Postimpact earliest Paleogene warming shown by fish debris oxygen isotopes (El Kef, Tunisia)

See allHide authors and affiliations

Science  29 Jun 2018:
Vol. 360, Issue 6396, pp. 1467-1469
DOI: 10.1126/science.aap8525

Warming after the big one

The Chicxulub impact 65 million years ago, which caused the mass extinction at the Cretaceous-Paleogene boundary, also initiated a long period of strong global warming. Using data from phosphatic microfossils, including fish teeth, scales, and bone, MacLeod et al. estimated global average temperature. Immediately after the asteroid strike, temperatures increased by ∼5°C and remained high for about 100,000 years (see the Perspective by Lécuyer). These results are relevant to current climate projections, because the Chicxulub impact perturbed Earth systems on time scales even shorter than the current rate of change.

Science, this issue p. 1467; see also p. 1400


Greenhouse warming is a predicted consequence of the Chicxulub impact, but supporting data are sparse. This shortcoming compromises understanding of the impact’s effects, and it has persisted due to an absence of sections that both contain suitable material for traditional carbonate- or organic-based paleothermometry and are complete and expanded enough to resolve changes on short time scales. We address the problem by analyzing the oxygen isotopic composition of fish debris, phosphatic microfossils that are relatively resistant to diagenetic alteration, from the Global Stratotype Section and Point for the Cretaceous/Paleogene boundary at El Kef, Tunisia. We report an ~1 per mil decrease in oxygen isotopic values (~5°C warming) beginning at the boundary and spanning ~300 centimeters of section (~100,000 years). The pattern found matches expectations for impact-initiated greenhouse warming.

The Cretaceous/Paleogene (K/Pg) mass extinction is unique among major extinction events in that its ultimate cause (the Chicxulub impact) perturbed Earth systems on time scales shorter than those of current anthropogenic changes (1). Study of the impact aftermath provides a perspective on the response of Earth systems to extremely rapid global perturbations, making the K/Pg event an unusually relevant natural experiment to compare to modern climatic and environmental changes. Prominent among the proposed effects of the Chicxulub impact are wide swings in temperature that include a brief infrared pulse due to frictional heating of ejecta re-entering the atmosphere, lasting for as little as 10 minutes (2, 3); an impact winter due to atmospheric loading of dust, soot, and sulfate aerosols lasting for months to years (46); and greenhouse warming lasting ~100,000 years or more due to increased atmospheric CO2 concentrations from impact-volatilized carbonates and wildfires (7, 8).

These shifts have been hypothesized for decades and are invoked as examples that should inform thinking about the collateral consequences of activities as different as nuclear war and anthropogenic emissions (1, 6, 9), but testing predictions is difficult. K/Pg changes occurred over short time scales relative to the typical resolution of the rock record, and samples suitable for generating meaningful paleotemperature estimates are scarce. Thus, progress has been slow, and data are often ambiguous.

Observational evidence cited in support of the extreme heat pulse are high concentrations of soot above the boundary and selective survivorship among terrestrial taxa (3). For impact winter, an immediately post-K/Pg, 2° to 4°C cold pulse is suggested by some TEX86-based paleotemperature estimates (6, 10). Evidence for postimpact greenhouse warming using samples that meet current criteria for quality of sample preservation has not been found. In fact, postimpact cooling lasting thousands of years has been suggested from paleontological data at the El Kef K/Pg Global Stratotype Section and Point (GSSP) and nearby sections (11, 12).

In this study, we report oxygen isotopic values (δ18O) of phosphate isolated from sand-sized (~0.1 to 2 mm) remains of fish teeth, scales, and bone (herein “fish debris”) from El Kef, Tunisia (Fig. 1), indicating ~5°C warming beginning at the K/Pg boundary and lasting for ~100,000 years. The El Kef section is dominated by hemipelagic marls (~40% CaCO3) and was deposited at ~20°N in waters 200 to 400 m deep on the Tethyan margin of northern Africa. The boundary is placed at the base of a thin red layer that contains geochemical, mineralogical, and sedimentological indicators of impact and coincides with the level of the K/Pg mass extinction. Above the red layer is a 50-cm-thick, carbonate-poor (<10%) claystone in which bulk carbonate δ13C values exhibit the expected K/Pg negative excursion. Carbonate content increases back to pre-impact levels gradually over a ~2-m interval above the boundary claystone (13, 14)

Fig. 1 K/Pg paleogeography.

K/Pg paleogeography (21) showing El Kef and Chicxulub.

We analyzed samples from 2 m below to 6.6 m above the boundary. The El Kef section is the GSSP for the K/Pg Boundary and, thus, is formally accepted as being stratigraphically complete with well-resolved chronostratigraphic control. Average sedimentation rates are ~3.5 cm/1000 years from 12 m below the boundary to 7 m above the boundary (Fig. 2). 3He concentrations suggest that the clay layer accumulated more rapidly than suggested biostratigraphically (15), but this detail does not affect our analysis as observed δ18O changes span several dated events.

Fig. 2 Age-depth model of El Kef foraminiferal events.

Age-depth model of El Kef foraminiferal events (22) dated relative to a K/Pg boundary (23, 24) using a boundary age of 66.043 million years (Ma) (25). LO, lowest occurrence; HO, highest occurrence.

In the El Kef section, carbonate microfossils are recrystallized (13), and organic molecules useful for paleotemperature measurements are absent (12). However, well-preserved fish debris is commonly present. The δ18O value of fish debris varies as a function of temperature at the time the phosphate was secreted and is more resistant to diagenetic alteration than carbonate δ18O values (16, 17). Thus, fish debris from El Kef is an excellent material with which to examine post-K/Pg temperature history.

Fish debris δ18O values show little variability in the uppermost 2 m of the Cretaceous [avg. = 20.7‰VSMOW (per mil) on the Vienna standard mean ocean water scale ± 0.33; n = 10], decrease by ~1‰ at the boundary, and remain low over the first 3 m of the earliest Paleogene (avg. = 19.5‰VSMOW ± 0.42, n = 21), and increase by ~1‰ (avg. = 20.6‰VSMOW ± 0.63, n = 9) in the top 4 m of section studied. The mean value for the older Paleocene samples is statistically distinct from the means of the subjacent and superjacent groups, whereas the means of Cretaceous samples and younger Paleogene samples are not statistically distinct (Fig. 3 and supplementary materials). A ~1‰ decrease in δ18O values across the boundary suggests ~5°C of warming, assuming seawater δ18O values remained approximately constant.

Fig. 3 El Kef fish debris δ18O values plotted against depth.

Samples are grouped into three stratigraphic bins divided at the K/Pg boundary and 3 m above the boundary. Averages ± 1 SD for each bin are shown to the right with P values of t tests comparing means (supplementary materials). The difference in means between the earliest Paleogene (orange) and both the underlying Cretaceous (blue) and overlying Paleogene (green) is significant at the 5σ and 4σ level, respectively. The difference in means between the blue and green bins is not significant. Ages in thousand years (Kyr) are based on Fig. 2 age model; temperature calculated using (26), assuming seawater δ18O = –1‰VSMOW; error bar ± 0.3‰ (supplementary materials).

We treat the δ18O results as an integrated signal of the outer shelf water column. Fish are mobile organisms, and different individuals could record conditions from a range of environments, depths, and seasons. Mitigating these concerns are estimated water depths (200 to 400 m) that restrict the depths at which the fish likely lived, and the tropical paleolatitude of the site, so expected seasonal variability is low. In addition, each measurement was based on separates of typically 50 to 75 individual microfossils, and the measurements were grouped into three stratigraphic bins—2 m of section representing deposition during the last 50,000 years of the Cretaceous, 3 m of section representing the first 100,000 years of the Paleogene, and 4 m of section representing 100,000 to 300,000 years after the boundary—of 10, 21, and 9 samples, respectively. The average of measurements within each bin are compared to assess temperature history (Fig. 3 and supplementary materials).

Potential sampling biases and temporal variability should still be considered despite large sample sizes. The high standard deviation among the samples from 3 to 7 m above the boundary suggest that uncertainty is greatest for this bin. Additional analyses might reduce apparent variability, might reveal stratigraphic trends within the interval, or might reinforce the impression that high variability among samples is a feature of this part of the section. In contrast, the low standard deviation of the δ18O measurements (close to analytical precision of 0.3‰) in the lower two bins argues that these bins have quite stable isotopic signatures that our sampling scheme adequately captured.

Evidence for increased temperatures begins at the K/Pg boundary. We see no evidence of an impact winter, but finding evidence for this ≤decade-long interval was unlikely. Collected samples spanned 2.5 to 10 cm of section (i.e., on average, each represents ≥1000 years of deposition), and bioturbation and physical reworking (14) would introduce additional time averaging. Reworking could also have smeared the evidence for the initiation of greenhouse warming. However, the sharp K/Pg δ18O decrease found (Fig. 3) indicates reworking was not that severe, a finding consistent with the well-resolved biostratigraphic and geochemical records that contributed to the selection of El Kef as the K/Pg GSSP.

Our results correspond remarkably well with trends in δ18O values predicted for greenhouse warming starting within decades after the impact, but they contradict conclusions based on paleontological data suggesting thousands of years of postimpact cooling (11, 12). We offer facies control as an alternative explanation for the paleontological results. Relatively high abundance of taxa with boreal affinities (the observation supporting the inference of cooler temperatures) are restricted to the 50-cm-thick boundary claystone. The K/Pg event could have perturbed many variables (e.g., precipitation, nutrient loading, pH, oxygen levels, carbonate production, carbon cycling, niche occupancy, and food web structure) that would plausibly affect both lithology and microfossil abundance without a priori temperature implications. The lack of a lithological shift at the end of the isotopic excursion, in contrast, renders ad hoc explanations that would explain away apparent warming as an artifact of a stratigraphically restricted interval of altered ocean circulation, changed moisture balance, and/or diagenesis (supplementary materials).

Our results suggest ~5°C of postimpact warming lasting ~100,000 years (Fig. 3). This magnitude and duration of warming matches well with modeled climate response to post-K/Pg increase in atmospheric CO2 concentrations to ≥2300 parts per million (ppm) from background levels of 350 to 500 ppm, as estimated from stomatal densities on fossil leaves (7). In contrast, paleosol paleobarometry (18), as well as modeling (19) and experimental results (20) of impact-induced volatilization of carbonates, suggest more modest CO2 increase (≤50%). If the match between stomatal estimates of CO2 change and El Kef temperature increase are meaningful, widespread wildfires (4, 8) or other large CO2 sources seem required to augment CO2 from impact volatilization, and paleosol-based interpretations implicitly would be questioned. Alternatively, if CO2 concentrations only increased modestly, either our interpretation of the El Kef results greatly overestimates global post-K/Pg warming or the climate sensitivity used in postimpact warming models is too low. Discriminating among these alternatives would be a step forward in understanding the post-K/Pg world, with implications for temperature responses to modern climatic perturbations (especially increasing atmospheric CO2 levels) imposed on very short time scales.

Supplementary Materials

Materials and Methods

Supplementary Text

Fig. S1

Tables S1 and S2

References (2730)

References and Notes

Acknowledgments: We thank I. Arenillas, J. Arz, M. Katz, and J. Smit for useful discussion; two reviewers for editorial suggestions; and S. Haynes, G. O’Neil, L. Alegret, S. Bey, M. Baroumi, M. Giron, R. Summons, H. Theyri, and J. Wendler for laboratory and/or field help. Funding: NSF-EAR 1323444 (K.G.M.), MU Undergraduate Research Fund (K.G.M.), and the MIT MISTI-Spain Program (J.S.). Author contributions: K.G.M. conceived the study, processed and analyzed samples, interpreted results, and wrote and edited the manuscript; P.C.Q. processed and analyzed samples and participated in interpretation, writing, and editing of the manuscript; and J.S. and M.H.N. collected samples and participated in interpretation, writing, and editing of the manuscript. Competing interests: None declared. Data and materials availability: All data are presented in the supplementary materials.

Stay Connected to Science

Navigate This Article