LINKAGE BETWEEN THE LARGEST PHANEROZOIC MASS EXTINCTION AND VOLCANISM: INSIGHTS FROM Hg-Ni-Cu-Zn-Zr CHEMOSTRATIGRAPHY AND C, Hg ISOTOPES AT THREE CLASSICAL P–TR TRANSITION SECTIONS

Geochemical anomalies of mercury, nickel, copper, zinc and zirconium have been proposed as potential proxies to tracking down volcanism or bolide impact events coeval with one of the Phanerozoic Big five mass extinctions. In this study, three classical P–Tr transition sections (the carbonate-dominated sections at Meishan in China, and Zal in Iran, and the siliciclastic-dominated Guryul Ravine section in India) display a larger Hg/TOC positive spike at the Late Permian mass extinction (LPME; 251.94 Ma) and a smaller one within the LPME and the P–Tr boundary interval. The Guryul Ravine section exhibits two additional Hg/TOC peaks, one at the P–Tr boundary and another at the early Triassic mass extinction (ETME) stratigraphic level. In the three sections, the LPME horizon shows a prominent negative δ¹³C_carb spike. At the Meishan and Guryul sections, Δ¹⁹⁹Hg values for the LPME horizon fall within the range typical for LIPs (-0.02 ± 0.06 ‰) and are compatible with Hg from primitive-mantle source, an assumption supported by Δ¹⁹⁹Hg‰ (MIF) vs δ²⁰²Hg ‰ (MDF) and Δ²⁰¹Hg‰ (MIF) vs Δ¹⁹⁹Hg ‰ (MIF) cross-plots. In the Zal section, Δ¹⁹⁹Hg for the LPME horizon shows a higher value (0.08 ‰), exceeding the typical range associated with LIPs. The Δ¹⁹⁹Hg values for a Hg/TOC spike in the pre-LPME interval and within the interval between the LPME and the P–Tr boundary in the three sections also slightly exceed the range for Δ¹⁹⁹Hg and outside the range for δ ²⁰²Hg (-1.85 ±0.30 ‰) expected from a primitive-mantle source. This observation raises the possibility of some role played by recycled Hg from subducting oceanic crust in an island-arc system, characterized by Δ¹⁹⁹Hg > 0 which reflects mass-independent fractionation of mercury isotopes. The second stage of the magmatic activities in the Siberian Traps Large Igneous Province (STLIP) is usually blamed responsible for the killing at the LPME and within the extinction interval. Although Hg isotopes apparently support this assumption, the same cannot be unequivocally said about the extinction recorded at the LPME and within the extinction interval at the Zal section, perhaps, linked to coeval nearby island-arc volcanism or to intermediate/felsic volcanic centers along the Simao–Indochina convergent continental margin. At the Guryul sections Hg/TOC, Ni/TOC, Cu, Zn and Zr chemostratigraphic pathways exhibit a positive spike at the LPME and ETME horizons, while at Meishan these spikes are present only at the LPME horizon. At Guryul, three positive spikes are observed in each of the Hg/TOC, Ni/TOC, Zn and Zr chemostratigraphic curves in the interval between the LPME and the P–Tr boundary at corresponding stratigraphic levels while at Meishan, the Hg/TOC, Ni/TOC and Zn chemostratigraphic curves exhibit only two positive spikes at this interval, and Zr shows none. This difference likely resulted from several factors, including: (a) depths of sedimentary deposition, gradually deeper at Guryul due to transgressional event during the extinction interval; (b) variation in continental weathering and terrigenous influence; and (c) differences in redox conditions. The use of Hg chemostratigraphy and Hg isotopes to fingerprint volcanism is a well-established proxy, but same cannot be said about the chemostratigraphy and isotopes of the other elements in this study, still in their early stages. Nickel-rich gases and aerosols released from LIPs, globally dispersed into the atmosphere, make Ni a most promising one.