VOLUMINOUS LOW-d18O MAGMAS IN HEISE VOLCANIC FIELD, IDAHO: IMPLICATIONS FOR THE FATE OF YELLOWSTONE HOTSPOT CALDERAS

Low-d¹⁸O magmas are conspicuous features of Snake River Plain caldera complexes. Heise volcanic field, Yellowstone's immediate predecessor, contains the largest single known low-d¹⁸O unit (Kilgore tuff, 1800 km³) in the Snake River Plain and perhaps worldwide. Here we present new oxygen isotope analyses of phenocrysts and U-Pb ages of zircons (quoted) from major caldera-forming ignimbrites and post-caldera lavas of Heise volcanic field. The 6.9-2 Ma volcanism at Heise is represented by three normal-d¹⁸O caldera-forming ignimbrites (Blacktail Creek, 6.92±0.28 Ma; Walcott; Conant Creek 5.70±0.19 Ma) with d¹⁸O_Qz = 6.4‰ and d¹⁸O_Zircon = 4.8‰, followed by a dramatic ¹⁸O depletion in the final Kilgore cycle (4.45-2.07 Ma) with d¹⁸O_Qz = 4.3‰ and d¹⁸O_Zircon = 1.5‰. Post-Kilgore lavas exhibit equally low d¹⁸O compositions and mark the final stage of low-d¹⁸O volcanism at Heise volcanic field prior to the onset of normal-d¹⁸O volcanism at Yellowstone Plateau volcanic field. One low-d¹⁸O post-Kilgore unit (rhyolite of Sheridan Reservoir, 2.07±0.19 Ma), is coeval with the first, normal-d¹⁸O caldera-forming ignimbrite in the Yellowstone Plateau field (Huckleberry Ridge Tuff, 2.05±0.006 Ma). The U-Pb ages of zircons in Kilgore's post-caldera lavas did not reveal any inheritance, nor d¹⁸O zoning, from previously erupted volcanic units, in contrast to what has been found at Yellowstone. However, zircons from one post-Kilgore rhyolite (Indian Creek, 4.46-3.96 Ma) contain significant (0.5 Ma) zoning between older cores and younger rims. At both Heise and Yellowstone, low-d¹⁸O magmas occur exclusively in the later stages of individual volcanic cycles. The abrupt d¹⁸O depletions of late stage volcanic units in the caldera complexes at Heise and Yellowstone fields indicate that an ¹⁸O depleting mechanism must be present, excluding the possibility of derivation from a low-d¹⁸O mantle source or crustal protolith, in which case all volcanic units would be low-d¹⁸O. We argue that low-d¹⁸O magmas at Heise (and Yellowstone) are generated by heat of the mantle plume remelting shallow hydrothermally altered intracaldera rocks and terminate once volcanism begins at a new place tapping fertile, normal-d¹⁸O crust. This trend may be applicable to older caldera complexes in the Snake River Plain that are poorly exposed.