Si isotope ratio of radiolaria across Triassic–Jurassic transition in a pelagic deep-sea bedded chert (Inuyama, Japan)

Abstract

The end-Triassic extinction event (ETE) marks one of the “Big five” mass extinction events of the Phanerozoic, yet the timing and nature of environmental changes on a global scale remain elusive. Here we report a mm-scale high-resolution δ30Si profile of sea surface-dwelling radiolaria, preserved as moulds, spanning the end-Triassic radiolarian turnover interval of the deep-sea succession at the Katsuyama section, Inuyama, Japan. The δ30Si of the Triassic-Jurassic radiolaria between − 0.6 ± 0.5‰ and 2.6 ± 0.3‰ overlap with available radiolarian δ30Si data for Triassic to Cenozoic, including modern samples. The mass balance of Si with dominant radiolaria up to 90% in chert supports negligible impact of diagenesis on δ30Si of radiolarian moulds. The cm-scale δ30Si variations up to 2 ‰ are overall associated with changes in silica contents. Here we hypothesize that the δ30Si of radiolarian reflect radiolarian productivity. We detected negative silicon isotope excursions within and above the end-Triassic radiolarian turnover interval, named here the SIE 1 and 2. The first SIE of 2 ‰ is detected from 0 mm to 10 mm above the base of purple chert interval, probably linked with the initial massive volcanism in the Central Atlantic magmatic provinces (CAMP) and associated deep-sea acidification and the onset of the end-Triassic radiolarian turnover. Subsequent positive SIE up to 2 ‰ ca. 10 mm above the first negative SIEs may record recovery of biosiliceous productivity, which could may be related with the first appearance of several Jurassic radiolarian taxa within the radiolarian turnover interval. We also detected data that is suggestive of SIE 2 in the lowermost dusky red chert bed, suggesting similar changes in productivity. These negative SIEs occurred within ~ 1 mm interval, implying the occurrence of kyr- or shorter-scale drastic environmental perturbations across the end-Triassic radiolarian turnover interval.