Dolomite genesis in bioturbated marine zones of an early-middle Miocene coastal mud volcano outcrop (Kuwait)

The origin of spheroidal dolomitized burrow from Al-Subiya sabkha in Kuwait was previously described as enigmatic as no evidence of precursor calcium carbonate was found in the siliciclastic sediment. An assumption for the genesis of spheroidal dolomite from the same area was attributed to hydrocarbon seepage but no evidence was provided. Here, we investigated a recently discovered early-middle Miocene coastal mud volcano outcrop in Al-Subiya sabkha where dolomitized burrows and spheroidal dolomite are found in bioturbated marine zones, and associated with traces of salt. Conversely, the continental zone lacks bioturbation features, dolomite and traces of salt, which together contrast with bioturbated rich marine zones. Geochemical signatures of Rare Earth Elements + Yttrium show a true positive Ce anomaly (Ce/Ce* > 1.2) and positive Eu/Eu* anomaly of spheroidal dolomite indicating strictly anoxic conditions, and sulphate reduction to sulphide, respectively. Our results are suggestive of a relationship between dolomite formation and interdependent events of hydrocarbon seepage, flux of hypersaline seawater, bioturbation, and fluid flow in the marine zones of the mud volcano. The bioturbation activity of crustaceans introduced channels/burrows in the sediment–water interface allowing for the mixing of seeped pressurized hydrocarbon-charged fluids, and evaporitic seawater. In the irrigated channels/burrows, the seeped pressurized hydrocarbon-charged fluids were oxidized via microbial consortia of methanotrophic archaea and sulphate-reducing bacteria resulting in elevated alkalinity and saturation index with respect to dolomite, thus providing the preferential geochemical microenvironment for dolomite precipitation in the bioturbated sediment.


Site description and sampling
The study area is located in the northeast arch of Kuwait bay within the post Eocene Kuwait Formation, a representative strata of the traditional tripartite subdivision of Ghar, Lower Fars, and Dibdibba formations 39,40 . The subsurface formation of this area was recently reinterpreted in relation to prolonged tectonic compression and strike-slip tectonics. Thin-skinned tectonics by bedding parallel slippage of the Dammam and Rus formations from the south of Kuwait caused northward transgression of the Jal-Azour escarpment and opened up splay faults for subsurface fluids and gases to escape 41 . Within the basal part of Kuwait Formation, an early-middle Miocene mud volcano outcrop was the focus of our study. The current exposures of mud volcano features after seawater regression demonstrate pseudo-bioherm formations that display cratered elliptical-shapes indicative of seismicity and the consequent ascending fluids from subsurface plumbing systems 27 . Geometries of chimneylike structures were described as evaporitic mud pipes and canonical mounds that were formed due to extensive gas venting 27 . The burrows are restricted to the marine zones while the pockmarks are ubiquitous throughout the complex. Rocks were sampled from three different zones, two marine and one continental, separated by a remnant tidal channel. The coordinates for each sampling locations are (29°38′07.1"N 47°59′43.4"E) for site 1, (29°36′56.2"N 48°00′57.6"E) for site 2, and (29°36′07.0"N 48°01′53.7"E) for site 3. The rocks are labelled as MV1, MV2, and MV3 corresponding to zones 1, 2, and 3, respectively. Figure 1 shows the location of the mud volcano outcrop and the sampling sites. An overview of the features of the investigated area and the sampled rocks is displayed in Fig. 2. Figure 3 shows burrows from the marine zones and these burrows are comparable  www.nature.com/scientificreports/ in size, shape, and mineralogy to published data on the same site 25,28 . The detailed description of the biota and the associated ichnofabric has been previously described 28 .

Methods and analytical techniques
Powder X-ray diffraction (XRD). The mineralogy of the burrows and rocks was determined using an X-ray diffractometer. The cores of the burrows were scooped out using a chisel and hammer and placed in sterile bags. MV1, MV2, and MV3 rocks were sampled and a representative fragment of each rock was cut using a wire diamond saw. All samples were pulverized into a fine powder using pestle and mortar pre-washed with isopropanol. The powdered samples were then loaded into the sample holder and analyzed on D8 Advance XRD (BRUKER, USA) for the burrows and a PHILIPS Analytical X-ray B.V. (PHILIPS, Netherlands) for the rocks.
The detailed scanning settings were as follows: scan type Gonio with continuous mode, 2 θ scanning starting from 20° to 60° with Cu-Kα radiation, and step size of 0.02° with scan speed of 0.020°/s. X'Pert Quantify software was used for scanning, HighScorePlus software was used for peak analysis, and comparisons were made with the database of the International Center for Diffraction Data (ICDD). Spectra were plotted using Origin 2018 software (OriginLab Corporation, MA, USA).

Scanning electron microscopy with energy dispersive spectroscopy (SEM-EDS). SEM-EDS
technique characterizes the morphology and texture of rock grains in sub-micrometer spatial resolution with semi-quantitative elemental microanalysis. Small grains from the rocks were transferred on a double carbon tape on an aluminium stub and coated with 10 nm of osmium tetroxide (OsO 4 ) (Filgen, Japan) before imaging. Samples were examined using a Hitachi S530 SEM (HITACHI, Japan) with emission current adjusted to 60 µA and voltage to 20 kV. Quartz PCI version 8 image management system (QUARTZ, Canada) was used for image acquisition. Elemental analysis of rock grains by EDS was performed using a JEOL JSM-6610LV SEM (JEOL, Japan) and equipped with the X-Max EDS system (Oxford Instruments, High Wycombe, UK). INCA software (Oxford Instruments, High Wycombe, UK) was used for data acquisition and processing.
Cathodoluminescence microscopy (SEM-CL). SEM-CL was used to inspect dolomite concentric radial fabric from thin sections of the rocks. A JEOL JSM-6610LV SEM (JEOL, Japan) was used with a cathodoluminescence detector (Gatan-MiniCL). Thin sections were prepared from the Epoxy-embedded samples used for EPMA. The embedded samples were glued on frosted petrographic glass slides. The bulk portion of the sample was cut by a diamond saw (IsoMet 4000) to obtain a thin layer on the slide. Grinding of the samples was processed using a PetroThin Thin Section Machine to attain a flat surface on the slide. Slides were then placed in a lapping and polishing machine (LOGITECH, LP30) using silicon carbide of 600 grit size (15 µm grain size) in order to obtain a sample thickness of 30 µm. An automated polishing system with an oil-based diamond suspension was applied in 3 steps: at 9 µm, 3 µm, and 1 µm consecutively, each step consisting of 5 min intervals (MetaServ 250 Grinding/Polishing-Buehler) at a speed of 200 rpm.
Laser ablation inductively coupled plasma mass spectrometry (LA-ICPMS). The concentration of the Rare Earth Elements, Yttrium (Y), Uranium (U), and Thorium (Th) were determined in 10 sub-samples of the rocks. Samples were prepared following the same protocol for EPMA but with two additional polishing steps of 0.3 mm Alumina and 0.06 mm colloidal silica using a polishing cloth. A pre-quantification of major elements (Ca, Mg, and Si) was initially performed by EPMA in order to correct LA-ICPMS primary results. LA-ICPMS analysis was performed using UP-213 laser ablation system coupled with VG PQ ExCell inductively coupled plasma mass spectrometry (ICPMS) system. Helium was used to transfer the ablated samples to ICPMS system and Argon was used to cool the torch. The settings of the laser were as follows: spot size ranged from 55 to 100 mm, scan speed was 20 mm/s, repetition rate was 10 Hz, and the laser percentage output power was 65%. The energy of the laser beam ranged from 0.10 to 0.40 mJ and the energy fluence rate was 4.19 J/cm 2 to 6.51 J/ cm 2 . The machine was tuned for maximum sensitivity and low oxide production to avoid the interference of www.nature.com/scientificreports/ oxides with the results. The oxides were standardized to be less than 2% by monitoring Th 232 /ThO +248 to be less than 2% and U 238 /Th 232 ~ 1%. Before starting the analysis, standardization of the values was performed based on the standard NIST610 synthetic glass for correcting data drift. Two lines were ablated on the standard glass before analysis and two at the end of the analysis. The values of the measured elements were corrected after EPMA quantification and normalized to shale average of Post Australian Archean Shale (PAAS) 42 . GraphPad Prism software version 8 (GraphPad, CA, USA) was used for graphical representations.

Results
Mineralogical composition of authigenic carbonates. The mineralogic contents of the burrows, MV1, and MV2 are mainly calcite, halite, and dolomite while MV3 is dominated by calcite with the absence of halite and dolomite. The burrows from marine sites adjacent to MV1 and MV2 samples have a similar XRD pattern ( Fig. 3) in comparison to MV1 and MV2 rocks ( Fig. 4) but the peaks display varying intensities. Dolomite The details of mineral abundance in the rocks are analysed from EPMA data ( Table 1) by calculating the normalized area to percentage from the compositional mineral phase maps. Figure     In MV1, microcrystalline dolomite forms a mesh-like fabric in the calcite matrix ( Fig. 5a) while dolomite in MV2 exhibits replacive cement of the pre-existing calcite matrix with clear delineation (Fig. 6a). Dolomitized fossils in MV2 are distinctive from Mg mapping (Fig. 6c) and their structures are recognizable in Na and Cl mapping (Fig. 6d,e). Interestingly, the mapping of Mg, Na, and Cl in MV2 are clearly superimposed. In MV3, the intensity of Mg (Fig. 7c) is very low in comparison with MV1 and MV2.
Single point analyses with EDS in MV1 (Fig. 8) show an example of microcrystalline dolomite within calcite, large spheroidal dolomite replacing calcite in MV2 (Fig. 9), and dolomite absence in MV3 (Fig. 10). Dolomite stoichiometry in MV1 and MV2 were determined from single point analyses where Ca, Mg, Mn, and Fe of dolomite were quantified and compared to standard dolomite ( Table 2). All analysed single points of dolomite show minor calcium enrichment in agreement with the XRD results ( Fig. 4) with traces of Mn and Fe.
The electron photomicrograph images with EDS analysis show planar texture of euhedral-subhedral microcrystal dolomite rhombs with size ranging from 2 to 5 μm. Dolomite rhombs are generally characterized by well-developed crystal faces and sharp boundaries; some rhombs demonstrate less distinctive characteristics. The growth of dolomite at the expense of calcite is noted from dolomite crystallization on a partially dissolved calcite (Fig. 11).
Petrographic characterization. The fabric of spheroidal dolomite obtained via cathodoluminescence microscopy reveals radial concentric fabric with alternating layers of bright and dark luminescence (Fig. 12). Dolomite radial zonation shows visible discontinuities making each zone distinctive from the adjacent zone.
Concentration of rare earth elements and trace elements. The concentrations of Rare Earth Elements and Yttrium (REY), Th, and U are listed in Table 3 with anomalies of PAAS-normalized (Ce/Ce*)sn, (Eu/  Figure 13 shows the reducing conditions based on the results of (Ce/Ce*)sn and (Pr/Pr*) sn 45 . The PAAS-normalized REY patterns have shown different trends among the investigated rocks. Figure 14a denotes the results of three sub-samples from the MV1 rock with no discrepancy between dolomite and calcite. All sub-samples exhibit (Ce/Ce*)sn in the range between 0.8 to 1.  www.nature.com/scientificreports/ anaerobic oxidation of methane resulted in increasing alkalinity; consequently precipitating dolomite 55 . Cross sections of these dolomitized cylindrical concretions have similar features to the spheroidal dolomite found within MV2 sample in this study suggesting that the hollow cores in MV2 are probably the channels of methane emissions (Fig. 9a). Furthermore, spheroidal dolomite in MV2 has similar structural features of spheroids of isopachous chains found in calcite spar from a Miocene methane seep in Italy, which were assumed to be   www.nature.com/scientificreports/ anaerobic microbial consortia responsible for methane oxidation 56 . Although important, hydrocarbon seepage does not explain the absence of dolomite in the continental zone of our studied mud volcano outcrop. An important factor for dolomite formation in hydrocarbon seepage environments is the fluid of hypersaline seawater. Dolomite formed in areas of marine hydrocarbon seepage where sub-surface methane-charged fluids penetrated upwards thereby mixing with the shallow seawater in the anaerobic sediment where methane was oxidized by sulphate 53 . This finding may substantiate the preferential dolomitization of authigenic carbonates in the marine zones of our study area which is related to the evaporitic sabkha. The evaporated seawater in the sabkha contributes to high Mg/Ca ratio 17 and is known to trigger dolomite formation 57 . Our results show clear evidence that the dolomite-comprising burrows and bioturbated MV1 and MV2 samples are from marine zones which is indicated by halite contents in XRD analyses (Figs. 3, 4) and foraminifera (Fig. 6e). Also, mineral phase abundance obtained from EPMA (Table 1) for the rocks demonstrate a close relationship between dolomite and NaCl. MV1 and MV2 constitute a considerable amount of dolomite and NaCl, while MV3 lacks dolomite and NaCl. Furthermore, the precipitation of megacrystic gypsum within MV1 zone is a reliable sign of high evaporation 58 . The microcrystalline dolomite in MV1 exhibits a mesh-like fabric represented by Mg mapping (Fig. 5c) possibly indicating a dissolution of susceptible dolomite 41 . Interestingly, a nanoscale in-situ observation using atomic force microscopy (AFM) has shown that the precipitation of gypsum is coupled with dolomite dissolution 59 . The absence of fossils in MV1 can be related to dolomitization in the upper tidal flat environment 60 . MV1 is possibly a re-worked carbonate material from the vents during sediment extrusion thereby abrading preexisting earlier dolomite of spheroidal nature. In MV2, the superimposed distribution of Mg, Na, and Cl (Fig. 6c, d, e, respectively) suggest an interdependent relationship between dolomitization and hypersalinity. Indeed, high   61 . Thus, the co-existence of high salt concentration, marine fossils, and dolomite is an indicator for the involvement of hypersaline seawater in dolomitization. On the other hand, the absence of NaCl has been found in the continental MV3 sample together with the low intensity of Mg in MV3 (Fig. 7c).
The impact of bioturbation on dolomite formation and diagenesis. Bioturbation refers to displacement within sediments introduced by organisms while ichnofabric is the recorded texture and structure in the sediment from the bioturbation activity 62 . Dolomite in this study was found in the bioturbated sediment and the burrows. Commonly, dolomite in burrows is associated with reducing conditions 63 , sulphate-reducing   The bioturbation structures in authigenic carbonates of the investigated mud volcano in this study provides evidence that the marine zones are the host of the dolomitized crustacean burrows whose origins were previously described as perplexing 25 . Remarkably, dolomite stoichiometry in our samples (Table 2) is almost identical to the composition (Ca 52 Mg 48 and Ca 57 Mg 43 ) of selectively dolomitized crustacean burrows described by Gunatilaka et al. (1987) 25 .
The selectively dolomitized burrows result from the effect of crustaceans on modifying the physical and chemical properties of the sediment where burrows act as a conduit for the sediment-water interface 68 . These conduits provide channels for the flow of fluids from sub-surface hydrocarbon reservoirs 32 and supply Mg 2+ Table 3. Concentrations (ppm) of rare earth elements, Yttrium, Thorium and Uranium in the sub-samples of carbonate rocks from Miocene-age mud volcano outcrop. Y is inserted between Dy and Ho due to the chemical similarity 97 . Anomalies are normalized to PAAS 42 .  www.nature.com/scientificreports/ to the burrows from interstitial marine sources 68 . In the microenvironment of the burrows, dolomite is generally prompted by a consortium of methane-oxidizing archaea and sulphate-reducing bacteria that mediate the sulphate-driven anaerobic oxidation of methane (SD-AOM), and produce bicarbonate and dissolved sulphide 36 . Moreover, sulphate reduction in the burrows can release Mg 2+ ions from neutral ion pairs which raises the concentration of Mg 2+ for the dolomite-filled burrows 63 . The patchy and mosaic dolomite in MV1 and MV2, respectively, created diagenetic textural heterogeneities depicted by comparing the mapping of dolomite (Figs. 5a, 6a) with the corresponding backscattered images (Figs. 5i, 6i). The textural heterogeneities of the calcite matrix contemporaneous to dolomitization is an indication of burrow-associated dolomite 69 . Additionally, the dolomitized fossils of fragmental shells and probably foraminifera (Fig. 6e) is direct evidence of diagenesis 70 where the microenvironment of burrows provides the favourable geochemical conditions for dolomitization 68 , and microfossils seem to act as a loci for dolomitization.  www.nature.com/scientificreports/ The preserved annular radial pattern in MV3 rock (Fig. 2f) sampled from the continental zone implies no diagenetic features and exhibits homogeneous texture (Figs. 7h, i). Not surprisingly then, that the diagenetic fabrics are indicative for dolomitizing fluids that seeped through burrow networks 68 , which explains the genesis of dolomite in the marine mud vents of Al-Subiya where prolific Ophiomorpha and Thalassinoides were described 28 . The preserved radial concentric Cathodoluminescence (CL) zonation of dolomite in MV2 (Fig. 12) can be recognized in textural association with multiple episodes of meteoric diagenesis related historically to transgression/regression 71 . Furthermore, dolomite formation in continental evaporitic environment can have a dual source of mixed meteoric/seawater such as in coastal lakes and lagoons of the Coorong region of south Australia 72 . The dull and luminescent zonation is controlled by varying concentrations of Mn and Fe 73 which describe the geochemical changes throughout dolomite diagenesis 74 . An unconfined aquifer would be oxygenated via vadose zone recharge resulting in Fe and Mn oxidation and dull luminescence. In contrast, the luminescence becomes bright when no recharge from vadose zone reaches the confined aquifer 71 . Nevertheless, we believe that the oxygenation of the aquifer in our case cannot be solely related to vadose recharge as methane seepage is part of the studied environmental setting. As such, methane pulsive seepage can also affect the pattern of CL zonation by changing the redox state of the environment and consequently the availability of Mn and Fe 75 .
Geochemical signatures of authigenic carbonates. The distinct distribution of PAAS-normalized seawater Rare Earth Elements and Yttrium (REY) can be authentically preserved in carbonate sediment and rocks 76 . This section provides evidence of the geochemical conditions governing the process of carbonate formation in the studied mud volcano. Rare Earth Elements and Yttrium (REY) have been considered reliable proxies for understanding the geochemistry of marine environments 77,78 .
A fundamental geochemical signature of the reducing environment is Ce anomaly; a proxy used for constructing oxic, suboxic, and anoxic conditions of ancient marine environments 79,80 . Under oxidizing conditions, Ce 3+ is oxidized to less soluble Ce 4+ scavenged by suspended particles and settling through the water column 77 . Therefore, the carbonates that have precipitated in equilibrium with seawater commonly demonstrate removed Ce and thus negative Ce anomalies 81 . The reducing environmental conditions were determined from the plot of Ce and La anomalies calculated from Pr/Pr* = Pr sn /(0.5Ce sn + 0.5Nd sn ) and Ce/Ce* = Ce sn /(0.5Pr sn + 0.5La sn ) 82 (Fig. 13). Moreover, a positive Ce anomaly can be due to intensive alkaline water conditions 83 since highly alkaline water can transform insoluble Ce to soluble complexes thereby resulting in positive Ce anomaly in carbonate sediments 84 . The implications of anoxic conditions from Ce anomalies may indicate a presence of methane related to hydrocarbon seepage 85 where the pockmarks found in the mud volcano outcrop (Fig. 2i) are indicative of methane bubbles 86 . Indeed, the formation of authigenic carbonates including dolomite related to methane seepage under anoxic conditions is well-documented in the literature 34,53,68 .
The patterns of REY trends in MV1 sub-samples (Fig. 14a) are inconsistent with the presence of inexplicable spikes; possibly due to the alteration of carbonate rock composition as a result of diagenesis 74 , admixture of dust and detrital organic matter 87 , or terrigenous contamination 88 . These patterns of MV1 sub-samples are referred to as the heterogenous matrix with no discrepancy between dolomite and calcite. The patterns of dolomite and calcite sub-samples in MV2 (Fig. 14b) show strong Eu anomalies which may denote the input of hydrothermal fluids to seawater 82,89 . The implication of the positive Eu anomalies is supported by the finding of upward movement of hydrothermal fluids observed nearby our study site at Bahrah field which is linked to seal breach in subsurface carbonate petroleum systems 90 . Additionally, the positive Eu anomaly indicates sulphate reduction to sulphide 91 which is compatible to the metabolic pathway of sulphate-reducing bacteria contributing to anaerobic methane oxidation 53 . Interestingly, the patterns of MV2D.1 and MV2D.2 of dolomite are identical to that of Congo Fan seep carbonate that formed under sulphidic conditions and significant anaerobic oxidation of methane 50  An additional important proxy for reducing conditions is the Th/U ratio where lower ratios reveal more reducing conditions 92 . The average Th/U for MV1, MV2, and MV3 is 1.26, 0.59, and 1.09, respectively, where the minimum ratio in MV2 shows the most reducing conditions matching the results of real positive Ce anomalies (Fig. 13).
Our model of dolomite formation in the early-middle Miocene coastal mud volcano outcrop in north Kuwait is depicted in Fig. 15. In summary, dolomite in the studied early-middle Miocene mud volcano outcrop is associated with the activity of decapod crustaceans in the marine zones. The activity of burrowing crustaceans is well documented in seep carbonates worldwide 93,94 where the burrows serve as channels of fluid for vertical flow within the sediment 69 . Burrowing crustaceans can thrive in environmental stresses of high salinity and low oxygen levels exemplified by a sabkha environment in the Mississippian Debolt Formation located in Northwestern Alberta, Canada. These conditions have created dolomitizing bioturbated fabrics and the burrows themselves are dolomitized in the presence of microbial sulphate-reduction 95

Conclusion
We have re-evaluated the genesis of dolomite in north Kuwait by investigating recently discovered early-middle Miocene mud volcano outcrops within Al-Subiya sabkha. Our results suggest that interdependent events of hydrocarbon seepage, hypersaline seawater, and bioturbation led to dolomite formation in the coastal mud volcano. Our findings show that burrows can provide a preferential geochemical microenvironment for dolomite genesis within marine mud volcanoes due to the mixing of ascending methane-charged fluids with overlying seawater, in the bioirrigated sediment. The involvement of methane in dolomite formation is implied from the positive Ce anomaly that reflects strictly anoxic reducing conditions. Furthermore, the positive Eu anomaly of dolomite suggests that sulphidic conditions is likely coupled with anaerobic oxidation of methane. The importance of high salinity for dolomitization was supported by the presence of dolomite and NaCl in the marine zones and the distorted rock fabric is indicative of diagenesis. These results propose a new diagenetic model of dolomitization in the Arabian Gulf coastlines where most other diagenetic models were reported to be consequences of sabkha flood recharge/reflux, and concentrated seawater.