TEMPERATURE CHANGE ACROSS THE CRETACEOUS/PALEOGENE BOUNDARY INFERRED FROM OXYGEN ISOTOPIC RATIOS IN FISH DEBRIS FROM EL KEF, TUNISI

The Cretaceous/Paleogene (K/T) mass extinction is unique among major extinction events in that its ultimate cause (the Chicxulub impact) perturbed Earth systems on time scales shorter than the current rate of anthropogenic changes. Thus, the K/T record provides a perspective on the response of Earth systems to extremely rapid, global perturbations not available from studying other deep time events. Temperature change is a predicted consequence of both impact and pre-K/T volcanism; however, whereas Deccan-related warming has been documented, few empirical constraints exist on temperature change related to impact. This shortcoming is due largely to the facts that preservation is not good enough in most early Paleogene samples to apply the carbonate-δ¹⁸O paleothermometer with confidence and most K/T sections are not demonstrably complete and expanded enough to resolve changes on relevant time scales.

With these concerns in mind, we addressed the question of the temperature record of the K/T interval by measuring the oxygen isotopic ratios of fish debris from the El Kef section in Tunisia. El Kef is the Global Stratotype Section and Point for the K/T boundary with chronostratigraphic events documented in the section that allow for recognition of changes occurring on 10³ to 10⁴ year time scales. Fish teeth, bones and mineralized scales are composed of bioapatite, a phase in which original δ¹⁸O values are more resistance to diagenetic alteration than are the δ¹⁸O values of carbonates. In 40 samples spanning the highest 2 m of the Cretaceous and lowest 9 m of the Paleogene, we found relatively stable and high values in the Cretaceous, a 1 to 1.5‰ decrease in δ¹⁸O values within the lowest 25 cm of Paleogene strata, an interval of relatively stable and low values from 0.3 to 3 m above the boundary, and an increase to pre-boundary δ¹⁸O values at 4.5 m that extends to at least 9 m above the boundary. Using published events to construct an age model and assuming the δ¹⁸O changes are due solely to temperature, these results suggest 4-6°C warming starting within the first few thousand years after the impact and lasting for ~100,000 years. This conclusion matches well with the predicted pattern of a CO₂-induced interval of greenhouse warming initiated by the impact.