Massive Dolomitization of Interior and Slope to Basin-Margin Facies of the Triassic Yangtze Platform Through Superposed Earth-Surface and Burial Mechanisms, Nanpanjiang Basin, South China

Abstract

Triassic strata of the Yangtze Platform at Guanling contain a dolomitized interior, undolomitized margin, and partially dolomitized slope to basin margin. Dolomitized microbial laminate caps of peritidal cycles and massive dolomite with associated evaporite nodules and solution collapse breccias are consistent with penecontemporaneous tidal flat and evaporative dolomitization in the platform interior. The preferential dolomitization of the slope and basin margin (up to 7 km basinward of the margin), dolomitization along fractures, and selective dolomitization of the matrix in slope breccia that diminishes toward the margin are interpreted to have resulted from the incursion of basin-derived fluids during burial. Integrated analysis of fluid-inclusion microthermometry, oxygen, carbon, and strontium isotopes, trace element geochemistry, U-Pb age dates of carbonate phases, and burial history support the recrystallization of interior dolomite and slope to basin-margin dolomitization by brines at high temperatures during burial. The Yangtze Platform at Guanling provides an excellent example of widespread stratiform dolomitization resulting from the superposition of multiple mechanisms, including penecontemporaneous dolomitization by evaporative seawater brines, high-temperature dolomitization of the slope and basin margin by basinal brines, and high-temperature recrystallization of dolomite by brines during burial. This study provides an example that suggests that widespread stratiform dolomite may result from superposed Earth surface and high-temperature burial dolomitization processes and provides a valuable analog for other carbonate platforms in which the margin remains undolomitized while the interior and basin margin are dolomitized. Similar mechanisms likely contributed to the widespread dolomitization of platforms across the Nanpanjiang and Sichuan basins.

1. Introduction

Dolomite [CaMg(CO₃)₂] is a carbonate mineral and diagenetically modified sedimentary rock that is widely used in construction materials and is useful in providing porous and permeable formations important for fluid extraction or storage in the subsurface (aquifers, hydrocarbons reservoirs, carbon sequestration, and hydrogen storage). The origin of dolomite and the mechanisms responsible for the massive dolomitization of carbonate platforms have been the subjects of intensive research, yet they remain topics of considerable controversy e.g., [1,2,3,4]. The “dolomite problem” is the paradox that massive widespread dolomite is common in the geologic record of carbonate platform facies predominantly interpreted to have been deposited in shallow-marine settings at Earth surface temperatures, whereas dolomite is rare in Holocene sediments and laboratory experiments have failed to produce dolomite under these conditions [1,3,5,6].

A wide range of models have been proposed to explain the massive dolomitization of carbonate platforms including a variety of fluid compositions, mechanisms at Earth surface temperatures or high temperatures, and a variety of hydrological fluid flow mechanisms [1,2,3,7]. Seawater and evaporatively concentrated seawater dolomitization models are widely supported, as seawater provides an abundant supply of Mg and the large volumes of rock–water interaction needed for widespread dolomitization cf. [3]. However, penecontemporaneous dolomitization in modern tidal flats is considered to produce minor volumes of dolomite [3] and the viability of evaporative seepage reflux has been questioned in case studies based on the lack of mineralogical or geochemical trends that are expected with reflux dolomitization [8,9]. High-temperature burial or hydrothermal dolomitization offers the advantage of overcoming the kinetic barrier to readily form ordered dolomite [1], but hydrothermal dolomitization has also been met with skepticism as a regional dolomitization mechanism as some authors consider amounts of hydrothermal dolomite to be volumetrically minor and concentrated along fractures and adjoining permeable stratigraphic conduits [3,10]. However, convincing textural, geochemical, geothermometric, and modeling evidence has been presented in support of widespread hydrothermal dolomitization [7,11,12,13,14,15,16,17,18,19].

There are several potential obstacles to resolving the dolomite problem from outcrop studies that have been discussed in review papers [1,2,3,20,21] including the following: (1) pervasive fabric destructive dolomitization that obscures paragenetic sequence, fabrics, and cross-cutting relationships that would provide constraints on the environment and relative timing of dolomitization; (2) a limitation of outcrop exposures to only one part of a carbonate platform such as an interior lagoon, whereas exceptionally exposed platforms with preserved transects of environments from shelf to basin and cross-sectional exposure are needed to provide spatial and temporal constraints to test models; (3) particular geochemical and petrographic proxies to evaluate fluids and temperatures of dolomitization are subject to multiple interpretations and alteration; and (4) studies that focus on a particular model for dolomitization (e.g., seawater dolomitization, hypersaline tidal flat or reflux dolomitization, freshwater–seawater mixing zone dolomitization, and burial or hydrothermal dolomitization) rather than evaluate the possibility of multiple overprinted mechanisms.

Recent outcrop and subsurface studies now routinely apply a diverse range of petrographic, mineralogical, geochemical, and geothermometric techniques in an attempt to unravel mechanisms, and several studies have documented the occurrence and complexity of overprinted multiple stages of dolomitization including Earth-surface and high-temperature burial mechanisms, e.g., [22,23,24,25,26,27,28,29,30,31,32]. However, despite the recent advances, the lack of consensus on the primary mechanisms responsible for the widespread massive dolomite in the rock record remains.

The goal of this study is to apply field observations, thin-section petrography, geochemical and geothermometric techniques, considerations of burial history, and LA-ICP-MS U-Pb dating to assess the environments and timing of dolomitization of the Lower to Middle Triassic platform to basin transect of the Yangtze Platform in the Guanling area of Guizhou Province, south China. The Yangtze Platform strata in the Nanpanjiang Basin of south China offer several advantages for the outcrop study of dolomitizing mechanisms to help unravel multiple overprinted mechanisms. The Yangtze Platform and isolated platforms in the Nanpanjiang Basin contain exceptionally widespread dolomitized formations, and the strata are partially dolomitized and exposed in continuous 2-D cross-sections with platform–basin transition paleobathymetry and environmental transects [33,34,35,36,37]. Further, the exceptional cross-sectional exposure and mature stratigraphic framework enable relative age control and aid in the reconstruction of burial history, e.g., [35]. The results show that the Yangtze Platform was dolomitized through the superposition of multiple mechanisms including Earth surface penecontemporaneous dolomitization of tidal flats and lagoon facies by evaporatively concentrated seawater in the platform interior followed by the high-temperature dolomitization of the slope and basin margin by the incursion of basinal brines during burial, and the high-temperature recrystallization of earlier dolomite phases during burial.

2. Geologic Setting

The Nanpanjiang Basin formed a deep-marine embayment in the southern margin of the south China tectonic block (present coordinates; Figure 1a inset). Plate reconstructions indicate that the south China block drifted northward across the eastern Tethys to approximately 12° north by the beginning of the Middle Triassic and docked with the north China plate during the Late Triassic [38,39,40,41]. The Siamo and Indochina blocks converged upon and collided with the southern margin of the south China block sometime during the Late Paleozoic or Early Triassic [42,43].

During the Permian and Triassic Periods, the Nanpanjiang Basin was bordered by the Yangtze Platform, a vast shallow-marine carbonate platform that stretched across the south China block (Figure 1a inset). The Nanpanjiang Basin spans 276,000 km² and contains several isolated carbonate platforms (Great Bank of Guizhou, Chongzuo-Pingguo, Heshan, and Debao) that developed within the basin during the Triassic (Figure 1). The Yangtze Platform and isolated platforms are dissected by faulted folds that preserved complete 2-D cross-sections through the platform interior to margin, slope, and basin architecture (Figure 1 and Figure 2).

During the Early Triassic, Induan age, the Yangtze Platform and isolated platforms developed low-relief ramps with oolitic margins and pelagic carbonate slopes. Inner-ramp lime mud shallowed upwards to oolitic and microbial to peritidal cycles in the platform interior (Figure 1). During the Early Triassic, Olenekian age, the Yangtze Platform and isolated platform margins steepened and produced a relatively flat-topped geometry with margin shoals, restricted interior facies, and pelagic carbonate slopes with debris-flow breccia and carbonate turbidites [33,35,44]. Platform margin facies consisted of oolitic shoals with tepee-disrupted intervals representing back shoal islands. Platform interior facies are composed of burrowed skeletal peloidal and oncolitic subtidal facies, and peritidal cycles with stromatolites, fenestral laminites, and tepee-disrupted peritidal facies [33,34]. Evaporitic conditions in the platform interior lagoons are inferred from common gypsum molds and widespread evaporite dissolution breccias in the middle and upper Olenekian stages (Figure 1) [33,35]. The Olenekian margin and interior facies are extensively dolomitized across the basin in the Anshun Formation of the Yangtze Platform and Great Bank of Guizhou, and the Beisi Formation of the Chongzuo-Pingguo, Heshan, and Debao platforms (Figure 1).

During the Middle Triassic, Anisian age the Yangtze Platform and Great Bank of Guizhou developed prograding and progressively steepening margins rimmed with Tubiphytes reefs and restricted interior lagoon and tidal flat facies. The Guanling Formation in the platform interior is commonly dolomitized, and slope facies are partially dolomitized, whereas the Tubiphytes boundstone remains limestone (Figure 1c and Figure 2a,b). Evaporite crystal molds and evaporite solution collapse breccias indicate evaporitic conditions in the platform interior (Figure 1) [33]. During the Middle Triassic, Ladinian age the basin rapidly subsided, and the Yangtze Platform developed laterally variable margin architectures that included prograded margins, aggraded margins, high-relief escarpments, and tectonically back-stepped or collapsed margins (Figure 1) [34,35,36]. A thick succession of siliciclastic turbidites (Xinyuan, Xuman, and Bianyang formations, up to 5 km in total) infilled the basin during the Middle Triassic (Figure 1 and Figure 2).

During the Late Triassic, subsidence increased in the basin resulting in the drowning of the western sector of the Yangtze Platform, including the Guanling area [33,35,36,45]. Drowning was marked by an abrupt shift to the thin-bedded, dark-colored pelagic limestone and black shale of the Zhuganpo and Wayao formations (Figure 1) [33,36,37]. The drowning succession on the Yangtze Platform is overlain by a thick succession of Upper Triassic siliciclastic turbidites (Laishike Fm., ~900 m thick) that grades upward to shallow marine and fluvial siliciclastic strata (Banan, Houbachong, and Erqiao formations, >1500 m thick) [33] and Jurassic terrestrial deposits [46].

3. Materials and Methods

To evaluate mechanisms of the dolomitization of the Lower to Middle Triassic strata of the Guanling Margin of the Nanpanjiang Basin, high-resolution geologic maps, field relationships, stratigraphic sections, petrographic relationships, isotope and elemental geochemistry, fluid-inclusion microthermometry, and U-Pb isotope age dates were analyzed. Data generated in this study are available at the following data repository: https://doi.org/10.7910/DVN/S0KKWJ (data archive created on 20 January 2025).

3.1. Field Work

Geologic maps at 1:50,000 and 1:200,000 scale were overlain on satellite images including basin-wide Landsat TM images, 50 cm-per-pixel Worldview images, high-resolution satellite images in Arc-GIS and Google Earth, and digital elevation models. The maps were used to assist with field mapping and to plot dolomite sample distribution within the Yangtze Platform and its associated slope units at Guanling (Figure 2a,c). Additional field work in the Guanling area included observation of dolomite, sedimentary and diagenetic; structural features; and basic stratigraphic and cross-cutting relationships to constrain environments and the relative timing of dolomitization.

Stratigraphic sections spanning the platform-to-basin transition of the Yangtze Platform at Guanling were measured at Laowai, Natau, and Xitouzhai and integrated with a section previously measured at Hongyan (Figure 2d) [37]. Scales of stratigraphic sections ranged from 1:40 to 1:200. To correlate sections, spectral gamma-ray measurements were made at each section with a handheld spectrometer (Radiation Solutions model RS-125; Mississauga, ON, Canada) [37]. Samples of dolomite with representative textures and structures were collected from the Lower through Middle Triassic facies for geochemical analysis and petrography. A few samples were collected from the dolomitized uppermost Permian to basal Triassic (Permian–Triassic boundary) beds in the basin margin at Natau and Xitouzhai sections (Figure 2d). Additional samples were collected in traverses through the Lower Triassic, platform-interior Anshun Formation at the Doupengzhai and Hongyan north sections (Figure 2c). The sample location was denoted in the field using GPS and plotted on geologic maps and satellite images to visualize the stratigraphic position (Figure 2c).

3.2. Petrographic Methods

Thin section petrography on 150 polished and unpolished thin sections was conducted and aided with alizarin red staining to highlight calcite versus non-stained dolomite phases. Polished thin sections were prepared at low temperatures for use with fluid-inclusion microthermometry. Low-power Leica MZ-165 and high-power Leica DMP-2500 microscopes (Deerfield, IL, USA) were used for petrography. Dolomite phases were classified using the textural classification scheme of [47]. Petrographic observations also were used to determine the overall paragenetic sequence, helping to characterize the relative timing of dolomitic phases based on cross-cutting relationships.

3.3. Burial History

Burial history analysis was performed using the measured thicknesses of the Yangtze Platform and overlying strata from [33] and the stratigraphic thicknesses of the Guanling margin from our sections (Figure 2). Burial temperatures were estimated using an average thermal gradient of 30 °C/km, and the assumption of an Earth surface temperature of 30 °C consistent with [48,49]. Thermal history was further constrained using the conodont color alteration index [50] for conodont samples from the nearby Great Bank of Guizhou, which has a similar burial history to the Yangtze Platform [35].

3.4. Fluid-Inclusion Microthermometry

Fluid-inclusion microthermometry was conducted on 47 samples containing primary fluid-inclusion assemblages (FIAs) in the crystals of replacement dolomite, euhedral dolomite cement, late-stage saddle dolomite and euhedral zoned dolomite, and vein-filling calcite. Fluid-inclusion microthermometry was not possible for aphanitic or very finely crystalline dolomite (≤150 µm) because the inclusions are too small to measure. Fluid inclusions were analyzed in doubly polished 60 µm thick sections prepared with care to avoid heating. Fluid-inclusion microthermometry was performed with a Linkham THM 600 stage (Salfords, Redhill, UK) with a 40× or 100× objective at Trinity University and with a Fluid Inc.-adapted USGS design gas-flow heating and freezing stage, and 40× Nikon objective (Melville, NY, USA) at the University of Wisconsin—Green Bay.

Petrography was used to identify the primary fluid inclusions trapped in crystal growth zones or within the interior of dolomite crystals. Fluid-inclusion homogenization temperatures were used to assess the temperature at the time of entrapment of fluids during dolomitization [51,52]. These temperatures were collected from samples spanning the platform interior, platform margin, slope, and basin for each discernable dolomite phase. We grouped petrographically associated fluid-inclusions into fluid-inclusion assemblages (FIAs) to reduce inter-sample variability and to achieve the most consistent results to interpret the temperature of dolomite formations using homogenization temperature (Th). The stages were calibrated using synthetic fluid inclusions between −120 °C to 400 °C. Freezing–heating experiments on fluid inclusions yield freezing point depression (Tm_ice) and ice nucleation temperatures (Tn), which constrain the salinity of dolomite formation. T_h and Tm_ice are measured and reported with a precision of 0.5 °C and 0.1 °C, respectively. Tm_ice of fluid inclusions in dolomite were further converted to NaCl equivalent salinity following [53]. For inclusions in which Tm_ice data could not be obtained due to metastable vapor bubble nucleation, the T_n of ice was used to estimate the salinity of the inclusions using the method of [54].

3.5. Geochemistry

The samples were drilled for specific carbonate phases (replacement dolomite, saddle dolomite in vugs, or calcite veins) using a microdrill with a diameter of 0.3 mm. Only the phases that could be clearly separated were microdrilled. Replacement dolomite was tested with acid first to minimize the mixing of calcite and dolomite phases. Carbon, oxygen, and strontium isotopes were analyzed in the context of environmental gradients across the preserved platform interior to basin profiles. Stable isotopes of carbon and oxygen were measured at the W. M. Keck Paleoenvironmental and Environmental Stable Isotope Laboratory at the University of Kansas. Approximately 75 µg of each sample was weighed into a stainless steel boat using a Mettler Toledo microbalance (Columbus, OH, USA). Stainless steel boats with samples were placed in a brass convoy and samples were vacuum roasted at 200 °C for 1 h to remove volatiles. Samples were analyzed, and sample CO₂ was generated by reaction with 3 drops of 105% H₃PO₄ at 70 °C for 540 s using a Thermo Scientific Kiel IV Carbonate Device (Waltham, MA, USA) interfaced to the inlet of a ThermoFinnigan MAT 253 dual inlet mass spectrometer (Waltham, MA, USA). Carbonate δ¹³C and δ¹⁸O isotope data were calibrated relative to the VPDB scale through daily analysis of primary standards. NBS-18 and/or NBS-19 and analytical performance were monitored with additional analysis of secondary standards TSF-1, Sigma Calcite, and Dolomite. The results have a precision better than 0.10‰ for both δ¹³C and δ¹⁸O.

Strontium isotopes were measured at the University of Kansas Thermal Ionization Mass Spectrometry (TIMS) Laboratory. ⁸⁷Sr/⁸⁶Sr ratios were determined using a VG Sector 54 (run at a signal strength of Sr⁸⁸ = 4 V) thermal ionization mass spectrometer (Wythenshawe, Manchester, UK) operating in dynamic mode. The samples were dissolved in 3.5 N HNO₃, and the strontium was eluted through ion exchange columns filled with strontium-spec resin. We used Eichrom strontium-spec resin to isolate the Sr in the samples. The samples were spiked with an ⁸⁴Sr tracer. We applied an isotope fractionation correction using ⁸⁶Sr/⁸⁸Sr = 0.1194 and an exponential correction. Isotope ratios were adjusted to correspond to a value of 0.710248 on NBS987 for ⁸⁷Sr/⁸⁶Sr to be consistent with the [55] normalization. Internal precision was 15 ppm, and external precision was 3 ppm in ratio to standards during the duration of analyses.

Additional Sr isotope analyses were measured at the FIRST lab at Stony Brook University. Sr isotope analyses were performed on an Isotopx Phoenix X62 thermal ionization mass spectrometer (TIMS) (Middlewich, Cheshire, UK). The samples were dissolved, and Sr was isolated in each sample using Eichrom Sr-Spec ion exchange resin. All the separation chemistry and dissolutions were performed in a class 10 clean room using ultra-pure chemicals. Approximately 200 ng of Sr was loaded onto zone-refined Re filaments with a Tantalum “loader” solution for mass spectrometer analysis. Several standards (NIST SRM 987) [56] are run with every set of samples, ~4 per sample set. The long-term average is used to correct the sample to the certified NIST value. An in-run mass bias correction was applied for all the sample and standard analyses using the in-run measured ⁸⁶Sr/⁸⁸Sr ratio which is corrected to the natural abundance ratio of 0.1194 using an exponential relation. Internal precision for a single run is less than 0.0004% so it is a negligible contribution to the analytical uncertainty. Analytical uncertainty is determined from the 2-sigma uncertainty of repeated analyses (the average long-term, 2-standard deviation reproducibility of the lab is currently 11 ppm) of SRM 987 before and during the time of the analyses.

Trace element analysis and preliminary U-Pb age dates were performed at the Facility for Isotope Research and Student Training (FIRST lab) at Stony Brook University. The instruments used for this study were an Agilent 7500cx quadrupole inductively coupled plasma mass spectrometer (ICP-MS) (Santa Clara, CA, USA) coupled to a solid-state 213UV New Wave Instruments laser ablation system (Fremont, CA, USA). A circular laser spot of either 80 or 160 µm was selected depending on the concentration of U and Pb in the carbonate sample, and a NIST612 [56] glass was selected as the primary standard to maximize signal while minimizing oxides and doubly charged ions. A final tune-up was performed with WC-1 based on the gas flow and torch position to adjust for the ²³⁸U/²⁰⁶Pb ratio, which should be around 23 based on isotope dilution [56]. For each session, WC-1 [56] was run, and Barstow [57] was used as a secondary standard. Data were then viewed in Iolite4 [58] to ensure that the ratios on WC-1 were reasonable and that the spread in Barstow yielded the correct age within uncertainty. Laser lines were created by moving the sample at a rate of 15 microns per second under the laser, and helium was used as the carrier gas at a flow rate of about 1 L per minute. A pause for 30 s between lines was performed to wash out any residual material. Geochronologic data were reduced using the U-Pb Geochronology DRS tool in Iolite4 [58] by subtracting the baseline, correcting for downhole fractionation (which is negligible with line scans), and making a final correction for fractionation and drift in the signal. Concentration data were reduced using the Trace Elements DRS tool in Iolite4 [59]. Minor and trace elements measured included Mg, Si, Sr, Li, B, Fe, Mn, U, Th, and REEs. The [60] approach used for extracting and pooling pixels was used in Iolite4 [58] to provide a framework for interpreting LACarb ages. For example, to select for calcite, a criterion of elevated counts of Ca and low counts for Si was chosen. The pixels were subdivided into 50 pools based on probability on ²³⁸U/²⁰⁸Pb, allowing for the greatest spread in U-Pb without the bias that might be introduced by the use of ²³⁸U/²⁰⁶Pb. The 10% from the lowest and highest ends was removed as recommended by [60], and the remaining data points were plotted on a Tera–Wasserburg plot using IsoplotR model 3 [61].

The updated age dates presented herein were collected at the University of Arizona LA-ICP-MS lab. The samples were analyzed using the methodologies outlined in [60], which involved image-based mapping of uranium (U) and lead (Pb) across a sample through successive linear scans to maximize the variation in U-Pb ratios for creating Tera–Wasserburg diagrams. Laser ablation, along with isotopic and elemental measurements, was carried out using an ESL NWRfemto system with a 257 nm femtosecond laser coupled to a ThermoFisher Element2 mass spectrometer (Waltham, MA, USA). The data processing utilized Iolite 4 [58] for baseline and drift corrections, as well as normalizing U and Pb to a primary reference material (NIST614 glass). The ASH-15 carbonate standard [62] served for the matrix matching of ²³⁸U/²⁰⁶Pb and age correction, while JT vein calcite [63] or the WC1 standard were employed as secondary standards. Basic parameters were used to filter areas of analysis including 0 < ²³⁸U/²⁰⁶Pb < 1000, 0< ²⁰⁷Pb/²⁰⁶Pb < 1, U > 0.01 ppm, and Th < 0.5. The resulting areas of analysis were segmented and pooled into pseudo-spots based on the probability of similar ²³⁸U/²⁰⁸Pb ratios, with each accounting for approximately 20 s of analytical time [60].

4. Results

4.1. Dolomite Distribution and Field Observations of Dolomitized Facies

Throughout the Lower to Middle Triassic, Anisian stratigraphic succession, dolomite is widespread in the platform interior and on the slope to basin-margin facies. In contrast, the platform margin facies is undolomitized (

Table 1. Stratigraphic units, interpreted environments, degree of dolomitization, and dolomite types.

Age and Stratigraphic Unit	Position on Depositional Profile	Lithofacies	Interpreted Depositional and Diagenetic Environments	Degree of Dolomitization	Dolomite Types and Interpreted Stages of Dolomitization
Middle Triassic, Anisian
Guanling Fm.	Interior	Meter-scale cycles, skeletal oncolitic wk to pk bases, fenestral microbial laminate caps, meniscus cements.	Tidal flats, cycles with subtidal bases, and subaerial exposure on cycle caps. Dolomitization by evaporitic brine on cycle caps *.	Partial; only the subaerially exposed caps of cycles are dolomitized. Fabric preserving.	Syndepositional, replacement anhedral very fine to finely crystalline non-planar. Later coarse crystalline dolomite cement lining fenestral pores.
		Massive bioturbated skeletal md to wk, stratiform breccias, angular clasts, cauliflower-shaped molds.	Restricted subtidal lagoon, evaporitic conditions, evaporite crystal molds, and bedded evaporites (represented by dissolution collapse breccia). Dolomitization by evaporitic brines *.	Pervasively dolomitized in the Guanling area. Largely fabric destructive.	Replacement anhedral to subhedral, medium to coarsely crystalline with euhedral crystals facing open pores. Later coarse dol. cement and saddle dol. lining vugs and fractures in breccia, cutting across earlier phases.
Poduan Fm.	Margin	Tubiphytes boundstone, bladed isopachous, and botryoids of cement.	Subtidal marine and extensive marine cement.	Undolomitized except for very minor dolomitization of micritic elements at contact with foreslope.
Qingyan Fm.	Slope	Thin bedded, laminated carbonate mudstone, interbeds of gr, pk, and polymict breccia.	Deep-marine pelagic rain, mass gravity deposits (carbonate turbidites, grain flows, and debris flows). Dolomitization by basin-derived fluids during burial.	Partially dolomitized lime mudstone (fronts cut across beds), micritic matrix in breccias. Largely fabric retentive, locally destructive.	Replacement anhedral to subhedral, medium to coarsely crystalline, and coarsely crystalline dolomite with local euhedral crystals facing open pores in gravity flow deposits. Later coarse crystalline dolomite cement lining pores in gravity flow deposits and vugs in carbonate mudstone. Saddle dolomite lining vugs and fractures that cut across earlier phases of replacement dolomite.
Xinyuan Fm.	Basin margin
Lower Triassic (Induan-Olenekian)
Anshun Fm. (Olenekian)	Interior	Meter-scale cycles, wk to pk bases, fenestral microbial laminate caps; Massive bioturbated skeletal md to wk, stratiform breccias, angular clasts, cauliflower-shaped molds.	Tidal flats and restricted subtidal lagoon. evaporitic conditions, evaporite crystal molds, and bedded evaporites (represented by dissolution collapse breccia). Dolomitization by evaporitic brines of tidal flat and lagoon facies *.	Pervasively dolomitized. Largely fabric destructive.
Gujiao Fm. (Olenekian)	Margin	Oolite grainstone.	Subtidal marine and extensive marine cement	Undolomitized, stratigraphic equiv. of Anshun Fm. dolomitized.
Luolou Fm. (Induan-Olenekian)	Basin margin	Thin bedded, laminated carbonate mudstone, interbeds of gr, pk and polymict breccia.	Deep-marine pelagic rain, mass gravity deposits (carbonate turbidites and debris flows). Dolomitization by basin-derived fluids during burial.	Partially dolomitized lime mudstone (fronts cut across beds), micritic matrix in breccias. Largely fabric retentive, locally destructive.
Upper Permian, Changhsingian
Wujiaping Fm.	Basin margin	Skeletal wackestone to packstone.	Shallow-subtidal, open marine (prior to transgressive backstep of margin). Dolomitization by basin-derived fluids during burial.	Pervasively dolomitized locally at the basin margin. Fabric destructive.

* syndepositional dolomite was recrystallized and overprinted by additional dolomite phases during burial. Md, wk pk, gr = Dunham texture terms.

; Figure 2b,d).

Platform-interior facies of the Lower Triassic (Olenekian) Anshun Formation contains widespread dolomite immediately platformward of the oolitic margin facies of the Doupengzhai section (Figure 2b,c) and in platform interior lagoon and tidal flat facies (Hongyan north section; Figure 2c). Further landward, carbonate facies of the Yongningzhen Formation are limestone (Figure 1c). Dolomite in the Anshun Formation includes massively dolomitized beds with fabric-destructive dolomite, stratiform dolomitized breccias, and pervasively dolomitized tidal flat facies (Table 1). The dolomitized breccia intervals contain cauliflower-shaped molds and stratiform breccias with angular chaotically oriented clasts (Table 1; Figure 3a,b).

Platform-interior facies of the Middle Triassic, Anisian Guanling Formation contains dolomite in the immediate backreef, platform interior lagoon, and tidal flats (Figure 2). Further landward, carbonate facies of the Guanling Formation are limestone (Figure 1c). Peritidal cyclic facies in the immediate backreef are partially dolomitized with dolomitized fenestral and microbial laminate caps (Table 1; Figure 3c). Dolomite in the platform interior lagoon facies of the Guanling Formation is composed of massively dolomitized thick beds (Figure 3d) and stratiform dolomite breccias with angular chaotically oriented clasts (Table 1; Figure 3e).

Platform-margin facies of the Middle Triassic Poduan Formation is composed of tightly cemented Tubiphytes boundstone with bladed marine cement and is undolomitized (Table 1; Figure 2b,d). Platform margin facies of the Lower Triassic Anshun Formation is locally dolomitized, but the oolitic facies is mapped as the undolomitized Gujiao Formation [46] likely due to the substantial bladed marine cements present in this margin facies cf. [37]. At the margin adjacent to the Doupengzhai section, the Lower Triassic oolitic facies of the Gujiao Formation is undolomitized (Table 1; Figure 2c).

Slope and basin-margin facies spanning the uppermost Permian to the Middle Triassic are partially dolomitized (Table 1; Figure 2a,d). Dolomitized facies were observed up to 7 km basinward from the platform margin at Xitouzhai (Figure 2a). Dolomitized facies include foreslope debris-flow breccia, distal thin debris-flow breccias, and carbonate turbidites (Figure 4a–c) at the basin margin as well as planar laminated, thin-bedded pelagic limestone and slump-folded pelagic limestone along the slope and basin (Figure 4d–g). The proportion of dolomite is greater in the basinward sections, with the proportion decreasing up the slope towards the platform margin (Figure 2a,b,d). For example, at the distal basin margin Xitouzhai section, the proportion of dolomite by stratigraphic thickness is 34%, whereas the proportion in the more proximal Natau section is 13% (Figure 2d). The proportion continues to decrease up the slope, until the proximal foreslope contains only minor matrix dolomitization, and the margin reef is entirely undolomitized (Figure 2b).

In the basin margin and slope, irregular, non-stratiform contacts between dolomite and host limestone (interpreted as dolomitization fronts) cross-cut bedding as dolomite follows bedding-parallel and across-bedding fractures (Figure 4d,e). Pelagic carbonate mudstone and thin carbonate turbidite beds are completely dolomitized in the basin-margin sections at Xitouzhai and Natau (Figure 4g), whereas this facies is only partially dolomitized in slope sections (Figure 4d–f). Closer to the platform margin at Laowai and in the foreslope adjacent to the platform margin, dolomite decreases in abundance and preferentially replaces the lime–mud matrix of debris-flow breccia (Figure 4a–c). In the immediate forereef and basinward edge of the Tubiphytes reef, very minor amounts of dolomite occur in the matrix of breccia and in fractures in the boundstone, as well as in minor micritic framework and grains of the otherwise undolomitized boundstone (Figure 5a).

4.2. Petrography

4.2.1. Precursor Limestone Facies

Field observations combined with thin-section observations of fabric retentive dolomite, partially dolomitized facies, and undolomitized host rock from Anshun and Guanling formations revealed seven end-member precursor limestone facies that were later partially or fully dolomitized:

• Interior microbial laminated, fenestral packstone with vadose cements (Figure 3c and Figure 5b).
• Interior pervasively dolomitized massive thick-bedded to massive intervals, which are inferred to have originally been mudstone to wackestone. Locally, this facies contains fabric observable in outcrop including burrows, oncoids, and skeletal fragments (Figure 3d).
• Interior stratiform carbonate breccia intervals that are pervasively dolomitized and interpreted to be evaporite solution collapse breccia (Figure 3d,e). The stratiform breccia contains angular clasts, often displays concave–convex contacts between clasts, and contains molds of evaporite nodules (Figure 3a,e).
• Platform margin Tubiphytes boundstone that is tightly occluded by bladed marine cement and locally contains minor dolomitized micritic elements (Tubiphytes and grains) at the transition from the reef facies to the foreslope (Figure 5a).
• Basin-margin fine-grained, laminated pelagic carbonate mudstone (Figure 4d–g and Figure 5c).
• Basin-margin grainstone to packstone beds (Figure 4g and Figure 5d).
• Slope debris-flow breccia pervasively dolomitized at the basin margin and partially dolomitized in the upper slope and proximal foreslope facies (Figure 4a–c).

4.2.2. Dolomite Types and Paragenetic Sequence

Dolomite types were defined based on crystal size and shape, fabric retentive, or destructive texture, and position within the fabric of the rock (e.g., replacement of matrix, grains, cements, or occurrence in pores, fractures, or vugs) as described below:

• Anhedral very fine to finely crystalline (0.01–0.06 mm) non-planar crystals that replace micrite, calcareous grains, and vadose cements in the microbial caps of peritidal cycles in the platform interior (Figure 5b) as well as a complete replacement of pelagic lime mudstone facies in the basin margin and slope (Figure 5c and Figure 6a).
• Anhedral to subhedral, medium to coarsely crystalline (0.06–0.5 mm) non-planar to planar-s fabric-destructive dolomite that completely replaced massive thick beds and evaporite solution collapse breccia beds in the platform interior (Figure 3d,e) and fabric retentive dolomite that partially or completely replaces packstone–grainstone carbonate turbidite beds in the slope and basin margin (Figure 5d). This dolomite contains cloudy cores with abundant inclusions.
• Coarse euhedral replacement dolomite (0.3–0.8 mm) that occurs sporadically in massive replacement dolomite (described above) and locally has euhedral crystal surfaces facing extant intergranular or vuggy porosity. This dolomite contains cloudy cores with abundant fluid inclusions and outer zones highlighted by variation in inclusion density.
• Coarse euhedral dolomite cement (0.3–0.8 mm) lining extant fenestral or intergranular porosity in platform interior facies and in slope grainstone beds (e.g., Figure 5b).
• Saddle dolomite and euhedral zoned dolomite cement crystals up to 1 mm infilling voids between clasts in platform interior breccia and late-stage vugs or fractures in slope and basin-margin facies (Figure 6b–e).

Observations of dolomite and other diagenetic phases and their cross-cutting relationships revealed a relatively consistent paragenetic sequence of phases from Lower to Middle Triassic strata from the platform interior and slope. Diagenesis began with the formation of syndepositional cements, including bladed marine cements and micritic vadose cements (Figure 5a,b), followed by early replacement dolomitization (Table 1).

Early replacement dolomitization was observed petrographically to occur in two or three ways. In the platform interior facies, it is represented as a very fine to finely crystalline anhedral replacement of micrite, carbonate grains, and micritic vadose cements in peritidal facies (Table 1; Figure 3c and Figure 5b); minor dolomitization of micritic elements in the reef front boundstone; and medium to coarsely crystalline anhedral to subhedral complete replacement of massive beds and stratiform breccia in massive interior lagoon facies (Table 1; Figure 4a,b and Figure 5a). The early replacement dolomite in the platform interior is most likely overprinted by phases that occur in different facies and include the superposed recrystallization of earlier phases as discussed in later sections (Table 1). In the slope and basin margin early replacement dolomite is represented by very fine to finely crystalline partial to complete replacement of pelagic carbonate mudstone and micrite matrix in debris-flow breccias (Table 1; Figure 4, Figure 5c, and Figure 6c) and as medium to coarse partial to complete replacement of packstone–grainstone and breccia beds (Table 1; Figure 4 and Figure 5c,d).

In both the platform interior and slope to basin margin, the replacement dolomitization phases were followed closely by a stage of euhedral dolomite cement lining extant depositional pores (Table 1; Figure 5b).

The replacement dolomite and dolomite cements were cross-cut by multiple generations of fractures and the development of vuggy porosity (Figure 6). Fractures and vugs contain a yet later stage of saddle dolomite (Figure 6c) and strongly zoned euhedral dolomite cement (Table 1; Figure 6d,e). The saddle dolomite contains characteristic curved crystal phases and exhibits sweeping extinction (Figure 6b,c). Additional generations of fractures cross-cut the saddle and zoned euhedral dolomite and are filled with calcite (Figure 6f). The vein-filling calcite contains twinning indicating subsequent structural deformation (Figure 6a,f). Stylolites are late diagenetic features and cross-cut all dolomite and twinned calcite vein phases (Figure 6f).

4.3. Constraints on Burial History and Temperatures

A basic constraint on burial depth can be made by using the present-day thicknesses of the Triassic strata at the Hongyan section (Figure 2d) and adding the overlying Upper Triassic and Jurassic strata using the regional stratigraphic thicknesses from [33] (Figure 7a). Estimates do not include decompaction corrections, therefore making these minimum burial estimates. Post-Jurassic strata are also not preserved, so additional burial is not known. Using burial depths, temperature estimates were made using a geothermal gradient of 30 °C/km and a mean annual surface temperature of 30 °C. Tropic sea surface temperatures reported up to about 35 °C during the Early Triassic [48]. Ref. [49] reconstructs approximately 31 °C in the Olenekian, and 28 °C in the Anisian. Thus, for burial temperature estimates using a mean annual surface temperature of 30 °C is adequate. Given these parameters, by the end of the Triassic, the specified stratigraphic levels would have reached the following depths and temperatures (Figure 7a): Permian–Triassic boundary (4405 m; 162 °C), top of Lower Triassic (3585 m; 138 °C), and top of Middle Triassic, Anisian (2705 m; 111 °C). Further, by the end of the Jurassic, the specified stratigraphic levels would have reached the following depths and temperatures: Permian–Triassic boundary (5005 m; 180 °C), top of Lower Triassic (4185 m; 155 °C), and top of Middle Triassic Anisian (3305 m; 129 °C).

A similar analysis was performed on the Great Bank of Guizhou on an isolated platform approximately 100 km southeast of the Guanling area (Figure 7b). The results indicate the burial depths and temperatures reached by the end of the Jurassic for the horizons as follows (Figure 7b): Permian–Triassic boundary (5003 m; 180 °C), top of Lower Triassic (4112 m; 153 °C), and top of Middle Triassic Anisian (3377 m; 130 °C) and top of Middle Triassic Ladinian (2227 m; 97 °C). For the Great Bank of Guizhou, we used the conodont color alteration index (CAI) [50] to constrain the burial temperatures and found conodonts with CAI grade 4 (190–300 °C) at the Permian–Triassic boundary at the base of the platform whereas conodonts at the top of the Middle Triassic Ladinian at the top of the platform had grade 1 (50–80 °C). Although the CAI grades have a wide range in possible temperatures, the minimum temperature for the base of the platform indicated by CAI grade 4 (190 °C) is consistent with the estimated burial temperature (180 °C), and the CAI at the top of the platform, in the uppermost Middle Triassic Ladinian, is roughly equivalent to the estimated burial temperature.

Twinning in the calcite filling late-stage fractures (calcite veins, e.g., Figure 5e and Figure 6a) falls in the “thin twin” or type 1 category of [64]. Thin twins (type 1) [64] are characterized as having a black line appearance in thin sections and being on average less than 2 µm in width, as opposed to thick twins (type 2) [64] that appear as a finite thickness of twinned material. We measured twin thicknesses < 2 µm and a density of 54 to 100 twins/mm (e.g., Figure 5e). The dominance of thin twins, width < 2 µm and density > 40/mm, indicate temperatures of deformation less than 170 °C [64] (Figure 2). In some thin sections, curved twins and crenulation of twins occur (Figure 5e) indicating complex deformation.

4.4. Fluid-Inclusion Microthermometry

Fluid-inclusion microthermometry was performed on natural primary fluid-inclusion assemblages (FIAs) within individual dolomite crystals. Some FIAs in individual crystals contained as many as 15–25 inclusions. For replacive dolomite (e.g., DP-33 and DOL-200; Figure 8a,b inset photos), crystals were typically weakly zoned or had a cloudy, inclusion-rich core enveloped by a clear outer rim. Most inclusions measured range from 1 to 6 µm, though rare inclusions up to 22 µm were measured. Inclusions have consistent, low vapor-to-liquid ratios and are distributed throughout the crystals. Many of these inclusions contained moving vapor bubbles, especially at higher temperatures, which allowed for small inclusions to be measured more easily.

Inclusions were considered primary to the replacement or precipitation of the dolomite because they are contained within zoned crystals or in material that preceded the outer clear growth zone cement (e.g., Figure 9 inset photos). Evidence for necking after a phase change was not observed. In some cases, unmeasurably small but abundant ≤ 1 µm inclusions with moving vapor bubbles were observed in the same temperature range as larger inclusions, which also supports the interpretation that the inclusion properties are representative of replacement or cementation rather than the stretching of larger inclusions or entrapment along unrecognized fractures.

For saddle dolomite crystals and euhedral-zoned dolomite cement (e.g., DP-11, DS-1a), all the inclusions occurred along specific inclusion-rich growth zones that were bounded by inclusion-poor dolomite (Figure 8c and Figure 9 inset photos). The inclusions were interpreted as primary because they occur along an obvious growth zone and are considered representative samples of the fluid present during precipitation.

Fluid-inclusion microthermometry data are presented in histograms organized by position on the platform-to-basin transect and stratigraphic age with individual measurements color-coded by FIA and segregated by diagenetic phase (Figure 8, Figure 9, Figure 10 and Figure 11). Homogenization temperatures (T_h) from FIAs in platform interior replacement dolomite, and associated dolomite cement including both the Lower Triassic and Middle Triassic Anisian samples were found to average 117 °C, SD (standard deviation) = 24.1 °C, with a range of 66–186 °C, N (number of measurements) = 73, (Figure 8a). T_h from FIAs in replacement dolomite from the Anisian foreslope facies were found to average 118 °C (SD = 16.6 °C) in a range of 85–141 °C (N = 68, Figure 8b). Measurements from N = 20 inclusions in a FIA in a single crystal from the Anisian foreslope replacement dolomite yielded an average T_h of 121 °C (SD = 12.9 °C) in a range of 92–141 °C (Dol-200, Figure 8b). T_h from Lower Triassic (Induan and Olenekian) replacement dolomite FIAs in the basin-margin samples from Natau and Xitouxhai sections were found to average 137 °C (SD = 21.5 °C) in a range of 67–189 °C (N = 72, Figure 8c). Finally, saddle dolomites from the platform interior, slope, and basin-margin samples were found to have T_h from FIA’s in a range of 90–185 °C, average 141 °C (SD = 21.7 °C) (N = 61, Figure 9).

Salinities of the inclusions from the platform-interior replacement dolomite and associated dolomite cement for both the Lower and Middle Triassic Anisian strata using both Tm_ice and T_n measurements were found to average 9.7 wt. % NaCl equiv. (SD = 2.8 wt. %) in a range of 5.2–15.1 wt. % (N = 21, Figure 10a). From Tm_ice alone, the salinities were calculated to average 10.4 wt. % NaCl equiv. (SD = 1.2 wt. %) in a range of 9.2–10.4 wt. % (N = 3, Figure 10a).

Salinities of the inclusions in the replacement dolomites of the foreslope (Middle Triassic, Anisian) from the Tm_ice and T_n measurements were calculated to average 12.1 wt. % NaCl equiv. (SD = 3.83 wt. %), in a range of 2.3–18.6 wt. % (N = 30, Figure 10b). From Tm_ice alone, the salinities were calculated to average 12.5 wt. % NaCl equiv. (SD = 3.4 wt. %), in a range of 5.9–18.4 wt. % (N = 16, Figure 10b). Salinities of the inclusions in the replacement dolomites in the basin margin (Lower Triassic) using both the Tm_ice and T_n measurements were calculated to average 11.25 wt. % NaCl equiv. (SD = 3.2 wt. %), in a range of 3.4–16.4 (N = 24, Figure 10c). From Tm_ice alone, the salinities were calculated to average 10.9 wt. % NaCl equiv. (SD = 3.9 wt. %), in a range of 3.4–16.4 wt. % (N = 10, Figure 10c).

Salinities from the inclusions in saddle dolomite from the platform interior, slope, and basin-margin samples calculated from Tm_ice and T_n yielded an average of 14.1 wt. % NaCl equiv. (SD = 4.2 wt. %), in a range of 6.4 to 19 wt. % (N = 27, Figure 10).

T_h from the late-stage vein-filling calcite were found to average 137.7 °C (SD = 7.1 °C) in a range of 125 to 145 °C (N = 18; Figure 11a) whereas salinities in calcite calculated from Tm_ice average 4.4 wt. % NaCl equiv. in a range of 2.6 to 6.4 wt. % (N = 4; Figure 11b).

4.5. Oxygen, Carbon, and Strontium Isotopes

Oxygen and carbon isotope data for 74 dolomite samples (33 from the platform interior and 41 from the slope and basin margin) and 12 samples of vein-filling calcite are presented in Figure 12. The δ¹⁸O and δ¹³C values show two distinct populations for dolomite and calcite with the values for platform interior and slope-basin dolomites plotting in a cluster of similar values (Figure 12).

The δ¹⁸O values of platform interior replacement dolomites range from −7.15 to 2.14‰ and average −3.14‰ VPDB whereas the δ¹⁸O values of slope and basin margin replacement dolomites range from −7.68 to 0.75‰ and average −2.61‰ VPDB (Figure 12). δ¹³C of platform interior replacement dolomite ranges from −2.37 to 2.6‰ and average 1.42‰ VPDB and from slope to basin margin replacement dolomite range from 0.75 to 4.0‰ and average 2.34‰ VPDB (Figure 12). Two saddle dolomite samples analyzed include a basin-margin specimen (δ¹⁸O −1.34‰ and δ¹³C 1.49‰ VPDB; similar to values of other dolomite samples) and a platform-interior specimen (δ¹⁸O −10.88‰ and δ¹³C 0.41‰ VPDB) which has more negative δ¹⁸O values than the other dolomite samples (Figure 12). The δ¹⁸O values of the vein-filling calcite samples range from −18.4 to −5.9‰, average −12.78‰ VPDB, while δ¹³C ranges from −6.05 to 3.44‰, average −0.62‰ VPDB (Figure 12).

⁸⁷Sr/⁸⁶Sr ratios for 21 dolomite samples yield values ranging from 0.70767 to 0.70860 (Figure 13). The platform interior samples have ratios ranging from 0.70796 to 0.70834, whereas the slope samples range from 0.70799 to 0.70860, and the basin-margin samples range from 0.70768 to 0.70846 (Figure 13).

4.6. Minor and Trace Elements

Trace elements from individual carbonate phases using LA-ICP-MS reveal that calcite typically has higher Sr content (412–1565 ppm) than dolomite with concentrations ranging from 207 to 476 ppm, except sample NTU-15 which contains a Sr concentration of 1061 ppm (

Table 2. Trace element concentrations in dolomite and calcite diagenetic phases.

Replacement dol	Sr	Th	238U	207Pb	Li	B	Fe	Mn	Ce/Ce *
Sample	ppm	ppm	ppm	ppm	ppm	ppm	ppm	ppm
HY-23 (coarse)	478	2.31	7.49	7.69	10.01	15.83	4856	604	0.94
NTU-15	1061	0.35	0.93	2.15	1.36	5.25	883	80	0.31
NTU-62.5 (coarse)	268	0.2	2.85	1.31	0.97	5.8	685	269	0.42
NTU-209.7 (coarse)	286	0.93	2.81	3.18	2.86	23.78	2839	241	0.9
ST-139 (fine)	239	7.72	2.79	17.46	11.3	84.14	6573	644	0.9
Saddle dol	Sr	Th	U	207Pb	Li	B	Fe	Mn	Ce/Ce *
Sample	ppm	ppm	ppm	ppm	ppm	ppm	ppm	ppm
DP-03 inner zone	207	0.21	2.34	9.73	n/a	n/a	n/a	n/a	n/a
DP-03 outer zone	235	0.07	0.15	1.86	n/a	n/a	n/a	n/a	n/a
DP-11 inner zone	240	n/a	n/a	n/a	0.66	10.96	168	143	1.02
NTU-209.7	407	0.05	0.12	0.59	0.67	12.79	1049	217	0.92
Calcite	Sr	Th	U	207Pb	Li	B	Fe	Mn	Ce/Ce *
Sample	ppm	ppm	ppm	ppm	ppm	ppm	ppm	ppm
HY-23	536	0.96	0.43	0.31	0.49	3.47	618	278	0.81
NTU-15	1565	0.5	1.79	1.94	1.57	8.05	326	84	0.32
NTU-62.5 (tw)	352	0.09	0.72	2.33	0	2.77	16	77	0.16
NTU-62.5 (untw)	412	0.04	0.38	3.31	0	2.45	70	98	0.25

tw: twinned. untw: untwinned; * Ce anomaly

). Replacement dolomite and saddle dolomite phases also tend to possess elevated Fe concentrations (186 to 6573 ppm, average of 2436 ppm) in comparison to calcite (16–618 ppm, average of 257 ppm).

Replacement dolomite also has high concentrations of ²³⁸U with concentrations ranging from 2.79 to 7.49 ppm, except sample NTU-15 which has an anomalously low ²³⁸U of 0.93 ppm (Table 2). Saddle dolomite contains variable concentrations of ²³⁸U ranging from 2.34 ppm in the core of a saddle dolomite crystal in DP-11 to values of 0.15–0.12 ppm in the outer zones of the crystal in DP-11 and in NTU-209.5 (Table 2). Calcite tends to contain lower ²³⁸U concentrations (0.38–1.79 ppm), and Mn concentrations are variable in both dolomite and calcite, with higher average concentrations occurring in dolomite (Table 2).

Dolomite and calcite phases have large ranges in values for Li and B, with replacement dolomite tending to have higher concentrations of Li (0.97–11.3 ppm) and dolomite overall (including replacement and saddle phases) tending to have greater B concentration (5.25–84.14 ppm) compared to calcite (Li = 0.05–1.57 ppm; B = 2.45–8.05 ppm; Table 2). Replacement dolomite samples HY-23 and ST-139 that have notably high Li concentrations (10.01–11.3 ppm) also have high B concentrations (15.83–84.14 ppm; Table 2).

Rare earth element (REE) concentrations were cast as shale-normalized profiles for the samples measured from replacement dolomite, saddle dolomite, and calcite phases (Figure 14). Several samples, including both dolomite and calcite phases, exhibit a negative Ce anomaly and a profile of increased middle REE concentrations (Figure 14).

4.7. U-Pb Age Dates Using LA-ICP-MS

The U-Pb age dating of carbonate diagenetic phases was performed using the image mapping approach of [60] which allowed the collection of data within particular carbonate phases such as dolomite, saddle dolomite, and vein-filling calcite (Figure 15). All of the samples analyzed in Figure 15 have a Lower Triassic stratigraphic age except for NTU-209.7 which is from Lower Anisian strata (Figure 2d).

NTU-15 is a replacement dolomite with pixels selected on the basis of high U content and corresponding to replacement dolomite domains that exclude cross-cutting calcite veins (Figure 15a). The age date is 185 ± 38 Ma, indicating that the dolomite formed sometime from the Late Triassic, Norian to Jurassic.

NTU-62.5 and NTU-209.7 are both replacement dolomite, selected in maps based on high U concentration and excluding cross-cutting calcite veins, and in NTU-209.7, excluding a vug that contains saddle dolomite with low U concentration (Figure 15b,c). These replacement dolomite phases yielded ages of 255 ±1.7 Ma and 262 ± 4.2 Ma, respectively (Figure 15b,c). The age dates are Upper Permian within error, indicating a problem with the age dates (that is, the dolomite U-Pb age is older than the stratigraphic age of the rock; see Section 5.6).

DP-03CZ and DP-03A are euhedral dolomite cement crystals in which high uranium concentration was used to select pixels and age date particular zones within the crystals (Figure 15d,e). Age dates for these samples are 243.5 ± 5.9 Ma and 242 ± 6 Ma, respectively, both indicating that the dolomite precipitated sometime between near the end of the Early Triassic (Latest Olenekian) and the Late Triassic, Carnian.

For HY-23, a polygon was used to select pixels with relatively low U content within a calcite vein to provide an age date for calcite-infilling late-stage fractures (Figure 15f). The age date is 119 ± 15 Ma, indicating a Cretaceous age for the calcite.

5. Discussion

5.1. Modes and Relative Timing of Dolomitization from Field Relationships

The Lower and Middle Triassic platform interior and the slope and basin-margin facies of the Yangtze Platform are variably dolomitized whereas the platform margin facies escaped dolomitization. We interpret that the extensive early marine cementation of the Tubiphytes reef facies prevented the later permeation of dolomitizing fluids originating both in the platform interior and basin. Likewise, for the oolitic margin shoals that are present within the Lower Triassic Olenekian margin of the Yangtze Platform at Guanling, we interpret that early marine cementation locally prevented the permeation of dolomitizing fluids and prevented the dolomitization of the margin facies, while the interior and slope facies were massively dolomitized (Table 1; Figure 2, Figure 3 and Figure 4).

The dolomitization of the platform interior facies in backreef lagoon positions as well as the occurrence of dolomitized microbial–fenestral facies in peritidal cycles (Table 1; Figure 2 and Figure 3c) in the Anshun and Guanling formations are consistent with penecontemporaneous tidal flat dolomitization, e.g., [67,68,69]. The association of massive dolomite with stratiform evaporite dissolution breccia and evaporite crystal molds (Table 1; Figure 3a,b,e) in the Anshun and Guanling formations are consistent with the development of restricted evaporitic conditions in the platform interior which would be expected to produce magnesium-rich brines capable of driving syndepositional dolomitization in the lagoon, e.g., [8] or the evaporative reflux dolomitization of platform interior strata, e.g., [3,70,71].

In the slope to basin-margin facies, we observe a greater proportion of dolomitization in basinward sections, with the proportion decreasing up-slope toward the platform margin (Figure 2). We interpret this relationship to indicate that the dolomitization of the slope facies resulted from basinal fluids, likely expelled by the compaction of shales and siliciclastic turbidites of the Xinyuan and Biangyan formations during burial (Figure 1). The Xinyuan and Bianyang formations are exceedingly thick ranging up to 876 m and 2764 m, respectively, in the western Nanpanjiang Basin adjacent to the Guanling area; or up to a maximum turbidite thickness of 5220 m within the basin in Guizhou province including the stratigraphic equivalent Xuman Formation [33,72]. The compaction of these thick shale and sandstone basinal turbidite units could yield a large volume of basinal fluids capable of dolomitizing the volumetrically much smaller slope and basin-margin carbonate units (e.g., see Figure 1 and Figure 2).

Furthermore, the observation of the extensive dolomitization of pelagic carbonate beds, carbonate turbidites, and debris-flow beds at distances of up to 7 km basinward of the platform margin makes it highly unlikely that platform-derived fluids (e.g., evaporative brines) could reach that far into the basin [73], leaving basin derived fluids as the most viable explanation. Our interpretation is consistent with previous studies that have inferred dolomitization of slope and basin-margin facies by basinal burial brines, e.g., [2,74].

Further evidence for the dolomitization of the slope and basin-margin facies by burial fluids comes from dolomite fronts that cut across bedding and occur in bedding plane parallel and across-bedding fractures in distributions consistent with platformward migration of basin-derived dolomitizing fluids during burial (Table 1; Figure 4d–g). In foreslope debris-flow breccia beds, dolomitization changes in character from the massive replacement of breccia in positions farther from the platform margin to the selective replacement of the micritic matrix proximal to the margin (Figure 4a–c). Dolomite progressively diminishes up-slope, with very minor dolomitization of the matrix in breccias observed at the proximal margin and within the fractures and micritic elements of the most basinward part of the Tubiphytes boundstone of the reef (Figure 5a).

5.2. Paragenesis of Dolomite and Other Diagenetic Phases

Thin-section observations indicate that diagenesis began early in the depositional environment, especially in the platform margin facies where extensive marine cementation took place, and in the platform interior where early marine cementation, vadose cementation, and dolomitization took place (Table 1, Figure 16). Early marine cementation is particularly evident in the Tubiphytes boundstone that makes up the Middle Triassic Anisian margin of the Yangtze Platform. In outcrop, the boundstone appears very tightly cemented and contains large volumes of isopachous bladed marine cement and cement fans that fill nearly all framework porosity [33,35,36,75]. All the boundstone samples observed in the Guanling margin are tightly occluded with bladed marine cement (e.g., Figure 5a). The only recognizable porosity that remained unfilled is local irregular cavities in the reef framework that later were filled during a burial with coarse equant calcite, e.g., [75]. The Anisian Tubiphytes margin facies of the Great Bank of Guizhou is likewise extensively filled with marine cement [76,77]. The Lower Triassic, Olenekian oolitic margin facies are also undolomitized in the Guanling area at Doupengzhai, and in other areas of the Yangtze Platform and Great Bank of Guizhou where it is mapped as the Gujiao Formation, and where extensive marine cementation of the oolitic facies has been observed (Table 1) [37,46].

Early marine bladed calcite cement and micritic vadose cements also occur partially infilling fenestral and intergranular porosity in platform interior tidal flat facies (Figure 5b). The earliest dolomitization is interpreted to have occurred very shortly after deposition as penecontemporaneous dolomite in tidal flats, represented by peritidal cycles with dolomitized intertidal to supratidal cycle caps (Table 1; Figure 3c and Figure 16). The dolomite replaced microbial laminate, micrite, grains, and micritic vadose cements (Figure 5b). Similar selectively dolomitized cycle caps are common in the peritidal platform-interior facies of the Yangtze Platform and dolomitized clasts of intertidal to supratidal facies have been observed as reworked into the subtidal bases of overlying cycles demonstrating penecontemporaneous dolomitization [33,45].

Widespread massive dolomitization formed sometime after early marine diagenesis and before brittle fracturing in both the platform interior and slope to basin margin as evidenced by the thin-section observations (Figure 16). During burial, early syndepositional dolomites of the platform interior were extensively recrystallized and slope to basin-margin facies were replaced by anhedral to subhedral dolomite (Table 1, Figure 16) non-planar and planar-S dolomite types of [46]. The dolomitization of slope and basin facies shows a significant influence of depositional texture on the crystal size of subsequent dolomite. Fine-grained pelagic carbonate mudstone is replaced by finely crystalline, anhedral dolomite; whereas coarser-grained, packstone to grainstone carbonate in turbidites and debris-flow deposits are replaced by more coarsely crystalline anhedral to subhedral dolomite (e.g., Figure 5c,d). The influence of precursor depositional texture on dolomite texture can be explained by the greater permeability and greater space for dolomite crystal growth in coarser-grained, grainstone depositional facies, e.g., [78].

During the replacive dolomitization or shortly after, euhedral cements were precipitated, lining fenestral and large intergranular pores in breccia (Table 1; Figure 5b and Figure 16). The dolomite cement consists of equant crystals that are greater in size (up to 0.8 mm) than the associated replacement dolomite. Both in the platform interior and slope facies, the replacement dolomite is followed by the development of dissolution vugs and fractures (Figure 6c–e and Figure 16). Late-stage crystals of zoned euhedral dolomite and saddle dolomite were precipitated in the vugs and fractures (Table 1; Figure 6b–e and Figure 16). Saddle dolomite is typically interpreted to be a high-temperature phase precipitated during burial [79].

Subsequent fracture sets cut across the vugs and fractures and are filled with calcite (Table 1; Figure 6a,f and Figure 16). Most remaining pore space in fenestral, intergranular, and vuggy porosity is likewise infilled with calcite (e.g., Figure 5b). The calcite within fractures and other pores developed twinning by subsequent deformation (e.g., Table 1; Figure 5e, Figure 6a, and Figure 16). All the dolomite phases and fracture-filling twinned calcite phases are finally cross-cut by stylolites (Table 1; Figure 6f and Figure 16). Stylolites were one of the last diagenetic features to have formed during burial, only late-stage fractures and untwinned calcite filling fractures postdate the stylolites (Figure 16).

5.3. Burial and Thermal History

Burial history reconstructed from local and regional stratigraphic thicknesses indicate that the Lower to Middle Triassic strata reached minimum burial depths and temperatures of 2705 m (111 °C) to 4405 m (162 °C) by the end of the Late Triassic, Carnian, or 3305 m (129 °C) to 5505 m (180 °C) by the end of the Jurassic. These minimum burial depth estimates do not include compaction corrections, and the maximum depth included here is the depth of burial at the top of the Lower Triassic, Olenekian. The burial temperatures from these estimates are consistent with the microthermometry of fluid inclusions in the dolomite presented in this study.

Dolomite is cross-cut with fractures that are filled with calcite, which was subsequently twinned during deformation (e.g., Figure 5e and Figure 6a). The thickness and characteristics of calcite twinning can be used as a proxy for the temperature during deformation [64,80]. In this case, “thin twins” or type 1 calcite twinning [64] was dominant, observed as sharp lines and an average twin width of less than 2 µm indicating that deformation took place at <170 °C. The temperatures are consistent with the burial history and fluid-inclusion microthermometry results. The burial depths and temperatures are consistent with our interpretations that earlier penecontemporaneous tidal flat and evaporative dolomites in the platform interior were recrystallized during burial and that the basin-margin and slope facies were dolomitized by high-temperature brines that were expelled from the basin during burial.

5.4. Temperature and Salinity of Dolomitizing Fluids from Fluid-Inclusion Analysis

Fluid-inclusion petrography and its relationship to growth zoning suggest that the FIAs are primary to replacement and/or precipitation and are, therefore, representative of entrapment conditions during dolomitization. Evidence for the necking down of inclusions after a phase change was not observed cf. [81]. No correlation was found between T_h and inclusion size, which would be expected if stretching from internal pressures had occurred during later burial. Taken as a whole, the measured inclusions span a wide range of temperatures. However, analysis of individual FIAs reveals narrower temperature distributions (Figure 8 and Figure 9). Some FIAs have the majority of their T_h values falling within a 15 °C range (characterized as “consistent” by [81]), but others have a wider range of Th. Salinity data (Figure 10) also show similar patterns, with the values preserved in each FIA typically clustering in a narrow compositional range. Because the samples do contain fractures, a few cases of unrecognized entrapment of secondary inclusions may have occurred, especially those with outlying T_h or salinity values, but these probably represent <10% of the total inclusions. Therefore, our interpretation is that the fluid-inclusion microthermometry data reliably recorded dolomitization over a range of temperatures and salinities as dolomitization progressed. Each FIA recorded the conditions present at that specific place and time as the recrystallization of the existing dolomite and formation of new dolomite occurred.

The consistently high homogenization temperatures of primary FIAs in replacement dolomite, zoned dolomite cement, and saddle dolomite indicate widespread high-temperature dolomitization across the platform-to-basin transition (Figure 8 and Figure 9). Fluid-inclusion T_h measurements from replacement and saddle dolomite all have similarly high averages and ranges in temperatures from 121 °C (66–186 °C) in the platform interior to 125 °C (85–185 °C) along the slope, and 136 °C (67–189 °C) in the basin margin, indicating high-temperature dolomitization across the interior to basin-margin transect (Figure 8).

Salinity calculations from Tm_ice and T_n measured from primary FIAs in dolomite indicate that most dolomite crystals formed from brines denser than seawater (3.4 to 18.7 wt. % NaCl equivalent) with the vast majority of measurements indicating saline brines ranging from 8 to 16 wt. %, NaCl eq. (Figure 10). There was only one measurement of Tm_ice from dolomite indicating a salinity less dense than seawater (2.3 wt. %, Figure 10b, Dol-5a1). This outlier probably does not represent dolomitizing fluids and instead is an unrecognized secondary fluid inclusion from entrapment of later diagenetic fluid. Notably, the salinities calculated from Tm_ice in primary FIAs from late-stage calcite vein fills also include values less dense than seawater (values as low as 2.6 wt. % NaCl eq.; Figure 11). We interpret that the one low-density inclusion recorded in Dol-5A1 (Figure 10b) is a secondary inclusion that represents fluids responsible for the calcite precipitation.

If platform interior strata initially underwent penecontemporaneous dolomitization in tidal flats and evaporative brine dolomitization as we have interpreted from field relationships, then the implication from the high-temperature fluid inclusions in the interior dolomite is that the earlier penecontemporaneous and evaporative dolomites were recrystallized at high temperatures during subsequent burial, e.g., [82]. We did not find any evidence for single-phase inclusions in the dolomite, which would be indicative of dolomitization at Earth’s surface temperatures [81]. The absence of single-phase inclusions indicates widespread recrystallization and cementation during burial, thus eliminating earlier dolomite phases that would be expected to contain such inclusions. We acknowledge the limitations of fluid-inclusion analysis as minute inclusions that could be single phases may be too small to detect a vapor bubble or too small to have nucleated a vapor bubble [81]. However, in this study inclusions smaller than 1 µm were routinely observed in dolomite primary FIAs to have moving vapor bubbles, suggesting that unrecognized all-liquid aqueous inclusions are not present.

Although some platform systems have been interpreted such that evaporative fluids from the platform interior refluxed into and dolomitized foreslope strata, e.g., [83], field relationships in our study indicate encroachment of brines from the basin into the basin margin and foreslope. Further, the presence of an impermeable margin and the distal basin margin position of dolomite observed in the Guanling margin (up to 7 km basinward of the platform margin) indicates that this area was likely beyond the reach of evaporative brines. This interpretation is consistent with the variable-density groundwater flow modeling results showing that evaporative reflux brine is constrained in the platform interior and upper slope and does not migrate to the distal basin margin [84]. Therefore, we interpret the dolomite in the basin margin to have formed by high-temperature brines that originated in the basin and migrated into the basin margin and slope sediments at high temperatures during burial.

5.5. Geochemical Constraints on Origin of Dolomitizing Fluids

The dolomite phases from the platform interior, basin margin, and slope contain δ¹⁸O values clustered in a range from −7.68 to 2.14‰ VPDB. The range includes δ¹⁸O values that are both isotopically lighter and heavier than the δ¹⁸O values −5.5 to −1.9‰ VPDB expected from carbonate sediments precipitated in Triassic seawater [48] (Figure 12).

Values of δ¹⁸O in dolomite exceeding expected seawater values may be explained by evaporative enrichment in δ¹⁸O, owing to evaporative conditions associated with penecontemporaneous tidal flat dolomitization or with the generation of evaporative brines in platform interior lagoons, e.g., [85], albeit evaporative enrichment in such a setting would be expected to yield values of δ¹⁸O at least 2 to 4‰ greater than seawater, e.g., [20,86]. The Tm_ice and T_h measurements of primary inclusions are consistent with dolomite precipitation from hypersaline brines. The lesser degree of enrichment in δ¹⁸O and the spread, including δ¹⁸O more negative than that expected of Triassic seawater, can be explained by the thermal effect of dolomite recrystallization and precipitation at high temperatures from brines during burial, e.g., [85]. Ref. [27] interpreted a similar cause for Triassic dolomites on the Great Bank of Guizhou to the southeast in the Nanpanjiang Basin.

To test this hypothesis, the δ¹⁸O_DOL values were used in conjunction with the available fluid-inclusion T_h values for the same or neighboring samples to estimate the δ¹⁸O_FLUID of the dolomitizing fluid using equations from [87]. The estimated δ¹⁸O_FLUID values for the dolomitizing fluids of platform interior, slope-basin, and saddle dolomites all fall in the range of +4.0‰ to +10.1‰ VSMOW. These results are very similar to those reported by [27] for Lower Triassic dolomites in the Great Bank of Guizhou. Our interpretation is that the initial dolomitizing fluids were brines that became isotopically enriched through evaporation, which is further supported by calculated fluid-inclusion salinities and field evidence for evaporites. However, as mentioned above, the dolomite was recrystallized during burial, thus partially altering the δ¹⁸O signal, e.g., [82]. Basinal brines are commonly considered to be modified seawater connate fluids; thus, the isotopic enrichment of brines in the Nanpanjiang Basin is not surprising given evidence for evaporative conditions during the Triassic on the Yangtze Platform and Great Bank of Guizhou, as well as in the Sichuan Basin c.f. [27,33,88].

Calcite-infilling late-stage fractures contains significantly lower δ¹⁸O values and a broad range in δ¹³C in a field distinct from dolomite (Figure 12). Equations governing the fractionation of δ¹⁸O_CAL from δ¹⁸O_WATER are different than those for dolomite. Limited fluid-inclusion T_h data were available for calcite in this study (Figure 11), but similar calculations were performed to estimate the δ¹⁸O_FLUID using values from [89]. We used reasonable ranges that match fluid-inclusion data from this study, as well as similar data for calcite from the Great Bank of Guizhou [90]. For calcite precipitation temperatures between 110 °C and 150 °C, calcite isotopic compositions predict δ¹⁸O_WATER values that are between −1.9‰ and +8.0‰ VSMOW, respectively (average +0.5 to +3.9‰, respectively). Although the calcite values suggest δ¹⁸O_FLUID that may be somewhat less enriched than the dolomitizing fluid, this is consistent with the salinity data available at both locations in the Nanpanjiang Basin. Coupled with the timing of calcite precipitation near peak burial, these values are interpreted to represent the precipitation of the calcite by high-temperature saline fluids with a moderate to significant signal of isotopic enrichment, and some possible dilution by a meteoric contribution during tectonism and denudation.

⁸⁷Sr/⁸⁶Sr is widely used as a proxy for seawater composition for the precipitation of carbonate sediments and diagenetic phases because it does not fractionate under temperature, pressure, evaporative, or biochemical processes and instead reflects the source of depositional or diagenetic fluids [91,92]. The ⁸⁷Sr/⁸⁶Sr ratio in seawater is well characterized over geologic time [93,94,95]. The Triassic seawater ⁸⁷Sr/⁸⁶Sr curve reconstructed from the conodont samples collected adjacent to the Great Bank of Guizhou [66] is shown in Figure 13 with the ⁸⁷Sr/⁸⁶Sr ratios from the analyses of dolomites from the platform interior, slope, and basin margin of the Yangtze platform margin at Guanling. The seawater curve shows lower ⁸⁷Sr/⁸⁶Sr values (0.7068–0.7074) in the Upper Permian followed by a steep rise in ⁸⁷Sr/⁸⁶Sr in the Lower Triassic Induan and Olenekian (0.7074 to 0.7085) and then Middle Triassic Anisian ⁸⁷Sr/⁸⁶Sr values ranging from 0.7078 to 0.70855). Most dolomite samples contain ⁸⁷Sr/⁸⁶Sr that are also consistent with Triassic seawater, although a few samples contain ⁸⁷Sr/⁸⁶Sr ratios higher than seawater likely due to a radiogenic contribution (Figure 13). Notably, the four samples that have a ⁸⁷Sr/⁸⁶Sr ratio elevated above Triassic seawater values all come from the basin margin, whereas all of the platform interior samples have ⁸⁷Sr/⁸⁶Sr ratios that overlap with seawater (Figure 13). This pattern is consistent with our interpretation of dolomitization by evaporatively concentrated seawater in the platform interior and by basinal brines in the basin margin. We interpret that the seawater ratio was imparted to the dolomite through the formation of dolomite in the platform interior by evaporatively concentrated seawater and that the recrystallization of the dolomite occurred at high temperatures in a rock-dominated system, thus retaining the original seawater ⁸⁷Sr/⁸⁶Sr signal. The high-temperature dolomitization of the slope and basin-margin facies by basinal brines that were essentially seawater buried in the basinal siliciclastic turbidite formations, and expelled into the basin margin and slope carbonates during compaction, yielded a ⁸⁷Sr/⁸⁶Sr signal similar to Triassic seawater but with a radiogenic contribution resulting from the diagenetic interaction of brines with siliciclastic basinal sediments during compaction and fluid migration.

The dolomite from the interior, slope, and basin margin sections have relatively high Fe and Mn concentrations, likely indicating precipitation in a reducing burial environment (e.g., Fe > 500 ppm, Mn > 200 ppm; Table 2) [2,96,97]. Elevated Li and B in some samples may indicate concentration by seawater evaporation or in burial brines cf. [98]. REE patterns have a negative Ce anomaly similar to seawater (Figure 14) [99,100]. This likely resulted from the burial precipitation of dolomite in a rock-dominated system wherein the signal was inherited from early marine dolomite or limestone, e.g., [101,102,103]. Other replacement dolomite or saddle dolomite samples have either a flat “clay signature” or a slight positive Eu anomaly (Figure 14) [99,100], suggesting the modification of the REE signature by input of fluids that have interacted with silicates, e.g., [19,101,104].

5.6. Absolute Age Constraints on Dolomitization and Calcite Diagenesis

The age dates of replacement dolomite and zoned dolomite cement indicate that the dolomite formation occurred during burial (Middle Triassic to Jurassic) which is consistent with the high-temperature recrystallization of earlier syndepositional replacement dolomite in the platform interior and the dolomitization of slope facies during burial and later dolomite cementation as indicated by fluid-inclusion homogenization temperatures (Figure 15 and Figure 16). The NTU-15 sample is replacement dolomite from Lower Triassic, Olenekian strata from the basin margin that yielded a U-Pb age date of 185 ± 38 Ma, with an error range placing dolomitization between Late Triassic Carnian to Jurassic. The DP-03 samples are zoned dolomite cements from the Lower Triassic, Olenekian strata of the platform interior that yielded U-Pb age dates of 243.5 ± 5.9 Ma and 242 ± 6 Ma, with errors placing dolomitization between the end of the Lower Triassic, Olenekian to the Late Triassic Carnian.

Replacement dolomites in the samples NTU-62.5 and NTU-209.7 yielded Late Permian ages (255 ±1.7 Ma and 262 ± 4.2 Ma) that are too old to be possible given that the samples originate from Lower Triassic Olenekian and Middle Triassic, Anisian strata, respectively (Figure 15b,c). Stratigraphic positions greater than 50 or 200 m above the Permian–Triassic boundary in strata that are devoid of Permian fossils make it highly unlikely that the stratigraphic age is incorrect. A more reasonable explanation is that the dates were skewed to older ages by the incorporation of an inherited clay component. The samples both come from the basin-margin strata, where a siliciclastic clay component is common in pelagic sediments. The LA-ICPMS elemental maps of NTU-62.5 and NTU-209.7 show scattered spikes in the Si content within the replacement dolomite (with silica ranging up to 1.4%) possibly representing clay inclusions.

The age date for the calcite vein (119 ± 15 Ma) is Cretaceous (Figure 15f). This young age is consistent with the observation that calcite occurs in late-stage fractures that cut across all the stages of dolomitization (Figure 16). A Cretaceous age is also consistent with the interpretation that the calcite-filled fractures formed around the same time as the major structural deformation (major folds and faults) and uplift in the region as evident in regional geologic maps, e.g., [46,105].

5.7. Dolomitization Model and Implications

Our model for the dolomitization and diagenetic evolution of the Yangtze Platform and its basin margin (

Table 3. Model for the dolomitization and diagenetic evolution of the Yangtze Platform at Guanling.

Diagenetic History	Interpreted Environments	Supporting Evidence
1. Sedimentation	The spectrum of carbonate depositional environments across platform interior to basin margin	See the Geologic Setting Section and precursor carbonate facies in the Petrography Section.
2. Syndepositional cementation	Subtidal marine cement in reef, vadose cement on tidal flats.	Bladed marine cement infill of reef porosity, vadose meniscus micritic cements.
3. Penecontemporaneous dolomite	Subaerially exposed tidal flats, evaporatively concentrated seawater dolomitizes tidal flat surface.	Preferential dolomitization of tidal flat cycle caps in platform interior.
4. Replacement dolomite formed by evaporatively concentrated seawater	Hypersaline lagoon, evaporatively concentrated seawater dolomitizes interior lagoon sediment in situ or via evaporative reflux dolomitization.	Preferential dolomitization of the interior lagoon (not landward or margin facies), association with evaporite crystal molds and evaporite solution collapse breccia, 87Sr/86Sr consistent with seawater source, δ18O consistent with evaporated seawater.
5A. Burial dolomitization (recrystallization of earlier dolomite phases), and associated dolomite cement	High-temperature burial conditions (up to 4.5 km burial, by the end of Triassic).	Fluid inclusion Th (avg. 115 °C, 66–186 °C) containing entrapped brines (avg. 9.7 wt. %, 5.2–15.1 wt. % salinity). Trace elements consistent with reducing conditions. REEs consistent with authigenic carbonate signal reflecting recrystallization in the rock-dominated system.
5B. Replacement dolomite and associated dolomite cement by basin-derived brines at basin margin and slope	High-temperature burial conditions (up to 4.5 km burial, by the end of Triassic), dolomitizing fluids migrate from the basin into the basin margin to foreslope; fluids expelled from compacting basin siliciclastic shale and turbidite units of Xinyuan, Xuman, and Bianyang formations.	Spatial distribution of replacement dolomite from basin margin to slope, fluid inclusion Th (avg. 127 °C, 85–190 °C) containing entrapped brines (avg. 11.7 wt. %, 5–19 wt. % salinity), U-Pb age (Late Triassic-Jurassic) consistent with burial timing. Trace elements consistent with reducing conditions. 87Sr/86Sr and REE signal consistent with the interaction of fluids with basinal siliciclastic facies.
6. Fracturingand dissolution (vugs)	High-temperature burial conditions.	Fractures and vugs petrographically postdate earlier dolomite phases.
7. Saddle dolomite and zoned euhedral dolomite cement precipitation in fractures, vugs and intergranular porosity.	High-temperature burial conditions (up to 3.6 to 5 km burial by end of Jurassic).	Saddle dolomite is considered a high temp phase [79], fluid inclusion Th (avg. 141 °C, range of 90–185 °C) containing entrapped brines (avg. 14.1 wt. %, 6.9–19 wt. % salinity).
8. Additional generations of fractures, calcite vein fill of fractures	High-temperature burial conditions.	Calcite-filled fractures petrographically postdate all the dolomite phases. U-Pb age date indicates Cretaceous age.
9. Development of deformation twinning in calcite	High-temperature burial conditions.	Calcite twins of type 1 (thin twins, >40 twins/mm) indicate deformation at temperatures up to 170 °C [64] deformation consistent with regional deformation.
10. Stylolitization	High-temperature burial conditions.	Petrographically postdate dolomite and twinned calcite phases.

) interprets that a large proportion of the platform-interior facies were initially dolomitized in penecontemporaneous tidal flats (e.g., sabkha type or penecontemporaneous dolomitization by the evaporative concentration of seawater on the tidal flat surface) and via evaporatively concentrated fluids in interior lagoons (either syndepositional dolomitization by the evaporitic concentration of brines in the lagoon or evaporative reflux dolomitization). Syndepositional dolomite was later recrystallized at high temperature during burial. We suggest that vast areas of stratiform dolomite in the interior of the Yangtze Platform and in the interiors of isolated platforms in the Nanpanjiang Basin are likely formed by a combination of these mechanisms (e.g., Figure 1).

We interpret that the basin margin and slope facies of the Yangtze Platform at Guanling were also dolomitized during burial by high-temperature brines that were essentially modified seawater, mg-bearing brines that were deposited with the basinal sediments and expelled during the compaction of siliciclastic turbidite facies and permeated into the basin margin and slope facies (Table 3; Figure 2b). Constraints on the burial history and the elevated temperatures of later diagenetic phases such as saddle dolomite point to continued diagenetic evolution during burial at high temperatures (Table 3, Figure 16).

This study provides a useful example of stratiform dolomitization of vast platform areas that were overprinted by later burial recrystallization. Moreover, the findings of this study are important as analogs for the spatial distribution of dolomite in other carbonate platform systems in which the margin and upper slope porosity were occluded by marine cement and remained undolomitized, whereas the platform interior and/or slope-to-basin margin succession are massively dolomitized and contain secondary porosity such as the Permian Capitan system [106,107], Triassic Latemar Platform [108], Carboniferous Tengiz Platform [108], and the lowermost Fammenian of the Lennard Shelf [109]. The spatial patterns of the dolomitization and distribution of associated secondary porosity are important for the development of reservoirs for hydrocarbons, carbon sequestration, or hydrogen storage.

One of the challenges in solving “the dolomite problem” probably lies in the fact that many dolomites form by multiple overprinted mechanisms including both Earth surface and deep burial processes, e.g., [22,23,24,25,26,27,28,29,30,31,32]. A combination of early disordered protodolomite formation at Earth surface conditions followed by stabilization to ordered dolomite during burial may explain widespread massive dolomite. Recent studies by [4,110] used the analysis of dolomite stoichiometry and cation ordering across the Phanerozoic to support the hypothesis that recently deposited and young dolomites (<30–40 Ma) exhibit a lower degree of cation ordering, and that stoichiometry and cation ordering increases with time, thus providing a mechanism of the conversion of syndepositional seawater and evaporative dolomites initially formed at Earth surface temperatures in carbonate platforms to become ordered dolomite at higher temperatures during burial.

The Yangtze Platform and isolated platforms of the Nanpanjiang Basin of south China provide a valuable example of widespread massive stratiform dolomite formed by superposed mechanisms, starting with the platform interior dolomitization by evaporatively concentrated seawater, followed by the recrystallization of interior dolomite and the dolomitization of the slope to basin-margin by basinal fluids at high temperatures during burial. We suggest that many massive stratiform dolomites in the rock record formed through a combination of penecontemporaneous dolomitization from seawater or modified seawater in a platform interior lagoon and tidal flat environments and later recrystallization at elevated temperature during burial.

6. Conclusions

Multiple overprinted mechanisms of dolomitization are interpreted to have contributed to massive dolomitization of the platform-interior and slope-to-basin-margin facies of the Yangtze Platform. Early marine diagenesis, especially the extensive marine cementation of the platform margin facies, impacted the pattern of dolomitization by preventing dolomitizing fluids from permeating the margin. The presence of preferentially dolomitized microbial laminate caps of peritidal cycles and the association of massive dolomite in the platform interior with evaporite nodules and evaporite dissolution breccia supports platform top dolomitization by penecontemporaneous evaporatively concentrated seawater at Earth surface conditions. Strontium isotope ratios similar to Triassic seawater, highly positive δ¹⁸O_FLUID values, and the elevated salinity in primary fluid inclusions in the dolomite are consistent with the retention of a signal of dolomitization from evaporatively concentrated seawater.

Consistently high homogenization temperatures (86–189 °C) of primary fluid inclusions in dolomites indicate the high-temperature recrystallization of the platform interior dolomite, as well as the formation of dolomite in the slope and basin-margin facies over a range of temperatures during burial. The presence of late-stage saddle dolomite in both the platform interior and slope-to-basin-margin successions also indicates high-temperature precipitation during burial. The U-Pb age dates of dolomite in the interior and slope to basin margin place dolomitization during Triassic to Jurassic burial, and the reconstruction of burial history produces depths and temperatures consistent with fluid-inclusion homogenization temperatures and other geothermometric proxies.

We interpret the slope to basin-margin facies to have been partially dolomitized by hot brines expelled from basin strata during compaction on burial. The massive dolomitization of basin-margin facies up to 7 km from the platform margin and the presence of an impermeable margin make it unlikely that these dolomitizing fluids originated on the platform top. This interpretation is supported by the progressive decline of dolomitization from the basin-margin up-slope toward the platform margin. Breccia beds are pervasively dolomitized in the basin margin but are only partially dolomitized with matrix-selective dolomitization in the proximal slope. Dolomitization fronts occur along a bedding plane and cross-cutting fractures in pelagic carbonate mudstone indicating post-lithification incursion of dolomitizing fluids during burial. Samples with Sr isotope ratios more radiogenic than expected of Triassic seawater solely originate from the basin-margin facies, indicating that dolomitizing fluids from the basin contain radiogenic Sr, possibly due to interaction with siliciclastic basin sediments.

The Yangtze Platform provides an excellent example of a system in which massive dolomitization of the platform interior and slope to basin margin resulted from multiple superposed dolomitization mechanisms including dolomitization by near-surface, evaporatively concentrated brines as well as high-temperature recrystallization and slope dolomitization by high-temperature basinal fluids during burial.