Electrochemical survey of electroactive microbial populations in deep-sea hydrothermal fields

Abstract

Electric discharge in deep-sea hydrothermal fields leads us to expect the existence of electroactive microbial ecosystems in the environments. Electrochemical properties such as electric field distribution on the seafloor and electrical conductivity of the rock can be useful indicators of searching electroactive microbial community in natural environments. We performed electric field measurements in deep-sea hydrothermal fields and collected rock samples by a remotely operative vehicle (ROV) operation. Several spots on the seafloor with strong electric fields were detected, which included both active hydrothermal vent areas and inactive sulfide deposits far from the vents. The electrical conductivity of the rock samples was correlated with the copper and iron sulfide content. Microbial community compositions of the rock samples were characterized by small subunit (SSU) rRNA gene amplicon sequencing analysis. The abundance of several microbial components, which are highly related to electroactive microorganisms such as Geobacteraceae and Thiomicrorhabdus, was affected by the electrical properties of rock samples. The results suggested that electrochemical properties on the seafloor would be related to the abundance of possible electroactive microbial populations, and that the electrochemical survey may be a powerful tool for exploring electroactive ecosystems.

1 Introduction

Microbial electrosynthesis refers to the electrochemical reduction of carbon dioxide to produce multiple organic compounds and is an attractive technique for the effective industrial production of valuable organic compounds (Nevin et al. 2010). The term electrosynthesis or electroautotrophy is also used as a counterpart concept to photosynthesis and chemosynthesis and is considered to the 3rd energy conversion system of ecosystem using electricity as the energy source rather than light or reduced chemical compounds to fix CO₂ (Joen et al. 2012; Ishii et al. 2015). However, the extent of contribution of these electrosynthetic microbial populations and functions to the whole microbial communities is highly unclear in natural environments. Although electricity is thought to occur universally in various environments, such as in redox boundaries in soils and sediments, it is difficult to distinguish between electrotrophic and chemotrophic metabolisms and populations in the quite limited space and the fragile physical and chemical conditions.

Deep-sea hydrothermal fields are known as natural power plants after the discovery of the spontaneous electric discharge (Yamamoto et al. 2017). Namely, differences in redox potential and temperature between the hydrothermal fluid beneath the seafloor and the seawater bordered by conductive sulfide mineral deposits generate electron flow from the subseafloor to the seafloor surface through the sulfide rocks (Nakamura et al. 2010; Yamamoto et al. 2013; 2017; Ang et al. 2015). Electric discharge is widespread on the seafloor in hydrothermal fields (Yamamoto et al. 2017), and it has been expected that an electrosynthetic microbial ecosystem, or at least electrosynthetic microbial populations, can be sustained on the electrically charged seafloor environments (Nakamura et al. 2010; Yamamoto et al. 2018). In fact, our previous study successfully enriched a putative electrosynthetic bacterium, “Candidatus Thiomicrorhabdus electrophagus” during electrochemical microbial cultivation with an active artificial vent in the deep-sea hydrothermal field using the natural electromotive force (Yamamoto et al. 2023). It was the first example showing that electrosynthetic microorganisms are latent in deep-sea hydrothermal fields and can be dominant when suitable habitats are formed.

It is well known, however, that chemosynthetic populations are generally dominate in microbial communities near the hydrothermal vents because hydrothermal fluid emission mixes with the surrounding seawater, resulting in the formation of a nonequilibrium redox environment with abundant reductants and oxidants (Nakamura and Takai 2014). Thus, even in the deep-sea hydrothermal fields, it is difficult to detect electrosynthetic populations and functions near active vents since the predominant chemosynthetic populations and functions would mask the electrosynthetic signatures. On the other hand, the hydrothermal fluid discharging zones and activities are specially and temporally limited in the whole deep-sea hydrothermal field as compared to the electrically charged zones (Yamamoto et al. 2017). Thus, possible preferred habitats to the electrotrophic populations and metabolisms may be spatially and temporally more widespread than those to the chemolithotrophic ones in the whole spaces and lifetimes of deep-sea hydrothermal fields.

Electrochemical survey to determine strong electrically discharged areas in the deep-sea hydrothermal field may be an aid to detect possible electrosynthetic or electroactive microbial communities. It has been reported that the deep-sea hydrothermal fields have anomalous distributions of electric fields (Kawada and Kasaya 2017). The generation of electric current caused by conductors in the redox gradients such as subseafloor sulfide deposits are called “geobatteries”, which result in a negative potential on the seafloor surface (Bigalke and Grabner 1997). An electrochemical survey called the self-potential (SP) method can be used to illustrate the electric field above the seafloor via the potential difference variation in seawater (Heinson et al. 2005). Geobatteries were assumed in deep-sea hydrothermal deposits both with and without active vent sites using the SP method (Kawada and Kasaya 2017). The electric field survey with the SP method may also be effective in exploring the yet-unidentified electrosynthetic or electroactive microbial communities depending on the geobatteries. We hypothesize that clearer signs of electrosynthetic or electrotrophic populations and functions should be detected in the seafloor areas that are more electrically charged but more distant from the influence of hydrothermal fluid emissions. In this study, we conducted in situ electric field measurements of hydrothermal fields and sampling of sulfide rocks in the strongly charged seafloor by remotely operative vehicle (ROV) operation, and tried to examine the abundance of possible electrosynthetic or electrotrophic populations by the small subunit (SSU) rRNA gene amplicon sequencing analysis.

2 Methods/experimental

2.1 Deep-sea research expedition

From the 1st to the 13th of November 2018, a research cruise (KR18-14 Leg 2) was conducted by the research vessel Kairei (JAMSTEC) and the ROV Kaiko MK-IV (JAMSTEC) in three deep-sea hydrothermal fields in the mid-Okinawa Trough (MOT). The three hydrothermal fields referred to as Field A, B and C in this work were the original site of Iheya North field (Kawagucci et al. 2011), the JADE site of Izena Hole field (Halbach et al. 1989) and the Ieyama field (Ishizu et al. 2022), respectively (Fig. 1). Seven dives were performed at three hydrothermal fields for this study (Fig. 2 and Additional file 1: Table S1). To measure the conductivity, temperature and depth in seawater, a conductivity, temperature, and depth (CTD) sensor (SBE 49 FastCAT, Sea-Bird Scientific, WA, USA) was installed on the ROV during the dive. A dissolved oxygen (DO) sensor (RINKO II, JFE Advantech Co., Ltd., Hyogo, Japan) was also installed to measure the DO concentration in seawater (Additional file 1: Fig. S1).

Fig. 1

Survey area of this research cruise. The red circles indicate the deep-sea hydrothermal fields surveyed in this study. Fields A, B and C are the original site of Iheya North field, the JADE site of Izena Hole field and the Ieyama field, respectively

Fig. 2

Electric fields in deep-sea hydrothermal fields. Dive track of ROV and electric field intensity along the track are shown.  A, B and C show the results at Fields A, B and C, respectively. Yellow triangles indicate location of active hydrothermal vents with hydrothermal fluid emissions. Green circles indicate the locations at which the rock samples were collected, and blue number means the sample number. The number on a contour line shows the water depth

2.2 Dive tracking

The dive track for each dive is constructed using the acoustic transponder data of the ROV. The original data, acquired at a 2-s interval, contains scattering and may be shifted each other or even during a single dive. The data are thus processed through the following steps. Firstly, a cosine low-pass filter of 180 s (for dive KK#804 with severe scattering) or 60 s (for all other dives) is applied using ‘filter1d’ and ‘sample1d’ of the Generic Mapping Tools (GMT; Wessel et al. 2019), which smooth the data as well as interpolating them to a 1-s interval. Second, the data are trimmed to match the time interval of the corresponding dive. Finally, deviations among dive tracks that encountered the same location in video logs are estimated by comparing the smoothed data of the dive track involved, and then, the dive tracks are shifted with respect to one of them so that they point to the same value at the landmark. Although we have found that dive tracks deviated from location to location, the smoothed data are used as such for field-scale plots. On the other hand, for close-up plots around sampling points, shifts in the horizontal directions among the corresponding dive tracks are determined for each of close-up plots. The shift is up to 30 m for each direction.

2.3 Self-potential (SP) method

We conducted continuous measurements of the electric field using five pairs of silver/silver chloride nonpolarized electrodes for deep-sea use (Filloux 1973) connected to a multi-channel (5 ch) and high-precision (24 bit) electrometer (V-CUBE, Clover Tech. Co., Ltd., Tokyo, Japan; Kasaya et al. 2019) that was originally developed for the receiver of an active electromagnetic survey system. The acquired data are in 50 Hz in the present study but are reduced to 1 Hz in the analysis. In this work, we employed the data for a single pair of electrodes that were mounted vertically with 0.555 m distance on the starboard of the ROV (Additional file 1: Fig. S1). For each dive, the 1-Hz reduced data, the electrostatic potential between the electrodes, is processed using the following steps similar to those applied to the acoustic positioning data: Firstly, the data are trimmed to the dive track; second, a cosine low-pass filter of 15 s is applied using ‘filter1d’ of GMT (Wessel et al. 2019), and a linear trend with a constant offset are removed using least-squares fitting, and; finally, the electrostatic potential is divided by the distance between the electrodes and a negative sign is added to make it the vertical electric field (Fig. 3 and Additional file 1: Fig. S2). Note that the vertical electric field in the present study is defined as upward positive. With this definition, if the seafloor is flat, a negative (positive) vertical electric field predicts that a negative (positive) current source at or below the seafloor. Thus, a typical geobattery that forms a negative potential near the seafloor may give a negative vertical electric field (e.g. Kasaya et al. 2021; Yamamoto et al. 2017).

Fig. 3

An example of the electric field anomaly observed from 10:17:30 to 10:25:00 during dive KK#804. The two upper left panels indicate the time course of the electric field (EF) and the ROV altitude from seafloor. The lower left panel indicates the electric field in color along the ROV track. The right photos show the seafloor at representative time

2.4 Rock sampling

Rock samples were collected from the seafloor using a manipulator of the ROV and were placed in sample boxes on the basket of the ROV. On the shipboard, the rock mass sample was divided into several pieces. Some pieces randomly chipped from the surface of the rock sample were stored at − 30 °C for DNA extraction, and remaining lumps of the rock sample were dried and stored at room temperature for petrographic observation, chemical analysis, and electrical conductivity measurements.

2.5 XRD analysis

Some pieces randomly chipped from the surface of the rock sample were used for X-ray diffraction (XRD) analysis. The pieces were crushed and pulverized using an agate mortar and pestle. Diffraction data were acquired using a MiniFlex II instrument (Rigaku Corporation; Tokyo, Japan) with a Cu source, a 30 kV generator voltage, and a 15 mA current. The XRD operating conditions included stepwise scans from 3° to 90° 2θ in 4,350 steps at 2° 2θ/min with a 1.25° divergence slit and 0.3 mm analytical slit. The diffraction data were analyzed using the manufacturer’s diffraction evaluation software (PDXL II) in combination with a crystal database from the International Centre for Diffraction Data (https://www.icdd.com/).

2.6 ICP‒MS analysis

Bulk major and trace elemental analysis was performed by inductively coupled plasma‒mass spectrometry (ICP‒MS) on an Agilent 7500ce instrument. Powdered rock samples used for XRD analysis were subjected to ICP‒MS analysis according to the HF-HNO₃-HClO₄ digestion method and appropriate dilution (Nozaki et al. 2021). The details of these analytical procedures, including instrumental drift and mass interference correction methods, have been described previously (Takaya et al. 2018).

2.7 DNA extraction

Several pieces of rock sample (~ 5 g) were collected in a plastic tube, and RNA (200 mg from yeast, Roche) was added in to increase the DNA yield during extraction (Ikeda et al. 2008). The rock pieces were broken using a Multibeads Shocker (MB2000, Yasui Kikai, Osaka, Japan). DNA was extracted from the broken rock sample using a DNA extraction kit (Power Max soil DNA isolation kit, Qiagen, Venlo, Netherlands). During the DNA extraction process, RNA was subjected to ribonuclease degradation (4 µL of 10 µg/µL solution; Nippon Gene Co. Ltd., Tokyo, Japan) and incubated for 30 min at 37 °C. The obtained DNA solution was purified through isopropanol precipitation.

2.8 Microbial community composition analysis

Sequences of the SSU rRNA gene amplicons were obtained according to the previous method (Yamamoto et al. 2023). The resulting sequences were analyzed using the QIIME2 v2019.4.0 pipeline (Bolyen et al. 2019). The generated operational taxonomic units (OTUs) at the 97% identity were assigned to taxa using the SILVA 138.1 database.

2.9 Multivariate analysis

The table of elemental abundance in the rock samples based on Fig. 4 was used for multivariate analysis. The OTU count profiles were converted to the proportion for each sample. The environmental data (water depth, temperature, temperature increase, salinity, DO concentration, electric field, and conductivity) shown in Table 1 were also used for the analysis. The conductivity was logarithmically transformed. Ordinations of the nonmetric multidimensional scaling (NMDS) and canonical correlation analysis (CCA) were calculated with the Bray‒Curtis distances using the vegan package in the R (http://www.R-project.org).

Fig. 4

Electrical conductivity and elemental composition of the rock samples. The left panel indicates the electrical conductivity of the rock sample (orange). The error bar represents the standard error (1SE). Only the conductivity of the A4 sample is shown along the 10 times higher axis (cyan). The right panel indicates the weight percent of atoms included in the rock sample, which was measured via ICP‒MS

Table 1 Rock sample list

3 Results

3.1 Electric field surveys in deep-sea hydrothermal fields

We conducted expeditions at three deep-sea hydrothermal fields (Fields A, B, and C) in the MOT (Fig. 1 and Additional file 1: Table S1), where previous researches reported the SP anomaly associated with sulfide minerals geobatteirs. Fields A and B are well-studied hydrothermal fields in the MOT, where sulfide mineral deposits have been reported to be widespread on the seafloor (Kawagucci et al. 2011; Halbach et al. 1989). It was reported that electric field anomalies were detected using the SP method on the geobatteries formed in the deep-sea hydrothermal sulfide deposit (Kasaya et al. 2020). Field C is a deep-sea hydrothermal field recently discovered using the SP method, and sulfide deposits forming geobatteries were predicted under the seafloor (Ishizu et al. 2022). Dive surveys were performed seven times for electric field measurements and rock sampling at the three hydrothermal fields. Electric field measurements were performed along the dive tracks using the SP method (Fig. 2 and Additional file 1: Fig. S2). Strong electric fields with negative (blue) and positive (red) were observed alternating over a narrow range of several meters not only around active hydrothermal vents but also in some areas away from hydrothermal vents. A typical example is taken from dive KK#804 of Field A (Fig. 3), during which the ROV climbed a gentle slope of an altered seafloor in the NW direction. The electric field anomalies do not necessarily coincide with hydrothermal fluid signatures. For example, at 10:18:15, a white-colored seafloor shows no electric field anomaly (Fig. 3 and Additional file 1: Fig. S3). At 10:21:00 and 10:22:15 with similar seafloor characteristics, the former is accompanied by a negative electric field but the latter is not. At 10:22:45 and 10:23:15, a similar situation occurs, in which one gives a negative electric field and the other does not. In addition, although the electric field anomalies appear to be somehow related to the ROV altitude, they clearly change independently of the altitude (Fig. 3). For example, the negative anomaly observed at around 10:22:45 recovers at 10:23:15, but the altitude remains almost constant. To conclude, the electric field anomaly is not simply influenced by surface hydrothermal fluid signatures or the ROV altitude but is rather reflected probably by subsurface features such as the geobattery. Kawada and Kasaya (2017) reported that the SP method can be used to detect subseafloor sulfide body concealed under the sediment without active vents as geobatteries with SP anomalies in a wide area (km range) survey. In this study, our results indicated that the SP method is effective at detecting hidden geobatteries within a small area (100 m range), even at a mound containing active vents. This result is consistent with the previous work in which the redox potential at the seafloor surface was measured in a deep-sea hydrothermal field (Yamamoto et al. 2017).

3.2 Analyses of the mineral and element components and conductivity of rock samples

One of the major factors in the formation of geobatteries is the presence of conductive solid materials such as sulfide minerals located in a vertical redox gradient from the seafloor to the subseafloor. Microbial species that can utilize the geobattery as an energy source are expected to inhabit on the surface of conductive rocks. We attempted to collect rock samples to ascertain the presence of these microbial species. Seventeen rock samples were collected from three deep-sea hydrothermal fields through the seven dives (Additional file 1: Fig. S4 and Table 1). A portion of each sample was crushed into powder, after which the constituent minerals (Additional file 1: Fig. S5 and Table S2) were identified by XRD and elemental compositions (Fig. 4) were analyzed by ICP‒MS. The constituent minerals and their modes of occurrence tended to vary from field to field. Most of the rock samples collected from Field A contained abundant sulfide minerals, such as sphalerite (ZnS), wurtzite (ZnS), galena (PbS), pyrite (FeS₂), and chalcopyrite (CuFeS₂). Only the A3 rock sample comprised quartz (SiO₂) and illite (KAl₂AlSi₃O₁₀(OH)₂). The samples from Field C contained sulfide minerals, which were mainly sphalerite and galena. On the other hand, the samples from Field B mainly comprised sulfate mineral of barite (BaSO₄).

The conductivity significantly varied among the rock samples (Fig. 4). The A4 rock sample from Field A showed by far the highest conductivity in this study. The A6 and A2 samples exhibited relatively high conductivity, although their conductivity was one digit lower than the A4 sample. A piece of the A4 rock was polished and observed by a microscope under the reflected light, revealing that covellite (CuS) was commonly present on the rock surface (Fig. 5). Covellite is known as a sulfide mineral with high conductivity matching that of metals. Covellite is often formed during the oxidation of chalcopyrite (Todd et al. 2003). The B4 rock showed higher conductivity even though it was mainly composed of barite. Since covellite was detected in sample B4 by XRD analysis, this feature seems to contribute to the conductivity of the rock. Other rock samples exhibited lower conductivities. Sample A3, which was comprised of quartz, exhibited the lowest conductivity in this study.

Fig. 5

Observation of the highly conductive rock sample. The A4 rock sample was the most conductive rock in this study. A Complete rock A4 collected from the seafloor. B Polished piece of the rock for microscopic observation. C A microscopic image under the reflected light. Cp: Chalcopyrite (CuFeS₂), Cv: Covellite (CuS), Gn: Galena (PbS), Py: Pyrite (FeS₂), Sp: Sphalerite (ZnS)

To estimate the relationship between the conductivity and the bulk chemical components of the rock samples, we performed multivariate analysis using the NMDS method (Additional file 1: Fig. S6), in which a small distance between each plot position indicates the similarity of the elemental compositions between samples. The plot positions of the Field A samples were close to each other except for that of the A3 sample. For the samples from Fields B and C, the plot positions of the samples from the same field were close to each other, but B4 and C5 were slightly deviated from the field groups. The plot positions of the highly conductive rock samples, A4, A6 and A2 coincided with the directions of the vectors for copper and iron, indicating that these elements contribute to conductivity. The Cu concentration strongly correlated with and electrical conductivity of the rock (R² = 0.54). However, any correlation was not found between rock electric conductivity or Cu concentration with the electric field where the rock sample collected.

3.3 Microbial community composition analysis of rock surfaces

We extracted genomic DNA from the rock samples and analyzed the potential microbial compositions based on SSU rRNA gene amplicon sequencing data (Fig. 6A). Members of Gammaproteobacteria and Bacteroidota were the major components in all samples. In several samples, such as A4, A6, C2 and C3, which were collected from bottom parts of active vent chimneys, the abundance of Nitrospirota members was greater than other samples. In several samples, such as A1, A2, A3, A5, C1 and C4, the abundances of Campyrobacterota, Chloroflexi and Desulfobacterota members were greater than that in the other samples. Since Campyrobacterota members are known to be the predominant primary producers in chemosynthetic microbial communities in deep-sea hydrothermal fields, especially near the vent fluids (Nakagawa et al. 2005; Yamamoto and Takai 2011), these samples were considered to be strongly influenced by the hydrothermal fluid emissions.

Fig. 6

Microbial community compositions of the rock samples. (A) Microbial community compositions based on the SSU rRNA gene amplicon sequencing data. The composition was represented by the phylum level classification basically. The letter color of rock samples correspond to the grouping in the clustering analysis in panel B. (B) NMDS analysis of the microbial community composition. The 17 rock samples are plotted with circles. The color in the circle indicates the conductivity of the rock sample. Four color lined ovals indicate the groupings from the clustering analysis using the k-means method. Con: electric conductivity of the rock, Dep: water depth, EF: electric field, and TI: temperature increase in ambient seawater

To estimate the correlation between the microbial community composition (SSU rRNA gene phylotype composition) in the rock samples and environmental factors, we performed NMDS analysis using the OTU data and environmental condition data (Fig. 6B). In the ordination, microbial community composition could be divided into four groups. The group including A1, A3 and C4 samples seemed to be strongly influenced by hydrothermal fluid emissions due to the abundance of Campyrobacterota members, and the increase in ambient seawater temperature. The group composed of B1 to B4 samples seemed to be clustered by the specific geological setting of Field B such as water depth. Field B is located on the deeper seafloor than Fields A and C, and the direction of water depth vector represented for the group. Microbial community compositions of highly conductive rocks (A4, A6, and A2 samples) were clustered in a group that included those of other rock samples with the lower conductivities. Since the relationship between the electrical properties such as the electric field and conductivity of the rock on the microbial composition was previously unclear, we performed CCA to determine the response of microbial community composition to environmental factors (Fig. 7). The microbial compositions were clustered into five groups by direct gradient analysis according to three environmental factors (water depth, temperature increase, and electrical conductivity), including the groups with very high conductivity (A4) and moderately high conductivity (A2, A6 and A7). The indicator species (phylotypes) characterizing the microbial communities in the clustered groups were determined using the IndVal method (Dufrene and Legendre 1997) with the following thresholds: abundance ratio ≥ 1%, P value ≤ 0.05. Table 2 shows the indicator species in the highly conductive rock groups, including 16 OTUs in the highly conductive rock group (A4) and 4 OTUs in the moderately high conductive rock group (A2, A6 and A7).

Fig. 7

CCA of the microbial composition of rock samples. Three environmental factors (Con: electric conductivity of the rock, Dep: water depth, and TI: temperature increase) were used for direct gradient analysis. The 17 rock samples are plotted with circles. The color in the circle indicates the conductivity of the rock sample. The dots indicate the OTUs, and the color indicates the maximum abundance ratio

Table 2 Indicator species on the electrically conductive rocks

4 Discussion

4.1 Electric fields formed in deep-sea hydrothermal fields

In this work, we carried out an electric field surveys in three deep-sea hydrothermal fields and found several seafloor points that are characterized by strong electric field but distant from the influence of hydrothermal fluid emissions (Fig. 2). Electric fields are generated around hydrothermal fluid emissions due to the local geobatteries caused by electrochemical redox interactions between minerals and fluids on the surfaces (Nakamura et al. 2010; Yamamoto et al. 2017). However, these conditions can be attained not only around active vents but also in the distant spots electrochemically connected to the subseafloor hydrothermal fluids as geobatteries (Fig. 3). In fact, the distribution of dispersed electrically discharging spots in several deep-sea hydrothermal fields was found in the previous studies (Kawada and Kasaya 2017; Yamamoto et al. 2017) and in this study as well (Fig. 8). Alternating negative and positive electric fields detected in a narrow range suggests that small geobatteries (several meters range) are formed due to the electrochemical conditions based on the geochemical settings under the seafloor, such as rock compositions and hydrothermal flow paths. These local geobatteries and electrically charged seafloor can prepare the preferred habitats to host more abundant electrotrophic and less abundant chemotrophic populations and metabolisms.

Fig. 8

Conceptual model of the study. Geobatteries are formed with electromotive force and highly conductive minerals in the seafloor of deep-sea hydrothermal fields. Chemolithotrophic microbial communities are predominant in mixing zone near active vents and hydrothermal fluid emissions. The relative abundance and dominance of electrotrophic populations may increase in the seafloor geobatteries that are more electrically charged but more distant from the influence of hydrothermal fluid emissions

4.2 Characteristics of the rock samples from deep-sea hydrothermal fields

On deep-sea hydrothermal vents, rock structures called chimneys are formed. During their formation process, anhydrite first forms as the initial backbone and gradually the anhydrite backbone is replaced with sphalerite or galena. Over the time, pyrite and chalcopyrite settle in the gaps (Haymon 1983; Janecky and Seyfriend 1984; Ohtomo 1996). The collected rock samples in this study exhibited different trends depending on the fields. Rocks from Field A, except for sample A3, contained abundant sulfide minerals with Zn, Pb, Cu and Fe (Fig. 4 and Additional file 1: Table S2), and where in particular characterized by high abundances of chalcopyrite (CuFeS₂) and pyrite (FeS₂). Rocks from Field C were rich in sphalerite (ZnS) and galena (PbS) but poor in chalcopyrite and pyrite. It has been suggested that microsized pyrite and chalcopyrite crystals in hydrothermal field rocks contribute to high electrical conductivity (Nakamura et al. 2010). There was a high correlation between the abundance of copper in the rock and the conductivity of the rock (see Results). In addition to the abundance of copper, the copper mineral composition also seemed to affect the conductivity. The A4 rock sample exhibited particularly high conductivity. Covellite was formed on the surface of the A4 rock (Fig. 5). Covellite shows very high conductivity, similar to metals, and is often formed during the oxidation of chalcopyrite in seawater (Todd et al. 2003). Therefore, the A4 rock is considered to have acquired very high conductivity because covellite was formed on the rock surface by the oxidation of chalcopyrite exposed to seawater. Covellite was also detected in the B4 sample via XRD analysis (Additional file 1: Fig. S5 and Table S2). The B4 rock showed higher conductivity, although the abundance of iron and copper was low (Fig. 4). Since the conductivity of covellite is extremely high, the formation of a thin layer of covellite on the surface of rock endows the rock with high conductivity. This suggests that the mineral formation due to environmental redox conditions is also important for determining conductivity.

Rocks from Field B were formed by mainly sulfate minerals such as barite (BaSO₄) and anhydrite (CaSO₄), known to show low conductivities (Guinea et al. 2012; Wypych 2021). The high abundances of barite in the rocks collected from the Field B were consistent with the previous results, indicating that the hydrothermal fluid temperatures were low (less than 200 °C) and that the metal ratio per unit of sulfur in the fluid was also low (Marumo and Hattori 1999). These results and observations suggested that the formation and longevity of high-temperatures hydrothermal fluid flow driven by the large and stable heat source would be coupled with the widespread distribution of geobatteries and high electric-conductive metal sulfide deposits, and even may be associated with the occurrence of suitable habitats for dominance of electrotrophic populations and metabolisms.

Although strong positive correlation is expected between the electric field and the rock electric conductivity due to the geobattery effect enhanced by high conductivity, no correlation between them was observed in this work. This may be due to low conductivity rocks covering the conductive rocks which form geobattery at subseafloor. In addition, there is also the issue of spatial resolution of the electric field measurement. It is probably because the electric field difference even within a small distance between the rock and the electrodes, and the measured value of electric field does not represent the true value of at the location of the rock. It suggests that electric field survey should be used for the first screening of hotspots of the electric field, or development of technology with higher spatial resolution is required.

4.3 Microbial community compositions formed on the rock samples

The microbial community compositions in the surfaces of the 17 rock samples were characterized by SSU rRNA gene amplicon sequencing. The microbial community compositions of several rock samples were characterized by the abundant populations of Campyrobacterota members and these rock samples were obtained from the seafloor zones, where the ambient seawater showed relatively higher temperatures based on the CTD data. The predominance of Campyrobacterota is likely influenced by the considerable input of hydrothermal fluid emission (Fig. 6). The previous study revealed that the negative redox condition with a higher hydrothermal fluid ratio per seawater would be advantageous for the energy metabolisms of Campyrobacterota than those of Gammaproteobacteria (Yamamoto and Takai 2011). Thus, it is verified in this study that the extent of hydrothermal fluid input to the environment (redox state) is one of the most important factors determining the microbial community development on the seafloor of deep-sea hydrothermal fields although the seafloor rock property varies in different sites, as previously theorized (Nakamura and Takai 2014).

Electric properties of rock samples seem to affect the surface microbial community compositions as well (Fig. 7). Statistical analyses indicated that several microbial members increased the relative abundances in the rocks with increasing Cu content and conductivity and were likely associated with the electrical conductivity of rock (Table 2). Some of them are phylogenetically related with the known electroactive microorganisms. Members of Geobacteraceae includes Geobacter are well known for their electroactivity (Lovley et al. 2011). One strain of Thiomicrorhabdus was enriched by in situ electrochemical cathodic microbial cultivation, and the metagenome-assembled genome (MAG), namely “Candidatus Thiomicrorhabdus electrophagus”, suggested that the strain has a putative extracellular electron transfer pathway and can grow electrosynthetically (Yamamoto et al. 2023). In addition, some members of Hydrothermarchaeales also exhibit potential for extracellular electron transfer due to the high number of extracellular multiheme c-type cytochromes (MHCs) encoded in the MAGs (Kato et al. 2019). It was also reported that related species of Bactrroidetes, Nitrosococcales, Nitrospirota, Ignavibacteriales, and Thermodesulfovibriona were possible electroactive bacteria (Iino et al. 2015; Ding et al. 2017; Caser et al. 2020; Yu and Leadbetter 2020; Gulay et al. 2023). Interestingly, we observed that a strain of the same species to “Ca. T. electrophagus” was enriched in laboratory using the electrochemical microbial cultivation chamber simulating the deep-sea hydrothermal discharge phenomenon with the A4 rock sample as the electrode and the microbial inoculum source (unpublished data). The microbial cultivation was performed under the electrosynthetic conditions using carbon dioxide and the rock electrode as the sole carbon source and the solo electron source, respectively. This result strongly suggests that at least one of the indicator species observed in the A4 rock sample in this study represents the electrosynthetic bacterium. Although there is no other direct evidence that these microbial populations responding to the strong electric field distribution and the high conductivity of rock are involved in the electroactive and/or electrosynthetic metabolic potentials and functions, these observations suggest that the microbial populations represent possible indicator species of electroactive and/or electrosynthetic microbial communities in the deep-sea hydrothermal fields.

5 Conclusions

In this work, we showed the first example of the relationship between the electrochemical properties and the possible electroactive microbial populations of the seafloor in the deep-sea hydrothermal fields. The electrical conductivity of the seafloor rock affected the microbial community composition, particularly the relative abundances of indicator species populations, potentially associated with electroactive and/or electrosynthetic metabolisms and functions. This study points out that a field scale of electrochemical survey seems effective as the first step of exploring electroactive and/or electrosynthetic communities in deep-sea hydrothermal systems (Fig. 8).

Improved techniques for electrochemical surveys will provide multiple lines of evidence to justify the relationship. To obtain higher resolution and accuracy of electric field measurements, the operational techniques of surveys such as keeping constant altitude and horizontal angle and real-time communication should be sophisticated. Next, the analytical aspects of improvement such as the chemical stability of electrodes and combined measurements of electric field, electrical conductivity and redox potential are required in future electrochemical surveys of exploring electroactive and/or electrosynthetic microbial communities in the deep-sea hydrothermal fields. In addition, we need the devising ways to collect and bring back a large number of rock samples separately from seafloor in various electric fields in order to better clarify the relationship between the geobattery effect and the microbial community. Furthermore, measurement of the chemical component concentrations in the seawater around the seafloor coupled with the electrochemical survey will provide a clear understanding of the energy source for the microbial communities formed on the seafloor. Once the applicability for searching electroactive and/or electrosynthetic microbial communities will be justified in the future, the electrochemical surveys may also be used for detection of extraterrestrial life and habitability in our solar system in the future astrobiological missions. This is because electricity can be an alternative energy source to sunlight and reductive compounds such as H₂, H₂S and CH₄, which have been considered as the exclusive energy source available to living organisms. Geobatteries can be formed in any environment with conductors and redox gradient, and so they are expected to be ubiquitous even on extraterrestrial bodies.