Ocean deoxygenation after the Sturtian Snowball

The Cryogenian Period (ca. 717–635 Ma) includes the Sturtian and Marinoan Snowball Earth glaciations separated by a nonglacial interlude (~661–650 Ma) during which multicellular eukaryotes, including algae and possible sponges, rose toward ecological dominance. This interval shows sustained high carbonate carbon isotope values (δ13Ccarb) punctuated by negative excursions and has prompted hypotheses of a large marine dissolved organic carbon (DOC) pool due to sulfate-poor deep oceans. Conflicting proxies have suggested both widespread euxinia (Mo isotopes) and transient oxygenation (U isotopes), obscuring mechanistic links among climate, ocean redox, nutrient dynamics, and biological evolution after the Sturtian deglaciation. To resolve these uncertainties, the study re-examines well-preserved carbonates from the lower Taishir Formation (Zavkhan Terrane, western Mongolia), deposited for several million years after the Sturtian deglaciation, using a multi-proxy approach (C, S, Zn, Sr, U isotopes; REE; P and iodine) to reconstruct seawater chemistry, redox structure, nutrient supply, and weathering signals.

Prior work indicates enhanced volcanism, brief intense chemical weathering, and global cooling following a postglacial super-greenhouse. Iron speciation suggests a short-lived euxinic interval, Mo isotopes from siliciclastics support extensive marine euxinia, while U isotopes from carbonates indicate transient oxygenation—an apparent contradiction. High δ13Ccarb has been linked to increased organic burial and oxygenation via glacially induced nutrient surpluses, and to possible DOC buildup in sulfate-poor deep oceans. Sr isotope anomalies over cap carbonates have been tied to meltwater plumes in some regions, but the Taishir interval lacks cap carbonate features and shows Sr values consistent with least radiogenic records, suggesting minimal meltwater plume influence. Previous datasets from Twitya and Rasthof, and age models for the nonglacial interlude, frame expectations for redox stratification, nutrient feedbacks (especially P recycling under euxinia), and asynchronous atmosphere–ocean oxygenation.

Field sampling targeted bulk limestones from the lower Taishir Formation; clean chips without veins or weathered surfaces were powdered. Sequential leaching isolated carbonate fractions for REE and P: ~50 mg powder prewashed with neutral 1 M ammonium acetate, rinsed, and partially leached with 0.3 M acetic acid; Ca–Mg–Fe–Mn–Al–P–Sr–Ba measured by ICP-OES (Varian 720) and REE by ICP-MS (Agilent 7900). REE were normalized to PAAS; Ce anomaly calculated as Ce/Ce* = CeN×NdN/PrN²; matrix/oxide interferences monitored and corrected; precision better than 5%. Carbonate-associated iodine (CAI) was extracted by brief 3% HNO3 leaching of ~10 mg powder, stabilized with EDTA/ammonium hydroxide; Ca and Mg measured by ICP-OES; iodine measured by ICP-MS (Agilent 7900) with potassium iodate standards; I/Ca detection limit ~0.1 µmol/mol; external RSD ~3%; standards (JCp-1, CRM-393) validated accuracy. Carbonate-associated sulfate (CAS) was isolated via multi-step oxidative and ionic cleaning (NaClO, H2O2, NaCl), followed by dissolution in 6 M HCl and BaSO4 precipitation with BaCl2; δ34S of BaSO4 measured by EA-IRMS (Iso-Analytical) calibrated to VCDT with multiple standards; CAS concentrations computed from precipitate mass and purity; five samples yielded sufficient S for isotopes. Uranium isotopes: ~500 mg powder partially leached with 0.2 M acetic acid (24 h), filtered; elemental concentrations by ICP-OES/ICP-MS; ~40 ng U spiked with IRMM-3636 double spike, purified on UTEVA resin; δ238U measured by MC-ICP-MS (Neptune Plus, Aridus III) at Royal Holloway; reported relative to CRM-112A; typical propagated uncertainty 0.03–0.08‰ (2 SE); procedural blanks negligible. Zinc isotopes: ~200 mg powder leached with buffered 1 M ammonium acetate (pH 5, 24 h), elemental analysis by ICP-OES/ICP-MS; 64Zn–67Zn double spike added; Zn purified on AG MP-1M resin; δ66Zn measured on Nu Plasma 3 MC-ICP-MS (Aridus III) at UCL using standard–sample bracketing and double-spike correction; reported relative to AA-ETH and converted to JMC-Lyon with +0.28‰ correction; external reproducibility constrained by secondary standards and CRM-393. Petrographic screening and diagenetic assessments support primary seawater trends in stratigraphic profiles; age and sequence stratigraphic correlations follow published models.

• Local redox: Ce anomalies with seawater-like REE patterns indicate a transition from largely suboxic/anoxic seafloor conditions in the lower part (0–32 m; mean Ce anomaly ~1.11) to more oxic conditions in the upper part (50–120 m; mean ~0.78). Low sulfur in decarbonated residues suggests mostly non-sulfidic local waters during deposition of the lower part.
• Shallow-water redox: Carbonate I/(Ca+Mg) values are low (<0.5 µmol/mol), consistent with Proterozoic records indicating a redox-stratified water column with a shallow oxycline; diagenetic loss of iodine cannot be fully excluded.
• Global redox: Carbonate δ238U from integrated records shows a decreasing trend after the Sturtian deglaciation, indicating significant expansion of anoxic waters; coupled with extremely high δ34S_CAS (mean ~+63‰), implying enhanced microbial sulfate reduction and extensive pyrite burial. Later, δ238U shifts toward higher values (mean ~−0.49‰), interpreted as a return to less reducing conditions rather than widespread oxygenation, possibly due to low productivity under ferruginous conditions; concurrent decreases in δ34S_CAS and pyrite δ34S suggest increasing sulfate levels, though overall sulfate remained low (low Δ34S and low [CAS]).
• Zinc isotopes and nutrients: Carbonate δ66Zn shifts from +0.74‰ to ~+0.45‰, remains near +0.45‰ in the lower part, then increases upsection to ~+0.68‰. The lower δ66Zn coincides with rising 87Sr/86Sr, indicating elevated postglacial weathering and input of isotopically light Zn from continents; Zn/(Ca+Mg) increases in the lower part, inconsistent with a dominant euxinic sink. Upsection δ66Zn increases prior to redox easing, suggesting diminishing weathering-derived light Zn input and relatively increased burial of isotopically light organic-rich sediments.
• Phosphorus proxy: P/(Ca+Mg) is higher in the lower part (mean ~0.11 mmol/mol) with a peak near 0.2 mmol/mol toward its top, versus lower values in the upper part (mean ~0.08 mmol/mol). The peak aligns with high δ34S_CAS and low δ238U, indicating elevated dissolved phosphate due to intensified recycling and low burial efficiency under euxinic/sulfidic conditions.
• Weathering and climate context: Rapid postglacial silicate weathering of reactive rock flour on a low-relief supercontinent caused atmospheric CO2 drawdown, followed by prolonged cooling due to unusually low arc and ridge degassing; this drove initial eutrophication and later subdued nutrient/sulfate delivery.
• Carbon cycle and DOC: Sustained high δ13Ccarb is partly explained by buildup of a recalcitrant deep-ocean DOC pool during expanded anoxia and later nutrient limitation/cooling, consistent with muted δ13Corg shifts and buffering of oxygenation/CO2 drawdown.
• Biosphere: Lessening of marine anoxia, cooling, and limited nutrients expanded habitable space, supporting a rise in biotic complexity; severe postglacial environmental stress likely delayed diversification, with continued largely anoxic oceans favoring aerobic taxa with low oxygen demands.

The multi-proxy dataset resolves conflicting views of post-Sturtian ocean conditions by demonstrating a transient deoxygenation marked by expanded low-sulfate euxinia, followed by a shift to less reducing states. Uranium and sulfur isotopes jointly indicate widespread anoxia and intense pyrite burial immediately after deglaciation, while later δ238U increases reflect reduced reducing conditions rather than a globally oxygenated ocean, consistent with ferruginous influences and productivity controls. Zinc and phosphorus proxies show that intense postglacial weathering delivered nutrients (P, Zn) that triggered eutrophication, positive feedbacks via P recycling under euxinia, and enhanced organic export, collectively elevating oxygen demand and maintaining ocean deoxygenation despite pulses of oxygen production from organic burial. As weathering waned and climate cooled, oxygen solubility increased and nutrient recycling efficiencies dropped, easing marine euxinia and favoring less reducing conditions. The carbon cycle response includes accumulation of a deep-ocean DOC reservoir that helps explain sustained high δ13Ccarb and modulates oxygenation trajectories, with episodic DOC oxidation events linked to coupled negative δ238U–δ13Ccarb shifts. Biologically, the transition to cooler, less reducing but nutrient-limited oceans likely expanded habitable niches for low-oxygen-demand aerobic organisms, aligning with biomarker records of increasing ecological complexity while highlighting asynchronous atmosphere–ocean oxygenation.

This study integrates carbonate-based proxies (Ce anomalies, I/(Ca+Mg), δ238U, δ34S_CAS, δ66Zn, P/(Ca+Mg), 87Sr/86Sr) from the Taishir Formation to show that the immediate aftermath of the Sturtian deglaciation experienced transient ocean deoxygenation via expansion of low-sulfate euxinia and enhanced pyrite burial, driven by weathering-induced nutrient influx and eutrophication. Subsequently, diminishing weathering fluxes and climatic cooling reduced marine euxinia and returned the ocean to less reducing conditions without widespread oxygenation. Nutrient dynamics and redox feedbacks likely promoted a deep-ocean DOC buildup that contributed to sustained high δ13Ccarb and buffered oxygenation. These environmental trajectories help explain delayed biological diversification and the rise of aerobic taxa with low oxygen demands during the Cryogenian nonglacial interlude. Future work should refine uranium isotope fractionation under ferruginous, low-productivity regimes, expand geographically distributed high-resolution multi-proxy records, quantify sulfate concentrations and Δ34S through improved CAS recovery, and couple proxy-constrained models of weathering, nutrient delivery, DOC dynamics, and redox evolution to assess thresholds for ocean–atmosphere oxygenation.

• The article lacks a formal abstract and precise publication date within the provided text.
• Potential transient ocean heterogeneity immediately post-deglaciation (e.g., meltwater plumes) could affect regional chemostratigraphy, although the studied interval shows no cap carbonate features and Sr isotopes consistent with global baselines.
• Diagenesis may affect carbonate proxies, notably I/(Ca+Mg), and neomorphism could have reduced iodine; the study mitigates diagenetic overprints, but cannot fully exclude them.
• CAS extraction yielded sufficient sulfur for isotopic analysis in only five samples, limiting δ34S_CAS coverage; overall low [CAS] complicates precise sulfate reconstructions.
• U isotope interpretations under ferruginous/suboxic conditions are sensitive to variable fractionation linked to productivity and organic loading; global inference remains uncertain.
• REE patterns in the lower part include non-seawater-like signatures in some samples, requiring exclusions and careful interpretation.
• Age model choices and sequence stratigraphic correlations introduce uncertainties in temporal alignments across records.