top

Abstract

Figures and Tables

top

Introduction

The P-E boundary is marked by a negative CIE in marine and terrestrial records and 5-8 °C global warming (Kennett and Stott, 1991

Kennett, J.P., Stott, L.D. (1991) Abrupt deep-sea warming, palaeoceanographic changes and benthic extinctions at the end of the Palaeocene. Nature 353, 225-229.

; see McInerney and Wing, 2011

McInerney, F.A., Wing S.L. (2011) Thermal Maximum: A Perturbation of Carbon Cycle, Climate, and Biosphere with Implications for the Future. Annual Review of Earth and Planetary Sciences 39, 489-516.

, for review). Although the source(s) and magnitude of low δ¹³C carbon released into the ocean-atmosphere system at the P-E boundary are debated, impact ejecta recently discovered at the CIE onset indicates that an extraterrestrial impact played a role. Glassy spherules are found in a discrete stratigraphic layer within the CIE onset on the U.S. Atlantic Margin at multiple sites spread over 1,000 km (Schaller et al., 2016

Schaller, M.F., Fung, M.K., Wright, J.D., Katz, M.E., Kent, D.V. (2016) Evidence of an Extraterrestrial Impact at the Paleocene-Eocene boundary. Science 354, 225-229.

). The ejecta at Wilson Lake B (WL; Schaller et al., 2016

Schaller, M.F., Fung, M.K., Wright, J.D., Katz, M.E., Kent, D.V. (2016) Evidence of an Extraterrestrial Impact at the Paleocene-Eocene boundary. Science 354, 225-229.

) and at Randall’s Farm (RF; this study, Fig. S-1) are formed as either solidified melt-droplets or vapour condensates. On the palaeo-continental shelf, the CIE onset coincides with the base of the thick Marlboro Clay that extends from Virginia to New Jersey (Fig. 1).

Full size image | Download in Powerpoint

In this study, we document unusually abundant microscopic (~100 μm) black charcoal shards near the base of the Marlboro Clay at WL and RF; these shards bear characteristics of well-preserved charred plant material and their distribution in the sediment column forms an abrupt peak immediately above the spherule level. SEM imaging – the most diagnostic method of studying charcoalified plant material (Scott and Collinson, 1978

Scott, A.C., Collinson, M.E. (1978) Organic sedimentary particles: results from SEM studies of fragmentary plant material. In: Whalley, W.B. (Eds) SEM in the Study of Sediments. Geoabstacts, Norwich, 137–167.

) – reveals three-dimensional plant features with excellent preservation, including homogeneous cell walls and honeycomb structures (Figs. 2, S-2). This paper explores two hypotheses for the origin of the charcoal following the impact ejecta within the CIE: 1) warming-induced, continental drying and ecosystem change; and 2) as a direct consequence of the impact.

Full size image | Download in Powerpoint

top

Evidence for Synchronous, Widespread Intense Wildfires

Charcoal is the most diagnostic artefact of wildfires (Belcher et al., 2003

Belcher, C.M., Collinson, M.E., Sweet, A.R., Hildebrand, A.R., Scott, A.C. (2003) Fireball passes and nothing burns - The role of thermal radiation in the Cretaceous-Tertiary event: Evidence from the charcoal record of North America. Geology 31, 1061-1064.

). Overlying the P-E impact ejecta, we observe a 24-fold increase in charcoal pieces (>63 µm)/gram at WL, and 21-fold increase at RF, a remarkable change above a low but consistent background (Figs. 3, S-3). The material is typically brittle, black, opaque in thin section, and preserves anatomical detail. Raman spectroscopy is widely used to characterise carbonaceous materials (Ulyanova et al., 2014

Ulyanova, E.V., Molchanov, A.N., Prokhorov, I.Y., Grinyov, V.G. (2014) Fine structure of Raman spectra in coals of different rank. International Journal of Coal Geology 121, 37-43.

; Wang et al., 2014

Wang, S., Cheng, H., Jiang, D., Huang, F., Su, S., Bai, H. (2014) Raman spectroscopy of coal component of Late Permian coals from Southern China. Spectroscopy 132, 767-770.

), and helps confirm the presence of charcoal (Fig. S-4). Wavelength dispersive X-ray spectroscopy (WDS) indicates that the black shards are 58-84 % carbon by weight. Charcoal exhibits high reflectance (R_o) compared to other maceral groups  as the result of heating; higher reflectance indicates higher combustion temperatures (Scott and Jones, 1991

Scott, A.C., Jones, T.P. (1991) Microscopic observations of Recent and fossil charcoal. Microscopy and Analysis 25, 13-15.

). Samples yield average R_o values between 0.38 % and 2.99 %, corresponding to temperatures between 340-550 °C (Scott and Jones, 1991

Scott, A.C., Jones, T.P. (1991) Microscopic observations of Recent and fossil charcoal. Microscopy and Analysis 25, 13-15.

) (Table S-1). Lowest R_o values indicating vitrinite (uncharred plant material) are found below the boundary; highest R_o values occur within the charcoal peak at both RF (1.24 %) and WL (2.99 %), and decrease above; this distribution of R_o values supports a widespread and intense singular wildfire event.

Full size image | Download in Powerpoint

The ejecta found at the base of the Marlboro Clay, have all the characteristics of instantaneous air-fall deposit (Schaller et al., 2016

Schaller, M.F., Fung, M.K., Wright, J.D., Katz, M.E., Kent, D.V. (2016) Evidence of an Extraterrestrial Impact at the Paleocene-Eocene boundary. Science 354, 225-229.

), and have been dated by K-Ar to 54.9 ± 3 Ma (Schaller and Fung, 2018

Schaller, M.F., Fung, M.K. (2018) The extraterrestrial impact evidence at the Paleocene-Eocene boundary and sequence of environmental change on the continental shelf. Philosophical Transactions of the Royal Society A, 376.

). Because the depositional and cooling ages of the spherules appear to be indistinguishable, the impact ejecta may provide a time-equivalent marker and baseline for the start of the CIE across the mid-Atlantic Salisbury Embayment. The charcoal peak at both localities occurs stratigraphically above both the spherule layer and the start of the CIE; based on this sequence, we infer that: 1) the charcoal was deposited after the impact; and 2) synchronous wildfires produced the charcoal.

Several lines of evidence indicate that the charcoal-rich layers are neither artefacts of reworked older sediments nor simply due to sedimentation change. (1) The charcoal peak lies 12 and 104 cm above the ejecta layer at WL and RF, respectively, indicating that the charcoal is not older burned material reactivated by the impact event. (2) The charcoal increases substantially above the change from sands to clayey silts, and decreases to near-background levels well within the clay. (3) There is a single observed pulse of charcoal, not multiple peaks as would be expected from reworking of charcoal.

There is a stratigraphic and causal link between our charcoal observations and an unusual and widespread abundance of magnetic nanoparticles in the Marlboro Clay (Lanci et al., 2002

Lanci, L., Kent, D.V., Miller, K.G. (2002) Detection of Late Cretaceous and Cenozoic sequence boundaries on the Atlantic coastal plain using core log integration of magnetic susceptibility and natural gamma ray measurements at Ancora, New Jersey. Journal of Geophysical Research 107, 2216.

), which resulted from soil pyrogenesis (Kent et al., 2017

Kent, D.V., Lanci, L., Wang, H., Wright, J.D. (2017) Enhanced magnetization of the Marlboro Clay as a product of soil pyrogenesis at the Paleocene-Eocene Boundary? Earth and Planetary Science Letters 473, 303-312.

). The stratigraphic level of the charcoal increase at RF and WL coincides with this unique change in magnetisation (Kopp et al., 2009

Kopp, R.E., Schumann, D., Raub, T.D., Powars, D.S., Godfrey, L.V., Swanson-Hysell, N.L., Maloof, A.C., Vali, H. (2009) An Appalachian Amazon? Magnetofossil evidence for the development of a tropical river-like system in the mid-Atlantic United States during the Paleocene-Eocene thermal maximum. Paleoceanography 24, PA4211.

and Kent et al., 2017

Kent, D.V., Lanci, L., Wang, H., Wright, J.D. (2017) Enhanced magnetization of the Marlboro Clay as a product of soil pyrogenesis at the Paleocene-Eocene Boundary? Earth and Planetary Science Letters 473, 303-312.

, respectively; Fig. 3), indicating that they share a common origin and were deposited contemporaneously. The unique magnetisation is observed in ten sites spanning ~400 km in the Salisbury Embayment (Kopp et al., 2009

Kopp, R.E., Schumann, D., Raub, T.D., Powars, D.S., Godfrey, L.V., Swanson-Hysell, N.L., Maloof, A.C., Vali, H. (2009) An Appalachian Amazon? Magnetofossil evidence for the development of a tropical river-like system in the mid-Atlantic United States during the Paleocene-Eocene thermal maximum. Paleoceanography 24, PA4211.

), providing an independent line of evidence indicating that the wildfire activity was regionally extensive (See SI). Elevated fern spores and charred plant material within the CIE at the Mattawoman Creek-Billingsley Road site in the southern Salisbury Embayment (Self-Trail et al., 2017

Self-Trail, J.M., Robinson, M.M., Bralower, T.J., Sessa, J.A., Hajek, E.A., Kump, L.R., Trampush, S.M., Willard, D.A., Edwards, L.E., Powars, D.S., Wandless, G.A. (2017) Shallow marine response to global climate change during the Paleocene-Eocene Thermal Maximum, Salisbury Embayment, USA. Paleoceanography and Paleoclimatology 32, 710-728.

) provides further evidence that the wildfires were widespread and contemporaneous (within the level of resolution possible). The dramatic increase and dominance of fern spores in the Marlboro Clay is consistent with the opportunistic nature of ferns immediately following wildfire (Ainsworth and Kauffman, 2009

Ainsworth, A., Kauffman, J.B. (2009) Response of native Hawaiian woody species to wildfires in tropical forests and scrublands. Plant Ecology 201, 197-209.

).

We use the nadir of the CIE onset as a tie point to correlate the sequence of events recorded at WL and RF (Fig. 3). Remarkably, the highest charcoal abundance occurs prior to the CIE nadir, highlighting the following sequence of events: 1) impact event (deposition of spherule layer); 2) widespread wildfires (significant increase in charcoal); and 3) CIE nadir. The significant increase in charcoal invites the obvious question: what triggered the intense, widespread wildfires that produced the charcoal spike? Here, we consider two scenarios to explain the elevated charcoal abundance at the P-E boundary: 1) continental drying and subsequent wildfires; and 2) impact-induced wildfires.

Scenario 1: Warming-induced, continental drying and subsequent wildfires

Large scale continental drying associated with Paleocene-Eocene Thermal Maximum (PETM) warming could put vegetation at risk for wildfires. These wildfires could be triggered by lightning strikes and be regionally widespread. The first of these many fires would burn extensively, probably contemporaneously, over a wide area, be very intense, and lead to deposition of the large charcoal peak we observe in the sedimentary record. This would likely be followed by subsequent smaller wildfire events producing repeated charcoal pulses, which we do not observe.

Under this scenario, continental interiors experienced aridification. Studies from the U.S. western interior indicate rainfall decreased by ~40 % near the PETM onset, but do not conclude if this aridification was more than regional (Wing et al., 2005

Wing, S.L., Harrington, G.J., Smith, F.A., Bloch, J.I., Boyer, D.M., Freeman, K.H. (2005) Transient floral change and rapid global warming at the Paleocene-Eocene boundary. Science 310, 993-996.

). Palaeosols in Wyoming also indicate decreased precipitation at the onset (Kraus and Riggins, 2007

Kraus, M., Riggins, S. (2007) Transient drying during the Paleocene–Eocene Thermal Maximum (PETM): analysis of paleosols in the bighorn basin, Wyoming. Palaeogeography Palaeoclimatology Palaeoecology 245, 444-461.

). Conversely, modelling results from northern and mid-latitudes (including Wyoming) indicate a 20-25 % increase in soil and atmospheric moisture during the PETM (Bowen et al., 2004

Bowen, G.J., Beerling, D.J., Koch, P.L., Zachos, J.C., Quattlebaum, T. (2004) A humid climate state during the Palaeocene/Eocene thermal maximum. Nature 432, 495-499.

), and Kopp et al. (2009)

Kopp, R.E., Schumann, D., Raub, T.D., Powars, D.S., Godfrey, L.V., Swanson-Hysell, N.L., Maloof, A.C., Vali, H. (2009) An Appalachian Amazon? Magnetofossil evidence for the development of a tropical river-like system in the mid-Atlantic United States during the Paleocene-Eocene thermal maximum. Paleoceanography 24, PA4211.

speculated on an “Appalachian Amazon” in Eastern North America. The latter is consistent with the apparent runoff increase responsible for the Apectodinium acme at the P-E onset in the Marlboro Clay (Sluijs et al., 2007

Sluijs A., Brinkhuis, H., Schouten, S., Bohaty, S.M., John, C.M., Zachos, J.C., Reichart, G.-J., Sinninghe Damsté, J.S., Crouch, E.M., Dickens, G.R. (2007) Environmental precursors to rapid light carbon injection at the Paleocene/Eocene boundary. Nature 450, 1218-1221.

). Hydrologic studies on the PETM thus have not provided clear evidence for a pattern of dryer or wetter conditions, but observations suggest the mid-Atlantic margin was wet.

Scenario 2: Impact-induced wildfires

The charcoal peaks at WL and RF occur stratigraphically above the ejecta layer; therefore, the wildfires could have been ignited by the P-E boundary impact. In this scenario, the massive thermal energy of an impact vapour plume and subsequent ejecta fallout (Melosh et al., 1990

Melosh, H.S., Schneider, N.M., Zahnlek, J., Latham, D. (1990) Ignition of global wildfires at the Cretaceous/Tertiary boundary. Nature 343, 251-254.

) ignited widespread wildfires, producing abundant charcoal that was transported to the shelf via the next major hydrologic event, resulting in the observed depositional sequence. Wildfires following the Cretaceous-Paleogene boundary impact event (see SI for review) provide a model for the P-E boundary impact and potential ignition mechanism, though we note that the P-E boundary impact was smaller.

The δ¹³C_charcoal at WL and RF provides insight into our two proposed wildfire ignition scenarios and sequence of events (Fig. 3). The δ¹³C_charcoal from the charcoal peak exhibits pre-excursion values, and δ¹³C_charcoal remains constant through the start of the CIE onset in bulk carbonate. Pre-excursion δ¹³C_charcoal values indicate that: 1) the charcoal peak resulted from biomass that grew prior to the impact event and climatic perturbation; and 2) there was no vegetation growth within the time represented by the section of steady δ¹³C_charcoal values through the CIE onset and charcoal peak. If the vegetation had grown during the CIE onset, as expected during a millennial-scale event, the δ¹³C_charcoal would display excursion values because new plant growth would record the changing atmospheric δ¹³C. In contrast, our δ¹³C_charcoal decreases above peak charcoal abundance and above the start of the CIE onset in bulk carbonate. At WL, two low charcoal abundance samples above the peak exhibit pre-excursion δ¹³C_charcoal values, hence the pre-excursion δ¹³C_charcoal values in the charcoal peak do not represent a mixture of pre-excursion and excursion carbon, diluted by an over-abundance of pre-excursion vegetation that burned along with it. Because there was apparently no time for new plant growth, we conclude that sediment deposition between the impact ejecta and the δ¹³C_charcoal excursion was relatively fast (no more than centennial scale).

top`>Climate Change or Impact Event?
Under the drying scenario, the climate must have become hotter and drier as a result of the carbon cycle event. The environmental perturbation and warm temperatures associated with the PETM created conditions favourable for wildfires, yet peak charcoal occurs within the start of the onset and before the nadir of the excursion, which coincides with most observations of environmental perturbation. This suggests that the PETM climate perturbation did not cause the widespread wildfires and charcoal deposition; rather, wildfires were the consequence of an earlier event. The drying scenario is also inconsistent with the apparent massive increase in sedimentation rate at Salisbury Embayment shelf sections, which would have required an accelerated hydrologic cycle and increased precipitation and runoff, although we note that wildfires would result in significant soil loss. Interestingly, fern spikes of up to 90 % within the Marlboro Clay in the southern Salisbury Embayment suggest warm, wet environments (Self-Trail et al., 2017

Self-Trail, J.M., Robinson, M.M., Bralower, T.J., Sessa, J.A., Hajek, E.A., Kump, L.R., Trampush, S.M., Willard, D.A., Edwards, L.E., Powars, D.S., Wandless, G.A. (2017) Shallow marine response to global climate change during the Paleocene-Eocene Thermal Maximum, Salisbury Embayment, USA. Paleoceanography and Paleoclimatology 32, 710-728.

) because ferns need moist environments to complete their lifecycle (Mehltreter et al., 2010

Mehltreter, K., Walker, L., Sharpe, J. (2010) Fern ecology. Cambridge University Press, Cambridge.

), making them particularly rare in arid climates (Aldasoro et al., 2004

Aldasoro, J.J., Cabezas, F., Aedo, C. (2004) Diversity and distribution of ferns in sub-Saharan Africa, Madagascar and some islands of the South Atlantic. Journal of Biogeography 31, 1579-1604.

). If drying were the source of the P-E boundary wildfires, then there should be evidence of fires during subsequent Early Eocene Hyperthermal events, yet charcoal studies through the Early Eocene Climatic Optimum in Germany do not indicate any increase in fire activity (Robson et al., 2014

Robson, B.E., Collinson, M.E., Riegel, W., Wilde, V., Scott, A.C., Pancost, R.D. (2014) A record of fire through the Early Eocene. Rendiconti Online Societa Geologica Italiana 31, 187-188.

).

Although our findings are most consistent with impact-induced wildfires at the P-E boundary, either ignition scenario is feasible. Importantly, both continental drying and impact-induced wildfires are in agreement with a fast scenario (Wright and Schaller, 2013

Wright, J.D., Schaller, M.F. (2013) Evidence for a rapid release of carbon at the Paleocene-Eocene thermal maximum. Proceedings of the National Academy of Sciences of the USA 110, 15908-15913.

), supported by δ¹³C_charcoal analyses. In contrast, a 4 kyr onset (Zeebe et al., 2016

Zeebe, R.E., Ridgwell, A., Zachos, J.C. (2016) Anthropogenic carbon release rate unprecedented during the past 66 million years. Nature Geoscience 9, 325-329.

) requires the absence of plant growth or wildfires in the hinterlands for thousands of years; this is unlikely because fire has a revitalising effect on ecosystems and vegetation growth is quickly renewed following wildfires (Ahlegren, 1974

Ahlegren, C.E. (1974) Introduction. In: Kozlowski, T.T., Ahlgren, C.E. (Eds) Fire and Ecosystems. Academic Press, New York, 1-5.

). A sustained period without vegetative growth or wildfires for millennia is highly improbable (Bond and Wilgen, 1996

Bond, W.J., van Wilgen, B. (1996) Fire and plants. Chapman and Hall, London.

) when the lifetime of a typical large woody plant is less than a few centuries; therefore, we would expect resumption of at least background charcoal flux with post-CIE δ¹³C values during those millennia, which is not observed (Fig. 3). The charcoal record could consist of a mixture of charcoal that burned over a number of years; decades seem feasible, yet thousands of years is extremely unlikely, because we would expect to see multiple pulses of charcoal in the record (especially in proximal site RF), which we do not. The charcoal is so well preserved and displays such delicate features that we find it unlikely the material remained exposed on the landscape for thousands of years before being aggregated by floods.

A more realistic explanation of the charcoal record is that the P-E boundary sediments were deposited relatively rapidly during the start of the CIE onset. This explanation is consistent with mechanics of charcoal production and transport (see SI). Extremely high sedimentation rates are common in response to widespread landscape clearance by forest fires, resulting in massive post-wildfire debris flows and deposition of large volumes of sediment (Wells, 1987

Wells, W.G. II. (1987) The effects of fire on the generation of debris flows in southern California. Reviews in Engineering Geology 7, 105-113.

). Modern fires increase soil erosion rates by ~30-fold compared to pre-fire levels (Swanson, 1981

Swanson, F.J. (1981) Fire and geomorphic processes. In: Mooney, H.A., Bonniksen, T.H., Christensen, N.L., Lotan, J.E., Reiners, W.A. (Eds) Fire Regimes and Ecosystem Properties. USDA Forest Service General Technical Report WO-26.

). In addition, rapid mud accumulation rates are observed in the modern Amazon shelf (Kuehl et al., 1986

Kuehl, S.A., DeMaster, D.J., Nittrouer, C.A. (1986) Nature of sediment accumulation on the Amazon continental shelf. Continental Shelf Research 6, 209-225.

), demonstrating that high sedimentation rates during the CIE onset are plausible, especially during an enhanced hydrologic cycle (Kopp et al., 2009

Kopp, R.E., Schumann, D., Raub, T.D., Powars, D.S., Godfrey, L.V., Swanson-Hysell, N.L., Maloof, A.C., Vali, H. (2009) An Appalachian Amazon? Magnetofossil evidence for the development of a tropical river-like system in the mid-Atlantic United States during the Paleocene-Eocene thermal maximum. Paleoceanography 24, PA4211.

) with ample sediment supply. Effects of wildfires in coastal California show a significant increase in sediment yield, and up to ~35 times greater than average during the subsequent wetter year (Warrick et al., 2012

Warrick, J.A., Hatten, J.A., Pasternack, G.B., Gray, A.B., Goni, M.A., Wheatcroft, R.A. (2012) The effects of wildfire on the sediment yield of a coastal California watershed. GSA Bulletin 124, 1130-1146.

). These studies highlight how an increase in sedimentation and erosion rates occurs during the next major hydrologic event and is therefore an immediate (decadal scale) consequence of wildfires; this provides a framework for our fast sedimentation model. Regardless of the trigger mechanism, burnt terrestrial plant matter and eroded forest soils were transported to continental shelves in a sediment deluge during the next storm event, producing the base of the Marlboro Clay (see SI).

To summarise, a synchronous, more than twenty-fold increase in charcoal abundance is identified at the P-E boundary in two locations, indicating that intense widespread wildfires occurred during a climatic warming event. We consider two hypotheses to explain the source of the wildfires: 1) warming-induced, continental drying during the PETM; and 2) thermal radiation and/or ejecta fallout from the P-E boundary impact event. Although we cannot definitively rule out either scenario, our data strongly support the impact-induced wildfire hypothesis. Regardless of the trigger mechanism, a critical finding of this study is that the carbon isotope composition of individual charcoal grains exhibit pre-excursion values and remain constant through the onset of the carbon isotope excursion, indicating a rapid onset of the CIE that challenges the rate paradigm for the very early stages of the event by thousands of years.

top

Acknowledgements

We thank D. Kent, K. Miller, and W. Broecker for helpful discussions, and two anonymous reviewers for their thoughtful suggestions. We express our gratitude to M. Mastalerz from Indiana University for reflectance measurements. Funding was provided by the Comer Family Foundation, and NSF award #1737100.

Editor: Sophie Opfergelt

top

References

top

Supplementary Information

The Supplementary Information includes:

• Materials and Methods
• Supplementary Text
• Tables S-1 and S-2
• Figures S-1 to S-5
• Supplementary Information References

Download the Supplementary Information (PDF)