292-14 Climate Across the K-Pg Boundary: Evidence from δ¹⁸O of Freshwater Unionid Bivalves from Montana

Abstract:

Terrestrial climate change during the K-Pg extinction and subsequent recovery is not well constrained. Mass extinction models often emphasize mean annual temperature change, but seasonal extremes may have more direct biological effects on organisms. Individual freshwater unionid bivalves record seasonal δ¹⁸O and δ¹³C variation within their shells, sometimes truncated during peak summer or winter. We examine these shells (n=54) as an archive of seasonal isotopic change across the Cretaceous-Paleogene (K-Pg) boundary in the Hell Creek and Fort Union (Tullock, Lebo, and Tongue River members) formations of central and eastern Montana.

Unionid shells are often geochemically well-preserved, but many localities, particularly in the Lebo and Tongue River members, yield only fragments rather than whole valves despite pristine aragonite mineralogy. We tested the isotopic fidelity of fragmentary material by serially sampling whole valves, then pulverizing the remaining half and sampling a random array of fragments (≥30 per specimen). Fragmentary sampling reliably reproduces mean δ¹⁸O, but recovery of seasonal extrema is variable and does not consistently match serially-derived ranges, complicating extrema interpretation where records rely solely on fragments.

We find a statistically significant decrease in δ¹⁸O extrema from Hell Creek to Tongue River: average δ¹⁸O maxima decrease from -9.4‰ to -12.2‰ over the record (Jonckheere-Terpstra test on n=54,  p<0.01), and average δ¹⁸O minima decrease from -14.6‰ to -19.3‰ (p<0.01). Mean δ¹⁸O shell shows no significant temporal trend, confining this signal of environmental change specifically to extrema. δ¹⁸O distribution shape shifts through the section, indicating the persistence of a bimodal δ¹⁸O distribution into the Paleocene, apart from the Tullock specimens which are distinctly unimodal.

Shells with more negative δ¹⁸O means tend to have larger δ¹⁸O ranges, a pattern present in both serially- and fragment-sampled shells, a pattern that persists after controlling for formation identity and sampling density. δ¹³C shows no comparable relationship.

δ¹³C remains unimodal and stable (mean ~-5‰) throughout, with no obvious temporal trend, though these values cannot be reliably interpreted in fragmentary material due to vital effects on the inner shell layer. Together, these results support changing hydrology and/or regional climate across the K-Pg boundary in this basin.

© Copyright 2026 The Geological Society of America (GSA), all rights reserved.