DM is commonly present throughout the upper Changping Formation and consists of muddy (mostly approximately 10 μm crystals) and generally non-planar-anhedral (nP-A) dolomite (Figure 4A). It commonly consists of light grey, parallel laminated beds, sometimes with massive structure. These DM samples have a red to dark red CL colour (Figure 4B) and low degrees of order (between 0.47 and 0.62, shown in Supplementary Appendix S1), indicating rapid crystal growth from the intertidal and subtidal zones (Goldsmith and Graf, 1958; Hood and Wallace, 2015). Spherical dolomite is observed (Figure 4C), commonly associated with mD or ssD of microbial-like textures in the EHST (Figure 4D), representing microbial induction (You et al., 2011).

mD is mainly represented by the EHST, with low abundance in the LHST. It consists of very fine crystalline (mostly 10–20 μm crystals) and mostly nP-A to planar-euhedral (P-E) dolomite (Figure 4E). Sedimentary structures are well-preserved in mD, such as unattached spherical structures in oncolites (Figure 4D), laminated structures in stromatolites (Figure 4E), and blotchy structures in thrombolites (Figure 4F). The mD exhibits a CL colour similar to that of DM, and a narrowly distributed degree of order, varying from 0.50 to 0.55 (Supplementary Appendix S1), indicates an irregular Mg²⁺ arrangement in the crystal lattice influenced by microbial metabolism (You et al., 2011; Baniak et al., 2014).

ssD primarily occurs in the whole upper Changping Formation. In the EHST, the degree of dolomitization is lower, about 30%, and the ssD can account for about 70% of the EHST dolomite. While in the LHST, the degree of dolomitization can be as high as 70%, and 80% of the dolomite can be ssD.

In the EHST, ssD and mD are in gradual contact (Figure 4G) on the hand specimen, and the fine to medium crystalline (mostly approximately 50 μm crystals), P-E and planar-subhedral (P-S) ssD display massive structures and obliterative microbial-like textures (Figure 4H). This blurred-ring dolomite has a dark red CL and medium-to high-order degree values with a narrow distribution (approximately 0.69–0.78, shown in Supplementary Appendix S1).

In the LHST, ssD features a foggy centre and bright edge (Figure 4I). Its noticeable ring effect occurs as an orange-red CL colour (Figure 4J), and the degree of order is widely distributed and high (approximately 0.75–0.89, shown in Supplementary Appendix S1). These two sets of high values may indicate lattice structure adjustment during the burial period (Machel, 2014), yet different zonal characteristics in CL images reveal distinct dolomitization processes in the upper Changping Formation.

MCD displays massive and saccharoidal structures and light grey to milky white colouration in the specimen (Figure 4K), and is primarily present in the LHST, particularly near the top karst surface with breccias (Figure 4L), but is not associated with the spatial distribution of fault locations. These well-developed paragenetic breccias suggest the impact of atmospheric water. Microscopically, MCD consists of medium to coarse crystalline, P-S, and nP-A dolomites, whose sizes range between 50 and 300 μm (Figure 4L), with bright red colour and an even bright red and yellow colour under CL (Figure 4M). A high degree of order, widely varying from 0.76 to 0.91 (Supplementary Appendix S1), corresponds to multiple adjustments of dolomitization (Jiang et al., 2016; Bai et al., 2021a).

Milky white, structureless SD is mainly observed in the vicinity of fractures in the EHST, outside of which SD coexists with nearby ssD (Figure 4N). SD comprises coarse crystalline, nP-A dolomite (crystal diameters ranging from 300 to 1,500 μm), and mostly undulose extinction (Figure 4O), and shares a similar CL colour with MCD. The degree of order is widely distributed with high values (approximately 0.61–0.79) (Supplementary Appendix S1), which may imply incomplete and rapid dolomitization (Huang et al., 2006; Machel, 2014; Bai et al., 2021a).

Fluid inclusions are commonly used for dolomitization fluid property analysis, and FIAs have been defined following the rules of Goldstein and Reynolds (Goldstein and Reynolds, 1994). More than 90% of FIAs homogenization temperature (T_h) data from a given paragenetic type of mineral fall within a range of less than 10%–15%. Therefore, it is unlikely that thermal re-equilibration has occurred (Goldstein and Reynolds, 1994).

The results show that only aqueous single-phase saltwater inclusions approximately 2 μm in diameter are observed in DM and mD. This means that the formation temperature was lower than 50°C (Goldstein and Reynolds, 1994; Jiang et al., 2016). Fewer pure liquid inclusions are found in the EHST ssD than in the LHST ssD, probably suggesting less surface seawater dolomitization transformation in the EHST. These samples have narrow, bimodal T_h ranges of 60°C–80°C and 100°C–120°C near the faults (Figure 5A). Correspondingly, salinity data of the normal ssD derived from ice-melting temperatures are predominantly between 8 and 18 wt%. Under the influence of fault fluids, the salinity can reach approximately 14–32 wt% (Figure 5B). Many two-phase, liquid–gas inclusions are observed in the LHST ssD, showing an equally low and narrow T_h range from 60°C to 80°C, with a lower salinity range from 6% to 20%. The ssD approaching the faults grew at higher temperatures and salinities (approximately 90°C–120°C and 14 to 28 wt%). A few LHST samples near the top show the effect of freshwater, with T_h and salinity values of approximately 60°C–70°C and 6 to 8 wt%, respectively (Figures 5A, B). The SD values range from about 100°C to 200°C, and have relatively high salinities of approximately 14–30 wt%. However, the T_h (80°C–180°C) and salinity data (8–24 wt%) of the MCD samples both have wide ranges (Figures 5A, B), mainly indicating multiple fluids in the dolomitization process.

Two phases of oil and gas inclusions can be identified. Phase I is mainly developed in the LHST ssD (frequency of oil inclusions = 5%, hereafter FOI), whose inclusion component is mainly bitumen (Figure 5C), and phase II in SD and EHST ssD is associated with calcite (FOI = 8%), dominated by bitumen and CH₄ inclusions (Figure 5D). Distinct inclusion characteristics correspond to the dual hydrocarbon discharge events described previously.

The ground temperature gradient in the study area is 25°C/km, and the surface temperature is 20°C (Yan, 2019). Combining the homogeneous temperature and salinity, the capture depth and dolomite formation time are determined, with two accumulation events separating the overall burial history into three phases: 1) DM and mD have burial depths <1,200 m and were formed approximately during the (pene) contemporaneous period. 2) ssD has formation depths of 1,600–2,400 m and may have originated from moderate salinity fluids in the shallow-middle burial period. 3) The burial depths of MCD and SD are 2,400–6,400 m and 3,200–7,200 m, respectively, countering the middle-deep burial period. The “broad-peaked” temperature and salinity of MCD may indicate the influence of a wide range of fluids with different properties.

REE parameters and patterns have good referential significance for dolomitization fluid determination (Bau and Dulski, 1965; Brand and Veizer, 1980; Bau, 1991; Muhammad et al., 2020); for example, δCe and δEu values can be used to distinguish seawater from meteoric water and a low-temperature dolomitization environment from a high-temperature environment (Banner et al., 1998; Bau and Alexander, 2006). Here, the REE distribution relationships and other correlative contents of dolostone, including associated limestone, were analysed (Figure 6; Supplementary Appendix S2). The analysis shows that all the limestone, DM, and mD exhibit nearly flat distribution characteristics of light REE (LREE) enrichment and heavy REE (HREE) loss (Figure 6A). Correspondingly, negative anomalies of δCe and δEu (<1) both indicate that those samples are mainly of seawater origin (Supplementary Appendix S2) (Azmy et al., 2011).

Some ssDs in the EHST display characteristic left-leaning patterns and are typical of seawater genesis (Kawabe et al., 1998) (Figure 6B). Samples associated with fractures and veins have δEu values greater than 1.15 (Supplementary Appendix S2). Because Eu²⁺ can enter lattices affected by hydrothermal fluids and replace internal Ca²⁺, Eu features a positive anomaly (>1) in such circumstances (Alibo and Nozaki, 1999; Kucera et al., 2009; Swart, 2015). Therefore, these data suggest a representative hydrothermal environment.

The ssD in LHST exhibits a similar distribution pattern compared to the marine carbonates (Figure 6B). However, compared to the EHST samples, the LHST samples show a reduced enrichment in LREEs and an increased enrichment in HREEs. In addition, the ssD in the vicinity of hydrothermal-genetic veins has a low positive anomaly of less than 1.1 (Supplementary Appendix S2). Therefore, this type of dolomite may have been modified by fluids other than seawater and hydrothermal liquids (Morad, 1998; Morad et al., 2010). Three Ce anomalous samples, including DH-8-1, DL-26–1, and DS-22, are identified (Supplementary Appendix S2). They have δCe values slightly greater than 1, which may indicate the influence of atmospheric freshwater (Kucera et al., 2009; Swart, 2015). The remaining two samples, which are far from the fault and unmodified by hydrothermal fluids and freshwater, have high overall REE contents (Supplementary Appendix S2) (Haas et al., 1995; Friedman, 2007; Morad et al., 2010; Azmy et al., 2011; Kareem et al., 2021).

MCD and SD are common hydrothermal dolomite crystal forms in petrology, yet atypical REE characteristics (Figure 6C) shed light on the lack of associated hydrothermal minerals (Figures 3, 4) (Shields and Stille, 2001). Their flat REE distribution patterns are similar to those of limestone, indicating the effective influence of seawater-related diagenetic fluids. Even though Eu in some samples shows different degrees of enrichment, compared with typical magmatic rock values (mainly w (Sc) ≈ (5–10) × 10⁻⁶; w (Th) ≈ (2–50) × 10⁻⁶; w (Hf) ≈ (2–5) × 10⁻⁶; and w (Zr) ≈ (50–200) × 10⁻⁶, given in Budd, 1997; Huang, 2014; Hao et al., 2020), crustal elements are always present at less than 1.00 × 10⁻⁶. Only one MCD sample has a Th value of 2.77 × 10⁻⁶, which is in the normal range (Supplementary Appendix S2). As this sample is close to the unconformity and has the highest REE values among all the samples, the influence of terrigenous material and atmospheric freshwater on REE values should not be ignored, which could explain the relatively low Eu anomaly values in the MCD compared to the SD in the lower strata (Supplementary Appendix S2). However, we do not observe positive Ce anomalies in these atypical hydrothermal dolomites, unlike in the ssD, which may indicate that the hydrothermal concentrations can mask the presence of meteoric water.

The relative contents of the major compounds MgO and CaO are effective parameters used to assess the dolomitization mechanism and degree of metasomatism (Zhao and Jones, 2012a; b). The results (Figure 7A; Supplementary Appendix S3) show that the ratios of MgO/CaO in DM and mD are primarily concentrated between 0.665 and 0.710, close to those in theoretical dolomite, CaMg(CO₃)₂ (0.714, according to Qian et al., 2019). Some DM samples were affected by terrigenous components, and their ratios reached 0.728. As shown in Figure 7A, DM and mD mainly feature typical positive correlations, indicating that Ca²⁺ and Mg²⁺ can be added into the dolomite lattice synchronously. This feature is typical of (pene) contemporaneous dolomitization in an open environment (Kırmacıa et al., 2018).

The ssD ratios in the EHST are lower than 0.400 (Figure 7B; Supplementary Appendix S3), which, combined with the overall low degree of dolomitization in this tract, suggests inadequate dolomitized fluids. Correspondingly, the negative MgO-CaO correlation indicates that as the process of dolomitization progresses, the amount of Mg²⁺ in the diagenetic environment gradually decreases and, therefore, produces restricted dolomitization conditions (Kaczmarek and Sibley, 2011). The ssD ratios in the LHST have a wide range from 0.446 to 0.720 (Figure 7B; Supplementary Appendix S3), probably representing a rock-forming environment between closed and open conditions. These ratio values are mainly arranged in a dual-correlated manner (Figure 7B), indicating multiple dolomitized settings.

SD yields lower values (mainly <0.35), and the correlation is strong but negative (Figure 7C; Supplementary Appendix S3). These values suggest limited Ca²⁺ supplements in a restricted burial environment (Vahrenkamp and Swart, 1990). The MgO/CaO ratios of MCD feature a broad peak characteristic (0.239–0.698) (Supplementary Appendix S3). The low values rank negatively (Figure 7C), indicative of a confined diagenetic environment. Two points of high values (>0.60) are evidence of open dolomitization conditions with sufficient fluid exchange transformed by faults and cracks (Tortola et al., 2020).

Trace elements Fe and Mn are sensitive to fluctuations in the dolomitization environment because these two elements are more likely to enter the dolomite lattice than Sr and Ba in the buried environment, which can reflect higher values in the buried dolomite than in the oxidized, (pene) contemporaneous dolomite (Hiatt and Pufahl, 2014; Huang, 2014). A relatively strong regularity can be observed in Supplementary Appendix S3, where the DMs and mDs yield minimal Fe and Mn values down to approximately 400 and 40 on average, indicating near-surface oxidation dolomitization (Al-Aasm, 2000; Al-Aasm and Crowe, 2018). Notably, diagenesis related to the weathering crust has increased the values of nearby samples.

The ssD values are moderate and have a wide range, showing the transitional formation (Supplementary Appendix S3). The MCD and SD display relatively broad values of multiple dolomitization effects, with Fe contents ranging between 500 and 1700 and Mn contents ranging between 12 and 162, respectively (Supplementary Appendix S3). However, Fe and Mn values of some samples are significantly lower than those of other buried dolomites (mostly Fe > 1,000 μg/g, Mn > 100 μg/g, Budd, 1997). In addition, strongly reduced burial dolomitization is indicated only when Fe >1,000 μg/g and Mn > 10–100 μg/g. Therefore, MCD and SD are not totally derived from typical burial diagenesis (Budd, 1997; Hiatt and Pufahl, 2014).

Stable carbon and oxygen isotopes are commonly used as indicators of palaeo-oceanic environments, rock-formation intensity analysis, and tracers of carbonate fluid origins (Carpenter et al., 1991). The carbon and oxygen isotopes in limestone samples display a relatively narrow range of values, namely, between 1.3‰ and 1.6‰ for δ¹³C and between −6.9‰ and −7.1‰ for δ¹⁸O, which are largely within the scope of contemporaneous Cambrian seawater calcite (approximately −2.7‰ < δ¹³C < 1.6‰ and −7.0 < δ¹⁸O < −10.0‰, according to Huang, 2014; William, 1989) (Figure 8; Supplementary Appendix S4).

On the basis that the δ¹⁸O_PDB value of seawater dolomite is 1.5‰ ∼ 3.5‰ higher than that of calcite (William, 1989; Huang, 2014), the δ¹⁸O value of marine dolomite in the upper Changping Formation should be between −5.4‰ and −3.6‰ (Figure 8; Supplementary Appendix S4). With this reference, we explore the fluid characteristics of various dolomites.

The DM and mD values are mainly within the range of those of contemporaneous seawater dolomite. Some DM values are close to the mD values, implying genesis similarities during the (pene) contemporaneous period. Microbial metabolism and methane oxidation can cause a decrease in δ¹³C values (Jiang et al., 2016; Jiang et al., 2019), e.g., δ¹³C values for DM can be concentrated at approximately 1.8‰, while those of mD are as low as approximately −1.5‰ (Figure 8; Supplementary Appendix S4).

The ssD of the EHST and LHST show distinct differences in carbon and oxygen isotopes. The narrow range of EHST samples reflects simple diagenesis. Its overall extent is closer to that of the limestone than to those of the DM and mD. This pattern may indicate that the less dolomitized EHST stratum has retained seawater-forming limestone features. Such low values of oxygen isotopes could also reflect burial genesis. The ssD of the LHST, with its wide range of carbon and oxygen isotopes, may indicate multiple geneses (Figure 8; Supplementary Appendix S4).

The δ¹⁸O values for the MCD and SD are similar to those of seawater, suggesting that they were influenced by residual seawater. However, the δ¹³C values for some MCD and SD are significantly positive, essentially above 2.0–2.5‰ (Figure 8; Supplementary Appendix S4). This is because high-temperature hydrothermal fluids could cause low oxygen contents in seawater, leading to increased carbon burial rates and consequently high δ¹³C values (Ahmad et al., 2021; Mansurbeg et al., 2021).

The Keith & Weber salinity index formula, Z = 2.048 (δ¹³C + 50) + 0.498 (δ¹³O + 50), has been applied mainly to identify the salinity of dolomitization fluids (generally Z > 120 for seawater diagenesis and Z < 120 for freshwater diagenesis) (Keith and Weber, 1964). Combined with the δ¹³C and δ¹⁸O values, it is suggested that mD was basically formed in a typical seawater environment, and its Z value is mainly stable at approximately 120. This result illustrates that microorganisms can form dolomite in a normal seawater environment. DM is likely influenced by both regular and saline seawater, with Z values between 127 and 129 (Supplementary Appendix S4).

Excluding the effect of hydrothermal fluids, the EHST ssD, which retains the characteristics of seawater, has the highest Z values between 129 and 131, demonstrating the presence of concentrated brine during the burial period. Some of the values of LHST ssD samples near the stratigraphic interface are below 120, showing freshwater influence (Supplementary Appendix S4).

The MCD and SD have Z values as high as 130, indicating the interaction of high-salinity hydrothermal fluids with stratum brines. Sample DK-9-1B, which has values as low as 123, may have been affected by atmospheric freshwater (Supplementary Appendix S4).

The composition of Sr isotopes in seawater is mainly controlled by terrestrial Sr-rich fluids and mantle-derived Sr-poor fluids and has been globally homogenized over a certain geological period (Burke et al., 1982; Denison et al., 1997). Because this fractionation is not affected by biology, salinity, temperature and pressure, the ⁸⁷Sr/⁸⁶Sr values may reflect dolomitization fluid properties and the environment (Burke et al., 1982; Denison et al., 1997). As shown in Supplementary Appendix S4, the ⁸⁷Sr/⁸⁶Sr values in the Changping Formation are largely within the range of Cambrian seawater (0.7087–0.7094, according to Burke et al., 1982; Huang, 2014), with a strong regularity, namely, mD < DM < MCD < SD < ssD (EHST) < ssD (LHST) (Supplementary Appendix S4), suggesting a gradual weakening of the influence of open seawater dolomitization systems (Burke et al., 1982; Carpenter et al., 1991). In addition, some ssD, MCD and SD samples have higher values with the influence of radioactive strontium in hydrothermal and Sr-rich atmospheric fluids (Figure 9; Supplementary Appendix S4).

Dolomite leads to higher δ²⁶Mg values (Mavromatis et al., 2013; Mavromatis et al., 2014), and its fractionation characteristics are mainly influenced by fluid properties, crystal size and crystallization order (Pogge von Strandmann, 2008; Pogge von Strandmann et al., 2008; Hippler et al., 2009; Schott et al., 2016). The mean values of δ²⁶Mg and crystal diameters are, respectively, ranked in the order of mD > ssD (EHST) > DM > MCD > ssD (LHST) > SD and SD > MCD > ssD (LHST) > ssD (EHST) > mD > DM (Supplementary Appendix S4), and the two factors are not strongly correlated. The preceding contents have shown that mD and DM crystallized the earliest, while MCD and SD crystallized the latest. This order does not correlate significantly with δ²⁶Mg values, so crystal sizes and crystallization order in the Changping Formation may not be the main factors influencing the values.

Compared to typical hydrothermal dolomites, the SD and MCD have higher δ²⁶Mg values (δ²⁶Mg < −2.0‰, Tipper et al., 2006a; Tipper et al., 2006b). The MCD even shares similar values with the ssD of seawater genesis. Thus, δ²⁶Mg-δ¹⁸O and δ²⁶Mg-⁸⁷Sr/⁸⁶Sr covariograms were analysed to identify the dolomitization fluids (Figure 9).

Mg isotopes are not affected by temperature and salinity, and O isotope values are mostly proportional to salinity. Therefore, the positive correlation in the δ²⁶Mg-δ¹⁸O covariogram indicates that the enrichment of Mg isotope and Mg²⁺ synchronized with the dolomitization process is synchronized with the O isotope enrichment. In other words, Mg²⁺ is constantly trapped into the lattice as the diagenetic fluid is gradually concentrated. These characteristics reflect an abundant supply of Mg²⁺ in the diagenetic environment rather than the Mg²⁺-limited closed environment. That is, Mg²⁺ may be essentially derived from evaporated seawater. Furthermore, both DM and mD have a positive correlation. Still, the slope of DM is steeper (Figure 9A), which may indicate that the salinity of the dolomitization fluid has a greater influence on DM formation, while mD may also have other more efficient mechanisms for low-salinity dolomitization. The δ²⁶Mg-⁸⁷Sr/⁸⁶Sr covariograms of DM and mD also support this view from another perspective, as their negative correlations indicate simultaneous enrichment of Mg isotope and Mg²⁺, and Sr depletion caused by the lighter mantle-derived strontium continuously trapped into the dolomite lattice. This evidence suggests that DM and mD originated from seawater. The mD slope is steeper (Figure 9B), which may indicate that more seawater is needed to obtain enough Mg²⁺ during the dolomitization process. This is related to the lower salinity of dolomitization fluid that formed mD.

The ssD values of EHST and LHST show different correlations in the δ²⁶Mg-δ¹⁸O plot (Figure 9A). The EHST samples with negative correlations reflect a buried dolomitization environment. The closed diagenetic system is unable to replenish Mg²⁺, so as the brine continues to concentrate, increasingly less Mg²⁺ can enter the lattice (Heydari, 1997; Whitaker and Xiao, 2010). The Mg and Sr isotopes are not correlated (Figure 9B). This suggests fluid modification from various diagenetic environments. Combined with the C and O isotope and Z value signatures (Figure 8; Supplementary Appendix S4), the (pene) contemporaneous fluids might not cause dolomitization under open conditions, and dolomitization in the EHST ssD likely occurred only in the burial period and was accompanied by phase II hydrocarbon accumulation.

The LHST samples with two distinct slopes of positive δ²⁶Mg-δ¹⁸O relationships exhibit different dolomitization mechanisms in an open environment (Figure 9A). The higher slope values typically result from open (pene) contemporaneous dolomitization events. The samples with slope values of almost zero reflect continuous dolomitization at nearly constant salinity. These phenomena are indicative of outer freshwater injection preventing the continued concentration of stratigraphic brine in a confined environment. If attributed to terrigenous water alone, these samples probably pose a negative distribution in the δ²⁶Mg-⁸⁷Sr/⁸⁶Sr aspect, as terrestrial-derived Sr is heavier (Burke et al., 1982; Carpenter et al., 1991). However, the random distribution of LHST ssD samples in the δ²⁶Mg-⁸⁷Sr/⁸⁶Sr diagram presents multiple origins (Figure 9B). Therefore, the seawater might still be replenished during the dolomitization process, although it is likely to be highly saline.

The MCD and SD samples in the δ²⁶Mg-δ¹⁸O diagram display an obscure negative correlation (Figure 9C), which may indicate an unconfined diagenetic environment affected by fractures and unconformity surfaces. Thus, infiltrating seawater is still involved in the dolomitization process. Additionally, these samples are not correlated in the δ²⁶Mg-⁸⁷Sr/⁸⁶Sr plot (Figure 9D). Combining the results of ssD may reveal that strontium and magnesium isotope covariation graphs are not very sensitive to mixed-source dolomitization fluids.

Based on the above lithological features, the results of microtemperature measurements and relevant geochemical analyses, it is concluded that mD and DM are of (pene) contemporaneous marine origin. The EHST ssD has both (pene) contemporaneous diagenetic characteristics with shallow-medium burial dolomitization, but hydrothermal effects cannot be completely excluded. The LHST ssD is the dolomitization product of seawater, strata brine and atmospheric freshwater. The MCD and SD are of burial origin, with the former mainly consisting of strata brine, hydrothermal and atmospheric fluids, and the latter chiefly consisting of residual concentrated seawater and hydrothermal liquids.

The Z values in Supplementary Appendix S4 show that mD can form in a typical seawater environment. Correspondingly, many single-phase inclusions in the Changping Formation indicate that they likely formed in a low-temperature surface environment rather than in a high-salinity burial environment (Jiang et al., 2016). This is because the extracellular polymeric substances produced from microorganisms can overcome the kinetic barriers posed by low-salinity seawater and increase the alkalinity of the local water environment, leading to dolomite supersaturation and ultimately to dolomite precipitation (Vasconcelos et al., 1995; Vasconcelos and McKenzie, 1997; Sumrall et al., 2015). Correspondingly, the negative carbon isotopic drift in mD (Figure 8) indicates vigorous metabolic activities (Jiang et al., 2016). Their REEs and major and trace elements (Figures 6, 7; Supplementary Appendixes S2, S3) reflect unrestricted environmental characteristics (Banner et al., 1998; Azmy et al., 2011). Associated values of oxygen isotopes are within the range of those of sea-derived dolomites (Figure 8), also suggesting that primary seawater may have been the main fluid for microbial dolomitization.

Microbial dolomitization also occurred in DM. Spherical and normal dolomites are observed in symbiosis (Figure 4C) and have geochemical indicators similar to those of mD (You et al., 2011). However, dolomicrite does not reach the level of microbialites, and the absence of a negative shift in carbon isotope values (Figure 8) implies a strong influence of (pene) contemporaneous sabkha dolomitization (Baniak et al., 2014; Ahmad et al., 2021).

Early diagenetic gypsum is not observed in the Changping Formation, so in a strict sense, there is no typical sabkha environment in the Dingjiatan area (Mei, 2011; Bai et al., 2021b), i.e., mD and DM that formed in the near-surface and (pene) contemporaneous phases are also the products of low-salinity seawater (Figure 10A).

This phenomenon depends to some extent on microbial activity and induction, and on the other hand, research in recent years has suggested that low-salinity seawater can cause large-scale dolomitization (excluding microbial dolomitization) under early low-temperature conditions (Jones and Xiao, 2005). For example, for DM with open, sea-derived geochemistry, lithological characteristics demonstrate that microbial action does not have the capacity for large-scale dolomitization (Figure 4C; Figures 6–9; Supplementary Appendixes S2–S4). Note that compared to typical microbial dolomitization, products of sabkha dolomitization have a slightly higher Mg²⁺/Ca²⁺ ratio (Figure 7; Supplementary Appendix S3) without reaching the salinity of gypsum precipitation, which may be related to frequent sea-level fluctuations in the early Cambrian North China Plateau (Mei, 2011). Reflux events are reflected in the fact that ssD retains the isotopic and REE information of seawater, and exhibits geochemical characteristics with a gradual increase in salinity (Figures 6–9; Supplementary Appendixes S2–S4). The carbon and oxygen isotopic ranges of ssD overlap with those of DM and mD and have lower oxygen isotope values (Figure 8), indicating that reflux is a continuation of sabkha evaporation near the surface. Lithologically, it is also acceptable that porous rocks in the high-energy zone can allow surface brine to pass through and penetrate deep into the strata (Hein et al., 1992; Jones and Xiao, 2005; Mahboubi et al., 2016), and the relevant evidence is that the LHST ssD shows zonal growth, with the earlier dolomite as a core (Figure 4I).

However, not all back-flow liquids can cause reflux dolomitization. The original rocks of the EHST ssD underwent (pene) contemporaneous transformation. However, the timing of effective dolomitization is within the shallow-medium burial period (Figures 10, 11). The lack of large-scale (pene) contemporaneous dolomitization is due to a high sea level and overall low evaporation in the EHST period. The resulting low salinity of the infiltrating seawater is not sufficient for reflux dolomitization to occur after surface dolomitization is completed. Additionally, the EHST has a low degree of dolomitization, indicating that (pene) contemporaneous sabkha and reflux phenomena both play important roles in large-scale dolomitization. By comparison, the overall drop in sea level during the LHST resulted in greater evaporation, causing brine downwelling and reflux dolomitization in LHST strata (Figure 10B).

Related studies in other basins have shown that (pene) contemporaneous infiltrated brine was continuously preserved in the formation and interacted with the surrounding strata during the shallow-middle burial period in a water-rock interaction (Jiang et al., 2016; Lukoczki et al., 2019), as evidenced by the dual isotopic geochemical characteristics of both evaporated seawater and burial environment in the LHST ssD (Figures 7–10), which is a typical reflux product.

However, the EHST and LHST ssD samples show obvious differences. The LHST specimens have a broad range of MgO/CaO ratios (Figure 7B). The higher values of positive correlation reflect a (pene) contemporaneous open environment, while the lower values tend to be negatively correlated, indicating dolomite lattice adjustment during the shallow-middle burial stage. The high XRD data (Supplementary Appendix S1) with typical seawater isotopic information (Figures 5–9) also show pore water interactions following (pene) contemporaneous dolomitization, and atmospheric freshwater can also enter the LHST strata and participate in ssD (Figures 5, 9; Supplementary Appendixes S2–S4). In contrast, the EHST samples have only one set of negatively correlated low values, indicating that they represent the direct production of shallow to medium burial dolomitization. In other words, no (pene) contemporaneous dolomitization event occurred in the EHST ssD, as mentioned in chapter 5.1.2. These two types of burial dolomites are also present in the Lower Ordovician Penglaiba Formation in the Tarim Basin (Qiao et al., 2021).

The burial history (Figure 11) shows that during the middle-deep burial phase, the Changping Formation was buried at depths up to 3,500 m and burial temperatures up to 120°C. Such conditions were favourable for hydrothermal dolomitization. The MCD and SD also show noticeable hydrothermal features of lithology (Figures 4K, 4N, 4O, 5) and geochemistry (Figures 8, 9).

However, the δ²⁶Mg values and crustal element contents of the MCD and SD differ significantly from those of typical hydrothermal rocks (Figure 9; Supplementary Appendixes S2–S4), and the Fe and Mn contents indicate that they are also not typical burial products (Supplementary Appendix S3), while the REEs are instead characteristic of seawater geochemistry (Supplementary Appendix S2). These results may indicate that some of the Mg²⁺ in the dolomitization process came from pore water in the middle-shallow burial period.

At the same time, the Indo-Yanshanian rift tension and fracture slip facilitated the upwelling of magmatic water and the flow of stratigraphic brine, and also the superimposed modification of burial and hydrothermal action in the fractured sections, resulting in the formation of dolomite with special chemical indicators (Jiu et al., 2020). Relative hydrothermal action also affected ssD. Notably, the LHST MCD has also been affected by atmospheric freshwater compared to the EHST saddle dolomites because of karstification and the influence of fault communication with the surface (Figure 10C).

Microbial dolomitization in the (pene) contemporaneous stage occurred largely in the EHST period, and massive microbialites were deposited with a negative δ¹³C shift. Microbial rocks have been shown to form potential reservoirs. However, according to the observation results (Figures 4D–F), the Changping Formation mD does not produce (pene) contemporaneous dissolutive pores, despite frequent sea-level fluctuations. The lack of a long-term unconformity in the EHST also prevents mD from forming coherent pores, and therefore, the EHST microbialites are not considered high-quality reservoirs.

DM is influenced by microbial metabolism, but mainly exhibits low-salinity sabkha dolomitization characterized by low porosity and permeability (Jiang et al., 2016), as indicated by the investigated physical data (Supplementary Appendix S5). Reflux dolomitization, however, could cause a more complex reservoir formation mechanism. Examples include the Sichuan Basin in China (Cai et al., 2014; Liu et al., 2020) and the Permian Basin in the United States (Saller, 2004). Near the restricted environment, the brine that converted directly from the sabkha and reflux effect contains sufficient or even excessive Mg²⁺, leading to excessive dolomitization and a theoretical 75% increase in dolomite volume, accompanied by a significant decrease in porosity with a main reaction of 2 C a C O 3 + M g 2 + + C O 3 2 − → C a M g C O 3 2 (Weyl, 1960; Pray and Murray, 1965). The dolomitization fluids far from the limited environment have lower alkalinity, and the slightly deeper stratigraphic zone also tends to reduce Mg²⁺ as the permeation proceeds. The replacement reaction is dominated by equimolar replacement, resulting in a theoretical 13% decrease in dolomite volume and the formation of intergranular pores, with a reaction equation of 2 C a C O 3 + M g 2 + → C a M g C O 3 2 + C a 2 + (Halley and Schomoker, 1983; Lucia and Major, 1994). For example, ssD samples with (pene) contemporaneous features from the LHST of the Changping Formation at 17.3 m, 16.7 m and 15.8 m (No. DH-8-1, DL-26-1, and DS-22, respectively) are all located in the Dingjiatan section, and their position in the outcrop deepens in turn and gradually moves away from the confining zone. Correspondingly, the MgO/CaO ratio decreases from 0.737 above the theoretical dolomite to 0.579 and 0.446, while the porosity increases from 2.37% to 3.68%, and ultimately 6.38% (Figure 13A).

During the sedimentary period of LHST, the large amount of DM indicates the influence of the (pene) contemporaneous sabkha dolomitization, followed by ssD with primary pores and reservoir skeletons, indicating reflux dolomitization. Although the ssD of the LHST is characterized by the shallow-medium burial phase mentioned above, during which dolomite grew in the annular form, the pore distribution pattern still retains the (pene) contemporaneous characteristics (Figure 13A), and the intercrystalline pores are filled with phase I oil and gas (Figure 5). These hydrocarbons are adsorbed onto the crystal surface, preventing further dolomitization and facilitating pore preservation (Jiang et al., 2015; Jiang et al., 2016). In contrast, the EHST reservoirs rely solely on middle-shallow burial dolomitization. The associated pore reduction mechanism and the accompanying changes in MgO/CaO ratios are very weak in the Miaofengshan section with sample numbers of DH-21-1, DL-32-1, and DH-9-2, respectively (Figure 13B). Despite the tendency for further concentration, as well as brine infiltration, the amount of Mg²⁺ and associated fluid flow capacity in the burial environment are limited, thus resulting in incomplete dolomitization. Their slightly better physical properties compared to those of limestone (Figure 12; Supplementary Appendix S5), however, may indicate that a few equimolar replacement mechanisms still occurred during the shallow-medium burial period, with the formation of intergranular pores filling the phase II hydrocarbon.

During the middle-deep burial period and Yanshanian period, the strong tectonic movement produced fractures, constituted new reservoir spaces, and also provided migration channels. When the hydrocarbon and hydrothermal fluids flowed along the faults, dissolved pores were produced, and these dissolved dolomite liquids also reached the dolomite saturation state under the influence of residual high-salinity formation water. For example, in the EHST, SD partly features a poikilitic texture (Figure 4O), so the phase II oil and gas accumulated only in the residual intergranular pores and fractures. These high-temperature hydrocarbons and hydrothermal fluids also affected the early products, such as ssD in the EHST with hydrothermal geochemical indicators and petrological characteristics shown in Figure 4H, where irregularly mosaic crystals are visible in the upper left ssD.

In the LHST, apart from deep fluids, the fault that communicated with the top of the formation introduced atmospheric water into the reservoirs. Thus, the MCD underwent more abundant dolomitized fluids during the growth period compared to the SD in the EHST, and these MCD samples contain breccias with a lack of pores (Figure 4L). Although the MCD can be distributed beyond the faults, which is destructive to the nearby ssD reservoirs, the overall ssD reservoir preservation condition in EHST is still quite good as mentioned above.

In conclusion, the discovery of porous Cambrian reservoirs formed by multiphase dolomitization in this study strengthens the hypothesis that paleo-reservoirs may host significant petroleum resources for exploration. The pore-enhancement mechanism of the LHST in the early stage was conducive to the formation of high-quality reservoirs. Although some pores in ssD were removed during the shallow-middle burial period, the effective pore-retention capacity is still beneficial to the early hydrocarbon accumulation. By comparison, the less effective ssD reservoirs in the EHST were only formed in the shallow-middle burial stage with a lack of early optimal transformation. Even in the middle-deep burial stage, the late hydrocarbon could also be accumulated in the intergranular pores of SD in the EHST, the primary exploration target for the study area is still the ssD reservoir in the LHST followed by the ssD reservoir in the EHST, as the better porosity increase mechanism and considerable preservation condition of the former.