were found for SG and Nette, Kärlich, and Kobern, suggesting degassing at Goloring site and the three springs are linked either by a similar Rn source in the subsurface or by an unknown fault system. CO 2 fluctuations in SG mean and Nette and Kärlich mineral springs are directly and instantaneously linked. Correlations of He between SG mean and Nette indicate a linkage in NW-SE direction and between SG mean and Kärlich an additional linkage in NE-SW direction (Appendix E). This linkage between SG at Goloring site and the minerals springs is supported further by cross-correlations (Appendix F): SG mean and Nette (lag ≈ 16 h) and Kärlich (lag ≈ 40 h) are either directly linked or positively correlated with respect to CO 2 . Moderate cross-correlation between SG mean and Nette for He was observed (lag ≈ 3 days). Relations between NG and Nette and Kärlich (lag ≈ 80 h) were small and positive for CO 2 and small and negative for He (lag ≈ 80 h to 5 days). NG and Kobern were related with a lag of ≈ 88 h for He. No to only low relations were observed for Rn between NG and the three springs (Appendix F).
Degassing rhythms in NG and SG were equivalently periodic and exhibited infradian rhythms of 2, 3, 4, and 6 days. The same infradian rhythms were found for the three mineral springs investigated [28]. These results support our conclusion that there are tectonic linkages between Goloring and Nette in the NW-SE direction (present-day stress field) and between Goloring and Kärlich in the NE-SW (Variscan fault direction) directions [28].
The volcanic activity in the EEVF is dormant but not extinct. Furthermore, information on active tectonic fault zones is missing in the EEVF and especially in the Ochtendunger Fault Zone. Monitoring of geogenic gases suggesting statistical bias when samples are taken at large temporal intervals. Therefore, we recommend daily soil gas samplings for a minimum of one year to understand-in combination with the recommended mineral water sampling-the EEVF's tectonic, magmatic and degassing system also in relation to new developments in earthquake processes which are related to magmatic processes in the lower crust.

Conclusions
Combined analyses of ambient air (AA), ant-nest gases (NG), and soil gases (SG) measured in situ from 1 March-30 September 2016 were evaluated to determine composition, fluctuation patterns, temporal variations, degassing rhythms, and carrier-trace gas couples of geogenic gases (CO 2 , He, Rn,) and compared to gas concentrations in three nearby mineral springs. Results of continual sampling during 7 months (bi-weekly) and 4 weeks (every 8 h) were: He concentrations in NG were above the atmospheric standard. A coupled CO 2 -He system supported the hypothesis that red wood-ant nests can be used as biological indicators for actively degassing faults.
Radon anomalies in SG with peak concentrations of 163 BqL −1 identified a high local Rn potential for the Goloring site and conributed to the Radon potential map of LGB-RLP 2017 [44].
Equivalently periodic degassing infradian rhythms in the red wood-ant nest, soil, and three nearby minerals springs suggested a NW-SE tectonic linkage between Goloring and Nette spring and a NE-SW tectonic linkage between Goloring and Kärlich spring.
Meteorology and low-magnitude local earthquakes did not modulate degassing at Goloring. Analyses of fluctuation patterns revealed that only 25% of the probes were affected by Earth tides. Earth tides were associated with soil degassing of CO 2 , He, and Rn only in biweekly samples, suggesting statistical bias when samples are taken at longer temporal intervals.
Because volcanic activity in the EEVF is dormant, more detailed information on active tectonic fault zones is needed in the EEVF, especially in the Ochtendunger Fault Zone. We recommend continuous monitoring of geogenic gases in soil and RWA nests-in combination with the recommended mineral water sampling and isotopic investigations-for a minimum of one year to understand the EEVF's tectonic, magmatic, and degassing systems in relation to new developments in earthquake processes that are related to magmatic processes at the lower crust.
Furthermore, electrical measurements, for example, between the tips of the probe in the nest on the fault and between other ones in the soil tens of meters away, could provide information about fluctuations of electronic charge carriers, which would be stress-activated at depths below, before, or during earthquakes.
Author Contributions: G.M.B. conceived the idea, designed the study, performed the field work, carried out the statistical analysis and wrote the manuscript. M.B.B. performed the field work, analyzed the data and contributed to the manuscript. A.M.E. and C.W. analyzed the data and contributed to the manuscript. All authors edited the manuscript and approved the final version.

Funding:
The study is part of the research project "GeoBio-Interactions" funded by the Volkswagen Stiftung (grant numbers Az 93 403 and Az 94 626) within the initiative "Experiment!" -Auf der Suche nach gewagten Forschungsideen. The Volkswagen Stiftung had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to publish the results.