The degradation of exopolymeric substances (EPS) by heterotrophic bacteria, concomitant release of calcium ions and precipitation of carbonates were studied in a temperate mountain lake, Lac d'Ilay, Jura France. Phytoplankton blooms in this lake produced large amounts of exopolymeric substances (EPS; 1.8-3.0 mg L-1), probably inhibiting CaCO3 precipitation by binding Ca2+ as shown by the saturation index of calcite and aragonite remaining well below 1. EPS settled to the sediments, where additional polymeric substances were produced by the benthic community. The total amount of EPS decreased downcore from ca 50 mu g/g dry sediment near the surface to ca 1.5 mu g/g dry sediment at the bottom (120 cm depth). A decrease in acidity, protein and sugar content, and calcium-binding capacity of EPS with depth coincided with active calcite precipitation. Aerobic and anaerobic EPS-degrading heterotrophic enrichments were obtained from the top, middle and bottom of the core. Doubling times of aerobic cultures from the top were six times shorter than those of cultures obtained from the bottom of the core, but anaerobic growth rates were similar across all enrichments. Aerobic turnover rates of organic compounds decreased by a factor of 4-5 from top to bottom; anaerobic rates were similar at all depths, except for the turnover of polymers, which was negligible at the surface compared to rates at the middle and bottom. All enrichments released calcium when grown on EPS. Growth on calcium-saturated EPS in anaerobic cultures obtained from the bottom of the core was the slowest, but still released 26% of the Ca in 20 days. This release during EPS degradation explained an increase in free calcium ions with depth reported in a previous study and may account for a large fraction of the carbonate mud. This suggests that sediments should be considered as an important source of biogenic carbonates.

Barchanoid dunes fields are commonly observed in both desert and coastal areas. However, while they are relatively well known in arid settings, their internal architecture, dune morphology and evolution processes in coastal environments are rarely documented in the literature. This study investigates coastal barchanoid dunes along the Cap Ferret Spit (southwest France) by combining high-resolution LiDAR data (1 m spatial resolution) with an extensive ground-penetrating radar (GPR) survey (>70 km of profiles). Five distinct dune morphotypes are identified along an ocean-spit-lagoon continuum, as modern foredune, embryonic dunes, proto-barchans, barchanoid ridges, and isolated barchans. Morphometric analyses reveal systematic relationships between dune size, elevation, and internal structure, with crest elevations ranging from 5 m in proto-barchans to 40 m in barchanoid ridges and isolated barchans. GPR profiles image lee-side accretion strata, reactivation surfaces, superimposed structures, and dune-dune interaction features, demonstrating that proto-barchans grow through stacking before evolving into barchanoid ridges and subsequently fragmenting into isolated barchans. The Cap Ferret dune system enables observation of preliminary and intermediate developmental stages of barchanoid dunes rarely documented in the literature, from embryonic dunes through proto-barchans and compound forms to isolated barchans. Second-order morphologies, including superimposed bedforms and wavy patterns with transverse orientation, occur systematically across all dune types and record short-term reactivation processes driven by dominant winds and potential vegetation interactions. These findings provide a reference model for coastal dune morphodynamic transitions including interactions between aeolian, hydrologeological and biotic processes. It also contributes to refining conceptual and numerical models of dune-field evolution by incorporating coastal boundary conditions.

In many estuaries, biogeochemical investigations have often focused on transient diatom biofilms that form on low-energy intertidal flats. Studies on microphytobenthos in high-energy sedimentary environments are unusual. The present investigation focuses on the biogeochemistry to a depth of 6 m of a fluvio-estuarine point bar from the Garonne channel (SW France) impacted by both tidal current and tidal wave, where three sediment cores were taken. Porewater chemistry was analysed with microelectrodes (pH, oxygen and sulfide), ion chromatography and inductively-coupled-plasma spectrometry (for major elements) and colorimetric assays (for iron speciation). Porewater composition was compared to measurements of microbial activity including isothermal calorimetry and metabolic assays using triphenyltetrazolium chloride and fluorescein diacetate to determine the distribution of predominant microbial metabolisms in the sediment. Finally, bulk sediment chemistry was characterized through X-ray fluorescence core scanning. Sediments are heterolithic, made of decimetre to meter thick alternating sand and mud. The uppermost 60 cm of the point bar sediment show a mostly classical vertical succession of microbial metabolisms: (i) oxygenic photosynthesis occurs mostly in diatom biofilm forming in the uppermost millimetres; (ii) aerobic respiration between 0 cm and 1 cm, (iii) nitrate reduction between 6 cm and 16 cm, partially overlapping (iv) sulfate reduction between 10 cm and 25 cm, (v) manganese oxide reduction below 2 cm and (vi) iron oxide reduction below 16 cm. Measurements of metabolic activity, elevated in areas showing significant geochemical changes, confirmed the impact of microbial metabolism on the composition of pore water. The highest metabolic activity coincides with areas where oxygen, nitrate and sulphate concentrations are decreasing. Hydrolytic activity peaked in the zone of aerobic respiration, possibly in part due to enzymatic degradation of organic matter (e.g., extracellular polymeric substances) produced in surface diatom biofilm. Low concentrations of nitrates and sulfates were measured in sands at 1.3 to 1.6 m and 3.2 m depth, coinciding with a renewed increase in hydrolytic activity and metabolically active cells. Because of the sediment heterolithic composition and the point bar architecture made of laterally accreting layers, subsurface advection of porewater through permeable horizons could explain the local increases of nitrate and sulfate reduction. Impacts of microbial metabolism on early diagenesis were modelled using PHREEQC software and outcomes predicted the potential precipitation of metastable iron and/or sulfides. This was confirmed by X-ray fluorescence analyses showing a coinciding increase of sulfur, Fe and/or Mn at several depths (e.g., 15 to 60 or 560 to 580 cm). Based on our observations, we propose a biogeochemical model that links microbial metabolisms and early diagenesis to the complex vertical sedimentary architecture of an estuarine point bar. Our results show that high-energy estuarine point bars are subject to an active biogeochemical cycling of C, S, N, Fe and Mn quite similar to that of intertidal mudflat, but locally altered by the sedimentary architecture of the point bar, resulting in lateral advection of porewater.

Barchans are crescent-shaped dunes with their convex side facing the wind and their horns pointing downwind, commonly observed in deserts but also in coastal environments. This study investigates the dune field of the Cap Ferret Spit (5-km wide, 23-km long; Atlantic Coast, SW France) including of series of barchanoids dunes, currently stabilized by forest cover. The aims are to classify each dune into distinct typologies and characterize their morphology in order to analyze their evolution. LiDAR-based analysis identified four distinct barchanoid dune types. Moreover, it revealed superimposed bedforms on the dunes, including draping and oscillating patterns, reflecting the influence of consistent wind regimes during dune development and migration. This study proposes an evolutionary sequence in which proto-barchans evolve into barchanoid ridges and eventually as mature isolated barchans. These findings provide insight into the geomorphological evolution of barchanoid dunes in a coastal context, and particularly their migration pattern on a coastal spit system.

This study evaluates neoichnological characteristics of the Gironde Estuary and Arcachon Bay to test ichnological proxies of salinity conditions along a freshwater to marine gradient. Both the Gironde Estuary and Arcachon Bay are classified as mixed-influence (wave-tide) estuaries, but tidal currents are stronger and turbidity is higher in the Gironde Estuary funnel than in Arcachon Bay. Trace distributions and the size-diversity index (SDI) are determined at various localities in both systems and are then compared to trends established in prior studies. Results indicate that SDI provides a robust, quantitative correlation with salinity that is comparable to geochemical and biological proxies. Trace (ichno) diversity alone reflects salinity qualitatively but is limited by sampling and human (interpretation) biases and data variability. These findings contribute to better understanding the predictive value and limitations of SDI in modern estuaries. However, the application of SDI in the sedimentary record requires additional research as neither infaunal diversity nor vermiform burrows can be established reliably therein.

The analysis of carbonate rocks through petrographic methods has long posed significant challenges in geological sciences, particularly regarding the systematic description and quantification of thin sections. Recent developments in artificial intelligence have suggested promising avenues for automation; however, the field remains constrained by limited training data availability, as manual annotation requires considerable time investment and can result in variable interpretations. This study addresses these challenges by introducing SynSection, a method for generating synthetic pairs of carbonate thin section images with corresponding labels for both classification and segmentation tasks. This approach has been carefully designed to incorporate sedimentological heuristics, ensuring the geological validity of generated samples through the integration of three primary components: 2D grain packing, groundmass generation using texture synthesis, and image blending for image composition. The advantage of this methodology lies in its ability to exponentially expand a limited set of manually labeled images, thereby reducing the resources required for dataset creation while maintaining consistency in annotation. The method was evaluated through the generation of 100,000 synthetic image-annotation pairs, implementing a transfer learning strategy in which models were initially trained on synthetic data before being fine-tuned on real thin section images. This approach demonstrated significant improvements in both classification and segmentation tasks compared to conventional training methods that rely solely on limited real data. In classification tasks, the transfer learning strategy enhanced performance metrics by 7.2%, achieving a top accuracy score of 94.9%. The impact of the method was particularly notable in segmentation tasks, where the binary grain segmentation model achieved an Intersection over Union (IoU) of 77.2% and a point counting determination score of 93.2%. The multiclass segmentation achieved a mean Intersection over Union (mIoU) of 55.3% and a point counting score of 89.6%, reflecting the inherent complexity of grain type differentiation. These results, which are consistent with medical imaging applications, demonstrate the method's potential to optimize carbonate thin section analysis by reducing time allocated to descriptive tasks. The approach generates diverse training data that maintains geological validity, establishing a quantitative framework for automated petrographic analysis that aligns with standardized classification criteria and expert-defined features.

Summary Semantic segmentation is a discipline in computer vision that aims to identify structures of interest in an image by assigning a category to each pixel. This method, which closely aligns with petrographic description, holds significant value in accelerating a process often deemed lengthy and tedious. However, the substantial amount of data to label in machine learning methods remains the primary obstacle to achieve an acceptable generalization for a given problem. Such datasets are currently not well-known in this specific field. To address this issue, this study presents a method to segment thin section images using deep learning and based on a synthetic training dataset. The generation of these compositions is based on a sampling strategy of real objects within an initial observation scene and packing them in various new locations on a textured background. This aims to produce more robust models during inference. The proposed approach achieved an intersection over union (IoU) of 0.75 on a base of 20 real examples. These results can thus highlight the future of deep learning in the field of carbonate petrography for applications in different geological formations and reservoirs, even with a limited amount of labeled data.

Download This Paper Open PDF in Browser Add Paper to My Library Share: Permalink Using these links will ensure access to this page indefinitely Copy URL Copy DOI

Abstract. The surface of intertidal estuarine sediments is typically covered with a photosynthetic biofilm. A large fraction of the carbon that is fixed is in the form of exopolymeric substances (EPS), providing the biofilm matrix. The consumption of organic carbon within the sediment column by heterotrophs bacteria is stratified according to the availability of electron acceptors used for organic matter degradation. This sequential use of electron acceptors strongly impacts geochemical gradients and early diagenetic processes within the sediment. In most studies, the distribution and role of the predominant microbial metabolisms is deduced from porewater chemistry and restricted to the upper decimeters of the sediment column, but rarely from direct measurements of microbial activity, potentially leading to erroneous conclusions of biogeochemical processes. We measured geochemical gradients in three estuarine sediment cores to a depth of 6 meters. Geochemical analyses of porewater and sediment were combined with measurements of microbial activity. In situ microelectrode measurements were performed for pH, oxygen and sulfide. Porewater was extracted and analyzed for major elements using Ion Chromatography, Inductively-Coupled-Plasma, and colorimetric assays for iron speciation. Porewater chemistry was compared to measurements of microbial activity including isothermal calorimetry and metabolic assays (triphenyltetrazolium chloride (TTC) and fluorescein diacetate (FDA)) and concentrations of EPS (sugars, proteins) measured in a previous study on the same cores. Finally, sediment composition was characterized through X-Ray Fluorescence core scanning. Results show that: (i) aerobic respiration occurred between 0 and 1 cm, (ii) nitrate reduction between 6 and 16 cm, (iii) sulfate reduction between 10 and 50 cm, (iv) manganese oxide reduction between 2–6 and 35–50 cm and (v) iron oxide reduction between 16–18, 24–26 and 35–45 cm. This is concomitant with the area where the microbial activity is the highest. In contrast to the literature, we conclude that some reactions, for example sulfate and nitrate reduction, were locally coupled or at least occurred concomitantly. Impacts of microbial metabolism on early diagenesis have been modeled via PhreeQc and predicted potential precipitation of metastable iron and/or sulfides. This is confirmed by iron and sulfur increases in sediments characterized through XRF. All these observations have been used to propose a biogeochemical model linking microbial metabolisms and early diagenesis that can be used as a basis for the study of other geochemical profiles in the future.

<p>The construction and interpretation of digital outcrop models (DOM) from outcropping reservoir analogues enable to capture and characterize reservoir heterogeneities (in terms of facies, diagenesis, and petrophysical properties) from centimeter to kilometer scales, thus allowing to improve upscaling approaches in 3D reservoir modeling. Digital outcrop models offer a reliable tridimensional representation of sedimentary heterogeneities, which can strongly impact fluid flow and therefore geothermal reservoir exploitation.</p> <p> </p> <p>The Roda Sandstones (Lower Eocene) are considered as a world-class outcropping example for deltaic sedimentary systems outcropping in the Graus-Tremp Basin (South Pyrenean Basin). Thanks to the quality of its outcrops and to the drilling of 50 to 80-meter-long cores in their vicinity, the Roda Sandstones are commonly used for educational and research purposes (<em>Crumeyrolle et al., 1992; Martinius, 2012</em>).</p> <p> </p> <p>Over the past 15 years, a few digital outcrop models have been published from the Roda Sandstones (<em>Enge et al., 2007; Leren et al., 2010)</em>. These models were only constructed at a small scale (decimeter to hectometer) and did not allow to capture the large-scale architecture of the Roda fluvio-deltaic system. In this study, we built a complete photogrammetric model of one of the prograding sand wedges of the Roda Sandstones (also called Y body) from more than 11000 photos acquired by drone. The model is accurately georeferenced thanks to a dGPS campaign carried out simultaneously with the drone acquisitions. This outcrop model covers a total area of about 4km², and the pixel resolution ranges between 3 mm and 3 cm.</p> <p> </p> <p>A significant amount of quantitative and qualitative information could be extracted from this digital outcrop model, that helps at constraining the reservoir model. Its interpretation in a software dedicated to the geological interpretation of DOMs enabled to take measurements (e.g., dips, distances, etc.), to identify and to trace the main stratigraphic surfaces, locate the field observations and samples, allowing to precisely assess the architecture and the facies distribution of the Y sandbody.</p> <p> </p> <p>The results show a multiphase sandbody, made up of different prograding lobes, with variable progradation directions and a diversity of sedimentary structures formed by the competition between fluvial and tidal currents, contributing to the complexity of the sedimentary system. Paleocurrent directions, sediment thicknesses, numerical outcrops painted in facies, digitized sedimentological sections, and boreholes interpreted in facies were used as input data to build a static facies model. The geological static model was then filled with porosity and permeability properties and used as a base for fluid flow simulations in order to assess the impact of sedimentary heterogeneities in deltaic reservoirs for geothermal exploitation purposes.</p>

Chlorite is recognized worldwide as a key mineral that inhibits the development of quartz cement in deeply buried sandstone reservoirs. Iron-rich chlorite is mainly formed by the transformation of a precursor clay mineral; however, few studies have focused on the early stages before the crystallization of chlorite. This study analyzed shallowly buried (400-1000 m) coastal sandstones from within the Wealden Group of the Paris Basin. Berthierine, a 7 A trioctahedral clay mineral belonging to the serpentine group, approximatively with same chemistry as chlorite but a different crystal structure, has been identified in a 900-m deep well but not in a 400-600-m-deep well. Berthierine has mainly been observed as clay coatings around detrital grains with a typical honeycomb texture. Nanopetrographic observations suggest that the honeycomb textural organization of the clay particles found in deeper buried sandstone reservoirs (>1500 m) is acquired from a berthierine precursor at shallow depths. However, small amounts of quartz overgrowths are observed on the surface of detrital grains at shallow depths and low temperature (below 40 degrees C), and it is believed that precursor berthierine coatings are primarily responsible for the inhibition of quartz overgrowths before Fe-rich chlorite is formed. This suggests that the key mineral primarily controlling the reservoir quality of deeply buried sandstone reservoirs is berthierine rather than iron-rich chlorite, which challenges the commonly accepted assertion that chlorite coating is the main process that inhibits quartz overgrowths. The source-to-sink context of the Paris Basin during the Early Cretaceous was decisive with respect to the supply of sands and berthierine clay precursors (in particular kaolinite and iron-rich, hydroxy-interlayered clay minerals) to the center of the basin.

Summary Diagenetic modification of carbonate depositional systems is a dominant process changing their pore systems away from primary texture and responsible for their challenging multi-modal and multi-scale behavior. It is these pore system characteristics that control dynamic behavior across many scales from plug – to log – to reservoir scale. One common diagenetic product in many Middle East reservoirs is dolomite and is invoked to be associated with improved storage and excess permeability. Despite these observations, reliable spatial models of dolomite distribution are rare, especially at field or seismic scale. This paper documents how the dolomite distribution across an outcrop in Morocco was captured and validated using high resolution 3D photogrammetry combined with hyperspectral acquisition. It suggested that these, “remote” attributes can be combined and not only provide spatial rules but also point to scenarios for reconstruction of timing and process of dolomitization.

The surface of intertidal estuarine sediments is covered with diatom biofilms excreting exopolymeric substances (EPSs) through photosynthesis. These EPSs are highly reactive and increase sediment cohesiveness notably through organo-mineral interactions. In most sedimentary environments, EPSs are partly to fully degraded by heterotrophic bacteria in the uppermost millimeters of the sediment and so they are thought to be virtually absent deeper in the sedimentary column. Here, we present the first evidence of the preservation of EPSs and EPS-mineral aggregates in a 6-m-long sedimentary core obtained from an estuarine point bar in the Gironde Estuary. EPSs were extracted from 18 depth intervals along the core, and their physicochemical properties were characterized by (i) wet chemical assays to measure the concentrations of polysaccharides and proteins, and EPS deprotonation of functional groups, (ii) acid-base titrations, and (iii) Fourier transform infrared spectroscopy. EPS-sediment complexes were also imaged using cryo-scanning electron microscopy. EPS results were analyzed in the context of sediment properties including facies, grain size, and total organic carbon, and of metabolic and enzymatic activities. Our results showed a predictable decrease in EPS concentrations (proteins and polysaccharides) and reactivity from the surface biofilm to a depth of 0.5 m, possibly linked to heterotrophic degradation. Concentrations remained relatively low down to ca. 4.3 m deep. Surprisingly, at that depth EPSs abundance was comparable to the surface and showed a downward decrease to 6.08 m. cryo-scanning electron microscopy (Cryo-SEM) showed that the EPS complexes with sediment were abundant at all studied depth and potentially protected EPSs from degradation. EPS composition did not change substantially from the surface to the bottom of the core. EPS concentrations and acidity were anti-correlated with metabolic activity, but showed no statistical correlation with grain size, TOC, depth or enzymatic activity. Maximum EPS concentrations were found at the top of tide-dominated sedimentary sequences, and very low concentrations were found in river flood-dominated sedimentary sequences. Based on this observation, we propose a scenario where biofilm development and EPS production are maximal when (i) the point bar and the intertidal areas were the most extensive, i.e., tide-dominated sequences and (ii) the tide-dominated deposit were succeeded by rapid burial beneath sediments, potentially decreasing the probability of encounter between bacterial cells and EPSs.

Logging data are measurements of physical properties of the formation surrounding a borehole, acquired in situ after completion of coring (wireline logging) or during drilling (Logging-While-Drilling, LWD). The range of data (resistivity, gamma radiation, velocity, density, borehole images,…) in any hole depends on the scientific objectives and operational constraints.

Summary The limited accessibility to some parts of large-scale outcrops (e.g., cliffs), remains a challenge for the full assessment of the architecture of world-class sedimentary systems using conventional field geology. The development of drone technologies in the last decades, followed by the improvement of image processing, photogrammetry, and interpretation software, allows to accurately determine the 3D architecture of seismic scale outcrops. We applied a workflow allowing to accurately (meter scale) define the internal architecture of a Jurassic ooid-dominated carbonate ramp exposed on the Amellago cliffs (High Atlas, Morocco). By integrating multiple datasets (fieldwork, literature) with a high resolution, 15km long UAV-based photogrammetric model interpreted with VRGS, we refined the pre-existing sequence stratigraphic scheme with a precise (decimetre scale) geometry, thickness, and facies variations along the proximal-distal profile of this oolitic carbonate ramp. This new step in the technology and interpretation of seismic scale outcrops will improve our understanding of general sedimentary models and their reservoir analogs.

ABSTRACT Estuaries provide an excellent depositional environment to study the interaction between minerals (clays, quartz sands…) and biofilms. The estuary bottom is largely covered by biofilms that impact sediment stability, the mud and clay-coat content in sands, and sedimentary-structure stability, thus influencing sandstone properties during burial. Although numerous oil, gas, and geothermal reservoirs are exploited in estuarine heterolithic point bars, many questions remain about the origin of reservoir properties and heterogeneities in these sedimentary bodies. In order to better understand the sedimentary and microbiologic processes in estuarine systems and to better predict the reservoir quality of estuarine sandstones, this study characterizes a modern heterolithic point bar located in the Garonne estuarine channel at various scales, ranging from the microscopic (thin section) to the macroscopic (core) scale. Three piston cores 4.5 to 6.8 m long were drilled in the Bordeaux North Point Bar. Three main facies were identified in these cores: 1) sandy gravel, 2) heterolithic, medium-grained sand dunes, and 3) thin heterolithic, fine-grained sand beds with mud drapes. The sands are classified as lithic arkoses to feldspathic litharenites. Detrital clay grain coats, which at deep burial depths are transformed to permeability preserving authigenic chlorite coatings, are observed from the base to the top of the point bar. These detrital clay grain coats are mainly composed of smectite, illite, kaolinite, and chlorite, intermixed with other components, such as diatoms or pyrite. Biofilms of exopolymeric substances (EPSs), mostly produced by diatoms, are believed to control the adhesion of the clay coats to the surface of sand grains. Quantification by thin section shows that on average about 30% of the sands are coated in the point bar. The proportion of clay-coated grains appears to be independent of facies. Radiocarbon age dating measured on organic matter points to significant vertical mixing, highlighting the significance of erosion and redeposition. The activities of 137Cs and 210Pb indicate a vertical sedimentation rate of ca. 0.02 m.yr–1 in the muddy chute channel. These ages, coupled to historical maps, suggest that the present-day point bar has developed over the last 300 years with a vertical sedimentation rate ranging from 0.015 to 0.036 m.y–1 and a lateral migration rate of about 1 m.y–1. The combination of sedimentary geology, thin-section petrography, and mineralogy at high spatial and temporal resolutions highlights the potential of this study area as a modern analogue for ancient tidally influenced point-bar deposits associated with clay coatings.

The Messinian microbialites of the Terminal Carbonate Complex (TCC) from the Neogene basins of southeastern Spain show both diversified morphologies and an excellent preservation of primary microbial microstructures. Their stratigraphic architecture, fabric (micro-, meso-, and macro-fabric), and mineralogical composition were investigated in eight localities from three sedimentary basins of southeastern Spain: The Sorbas and Bajo Segura basins and the Agua Amarga depression. Two recurrent microbialite associations were distinguished. Laterally linked low relief stromatolites predominated in Microbialite Association 1 (MA1), which probably formed in low energy lagoons or lakes with fluctuating normal marine to hypersaline water. The microfabrics of MA1 reflected the predominance of microbially induced/influenced precipitation of carbonates and locally (Ca)-Mg-Al silicates. Microbialite Association 2 (MA2) developed in high energy wave and tidal influenced foreshore to shoreface, in normal marine to hypersaline water. High-relief buildups surrounded by mobile sediment (e.g., ooids or pellets) dominated in this environment. MA2 microbialites showed a significant proportion of thrombolitic mesofabric. Grain-rich microfabrics indicated that trapping and binding played a significant role in their accretion, together with microbially induced/influenced carbonate precipitation. The stratigraphic distribution of MA1 and MA2 was strongly influenced by water level changes, the morphology and nature of the substratum, and exposure to waves. MA1 favorably developed in protected areas during third to fourth order early transgression and regression phases. MA2 mostly formed during the late transgressions and early regressions in high energy coastal areas, often corresponding to fossil coral reefs. Platform scale syn-sedimentary gypsum deformation and dissolution enhanced microbial carbonate production, microbialites being thicker and more extended in zones of maximum deformation/dissolution. Microbial microstructures (e.g., microbial peloids) and microfossils were preserved in the microbialites. Dolomite microspheres and filaments showed many morphological similarities with some of the cyanobacteria observed in modern open marine and hypersaline microbialites. Dolomite potentially replaced a metastable carbonate phase during early diagenesis, possibly in close relationship with extracellular polymeric substances (EPS) degradation. Double-layered microspheres locally showed an inner coating made of (Ca)-Mg-Al silicates and carbonates. This mineral coating could have formed around coccoid cyanobacteria and indicated an elevated pH in the upper part of the microbial mats and a potential dissolution of diatoms as a source of silica. Massive primary dolomite production in TCC microbialites may have resulted from enhanced sulfate reduction possibly linked to the dissolving gypsum that would have provided large amounts of sulfate-rich brines to microbial mats. Our results open new perspectives for the interpretation of ancient microbialites associated with major evaporite deposits, from microbe to carbonate platform scales.