Analyzing sources of uncertainty in terrestrial organic carbon isotope data: A case study across the K-Pg boundary in Montana, USA

https://doi.org/10.1016/j.palaeo.2021.110451Get rights and content

Highlights

  • δ13Corg values from hand trenches are more positive than those from cores.

  • Despite trenching, modern contamination/alteration affects δ13Corg values.

  • Negative δ13Corg excursions are associated with organic rich lithologies.

  • The K-Pg boundary associated δ13Corg excursion is affected by lithological variation.

Abstract

The Cretaceous-Paleogene boundary (KPB) in the Hell Creek area of Montana is recognized in several places by an iridium anomaly, which is typically identified at or very near the lithological contact between the Hell Creek Formation and the Tullock Member of the Fort Union Formation. Previous work in the area has argued that organic carbon isotope (δ13Corg) excursions can be used for chemostratigraphic correlation within these continental depositional environments, most importantly for the identification of the KPB where impact evidence is unavailable. However, it is unclear how modern surficial weathering affects δ13Corg values, particularly in terrestrial depositional settings, and whether standard sampling methods are sufficient to obtain unaltered rock material. We tested the fidelity of the terrestrial δ13Corg record with respect to surficial alteration and contamination through investigation of different field sampling techniques, including hand trenches, a backhoe-excavated trench, and sediment coring. We find that δ13Corg values in hand and backhoe trenched sections are more positive than δ13Corg values in cored sections, implying that modern surficial alteration affects δ13Corg values but not overall trends. A negative δ13Corg excursion associated with the KPB is present in most sections we analyzed, but it does not appear to be unique within our sections. The KPB excursion occurs within a coal layer, and we observe similar excursions within other carbon-rich lithologies. Given that we cannot disentangle local lithological effects from global atmospheric changes, we conclude that a negative δ13Corg excursion is not an unequivocal indicator of the KPB in the Hell Creek area.

Introduction

Many major events in Earth history, including mass extinctions and ocean anoxic events, are marked by significant changes to the carbon cycle that can be recognized and characterized by shifts in the carbon isotope ratios (δ13C) of a variety of organic and inorganic carbon reservoirs. These shifts, often dubbed carbon isotope excursions, can be used to correlate these events across the globe, particularly in the marine realm (e.g., Saltzman and Thomas, 2012). Additionally, the magnitude or direction of the δ13C excursions can reveal information about the nature of global carbon cycle changes. However, the greater spatial heterogeneity of continental environments compared to marine realms raises questions as to whether variability in continental δ13C records is necessarily driven by changes in atmospheric carbon isotope composition, especially on short temporal scales (Grandpre et al., 2013; Therrien et al., 2007).

The end-Cretaceous mass extinction resulted in the extinction of non-avian dinosaurs, ammonites, and many other important terrestrial and marine taxa (e.g., Fastovsky and Bercovici, 2016; Henehan et al., 2019; Landman et al., 2014; Longrich et al., 2012). At the Cretaceous-Paleogene (K-Pg) boundary (KPB) an approximately 2‰ negative excursion in δ13C is reported in both terrestrial organic carbon (Arens and Jahren, 2000; Arinobu et al., 1999; Bourque et al., 2021; Grandpre et al., 2013; Maruoka et al., 2007; Schimmelmann and DeNiro, 1984; Therrien et al., 2007) and marine inorganic carbon in many (but not all) sections (D'Hondt, 2005; Keller and Lindinger, 1989; Schulte et al., 2010; Sepúlveda et al., 2019). This excursion is interpreted to have been the result of a significant perturbation to the carbon cycle associated with the asteroid impact that substantially or entirely drove the K-Pg mass extinction (Archibald et al., 2010; Hull et al., 2020; Schulte et al., 2010).

Here we present results from sedimentary organic carbon isotope (δ13Corg) sampling across the KPB from the Hell Creek region of Garfield County in northeastern Montana, USA (Fig. 1). The Hell Creek region has featured prominently in debates regarding the timing and causes of the K-Pg mass extinction in the terrestrial realm (Archibald and Clemens, 1982; Clemens et al., 1981; Smit and Van Der Kaars, 1984; Wilson, 2014). In this region, the Hell Creek Formation (Khc; mostly Cretaceous) and Tullock Member of the Fort Union Formation (Pgft; mostly Paleogene) compose a record of nearly continuous alluvial, lacustrine, and palustrine deposition that spans the KPB, with the boundary clay itself often being recognized locally (within our sampling area, Fig. 1) as nearly synchronous with the lithological contact between the Khc and Pgft (Fig. 2) (Baadsgaard et al., 1988; Fastovsky and Bercovici, 2016; Hartman et al., 2014; Moore et al., 2014). As such, the Hell Creek region has a rich history of paleontological (Archibald, 1982; Horner et al., 2011; Lofgren, 1995; Scannella et al., 2014; Smith et al., 2018; Wilson, 2014; Wilson, 2013; Wilson, 2005) and geological research (Fastovsky, 1987; Fendley et al., 2019; Hartman et al., 2014; Ickert et al., 2015; Noorbergen et al., 2018; Sprain et al., 2018; Sprain et al., 2015; Tobin et al., 2014).

Previous δ13Corg studies in this region (Arens et al., 2014; Arens and Jahren, 2002; Arens and Jahren, 2000; Gardner and Gilmour, 2002; Maruoka et al., 2007) have argued that a δ13Corg excursion at or near the KPB is correlative with the globally recognized marine carbon isotope excursion, though the interpretation of these records as globally relevant has been questioned (Grandpre et al., 2013). The duration of the proposed KPB δ13C excursion differs by several orders of magnitude depending on the location examined (Arens and Jahren, 2000; Sepúlveda et al., 2019), which makes it difficult to determine which records reflect global atmospheric effects, and which are driven by local conditions. In the Hell Creek region, the proposed δ13Corg excursion likely lasted less than 10 k.y. before returning to pre-excursion values (Renne et al., 2013), but many marine records show carbon isotope recovery that lasted 100–1000 k.y. (e.g., D'Hondt, 2005). This difference in duration between continental and marine records has been used to argue that the terrestrial biosphere recovered more quickly from the extinction (Arens and Jahren, 2000), and these short duration excursions have guided modeling of this time period (e.g., Milligan et al., 2019). However, substantial decoupling between the duration of marine versus terrestrial carbon isotope excursions has not been noted in other global carbon cycle disruptions; for example, continental and marine carbon isotope records from the Paleocene-Eocene Thermal Maximum have similar durations, on the order of 100 k.y. (Tipple et al., 2011).

This study tests the importance of two factors that may complicate bulk sedimentary δ13Corg records, particularly those taken from continental paleoenvironments. Specifically, we test the effects of (1) modern surficial weathering and/or contamination, and (2) changing depositional environments, as reflected by differences in lithology. We then test whether these factors complicate interpretation of δ13Corg records across the KPB. There are other factors that may affect δ13Corg values, including burial diagenesis, but our tests focus on sections that are within 3 km of each other and have experienced an identical burial history.

First, modern surface conditions may affect sedimentary organic carbon records obtained from bulk sediment. Depositional organic carbon may be altered or removed through a variety of processes, including oxidation, dissolution, or bacterial metabolism (See supplementary material for quantitative discussion of these processes). These processes have the potential to fractionate carbon, altering the δ13Corg value of the remaining organic material. Additionally, modern plants may contribute carbon to the sedimentary organic pool through root infiltration, which if unrecognized will bias the δ13Corg value of sedimentary organic material. This effect can be strong because modern C4 plants (approximately −13‰ VPDB) have much higher δ13Corg values (e.g., Morgan et al., 1994) than Cretaceous plants, which exclusively used the C3 photosynthetic pathway (approximately −27‰ VPDB).

Second, bulk sediment from different continental depositional environments may record different δ13Corg values due to the deposition of different plant species, or even different tissues from the same plant (e.g., leaves vs. woody debris), which can have measurably different isotopic values (e.g., Bögelein et al., 2019; Graham et al., 2019). River migration or avulsion can rapidly change the local environment from fluvial to palustrine or lacustrine (e.g., Hupp et al., 2019). If these different environments accumulate carbon with different δ13Corg values, these geologically rapid facies shifts may be accompanied by equally rapid δ13Corg changes without any shift in the δ13C value of atmospheric CO2. The spatial heterogeneity of continental environments can potentially affect not only the stratigraphic (and therefore temporal) record of δ13Corg, but also the spatial record, meaning temporally correlative sections could record different δ13Corg patterns (see Beerling and Royer, 2002 for further discussion of these issues).

Previous work in the Hell Creek region (Arens et al., 2014; Arens and Jahren, 2002; Arens and Jahren, 2000) argues for two hypotheses that we test here: (1) original δ13Corg values recovered from hand trenches are not altered due to surface weathering or contamination, and (2) the δ13Corg of terrestrial organic matter is regionally consistent, regardless of local-scale depositional (and therefore lithological) variation. Those authors concluded that both hypotheses were supported, and subsequently that negative δ13Corg excursions can be used as chemostratigraphic markers, most notably to recognize the KPB (see Section 5.2).

To test whether modern surficial processes can alter sedimentary δ13Corg values, we employed three different sampling methods at nearly the same location to test whether surface diagenetic effects create measurable differences in the δ13Corg values (see Section 5.1). Specifically, sediment samples were obtained from shallow trenches dug by hand, a deep trench dug by an excavator, and sediment cores.

We also tested the effects of depositional control on δ13Corg values by applying the same hand-trenched sampling method to three sites across a ~ 3 km2 area and controlling for lithology in our data analysis. We then compared our results with previously published data to test whether terrestrial δ13Corg records can be used to unambiguously identify the KPB and constrain atmospheric conditions in the immediate aftermath of the Chicxulub bolide impact (see Sections 5.2.1, 5.2.2).

Section snippets

Geologic overview

The Khc is largely composed of meter-bedded, drab-colored mudstone and sandstone, typically with gradational contacts between beds. In contrast, the Pgft is largely composed of dm- to m-bedded, more brightly-colored mudstone and sandstone beds with sharp, clearly defined contacts between beds. Bedding contacts are particularly evident at laterally continuous lignite and coal layers; the maturity of these lithologies varies, but hereafter we will use “coal” to indicate these lithologies (see

Field sampling

Hand trenches at Nirvana, Iridium Hill Annex and Worm Coulee were dug with rock hammers and picks, and the deeper trench was excavated with a backhoe at Nirvana. Rock samples were obtained from trenches in the field using standard field tools (e.g., hammers, trowels) and placed in aluminum foil to avoid organic contamination from sample bags. Samples were placed in their stratigraphic context as the section was measured (Jacob's staff and Abney level) and logged, typically at dm resolution. Two

Nirvana hand trench

A total of 28 samples were obtained from 7.7 m of exposed stratigraphic section, at a sampling interval of 30 cm, with an increase to 10 cm near the IrZ coal (Fig. 3D). Most samples (n = 25) had sufficient organic carbon to measure reliable δ13Corg values, but a small number (n = 3) were too low to obtain reliable values given the maximum permitted sample size (see Section 3.2). The low organic samples were usually from sandstones or other sandier lithologies, a pattern previously observed in

Testing modern surficial effects on δ13Corg

The offset in δ13Corg values between trenched and cored samples (Fig. 7, Fig. 8) implies that trenched samples must have been more affected by modern surface conditions, despite efforts to recover superficially unaltered material. It is beyond the scope of this study to unequivocally determine the cause of this δ13Corg offset, but it could result from either in situ alteration of existing organic material or contamination with modern organic carbon from several sources (or both). This

Conclusions

Modern surface weathering and/or contamination affects absolute δ13Corg values, at least within the Khc and Pgft, and therefore absolute δ13Corg values should be interpreted with some caution, especially when recovered from hand-dug trenches. However, the pattern and magnitude of δ13Corg changes is largely preserved despite a roughly 1.0‰ offset towards more positive values in surface trenches, though the magnitude and direction of offset generated by surficial alteration is likely to be

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

We acknowledge that the work presented in this paper was performed on lands that are the traditional territory of the Fort Belknap Assiniboine & Gros Ventre Tribes and Fort Peck Assiniboine & Sioux Tribes. We dedicate this paper to the memory of William A. Clemens, and thank him for advice throughout this project, and his lifelong contributions to paleontology in the Hell Creek region.

We would like the thank Dale and Jane Tharp for their hospitality, access to and through their land, equipment

References (58)

  • J.D. Archibald

    Fossil mammalia and testudines of the Hell Creek formation, and geology of the Tullock and Hell Creek formations, Garfield Co., Montana

    Univ. Calif. Publ. Earth Sci.

    (1982)
  • J.D. Archibald et al.

    Late cretaceous extinctions: the fossil evidence currently available suggests that changes across the Cretaceous-Tertiary boundary were the result of gradual processes and not of a single catastrophic event

    Am. Sci.

    (1982)
  • J.D. Archibald et al.

    Cretaceous extinctions: multiple causes

    Science

    (2010)
  • N.C. Arens et al.

    Carbon isotope excursion in atmospheric CO2 at the Cretaceous-Tertiary boundary: evidence from terrestrial sediments

    Palaios

    (2000)
  • N.C. Arens et al.

    Chemostratigraphic correlation of four fossil-bearing sections in southwestern North Dakota

  • N.C. Arens et al.

    Carbon isotope stratigraphy and correlation of plant megafossil localities in the Hell Creek Formation of eastern Montana, USA

  • T. Arinobu et al.

    Spike of pyrosynthetic polycyclic aromatic hydrocarbons associated with an abrupt decrease in δ13C of a terrestrial biomarker at the Cretaceous-Tertiary boundary at Caravaca, Spain

    Geology

    (1999)
  • H. Baadsgaard et al.

    A radiometric age for the Cretaceous-Tertiary boundary based upon K-Ar, Rb-Sr, and U-Pb ages of bentonites from Alberta, Saskatchewan, and Montana

    Can. J. Earth Sci.

    (1988)
  • D.J. Beerling et al.

    Fossil plants as indicators of the phanerozoic global carbon cycle

    Annu. Rev. Earth Planet. Sci.

    (2002)
  • R. Bögelein et al.

    Differences in carbon isotope leaf-to-phloem fractionation and mixing patterns along a vertical gradient in mature European beech and Douglas fir

    New Phytol.

    (2019)
  • R.W. Brown

    Distribution of plant communities in Southeastern Montana badlands

    Am. Midl. Nat.

    (1971)
  • W.A. Clemens et al.

    Out with a Whimper not a Bang

    Paleobiology

    (1981)
  • A.J. Collier et al.

    The Coal Resources of McCone County, Montana

    (1939)
  • S. D’Hondt

    Consequences of the cretaceous/Paleogene mass extinction for marine ecosystems

    Annu. Rev. Ecol. Evol. Syst.

    (2005)
  • D.E. Fastovsky

    Paleoenvironments of vertebrate-bearing strata during the Cretaceous-Paleogene transition, eastern Montana and western North Dakota

    Palaios

    (1987)
  • D.E. Fastovsky et al.

    Sedimentology, stratigraphy, and extinctions during the Cretaceous-Paleogene transition at bug Creek, Montana

    Geology

    (1986)
  • A.F. Gardner et al.

    Organic geochemical investigation of terrestrial Cretaceous-Tertiary boundary successions from Brownie Butte, Montana, and the Raton Basin, New Mexico

  • H.V. Graham et al.

    Canopy structure in Late Cretaceous and Paleocene forests as reconstructed from carbon isotope analyses of fossil leaves

    Geology

    (2019)
  • J.H. Hartman et al.

    Context, naming, and formal designation of the Cretaceous Hell Creek Formation lectostratotype, Garfield County, Montana

  • Cited by (3)

    View full text