Magnitude of the carbon isotope excursion at the Paleocene–Eocene thermal maximum: The role of plant community change

Carbon-isotope measurements (δ13C) of leaf-wax n-alkanes from the Paleocene–Eocene Thermal Maximum (PETM) in the Bighorn Basin, Wyoming, reveal a negative carbon isotope excursion (CIE) of 4–5‰, which is 1–2‰ larger than that observed in marine carbonate δ13C records. Reconciling these records requires either that marine carbonates fail to record the full magnitude of the CIE or that the CIE in plants has been amplified relative to the marine. Amplification of the CIE has been proposed to result from an increase in available moisture that allowed terrestrial plants to increase 13C-discrimination during the PETM. Leaf physiognomy, paleopedology and hydrogen isotope ratios of leaf-wax lipids from the Bighorn Basin, however, all suggest that rather than a simple increase in available moisture, climate alternated between wet and dry during the PETM. Here we consider two other explanations and test them quantitatively with the carbon isotopic record of plant lipids. The “marine modification” hypothesis is that the marine carbonate record was modified by chemical changes at the PETM and that plant lipids record the true magnitude of the CIE. Using atmospheric CO2δ13C values estimated from the lipid record, and equilibrium fractionation between CO2 and carbonate, we estimate the expected CIE for planktonic foraminifera to be 6‰. Instead, the largest excursion observed is about 4‰. No mechanism for altering marine carbonate by 2‰ has been identified and we thus reject this explanation. The “plant community change” hypothesis is that major changes in floral composition during the PETM amplified the CIE observed in n-alkanes by 1–2‰ relative to marine carbonate. This effect could have been caused by a rapid transition from a mixed angiosperm/conifer flora to a purely angiosperm flora. The plant community change hypothesis is consistent with both the magnitude and pattern of CIE amplification among the different n-alkanes, and with data from fossil plants. This hypothesis predicts that the magnitude and pattern of amplification of CIEs among different n-alkanes will vary regionally and systematically depending on the extent of the replacement of conifers by angiosperms during the PETM.