Evidence of Arctic methane emissions across the mid-Pleistocene

The study investigates whether changes in Arctic ice sheet dynamics during the Quaternary, particularly across the mid-Pleistocene transition (MPT; ~1.3–0.7 Ma), influenced submarine methane emissions. The MPT marks a shift from 41-kyr to 100-kyr glacial cycles, accompanied by thicker ice sheets without a corresponding change in orbital forcing, implying strong internal climate feedbacks (ocean circulation, sea ice extent, ice sheet dynamics) and decreased atmospheric CO2. The Barents and Kara Seas shelves were prime locations for large Eurasian marine-based ice sheets, with evidence for intensified glaciation and iceberg grounding between ~950 and ~800 ka. While methane stored as free gas and hydrates in marine sediments is sensitive to pressure-temperature changes from glaciation, direct evidence for pre-LGM methane outbursts in the circum-Arctic has been lacking. This study aims to determine if ‘surplus’ glaciations and evolving thicker ice sheets prior to and across the MPT impacted the stability of Arctic gas hydrates and triggered episodic methane leakage.

Previous work documents widespread methane and gas hydrate reservoirs on formerly glaciated continental margins and major methane expulsion during rapid deglaciations after the Last Glacial Maximum (~22–18 ka BP) and penultimate deglaciation (~140 ka BP) in the Arctic. The MPT involved a transition to thicker ice sheets and enhanced glaciations, with evidence of increased Antarctic and Arctic ice volumes and expanded sea ice. Despite these insights, the effect of MPT glaciations on methane hydrate stability in the Arctic remained unknown due to limited greenhouse gas records, especially from older ice cores. Seismic studies in the Yermak Plateau region indicate gas accumulations and bottom-simulating reflections (BSRs), and modeling suggests episodic release of fossil hydrocarbons since ~1.5 Ma, potentially linked to tectonic stress changes associated with glacial isostatic adjustments.

• Study area and core: Ocean Drilling Program (ODP) Leg 151, Hole 912A (79°57.557′N, 5°27.360′E; 1037 m water depth) on the southern Yermak Plateau, NW Svalbard. A 145.4-m-thick sequence of silty clays and clayey silts representing the last ~2.2 Ma.
• Chronology and stratigraphy: Age model based on magnetostratigraphy (Matuyama, Jaramillo, Cobb Mountain, Olduvai Subchrons) integrated with biostratigraphy and high-resolution seismic correlation among ODP Holes 910C, 911A, and 912A. Sedimentation rates between magnetic boundaries: ~3–13 cm kyr−1 (constant ~8.7–9.2 cm kyr−1 between lower Subchron boundaries). Benthic δ18O (Cassidulina teretis–neoteretis) correlated to LR04 stack to identify interglacial MIS 47, 53, 55, 57 (~1.42–1.64 Ma) and constrain the MPT interval.
• Foraminiferal stable isotope analyses: Species analyzed included benthic (C. teretis–neoteretis, C. reniforme, Melonis barleeanus, Fursenkoina cf. acuta, Stainforthia feylingi) and planktonic (Neogloboquadrina pachyderma sinistral, N. atlantica). Sample counts: C. teretis–neoteretis (n=142), M. barleeanus (n=27), F. cf. acuta (n=10), S. feylingi (n=4), N. pachyderma sin (n=110), 10–20 tests per sample, >125 µm fraction. Analyses performed using Thermo Scientific MAT253 IRMS with GasBench II; precision ±0.03‰ for δ13C and ±0.08‰ for δ18O (relative to VPDB).
• Microscopy and mineralogical analyses: SEM and backscatter imaging of foraminiferal tests, with energy-dispersive X-ray spectroscopy (EDS) and electron backscattered diffraction (EBSD) to identify secondary methane-derived authigenic carbonate (MDAC) overgrowths (≥10 µm thick) and framboidal pyrite within tests. Instrument: Zeiss Merlin VP Compact SEM with Oxford Instruments X-Max 80 EDS; samples mounted and polished; analyses at 20 kV, 9–10 mm working distance.
• Biomarker analyses: Targeted AOM-related biomarkers including archaeol and sn-2-hydroxyarchaeol (archaea) and iso- and anteiso-C15:0 fatty acids (sulfate-reducing bacteria). Seven samples selected where foraminiferal assemblages changed and δ13C showed negative excursions. Lipids extracted from 20–50 g freeze-dried sediment; concentrations and δ13C measured to assess methane oxidation signatures.
• Total sulfur (TS) and ice-rafted debris (IRD): TS measured by LECO SC-632 after combustion at 1350 °C; IRD quantified in 500–1000 µm fraction after wet sieving, expressed as % of bulk.
• Seismic data: High-resolution profiles showing BSRs and high-amplitude reflections indicative of gas and gas hydrates; fault mapping to infer migration pathways near Sites 910C and 912A.
• Gas hydrate stability modeling: Simulated gas hydrate stability zone (GHSZ) at Site 912 between 0.7 and 2.0 Ma, incorporating transient sea-level variations and time-varying sedimentation/IRD deposition. Modeled base GHSZ depth shifts assessed to evaluate potential pore pressure changes and methane release mechanisms.

• Three distinct methane emission episodes (ME I–III) identified prior to and across the MPT: • ME I: ~1.66–1.57 Ma. • ME II: ~1.26–1.04 Ma. • ME III: ~0.81–0.72 Ma.
• Foraminiferal δ13C excursions indicative of methane influence: • Benthic species: values down to −28.91‰ (Fursenkoina cf. acuta) during ME II; C. teretis–neoteretis as light as −1.06‰ (ME II) and −0.78‰ (ME I). • Planktonic N. pachyderma: −3.24‰ to −12.91‰ during ME III.
• Microscopy/EDS/EBSD show pervasive secondary MDAC overgrowths (≥10 µm thick) on foraminiferal tests and framboidal pyrite within tests at ME I–III, consistent with AOM near the SMTZ.
• Biomarker evidence of AOM: • Presence of archaeal lipids (archaeol, OH-archaeol) and SRB fatty acids (iso-/anteiso-C15:0) within the three intervals. • Archaeol δ13C strongly depleted: −93.7‰ at ~1.6 Ma and −66.3‰ during the MPT; SRB fatty acids less depleted (iso-C15:0 −25.3‰, anteiso-C15:0 −26.3‰), suggesting organic matter degradation rather than direct AOM involvement for SRB. • Reported biomarker concentrations indicate enhanced microbial activity associated with methane oxidation within ME intervals (e.g., iso- and anteiso-C15:0 up to ~1.67 and 2.85 µg g−1 dry weight).
• Seismic evidence: BSRs and high-amplitude reflections adjacent to faults near Site 912 indicate shallow gas and hydrate accumulations fed by deeper reservoirs, with faults as likely conduits.
• GHSZ modeling shows base shifts mostly <20 m over 0.7–2.0 Ma due to sea level and sedimentation changes—insufficient to explain large, episodic seafloor emissions, pointing to tectonic/glacial stress as key triggers.
• Temporal alignment: Methane emission episodes coincide with phases of increased ice volume, shelf-edge glaciations, crestal erosion, and iceberg plough marks on the Yermak Plateau, and enhanced IRD/dropstones in Hole 912A.

The combined geochemical, microstructural, biomarker, seismic, and modeling evidence demonstrates episodic methane release on the Yermak Plateau aligned with Arctic ice sheet evolution prior to and across the MPT. The extreme negative δ13C in foraminiferal tests, MDAC overgrowths, pyrite, and AOM-specific biomarkers indicate syn-sedimentary AOM close to the SMTZ during ME I–III. Seismic BSRs and fault-associated reflectors, together with hydrate stability modeling, suggest that sea-level-driven GHSZ shifts alone cannot account for the magnitude and timing of emissions. Instead, glacially induced tectonic stress associated with glacial isostatic adjustments (forebulge formation, ice growth and retreat) likely reactivated faults, creating transient migration pathways from deep, old carbon reservoirs (e.g., gas hydrates and shallow gas accumulations) to the seafloor. These findings address the central question by linking Arctic ice sheet dynamics during the MPT to methane leakage episodes and imply that similar mechanisms may have operated across other circum-Arctic margins with extensive old carbon stores.

This study provides direct multiproxy evidence for three episodes of Arctic seafloor methane emissions prior to and across the mid-Pleistocene transition, temporally associated with major changes in Northern Hemisphere ice volume. By integrating foraminiferal stable isotopes, microtextural/mineralogical analyses, AOM biomarkers, seismic observations, and hydrate stability modeling, the work identifies glacially driven tectonic stress and fault reactivation as the primary mechanism for episodic methane leakage, rather than sea-level-induced hydrate stability changes alone. The results imply that widespread old carbon reservoirs beneath formerly glaciated Arctic margins were repeatedly perturbed during Pleistocene glacial cycles, leading to persistent seepage pathways. Future research should aim to constrain the atmospheric impact of these emissions via improved high-resolution greenhouse gas records (particularly older Arctic/Greenland ice cores), expand spatial assessments of seepage timing and magnitude across the Arctic, and refine models coupling GIA-induced stress, fault dynamics, and fluid flow.

• Atmospheric implications remain unresolved due to paucity of greenhouse gas data from older ice cores; direct linkage to past atmospheric CH4 is not established.
• Evidence is primarily from a single core/site (ODP Hole 912A) in the Yermak Plateau; regional extrapolation, while supported by seismic and modeling, requires broader sampling.
• Hydrate stability modeling indicates modest GHSZ base shifts (<20 m), necessitating assumptions about stress-driven fault reactivation to explain emissions.
• Biomarker interpretations acknowledge that SRB fatty acids are not strongly 13C-depleted, and AOM may be archaeal-only in these settings; microbial community composition through time is inferred rather than directly observed.
• Age model and δ18O correlations in Arctic sediments can be erratic; although multiple stratigraphic constraints were applied, age uncertainties remain.