Uranium isotopes in non-euxinic shale and carbonate reveal dynamic Katian marine redox conditions accompanying a decrease in biodiversity prior to the Late Ordovician Mass Extinction

he Late Ordovician mass extinction (LOME) is the first major mass extinction event in the Phanerozoic. However, global biodiversity started to decline during the Katian (prior to the LOME) and coeval global ocean redox conditions are not well understood. The deposition of Katian organic-rich sedimentary rocks, namely the calcareous Collingwood Member and overlying siliciclastic Rouge River Member, was triggered by the concurrent Taconic Orogeny in southern Ontario (Canada) near the interface between the epicratonic Appalachian and Michigan basins. In this study, we used geochemical elemental proxies and uranium isotope compositions (δ238U) to reconstruct local and global ocean redox states during deposition of both units at seven drill core localities throughout southern Ontario. Local redox conditions are revealed by several proxies (redox sensitive trace metals, Fe speciation, and Corg:P ratios). Collectively, these proxies suggest spatiotemporal redox variations for the Collingwood Member: O2-deficient conditions (non-euxinic) for the deep waters of the Appalachian and Michigan basins and mostly oxic-suboxic conditions for the shallow waters. In contrast, predominantly oxic-suboxic conditions are suggested for the Rouge River Member.

Global ocean redox conditions are inferred from the δ238U of both units. The coeval seawater δ238U value during Collingwood Member deposition is estimated to be from ∼−0.87‰ to ∼−0.64‰, based on both carbonate δ238U (samples leached with 1 N HCl) and δ238Ubulk-non-detrital data (corrected from the bulk δ238U). Particularly, the δ238Ubulk-non-detrital data of the Collingwood Member from multiple cores (both shallow and deep waters) are positively correlated with elemental redox proxies, suggesting a local redox control on δ238U offsets between sediments and water columns. In contrast, the δ238Ubulk-non-detrital of the suboxic Rouge River Member (average = −0.32 ± 0.12‰) from several cores do not correlate with local redox proxies. The coeval seawater δ238U during Rouge River Member deposition is estimated between −0.62 ± 0.12‰ and −0.42 ± 0.12‰ assuming a δ238U offset of 0.1–0.3‰ as observed between modern suboxic sediments and seawater. For this range of seawater δ238U, a three-sink U isotope mass balance model suggests a contraction of global euxinic seafloor area from deposition of the Collingwood Member (0.5–31.6%; median = 1.0–14.2%) to the Rouge River Member (0.2–2.0%; median = 0.3–1.0%), which could have been associated with the concurrent Taconic Orogeny. Collectively, global marine redox proxy data (i.e., δ98Mo, δ238U) from this and previous studies imply dynamic Katian ocean redox conditions during the decrease of metazoan biodiversity prior to the LOME. As a similar extent of global ocean euxinia during the LOME2 also occurred episodically during the Katian but did not result in LOME2-like mass extinctions, it is suggested that other factors (e.g., climate), besides expanded ocean euxinia, could have contributed to the second phase of the LOME.