Bacteria-like Ferruginous Structures in Carboniferous Limestones as Remains of Post-Variscan Hydrothermal Activity in Southern Poland

Abstract

Structures resembling iron-related bacteria (IRB) have been found in the Mississippian limestones that form part of the carbonate platform in the Moravo-Silesian Basin that surrounds the Upper Silesian Block, an eastern margin of the Brunovistulicum. Microfacial, petrological, and geochemical analyses were used to determine the bacteria-like structures that are present in narrow zones unrelated to bedding. We present here the morphology and chemistry of the studied microstructures showing their similarities to IRB from the present-day Sphaerotilus-Leptothrix group, the Galionella group, and the Mariprofundus ferrooxydans species. We suggest that bacterial growth occurred in the originally empty micropores of microfossil skeletons and shells, between bioclasts or in secondary voids formed during the selective dissolution of micrite or smaller sparite crystals. Hydrothermal solutions, associated probably with the post-Variscan magmatism in this area, provided iron compounds for the growth of the IRB.

1. Introduction

Single-celled organisms are able to actively participate in changes in environmental chemistry. They mediate oxidation and reduction reactions occurring in both aquatic and terrestrial environments, e.g., [1,2]. Such processes also occurred in the geological past and could be recorded in rocks, which we can read by deciphering fossilization processes, e.g., [3,4]. The fossilization of single-celled organisms is most likely if they precipitated extracellular inorganic coatings [5,6]. Such structures are usually left as the only remnants after the decomposition of bacterial cellular matter, providing valuable information on microbial life in ancient environments, e.g., [7]. The more resistant structures to later remobilization are metallic bacterial coatings such as those created by iron related bacteria (IRB) [8,9]. Such bacterial types are able to catalyze the oxidation of Fe (II) [10,11] and accumulate (oxyhydr)oxides, either outside the cells or in the surrounding extracellular polymeric matrices, e.g., [12].

IRB are well known from modern environments as well as from the fossil record. The earliest fossilized bacterial remnants, morphologically comparable with the modern IRB have been recognized so far from the well-known Precambrian Gunflint cherts [13,14], as well as in many Precambrian sedimentary successions. They also contributed significantly to the banded iron formations (BIFs) [3,13,15].

Bacterial iron encrustations have also been described from various sedimentary environments in many Phanerozoic successions, both freshwater, e.g., [16], and marine, e.g., [17,18,19]. They were recognized in fossil record from high-energy, shallow, e.g., [20], to calm, deep-water settings [7], where they can formed Fe-oolite sands [21,22], bacterial mats [17], stromatolites [23], or incrustations binding mud mounds [24]. Especially in marine environments, bacterially mediated amorphous iron oxyhydroxides combined with manganese oxyhydroxides and silica are commonly found in areas of mid-ocean and back-arc volcanic activity, covering vast areas of the ocean floor [25], also associated with hydrothermal fluid activity.

In the present study, we report morphological and chemical evidences for the presence of bacteria-like structures belonging to the IRB group in the Mississippian limestones along the Upper Silesian Block (Figure 1a,b), which is the northern part of the larger unit known as the Brunovistulicum, e.g., [26], and we try to indicate their origin and time of appearance in these rocks. A characteristic feature of the limestones consisting of the IRB-like structures is their red colour, which does not occur in underlying carbonate successions [more than 1 km thick], which formed on the epicontinental shelf stretching from the Western Europe to the Ukraine [27]. Previous studies suggested the sedimentary origin of the red colouration of these limestones, related to the admixture of hematite, which would be redistributed from the terrestrial environment [28]. Here, we test another hypothesis, which connects the occurrence of IRB-like structures with the activity of hydrothermal waters. In the study area, they could have occurred during two different geological periods. The first one could be related to the Pennsylvanian and Cisuralian post-orogenic granitoid plutonism and bimodal volcanism, e.g., [29,30]. And the second one could be related to the Late Triassic sulphide mineralization [mainly of Zn-Pb] of unknown fluid origin [31].

2. Geological Setting

The IRB-like remnants have been found in the uppermost part of the Upper Devonian–Mississipian succession of the ancient carbonate platform in the Moravo-Silesian Basin that surrounded the Upper Silesian Block, an eastern margin of the Brunovistulicum, e.g., [34,35,36]. This carbonate body developed approximately three hundred kilometres on a passive continental margin facing a deep-water basin of the Rhenohercynian system [37,38,39]. Due to sea-level rise and tectonic extension, successive drowning of this platform occurred [39,40], partly associated with its disintegration and the development of intra-platform troughs [41,42,43]. The closure of the Rhenohercynian ocean resulted in subsequent continental collision between Lugodanubian and subducting Brunovistulicum [44], which ultimately amalgamated the Upper Silesian Block with the Bohemian Massif [28].

During Pennsylvanian–Cisuralian post-orogenic stages, granitoid plutonism and bimodal volcanism took place in both foreland and the internal parts of the Variscan orogen, including the recent southern Poland [31,45]. Their numerous occurrences, e.g., [32,46,47,48,49,50,51,52], are associated with the Kraków–Lubliniec Fault Zone, which divides the Upper Silesian Block from the Małopolska Block (Figure 1b).

The Variscan basement containing the Devonian and Carboniferous sediments is covered discordantly by the succession of Permian–Mesozoic carbonate and siliciclastic sediments of the Central European Basin, dominated by the Triassic strata (Figure 1c). All of them dip toward the northeast, reflecting the vertical crustal mobility of this region [53]. The Middle Triassic limestones and dolomites (Muschelkalk) comprise Zn-Pb deposits, locally as resources of economic value, e.g., [31,54]. However, this Zn-Pb mineralization is hosted also by older (Devonian–Permian) and younger (Jurassic) rocks [55,56].

The studied Upper Devonian to Mississippian limestones of the shallow-water platform (bioclastic wackstones with brachiopods) are recently outcropped in the southernmost part of the Silesian-Kraków Monocline (southern Poland), 25 km west of Kraków (Figure 1c). In this area, the Middle Devonian–Mississippian carbonates (limestones and dolomites) are cut by several gorges containing, among others, the Racławka and Czernka valleys. Our microfacial and geochemical studies are based on samples from three sections located in these valleys. The section from the Racławka Valley (DR3; Figure 2) corresponds to the lower Tournaisian, whereas the two sections from the Czernka Valley (DC1, DC2; Figure 2) belong to the middle Viséan, based on analyses of foraminiferal assemblages [57]. Litho- and biostratigraphic studies of this carbonate succession have a history of over 160 years, as summarized in [57]. The assignment of sediments from the studied sections to lithostratigraphic units is based on the later classification [27]. Their chronostratigraphy is based on foraminiferal assemblages from the class Fusulinata studied in the same rock samples [57].

3. Material and Methods

3.1. Sampling

The criterion for selecting 18 samples for iron-related bacteria research was the reddish colour of the rocks in the fresh fracture or any ferruginous discolourations and coatings visible in subsequent microscopic observations. Samples were taken from three exposures in the selected area (Figure 1c).

Four samples of the Viséan biogenic limestones belonging to the Czerwona Ścianka Formation [27,55] were selected from the stratigraphically youngest (DC1) section (the Red Wall in the Czernka Valley (50°10′12.9″ N, 19°37′22.1″ E; Figure 2)). The characteristic feature of this location is the red colour of rock on the exposed surface (Figure 3a,b). Another distinguishing feature of the limestones is also the red colour on the borders around the bioclasts (Figure 3c,d).

Four rock samples were selected from the stratigraphically older (DC2) section (the Commune quarry in the Czernka Valley (50°10′09.7″ N, 19°37′24.2″ E; Figure 2)) representing also the Viséan limestones of the Czerna Formation [27,55]. This outcrop is located on the orographically left bank of the Czernka stream, near a local water intake. The profile contains grey to dark grey limestones on a fresh fracture, and beige to pinkish on a weathered surface. They are characterized by a lack of lamination, but within the outcrop scale, layering can be observed, with layers ranging from 1 to 3 m thick. There are sparse cracks (0.2 to 0.5 mm) secondarily filled with white calcite. In the thin-section view, limestones are pelitic mudstone-type and, locally, bioclastic wackstone. Lithoclasts composed of carbonate rocks are visible in individual samples. In the central part of the exposure, a succession of fine-bedded limestones with a nodular structure is visible. These limestones do not show the presence of macrofossils.

Four samples of massive dark grey Tournaisian limestones belonging to the Racławka Formation [27,57] were taken from the outcrop (DR3) on the slopes of the Komarówka Hill in the Racławka Valley (50°10′28.9″ N, 19°40′26.7″ E). In the axis of the main and lateral valleys, there are two faults along which tectonic blocks containing the Famennian and Tournaisian limestones were dislocated relative to each other (Figure 1c). The profile exposes pelitic limestones without lamination, light grey on the weathered surface and dark grey on the fresh fracture. Two-plane fractures dominate, ranging in thickness from 1 to 4 mm. Cracks filled with brick red calcite are visible in the upper part of the profile. In the fillings of larger cracks, the crack rim is brick red and the interior is filled with white, creamy calcite.

3.2. Petrology and Geochemistry Studies

Twenty standard uncovered petrographic thin sections with a size of 3 × 5 cm were made from the selected samples. All thin sections were cut perpendicular to the lamination. All thin sections were observed under a conventional transmitted light microscope (TLM) using a Nikon Eclipse LV100N Pol polarizing microscope with a digital camera (Nikon DS10) (Nikon Corporation, Yokohama, Japan). The thin sections are stored as Collection IRB-2-2024 at the Faculty of Geology, Geophysics, and Environmental Protection of the AGH University of Science and Technology, Kraków.

The chemical composition of the minerals was determined using the SEM-EDS method. Carbon-coated polished thin-section samples were analyzed using a JEOL 5410 microscope (JEOL Ltd., Akishima, Japan) with a Voyager 3100 (NORAN) EDS spectrometer (NORAN Instruments, Middleton, WI, USA) as well as a HITACHI S 4700 microscope (Hitachi High-Tech Corporation, Naka, Japan) with a Vantage (NORAN) spectrometer (NORAN Instruments, Middleton, WI, USA). The time of analysis was 100 s for a point at an acceleration voltage of 20 kV. The ZAF correction algorithm was used. These studies were performed in the Laboratory of Field Emission SEM and Microanalysis at the Institute of Geological Sciences, Jagiellonian University, Kraków. Additional scanning electron microscope observations regarding mineralogical studies of iron oxides and oxyhydroxides were carried out in low-vacuum mode, using a FEI 200 Quanta FEG microscope (Fei Company, Brno, Czechia) equipped with an EDS/EDAX spectrometer (at the AGH University of Krakow). The acceleration voltage was 20 kV, and the pressure was 60 Pa. The samples were not coated with any conductive layer.

Raman microspectroscopy measurements were performed with a Thermo Scientific DXR Raman Microscope (Thermo Electronic Scientific Instruments LLC, Madison, WI, USA). An Olympus (100×) objective lens (Olympus Corporation, Nagano, Japan)was used to focus the incident laser on a sample resulting in a spot size of ca. 0.7 μm. The spectra were recorded at room temperature using a 900 grooves/mm grating and a CCD detector. The excitation wavelength was 532 nm, and the power was 10 mW. An exposure time of 3 s and 100 data accumulations were used.

All these samples were analyzed for major and minor element concentrations at the Bureau Veritas Minerals Laboratories, Vancouver, Canada. Total abundances of the major oxides, several minor elements, rare Earth and refractory elements were analyzed by inductively coupled plasma (ICP) emission spectrometry, following lithium metaborate/tetraborate fusion and dilute nitric acid digestion. Loss on ignition (LOI) was determined by the weight difference after ignition at 1000 °C for >2 h. Moreover, separate 0.5 g samples were digested in Aqua Regia and analyzed by ICP mass spectrometry to determine the precious and base metals. The detection limits ranged from 0.002 wt% to 0.01 wt% for major oxides, from 0.1 ppm to 20 ppm for trace elements, and from 0.01 ppm to 0.1 ppm for the rare Earth elements. The CANMET- and USGS-certified reference materials were used as monitors of data quality. Normalized REE patterns are shown against a logarithmically scaled concentration axis. To calculate the Ce anomaly, we used an equation (Ce/Ce*)_N = Ce_N/(Pr_N × Pr_N/Nd_N) [58]. Gd anomalies were calculated using the Gd/Gd* ratio, where Gd* is the interpolated Gd concentration for a smooth Post Archean Australian Shale-normalized REE pattern, and Xn is the concentration of element X normalized to PAAS: Gd*_N = Sm_N^1/3 × Tb_N^2/3.

Based on standards and sample replicates, the precisions for abundances and element ratios (including Gd/Gd* ratios) are in most cases much better than 5% (2 RSD). Typical errors for gadolinium excesses in shells are assumed to be about 0.25 ng/g (2σ).

4. Results

4.1. Microstructure of Limestones with Ferruginous Pigments

The analysis of foraminiferal assemblages and microfacies showed that all samples studied are biogenic limestones that were deposited in a shallow marine environment [55]. They contain different percentages of calcareous bioclasts, which are mainly benthic foraminifers, mostly belonging to the fusulinids (Figure 4a–d). The remaining bioclasts are fragments of skeletal plates of echinoderms (Figure 4a). Other biotic particles are very rare. A characteristic feature of these limestones is sparite binder, which comprises crystals that vary in size (Figure 4e). The smallest calcite crystals (3–5 µm) occur in the recrystallized walls of foraminifera, and the largest (5–10 µm and larger) in the infill of originally empty chambers or pores in skeletons, as well as in the space surrounding bioclasts.

SEM-EDS images of selected surfaces of thin sections of rocks, which were characterized by the presence of redness, show their structural and chemical differences. Larger, light grey and dark grey areas seen in the BSE image (sometimes white-glowing points) correspond to the shell walls of foraminifers and other bioclasts, which is confirmed by TLM images (Figure 4b,c). Among them, white dots and streaks occur mainly around bioclasts or sparite crystals. In turn, dark grey areas indicate the filling of originally empty spaces in foraminiferal chambers and spaces between bioclasts. Chemical SEM-EDS microanalysis (Figure 5) indicates that light grey areas consist mainly of calcite, which is the main component of bioclasts. White streaks and glowing points contain an increased content of Fe oxides. In some places, these white peaks occur in microcrystalline quartz (Figure 6). On the other hand, dark grey areas contain Mg oxide and therefore may be partially composed of dolomite crystals (Figure 6).

Detailed microscopic analysis of all thin sections of the rocks studied show that only red Viséan limestones (section DC1) possess irregular rust to brown areas that could have been caused by iron-related bacterial activity (Figure 4e, Figure 5 and Figure 6). Microfacies analysis (TLM) shows that these pigmented areas contain iron oxides and oxyhydroxides, accompanied by individual iron-bearing microcrystals and/or aggregates. These ferrous discolourations can occupy areas of up to 40% of the microscopic view. They occur in various positions in relation to the sparite cement crystals. They are most common between crystals, on the edges and walls of crystals, forming a characteristic network (Figure 7). This is especially the case when the calcite crystals are small. Larger crystals contain individual oval structures of 1–2 µm in size enclosed within them (Figure 8 and Figure 9). Rust-coloured and brownish structures may form thin streaks between crystals or thicker coatings with raised, oval shapes and blurred edges, whereas when these structures are thicker, they become dark brown to black inside and have oval to convex shapes that can be single or clustered in larger black areas. In most cases, these areas were the site of growth of Fe-bearing crystals.

4.2. Mineralogy of Ferruginous Discolourations in Limestones

The microscopic images (TLM) show that the sources of the reddish and yellowish discolouration of the studied rocks, which gives the overall macroscopic colour and creates rusty coatings around fossils and bioclasts, are numerous veins filled with several generations of minerals (Figure 10a,b).

SEM observations and SEM-EDS analyses supported by Raman microspectroscopy indicate that the Fe oxyhydroxides and oxides present in the studied samples are goethite (α-FeOOH) and hematite (α-Fe₂O₃), both of which form very diverse morphological forms (Figure 10 and Figure 11). These are very finely dispersed crystals and indentations, several µm in size, which are responsible for the majority of the colour of the entire rock. Furthermore, pore and vein fillings are common, as are pseudomorphs after pyrite (e.g., Figure 11d) and rounded porous aggregates, likely of bacterial origin (Figure 11c). The collected Raman spectra indicate that goethite and hematite do not occur together; individual samples contain only one or the other mineral. Goethite is the dominant iron carrier in samples DC1.1 and DC1.2, which are characterized by a yellowish colour. Here, small (max 2–3 µm) crystallites (most often filling pores) can be encountered, with the acicular shape characteristic of this oxyhydroxide. In contrast, in specimens with a dominant reddish colour (samples DC1.3, DC1.4.1, DC1.4.2, DC1.5), the colouring agent is hematite, in places forming quasi-spherical or plate-like crystallites. Variation in colour within a single sample (e.g., DC1.4.2) appears to be unrelated to mineralogical diversity (in this particular case, the only Fe form detected is hematite), but rather to the degree of mineral accumulation in different locations. There is also no evidence of significant differences in chemical composition between zones of different colour within a single sample. However, subtle differences in the composition of iron compounds between samples are noticeable. Some samples do not contain EDS-detectable admixtures of other transition metals (as in sample DC1.3) but do contain admixtures of Zn (samples DC1.1, DC1.2, DC1.5), Ti (samples DC1.1, DC1.2), and Mn (in some locations in samples DC1.4.1 and DC1.5). In sample DC1.2, Mn oxides with significant Pb content are occasionally found alongside goethite. The chemistry and morphology of the fine crystallites suggest that it is coronadite Pb_1-1.4(Mn⁴⁺,Mn³⁺)₈O₁₆ (Figure 10f,g).

Other non-carbonate components of the limestones include Fe sulphides (pyrite), quartz, and kaolinite (Figure 10e and Figure 11c). Barite (Figure 11d), Ti oxides (rutile), and fluorapatite were also encountered. Pyrite most often occurs as scattered single euhedral or subhedral crystals, with sizes of up to several µm. Framboidal clusters (up to 10 µm in diameter) composed of crystals < 1 µm in size can also be found (sample DC1.1). Pseudomorphs of Fe oxides after FeS₂ are common alongside fresh, unoxidized pyrite crystals. Quartz and kaolinite typically occur in zones enriched in iron oxides and oxyhydroxides, as evidenced by elemental maps indicating correlations in Fe, Si, and Al concentrations. Both silicates are often euhedral, and euhedral quartzes with visible crystal growth zones are also found.

4.3. Geochemistry of Major Oxides and Trace Elements

The rocks studied show strong differences in the content of major elements (

Table 1. Content (wt%) of major elements in the rock samples studied. DC1—strong hydrothermally altered red limestones (the Czernka Valley—the Red Wall); DC2—weak/poorly hydrothermally altered pinkish grey limestones (the Czernka Valley—the Commune quarry); DR3—non-hydrothermally altered grey limestones (the Racławka Valley—slopes of the Komarówka Hill).

	DC1 (n = 4)	DC2 (n = 4)	DR3 (n = 10)
CaO	51.88–44.70 (av. 48.98)	55.51–49.72 (av. 53.85)	55.34–54.16 (av. 54.79)
Σ CaO + LOI	94.38–82.00 (av. 89.40)	98.82–90.22 (av. 95.57)	99.03–98.02 (av. 98.59)
Σ other elements	17.81–5.56 (av. 10.49)	9.83–1.19 (av. 3.44)	2.02–0.97 (av. 1.40)
SiO2	12.14–3.28 (av. 6.80)	8.04–0.59 (av. 2.58)	1.06–0.41 (av. 0.66)
MgO	1.26–0.30 (av. 0.68)	0.33–0.24 (av. 0.27)	0.66–0.40 (av. 0.58)
Fe2O3	1.69–0.44 (av. 0.92)	0.43–0.07 (av. 0.17)	0.13–0.05 (av. 0.07)
Al2O3	3.34–0.46 (av. 1.84)	0.88–0.02 (av. 0.33)	0.19–0.06 (av. 0.09)
K2O	0.14–0.03 (av. 0.09)	0.07–0.01 (av. 0.03)	0.06–0.01 (av. 0.02)
TiO2	0.10–0.02 (av. 0.06)	0.03–0.01 (av. 0.02)	0.01
MnO	0.07–0.02 (av. 0.05)	0.03–0.01 (av. 0.02)	0.01
P2O5	0.06–0.01 (av. 0.04)	0.01	0.01
Na2O	0.03–0.01 (av. (0.02)	0.01	0.01

). In the Viséan red limestones (DC1 section), the concentration of CaO is markedly lower (44.70–51.88 wt%), in contrast to the sum of other components that are not geochemically related to limestones, i.e., SiO₂, MgO, Fe₂O₃, Al₂O₃, K₂O, MnO, TiO₂, P₂O₅, and Na₂O. The content of SiO₂ is high (3.28–12.14 wt%), as well as the contents of Al₂O₃ (0.46–3.34 wt%) and Fe₂O₃ (0.44–1.69 wt%), while the amount of MgO is relatively small (1.26–3.28 wt.%). The contents of other elements are also clearly higher than the average contents for marine limestones [59].

The samples from the second section of the Viséan rocks (DC2) are characterized by the higher amount of CaO (49.72–55.51 wt%) and a lower sum of other components (1.19–9.83 wt%). The concentration of SiO₂ is lower (0.59–9.83 wt%), similarly to the content of Al₂O₃ (0.02–0.88 wt%), Fe₂O₃ (0.07–0.33 wt%), and MgO (0.24–0.33 wt%). The abundance of other elements is lower (K₂O < 0.07 wt%; TiO₂ < 0.03 wt%; MnO < 0.03 wt%).

The quite different composition of main oxides characterizes the Tournaisian limestones (DR3 section). The high content of CaO (54.16–55.34 wt%) and low amounts of SiO₂ (0.41–1.06 wt%), Al₂O₃ (0.05–0.19 wt%), and Fe₂O₃ (0.04–0.13 wt%) are typical of limestones. The only amount of MgO (0.40–0.66 wt%) is clearly higher in this succession. The abundances of other oxides like TiO₂, MnO, P₂O₅, and Na₂O do not exceed 0.01 wt%. Only the K₂O content is slightly higher (up to 0.06 wt%).

4.4. Correlation Between Main Elements and SEM-EDS Mapping

Pearson’s coefficient values, calculated for individual pairs of oxide content in all samples from the DC1 section (

Table 2. Pearson’s coefficient values, calculated for individual pairs of oxides in all samples from the DC1 section (the Czernka Valley—the Red Wall).

	SiO2	Al2O3	Fe2O3	MgO	CaO	Na2O	K2O	TiO2	P2O5	MnO
SiO2		0.93352	0.93112	−0.08947	−0.98424	0.25506	0.8845	0.92725	0.78691	0.7502
Al2O3	0.93352		0.96874	−0.10711	−0.95836	0.34937	0.9138	0.99673	0.93623	0.76123
Fe2O3	0.93112	0.96874		−0.012	−0.9598	0.35121	0.89023	0.96939	0.86091	0.70297
MgO	−0.08947	−0.10711	−0.012		−0.02716	0.1545	−0.01154	−0.09549	−0.09094	0.30166
CaO	−0.98424	−0.95836	−0.9598	−0.02716		−0.33029	−0.91739	−0.95411	−0.83796	−0.79583
Na2O	0.25506	0.34937	0.35121	0.1545	−0.33029		0.44729	0.39502	0.53825	0.32657
K2O	0.8845	0.9138	0.89023	−0.01154	−0.91739	0.44729		0.9092	0.85725	0.6789
TiO2	0.92725	0.99673	0.96939	−0.09549	−0.95411	0.39502	0.9092		0.94817	0.75895
P2O5	0.78691	0.93623	0.86091	−0.09094	−0.83796	0.53825	0.85725	0.94817		0.75491
MnO	0.7502	0.76123	0.70297	0.30166	−0.79583	0.32657	0.6789	0.75895	0.75491

), show strong linear correlation between SiO₂ and Al₂O₃, Fe₂O₃, K₂O, TiO₂, P₂O₅, and MnO. Moreover, there is a strong negative correlation between CaO and Fe₂O₃, SiO₂, Al₂O₃, K₂O, TiO₂, P₂O₅, and MnO. There is also significant negligible correlation between MgO and all of the above mentioned oxides, which is characteristic for all samples investigated. The same trend is distinctly visible for samples of all sections investigated (DC1, DC2, DR3) (Figure 12), although individual profiles were affected to a different extent by the impact of hydrothermal solutions, which was reflected in the different contents of individual elements.

The above tendency observed in the analytical results is also clearly depicted in the SEM-EDS mapping (Figure 6). The microanalysis of Si, Fe, Ca, Mg, Al, and K clearly shows that the micro-area of increased SiO₂ and Fe oxide and oxyhydroxide contents overlap. This dependence is clearly visible in the images from rocks in profile DC1, where the weight of both oxide groups is relatively high. The same area also coincides with the fields of Al and K oxides in the same samples.

The area of co-occurrence of SiO₂ and Fe oxides and oxyhydroxides and (Al, K) elements is generally separated from the area with increased CaO and MgO values, which is associated primarily with coarse crystalline cement filling the original porosity and generally does not contain Fe-bearing browning.

4.5. Trace Elements

The composition of trace elements in these limestones displays similar differences in the sections studied (

Table 3. The content of trace elements (ppm) in the rock samples studied. DC1—strong hydrothermally altered red limestones (the Czernka Valley—the Red Wall); DC2—weak/poorly hydrothermally altered pinkish grey limestones (the Czernka Valley—the Commune quarry); DR3—non-hydrothermally altered grey limestones (the Racławka Valley—slopes of the Komarówka Hill).

	DC1 (n = 4)	DC2 (n = 4)	DR3 (n = 10)
Sr	758.3–354.4 (av. 607.4)	155.0–109.2 (av. 140.7)	329.5–184.8 (av. 187.7)
Rb	5.4–3.0 (av. 3.7)	1.7–0.1 (av. 0.6)	2.6–0.4 (av. 1.1)
Ba	30–18 (av. 26)	30–4 (av. 11)	6–2 (av. 4)
Mo	1.1–0.4 (av. 0.6)	0.2–0.1 (av. 0.2)	0.3–0.1 (av. 0.17)
Cu	4.6–2.0 (av. 3.5)	2.8–1.1 (av. 1.8)	2.1.7–0.2 (av. 0.84)
Pb	191.0–8.8 (av. 83.5)	33.2–3.5 (av. 14.8)	5.0–0.7 (av. 2.3)
Zn	298–15 (av. 151)	86–17 (av. 36)	19–15 (av. 10.7)
Zr	35.2–5.0 (av. 19.2)	35.9–1.7 (av. 11.8)	7.8–2.8 (av. 2.3)
U	3.7–1.3 (av. 2.5)	2.6–0.7 (av. 1.6)	4.1–1.1 (av. 2.5)
Th	5.7–0.7 (av. 2.9)	3.3–1.0 (av. 1.8)	0.2
Hf	1.0–0.1 (av. 0.6)	0.9–0.1 (av. 0.3)	0.3–0.1 (av. 0.1)
ΣREE	86.6–20.9 (av. 47.8)	36.1–3.5 (av. 18.3)	13.1–2.5 (av. 6.5)

). The red limestones from the Czernka Valley (DC1, DC2) are particularly distinctive in this respect. The DC1 samples are characterized by enhanced content of several elements in relation to samples from the Racławka Valley (DR3). This applies to the following elements: Zn (up to 298 ppm), Pb (up to 191 ppm), Sr (up to 758 ppm), V (up to 36 ppm), Ba (up to 30 ppm), As (up to 15.1 ppm), Th (up to 5.7 ppm), Nb (up to 4,5 ppm), U (up to 3.7 ppm), Zr (up to 35.2 ppm), Mo (up. To 1.4 ppm), Hf (up to 1 ppm), Sb (up to 1.0 ppm), Ga (up to 1.3 ppm), and Hg (0.4–0.7 ppm).

In the second section of the Viséan limestones (DC2), the increased values refer only to one sample (DC2.08), where some trace elements, i.e., Zn (86 ppm), Ba (30 ppm), Zr (35.9 ppm), Th (3.3 ppm), U (1.6 ppm), Nb (1.6 ppm), Hf (up to 0.9 ppm) have high contents, comparable to the DC1 samples. The abundances of Pb (up to 33.2 ppm), Zn (up to 86 ppm), Cu (up to 2.8 ppm), and Mo (up to 0.2 ppm) are markedly lower.

The DR3 samples have small amounts of the most trace elements compared to the rock samples described above (Table 3). Only the content of Sr (184.8–329.5 ppm) is medium, and it is higher than that of the DC2 samples.

4.6. REE Signatures

The studied sediments display a variable content of REE (Table 3). In the red limestones (section DC1; av. 60.3 ppm), it is higher than in the grey limestones containing pink spots (section DC2; av. 23.2 ppm), and much higher than in the grey limestones containing ferruginous discolourations and coatings visible in microscopic observations (section DR3; av. 5.7 ppm). The chondrite- and PAAS-normalized REE + Y patterns were similar between the sections in the Czernka Valley (DC1 and DC2), emphasized by Ce- and Eu-negative anomalies with Gd- and Y-positive anomalies (Figure 13). This pattern for limestones from the Racławka Valley (section DR2) differs among the Light-REE in the lack of cerium anomaly. The materials studied from the Czernka Valley exhibit superchondritic Y/Ho ratios (av. 35–57 in particular sections). In turn, the lowest values (32–39) are characteristic of the DC1 section. The Ce/Ce* average values in limestones vary significantly between the sections from 0.73 (DC1 section) to 1.6 (DR3 section). The values of the Dy_N/Sm_N ratio, a signature of diagenetic enrichment in the MREE [60], are above 1.2, excluding a single sample in the Czernka Valley (DC1.01). The Gd/Gd* average values in limestones vary significantly between the sections from 20 (DC1 section) to 160 (DR3 section).

5. Discussion

5.1. Bacterial Origin of Rust-Coloured and Brown Pigments

The thin-section views of samples from the DC1 profile, using TLM at magnification x 1000, show spherical to oval dark brown structures measuring 1–2 μm across (Figure 4e, Figure 7, Figure 8, Figure 9, and Figure 14). They occur as single capsules or tubes enclosed in a spary binder (e.g., Figure 14a–c), arranged linearly (Figure 14a) between crystal walls (e.g., Figure 4e), or appear in groups of several specimens (Figure 11c and Figure 14c). The single capsule is hollow inside and has a wall that is about 0.1 µm thick. SEM-EDS and Raman analysis showed that the capsule walls contain a large amount of iron oxides and hydroxides (Figure 6 and Figure 11), which may be a remnant of the original bacterial sheath from the IRB group. Oval to elongated structures of various sizes resembling bacterial cells are often arranged linearly or in rounded consorcia (Figure 11c and Figure 14a–c). This indicates a similarity to IRB from the Sphaerotilus-Leptothrix group due to the arrangement of individual cells in bacterial sheaths. In some places, empty tubular structures are visible, which may also be remnants of specimens of this group of bacteria (Figure 14c). Some of the specimens possess stalked, twisted ribbon-like structure (Figure 14b) resembling IRB from the Galionella group or another Fe oxidizer like Mariprofundus ferrooxydans, e.g., [64,65,66].

Further evidence for the association of rusty discolourations with bacterial origin are oval, dark brown pigment clusters usually surrounded by rusty, translucent haloes with an unevenly coloured, floccular structure, visible under a light microscope (Figure 14d). These structures resemble the bacterial holdfasts, whose general outline and shape might be similar to that of bacteria from the Sphaerotilus-Leptothrix group. Particular similarity to the present-day species of Leptothrix lopholea and Leptothrix cholodnii can be seen [5,67].

The structures that occur in the examined samples from the DC1 outcrop can be classified as bacteria only on the basis of the above mentioned features because in the case of fossil material from the late Paleozoic, this is the only available criterion. It is not possible to assess their belonging to any group of bacteria (or even Archaea) by applying microbiological and chemical methods to modern material. For this reason, we stick to the term “bacteria-like″ or “resembling bacteria″ to fully reflect the above impossibility of classifying the above structures as bacteria. However, in some cases, especially when bacteria precipitate inorganic compounds such as iron oxides or calcium carbonate outside the cell membrane, in the fossil record, they may preserve unique shapes and various types of bacterial sheaths and filaments, corresponding to features of specific bacterial species cf. [66].

5.2. Timing for IRB Growth

The rocks in which the bacteria-like remnants are found have a long geological history. They belong to a nearly 1200 m long sequence of the Upper Paleozoic carbonate platform [27]. Carbonate sedimentation occurred in a relatively shallow part of the epicontinental Moravo-Silesian Basin (MSB), along the shelves bordering the Upper Silesian Block—a part of a larger tectonic unit—the Brunovistulicum [34,36,68]. The older basement of this tectonic structure is interpreted as an ancient piece derived from the Gondwana continent [69].

The taphonomic analysis shows that bacteria-like remnants are not associated with the sedimentation in shallow marine environment of the original limestones. First of all, bacteria are present in part of the limestone with sparite cement, which occurs in the studied rocks as narrow zones of several millimetre to several centimetre, unrelated to bedding and cutting rocks in different directions (Figure 10a,b), also across microfossils as different types of foraminifers or echinoderm plates (Figure 7, Figure 8 and Figure 9). Rusty pigmentation occurs especially in the originally porous parts of the rock, which are microfossil chambers or porous walls of skeletal remains (Figure 3c,d, Figure 8, Figure 9 and Figure 11e). Additionally, bacteria-like structures are not associated with bacterial mats, which could have formed in the sedimentary environment of the host rocks at that time [57].

In order to draw conclusions about the origin of bacteria in the studied rocks, it is important to note that bacterial remnants are present locally, only in samples from the DC1 profile. In other profiles (DC2, DR3), bacteria were not observed despite the fact that they represent similar lithotypes of shallow-water limestones.

The microfacies analysis of the reddish colourations inside the microfossil walls and around them (Figure 3c,d), as well as shape of the walls and their vicinity (Figure 8, Figure 9 and Figure 15), shows that their origin is partly related to the dissolution process. The thinnest infillings were developed as seams bordering bioclasts, which usually consist of calcite spare, while they were surrounded by micrite. Reddish (iron oxide) seams started to develop in places where calcitic bioclast contacted micrite. This shows that fissures can be opened by leaching micrite from the host limestone and successively filling them with iron oxides, which are associated with calcite spar and quartz crystals (Figure 15). The more resistant parts of foraminiferal chambers or echinoid plates were left as residue (Figure 11e). Additionally, the crystals of calcitic spar are rounded (Figure 7), indicating their stepwise dissolution and crystallization during subsequent hydrothermal pulses.

This may indicate that bacterial growth occurred in the originally empty micropores of the skeletons and shells of microfossils, between bioclasts, or in secondary voids formed during the selective dissolution of micrite or smaller sparite crystals. The spatial relationship of sparite crystals to iron deposits and bacteria-like remnants indicates that iron oxide precipitation and sparite crystallization were contemporaneous processes. For the development of IRB, bacteria had to be solutions with iron content increased to such a level that individual species of bacteria could encrust cell membranes during their life. Samples from the DC1 profile contain increased contents of Fe, and in addition to this, other elements constituting a trace of the influence of hydrothermal conditions on these rocks. These solutions had to be relatively cold if IRB could live during their circulation. Another indicator of low-temperature hydrothermal activity is enrichment in elements like Ba, Sr, As, and Sb, e.g., [70]. Significantly increased contents of these elements, compared to samples from other tested profiles, occur in the DC1 profile.

Among the studied platform succession sediments, the IRB-like structures were not found in the Tournaisian limestones (DC2 profile), whose location right next to the same fault as the red limestones from the DC1 exposure suggests the possibility of their infiltration by hydrothermal waters. However, among the oxides whose elevated content may indicate hydrothermal origin, the composition and weight fraction differs from that in the DC1 profile. Additionally, the content of iron oxides and hydroxides is over 50 percent lower. Differences in chemical composition suggest different origins, and perhaps different durations, of hydrothermal solutions in these nearby areas. At the same time, the lack of visible structures resembling IRB indicates low iron concentrations in the original solutions circulating in the rock, which were insufficient for bacterial formation to produce inorganic coatings. This may also indicate the higher temperature of the solutions interacting in this part of the carbonate complex.

In turn, the limestone succession from the DR3 profile is characterized by the typical marine limestone content of oxides, accessory elements, and rare earth elements. The biotic components and the state of preservation of the clasts indicate a high-energy sedimentary environment. Therefore, the sparite present in these samples likely crystallized in the sedimentary environment and does not originate from the subsequent recrystallization of micrite during the circulation of hydrothermal solutions.

The local scope of influence of hydrothermal solutions and the possibility of the temporary development of IRB in them only in the area of neptunian dikes is evidenced by the fact that total REE content measured from the studied limestones is comparable to that of normal seawater (25–26 ppm; Table 3; e.g., [71]) or even lower in the case of the Si-enriched dyke (20 ppm). The same conclusion is suggested here on the basis of the PAAS-normalized REE patterns of the material from the dykes that were more or less flat, excluding the Y enrichment (Figure 13). This is similar to carbonate and authigenic marine phases, which mainly produced a seawater-like REE pattern [72]. The REE of the host rock was slightly higher (ca 20%; Table 3), which may reflect the contamination of phosphates in the echinoderm-foraminiferal limestone observed in thin sections of the rock (Figure 11e).

5.3. Potential Sources of Hydrothermal Fluids

Here, we consider two potential fluid sources, the first of which could be related to post-Variscan magmatism and volcanism, and the second, to a stage of Mesozoic sulphide mineralization. The first of these episodes is well documented in the study area, and the second, in its immediate vicinity.

Variscan compressional deformations in this area, which took place at the turn of the Moscovian and Kasimovian (Westphalian and Stephanian), resulted in the formation of gentle folds and faults with a latitudinal course and strike-slip character [44]. In this phase, the Dębnik anticline was formed, among others (Figure 1c). Starting from the Gzhelian (Stephanian B), this area experienced a phase of accumulation of post-orogenic deposits (arkosic sandstones and conglomerates containing silicified Dadoxylon tree trunks) and the onset of terrestrial volcanism, which lasted until the Early Permian, e.g., [73]. Volcanic, subvolcanic, and volcanoclastic rocks occur in this area on a surface of approx. 200 km² in the form of lava flows and small intrusions in the form of domes, laccoliths, and veins, and local tuffs of an ignimbrite nature, e.g., [47,48,49,50,51,52,74]. The intrusions are local and do not form continuous covers, although they are associated with the Kraków–Lubliniec tectonic zone, which divides the Upper Silesian Block from the Małopolska Block (Figure 1b) [34,45]. This volcanism was bimodal, from andesitic to rhyolitic, and a history of more than 100 years of research indicates that both varieties of volcanites appear to be genetically related and derived from magmas originating from the lithospheric mantle and crust, e.g., [30].

Zircon U–Pb dating of various igneous rocks (granodiorites, dacites, lamprophyre, and diabase) from this igneous belt showed that magmatism spanned within a short-lived event, between 303.8 ± 2.2 and 292.7 ± 4.9 Ma [30]. Felsic calc-alkaline magmatism was a characteristic for the oldest interval (during the latest Carboniferous), followed by the slightly alkaline volcanism of mafic–intermediate composition (the oldest Permian). The last one concerns the lamprophyre and diabase dikes located near the study area, in the 20 km wide zone along the Kraków–Lubliniec tectonic zone [30]. The oldest of them is the so-called granodiorite II from the Zawiercie region located in the Lubliniec–Kraków Fault Zone, which has been dated (Sr-Rb) to 340 ± 4 Ma [75].

This type of magmatism is associated with the Cu-Mo mineralization in this area occurring both in porphyry-type deposits, and in the surrounding skarn-contact metasomatic and vein-type deposits [46,75,76,77,78,79,80,81,82,83], forming a system of quartz veins with ore minerals. This mineralization was related to felsic magmatism and could be the most likely source of the elevated trace element values found in the Czernka Valley. In turn, the younger alkaline volcanism is characterized by the REE mineralization of the metasomatic hydrothermal type [84], which was not confirmed in the examined deposits.

Another possibility of the origin of the fluid sources for IRB that we considered could be related to the Mississippi Valley-type Zn-Pb mineralization hosted mainly by the Middle Triassic limestones and dolomites (Muschelkalk), which is of economic importance in a neighbouring region, just west of the study area [31,54]. Most probably, these ore fluids have also been related to the Variscan igneous rocks, which gave rise to the polymetallic ores, and later, they became remobilized by heat flow and redeposited in various sedimentary successions [85]. The timing of the remobilization of older ores and the origin of fluids are still debated. They are most likely related to crustal extension during the Mesozoic plate–tectonic reorganization (Late Triassic according to [85] or Early Cretaceous according to [86]). This type of mineralization is also known in association with Devonian–Permian rocks, which occur close to the studied sediments, in the Czernka Valley [55]. This proximity to the studied DC1 and DC2 profiles, and the fact that Pb-Zn mineralization was associated with low-temperature hydrothermal fluids, leads us to consider this Fe source for the IRB growth. However, our geochemical data are insufficient in this respect, and geological publications contain data on trace element content without detailed information on whether they originate from mining fields or their transition zones, or from “pure″ limestones or dolomites, e.g., [87]. For this reason, we leave this issue open for further geochemical studies.

6. Conclusions

• In the Viséan part of the thick limestone succession located along the Upper Silesian Block (northern part of the Brunovistulicum), there are local zones of red-coloured rocks, which, based on morphological and chemical evidence, have been classified here as bacterial structures. These are spherical or oval forms, 1–2 µm in diameter, that occur as individual capsules or tubes arranged linearly between the walls of calcite crystals. The single capsule is hollow inside and has a wall that is approximately 0.1 µm thick, which contains a large amount of iron oxides and oxyhydroxides; this may be a remnant of the original bacterial sheath. Some oval, dark brown pigment clusters, surrounded by translucent halos with a floccular structure, resemble the bacterial holdfasts. We present here the morphology and chemistry of the microstructures showing their similarities to IRB from the present-day Sphaerotilus-Leptothrix group, the Galionella group, and the Mariprofundus ferrooxydans species.
• The taphonomic analysis shows that bacteria-like remnants are not associated with the sedimentation of shallow marine carbonate mud. They are present only in narrow zones that are a few millimetres to several centimetres wide, unrelated to bedding, and they also occur across various microfossils. Based on microfacies analysis, we indicate that bacterial growth occurred in the originally empty micropores of microfossil skeletons and shells, between bioclasts, or in secondary voids formed during the selective dissolution of micrite or smaller sparite crystals. The dissolution of primary micrite and the subsequent filling of spaces with iron oxides/xyhydroxideswere associated with repeated hydrothermal pulses. The solutions must have been relatively cold, which is evidenced by the increased content of Ba, Sr, As, and Sb.
• Taking into account the types of mineralization associated with post-Variscan plutonism and volcanism, well documented in the immediate vicinity of the studied Mississippian sedimentary succession, we suggest that the hydrothermal pulses that associate with the formation of bacterial structures are probably related to felsite magmatism (Pennsylvanian), which was characterized, among other, by Cu-Mo mineralization, typical of many neighbouring porphyry deposits and skarn metasomatic deposits. However, this issue requires further geochemical studies on these sediments, in conjunction with analogous studies on carbonate rocks around the Zn-Pb mineralization zones, but outside the ore bodies and their transition zones.