Pioneer species of Cyanobacteria in hot springs and their role to travertine formation: The case of Aedipsos hot springs, Euboea (Evia), Greece

Abstract

Cyanobacteria are considered to be among the first microorganisms to settle in hot springs where they form a favourable environment for further biological establishment. Nevertheless, the exact pioneer species and how early they start participating in the biomineralisation processes remain unknown. The aim of the present study was twofold, that is, to identify the pioneer Cyanobacteria in hot springs (i.e. Aedipsos area, Greece) and to record their early biomineralisation processes. The in situ experimental approach included the setup of sterile glass and/or plexiglass slides in several locations to facilitate colonisation by Cyanobacteria, and removal of slides for study after 48–202 h. Synechococcales (37%) and Oscillatoriales (33%) were the dominant orders, followed by Chroococcales (15%) and Spirulinales (11%); whereas Chroococcidiopsidales (4%) was found only in a few sites. The order Nostocales was not observed at the early stages of colonisation although it was present in mature stages. Forty‐three species of Cyanobacteria were identified as pioneer microorganisms, with Spirulina subtilissima being the most frequently found. The most common pioneers were multicellular filamentous Cyanobacteria, that is, organisms with a large surface area able to form significant amounts of extracellular polymeric substances. Among the pioneers, thermophilic species of Cyanobacteria were typical such as Chroococcidiopsis thermalis, Chroococcus thermalis, Leptolyngbya thermalis, S. subtilissima and Symploca thermalis, as well as typical limestone substrate species such as Chroococcus lithophilus and Leptolyngbya laminosa. Temperature seems to affect biodiversity. Also, pioneers were found to contribute to the biomineralisation processes from their first appearance. In the studied samples, three biomineralisation processes were identified, that is, (i) calcification of cyanobacterial sheaths, (ii) trapping of carbonate crystals on a crystal retention lattice formed by extracellular polymeric substances and filaments and (iii) trapping and confinement of carbonate crystals around filamentous Cyanobacteria.

INTRODUCTION

Extreme environments have attracted the interest of both geologists and biologists due to the geological processes taking place there and because of the living microorganisms that manage to deal with these harsh conditions. Study of extreme environments requires an interdisciplinary research approach, that is, geomicrobiological, which can address fundamental questions such as the origin of life or evolution and which might result in practical applications (Lewin et al., 2013; Schlesinger, 1990; Shu & Huang, 2021). In geomicrobiological studies, it is possible to assess the unknown biodiversity, the related geo-bio-processes and the prevailing physicochemical conditions (Fouke, 2011; Fouke et al., 2003; Kuang et al., 2021; Merino et al., 2019; Meyer-Dombard et al., 2005; Pandey & Sharma, 2021; Power et al., 2018; Sharp et al., 2014; Shu & Huang, 2021; Thomas, 2015; Thomas & Ariztegui, 2019).

In harsh environments, such as hot springs, microorganisms from all three domains of life have found ways to survive. Cyanobacteria (also known as cyanophytes or blue-green algae) are among the microorganisms thriving there; they represent a group of early and worldwide-distributed photosynthetic prokaryotes (Whitton et al., 2012), surviving even in limited light conditions (Hubalek et al., 2016; Puente-Sanchez et al., 2018). Additionally, Cyanobacteria can actively or passively participate in the formation of minerals, such as carbonate minerals (Altermann et al., 2006; Arp et al., 1999; Bundeleva et al., 2014; Freytet & Verrecchia, 1998; Jansson & Northen, 2010; Kamennaya et al., 2012; Kupriyanova & Pronina, 2011; Merz, 1992). However, the exact mechanism through which Cyanobacteria can participate in the formation of carbonate minerals has not been fully elucidated and, until recently, it was believed that biomineralisation of carbonate minerals takes place only extracellularly. Couradeau et al. (2011) and Benzerara et al. (2014) showed that intracellular biomineralisation could also take place in several species, i.e. Candidatus Gloeomargarita lithophora (Moreira et al., 2017), Candidatus Synechococcus calcipolaris (Ragon et al., 2014) and several other members of the same phylogenetic clade, including Thermosynechococcus elongatus. Li et al. (2013 and references therein) and Görgen et al. (2021 and references therein) summarised all of the known biomineralisation processes of Cyanobacteria, such as extracellular polymeric substances (EPS) and the influence of bacterial cell walls, alkaline engine, etc. With carbonate rocks, combinations of biocontrolled, bioinduced, as well as abiotic mineralisation are possible in these harsh environments, resulting in the formation of hybrid carbonates (Riding & Virgone, 2020).

Cyanobacteria have contributed significantly to the foundation of ecosystems before the great oxidation event (Blank & Sánchez-Baracaldo, 2010; Golubic & Seong-Joo, 1999; Gumsley et al., 2017). Therefore, they have participated in both old and present C and Ca geochemical cycles (Riding, 2006; Zavarzin, 2002) while assisting fossilisation processes and the preservation of traces of ancient life (Li et al., 2013). Additionally, there are numerous possible practical applications based on carbonate biomineralisation, such as soil improvement (Whiffin et al., 2007), removal of heavy metals (Warren et al., 2001) and CO₂ capture and sequestration (Kamennaya et al., 2012; McCutcheon et al., 2014; Yasumoto et al., 2014).

Furthermore, Cyanobacteria are considered to be among the first settlers (pioneers), that is, organisms that settle on a surface and create suitable conditions leading to the development of other more complex bacterial consortia formed by heterotrophic organisms. The main source of energy for these biofilms is photosynthesis. Cyanobacteria use light as an energy source and reduce CO₂ by providing organic substrates and oxygen. The resulting complex biocommunity consists of microorganisms immersed in a hydrated microenvironment made of water (70%–95%) and EPS (Cuzman, 2009; Flemming, 1993; Wahl, 1989). Ιt is crucial to identify the microorganisms that act as pioneers since they contribute to further development of biocommunities in these areas and possibly to the formation of the first layer of thermogenic travertine (Jones et al., 1998; Guidry & Handley et al., 2005). Systematic identification of the pioneer microorganisms is often missing preventing a clear understanding of their role in mineral precipitation.

Greece presents an excellent region for this kind of study due to the geological setting, the active tectonics and the volcanoes, resulting in hundreds of emerging hot springs (Gkioni-Stavropoulou, 1983; Orfanos, 1985; Orfanos & Sfetsos, 1975; Sfetsos, 1988). The existing knowledge of Cyanobacteria from the wealth of Greek hot springs is rather limited (Anagnostidis, 1959, 1961, 1964, 1967a, 1967b, 1968, 1977; Anagnostidis & Economou-Amilli, 1978; Anagnostidis & Pantazidou, 1988; Anagnostidis & Roussomoustakaki, 1991; Bravakos et al., 2016; Economou-Amilli, 1976; Kanellopoulos et al., 2015; Panou & Gkelis, 2022; Radea et al., 2010; Roussomoustakaki & Anagnostidis, 1991). Additionally, several Greek hot springs currently deposit thermogenic travertine, and previous studies have shown the participation of Cyanobacteria in such biomineralisation processes (Kanellopoulos et al., 2015, 2022). However, further information is needed especially for the pioneer species. Determining how their biomineralisation processes contribute to the first stage of thermogenic travertine formation in hot springs could be an example of successful implementation of this.

Τhis study aims to fill this gap by identifying the pioneer Cyanobacteria in a group of hot springs (i.e. in Aedipsos area, North-West Euboea Island, Greece) and investigating how their possible biomineralisation processes contribute to the first stage of thermogenic travertine formation.

MATERIALS AND METHODS Sampling and analysis

Sampling of fresh material was made, and the experimental setups of glass or plexiglass slides were installed in several locations of Aedipsos hot springs, Euboea (Evia) Island, Greece (Figure 1). All sampling processes were conducted with sterile metal chisels and tweezers. All artificial sub-surfaces, that is, glass or plexiglass slides, were sterile before they were installed at the study sites.

FIGURE 1. Geological map of the Aedipsos area (AF = Aedipsos Fault) (after Kanellopoulos et al., 2020). The geographical coordinates are in EGSA 87.

From several sampling sites, two sub-samples were collected, but Cyanobacteria were identified only at 11 sites. The first sub-sample was incubated into sterile transparent vials in the field. The second sub-sample was stored in formaldehyde solution (2.5%). Enriched cultures were obtained in flasks and Petri dishes with BG11 and BG 11₀ culture media (Stanier et al., 1971).

All samples were examined under an optical microscope (Zeiss Photomicroscope III; Zeiss) and a stereo-microscope (Zeiss Stemi 2000C; Zeiss). Classical and recent literature was used for species identification (Komárek & Anagnostidis, 1989, 1999, 2005; Komárek et al., 2014 and references therein).

Selected dehydrated samples in an alcohol series (30%–100%) were critical point dried, gold-coated and observed under a scanning electron microscope (Jeol JSM 5600; Jeol USA, Inc.) equipped with an Oxford ISIS 300 (Oxford Instruments) microanalytical device (operating conditions were: accelerating voltage 20 kV, beam current 0.5 nA, time of measurement 50 s and beam diameter 1–2 mm).

Physicochemical parameters of the water (i.e. temperature, pH, salinity) were measured for each sampling site in situ using a portable device. Temperature was measured with a probe connected to the pH meter, with the error estimated to be less than ±0.3°C. Geographical coordinates for each sampling site were recorded using a handheld portable GPS.

Statistical analysis and use of GIS

Multivariate statistical analysis has been a valuable tool for environmental studies since it can simplify a dataset revealing the similarity between samples. Cluster analysis and principal component factor analysis (PCA) were applied using Primer 6/Permanova statistical software (Primer-e). Stacked bar diagrams for the percentage of the Cyanobacteria orders were created using Microsoft Excel 365 (Microsoft).

ArcGIS (ESRI, 2011; ESRI) was used to modify the geological map presented by Kanellopoulos et al. (2020).

GEOLOGICAL SETTING

In the Loutra area of Aedipsos (North-West Euboea Island, Greece) several hot springs occur (Figure 1) depositing thermogenic travertine. These hot springs, known for their healing properties since ancient times, are related to the recent local volcanism (Lichades volcanic centre, 0.5 Ma age, dated by K–Ar method - Fytikas et al., 1976) and active tectonics (Kanellopoulos et al., 2016, 2017, 2020). They are classified as sodium-chloride springs with almost neutral pH and temperatures from 37.2 to 82.2°C (Kanellopoulos et al., 2016, 2017, 2020).

Geologically, the area belongs to the Pelagonian geotectonic zone of the Hellenides (Aubouin, 1959; Jolivet et al., 2013; Mountrakis, 1986). The main geological formations in the area are: a metamorphic crystalline basement (pre-middle to middle Carboniferous age), a basic volcanoclastic complex series (Permian–Triassic age), shallow marine carbonate and clastic rocks (middle Triassic age) with volcanic rock intercalations (Katsikatsos et al., 1982; Scherreiks, 2000; Figure 1) overlain by alluvial and thermogenic travertine deposits.

The thermogenic travertine deposits are vast since, in many cases, depositional rates have been high, and the hot springs have been active since ancient times. The travertines present a great variety of facies (Kanellopoulos, 2012, 2013) and were used locally as a construction material. Ιn previous studies, the possible occurrence of biomineralisation processes was discussed (Kanellopoulos, 2011, 2012, 2014), and the first geomicrobiological study in the area presenting the Cyanobacteria biodiversity and relevant biomineralisation processes was made (Kanellopoulos et al., 2022).

RESULTS AND DISCUSSION Description of sampling sites and in-situ measurements

Successful sterile glass or plexiglass slides were installed in four main sites of Aedipsos. In the setup sites, water physicochemical measurements and field observations were taken in situ (Table 1). Based on Castany's (1963) classification, the studied sites are characterised as meso-thermal (T: 35°C up to 50°C) and hyper-thermal (T: over 50°C). Slides were installed in sites with temperatures over 70°C, but no Cyanobacteria were identified on them.

TABLE 1 Sampling locations, physicochemical parameters and classification of the studied hot waters

Code	Location	Time (h)	Classification (based on Castany, 1963)	T (°C)	Sal (‰)	pH
P1	EOT	105	Hyper-thermal	54.2	26	6.6
P2	EOT	175	Hyper-thermal	69	nm	6
P3	Platania	191	Meso-thermal	49.2	20	6
P4	Platania	191	Meso-thermal	49.2	20	6
P5	Platania	191	Meso-thermal	37.2	27	6.3
P6	Ilios	119	Meso-thermal	48.9	24	7.4
P9	Casino	48	Hyper-thermal	52.4	26	7.2
P10	Casino	48	Hyper-thermal	52.4	26	7.2
P11	Casino	48	Meso-thermal	40.1	25	7.3
P12	Casino	48	Meso-thermal	40.1	25	7.3
P14	EOT	202	Hyper-thermal	54.2	26	6.6

Abbreviation: nm = not measured.

At Platania, the main hot spring (T = 49.2°C, pH = 6.05, Sal = 20‰), with low water discharge, occurs in a cave. In that area, limited or no travertine is deposited. The first slide was installed exactly at the hot water outlet (area with limited sunlight; P4; Figure 2A,B) for 191 h. The second slide (P3; Figure 2A) was installed at the cave entrance and the beginning of the drainage channel (less than 0.5 m from the previous one, with slow water flow but no sunlight restriction) and also left for 191 h. The third slide (P5; Figure 2C,D) was installed at the end of the drainage channel, where a small pool occurs (T = 37.2°C, pH = 6.27, Sal = 27‰; no water flow, no sunlight restriction), and also collected after 191 h.

FIGURE 2. Field photographs of the sampling sites. (A) View of the cave where the main hot spring occurs in the Platania area. The site P4 is marked. No travertine deposition was identified at the bottom of the cave. (B) The slide P4, after 191 h. It was installed at the hot spring of Platania. (C) Small pool at the end of the drainage channel of Platania area. The P5 setup was installed at the bottom of the pool. No travertine deposition was identified. (D) The P5 setup, after 191 h. (E) The P6 setup at the drainage channel of Ilios hot spring. The rock bottom of the channel consists of laminated travertine. (F) The Casino drainage channel (open part). (G) The Ntamaria hot spring, at EOT location; the P2 installation site is marked. (Η, Ι) The EOT drainage channel; the P1 and P14 setup can be seen, at (H) just after the installation and at (I) after 202 h.

At Ilios, a hot spring occurs inside a cave. In that area, orange-laminated travertine is deposited. The slide (P6; Figure 2E) was set up at the drainage channel, a few metres from the cave entrance (T = 48.9°C, pH = 7.39, Sal = 24‰; slow water flow and no sunlight restriction), and was collected after 119 h.

At Casino, one main drainage channel (Figure 2F) exists where water from several hot springs is collected and discharged at sea. At the channel, orange-laminated travertine is deposited. Part of the drainage channel is open, and part of it is boxed under cement plates. The first two samples were collected from the open part (T = 52.4°C, pH = 7.16, Sal = 26‰; fast water flow, no sunlight restriction). A slide was installed vertically to the water flow. Due to the fast water flow and differences between the two sides, the front part of the slide (P9, opposing the water flow) was distinguished as a separate sample from the back part of the slide (P10). From the boxed part of the drainage channel (T = 40.1°C, pH = 7.26, Sal = 25‰; fast water flow, limited sunlight access), again, a slide was installed vertically to the water flow. The front part of the slide (P11, opposing the water flow) was distinguished as a separate sample from the back part of the slide (P12). All slides from that area were collected after 48 hours.

In the area of EOT, several hot springs occur in close proximity, with temperatures up to 80°C, and their water discharge is collected in drainage channels. In most hot springs of this area, travertine is deposited. Several slides were installed there, but Cyanobacteria were identified only in places with temperatures below ca 70°C.

At the Ntamaria hot spring (EOT area), a slide was installed only a few tens of centimetres from the hot spring (P2; T = 69 °C, pH = 6; Figure 2G) for 175 h. On that side, there was fast water flow, and two colour variations of dark red-white specular travertine occur. Kanellopoulos (2012) suggests that the reddish coloured area is iron-rich and is formed under the surface of the hot water, while the white coloured area is formed above the surface of hot water and is Ca-rich.

At one of the main drainage channels of the EOT area, a set of slides vertical to the water flow was installed (T = 54.2°C, pH = 6.6, Sal = 26‰; Figure 2Η–Ι). The first slide (P1) was collected after 105 h, while the second one (P14) was collected after 202 h. The water flow is fast, and in that area, orange-laminated travertine is deposited.

Diversity of Cyanobacteria and statistical analysis

Microscopic analysis of fresh and cultured material revealed a total number of 43 species of Cyanobacteria (Table 2). In sites where the temperature was over 70°C, no Cyanobacteria species were identified. In some sampling sites, diatoms have also been recorded, but their detailed identification remains a subject of future research. Among the identified species, typical thermophilic species have also been observed, such as Chroococcidiopsis thermalis (Figure 3A), Chroococcus thermalis, Leptolyngbya thermalis (Figure 3B), Spirulina subtilissima (Figure 3C) and Symploca thermalis (Figure 3D). Moreover, typical limestone substrate Cyanobacteria species were found, such as Chroococcus lithophilus (Figure 3E) and Leptolyngbya laminosa (Figure 3F). The most common pioneer species in the Aedipsos hot springs are: (i) S. subtilissima (Figure 3C) found in seven sites, (ii) Synechococcus elongatus (Figure 3G) found in six sites and (iii) Oxynema acuminatum (Figure 3H) found in five sites. Thus, the most common pioneer organisms are multicellular filamentous Cyanobacteria, that is, organisms with a larger outer surface which is generally related to their ability to adhere more easily to surfaces. These organisms also produce larger amounts of EPSs due to their surface area, resulting in greater calcium carbonate binding compared to unicellular Cyanobacteria (Lau et al., 2008). The only exception is the presence of Synechococcus elongatus, a unicellular species with a slightly varying cell form according to the environmental temperature. In a geomicrobiological study of the North-West Euboea hot springs, Kanellopoulos et al. (2022) also identified the species mentioned above and concluded that the most common was S. subtilissima.

TABLE 2 Identified Cyanobacteria species per sampling site

Species	Site
Aphanocapsa fuscolutea Hansgirg 1892	P1
Aphanothece cf. microscopica Nägeli 1849	P1
Chroococcidiopsis thermalis Geitler 1933	P10, P11, P12
Chroococcus lithophilus Ercegovic 1925	P3
Chroococcus minutus (Kützing) Nägeli 1849	P1, P5
Chroococcus subsphaericus Gardner 1927	P11, P12
Chroococcus thermalis (Meneghini) Nägeli 1849	P5
Chroococcus turgidus (Kützing) Nägeli 1849	P5, P14
Cyanobacterium minervae (J.J.Copeland) Komárek, Kopeck‡ & Cepák 1999	P14
Cyanosarcina cf. thalassia Anagnostidis & Pantazidou 1991	P12
Cyanosarcina cf. thermalis (Hindák) Kovácik 1988	P3, P9
Anagnostidinema amphibium (Agardh ex Gomont) Strunecký, Bohunická, Johansen et Komárek 2017	P10
Kamptonema formosum (Bory ex Gomont) Strunecký, Komárek & J.Smarda 2014	P3
Kamptonema laetevirens (H.M.Crouan & P.L.Crouan ex Gomont) Strunecký, Komárek & J.Smarda 2014	P3
Kamptonema sp.A	P3
Komvophoron halobium Anagnostidis 2001	P5
Leptolyngbya cf. ferruginea (G.S.West) Anagnostidis & Komárek 1988	P14
Leptolyngbya cf. thermobia Anagnostidis 2001	P9
Leptolyngbya laminosa (Gomont ex Gomont) Anagnostidis & Komárek 1988	P4
Drouetiella lurida (Gomont) Mai, J.R.Johansen & Pietrasiak 2018	P3, P14
Leptolyngbya orientalis (G.S.West) Anagnostidis & Komárek 1988	P1
Leptolyngbya sp.A	P12
Leptolyngbya tenuis (Gomont) Anagnostidis & Komárek 1988	P3
Leptolyngbya thermalis Anagnostidis 1988	P10, P11, P12
Oscillatoria cf. ornata Kützing ex Gomont 1892	P5
Oscillatoria nigroviridis Thwaites ex Gomont 1892	P5
Oxynema acuminatum (Gomont) Chatchawan, Komárek, Strunecky, Smarda & Peerapornpisal 2012	P1, P3, P4, P9, P14
Phormidium acidophilum J.J.Copeland 1936	P11, P14
Phormidium cf. chlorinum (Kützing ex Gomont) Umezaki & Watanabe 1994	P9
Phormidium cf. cortianum (Meneghini ex Gomont) Anagnostidis & Komárek 1988	P4
Phormidium corium Gomont ex Gomont 1892	P14
Phormidium sp.A	P3
Phormidium tergestinum (Rabenhorst ex Gomont) Anagnostidis & Komárek 1988	P3
Pseudanabaena minima (G.S.An) Anagnostidis 2001	P1
Schizothrix cf. lardacea Gomont 1892	P2, P14
Spirulina subsalsa Oersted ex Gomont 1892	P12
Spirulina subtilissima Kützing ex Gomont 1892	P1, P4, P5, P9, P10, P12, P14
Spirulina tenerrima Kützing ex Gomont 1892	P14
Symploca thermalis Gomont 1892	P4, P6, P12
Synechococcus bigranulatus Skuja 1933	P9
Synechococcus elongatus f. thermalis Geitler 1935	P3, P9, P10, P11, P12, P14
Synechocystis salina Wislouch 1924	P1, P3
Trichocoleus delicatulus (West & G.S.West) Anagnostidis 2001	P10
Diatoms	P3, P5

FIGURE 3. Identified species of Cyanobacteria. (A) Chroococcidiopsis thermalis, (B) Leptolyngbya thermalis, (C) Spirulina subtilissima, (D) Symploca thermalis, (E) Chroococcus lithophilus, (F) Leptolyngbya laminosa, (G) Synechococcus elongatus, (H) Oxynema acuminatum, (I) Leptolyngbya orientalis.

In order to evaluate the distribution of the Cyanobacteria microflora, stacked bar diagrams were created (Figure 4), presenting the orders of Cyanobacteria based on the number of species per sampling site. The identification was made according to the latest classification system (Hauer & Komárek, 2021). The Cyanobacteria identified were classified under the following orders, i.e. Chroococcales, Chroococcidiopsidales, Oscillatoriales, Spirulinales and Synechococcales. It can be deduced (Figure 4A) that Synechococcales (37%) and Oscillatoriales (33%) are the dominant orders, followed by Chroococcales (15%) and Spirulinales (11%), while Chroococcidiopsidales (4%) are found only in a few samples and always with the minimum percentage.

FIGURE 4. Stacked bar diagrams presenting the percentage of each Cyanobacteria order; in (A) A summary from all sampling sites, (B) P1 sample, (C) P2 sample, (D) P3 sample, (E) P4 sample, (F) P5 sample, (G) P6 sample, (H) P9 sample, (I) P10 sample, (J) P11 sample, (K) P12 sample and (L) P14 sample.

In the drainage channel of EOT, two samples were collected at different residence times (P1: 105 h vs P14: 202 h). In the first sample (P1; Figure 4B), the dominant order is Synechococcales (50%), followed by Chroococcales (25%), Spirulinales (13%) and Oscillatoriales (12%). After an additional 97 h (P14; Figure 4L), although less abundant (37%), Synechococcales remains the predominant order, followed by Oscillatoriales with the percentage contribution doubled to 27%. The percentage contribution of Chroococcales and Spirulinales remains unchanged at 18%.

At Platania, in the hot spring sample (P4; Figure 4E) with limited sunlight access, the dominant order was found to be Oscillatoriales (60%) followed by Synechococcales (20%) and Spirulinales (20%); whereas at the beginning of the drainage channel (P3; Figure 4D) with full sunlight access, Chroococcales are replacing Spirulinales, that is, Oscillatoriales (50%), Synechococcales (33%) and Chroococcales (17%).

In the drainage channel at Casino, the dominant order in all samples was Synechococcales followed by Oscillatoriales, Chroococcales and Spirulinales. A similar microalgal composition was found in areas either against the water flow or on the opposite side, whereas Chroococcidiopsidales were present only in samples with limited sunlight, that is, P9-P12 (Figure 4H–K).

In the drainage channel at Ilios (P6; Figure 4G), all identified Cyanobacteria belong to Oscillatoriales (100%). Whereas at the hot spring of Ntamaria (P2; EOT area, Figure 4C), all identified Cyanobacteria belong to the order Synechococcales (100%).

Cluster analysis and PCA revealed that only three samples present 40% similarity, that is, P10, P11 and P12. The sites with the most distinct biodiversity are P2 and P6 since they are restricted to species limited to one order (P2: 100% Synechococcales and P6: 100% Oscillatoriales; Figure 5A). While reducing similarity to 20%, two more groups were revealed: (i) P1 and P4 and (ii) P3, P9 and P14, which are all samples from drainage channels and with temperatures from 49.2 to 54.2°C (Figure 5B).

FIGURE 5. Statistical diagrams. (A) Cluster dendrogram based on the identified Cyanobacteria species per site. (B) PCA diagram based on the identified Cyanobacteria species per site.

Based on the data mentioned above, the biodiversity of pioneer Cyanobacteria seems to be regulated by the temperature, that is, no pioneer Cyanobacteria were identified in sites with temperatures over 70°C. Kanellopoulos et al. (2022), after a geomicrobiological assessment of the hot springs of North-West Euboea Island (including Aedipsos area), defined temperature and salinity as the two major controlling factors of biodiversity. Based on their results, only a few Cyanobacteria orders were observed in hot springs with high temperatures (over 62°C), that is, Synechococcales was dominant, followed by Spirulinales. In hot springs with high salinity, such as 35‰, that is, similar to the average salinity of the oceans, again only a few Cyanobacteria orders were observed, that is, Oscillatoriales were dominant, followed by Spirulinales and Chroococcales. During the present study, samples were collected from sites with a limited range of salinity, that is, 20–27‰. In future research, it would be interesting to study pioneer Cyanobacteria species in additional Greek hot springs with high salinity. Other geomicrobiological studies in hot springs worldwide have also suggested temperature as one of the major factors controlling biodiversity (Della Porta et al., 2021; Miller et al., 2009; Sharp et al., 2014; Ward et al., 2017). Although that is not a general rule, there are cases, such as in the Taupo Volcanic Zone hot springs (New Zealand), where temperature plays an important role, but in lower temperatures pH is the major factor controlling biodiversity (Power et al., 2018).

In the hot springs of North-West Euboea Island, Kanellopoulos et al. (2022) defined Oscillatoriales as the most abundant order, followed by Synechococcales. Both orders dominate among the pioneer Cyanobacteria, and in two samples they are present exclusively (P2: 100% Synechococcales and P6: 100% Oscillatoriales). It is worth noting that Kanellopoulos et al. (2022), studying the Aedipos hot springs, mention finding the same orders listed above, plus Nostocales. The order Nostocales was identified in sampling sites with specific salinity (25–27‰) suggesting that salinity controls its presence. Also, Chroococcidiopsidales are present at a higher percentage within the pioneer species when compared to the typical biodiversity of the local hot springs, that is, pioneers: 4% versus typical: 0.3% (Kanellopoulos et al., 2022).

Thus, it can be concluded that almost the same Cyanobacteria orders are present at the early settlement stages (i.e. in the first few hours) and at the mature stages. The orders Oscillatoriales and Synechococcales dominate at both early and mature stages.

Biomineralisation

The contribution of Cyanobacteria to biomineralisation processes in hot spring environments is known (Fouke, 2011; Fouke et al., 2003; Kanellopoulos et al., 2015, 2022; Pentecost, 2005). However, it is still unclear whether the biomineralisation processes start from the early stages (i.e. in few hours) or in mature stages. Examination of samples which were left for only a few hours underwater (Figure 6) revealed that the pioneer Cyanobacteria are participating in the calcium carbonate biomineralisation processes from the beginning of their colonisation and in several ways (Figure 6). Thus, along with the abiotic deposition processes, they result in the formation of hybrid travertines (Riding & Virgone, 2020).

FIGURE 6. Biomineralisation processes by Cyanobacteria under a stereoscope (A, B) and under SEM (C–I). (A, Β) Slide from the EOT drainage channel, after 30 days, where travertine shrubs can be seen; 7 mm thick. The pioneer Cyanobacteria can also be seen. The dominant species are Cyanobacterium minervae, Schizothrix lardacea and Oxynema acuminatum. (C) Filament of the genus Spirulina (red arrow) holding calcium carbonate crystals and a calcified sheath of filamentous Cyanobacteria (blue arrow). (D) EPSs coating calcium carbonate crystals complex. (E–H) EPSs, along with filamentous Cyanobacteria forming retention lattice of calcium carbonate crystals. (I) Filamentous Cyanobacteria retain crystals without the presence of EPSs.

Calcification of pioneer filamentous Cyanobacteria sheaths was observed in the studied samples (Figure 6C). It can be seen as a micritic encrustation leading to tube-forms of calcium carbonate minerals. These structures are not uncommon on the travertines (Freytet & Verrecchia, 1998; Pentecost, 1978, 2005; Schneider & Le Campion-Alsumard, 1999; Tavera & Komárek, 1996), but it is verified here that they start to appear only a few hours after colonisation. Kanellopoulos et al. (2022) found similar structures also in the Aedipsos hot springs and suggested that they could be related to some shrub facies of the Aedipsos travertines. Several suggestions exist about the processes responsible for the sheath encrustation (Couradeau et al., 2013; Dupraz et al., 2009), such as nucleating molecules like those composing cyanobacterial sheaths (Merz-Preiss & Riding, 1999) or S layer proteins (Thompson et al., 1997). Although the most commonly accepted process is associated with the oxygenic photosynthesis, which locally increases the pH in the cell vicinity leading to carbonate oversaturation and precipitation (Jansson & Northen, 2010; Riding, 2006).

Additionally, the presence of EPSs was usually observed in the studied samples, mainly consisting of polysaccharides, polymers and some nucleic acids, lipids and proteins (cf. Flemming & Wingender, 2010). The EPSs are commonly related to the carbonate mineral precipitation (Benzerara et al., 2006; Dupraz et al., 2004) since they represent the main component of biofilms (Flemming & Wingender, 2010). Also, EPSs could play the role of a low-energy substrate for crystal nucleation (EPS-mediated mineralisation; Della Porta et al., 2021). In pioneer samples of Aedipsos, EPSs and filaments commonly form a crystal retention lattice; thus, calcium carbonate crystals are trapped sticking on it (Figure 6D–H).

Finally, filamentous Cyanobacteria trap calcium carbonate grains/sediments precipitated by the highly supersaturated hot water, and hold them together (Figure 6C,I). This might be a way for Cyanobacteria to anchor at a specific site, especially in the early stages of attachment to the substrate.

CONCLUSIONS

The indigenous biodiversity of extreme environments, and the processes involved, such as biomineralisation, remain an interdisciplinary subject of high scientific interest. Nevertheless, knowledge is still limited on (a) the systematic identification of the pioneer microorganisms colonising an area and (b) the starting time and role of their participation in biomineralisation processes. In the present study, pioneer cyanobacterial colonists from extreme Greek environments (hot springs), developing on artificial slides in situ, were investigated.

Based on the acquired data, temperature seems to control biodiversity, while salinity and pH do not seem to play any major role in the early colonisation stages. While in mature stages the biodiversity was mainly controlled by temperature and salinity.

Synechococcales (37%) and Oscillatoriales (33%) are the dominant orders, followed by Chroococcales (15%) and Spirulinales (11%), while Chroococcidiopsidales (4%) were found only in a few samples and always with a minimum percentage. At the mature stages, the same orders are present with the additional presence of Nostocales, which is totally absent among the pioneers.

In total, 43 species of Cyanobacteria were identified as pioneer microorganisms in the thermal springs of Aedipsos. The most common pioneer species were found to be S. subtilissima, Synechococcus elongatus and Oxynema acuminatum. These species are also among the most common species in the later stages. The most common pioneers were found to be multicellular filamentous Cyanobacteria, that is, organisms with a larger outer surface with the exception of S. elongatus. The multicellular filamentous Cyanobacteria create larger amounts of EPSs due to their increased surface area, thus resulting in greater calcium carbonate binding ability compared to unicellular Cyanobacteria. Among the pioneers typical thermophilic species have been observed, such as C. thermalis, C. thermalis, Leptolyngbya thermalis, S. subtilissima and Symploca thermalis, as well as typical limestone substrate Cyanobacteria, such as C. lithophilus and L. laminosa.

The use of artificial slides—left only for a few hours in hot water sites—demonstrated that pioneer Cyanobacteria are participating in the formation of travertine from the beginning. Also, the observed biomineralisation processes include (i) calcification of Cyanobacteria sheaths, (ii) trapping of carbonate crystals on a crystal retention lattice formed by EPSs and filaments and (iii) trapping and confinement of carbonate crystals around filamentous Cyanobacteria. These processes, along with the abiotic formation processes, result in the formation of hybrid travertines from the point Cyanobacteria are first established.

Among the plans for future research is the study of pioneer species in extreme environments with different physicochemical conditions in an attempt to identify them and further understand the processes taking place in the first hours of colonisation.

AUTHOR CONTRIBUTIONS

C.K., V.L. and A.E.-A were involved in conceptualisation. C.K. and V.L were involved in sampling and writing of the paper and editing. A.P., V.L. and A.E.-A were involved in biological experiments and assessment. A.P., C.K. and P.V were involved in geological experiments and assessment. A.P., C.K. and V.L were involved in SEM analysis and geobiological assessment, statistical analysis and visualisation. C.K was involved in GIS.

A.E.-A., P.V., C.K. and V.L. were involved in major review of the paper. A.E.-A was involved in supervision.

ACKNOWLEDGEMENTS

The authors would like to thank the local population and authorities, and especially the Director of the Public Properties Company—Aedipsos branch, Ilias Siakantaris, for their co-operation during the fieldwork. The corresponding author would like to thank Dr. George Vougioukalakis from the Greek Geological Survey (IGME, present name Hellenic Survey of Geology and Mineral Exploration, HSGME) for his support and encouragement during this research. We acknowledge the constructive comments by the reviewer Assist. Prof. Marianna Kati and one anonymous reviewer and the editor that greatly improved the manuscript.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT

Data sharing is not applicable to this article as no new data were created or analyzed in this study.