Origin of the Dolomitic Ooids Formed in the Pliocene Shizigou Formation in the Qaidam Basin, Northern Tibet Plateau and Implications for Climate Change

Abstract

The concentric layers of ooids from the modern environment are mostly aragonite and those from the ancient are mostly calcite and Mg-calcite. Dolomitic ooids are rare and are usually formed via the replacement of aragonite or calcite. Here, dolomitic ooids were found in the Pliocene Shizigou Formation in the Qaidam Basin, Northern Tibet Plateau. This paper focuses on whether the dolomitic ooids is a primary precipitate. Optical microscope and scanning electron microscope observation, combined with X-ray diffraction and cathodoluminescence analyses, indicate that the primary mineral of the ooid cortices is poorly ordered dolomite. Extracellular polymeric substances and pyrite were found in the ooids, indicating that the microbe participation was involved in the formation of the ooids. Firstly, the ooids grow on the offshore lake floor. Due to the involvement of sulfate-reducing bacteria and dissolved silica, the nanominerals were precipitated on extracellular polymeric substances. Then, the ooids were transported to strong hydrodynamic surf zones, where the random nanominerals were abraded to form flattened plates as a new polished layer. In addition, a comparison between the carbon and oxygen isotopic compositions and minerals of ooids from different periods indicate that the Pliocene lakes had a lower salinity and were more humid than Pleistocene lakes. Therefore, ooids may be an effective proxy for reflecting the climatic change and uplift history of the Tibet Plateau.

1. Introduction

Ooids are spherical or ellipsoidal grains with concentric layers of carbonate, which accrete around a nucleus composed of a material such as a lithoclast, bioclast, or siliciclastic grain [1]. Ooids are found in both ancient and modern sedimentary environments [2,3,4,5]. However, whether the formation of ooids is an inorganic or organic process has been debated over several centuries. The widely accepted theory is that the formation is attributed to a physicochemical process, which emphasizes the formation of inorganic carbonates through direct chemical precipitation from water [6,7]. Recently, by applying new technologies and integrative approaches, it has been suggested that microbes play a key role in the formation of ooids [8,9,10,11,12]. Metabolic activities can change the water chemical condition, leading to carbonate mineral formation [8,9]. Although the origin of ooids is controversial, they contain abundant geochemical and petrographic information. Thus, ooids can be used as paleoclimatic and paleoceanographic proxies that reflect the water temperature, chemistry, redox state, salinity, and carbonate saturation state of their environment [13,14,15,16]. The mineral compositions of ooids may undergo variations through geological history [17]. Aragonitic layers dominate in modern ooids, whereas the concentric layers of ancient ooids are mostly composed of calcite and Mg-calcite [17,18]. The primary mineral of ooids may further provide a more accurate clue to understand paleo-water chemistry [17]. In order to reconstruct the paleoenvironmental conditions, the identification of an ooid’s primary mineral is necessary.

Here, we report the first discovery of dolomitic ooids in the Pliocene Shizigou Formation from the Qaidam Basin. Combining X-ray diffraction data, scanning electron microscope (SEM) observations, and carbon and oxygen isotopic compositions of the ooids, the primary mineral of the ooids was identified to be dolomite, and its origin is discussed. The Qaidam Basin lies in the northeastern Tibetan Plateau, which deposits thick layer of Cenozoic lake sediments and has recorded tectonic and climatic changes since the Paleocene epoch [19]. Previous studies have focused on geochemical, magnetic, mineral, grain size, and paleontology to study the climatic changes of the Qaidam Basin [19,20,21,22,23]. The climatic change and uplift history of the Qaidam Basin could have led to the oscillations of minerals by controlling the depositional environment [20]. Aragonitic ooids were often formed in the Pleistocene [23,24], but ooids are less reported in the Pliocene. In this study, we discovered dolomitic ooids in the Pliocene Shizigou Formation, which are different from the ooids formed in the Pleistocene. By discussing the origin of dolomitic ooids and comparing them with the ooids developed in the Pleistocene, it can be determined whether the ooids can reflect the climatic change and uplift history of the Tibet Plateau.

2. Study Area

The Qaidam Basin, located on the northeastern Tibetan Plateau (Figure 1a), is a large intermountain basin (~700 km × 300 km) [25]. It is surrounded by the Kunlun Mountains to the south, the Qilian Mountains to the north, and the Altyn Mountains to the northwest (Figure 1b). The basin sits at an elevation of 2800–3300 m and is a vast arid region [25]. The basin’s annual evaporation is over 20 times greater than its annual precipitation [21]. The mean annual temperature is about 0–5 °C [23]. The Qaidam Basin is the largest Cenozoic sedimentary basin within the Tibetan plateau [26]. Its formation was closely related to the India–Asia collision. The Cenozoic tectonic evolution history of the Qaidam Basin underwent three stages: the strike-slip basin stage (60–46 Ma), the foreland basin development (46–2.4 Ma), and the intermontane basin stage (2.45–0 Ma) [27]. The Cenozoic strata in the Qaidam Basin exceed a thickness of 12,000 m [19], and were deposited mainly in a fluvial-lacustrine environment that includes alluvial fans and fluvial, delta and lake sediments [28]. This Cenozoic stratigraphy can be divided into the following seven formations: the Lulehe Formation (E₁₊₂), the Xiaganchaigou Formation (E₃), the Shangganchaigou Formation (N₁), the Xiayoushashan Formation (N₂¹), the Shangyoushashan Formation (N₂²), the Shizigou Formation (N₂³), and the Qigequan Formation (Q₁) [19,29]. This study area lies on the Yahu Anticline, located in the central Qaidam Basin (Figure 1b). At this study location, the Shizigou Formation (5.2 Ma to 2.6 Ma) is dominated by marginal lakes and mud-flat environments with fluvial contribution [30]. The Qigequan Formation is dominated by evaporite lithofacies [30]. A well, Z1, was drilled in this area and a sample (Z1-15) was collected from the Shizigou Formation at a depth of 1500 m (Figure 2).

3. Materials and Methods

The sample Z1-15 was divided into three parts (A1, A2 and A3) for different analyses. Sample A1 was used for optical microscope and cathodoluminescence observations. Two polished thin sections were prepared from sample A1. One section was stained with Alizarin Red S and K ferricyanide and was imaged under a polarization microscope with a photo-collecting system (Zeiss Axioscope A1 APOL). The other section was used for cathodoluminescence (CL) analyses using an Olympus microscope equipped with a CL8200-MKSCLinstrument. To acquire more microinformation about the sample, sample A2 was used for SEM observation. Two polished thin sections and one fresh section were prepared for SEM. One thin section was prepared via Ar ion milling (685.C, Hitachi High-Tech) with an accelerating voltage of 5 kV and a milling time of 0.5 h. The other thin section and the fresh section were coated with carbon to render them conductive. Three sections were mounted on aluminium stubs and imaged using a Zeiss MERLIN Compact field emission scanning electron microscope (FESEM) with an X-ray energy dispersive spectrometer (EDS) (Xflash/30, Bruker). Lower accelerating voltages (1 kV) with working distances of approximately 5mm were used on the thin section milled by Ar ion. The other thin section and the fresh section were observed using the higher accelerating voltages of 5 kv, 10 kv, or 15 kv, and working distances of about 8 mm. Both secondary electron (SE) and angle-selective backscatter (ASB) images were acquired. The ooid size and cortical thickness were obtained from SEM micrograph images using the ImageJ software.

Sample A3 was crushed into 200 mesh powders by using an agate mortar for carbon and oxygen stable isotope analyses and whole-rock minerals. A minor amount of powders reacted with 100% orthophosphate at a temperature of 90°C to obtain CO₂. The carbon and oxygen isotope compositions of CO₂ were analyzed using a Thermo Fisher Delta V-GasBench II mass spectrometer. Carbon and oxygen isotope compositions of samples were expressed with respect to the Vienna PeeDee Belemnite (V-PDB) standards. Replicate measurements of the standard were 0.06‰ for δ¹³C and 0.08‰ for δ¹⁸O. Whole-rock minerals were analyzed using a Rigaku Ultima IV X-ray diffraction (XRD) analyzer with Cu-Kα radiation, a voltage of 40kV, and a current of 40 mA. The quartz (d₁₀₁ = 3.343 Å) was used as an inter-standard to correct the peak. The ratio of the intensity of the (015) reflection to the (110) reflection (R) was used for estimating the degree of cation ordering in the dolomite [31]. The Mg content of the dolomite was calculated by using the method of a diagram of MgCO₃ % vs. d₁₀₄, which was constructed by Fang and Xu [32].

4. Results

4.1. Mineralogy

The XRD patterns show that the sample is constituted of dolomite and calcite (Figure 3 and Figure S1). The (104) plane d-spacing (d₁₀₄) in the calcite of the sample is 3.029 Å. According to the method for calculating the Mg content of the calcite [32,33], the value of MgCO₃ is only 1.9 mol%. Generally, calcites containing >4 mol% MgCO₃ are considered to be high-magnesium calcite, and, conversely, those containing <4 mol% MgCO₃ are called low-magnesium calcite [34]. A linear relationship exists between the increase Mg²⁺ content and the decrease in the (104) d-spacing (larger 2θ value) in calcites [35]. Figure 3 shows that the (104) reflection in the sample (Figure 3c) is located between optical-grade calcite (Figure 3d) and high-magnesium calcite (Figure 3b). Thus, the calcite in the sample is considered to be low-magnesium calcite.

The (015) reflection is the most intense “ordering” reflection (indicating alternating layers of Mg and Ca) observed out of all the dolomite XRD patterns (Figure 3a) [36]. The (015) reflection displays decreasing intensity alongside disorder [31]. The ordered, stoichiometric dolomite shows nearly equal intensities between the two reflections (R = 1) in the XRD patterns (Figure 3a). The (015) reflection in this sample has a weak intensity (Figure 3c). The ordering state of the dolomite in this sample is 0.11, which means it is poorly ordered. The d₁₀₄ of the dolomite of the sample is 2.912 Å. The Mg content of the dolomite was calculated by using the diagram of MgCO₃ % vs. d₁₀₄ [32]. When the d₁₀₄ is 2.912 Å, the content of Mg ranges from 40 mol% to 55 mol% and is constrained by the lowest and highest cation ordering state. Combined with the order parameter of 0.11, the Mg content of the dolomite in the sample is about 50.5 mol%.

4.2. Characteristics of Ooids and Stable Isotopic Compositions

The sample consists of gray-white, well-sorted ooids (Figure 4a,b). The thin-section observation indicates that the sample is constituted of dolomitic ooids with calcite cement, and micritic ooids that are commonly dark and nearly opaque (Figure 4c,d). The calcite is red in the thin section, whereas the dolomitic ooids have no color (Figure 4d). All of the ooids display spherical or ovoid shapes (Figure 4c,d). The thin-section observation shows that the ooids have concentric or radial-concentric cortices (Figure 4c), whereas the dolomitic ooids display concentric cortices by observing the CL and ASB images (Figure 4e,f and Figure 5a). The ooids mainly consist of micritic dolomite (Figure 5a,b and Figure S2), whereas euhedral dolomite crystals develop on the surface of the ooids (Figure 5c,d). The minerals in the nucleus are mostly micritic dolomite with a minor clay mineral (kaolinite) and pyrite (Figure 5a,e,f and Figure S2), and some are quartz or albite (Figure 5b and Figure S2). The CL observation indicates that the calcite cement displays a light-orange luminescence and that the ooids show concentric dark and orange luminescence (Figure 4e,f). A total of 402 ooids were counted to identify grain sizes in the sample. The size of the ooids in the sample ranges from 144.7 µm to 1076 µm with a mean value of 431.3 µm, and with a majority of the ooids being a size from 244 µm to 544 µm (Figure 6a). The thickness of the cortices is from 41.4 µm to 312.7 µm with a mean value of 125 µm, and most have a thickness distributed from 71 µm to 161 µm (Figure 6b). The sample has a δ¹³C_VPDB value of −0.7‰ and a δ¹⁸O_VPDB value of −5.4‰.

5. Discussion

5.1. Primary Mineral of Ooids

The cortex layers of ooids from the modern marine environment are mostly composed of aragonite, and sometimes, they consist of high-Mg calcite in the marine hypersaline environment [37,38,39]. In contrast to marine ooids, the cortex layers of lacustrine ooids are generally aragonite [1]. As aragonite and high-magnesium calcite belong to the metastable mineral groups, they are almost completely transformed into low-magnesium calcite under the influence of diagenesis, therefor, ancient ooids are characterized by having calcite and low-magnesium calcite [1,17]. Dolomitic ooids are also found in ancient eras, such as the Neoproterozoic, the Paleoproterozoic, the Cambrian, etc. [40,41,42]. Dolomitic ooids are most commonly formed via the replacement of aragonite or calcite, and retain some information about the original mineral characteristics and textures [17]. In the Qaidam Basin, the dolomitic ooids was found in the Pleistocene and shows the dolomite replacement of the aragonite, indicating that the primary minerals of the cortex layers of ooids are aragonite [24]. If the Ca-Mg carbonate is ordered, then it is dolomite; otherwise, it is not dolomite [43]. From the XRD patterns, it was identified that the Ca-Mg carbonate of this study was ordered, even though the ordering state of the dolomite is low. Our studied ooids consisted of nonplanar micritic dolomite and outward planar-e (euhedral) dolomite crystals (Figure 5c,d). The SEM images of the ooids do not show aragonite and its textures in the cortex layers of the ooids (Figure 5a,b). A few calcite found in the ooids’ cortices occurred in some destroyed ooids, implying that the calcite was filled in during the later diagenesis, such as with the calcite cement in the sample, meaning that calcite is not the original mineral (Figure 4f). These features indicate that the primary minerals of the ooid cortices may be dolomite. The dolomite developed on the surface ooids and in the pores, presenting coarse euhedral crystals, which indicate that these dolomites were formed after the ooids were deposited.

Dolomite cannot be directly precipitated from modern seawater, and it is difficult synthesize in the laboratory under Earth surface conditions, which, due to hydration effects, hinder Mg²⁺ into the crystal lattice of the dolomite [44,45,46]. In contrast, recent studies show that dolomites can form at low temperatures through microbial-organic mediation with sulfide, silica, and clay minerals, which is supported by both laboratory experiments and field observations [47,48,49,50,51,52,53,54,55]. Sulfate-reducing bacteria play an important role in dolomite formation [47]. Sulfate reduction is an important redox process that assists in the conversion of SO₄²⁻ to HS⁻. Dolomite occurs alongside the pyrite formed by sulfate-reducing bacteria in modern lakes [56,57]. The ASB images show the presence of micritic dolomite mixed with pyrites in the sample (Figure 5a,b). The elemental mapping of the ooids shows that nutritional elements such as P and Fe were detected, indicating that microbial remnants are present in the ooids (Figure 7). These findings indicate that the formation of dolomite is associated with sulfate-reducing bacteria. The CL properties of carbonate minerals reflect the spatial distribution of Fe and Mn ions [58]. In carbonate mineral, Mn²⁺ is the most important activator of CL, whereas Fe²⁺ is a quencher [59,60]. The CL images of the dolomitic ooids in the sample show a dark luminescence, which reflects more Fe²⁺ or low Mn²⁺ distributed in the ooids (Figure 4e,f). Combined with the Fe elemental mapping and the occurrence of the pyrite, more Fe²⁺ contents should lead to the ooids having a dark luminescence. This result also proves that micritic dolomite was formed through a reducing condition. Microbial sulfate reduction can produce dissolved sulfide in water, which enhances the Mg²⁺ incorporation into the calcitic structure and result in the precipitation of dolomite [51]. In addition, dissolved silica in water can also promote Mg²⁺ incorporation into the Ca-Mg carbonates, promoting dolomite nucleation and growth [55]. The elemental mapping of the ooids shows amounts of Si in the core (Figure 7). Thus, dissolved silica may be another important factor that enhances dolomite formation, with the exception of the microbial sulfate reduction. Dolomite formed at room temperature is often disordered dolomite or protodolomite [36,51,61]. When the sediments are buried, the protodolomite is converted into ordered dolomite due to increasing temperature and pressure [36]. During an early burial, dolomite crystals formed on the surface ooids and in the pores (Figure 5d). With the dissolved silica and sulfide concentration drawn down in the pore water, the Mg²⁺ cannot enter into the calcitic structure, which leads to dolomite growth ceasing and calcitic cement occurring (Figure 5a).

5.2. Origin of Ooids

The mechanisms of ooid formation have been attributed to inorganic (physicochemical progress) or organic (biological progress) origins [1]. Based on the genesis of the dolomite above, it was found that dolomite is formed through microbial mediation. Therefore, the origin of ooids could be related to microorganisms. A biologically mediated process for ooid formation follows two processes: (1) a biologically induced mechanism (active mineralization) induces mineral precipitation though the microbial interaction with the environment; and (2) a biologically influenced mechanism (passive mineralization) provides mucilaginous material or extracellular polymeric substances (EPS) that act as a template for mineral precipitation [1]. SE images show that EPS was found in the cortex layers of ooids and nanodolomites mixed with EPS (Figure 8a). Through observations of the thin section milled by Ar ion, pyrites and EPS appear in the same area (Figure 8b). All together, these findings provide evidence that the formation of ooids is attributed to the biologically mediated mechanism. If the formation of ooids is entirely attributed to an organic process, they would grow in a quiet water environment with radial cortices [1]. Although some ooids present radial-concentric cortices, as seen in the thin section (Figure 4c), the dolomitic ooids display concentric cortices, as seen in the CL and ASB images (Figure 4e and Figure 5a). Scanning electron microscopy analyses of ooid cortices reveal dolomite with plate morphologies, and nanograins along the edges of the plate (Figure 8c,d). SE images of etched ooids show concentric laminations and nanograins aligned parallel to the laminae due to the nanocrystal dolomite aggregate running along lamellae directions (Figure 8e,f). These textural features indicate that the ooid cortices underwent textural changes. At the early stages of ooid accretion, the ooid cortices consisted of dolomitic nanograins with a random pattern (Figure 8g). The mutual impact of the ooids allows the nanograins to flatten and combine with each other to form plates that are oriented parallel to the laminae (Figure 8d,h). Therefore, ooid growth needs a strong hydrodynamic environment such as the surf zone, as this environment provides the force necessary for inter-grain collision.

The formation of ooids generally requires alternating agitated and resting stages [1,6]. These two stages often occur in the surf zone controlled by waves [1,6]. The textural characteristics of ooid cortices indicate that ooids are formed in a strong hydrodynamic environment, but microbial activity such as that of sulfate-reducing bacteria requires calm waters. Therefore, the surf zone of beaches with strong hydrodynamics does not meet the condition for ooid formation. Instead, the ooids could grow on the offshore lake floor away from the turbulent surf zones, where the sulfate-reducing bacteria and dissolved silica lead to nanograin mineral formation and the EPS matrix serves as the nucleation template for nanomineral precipitation. When the lake level draws down in the dry season, the ooids are transported from the calm microbially stabilized areas to surf zones, where the newly formed random nanograin minerals on the ooids are abraded to form flattened plates as a new polished layer. The sediment transport is repeated periodically until the ooids exceed a threshold size. This model of the ooid formation also occur in the marine environment and is named the “conveyor belt” model [11,62], which may be developed on windward beaches.

5.3. Implications for the Paleoenvironment

Although the lacustrine dolomites of the Quaternary age are found to span a wide range of geochemical and hydrological conditions, nearly all are from saline lakes [56]. The presence of dolomite ooids in this study suggests that the paleolake from the Pliocene may be a salt lake. However, the ooids formed in Pleistocene salt lakes in the Qaidam Basin consisted of aragonite and were cemented by halite and gypsum [24]. The occurrence of halite crystals, as the most common evaporative minerals, indicates that the Pleistocene lakes have a higher salinity than Pliocene lakes. Therefore, the shifting of ooid minerals from dolomite to aragonite may reflect the variation of the climate. The carbon and oxygen isotopic compositions of aragonitic ooids from the Pleistocene are similar to those of ooids from the Bahamas and New Zealand [63,64]. However, the carbon and oxygen isotopic compositions of our sample, which come from the boundary of the salt lake and belong to a brackish lake, are different from previously found compositions and present lower values (Figure 9). Variations of carbonate δ¹³C and δ¹⁸O are mainly attributed to the evaporation in the Qaidam Basin [25]. The evaporation of lake water was strengthened by climate drying, which results in the increase in the of δ¹³C and δ¹⁸O values of carbonates [65,66]. Differences in the mineral composition and isotopic composition indicate that the Pliocene lakes have a lower salinity and are more humid than Pleistocene lakes. Fang and Xu [67] also indicated that the oscillatory appearance in the types of sedimentary carbonate can be used to interpret paleoenvironmental changes. Differences in the silicate weathering can lead to variation in the input of silica into the lake [68]. Strong silicate weathering produces more silica in the lake and enhances the dolomite precipitation; conversely, weak silica weathering precipitates calcite or aragonite [68]. Therefore, the change from dolomite ooids in the Pliocene to aragonite ooids in the Pleistocene might also be due to dissolved silica concentration changes in the lake.

Climate change and tectonic activity in the Qaidam Basin can control the oscillations of authigenic minerals, such as various evaporite and clay minerals [20]. During the Pliocene–Quaternary, two intense main uplifts in the Qaidam Basin occurred at 5.3–4.5 Ma and ~3.6 Ma [19,69,70]. At 5.3 Ma, the climate is already dry, but the aridification is not serious [21]. The appearance of Pediastrum in the Yahu Anticline, which occurs frequently in freshwater environments, indicates relatively wet conditions during this period [21]. Heermance et al. [30] indicated that the lake in the central Qaidam Basin is a freshwater lake at 4.2 Ma. Thus, the lake water gradually changes from fresh water to brackish water during the Pliocene. It is consistent with the lake water conditions indicated by the carbon and oxygen isotopes of ooids in this study. During this period, the Qaidam Basin experienced a relatively high temperature [20], which enhanced the silicate weathering and provided more dissolved silica to be inputted into the lake [68]. When the water reached the chemical conditions that can precipitate dolomite, the dolomite ooids were formed due to the participation of sulfate-reducing bacteria and dissolved silica. A drought event occurred at 3.6 Ma owing to both global cooling and the Tibetan Plateau uplift [19,20], which caused a long-term stepwise drying of the Qaidam Basin after the late Pliocene [21]. A widespread distribution of evaporative minerals such as calcite, aragonite, and gypsum occurred during this period reflecting an abrupt change in the paleolacustrine chemistry [20]. Notably, during the Mid-Pleistocene Climate Transition event, global cooling led to the global ice volume increasing by ~15% [71], which resulted in decreasing strength of silicate weathering and, ultimately, less dissolved silica [68]. Reduced silica input into the lake resulted in aragonite precipitation in the lake water and the formation of aragonite ooids. The change in the climate led to the variation in the microbial activity and dissolved silica concentrations in the water, which could catalyze dolomite growth [55,67]. The variations in the mineral compositions and isotopes of ooids in the Pliocene and Pleistocene correspond to the different uplift stages of the Qaidam Basin. Therefore, ooids can be used as an indicator of climatic change and uplift history for the Tibet Plateau.

6. Conclusions

(1)

The ooids formed in the Pliocene Shizigou Formation in the Qaidam Basin mainly consist of micritic dolomite that is poorly ordered. The nuclear minerals are mostly micritic dolomite with a minor clay mineral and pyrite. The thin-section and SEM observations combined with CL indicate that the primary mineral of the ooid cortices is dolomite. The micritic dolomite was formed through a reducing condition, and its formation was related to the presence of sulfate-reducing bacteria and dissolved silica. The euhedral dolomite crystals were formed in the pore water after the ooids were deposited.

(2)

The formation of the dolomitic ooids in the Shizigou Formation occurred according to the “conveyor belt” model. The ooids grew on the offshore lake floor, where nanominerals were precipitated on the EPS due to the involvement of sulfate-reducing bacteria and other microbes. Then, the ooids were reformed under strong hydrodynamic surf zones, where the random nanograin minerals were abraded to form flattened plates as a new polished layer. The sediment transport was repeated periodically until the ooids exceeded a threshold size.

(3)

Ooids from the different uplift stages of the Qaidam Basin (in the Pliocene and Pleistocene) have variations in mineral compositions and the carbon and oxygen isotopes. The minerals of the ooid cortices changed from Pliocene dolomite to Pleistocene aragonite. The δ¹³C and δ¹⁸O values of the Pleistocene carbonates are higher than those of the Pliocene. These differences indicate that the Pliocene lakes had a lower salinity and were more humid than Pleistocene lakes. Ooids may be an effective proxy for reflecting the climatic change and uplift history of the Tibet Plateau.