End-Cretaceous extinction in Antarctica linked to both Deccan volcanism and meteorite impact via climate change

TL;DR

A new clumped-isotope temperature record from Seymour Island, Antarctica, shows a 7.8±3.3 °C warming synchronous with Deccan Traps onset and a smaller 1.1±2.7 °C warming at the KPg boundary, each coinciding with a pulse of species extinction. The authors conclude that the end-Cretaceous extinction at this site was linked to both Deccan volcanism and the Chicxulub impact via climate change.

Key findings

• Clumped isotope temperatures warm from ∼5 °C to ∼14 °C between 68.7 and 67.8 Ma, remain ∼9–12 °C for ∼1 Myr, cool to ∼4 °C from 66.9 to 66.25 Ma, then show a 7.8±3.3 °C warming ∼150 kyr before the KPg boundary and a second 1.1±2.7 °C warming at the boundary; temperatures return to pre-event levels ∼200–400 kyr after the initial spike.
• Extinction at Seymour Island occurred in two pulses: 10 of 24 species were eliminated before Chicxulub, coinciding with the large warming and Deccan onset, and 14 of 24 species were eliminated at the KPg boundary; the earlier pulse had 9 of 10 benthic species, whereas the later pulse had only 6 of 14 benthic species, and one ammonite species went extinct in the first event versus all remaining six at the KPg boundary.
• δ18Ow varied from −3.4 to +1.0‰ (mean −1.3±0.8‰) and was highly correlated with temperature; assuming a fixed δ18Ow gives only about half the variability of the clumped isotope record (7 °C versus 12.5 °C range for shells in this study).
• Deccan CO2 emissions between Chron 29R onset and the KPg boundary are estimated at 270–900 p.p.m. added to a background of ∼360–380 p.p.m.; post-KPB volcanism potentially added another 830–900 p.p.m., with modelling studies predicting 0.5–4 °C global warming and possible local amplification by ice-albedo feedback.

Methods

• Twenty-nine well-preserved bivalve shells representing five species were selected from the Zinsmeister collection, Units 9 and 10 of the López de Bertodano Formation, Seymour Island; most were sampled at the umbo and ventral margin for δ18O, δ13C, and clumped isotope composition (Δ47), with a larger set of 116 shells analyzed for δ18O and δ13C only.
• Shell preservation was screened by X-ray diffraction, cathodoluminescence, and trace element thresholds (Fe <500 p.p.m., Mn <200 p.p.m.); an age model was built from published magnetostratigraphy converted to the Zinsmeister stratigraphic framework and corroborated with new and published 87Sr/86Sr data.
• Temperatures were derived from Δ47 using the Defliese et al. calibration; δ18Ow was calculated from Δ47-derived temperature, shell δ18O, and aragonite-water fractionation; horizon means were inverse-variance weighted, and CO2 emissions were calculated from estimated lava volumes using 14 Tg CO2 km−3 lava.

Limitations

• The paper states that it is impossible at this point to separate the influences of the meteorite and the additional post-KPB Deccan volcanism in driving the second extinction pulse or the second warming event.
• The paper notes inherent uncertainty in the magnetostratigraphy-based age model from interpolation and the Geological Time Scale, which was not explicitly accounted for, and that the apparent rapid rate of Deccan warming may be an artifact of variable sampling resolution.
• The paper states that δ18O-derived temperatures assuming constant δ18Ow are biased and underestimate true variability; the clumped isotope sample set is limited to the best-preserved shells, and horizon averages combine species and shell positions without knowing which best represents mean annual temperature.