Figures

top

Introduction

Reconstructing past variations in continental weathering both in rate and intensity is crucial for understanding and modelling changes in atmospheric and oceanic pCO₂ levels. Lithium isotopes (expressed as δ⁷Li) in riverine and marine systems have increasingly been recognised as a robust proxy for silicate weathering intensity. Given the Li residence time in the ocean (∼1.5 million years), Li isotopes are particularly well suited for investigating the global carbon cycle over extended geological periods.

Numerous studies attribute the rise in Cenozoic seawater δ⁷Li values deduced from carbonates to significant changes in riverine Li flux and/or its δ⁷Li composition. The riverine δ⁷Li signature is primarily governed by the dissolution of silicate phases and the formation of secondary clays. A strong connection between continental silicate weathering and climate has thus been inferred from carbonate δ⁷Li records (e.g., Misra and Froelich, 2012

Misra, S., Froelich, P.N. (2012) Lithium isotope history of Cenozoic seawater: changes in silicate weathering and reverse weathering. Science 335, 818–823. https://doi.org/10.1126/science.1214697

; Vigier and Goddéris, 2015

Vigier, N., Goddéris, Y. (2015) A new approach for modeling Cenozoic oceanic lithium isotope paleo-variations: the key role of climate. Climate of the Past 11, 635–645. https://doi.org/10.5194/cp-11-635-2015

). On shorter time scales, such as during Oceanic Anoxic Event 2 (Pogge von Strandmann et al., 2013

Pogge von Strandmann, P.A.E., Jenkyns, H.C., Woodfine, R.G. (2013) Lithium isotope evidence for enhanced weathering during Oceanic Anoxic Event 2. Nature Geoscience 6, 668–672. https://doi.org/10.1038/ngeo1875

), sedimentary carbonate records also reveal rapid δ⁷Li fluctuations (≪1 Myr). These variations have been interpreted as a temperature driven acceleration of silicate weathering, subsequently impacting the carbon cycle.

At the Cretaceous-Paleogene (K-Pg) boundary (∼66 Ma), δ⁷Li values in foraminifera drop by ∼5 ‰ in less than 400,000 years, a shift attributed to a major disruption in the marine Li budget in response to large changes in continental fluxes (Misra and Froelich, 2012

Misra, S., Froelich, P.N. (2012) Lithium isotope history of Cenozoic seawater: changes in silicate weathering and reverse weathering. Science 335, 818–823. https://doi.org/10.1126/science.1214697

). The transition from the late Maastrichtian to the Paleogene is characterised by profound environmental upheavals and one of the most devastating mass extinctions in Phanerozoic history. This catastrophic event is closely linked to two pivotal geological phenomena: the Chicxulub impact and the peak activity of the Deccan Traps volcanism (Schoene et al., 2019

Schoene, B., Eddy, M.P., Samperton, K.M., Keller, C.B., Keller, G., Adatte, T., Khadri, S.F.R. (2019) U-Pb constraints on pulsed eruption of the Deccan Traps across the end-Cretaceous mass extinction. Science 363, 862–866. https://doi.org/10.1126/science.aau2422

). Both events released substantial volumes of CO₂ and SO₂ into the ocean-atmosphere system, with the impact triggering abrupt climatic shifts and the volcanism sustaining emissions over 100–150 thousand years (Schoene et al., 2019

Schoene, B., Eddy, M.P., Samperton, K.M., Keller, C.B., Keller, G., Adatte, T., Khadri, S.F.R. (2019) U-Pb constraints on pulsed eruption of the Deccan Traps across the end-Cretaceous mass extinction. Science 363, 862–866. https://doi.org/10.1126/science.aau2422

; Sprain et al., 2019

Sprain, C.J., Renne, P.R., Vanderkluysen, L., Pande, K., Self, S., Mittal, T. (2019) The eruptive tempo of Deccan volcanism in relation to the Cretaceous-Paleogene boundary. Science 363, 866–870. https://doi.org/10.1126/science.aav1446

). These processes likely induced significant changes in ocean chemistry, biogeochemical cycles, and global environmental conditions (Henehan et al., 2019

Henehan, M.J., Ridgwell, A., Thomas, E., Zhang, S., Alegret, L., et al. (2019) Rapid ocean acidification and protracted Earth system recovery followed the end-Cretaceous Chicxulub impact. Proceedings of the National Academy of Sciences 116, 22500–22504. https://doi.org/10.1073/pnas.1905989116

). However, the precise timing and magnitude of these events, as well as their effects, remain debated.

Throughout the Phanerozoic, marine δ⁷Li records have largely relied on fossil calcium carbonates, such as foraminiferal shells, dolostones, and shallow water marine carbonates (Misra and Froelich, 2012

Misra, S., Froelich, P.N. (2012) Lithium isotope history of Cenozoic seawater: changes in silicate weathering and reverse weathering. Science 335, 818–823. https://doi.org/10.1126/science.1214697

; Kalderon-Asael et al., 2021

Kalderon-Asael, B., Katchinoff, J.A.R., Planavsky, N.J., Hood, A.V.S., Dellinger, M., et al. (2021) A lithium-isotope perspective on the evolution of carbon and silicon cycles. Nature 595, 394–398. https://doi.org/10.1038/s41586-021-03612-1

; Liu et al., 2023

Liu, X.-F., Liu, X.-M., Wang, X.-K., Zhai, S., Liu, X. (2023) Dolostone as a reliable tracer of seawater lithium isotope composition. Communications Earth & Environment 4, 58. https://doi.org/10.1038/s43247-023-00711-x

). A fundamental assumption is that they faithfully reflect the seawater δ⁷Li at the time of their formation. However, biogenic carbonates represent a negligible Li sink, and biological or environmental factors may influence their δ⁷Li values (Vigier et al., 2015

Vigier, N., Rollion-Bard, C., Levenson, Y., Erez, J. (2015) Lithium isotopes in foraminifera shells as a novel proxy for the ocean dissolved inorganic carbon (DIC). Comptes Rendus. Géoscience 347, 43–51. https://doi.org/10.1016/j.crte.2014.12.001

; Chen et al., 2023

Chen, D., Thibon, F., Felbacq, A., Weppe, L., Metian, M., Vigier, N. (2023) Coupled survey of lithium isotopes and Li/Ca in biogenic and inorganic carbonates. Earth-Science Reviews 244, 104500. https://doi.org/10.1016/j.earscirev.2023.104500

; Poet et al., 2023

Poet, M., Vigier, N., Bouret, Y., Jarretou, G., Gautier, R., et al. (2023) Biological fractionation of lithium isotopes by cellular Na⁺/H⁺ exchangers unravels fundamental transport mechanisms. iScience 26, 106887. https://doi.org/10.1016/j.isci.2023.106887

). In contrast, marine clays represent the dominant oceanic Li sink. Authigenic smectite can form extensively under hydrothermal or low temperature conditions, through the alteration of volcanic glass or biogenic opal (Zhao et al., 2025

Zhao, S., Saad, E.M., Pickering, R.A., Liu, P., Zuo, H., et al. (2025) Rapid transformation of biogenic silica to authigenic clay: Mechanisms and geochemical constraints. Science Advances 11, eadt3374. https://doi.org/10.1126/sciadv.adt3374

). Marine clays are typically Li-rich compared to any other authigenic phases and they preferentially incorporate the lighter isotope (⁶Li). This inorganic process explains the modern open ocean δ⁷Li value of 31.1 ‰ ± 0.6 ‰ (Jouini et al., 2026

Jouini, A., Payant, L., Vigier, N. (2026) Advances in high-precision lithium isotopic measurements with the Neoma™ MC-ICP-MS. Journal of Analytical Atomic Spectrometry 41, 1004–1016. https://doi.org/10.1039/D5JA00426H

), significantly higher than that of its primary sources (riverine δ⁷Li: 23 ‰; hydrothermal fluids: 8 ‰).

In this study, we explore the potential of marine clays to serve as a robust proxy for reconstructing seawater δ⁷Li values at the time of their formation. We present a new high resolution δ⁷Li record spanning the K-Pg transition and discuss its implications for environmental changes during this critical period.

top

A Single Temperature Dependent Li Isotopic Fractionation Law

To establish marine authigenic clays as reliable proxies for palaeo-seawater δ⁷Li reconstruction, we first establish the seawater-clay Li isotopic fractionation factor at temperatures representative of early marine diagenesis (T < 90 °C). We then validate this approach by comparing results with modern seawater and recent carbonate shell records (0–11 Ma), a time frame particularly relevant to the long Li residence time in the ocean.

Previous experimental studies have quantified Li isotope fractionation during smectite formation at temperatures ranging from 90 °C to 250 °C, revealing significant enrichment of ⁶Li in clays (Vigier et al., 2008

Vigier, N., Decarreau, A., Millot, R., Carignan, J., Petit, S., France-Lanord, C. (2008) Quantifying Li isotope fractionation during smectite formation and implications for the Li cycle. Geochimica et Cosmochimica Acta 72, 780–792. https://doi.org/10.1016/j.gca.2007.11.011

; Wunder et al., 2010

Wunder, B., Deschamps, F., Watenphul, A., Guillot, S., Meixner, A., Romer, R.L., Wirth, R. (2010) The effect of chrysotile nanotubes on the serpentine-fluid Li-isotopic fractionation. Contributions to Mineralogy and Petrology 159, 781–790. https://doi.org/10.1007/s00410-009-0454-x

; Hindshaw et al., 2019

Hindshaw, R.S., Tosca, R., Gout, T.L., Farnan, I., Tosca, N.J., Tipper, E.T. (2019) Experimental constraints on Li isotope fractionation during clay formation. Geochimica et Cosmochimica Acta 250, 219–237. https://doi.org/10.1016/j.gca.2019.02.015

). The clay-solution Li isotopic fractionation factor is temperature dependent. However, experiments conducted below 90 °C have yielded inconclusive results. These challenges stem from lower crystallinity and the persistent presence of exchangeable Li (Vigier et al., 2008

Vigier, N., Decarreau, A., Millot, R., Carignan, J., Petit, S., France-Lanord, C. (2008) Quantifying Li isotope fractionation during smectite formation and implications for the Li cycle. Geochimica et Cosmochimica Acta 72, 780–792. https://doi.org/10.1016/j.gca.2007.11.011

). We therefore developed an optimised protocol for low temperature conditions. This protocol includes extended experiment durations to enhance crystallinity and more efficient removal of exchangeable Li, which may have a distinct isotopic composition (Yang et al., 2023

Yang, C., Yang, S., Vigier, N. (2023) Li isotopic variations of particulate non-silicate phases during estuarine water mixing. Geochimica et Cosmochimica Acta 354, 229–239. https://doi.org/10.1016/j.gca.2023.06.020

). Additionally, we investigated the influence of clay type and chemical composition on the fractionation factor (see the Supplementary Information for the protocol, chemical and mineralogical analyses by XRD, FTIR and TEM).

Collectively, the new and published data reveal a consistent temperature dependent trend for the clay-solution Li isotopic fractionation factor (Fig. 1). Over a range of 25 °C to 250 °C, the fractionation factor exhibits an inverse correlation with temperature, independent of solution chemistry and of the amount of Li incorporated into the crystalline structure. As a result, the δ⁷Li signature of smectite is primarily governed by two key factors: (1) temperature, and (2) the δ⁷Li value of the solution. This relationship provides a theoretical framework for reconstructing water δ⁷Li values from clay samples. Using weighted linear regression that accounts for analytical uncertainties, the empirical relationship derived from experimental studies can be expressed by the equation:

where T is in Kelvin. Although this relationship is primarily established for Mg smectite, the δ⁷Li values of kaolinite and Al smectite (formed at 200 °C) with distinct chemical compositions (see Table S-1) also adhere to this trend. Additionally, published data for various secondary phases and clay mixtures are also consistent with our new experiments (Fig. 1). This unity is explained by theoretical modelling (Dupuis et al., 2017

Dupuis, R., Benoit, M., Tuckerman, M.E., Meheut, M. (2017) Importance of a Fully Anharmonic Treatment of Equilibrium Isotope Fractionation Properties of Dissolved Ionic Species As Evidenced by Li⁺(aq). Accounts of Chemical Research 50, 1597–1605. https://doi.org/10.1021/acs.accounts.6b00607

), which identifies the strength of the Li-O bond and its corresponding vibrational energy as the primary factor controlling the preferential incorporation of ⁶Li into the clay structure. The Li-O bonding dynamics can overshadow potential isotopic effects arising from variations in solution or clay chemistry. This consistent linear correlation suggests that the Li isotopic fractionation during Li incorporation into clay octahedral sites is largely independent of clay type (kaolinite or smectite) and its chemical composition. Instead, temperature emerges as the dominant control.

Full size image

top

Authigenic Clays: a New Proxy of Past Seawater δ⁷Li Variations

To assess the applicability of the relationship in Figure 1 for reconstructing palaeo-seawater δ⁷Li variations, we studied smectite-rich sediments collected through international ocean drilling programmes (DSDP, ODP). These samples represent both modern (core top sediments) and “recent” (<11 Myr) formation. Modern seawater δ⁷Li values, extensively documented in the literature, are close to homogeneous (δ⁷Li = 31.1 ± 0.6 ‰; Jouini et al., 2026

Jouini, A., Payant, L., Vigier, N. (2026) Advances in high-precision lithium isotopic measurements with the Neoma™ MC-ICP-MS. Journal of Analytical Atomic Spectrometry 41, 1004–1016. https://doi.org/10.1039/D5JA00426H

). Given the Li residence time in the ocean, seawater Li concentration and δ⁷Li values can be considered stable over the past 4–5 million years. Furthermore, palaeo-environmental studies using beryllium 10 (Lenard et al., 2020

Lenard, S.J.P., Lavé, J.M., France-Lanord, C., Aumaître, G., Bourlès, D.L., Keddadouche, K. (2020) Steady erosion rates in the Himalayas through late Cenozoic climatic changes. Nature Geoscience 13, 448–452. https://doi.org/10.1038/s41561-020-0585-2

) also indicate steady erosion rates during the late Cenozoic, implying negligible disturbances in riverine Li flux and only minor variations in the oceanic Li budget during this period. Thus, our analysis of recent and sub-recent authigenic clays from open ocean environments provides a robust framework to verify whether the solution δ⁷Li values derived from the α-T relationship in Figure 1 are valid.

The core top smectite-rich clays from Valdivia VA13/2 96BL drill core yield an average δ⁷Li value of 5.6 ± 0.6 ‰ (see Table S-7). Using the equation given above and a temperature typical of the Central Pacific abyssal hill (3 °C), this corresponds to a seawater δ⁷Li value of 30.5 ± 0.6 ‰ that is consistent with that of modern seawater (δ⁷Li = 31.1 ± 0.6 ‰; Jouini et al., 2026

Jouini, A., Payant, L., Vigier, N. (2026) Advances in high-precision lithium isotopic measurements with the Neoma™ MC-ICP-MS. Journal of Analytical Atomic Spectrometry 41, 1004–1016. https://doi.org/10.1039/D5JA00426H

). Over the 0–11 Ma period, δ⁷Li records from foraminifera, brachiopod, bulk carbonates and dolostones document a roughly consistent trend (Misra and Froelich, 2012

Misra, S., Froelich, P.N. (2012) Lithium isotope history of Cenozoic seawater: changes in silicate weathering and reverse weathering. Science 335, 818–823. https://doi.org/10.1126/science.1214697

; Washington et al., 2020

Washington, K.E., West, A.J., Kalderon-Asael, B., Katchinoff, J.A.R., Stevenson, E.I., Planavsky, N.J. (2020) Lithium isotope composition of modern and fossilized Cenozoic brachiopods. Geology 48, 1058–1061. https://doi.org/10.1130/G47558.1

; Liu et al., 2023

Liu, X.-F., Liu, X.-M., Wang, X.-K., Zhai, S., Liu, X. (2023) Dolostone as a reliable tracer of seawater lithium isotope composition. Communications Earth & Environment 4, 58. https://doi.org/10.1038/s43247-023-00711-x

). Note that some of these data had been corrected for isotopic vital (biological) effects during shell growth. We analysed smectite-rich marine sediments from sites 800, 1218, and 1219 in the Tropical Pacific Ocean to reconstruct seawater δ⁷Li variations over the past 11 million years (see Table S-7). These samples were thoroughly characterised using scanning electron microscopy (SEM), laser ablation spectroscopy (LAS), and X-ray diffraction (XRD). Sediments with >80 % smectite content and no detectable opal or quartz were selected to minimise detrital inputs (see Supplementary Information).

Using the experimental α-T relationship (Fig. 1) and deep ocean temperature (Meckler et al., 2022

Meckler, A.N., Sexton, P.F., Piasecki, A.M., Leutert, T.J., Marquardt, J., et al. (2022) Cenozoic evolution of deep ocean temperature from clumped isotope thermometry. Science 377, 86–90. https://doi.org/10.1126/science.abk0604

), we calculated the δ⁷Li values of seawater in equilibrium with clays during their formation. Over the past 5 million years, smectite-rich clay fractions yield seawater δ⁷Li values ranging from 29.7 ± 0.3 ‰ to 30.5 ± 0.6 ‰, consistent with the relatively stable values observed in carbonate records (Fig. 2). From 5 to 10.6 million years ago, clay derived δ⁷Li values gradually decline to 28.1 ± 0.5 ‰, a trend that matches the general pattern established from carbonates. Deep sea water temperature reconstructions for this period may carry an uncertainty of 3–8 °C (Meckler et al., 2022

Meckler, A.N., Sexton, P.F., Piasecki, A.M., Leutert, T.J., Marquardt, J., et al. (2022) Cenozoic evolution of deep ocean temperature from clumped isotope thermometry. Science 377, 86–90. https://doi.org/10.1126/science.abk0604

). Considering this, the propagated uncertainty for seawater δ⁷Li values reconstructed from marine authigenic clays is ∼1.1 ‰ (see Supplementary Information). Additionally, uncertainty may arise from structural differences between the natural and experimentally synthesised clays. The time scale of authigenic clay formation in the deep sea has recently been constrained to <40 days (Zhao et al., 2025

Zhao, S., Saad, E.M., Pickering, R.A., Liu, P., Zuo, H., et al. (2025) Rapid transformation of biogenic silica to authigenic clay: Mechanisms and geochemical constraints. Science Advances 11, eadt3374. https://doi.org/10.1126/sciadv.adt3374

), which is close to our synthesis times at low temperatures. The faster reaction rates typically result in the formation of less crystalline minerals. Nevertheless, the complete removal of exchangeable Li and the consistent Δδ⁷Li values of −20.3 ‰ and −20.0 ‰ observed at 25 °C after 4 and 8 weeks strongly suggest that Li isotopic equilibration has been achieved in our experiments (Table S-1). We acknowledge that potential differences may exist between synthetic and natural authigenic clays, and that Li isotope fractionation during natural authigenic clay formation requires further investigation.

Full size image

There are advantages of using authigenic smectite-rich clays to reconstruct palaeo-oceanic δ⁷Li variations. Planktonic foraminifera underwent significant biotic crises, such as observed at the Cretaceous-Paleogene (K-Pg) boundary, and the influence of vital effects on the biogenic δ⁷Li values remains a subject of debate. Our clay based δ⁷Li proxy opens new potentials in settings where carbonate archives are absent or ambiguous. While diagenetic effects warrant further investigation, our study indicates that their impact is minimal, likely due to the negligible influence of exchangeable Li within clays. In contrast, marine carbonates display Li concentrations several orders of magnitude lower than those in clays, suggesting that their Li isotopic compositions may be more susceptible to modification.

top

Ocean δ⁷Li Variations across the K-Pg Boundary

During the transition from the Late Maastrichtian to the Danian, seawater δ⁷Li values reconstructed from planktonic foraminifera records initially exhibit a gradual increase between 69 and 66 million years ago, followed by an abrupt and pronounced decrease of ∼5 ‰ in less than 400,000 years (Fig. 3b). This rapid decline has been attributed to enhanced dissolution of continental soils and a massive increase in riverine Li flux (Misra and Froelich, 2012

Misra, S., Froelich, P.N. (2012) Lithium isotope history of Cenozoic seawater: changes in silicate weathering and reverse weathering. Science 335, 818–823. https://doi.org/10.1126/science.1214697

). However, the duration of this event is brief compared to the modern Li residence time in the ocean and the time scales required for changing the continental weathering regime. Notably, the dominance of thick lateritic soils during the Cretaceous also suggests a buffering effect against rapid changes (Vigier and Goddéris, 2015

Vigier, N., Goddéris, Y. (2015) A new approach for modeling Cenozoic oceanic lithium isotope paleo-variations: the key role of climate. Climate of the Past 11, 635–645. https://doi.org/10.5194/cp-11-635-2015

). The abrupt δ⁷Li decrease in foraminifera at 66 Ma cannot be explained by a shorter oceanic Li residence time, as the expected rapid return to the preceding K-Pg values is absent. Instead, environmental disruptions like meteorite impacts or massive volcanism may have altered foraminiferal growth. For instance, temperature and pH changes can alter growth rates, which in turn affects both the entrapment of trace elements (including lithium) in crystals and the kinetics of isotopic fractionation (e.g., Vigier et al., 2015

Vigier, N., Rollion-Bard, C., Levenson, Y., Erez, J. (2015) Lithium isotopes in foraminifera shells as a novel proxy for the ocean dissolved inorganic carbon (DIC). Comptes Rendus. Géoscience 347, 43–51. https://doi.org/10.1016/j.crte.2014.12.001

; Roberts et al., 2018

Roberts, J., Kaczmarek, K., Langer, G., Skinner, L.C., Bijma, J., Bradbury, H., Turchyn, A.V., Lamy, F., Misra, S. (2018) Lithium isotopic composition of benthic foraminifera: A new proxy for paleo-pH reconstruction. Geochimica et Cosmochimica Acta 236, 336–350. https://doi.org/10.1016/j.gca.2018.02.038

; Chen et al., 2023

Chen, D., Thibon, F., Felbacq, A., Weppe, L., Metian, M., Vigier, N. (2023) Coupled survey of lithium isotopes and Li/Ca in biogenic and inorganic carbonates. Earth-Science Reviews 244, 104500. https://doi.org/10.1016/j.earscirev.2023.104500

). Species specific characteristics may have also influenced vital (isotopic) effects, as marine biodiversity underwent substantial and rapid changes during the K-Pg transition.

Full size image

We reconstructed seawater δ⁷Li variations using smectite-rich sediments from DSDP Site 524 in the Southwest Atlantic across the K-Pg boundary. Samples were meticulously selected based on a comprehensive analysis of their chemical and mineralogical compositions (SEM, TEM, XRD, FTIR), which consistently demonstrated the dominance of authigenic smectite and a low degree of diagenesis (see Supplementary Information). We used the ocean bottom water temperature calculated from δ¹⁸O records from the Atlantic Ocean and from Mg/Ca ratios of benthic foraminifera (Hull et al., 2020

Hull, P.M., Bornemann, A., Penman, D.E., Henehan, M.J., Norris, R.D., et al. (2020) On impact and volcanism across the Cretaceous-Paleogene boundary. Science 367, 266–272. https://doi.org/10.1126/science.aay5055

; Fischer et al., 2025

Fischer, A., Batenburg, S.J., Bahr, A., Voigt, S., Rheinberger, A., Schmickal, S., Rheinberger, S., Greule, M., Rheinberger, S., Friedrich, O. (2025) Precession-paced late Maastrichtian bottom-water dynamics. Communications Earth & Environment 6, 239. https://doi.org/10.1038/s43247-025-02219-y

).

From 69 to 64 million years ago, the clay derived seawater δ⁷Li record can be divided into three distinct phases (Phases I, II, and III in Fig. 3). During Phase I (69–66.5 Ma), the seawater δ⁷Li trend closely follows the benthic foraminifera δ¹⁸O record, suggesting that the surface Li cycle was influenced by temperature. This phase is characterised by an initial increase in δ⁷Li, followed by a gradual decline over 2 million years. Phase II (66.5–65.5 Ma) coincides with the peak of Deccan volcanism, as well as the Ir-rich layer and the osmium isotope decline, which are often interpreted as evidence of the Chicxulub impact (though this remains debated). This period also includes rapid and pronounced δ¹⁸O fluctuations. Recent studies indicate that the main phase of Deccan volcanism lasted ∼1 million years and can be further divided into four distinct mega-pulses (e.g., ∼66.1–66.0 Ma and ∼65.9–65.8 Ma; Schoene et al., 2019

Schoene, B., Eddy, M.P., Samperton, K.M., Keller, C.B., Keller, G., Adatte, T., Khadri, S.F.R. (2019) U-Pb constraints on pulsed eruption of the Deccan Traps across the end-Cretaceous mass extinction. Science 363, 862–866. https://doi.org/10.1126/science.aau2422

; Sprain et al. 2019

Sprain, C.J., Renne, P.R., Vanderkluysen, L., Pande, K., Self, S., Mittal, T. (2019) The eruptive tempo of Deccan volcanism in relation to the Cretaceous-Paleogene boundary. Science 363, 866–870. https://doi.org/10.1126/science.aav1446

). Despite these volcanic and impact related events, seawater δ⁷Li derived from the clay record remains low, and only small variations are observed throughout this interval (Fig. 3). Finally, in Phase III (65.5–64 Ma), following the cessation of major Deccan volcanic activity, clay data show that the seawater δ⁷Li value gradually returns to its mid-Maastrichtian level (∼25 ‰).

During Phase I (69–66.5 Ma), the parallel trends in δ¹⁸O and δ⁷Li records suggest a long term climatic influence on continental silicate weathering, likely due to progressive soil evolution (e.g., Vigier and Goddéris, 2015

Vigier, N., Goddéris, Y. (2015) A new approach for modeling Cenozoic oceanic lithium isotope paleo-variations: the key role of climate. Climate of the Past 11, 635–645. https://doi.org/10.5194/cp-11-635-2015

). Fischer et al. (2025)

Fischer, A., Batenburg, S.J., Bahr, A., Voigt, S., Rheinberger, A., Schmickal, S., Rheinberger, S., Greule, M., Rheinberger, S., Friedrich, O. (2025) Precession-paced late Maastrichtian bottom-water dynamics. Communications Earth & Environment 6, 239. https://doi.org/10.1038/s43247-025-02219-y

attributed the temperature decline after 68.5 Ma to reduced CO₂ levels linked to variations in continental weathering and volcanic activity. They consider that Ninetyeast Ridge volcanism triggered a warming event at ∼69.2 Ma and the Mid-Maastrichtian Event, followed by cooling due to enhanced silicate weathering. The decrease in seawater δ⁷Li aligns with this interpretation.

In Phase II (66.5–65.5 Ma), clay derived seawater δ⁷Li reaches its lowest values with minimal fluctuation, indicating that the Deccan Traps volcanic pulses and the Chicxulub impact did not significantly perturb the riverine flux and the marine Li budget. Conversely, planktonic foraminifera exhibit a sharp δ⁷Li decline (∼5 ‰), likely due to vital effects influenced by changes in dissolved inorganic carbon (DIC) or pH (Vigier et al., 2015

Vigier, N., Rollion-Bard, C., Levenson, Y., Erez, J. (2015) Lithium isotopes in foraminifera shells as a novel proxy for the ocean dissolved inorganic carbon (DIC). Comptes Rendus. Géoscience 347, 43–51. https://doi.org/10.1016/j.crte.2014.12.001

; Roberts et al., 2018

Roberts, J., Kaczmarek, K., Langer, G., Skinner, L.C., Bijma, J., Bradbury, H., Turchyn, A.V., Lamy, F., Misra, S. (2018) Lithium isotopic composition of benthic foraminifera: A new proxy for paleo-pH reconstruction. Geochimica et Cosmochimica Acta 236, 336–350. https://doi.org/10.1016/j.gca.2018.02.038

; Poet et al., 2023

Poet, M., Vigier, N., Bouret, Y., Jarretou, G., Gautier, R., et al. (2023) Biological fractionation of lithium isotopes by cellular Na⁺/H⁺ exchangers unravels fundamental transport mechanisms. iScience 26, 106887. https://doi.org/10.1016/j.isci.2023.106887

). Supporting this, calcium isotopes reveal significant δ^44/40Ca fluctuations in foraminifera at the K-Pg boundary (Jouini et al., 2023

Jouini, A., Paris, G., Caro, G., Bartolini, A., Gardin, S. (2023) Constraining oceanic carbonate chemistry evolution during the Cretaceous-Paleogene transition: Combined benthic and planktonic calcium isotope records from the equatorial Pacific Ocean. Earth and Planetary Science Letters 619, 118305. https://doi.org/10.1016/j.epsl.2023.118305

), tied to disruptions in ocean carbonate chemistry (CO₃²⁻, DIC, Ω) and associated biological responses. Boron isotopes also demonstrate a rapid increase in seawater pH by ∼0.4 units after the K-Pg boundary (Henehan et al., 2019

Henehan, M.J., Ridgwell, A., Thomas, E., Zhang, S., Alegret, L., et al. (2019) Rapid ocean acidification and protracted Earth system recovery followed the end-Cretaceous Chicxulub impact. Proceedings of the National Academy of Sciences 116, 22500–22504. https://doi.org/10.1073/pnas.1905989116

). Based on the proposed pH-δ⁷Li relationship (Roberts et al., 2018

Roberts, J., Kaczmarek, K., Langer, G., Skinner, L.C., Bijma, J., Bradbury, H., Turchyn, A.V., Lamy, F., Misra, S. (2018) Lithium isotopic composition of benthic foraminifera: A new proxy for paleo-pH reconstruction. Geochimica et Cosmochimica Acta 236, 336–350. https://doi.org/10.1016/j.gca.2018.02.038

), this change is sufficient to drive the observed δ⁷Li decline in foraminifera.

Overall, these findings highlight the asynchronous behaviour of terrestrial and marine systems during this major ecological transition. The contrasting records between clays and foraminifera illustrate the role of long term climate change on continental weathering. At the boundary, rapid changes in carbonates δ⁷Li are consistent with large atmospheric carbon fluctuations profoundly impacting calcifier biology, with minimal impact on continental fluxes.

top

Acknowledgements

This work was funded by CNRS-INSU SYSTER, by the ANR INTOCC (Grant No. ANR-15-CE31-0013), and by the National Natural Science Foundation of China (Grant No. 42376057). AJ is funded by the ERC Advanced “SeaLi2Bio” (Grant #101097738). Many thanks to Bremen Core Collection team for access to Deep Sea Core Samples, to Ken Rubin, Denys Vonderhaar and Doug Pyle from Hawaii University, to Pierre Chansigaud from IC2MP for FTIR measurements, and to Philippe Telouk at CNRS-INSU national service (ENS-Lyon). We would like to thank the anonymous reviewer and Xiao-Ming Liu for their constructive comments, which have enhanced the quality of this manuscript.

Editor: Gavin Foster

top

References

top

Supplementary Information

The Supplementary Information includes:

• 1. Smectite and Kaolinite Synthesis
• 2. Li Adsorption/Saturation Experiments
• 3. Marine Sites
• 4. Li and δ⁷Li Measurements
• 5. Study of Crystal Chemistry
• Tables S-1 to S-9
• Figures S-1 to S-9
• Supplementary Information References

Download the Supplementary Information (PDF)

Download Table S-9 (xlsx)