Geochemistry of Altungol cap dolostones from the Tarim Basin, NW China

The Snowball Earth hypothesis postulates that the whole earth was covered by ice sheets for millions of years in the Neoproterozoic. In Tarim Basin, there are four intervals of glacial deposits recorded in the Quruqtagh area during the Neoproterozoic. In chronological order, they are Bayisi, Altungol, Tereeken, and Hankalchough glaciations. In this study, we analyzed carbon, oxygen, and strontium isotopes; whole-rock geochemistry; and acid-insoluble residua of the ∼5-m-thick Altungol cap dolostone (ten samples) in Quruqtagh unit of the Tarim basin. The geochemical data of whole rock are used to explain the ocean environmental changes, while the geochemical data of acid-insoluble residua reflect the changes of the continental environment. The δ13CPDB of cap dolostone varies from −2.5 to −1 ‰, which is likely due to the influence of organics. From the bottom to the top of the Altungol cap dolostones, the chemical index of alteration of acid-insoluble residua rises from 72 to 77, but the index of chemical variation drops from ∼3 to ∼1. Ni, Zn, Cu, and Y contents of acid-insoluble residua are enriched in 2.5 m. The values of iron maintain at a high level in all of the whole-rock samples (>5.15 %), but in the samples of acid-insoluble residua, the values of iron are higher than in the whole rock (19∼60 %) below 1.5 m. The variations in chemical composition were probably associated with the changes of continental weathering. The geochemistry suggests that ocean and continent environmental variations had taken place after the Altungol glaciation, including the destruction of ocean stratification and the changes of chemical weathering. During this period, the source of this area had changed. At first, the weathering residues of the parent rocks were exhausted by meltwater. Then, the vigorous weathering of freshly exposed continental crust provided a new source.

Most Neoproterozoic glacial deposits are capped by carbonates usually deposited in warm-water environments (Evans 2000;Porter et al. 2004;Hoffman 2011). The glacial deposits terminated abruptly with the deposition of cap carbonates, which constitutes the sudden change of lithofacies. This sudden change in lithofacies implies dramatic environmental changes during the thawing of a Snowball Earth event (Williams 1979;Hoffman et al. 1998). The trigger(s) of Snowball glaciation and deglaciation, as well as the resulting physical and chemical perturbations to the post-glacial ocean-atmosphere environment, remains a topic of considerable debate (Grozinger and Knoll 1995;Kennedy et al. 1998Kennedy et al. , 2001Hoffman et al. 1998;Fairchild and Kennedy 2007;Hoffman and Li 2009;Hoffman 2011). Carbon and sulfur isotope chemostratigraphic data imply rapid post-glacial changes in marine circulation and chemical stratification (Shen et al. 2005(Shen et al. , 2008(Shen et al. , 2010Giddings and Wallace 2009). Ratios of Fe/S are considered evidence for increased surface runoff from enhanced rainfall during a post-Snowball greenhouse environment Swanson-Hysell et al. 2010).
This study presents stable isotope, major, and trace metal concentrations of a Sturtian-aged cap dolostone overlying the Altungol glacial strata (Gao and Chen 2003;Xu et al. 2009;Hoffman and Li 2009;Gao et al. 2010). The data of δ 13 C carb chemostratigraphy reported by Kou et al. (2008) suggested that Altungol Formation may be similar to the most intensively studied Sturtian cap carbonate in the Rasthof Formation, Namibia (Hurtgen et al. 2002;Macdonald et al. 2010a, b;Hoffman 2011), and other post-Sturtian cap dolostones (Giddings and Wallace 2009).
In this study, new geochemical results of the whole rock and acid-insoluble residua of Altungol cap dolostones are reported. All of these geochemical data can be explained by a coupling environmental change in post-glacial ocean and continent, in the Tarim Basin. The isotopic and chemical composition of dolostones can be used to explain the environmental changes in post-glacial ocean. The elemental geochemistry of acid-insoluble residua reflects the environmental changes of post-glacial continent.

Regional geology
The Quruqtagh unit is located in the Northeastern edge of Tarim Block and is considered as the southeast branch of Tianshan mountains (Gao et al. 1980). A complete Neoproterozoic succession is distributed in the study area, containing four sets of glaciation sedimentary strata, referred to as Bayisi, Altungol, Tereeken, and Hankalchough Formations (Xu et al. 2009;Gao et al. 2010). They are correlated to Kiagas, Sturtian, Marinoan, and Gaskier Formations (Macdonald et al. 2010a, b;Gao and Chen 2003;Hoffman and Li 2009), respectively, based on the U-Pb zircon age of volcanic units in these formations.
In this study, the Altungol cap dolostone samples were collected from the southern Yaerdany Mountains I section (the YI section, 40.74°N, 88.94°E), which is located in the southeastern of Quruqtagh area (Fig. 1). Neoproterozoic strata of the YI section are described as the Quruqtagh Group (Gao et al. 1984(Gao et al. , 1993 and can be divided into the Bayisi, Altungol, Tereeken, Zhamoketi, Yukkengol, Shuiquan, and Hankalchough Formations. The Altungol Formation unconformably overlies the Bayisi Formation and is conformably overlain by Tereeken Formation. The Zhaobishan Formation, which is reported in the other sections (such as Xishankou section and Yukengou section) (Shen et al. 2008;Xu et al. 2009), is missing in the Yaerdany area. In the YI section, the Altungol Formation consists of a ∼45-m-thick tillite, a ∼5-m cap dolostone, and a ∼20-m shale. The Altungol cap dolostone consists of dark-gray medium-bedded dolomite with less sedimentary structure. The contact between the dolomite and the underlying tillite is uneven, with the ∼10-cm-thick black shale. The dolomite is conformably overlain by greyweathered shale, in the YI section.
A volcanic unit in the topmost Bayisi Formation gave a SHRIMP zircon U-Pb age of 725 ± 10 Ma (Xu et al. 2009). He et al. (2014) reported U-Pb zircon ages of 655.9 ± 4.4 and 654 ± 10 Ma from the upper Altungol Formation in Quruqtagh. Therefore, the age of the Altungol diamictites and cap dolostones is between 725 ± 10 and 654 ± 10 Ma, which is probably correlated with the Sturtian glaciations (Rooney et al. 2015).
Ten cap dolostone samples (numbered as YI-12-1 to YI-12-10) were collected at 0.5-m interval from the Altungol Formation at the YI section. Special sedimentary structures, caused by the giant wave ripples (Allen and Hoffman 2005), are not present in the Altungol cap dolostones (Fig. 2). These dolostones are also lack for other sedimentary structure. Therefore, it is thought that the Altungol cap dolostones were probably deposited below storm wave base. Photomicrograph of the Altungol cap dolostones shows that the samples are homogeneous dolomicrite and micro-powder crystal dolostones (Fig. 2b). There are no typical clastic particles and biological detritus in thin section.
75°C for 3 h, and CO 2 was extracted using a traditional offline technique (Craig 1953). The stable isotopic analysis was conducted using a MAT 253 mass spectrometer, at CNNC Beijing Research Institute of Uranium Geology, Beijing, China. The data were expressed as relative isotopic ratios for δ 13 C and δ 18 O, in per mil units relative to the VPDB international standard.

Strontium isotopes
For strontium isotopes tests, nine samples (except YI-12-2) of whole rock were cleaned and crushed into rock powder ∼5 mg for tests. Sample powder was digested with a 2.5 M hydrochloric acid and prepared for Sr isotopic analyses (Wang et al. 1988(Wang et al. , 2007. The strontium was separated using standard ion exchange techniques (Yang et al. 1997). Sr isotopic analysis was performed on a VG 354 mass spectrometer with five collectors at the Center of Modern Analysis, Nanjing University. Accuracy of Sr isotope was checked by running the Standard Reference Material NBS 987 (certified value 0.710340 ± 0.000260) with a mean 87 Sr/ 86 Sr value of 0.710241 ± 0.000007.

Geochemistry
In this research, both whole rock (nine samples, except YI-12-2) and acid-insoluble residua (nine samples, except YI-12-8) elemental geochemistry analyses were utilized. The preparation of acid-insoluble residua of cap dolostones is described as follows. The pure powdered samples were put into 1 mol/L hydrochloric acid. When the carbonate part was completely dissolved, the insoluble parts were extracted and cleaned (Wang et al. 1999).
For major element analysis, the sample powder and lithium metaborate flux were mixed in a 1:10 ratio and fused at 1050°C in a Pt-Au crucible. The resultant melt was cooled, and a glass disk was made for XRF analysis. For trace element and rare earth element (REE) analysis, about 25-mg sample powder was dissolved using a hydrofluoric acid (HF) + HNO 3 mixtures in a Teflon bomb at ∼80°C for 24 h and then evaporated. When the solutions were evaporated to nearly dry, 1.5 ml HNO 3 , 1.5 ml HF, and 0.5 ml HClO 4 were respectively added, and the beakers with solutions were capped for digestion within a high-temperature oven at 180°C for at least 48 h until the samples were completely dissolved. Finally, the solutions were diluted with 1 % HNO 3 to 50 ml for determination. The major and trace elements were detected at CNNC Beijing Research Institute of Uranium Geology, Beijing, China. The samples for major element test were analyzed using the Axios MAX X-ray Fluorescence Spectrometer. The trace and rare earth elements were analyzed using the ELEMENT XR plasma mass spectrometer.

Carbon, oxygen, and strontium isotopes
The carbon and oxygen isotopic data are presented in Table 1. The negative δ 13 C PDB values of post-Sturtian cap dolostone have been reported in many carbonate sections over the world. Such as the δ 13 C PDB average value of the Rasthof Formation in Namibian is −2.39 ‰ (Hurtgen et al. 2002), and two sections of the Sturtian Formation in South Australia show that the average value of δ 13 C PDB is −1.77 ‰ of dolostone (Giddings and Wallace 2009). In the study area, Kou et al. (2008) reported values of δ 13 C PDB ranging from −2.09 to −2.43 ‰. In the YI section, the values of δ 13 C PDB begin at −2.5 ‰ and increase to −1 ‰. The overall δ 13 C PDB is increasing from bottom to top but drops abruptly to −4 and −2.2 ‰ at 2 and 4 m (Fig. 3). The δ 18 O PDB values range from −8.1 to −4.3 ‰. The Strontium isotope compositions vary between 0.709125 and 0.709608, with an average error of ±0.000007 (Table 1 and Fig. 4c).

Elemental geochemistry
In this study, the elemental data of the cap dolostones acidinsoluble part were obtained (Tables 2, 3, 4, and 5). Huang and    Bausch (1999) indicated that most of the acid-insoluble residua in the carbonates are related to the terrigenous deposits. It is inferred that the composition of the acidinsoluble residua depends on the source and the corresponding alterations, like chemical weathering. Thus, the geochemical indicators of clastic rock (especially the mudstone) can be used to analyze acid-insoluble residua of cap dolostone, to reflect the environment changes of the source region. The major element geochemistry results of the whole rock are presented in Table 2. The Altungol cap dolostones have relatively high content of SiO 2 and Al 2 O 3 , which range from 2.27 to 5.67 % and 0.416 to 1.32 %. The Mg/Ca (mol/mol) ratios of most samples range from 0.768 to 0.829 (except the ratio of YI-12-3 which is 0.486). The iron content (wt%) maintains at a high level and increases from 5.15 to 10.21 % (Fig. 4e).
In the acid-insoluble part (Table 3), the ratios of Ni/ Ti, Zn/Ti, Y/Ti, Cu/Ti, and Zn/Y are relatively high in the lower part of the section and decrease upward ( Fig. 4a, b). These ratios have an obvious variation at 3 m, which possess the analogous change trend with δ 13 C PDB . The samples below 1.5 m show an iron enrichment in acid-insoluble residua (Fig. 4d).
The REE data of the Altungol cap dolostones, NASCnormalized REE patterns, and parameters (Haskin and Frey 1966) are presented in Table 4 and Fig. 5. The total REE contents of most samples range from 12.4 to 79.6 (except the value of YI-12-1 is 218). As various barium compounds can cause artificial positive Eu anomalies when measured by ICP-MS (Dulski 1994), the extremely high Ba concentration may result in the positive Eu anomaly of YI-12-10. The samples of the Altungol cap dolostone are characterized by LREE depletion relative to HREE (La N /Yb N = 0.124-0.671, except YI-12-1 La N /Yb N = 4.86), no Ce anomalies (Ce/Ce* ranges from 0.72 to 1.35), and negative Eu anomalies (Eu/Eu* ranges from 0.163 to 0.935, except YI-12-10 Eu/Eu* = 2.03).

Reliability of geochemical data
As the cap carbonates are significantly enriched in Mn and Fe, presumably originating from weathering products (Shen et al. 2005), ratios of Mn/Sr were not measured to evaluate diagenesis. Theoretically (Zhao et al. 2012). In addition, δ 13 C PDB has no linear relationship with δ 18 O PDB , which indicates that the samples have not been influenced by strong diagenesis and hydrothermal alteration, and the δ 13 C PDB compositions reflect depositional conditions (Derry et al. 1994).
REEs of the marine carbonates reflect the geochemistry of waters in which the carbonates were deposited (Webb and Kamber 2000;Nothdurft et al. 2004), and they are immobile during diagenesis (Webb et al. 2009). However, diagenetic exchange after deposition would also cause Ce-enriched, Eu-depleted, and low Dy N /Sm N ratios (Shields and Stille 2001). Thus, if the REE of carbonates samples has been changed by diagenetic exchange, the Ce/Ce* and Eu/Eu* will show a negative correlation relationship, while the Ce/Ce* and Dy N /Sm N will show the same relationship. In the YI section, there are no negative correlations of Ce/Ce*-Eu/Eu* and Ce/Ce*-Dy N /Sm N caused by diagenetic alteration (Figs. 7a, b), suggesting that the REE of the samples can represent the primary seawater and the geochemical data have not been changed after deposition (Ce/ Ce* = 2*Ce N / (La N + Pr N ), Eu/Eu* = 2*Eu N / (Sm N + Gd N )). The values of Eu and Ca (except YI-12-3) show a positive correlation, suggesting that the Eu anomalies may be caused by sedimentary environment (Fig. 7c). The La anomalies can influence the Ce anomaly data and must be removed by using Pr/Pr* values (Bau and Dulski 1999), before using the Ce/Ce* data. According to the binary diagram of the relationship between Ce/Ce* and Pr/Pr* (Pr/Pr* = Pr N / (0.5Ce N + 0.5Nd N )), the Ce anomalies can be divided into five fields (Fig. 7d): field A, neither Ce nor La anomaly; field B, positive La anomaly and no Ce anomaly; field C, negative La anomaly and no Ce anomaly; field D, true positive Ce anomaly; and field E, true negative Ce anomaly. Figure 7d shows that the samples above 3 m show true positive Ce anomaly.
Trace element criteria for assessing optimal strontium isotope of carbonate samples have been proposed for the Precambrian (Brand and Veizer 1980;Derry et al. 1989;Asmerom et al. 1991;Kaufman et al. 1993). Generally, samples with low Sr contents and high Fe/Sr or Mn/Sr are least likely to preserve the original marine 87 Sr/ 86 Sr (Miller et al. 2003). However, the cap carbonates have the characteristics of significantly high in Mn and Fe contents (Shen et al. 2005). The 87 Sr/ 86 Sr ratios are higher than 0.709, which is much greater than the Sr isotope of ∼0.707 or less for the Sturtianaged cap dolostones (Halverson et al. 2007), suggesting that the 87Sr/86Sr values do not represent the Sr isotope compositions of pristine meltwater and/or ocean water after the Sturtian glaciation but altered signals.

Changes of the ocean chemical stratification
In the YI section, the δ 13 C PDB values of cap dolostones display a distinct depletion from bottom to top and the Zn/Y of the acid-insoluble residua presents relatively high ratios from 0.5 to 1.5 m. The values of zinc and ytterbium usually display a correlation during weathering (Nesbitt and Markovics 1997). However, the ratios of Zn/Y in the acid-insoluble residua are 46.99, 98.53, and 262.53 at 0.5, 1, and 1.5 m, respectively, then drop to 9.39 at 2 m, and keep stable above 2 m with an average value of 7 (Fig. 4b). Nesbitt and Markovics (1997) found that the ratios of Cu/Ti, Ni/Ti, Zn/Ti, and Y/Ti always show similar variation during weathering. These ratios presented glacial/interglacial cycles together in the sediments of the South China Sea (Wei et al. 2004;Liu et al. 2012). The ratios are higher during interglacials and lower during glacials. In this section, these ratios in the acid-insoluble residua are high at the bottom but low at the top, while the Zn/Ti ratios and their decrease are different from the others (Fig. 4a). As Cu, Ni, Zn, and Y have similar values in the upper continental crust (Taylor and McLennan 1995), the significant difference of Zn/Ti could be inferred that zinc has a second source while Cu, Ni, and Y do not have. Kuss and Kremling (1999) considered that the excess fluxes of Cu, Ni, and Zn are linearly related to organic carbon in the northeast Atlantic sediments (Kuss and Kremling 1999). Especially, Zn is an important element in organisms and an essential element constituting kinds of enzymes (Anderson et al. 1978). For example, in algae, the contents of Cu, Ni, and Y are similar, but the content of Zn is about ten times of their (Guo et al. 2003;Xie et al. 2003). In the model of the ocean stratification, the deep water was enriched with organics during glacials (Shen et al. 2005,   Giddings and Wallace 2009). Therefore, the enrichment of zinc in the acid-insoluble residua is perhaps related to the organic-enriched deep water. The organic-enriched deep water is likely to be the second source of Zn in the acidinsoluble residua.
The δ 13 C negative shift in Sturtian cap dolostones is a global event and is also reported in Australia and Namibia (Hurtgen et al. 2002;Giddings and Wallace 2009). During Sturtian glaciation, the presence of large ice sheet isolated the air and seawater and built up the anoxic, alkaline, and organic-enriched deep water. When the snowball earth unfroze, a large number of freshwater flowed into the sea. The stratification of seawater was destroyed by cold freshwater and upwelling deep water, and organic carbon was oxidized (Giddings and Wallace 2009). The stratified and mixed states in the post-glacial ocean were also reported in Marinoan cap carbonates (Halverson et al. 2005(Halverson et al. , 2006. Therefore, we suggest the Altungol cap dolostones were deposited from the water which was rich in organic carbon. The cap dolostones which formed earlier show more negative δ 13 C PDB and higher ratios of Zn/Y and Zn/Ti of the acid-insoluble residua. As the organics had been consumed,  the δ 13 C PDB was rising, but the ratios of Zn/Y and Zn/Ti of the acid-insoluble residua were declining. The Eu/Eu* data of whole rock shows negative anomaly (except YI-12-10), especially in the samples below 2 m (Eu/ Eu* <0.55). As the Eu 3+ /Eu 2+ is controlled by oxidation (Kamber and Webb 2001;Bolhar et al. 2004;Frimmel 2009;Zhao and Jones 2013), the negative anomaly Eu/Eu* reflects an alkaline and oxic environment. Thus, the strong negative anomaly of Eu below 2 m may indicate an oxic sedimentary environment, which is probably caused by the oxic melting water. The negative Eu anomaly diminished in the samples above 2 m, suggesting the weakening of oxidation. In addition, the high ∑REE (>10) and no Ce anomalies of cap dolostones (the samples below 3 m) reflect the input of continental silicate detritus (Chen et al. 2015) carried by the meltwater. Alternatively, the samples above 3 m with the positive Ce anomaly may be due to anoxic condition (Ling et al. 2013). This anoxic condition probably reflects the rising sea levels caused by melting glaciers.
The chemical index of alteration (CIA = [Al 2 O 3 /(Al 2 O 3 + CaO + Na 2 O + K 2 O)] × 100) (Nesbitt and Young 1982) and chemical variation (ICV = (Fe 2 O 3 + K 2 O + Na 2 O + CaO + MgO + MnO + TiO 2 )/Al 2 O 3 ) (Cox et al. 1995) have been widely used to quantify the degree of chemical weathering and to reconstruct the paleoweathering conditions of terrigenous sediments (Bock et al. 1998;Roddaz et al. 2006;Schoenborn and Fedo 2011;Jian et al. 2013). The values of CIA increase with chemical weathering enhancement (Nesbitt and Young 1982). The CIA = 50∼65 means that the environment is dry and cold and chemical weathering is weak. The CIA = 65∼85 shows that the environment is warm and wet and chemical weathering is medium. The CIA = 85∼100 means that the environment is hot and humid with strong chemical weathering.
The CIA of the cap dolostones of acid-insoluble residua in this section ranges from 72 to 77 and rises from bottom to top, indicating that the samples are formed from a warm and wet environment and moderate chemical weathering condition (Fig. 4f). The values of ICV drop with chemical weathering enhancement, and 1 is the boundary between strong and weak chemical weathering. (Cox et al. 1995). The ICV value drops from ∼3 at bottom to ∼1 at top (Fig. 4g). Both ICV and CIA show the same trend that chemical weathering increases from bottom to top in the YI section.
Changes of continental climate Hoffman and Schrag (2000) suggested that when tropical oceans thawed, evaporated seawater would work along with carbon dioxide to produce a more intense greenhouse. Surface temperatures would soar to more than 50°C, driving an intense cycle of evaporation and rainfall (Hoffman and Schrag 2000). We consider that the evidence of freshwater can be found in the acid-insoluble residua. As the acid-insoluble Fig. 7 a Cross-plots of Ce/Ce*-Eu/Eu*. b Cross-plots of Ce/Ce*-Dy N /Sm N . c Cross-plots of Ca-Eu. d Cross-plots of Ce/Ce*-Pr/ Pr*. Field A, neither Ce nor La anomaly; field B, positive La anomaly and no Ce anomaly; field C, negative La anomaly and no Ce anomaly; field D, true positive Ce anomaly; and field E, true negative Ce anomaly residua depended on weathering residues on the surface of the earth, the CIA and ICV can reflect the weathering degree of earth surface after glaciation. The values of CIA and ICV in this study indicate a warm and wet environment and moderate chemical weathering (Nesbitt and Young 1982).
The iron enrichment in the cap carbonate has been reported in the Rasthof Formation after the Chuos glaciation in Namibia (Hurtgen et al. 2002) and the Etina Formation after Sturtian glaciation in Australia (Swanson-Hysell et al. 2010). Swanson-Hysell et al. (2010) believed that when ice sheets retreated from the Sturtian glaciation in Australia, the vigorous weathering of freshly exposed continental crust would result in a high proportional delivery of Fe to S into the ocean (Hurtgen et al. 2002). In this study, the values of iron maintain at a high level in whole rock. For the acid-insoluble residua results, the values of iron are higher in the samples of 0-1.5-m intervals than 1.5-5-m intervals (Fig. 4d, e). It means that the sources of iron were different between below 1.5 m and above 1.5 m. At the initial stage of Altungol cap dolostone formation, the main source of iron is the weathering residues of the parent rocks. With the continental regolith which accumulated during the ice age was exhausted by meltwater, contents of Fe, Zn, Ni, and Cu in acid-insoluble residua decreased. Meanwhile, the vigorous weathering of freshly exposed continental crust provided a new source of iron in the warm and wet environment.
The Altungol cap dolostones have high ratios of 87 Sr/ 86 Sr (∼0.7094) and the negative correlation between 87 Sr/ 86 Sr and δ 18 O PDB . Liu et al. (2013) suggested that the elevated 87 Sr/ 86 Sr ratios could be understood by the interactions with brines/saline groundwaters. These groundwaters were probably formed by meteoric water (McKirdy et al. 2001) or ice melt seeps (Harris et al. 2007). An alternative explanation for the elevated 87 Sr/ 86 Sr ratios of cap dolostones is that these dolostones are formed in the glacial meltwater (Shields 2005;Hoffman 2011;Liu et al. 2013). Both explanations reflect that the Altungol cap dolostones are formed with a large amount of meteoric water or meltwater.

Conclusions
In this study, we reported new geochemical results of the Altungol cap dolostones. The enriched Zn of the acidinsoluble residua and depleted δ 13 C of the whole rock is the evidence of the ocean stratification theory (Shen et al. 2008(Shen et al. , 2010Giddings and Wallace 2009;Shen et al. 2005). The element geochemistry (CIA, ICV, Cu, Ni, Zn, and Fe) of the acid-insoluble residua displays a warm and wet environment of post-glacial continent. The high ratios of 87 Sr/ 86 Sr and the negative correlation between 87 Sr/ 86 Sr and δ 18 O PDB reflect that the Altungol cap dolostones are formed with a large amount of meteoric water or meltwater. The high ∑REE (>10) and the absence of Ce-negative anomalies of cap dolostones also reflect input of meltwater. The changes of Eu/Eu* and the true positive Ce anomaly in the samples above 3 m probably reflect the rising sea levels caused by melting glaciers. All of these characteristics and the other geochemical data can be explained by coupling environmental changes in post-glacial ocean and continent, in the Tarim Basin (Fig. 8).
Following the Altungol glaciation, glaciers thawed and meltwater changed ocean chemistry. Oxidation of upwelling a During the Altungol glaciation, the ocean was covered by the ice sheet and the bottom water was anoxic, alkaline, and enriched in organics. b When the Snowball Earth thawed, meltwater destroyed the ocean stratification. Deeper water upwelled and was oxidized, causing the water to be enriched in Zn and depleted in δ 13 C. Weathering residues, enriched in Fe, Zn, Ni, and Cu, were carried into the ocean by surface runoff. c The intense greenhouse drove an intense cycle of evaporation and rainfall. The weathering residues of the parent rocks were exhausted by meltwater. The vigorous weathering of freshly exposed continental crust provided a new source of iron. The ratios of 87 Sr/ 86 Sr also reflect that the Altungol cap dolostones are formed with a large amount of meteoric water or meltwater organic carbon-rich deep-water caused δ 13 C depletion and Zn enrichment. On the continents, deeply weathered regolith, enriched in Fe, Zn, Ni, and Cu, was delivered to the ocean by surface runoff.