New Perspectives on the Bishop Tuff from Zircon Textures, Ages and Trace Elements

We present zircon textural, trace element and U–Pb age data obtained by secondary ion mass spectrometry (SIMS) (SHRIMP-RG: sensitive high-resolution ion microprobe, reverse geometry) from 15 stratigraphically controlled Bishop Tuff samples and two Glass Mountain (GM) lava samples (domes OD and YA). Bishop zircon textures divide into four suites: (a) dominant sector-zoned grains, with (b) subordinate grains showing bright rims [lower U, Th, rare earth elements (REE)] in CL imaging, (c) sparse GM-type grains (texturally similar to zircons from GM dome YA) and (d) sparse Mesozoic xenocrysts from Sierran granitoid country rocks. All Bishop zircons from suites (a)–(c) combined have a weighted mean age of 777·9 ± 2·2 ka (95% confidence) and a tail back to ∼845 ka. Our eruption age estimate using the weighted mean of 166 rim ages of 766·6 ± 3·1 ka (95% confidence) is identical within uncertainty to published estimates from isotope-dilution thermal ionization mass spectrometry (ID-TIMS) (767·1 ± 0·9 ka, 2σ) and ⁴⁰Ar/³⁹Ar (767·4 ± 2·2 ka, 2σ) techniques, the latter using the 28·172 Ma age for the Fish Canyon sanidine standard. We estimate also an eruption age for GM dome YA of 862 ± 23 ka (95% confidence), significantly older than the currently accepted 790 ± 20 ka K–Ar age. The oldest zircon cores from late-erupted Bishop material (including those with GM-type textures) have a weighted mean age of 838·5 ± 8·8 ka (95% confidence), implying that the Bishop Tuff system was active for only ∼80 kyr, and had effectively no temporal overlap with the GM system. Trace element variations in Bishop zircons are influenced strongly for many elements by sector zoning, producing up to 3× concentration differences between sides and tips within the same growth zone. Contrasting trends in molar (Sc + Y + REE³⁺)/P ratios between sides and tips indicate contrasting mechanisms of substitution in different sectors of the same crystal. Concentrations of Ti in tips are double those in the sides of crystals, hindering applicability of the Ti-in-zircon thermometer, in addition to variations inherent to the 0·15–0·67 range in values proposed for aTiO2. The bright-rim portions of grains are inferred to have crystallized from the same magma as that which generated the bright rims seen under cathodoluminescence or back-scattered electron imaging on quartz and feldspar, respectively. This less evolved, slightly hotter magma invaded the deeper parts of the chamber represented in the late-erupted northern units possibly up to ∼10 kyr prior to eruption, but invaded shallower levels only very shortly before eruption, as shown by our textural information and previously proposed from the sharp delineation of quartz bright rims. By obtaining a large number of analyses from zircon separates that systematically cover the entire Bishop Tuff eruption sequence we can produce an eruption age estimate using SIMS to the same precision and accuracy as ID-TIMS and ⁴⁰Ar/³⁹Ar techniques.