U–Pb zircon geochronology of silicic tuffs from the Timber Mountain/Oasis Valley caldera complex, Nevada: rapid generation of large volume magmas by shallow-level remelting

Abstract

Large volumes of silicic magma were produced on a very short timescale in the nested caldera complex of the SW Nevada volcanic field (SWNVF). Voluminous ash flows erupted in two paired events: Topopah Spring (TS, >1,200 km3, 12.8 Ma)–Tiva Canyon (TC, 1,000 km3, 12.7 Ma) and Rainier Mesa (RM, 1,200 km3, 11.6 Ma)–Ammonia Tanks (AT, 900 km3, 11.45 Ma; all cited ages are previously published 40Ar/39Ar sanidine ages). Within each pair, eruptions are separated by only 0.1–0.15 My and produced tuffs with contrasting isotopic values. These events represent nearly complete evacuation of sheet-like magma chambers formed in the extensional Basin and Range environment. We present ion microprobe ages from zircons in the zoned ash-flow sheets of TS, TC, RM, and AT in conjunction with δ18O values of zircons and other phenocrysts, which differ dramatically among subsequently erupted units. Bulk zircons in the low-δ18O AT cycle were earlier determined to exhibit ∼1.5‰ core-to-rim oxygen isotope zoning; and high-spatial resolution zircon analyses by ion microprobe reveal the presence of older grains that are zoned by 0.5–2.5‰. The following U–Pb isochron ages were calculated after correcting for the initial U–Pb disequilibria: AT (zircon rims: 11.7 ± 0.2 Ma; cores: 12.0 ± 0.1 Ma); pre-AT rhyolite lava: (12.0 ± 0.3 Ma); RM: 12.4 ± 0.3); TC: (13.2 ± 0.15 Ma); TS: (13.5 ± 0.2). Average zircon crystallization ages calculated from weighted regression or cumulative averaging are older than the Ar–Ar stratigraphy, but preserve the comparably short time gaps within each of two major eruption cycles (TS/TC, RM/AT). Notably, every sample yields average zircon ages that are 0.70–0.35 Ma older than the respective Ar–Ar eruption ages. The Th/U ratio of SWNVF zircons are 0.4–4.7, higher than typically found in igneous zircons, which correlates with elevated Th/U of the whole rocks (5–16). High Th/U could be explained if uranium was preferentially removed by hydrothermal solutions or is retained in the protolith during partial melting. For low-δ18O AT-cycle magmas, rim ages from unpolished zircons overlap within analytical uncertainties with the 40Ar/39Ar eruption age compared to core ages that are on average ∼0.2–0.3 My older than even the age of the preceding caldera forming eruption of RM tuff. This age difference, the core-to-rim oxygen isotope zoning in AT zircons, and disequilibrium quartz–zircon and melt-zircon isotopic fractionations suggest that AT magma recycled older zircons derived from the RM and older eruptive cycles. These results suggest that the low-δ18O AT magmas were generated by melting a hydrothermally-altered protolith from the same nested complex that erupted high-δ18O magmas of the RM cycle only 0.15 My prior to the eruption of the AT, the largest volume low-δ18O magma presently known.