By decoding carbon trapped in ancient plant matter, scientists track carbon fluxes back beyond ocean records.
Researchers have shown that fossilized plant matter, sampled from two rock formations in Wyoming dating back 50 to 60 million years, can reliably reconstruct what Earth’s atmosphere looked like at the time, closely matching independent estimates drawn from ancient ocean fossils. The finding offers scientists a promising avenue for tracing the planet’s carbon cycle back further in time than ocean records alone can reach.
Delta Carbon-13 (δ13C), a ratio of the two stable and most common forms of carbon, is a measure of carbon flux and the global carbon cycle. One form (Carbon-13) is slightly heavier than the other (Carbon-12), and so the δ13C ratio is used to determine the source of carbon in the atmosphere. Researchers have long used various marine proxies for analyzing how the carbon cycle has changed with the changing climate. These proxies include ocean sediments and the shells of organisms long passed, going back as far as the Cretaceous Period, about 145 Ma (million years ago). But even these records, starting from the Middle Eocene some 45 Ma, have suboptimal resolution and older records are inaccessible.
The Cretaceous and adjacent time periods harbor some of the most crucial climatic and environmental events, including the K-Pg extinction. Some climatic events during the Cretaceous, Paleocene and Eocene are similar to modern-day global warming, and understanding the impact of carbon fluxes seen in the past could help us fathom the challenges that lie ahead. But with δ13C data for these periods remaining fuzzy, so too is our understanding.
Nathan Sheldon and Katarina Keating of the University of Michigan (Ann Arbor) in the US and colleagues turned to fossilized plant and other organic matter to trace δ13C from the early Eocene all the way back to the late Paleocene (∼60–50 Ma). They exploited the fact that plants use and trap carbon dioxide from their environs to grow, and this terrestrial organic matter can be preserved over the geological timescale. Sheldon and Keating used this carbon trapped in the land to reconstruct atmospheric carbon fluxes in the past.
For this study, the researchers sampled sediments, going several meters underground, from two sites in the Hoback Basin in Wyoming. They processed and analyzed all samples for δ13C from terrestrial organic matter. Then the researchers compared their numbers with other global terrestrial records from the literature to determine values of δ13C in the atmosphere millions of years ago. They also compared these values to those from marine organisms fossilized over the Cenozoic period going back 66 Ma, to see if terrestrial values would make for a good proxy for atmospheric values of δ13C.
Sheldon and Keating found that their reconstructions of atmospheric δ13C values matched those previously derived from ancient marine organisms.
To understand how the climate or environs might be causing carbon fluxes, these reconstructed values are usually compared to a baseline. Volcanoes release carbon in a steady fashion, with a δ13C of −5.4‰; the reconstructed measure for the atmosphere during the Eocene, some 53 Ma, was −6.30‰, meaning it was lacking in the heavier form, Carbon-13. With lighter carbon abounding in the atmosphere, the period was warmer.
On the other hand, the same measure flipped to −4.37‰ during the late Paleocene, some 57 Ma. The researchers suggested that more light carbon was being locked away in the ocean or elsewhere during this period, helping to cool the climate, leaving behind heavier carbon in the atmosphere. Moreover, they noted similar values in late Paleocene records drawn from both terrestrial and oceanic records, implying a common process is responsible.
“Our results using this ancient plant matter align well with independent estimates using marine records,” write the researchers. “This provides a method for understanding the sources and sinks of carbon dioxide for as long as plants have existed on Earth.”
The team did come across some complications. When they compared the reconstructed values to those previously measured at other sites globally, the results varied with rainfall: sites that had been wetter in the past gave readings that had less heavy carbon, and drier sites had values like those from the Hoback Basin. The researchers suggest that precipitation and its impact need to be ruled out in wetter sites to get more accurate readings of carbon fluxes from fossilized plants. Despite this concern, “this new approach can supplement the marine-derived δ13Catm (atmospheric Delta Carbon-13) record in data-poor parts of the Cenozoic and Mesozoic,” add the researchers, “as well as enable deeper time reconstructions for time periods without any marine constraints.”
Reference: Katarina A. Keating, et al., Using Terrestrial Organic Matter to Reconstruct the Stable Carbon Isotope Composition of the Atmosphere, Paleoceonography and Paleoclimatology (2026). DOI: https://doi.org/10.1029/2025PA005329
Featured Image Credit: Eddie Mark Blair via Unsplash
