Shifts in magnetic mineral assemblages support ocean deoxygenation before the end-Permian mass extinction

Expansion of oceanic anoxia is a prevailing hypothesis for driving the marine end-Permian mass extinction and is mainly based on isotopic geochemical proxies. However, long-term oceanic redox conditions before the end-Permian mass extinction remain unresolved. Here we report a secular redox trend based on rock magnetic experiments and cerium anomalies through the Changhsingian and across the Permian-Triassic boundary at the Meishan section, China. Magnetic mineral assemblages changed dramatically at ca. 252.8 million years age (Ma), which indicates that oceanic deoxygenation started about 0.9 million years earlier than the end-Permian mass extinction. The magnetite-dominant post end-Permian mass extinction interval suggests a ferruginous dysoxic conditions with enhanced weathering in the earliest Triassic. Also, a gradual magnetite abundance decrease to pre-extinction levels is observed at ca. 251.5 Ma, coinciding temporally with the waning of Siberian Trap and arc volcanism. All of these observations demonstrate that environmental deterioration began much earlier than the end-Permian mass extinction and finally collapsed in the end-Permian. Oceanic deoxygenation began about one million years earlier than the marine End-Permian mass extinction, as indicated by variations in magnetic mineral assemblages and geochemical anomalies.