Duration and Intensity of End‐Permian Marine Anoxia

Abstract Ocean anoxia was an important kill mechanism in the end‐Permian mass extinction and uranium isotope data are among the most powerful tools for quantifying the global extent and duration of ocean deoxygenation due to the dependence of uranium isotope fractionation on bottom‐water redox conditions. Although coherent stratigraphic variation in uranium isotope ratios (δ 238 U) and uranium concentrations ([U]) indicative of prolonged deoxygenation beginning coincident with the extinction is well established, the precise extent of anoxia and associated uncertainty have yet to be quantified. Uncertainty arises from both noise in the data and imprecise knowledge of key parameters within the uranium cycle. In this study, we use the Monte Carlo method to explore a range of scenarios and their implications for the uranium cycle across the Permian‐Triassic boundary and through the first 1.7 million years of the Triassic. We then compare model predictions against measured data using principal component analysis to identify model runs and associated parameter values most compatible with trends in the observed data. The best‐fitting models indicate a pronounced increase in the extent of seafloor anoxia across the Permian/Triassic transition, reaching 18% of the seafloor (95% CI: [11%, 47%]), lasting anywhere from 20 kyr to 1.2 Myr. There is an inverse relationship between the extent and duration of anoxia in the set of best‐fitting models. This initial pulse of pronounced anoxia is followed by a prolonged aftermath, which continues through the remainder of the study interval, of less extensive, yet still expanded, anoxia covering 7.8% of the seafloor (95% CI: [1.6%, 48.9%]). Both expanded and protracted anoxia are required to fit existing data, with no indication of full re‐oxygenation during the study interval.