Introduction
Over the past two decades, gold geomicrobiology research has focused primarily on bacterial microcosm experiments or on the characterisation of gold particles sourced from tropical to sub-Arctic environments (see Shuster and Reith, Reference Shuster and Reith2018 and references there in). Characterisation-based studies have included the (semi)quantitative assessment of particle structure/chemistry, detection of amplified DNA to infer bacterial presence, and more recently the enrichment of bacteria directly from particles (Reith et al., Reference Reith, Falconer, Van Nostrand, Craw, Shuster and Wakelin2019; Sanyal et al., Reference Sanyal, Shuster and Reith2019; Reference Sanyal, Reith and Shuster2020a). These studies have demonstrated that the occurrence of pure (>99%) secondary gold on particles can be attributed, in part, to microbial activity (Sanyal et al., Reference Sanyal, Brugger, Etschmann, Pederson, Delport, Dixon, Tearle, Ludington, Reith and Shuster2020b). These bacterial-gold interactions are thought to occur on the surface of particles within concavities containing clays, residual organics and secondary gold structures; collectively, these materials have been called ‘polymorphic layers’ (Shuster et al., Reference Shuster, Reith, Cornelis, Parsons, Parsons and Southam2017). Therefore, bacteria contribute to gold biogeochemical cycling—the sum of gold dissolution and reprecipitation processes—leading to gold particle transformation (Fairbrother et al. Reference Fairbrother, Etschmann, Brugger, Shapter, Southam and Reith2013; Reith et al., Reference Reith, Brugger, Zammit, Nies and Southam2013; Rea et al., Reference Rea, Zammit and Reith2016; Sanyal et al., Reference Sanyal, Shuster and Reith2018; Shuster et al., Reference Shuster, Reith, Cornelis, Parsons, Parsons and Southam2017; Shuster and Reith, Reference Shuster and Reith2018). The biogeochemical cycling of gold is controlled by a number of complex interactions between climate, landscape, and sediment chemistry (e.g. Melchiorre et al., Reference Melchiorre, Orwin, Reith, Rea, Yahn and Allison2018; Roy et al., Reference Roy, Upton and Craw2018; Craw Reference Craw2018; Rea et al., Reference Rea, Shuster, Hoffmann, Schade, Bissett and Reith2019a,Reference Rea, Wulser, Brugger, Etschmann, Bissett, Shuster and Reithb). In light of this, there is value in continuing gold geomicrobiology research by studying particles from environments with different climatic and ecological conditions, such as the Kamchatka Krai peninsula located on the far eastern side of Russia (Okrugin et al., Reference Okrugin, Andreeva, Etschmann, Pring, Li, Zhao, Griffiths, Lumpkin, Triani and Brugger2014; Zinkevich and Tsukanov, Reference Zinkevich and Tsukanov2010). This work builds upon the most recent gold geomicrobiology studies by focusing on the characterisation of gold particles obtained from Kamchatka (Fig. 1). In doing so, these particles are compared to previously characterised particles from similar latitudes, e.g. Germany, Switzerland and the United Kingdom. More importantly, this study further highlights the benefit of interdisciplinary research to gain a holistic perspective of biogeochemical processes that contribute to particle transformation within natural environments.
The sampling area is situated between the Sredinny and Vostocny Ranges where the Central Kamchatka Depression meets the Eastern Volcanic Plateau. Placer particle transformation involves both physical reshaping and gold biogeochemical cycling. The former is primarily attributed to sedimentation within hydrological regimes (i.e. rivers and streams). The latter is attributed, in part, to the presence of microbes occurring on the surface of particles (see Reith et al., Reference Reith, Fairbrother and Nolze2010).
Materials and methods
Thirty four gold particles were obtained by panning within a minor tributary (stream) that flows into the Bystraya River in the Yelizovsky District, Kamchatka, Russia (Fig. 1). The nearest city is Petropavlovsk-Kamchatskiy, located ~100 km south east of the sampling site. The gold particle sampling procedures are described in Reith et al. (Reference Reith, Fairbrother and Nolze2010). The sampling site (53.439667N, 157.613917E) is located within a stream that drains a shallow (~5 m depth) placer buried under glacial deposits. The region was worked industrially in the 1960–70s, producing ~543 kg of placer gold with an average grade of 0.745 g/m3. Gold occurred as small to medium particles (0.25–3.0 mm) with rare individual quartz-rich nuggets weighing up to 1252 g (393 g contained Au).
The climate at the site is continental (Dsd; altitude ~400 m) and polar/alpine (ET) at higher altitudes, based on the Köppen–Geiger classification (Kottek et al., Reference Kottek, Grieser, Beck, Rudolf and Rubel2006). The average monthly precipitation is 108 ± 25 mm; the months of October to December generally receiving more precipitation compared to other months. The average daily temperature in January (winter) and August (summer) is –10.8°C and 13.8°C, respectively (Petropavlovsk-Kamchatskiy, Climate Data, 2021). Permafrost does not exist and water in the stream is sourced from both surface runoff and meltwater from the range (Jones and Solomina, Reference Jones and Solomina2015). High humidity, relatively low temperatures, short growing seasons and extremely heavy snowfall during the winter season creates distinct environmental conditions that support Ermann's birch (Betula ermanii) forests, more commonly known as ‘snow forests’ (Krestov et al., Reference Krestov, Omelko and Nakamura2008). The particles were collected in September 2015 near the peak of summer biological activity.
Ten gold particles were prepared for structural/chemical characterisation following a method adapted from Shuster et al. (Reference Shuster, Southam, Reith, Kenney, Veeramani and Alessi2019). Briefly, gold particles were placed in 2% glutaraldehyde aqueous solution and incubated at 5°C for 48 hours to fix any bacterial cells. After incubation, the particles were sequentially transferred and incubated for 10 to 15 min in ethanol solution (70 and 90%, 3×100%). Reagent grade hexamethyldisilazane (HMDS) was diluted to 50% using pure 200 proof ethyl alcohol. The particles were transferred to this solution and incubated at 23°C for 30 mins; this step was repeated using 100% HMDS. After the final incubation, the particles were removed from the solution and airdried for 12 hours before being placed onto aluminium stubs with carbon adhesive tabs and coated with a 10 nm thick layer of carbon or iridium. All particles were imaged in secondary electron (SE) and back-scatter electron (BSE) modes using a FEI DualBeam Scanning Electron Microscope (SEM) or a JEOL JSM-7100 SEM operating at 2 or 20 kV. Both microscopes were equipped with an energy dispersive spectrometer (EDS) for semi-quantitative micro-chemical analyses; data were collected in spot mode. The long axis and perpendicular short axis of each particle were measured on the basis of SEM imaging, and the surface area of particles that occurred as clay-filled concavities (polymorphic layers) was estimated by analysing BSE micrographs with ImageJ software (NIMH, 2018).
Twenty-four gold particles were used to detect (remnant) bacterial DNA by polymerase chain reaction (PCR) amplification of the 16S rRNA genes combined with next generation sequencing using the Illumina MiSeq platform (Bissett et al., Reference Bissett, Fitzgerald, Meintjes, Mele, Reith, Dennis and Brugger2016). Two-step nested PCR was performed using the universal primers 27F and 1492R in the first round, then further amplification using 27F and 519R (Lane et al., Reference Lane, Pace, Olsen, Stahl, Sogin and Pace1985; Lane Reference Lane1991; and Osborn et al., Reference Osborn, Moore and Timmis2000). Samples were submitted for sequencing at the Australian Genome Research Facility (AGRF) in Melbourne, Australia. The 16S amplicons were sequenced using 300 bp paired end sequencing. Sequences were merged using FLASH (Magoc and Salzberg, Reference Magoč and Salzberg2011) and homopolymer runs of >8 bp or reads containing Ns were removed using MOTHUR v1.34.1 (Schloss et al., Reference Schloss, Westcott, Ryabin, Hall, Hartmann, Hollister, Lesniewski, Oakley, Parks, Robinson and Sahl2009). Abundance profiles were built by mapping data to identified amplicon sequence variants (ASV) (see Bissett et al. (Reference Bissett, Fitzgerald, Meintjes, Mele, Reith, Dennis and Brugger2016) for detailed method). ASVs that were observed fewer than two times were removed and ASVs with one single read were discarded. ASVs were analysed statistically to identify different bacterial phyla. A maximum-likelihood phylogenetic tree with 1000 bootstrap replicates was constructed using GENEIOUS 2021.0.3. Sequences were deposited in GenBank with accession numbers MW563744–MW563751.
The measurements of gold particles and the detected bacterial phyla were compared to particles from Germany (60; Rea et al. Reference Rea, Shuster, Hoffmann, Schade, Bissett and Reith2019a), Switzerland (46; Rea et al., Reference Rea, Wulser, Brugger, Etschmann, Bissett, Shuster and Reith2019b), and the United Kingdom (50, Rea et al., Reference Rea, Standish, Shuster, Bissett and Reith2018). These comparative samples were selected because they are located at broadly similar latitudes (46° to 56°30′) and both physical and biogeochemical factors are known to influence particle structure, chemistry, and microbial communities on gold particles.
Results
The gold particles from Kamchatka were 100s of micrometres in size and demonstrated a range of morphologies. Nine of the particles had rounded perimeters which gave them a ‘nugget-like’ appearance. One particle occurred as a semi-octahedral platelet with distinct edges. The particles also demonstrated a broad range of surface textures attributed to mechanical re-shaping (smooth surfaces) as well as biogeochemical processes (dissolution features and secondary gold nanoparticles). One nugget-like particle had a ‘cracked’ texture and also contained mercury, based on EDS analysis (Fig. 2). All particle surfaces had concavities of different sizes that contain varying amounts of clay and secondary gold nanoparticles. These concavities covered an average 73% of the surface area. When compared to other particles from similar latitudes, the particles from Russia contained 5 to 7 times more clay-filled concavities on the surface (Switzerland 15%, Germany 19%, and United Kingdom 22%). In terms of particle morphology, all particles from each location had an average long:short axis ratio of 1:0.7 (Fig. 3).
Back-scatter electron micrographs of gold particles and their respective surface textures. While the majority of particles appeared nugget-like and contained smooth and rounded surfaces attributed to mechanical reshaping (a), one particle appeared more euhedral in morphology but contained a weathered surface texture (b). One nugget-like particle contained a surface texture that appeared cracked (inset, arrow). Mercury was detected, based on EDS analysis (c).
A comparison of placer gold particles. Russian particles contained more clay-filled concavities relative to particles from Germany (Rea et al., Reference Rea, Shuster, Hoffmann, Schade, Bissett and Reith2019a), Switzerland (Rea et al., Reference Rea, Wulser, Brugger, Etschmann, Bissett, Shuster and Reith2019b) and the United Kingdom (Rea et al., Reference Rea, Standish, Shuster, Bissett and Reith2018). The average long:short axis ratio of all particles was 1:0.7. Secondary gold occurred as nanoparticles of varying shape and size. The amount of nanoparticles within clay-filled concavities (polymorphic layers) was also variable (insert).
A total of 902 ASV counts were detected from the Russian gold particles. The most abundant phyla included Proteobacteria (545 ASVs; 80.5% of total sequencing reads), Acidobacteria (81 ASVs; 4.2% of total sequencing reads), and Bacteroidetes (69 ASVs; 4.0% of total sequencing reads). Collectively, these phyla constituted more than 75% of all detected ASV counts (Table 1). Within Proteobacteria, Gammaproteobacteria was the dominant class that constituted more than half of the counted ASVs (290; 54.8% of total sequencing reads). From this class, ASVs representing Pseudomonas sp. and Rahnella sp. were consistently detected with greater frequency than other ASVs; additionally, Serratia proteamaculans was also detected (Fig. 4). When comparing bacterial compositions between sites, Proteobacteria from Russian particles were 13, 15 and 20% higher than what was detected on particles from Germany, Switzerland and the United Kingdom, respectively. Similarly, Acidobacteria and Bacteroidetes were the second and third most abundant phyla of the other locations when taken as an average (Fig. 5).
The number of Amplicon Sequence Variants (ASVs) detected from all gold particles.
A maximum likelihood phylogenetic tree of representative 16S rRNA sequence data of Gammaproteobacteria on at least 50% of sequenced Russian gold particles. Shown are percentages of 1000 bootstrap values and Methanobrevibacter smithii used as the outgroup. A strain of Serratia proteamaculans has been isolated from Australian gold particles (yellow circle) and is known to contain heavy-metal resistant genes and to withstand up to 50 μM Au (see Sanyal et al., Reference Sanyal, Reith and Shuster2020a).
A comparison of bacterial phyla detected on Russian particles with those that have been detected on particles from Germany, Switzerland and the United Kingdom. On Russian particles, Proteobacteria was the dominant bacterial phylum (60% of ASV counts); of this phylum, Gammaproteobacteria was the dominant class. Overall, bacteria detected on all particles were diverse.
Discussion
With active geology and a diverse geography, Kamchatka is a ‘natural laboratory’ to study the dynamics of gold biogeochemical cycling from particles. Based on field observations and the use of dynamically created maps (Google Maps, Google, 2021), the sampling site is located downstream from eight circular clearings 20–25 m in diameter within the stream. A similar group of circular clearings occurred ca. 600 m and 900 m upstream. These circular clearings were probably shallow pits attributed to placer mining activity during the 1960–70s, which were gradually reclaimed by vegetation over time. Despite this historically recent mining activity, gold was still recoverable from the stream (Figs 1, 6).
A schematic of the physical and biogeochemical processes contributing to the transformation of placer gold particles and the dispersion of gold within the stream at Kamchatka, Russia.
It is reasonable to suggest that the majority of gold particles, obtained from the sampling site, were probably derived from the same primary source as all particles had a nugget-like morphology. The particle that occurred as an octahedral platelet presumably came from a different source or represents gold that was unrecovered from the upstream mining pits. In general, smooth surface textures on particles are indicative of mechanical reshaping within sediment, which is consistent with early geological studies highlighting particle movement within a fluvial environment (e.g. Townley et al., Reference Townley, Hérail, Maksaev, Palacios, de Parseval, Sepulveda, Orellana, Rivas and Ulloa2003) (Figs 2a,b, 6). Some placer gold particles can contain trace levels of mercury, which is indicative of their primary origin (Reith et al., Reference Reith, Fairbrother and Nolze2010). Additionally, aurihydrargyrumite (Au6Hg5) has been detected on the surface of some placer gold particles (Nishio-Hamane et al., Reference Nishio-Hamane, Tanaka and Minakawa2018). It has been suggested that, in natural environments, the formation of this gold–mercury phase is analogous to how gold-enriched rims develop on particles. Briefly, gold and mercury are dissolved from the particle surface but are immediately re-precipitated back onto the surface via self-electrorefining and forming aurihydrargyrumite (Nishio-Hamane et al., Reference Nishio-Hamane, Tanaka and Minakawa2018). Therefore, it is possible that some mercury-bearing placer particles from Kamchatka could contain coatings of aurihydrargyrumite. In this study, however, the mercury-bearing particle exhibited a smooth but cracked surface texture (Fig. 2c). This texture has been described as being indicative of environmental contamination as mercury has an affinity for gold and can alter particle surface texture and chemistry. It has been suggested that extensive mercury contamination completely alters gold particle morphology forming near-perfect round spheres, an uncommon morphology for placer gold particles (Sanyal et al., Reference Sanyal, Brugger, Etschmann, Pederson, Delport, Dixon, Tearle, Ludington, Reith and Shuster2020b). Anthropogenic activity such as small-scale artisanal mining has contributed to mercury contamination in a number of natural environments around the globe with devastating impacts (Veiga and Hinton, Reference Veiga and Hinton2002; Telmer and Veiga, Reference Telmer and Veiga2009). Mining activity occurred upstream of the sampling site during the 1960–70s; however, there are no known records indicating that mercury was used for amalgamation to concentrate the recovered gold. It has been suggested that the accumulation of mercury in lacustrine sediments from remote regions of Kamchatka is attributed mostly to global fossil fuel combustion (Jones et al., Reference Jones, Rose, Self, Solovieva and Yang2015). Therefore, it is reasonable to suggest that fossil fuel combustion used to operate machinery, upstream of the sampling site, probably contaminated the stream with mercury, thereby being dispersed downstream and interacting with gold to form an amalgam on the surface of some particles (Figs 6).
In terms of morphology, placer particles with near-equant axes tend to appear more nugget-like, which correlates with their deposition from the primary source (Townley et al., Reference Townley, Hérail, Maksaev, Palacios, de Parseval, Sepulveda, Orellana, Rivas and Ulloa2003; Shuster et al., Reference Shuster, Reith, Cornelis, Parsons, Parsons and Southam2017). The average axis ratio of Russian particles was consistent with the particles from Germany, Switzerland and the United Kingdom, suggesting that transport from their respective primary sources were potentially comparable. The most striking difference between the Russian gold particles was that they contained more clay-filled concavities (Figs 3, 6). Recent studies have shown that changes in climate, e.g. wetter winters and increased air temperatures, have contributed to increased weathering and thus more sediment movement and accumulation in rivers and lakes throughout Kamchatka (Jones et al., Reference Jones, Rose, Self, Solovieva and Yang2015; Jones and Solomina, Reference Jones and Solomina2015; Kuksina, Reference Kuksina2019). Therefore, greater sedimentation could explain the ‘dirtier’ appearance of these particles.
Overall, Proteobacteria and Gammaproteobacteria were the most dominant phylum and class, respectively, occurring on the surface of Russian particles (Table 1, Figs 4), which is consistent with previous studies (Reith et al., Reference Reith, Falconer, Van Nostrand, Craw, Shuster and Wakelin2019 and references therein related to bacteria associated with particles). Interestingly, all particles, regardless of locality, contained a diverse range of bacteria and the relative phyla proportions were similar (Fig. 5). Rea et al. (Reference Rea, Zammit and Reith2016) proposed that soluble gold, from particles experiencing dissolution, can act as a selective pressure for metal-resistant bacteria occurring on the surface of particles (Fig. 6). This was demonstrated by microcosm experiments involving bacteria grown from particles (Sanyal et al., Reference Sanyal, Shuster and Reith2019, Reference Sanyal, Reith and Shuster2020a). Additionally, Proteobacteria sp. has been detected and cultured from gold particles obtained from an environment impacted by anthropogenic heavy metals in South Africa (Sanyal et al., Reference Sanyal, Brugger, Etschmann, Pederson, Delport, Dixon, Tearle, Ludington, Reith and Shuster2020b). Furthermore, Serratia proteamaculans has been cultured from particles obtained from a historic and abandoned gold mine in Australia (Sanyal et al., Reference Sanyal, Reith and Shuster2020a). Comparative genome analysis of the Serratia sp. isolate contained a number of heavy-metal resistance and stress-response genes that were identical to those detected in Cupriavidus metallidurans CH34 – a Au-tolerant microbe (Reith et al. Reference Reith, Etschmann, Grosse, Moors, Benotmane, Monsieurs, Grass, Doonan, Vogt, Lai, Martinez-Criado, George, Nies, Mergeay, Pring, Southam and Brugger2009; Wiesemann et al. Reference Wiesemann, Bütof, Herzberg, Hause, Berthold, Etschmann, Brugger, Martinez-Criado, Dobritzsch, Baginsky, Reith and Nies2017). The presence of these genes highlights this microbe's genomic capacity to withstand high concentrations of heavy metals including gold (Sanyal et al., Reference Sanyal, Reith and Shuster2020a). Therefore, it is possible that the microbes detected on the surface of Russian particles reflect changes in environmental geochemistry including the impact of past anthropogenic activity within and along the stream. Indeed, other factors such as bioavailability of carbon and nitrogen, attributed to weathering rates, can also result in a shift in bacterial diversity (Lin et al., Reference Lin, Jia, Wang and Chiu2017; Reith et al., Reference Reith, Brugger, Zammit, Gregg, Goldfarb, Andersen, DeSantis, Piceno, Brodie, Lu, He, Zhou and Wakelin2012). In terms of particle transformation, the detection of ASVs directly from Russian particles highlights the types of bacteria that could contribute to gold biogeochemical cycling in the environment.
Conclusion
This investigation compared gold particles obtained from a distinct ecological environment (Snow Forest, Kamchatka, Russia) to particles obtained from regions at similar latitudes (Germany, Switzerland and the United Kingdom). While Russian particles had more variability in gold surface textures, overall morphologies of all particles were consistent, suggesting that transport was probably comparable. The accumulation of clays within concavities and detection of ASVs representing Gammaproteobacteria on particles could reflect localised changes in the environment attributed to anthropogenic activity. Further studies are needed to better understand how trace-element composition of particles could relate to changing bacterial compositions over time.
Acknowledgements
This publication is dedicated to Associate Professor Frank Reith. This research was supported by the Australian Research Council (ARC) Future Fellowship (FT100150200) awarded to F. Reith (c/o J. Shuster). Scanning electron microscopy analysis was performed at Adelaide Microscopy, The University of Adelaide (Adelaide, SA 5005, Australia) and the Centre for Microscopy and Microanalysis, The University of Queensland (St. Lucia, QLD 4072, Australia). Molecular analysis was performed at CSIRO Land and Water, Environmental Toxicology and Chemistry in association with the Australian Genome Research Facility (Urrbrae, SA 5064, Australia). We thank A. Bissett and C. Zammit for bioinformatics analysis and Stefan Ansermet (Lausanne) for assistance with the sampling in Kamchatka.
