Paleoatmospheric pCO₂ fluctuations across the Cretaceous–Tertiary boundary recorded from paleosol carbonates in NE China

Abstract

A dramatic change in atmospheric composition has been postulated because of global carbon cycle disruption during the Cretaceous (K)–Tertiary (T) transition following the Chicxulub impact and Deccan Trap eruptions. Here, pedogenic carbonates were collected from drill core of a borehole (SK-1 (N)) straddling the Late Cretaceous and early Paleocene strata in the Songliao Basin, northeast China, to reconstruct atmospheric CO₂ concentrations using a paleosol paleobarometer. Our estimates for atmospheric pCO₂ from paleosol carbonates range between 277 ± 115 ppmv and 837 ± 164 ppmv between 67.8 Ma and 63.1 Ma. One large (~ 66–65.5 Ma) and several small CO₂ spikes (~ 64.7–~ 64.2 Ma) during the latest Maastrichtian to earliest Danian are reported here and incorporated with previously published pCO₂ estimates also estimated from paleosol carbonates. These CO₂ spikes are attributed to one-million-year-long emplacement of the large Deccan flood basalts along with the extraterrestrial impact at the K–T boundary.

Introduction

The K–T boundary was marked by one of the largest mass extinctions during the past 500 million years (Peters, 2008), and several hypotheses have been proposed to explain the mass extinction at the K–T boundary. The impact hypothesis was introduced to account for the mass extinction (Alvarez et al., 1980), and increasing numbers of scientists attribute the mass extinction to the Chicxulub impact event (Hildebrand et al., 1991, Kring, 2007, MacLeod et al., 2007, Miller et al., 2010, Schulte et al., 2010). Global environmental consequences of the impact included release of large quantities of water, dust, and climate-forcing sulfuric and nitric acidic gases (Retallack, 1996), extensive combustion of biomass or fossil organic matter (Wolbach et al., 1988, Melosh et al., 1990, Ivany and Salawitch, 1993, Jones and Lim, 2000, Belcher et al., 2009), and mega-tsunami and ejecta debris deposition (Claeys et al., 2002). Alternatively, a continental flood basalt hypothesis has also been used to explain the K–T mass extinction pattern, due to abrupt global cooling resulting from the voluminous release of sulfur dioxide and dust into the atmosphere for single eruptive events in the Deccan flood basalt traps (Keller et al., 2008, Chenet et al., 2009, Courtillot and Fluteau, 2010), or to later greenhouse warming with increase of atmospheric CO₂ once dust, soot and aerosols fell to the ground (Duncan and Pyle, 1988, O'Keefe and Ahrens, 1989, Crowley and Berner, 2001). In addition, multicausal models including impact, volcanic activity, marine regression, and changes in global and regional climatic patterns have been linked to the extinction event (Keller, 2001, Keller et al., 2003, Keller et al., 2009, MacLeod, 2003, Archibald et al., 2010, Keller et al., 2010).

A significant perturbation of the global carbon cycle has been predicted from extinctions themselves, as well as from impact and volcanic eruption near the K–T boundary. It was hypothesized that atmospheric CO₂ would rise dramatically across the K–T transition due to massive amounts of CO₂ from Chicxulub's target carbonate-rich lithologies and the projectile (O'Keefe and Ahrens, 1989, Agrinier et al., 2001, Kring, 2007), from widespread large wildfires (Melosh et al., 1990, Wolbach et al., 1990, Ivany and Salawitch, 1993, Durda and Kring, 2004), from intruded or impacted coal or hydrocarbons (Belcher et al., 2005, Harvey et al., 2008, Belcher et al., 2009), from reduction in worldwide marine primary productivity (D'Hondt et al., 1998, Aberhan et al., 2007, Maruoka et al., 2007), and from degassing of mantle volatiles during several short eruptions of the Deccan Traps (Courtillot et al., 1986, Officer et al., 1987, Self et al., 2006, Kring, 2007, Chenet et al., 2009).

Estimated atmospheric CO₂ concentrations across the K–T transition are tests of these hypotheses. An abrupt pCO₂ fluctuation at the K–T boundary has been examined using both stomatal index of fossil plants (Beerling et al., 2002, Retallack, 2009a) and a paleosol barometer (Nordt et al., 2002, Nordt et al., 2003). However, disparity between magnitude and duration of CO₂ concentration in these studies highlights the need for more records with greater precision and temporal resolution (Arens and Jahren, 2002, Retallack, 2004). In the past, lack of information about key parameters such as soil respiration for the pedogenic CO₂ paleobarometer of Cerling (1991) have limited their precision in determining ancient CO₂ levels, but now a variety of proxies for soil respiration are available (Retallack, 2009b, Breecker et al., 2010, Royer, 2010, Cotton and Sheldon, 2012).

The Songliao Basin of China has thick sequences of Jurassic–Paleogene terrestrial strata (Wan et al., 2007), including carbonate-nodule-bearing paleosols suitable for determination of paleoatmospheric CO₂ (Huang et al., 2010). Our study uses selected carbonate paleosols in northeast China in order to: (1) estimate atmospheric pCO₂ levels across the K–T boundary using a paleosol CO₂ paleobarometer and supplement the global database of CO₂ concentrations; and (2) indicate the source(s) for the change in pCO₂, if any.