At the heart of this investigation lie branched glycerol dialkyl glycerol tetraethers (brGDGTs) and long-chain n-alkanes, molecular fossils derived from bacterial and terrestrial plant sources, respectively. These biomarkers serve as unique fingerprints that allow scientists to distinguish inputs of organic carbon originating from soils and vascular plants, versus that synthesized or deposited within marine systems. By compiling and analyzing brGDGT and n-alkane data from multiple PETM-aged sediment cores spread across geographically diverse sites—from the High Arctic to the Tasman Sea—the researchers bridged geographic and depositional contexts to create a comprehensive view of carbon sourcing.

The methodology builds on carefully extracted lipid fractions separated via microwave-assisted and accelerated solvent extractions, followed by advanced chromatographic and mass spectrometric techniques. Gas chromatography tandem mass spectrometry (GC-MS/MS) was leveraged for hydrocarbon analyses such as n-alkanes, while high-performance liquid chromatography mass spectrometry (HPLC-MS) targeted polar fractions containing brGDGTs. These high-resolution extraction and detection methods ensured robust molecular quantification critical for downstream proxy calculations.

Several key indices were calculated to differentiate terrestrial versus marine organic carbon sources. The branched-to-isoprenoid tetraether (BIT) index quantitatively captures the balance between soil-derived branched GDGTs and marine archaeal crenarchaeol, with soil endmember values derived from an extensive modern peat and soil database, and marine endmember values from distal ocean sediments. Complementarily, the terrestrial-to-aquatic ratio (TAR), based on the abundance of long-chain versus short-chain n-alkanes, serves as a proxy for vascular plant contributions. These indicators were integrated into a rigorous three-endmember mixing model, facilitating the quantification of soil, plant, and marine fractions in PETM sedimentary organic carbon.

A significant innovation in the study is how the mixing model accounts for conservative tracers insensitive to certain endmembers, enabling refined calculations even when biomarkers partially constrain source inputs. This non-negative least squares approach, solved within a Monte Carlo framework incorporating endmember uncertainty and allowing for minor contributions from uncharacterized sources, generates statistically robust source apportionment. The approach was validated against modern Gulf of Mexico sediments, demonstrating high fidelity in reconstructing known terrestrial versus marine carbon contributions.

Quantifying historical carbon mass accumulation rates (MARs) involved integrating fractionated organic carbon proportions with independently derived sedimentation rates and bulk densities from multiple PETM sedimentary records. Standardizing dry bulk density assumptions based on typical detrital and clay mineral densities ensured consistency across sites. This multi-site MAR assessment revealed nuanced shifts in terrestrial carbon burial efficiency during the PETM, reflecting changes in sediment supply, organic carbon preservation, and coastal-marine interactions under extreme warming conditions.

Climate modeling experiments provided the contextual framework for interpreting these geochemical findings. Utilizing the DeepMIP ensemble within the Paleoclimate Modelling Intercomparison Project (PMIP4) and the Coupled Model Intercomparison Project Phase 6 (CMIP6), simulations were executed with varying atmospheric CO₂ concentrations, ranging from pre-industrial levels to multiples up to sixfold. Normalized mean annual precipitation fields extracted from multi-model means enabled the assessment of hydrologic impacts of elevated greenhouse gas forcing on terrestrial and marine carbon dynamics.

Further nuance is provided by the Hadley Centre’s coupled atmosphere-land model configuration incorporating the MOSES2.1 land surface scheme and TRIFFID dynamic vegetation and carbon cycle model. Spanning two key atmospheric CO₂ scenarios bracketing the PETM transition, this framework allowed estimation of changes in vegetation distribution, soil carbon stocks, and ecosystem productivity under palaeoclimate boundary conditions. These models affirm the plausibility of enhanced terrestrial organic carbon export and subsequent marine burial under PETM warming, as observed in empirical sediment biomarker records.

The sophisticated integration of geochemical proxies, sedimentological data, and model simulations collectively advances our understanding of carbon cycling feedbacks during hyperthermal intervals. Accelerated burial of terrestrial organic carbon in marine sediments would have acted as an important carbon sink mechanism, potentially modulating atmospheric CO₂ levels and climate evolution through the PETM. This insight bridges a critical knowledge gap regarding organic carbon fluxes from continents to oceans during rapid warming events in Earth’s deep past.

Moreover, the study’s methodological framework—combining conservative tracer endmember mixing with uncertainty quantification—represents a transferable approach applicable to other geological intervals characterized by coupled climate-carbon perturbations. It highlights the vital role of molecular-level biomarker analyses in reconstructing paleoecosystem structure and sedimentary carbon sources, deepening our ability to predict carbon cycle sensitivity to future climatic changes.

These findings challenge simplistic interpretations of terrestrial carbon loss during warming, revealing instead that significant portions were transferred and buried in marine environments, emphasizing complex cross-system connectivity. This underlines the importance of integrated biogeochemical and climate modeling perspectives to reconstruct feedbacks governing Earth system resilience and tipping points.

Crucially, the validation of biomarker endmembers via modern analogs—such as the Mississippi–Atchafalaya River system—strengthens confidence in the approach’s accuracy and relevance. By demonstrating model outputs consistent with contemporary river-coastal sediment organic carbon distributions, the study underscores the robustness of proxy-based quantitative source apportionment.

This multidisciplinary effort pushes the frontier in paleoenvironmental science, with implications extending from fundamental deep-time carbon cycle reconstructions to informing projections of future organic carbon dynamics under anthropogenic climate forcing. It exemplifies how state-of-the-art analytical chemistry, sedimentology, and Earth system modeling converge to unravel complex Earth processes critical to understanding planetary climate stability.

Further research building on these results might explore finer temporal resolution biomarker studies, integration with isotopic measurements, and expanded geographic coverage to refine the spatial heterogeneity of organic carbon sources and burial efficiencies. Additionally, investigating mechanisms underpinning organic matter preservation in marine sediments under elevated temperatures could illuminate pathways controlling long-term carbon sequestration.

In conclusion, this seminal work presents unprecedented insights into the enhanced burial of terrestrial organic carbon in marine settings during the PETM, a period typified by rapid, profound climate change. Through an elegant fusion of biomarker geochemistry and climate simulation, it opens new vistas in deciphering Earth’s carbon cycle responses and offers a potent analog to contemporary global warming scenarios.

Subject of Research: Enhanced marine burial of terrestrial organic carbon during the Palaeocene–Eocene Thermal Maximum

Article Title: Enhanced marine burial of terrestrial organic carbon through the Palaeocene–Eocene Thermal Maximum

Article References:
Inglis, G.N., Hemingway, J.D., Stockey, R.G. et al. Enhanced marine burial of terrestrial organic carbon through the Palaeocene–Eocene Thermal Maximum. Nat. Geosci. (2026). https://doi.org/10.1038/s41561-026-02012-2

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s41561-026-02012-2