Globally limited but severe shallow-shelf euxinia during the end-Triassic extinction

The Triassic/Jurassic boundary interval (~201 Ma) records the end-Triassic mass extinction (ETME), one of the largest losses of complex marine life. The ETME has been closely associated with Central Atlantic Magmatic Province volcanism, linked to atmospheric carbon injection evidenced by negative carbon isotope excursions, as well as marine acidification and de-oxygenation. Multiple lines of evidence (biomarkers, δ³⁴S, prasinophyte blooms, carbonate uranium isotopes, widespread Early Jurassic black shales, elemental redox proxies, iron speciation) indicate locally sulfidic and oxygen-poor conditions in marginal marine settings of the Tethys and Panthalassa around the boundary. Yet most Late Triassic redox studies are local to sub-regional, leaving the global-scale distribution of anoxia poorly constrained. The isotopic composition of molybdenum in sediments is a powerful proxy for reconstructing local and global marine redox, with distinct behaviors under oxic, suboxic, ferruginous, and sulfidic conditions. No observations of δ⁹⁸Mo evolution existed across the TJB and ETME prior to this study. This work aims to resolve global versus local redox conditions during the ETME using Mo concentrations and isotopes from well-dated Tethyan shelf records to assess the role and spatial extent of de-oxygenation in driving marine extinction.

Previous research links the ETME to CAMP volcanism, carbon-cycle perturbations, and marine environmental stress including acidification and de-oxygenation. Local to regional evidence for oxygen-poor to sulfidic conditions comes from biomarkers and δ³⁴S indicating photic-zone euxinia in Tethyan and Panthalassan marginal seas; blooms of prasinophycean algae; negative excursions in carbonate δ²³⁸U; widespread Early Jurassic black-shale deposition; elemental redox proxies; and iron speciation. Uranium isotope studies have been interpreted to reflect expanded anoxia (uranium reduction) over ~8–20% of the global seafloor near/after the ETME, though they provide limited constraints on the prevalence of sulfidic conditions and may largely reflect marginal and intermediate-depth settings. Despite these advances, the global-scale pattern and severity of sulfidic anoxia across the TJB remained unconstrained due to the paucity of global proxies such as δ⁹⁸Mo spanning the interval.

Materials: Sedimentary successions spanning pre-, syn-, and post-ETME time from three Tethyan shelf cores were analyzed: Carnduff-2 (Larne Basin, Northern Ireland), Hebelermeer-2 (Germanic Basin, west Germany), and Schandelah-1 (Lower Saxony Basin, north Germany). These records comprise marine marls, sandstones, and organic-rich shales with established stratigraphy and biostratigraphic correlation. Analytical approach: The study used sedimentary molybdenum enrichment (Mo, MoEF) and molybdenum isotopes (δ⁹⁸Mo) to reconstruct local depositional redox and infer coeval seawater δ⁹⁸Mo (δ⁹⁸Mo_sw) to assess global redox. Detrital input effects were evaluated, with little correlation found between detrital proxies and δ⁹⁸Mo or total Mo, supporting a redox control on Mo records. Bulk inorganic geochemistry: ~50 mg per sample was digested with inverse aqua regia (3:1 HNO₃:HCl) followed by 1:2 HF:HNO₃ for silicates. Major elements were measured by ICP-AES (Perkin Elmer Optima 3300RL) and trace elements by ICP-MS (Perkin Elmer NexION 350D). Accuracy assessed with MAG-1 (±2.5% with Mo within certified uncertainty); precision from MAG-1 (±17% for most, Cr, Cu, Zn, Al ±<10%) and in-house mudrock standard PERN-1 (±21%). Mo enrichment relative to upper continental crust was calculated as (Mo_sample/Al_sample)/(Mo_crust/Al_crust). Molybdenum isotopes: Samples were spiked with a ⁹⁵Mo–¹⁰⁰Mo double spike and digested (3:1 HNO₃:HCl). Mo was separated using anion exchange chromatography. Isotope ratios were measured on a Thermo Neptune Plus MC-ICP-MS. Procedural blanks were 0.5–5.0 ng. Data are reported relative to NIST SRM 3134 (+0.25‰). Accuracy and precision were verified using the Open University Mo solution standard (−0.35 ± 0.15‰, 2 s.d., n=12) and multiple digests of USGS SDO-1 shale (1.04 ± 0.08‰, 2 s.d., n=6), consistent with published values. Stratigraphic, lithologic, and correlation frameworks followed prior studies cited for each core, including carbon isotope stratigraphy and biostratigraphic markers. Interpretive framework: Local depositional redox was inferred from Mo concentrations/enrichments and δ⁹⁸Mo behavior under suboxic, ferruginous, and sulfidic conditions, considering fractionation associated with oxyhydroxide adsorption, porewater sulfide availability (thiomolybdate formation), and potential intermediate thiomolybdate burial at H₂S <11 µM. Coeval seawater δ⁹⁸Mo_sw was estimated from upper-bound sediment δ⁹⁸Mo values deposited under non-euxinic conditions, applying a minimum fractionation of ~−0.7‰ between porewater sulfide-precipitated Mo and seawater, supported by independent iron speciation constraints.

• Global extent of sulfidic anoxia remained limited across the TJB: Maximum Late Triassic mudstone δ⁹⁸Mo values are ~−1.56‰ (Carnduff-2) and ~−1.63‰ (Hebelermeer-2) from levels indicating suboxic deposition. Given ≥0.7‰ fractionation between sediment and seawater under such conditions, Late Triassic δ⁹⁸Mo_sw was probably >2.3‰, similar to or heavier than modern seawater, implying sulfidic conditions covered only ~0.05–0.10% of the seafloor (similar to or less than today).
• Persistence into earliest Jurassic: Basal Jurassic (Carnduff-2) upper-bound δ⁹⁸Mo averages 1.47 ± 0.58‰ (n=3) under localized suboxia; thus δ⁹⁸Mo_sw was probably >2.2‰, indicating no substantial change in the global Mo cycle across the TJB and continued spatial limitation of sulfidic anoxia in the Early Jurassic open ocean.
• Regional, pulsed de-oxygenation on the Tethyan shelf: Schandelah-1 and Carnduff-2 show positive δ⁹⁸Mo shifts coincident with the main extinction interval (initial CIE), with increased Mo enrichment, indicating increased [HS⁻] and shoaling of the sulfate reduction zone toward the sediment-water interface. Hebelermeer-2 records elevated MoEF with low δ⁹⁸Mo near the base of the main extinction interval, consistent with intermediate thiomolybdate burial at low H₂S (<11 µM), indicating water-column sulfide presence.
• Multiple pulses of marginal-marine anoxia/euxinia: Pulses occurred during deposition of the basal and upper Westbury Formation (and equivalents) in the middle Rhaetian, coeval with photic-zone euxinia on the Tethyan shelf and episodic photic-zone euxinia in the Bristol Channel Basin and denitrification in parts of the Central European Basin. Further positive δ⁹⁸Mo and Mo_red shifts occur in the basal Jurassic of Carnduff-2 (and correlative shifts in Schandelah-1), indicating repeated shoaling of the sulfate reduction zone and transient oxygen-poor conditions.
• Coupling to biotic crises: Pulses of de-oxygenation are temporally coincident with Late Triassic extinction phases across multiple groups, suggesting a causal link between localized sulfidic events and shallow-marine extinctions.
• Reconciliation with U isotopes: While carbonate δ²³⁸U indicates expanded anoxia (uranium reduction) over ~8–20% of seafloor near/after the ETME, Mo isotope data show little increase in globally sulfidic waters, implying that much of the anoxia was non-sulfidic and concentrated in marginal/intermediate-depth settings rather than the open ocean.

The study resolves a key uncertainty in the ETME: whether anoxia, especially sulfidic conditions, expanded globally or was geographically restricted. High estimated δ⁹⁸Mo_sw (>2.2–2.3‰) across the Late Triassic into the earliest Jurassic indicates that globally sulfidic waters did not expand beyond modern-like areal extents, thus ruling out pervasive global euxinia as the primary kill mechanism. Instead, Mo isotope and enrichment records from Tethyan shelf cores reveal repeated, localized pulses of de-oxygenation with increased porewater and episodic water-column sulfide, synchronous with extinction phases and major carbon-cycle perturbations (initial CIE). These results reconcile with uranium isotope evidence for broader anoxia by suggesting that expanded low-oxygen conditions were primarily non-sulfidic and focused in marginal and intermediate-depth environments, while open ocean settings likely remained comparatively well oxygenated and functioned as refugia. Mechanistically, enhanced continental weathering and runoff during warming and ecosystem disruption likely drove coastal eutrophication and stratification, and expansion of oxygen minimum zones along margins, producing geographically localized sulfidic episodes that disproportionately impacted shallow, biodiverse marine ecosystems. The findings highlight that global marine biodiversity and ecosystem stability can be vulnerable to regionally confined anoxic events, with implications for modern anthropogenic de-oxygenation and nutrient enrichment in marginal seas.

Using molybdenum concentrations and isotopes from well-correlated Tethyan shelf cores, the study shows that during the end-Triassic extinction, sulfidic conditions did not expand globally but instead occurred as severe, pulsed, and geographically localized events in marginal marine settings. Estimated seawater δ⁹⁸Mo values comparable to or heavier than modern imply limited global euxinia, while positive δ⁹⁸Mo excursions and Mo enrichments mark repeated shoaling of the sulfate reduction zone and episodic sulfide development that coincided with extinction pulses. These results refine the role of redox change in the ETME, reconciling Mo and U isotope perspectives, and underscore that regionally localized de-oxygenation can drive global-scale biodiversity crises. The study further indicates that ongoing anthropogenic expansion of marine anoxia and nutrient loading could have outsized impacts on biodiversity and ecosystem stability, particularly in sensitive marginal marine environments.

• Spatial coverage focuses on three Tethyan shelf cores (Larne, Germanic/Lower Saxony basins), with limited direct open-ocean records; extrapolation to global conditions relies on interpreting upper-bound sediment δ⁹⁸Mo to estimate seawater values.
• Interpretation depends on assumptions about isotopic fractionation (minimum ~−0.7‰) between porewater sulfide-precipitated Mo and seawater under non-euxinic conditions; local redox variability may influence fractionation magnitudes.
• Low marine sulfate concentrations in the Late Triassic have been proposed; although argued here to be sufficient for thiomolybdate formation and not to alter broad δ⁹⁸Mo trends, uncertainties in sulfate levels could affect local Mo isotope systematics.
• Isotopically light δ⁹⁸Mo can arise from intermediate thiomolybdate burial at low H₂S, complicating simple distinctions between oxic adsorption and sulfidic conditions; multi-proxy integration is required.
• Differences between Mo and U isotope inferences indicate that while anoxia expanded, the extent and nature (sulfidic vs non-sulfidic) vary by proxy; constraints on exact areal extents of anoxic subtypes remain uncertain.