INSIGHTS INTO THE MAGMATIC ASSEMBLY OF A VOLUMINOUS, LOW δ18O, AND STRONGLY TRACE ELEMENT ZONED HIGH-SILICA RHYOLITE BODY: THE DEVINE CANYON TUFF, OREGON

The Devine Canyon Tuff is a crystal-rich (3 to 22%), large-volume (300 km³), high-silica rhyolitic ignimbrite that erupted ~9.7 Ma, from a presumed source in the Harney Basin of southeastern Oregon. The tuff erupted as a simple cooling unit after an initial Plinian phase with bulk rock and pumice analyses indicating that all of the tuff (>99%), except for the sparse dacite and dacite/rhyolite mixed pumices, are strongly trace-element zoned. Individual rhyolitic pumices extend the compositional range of bulk tuff samples [ppm]: e.g. Zr 611-1600, Nb 46-112, Rb 52-180, Y 75-204. All pumices analyzed fall into compositional clusters with or without gaps. Quartz and feldspar phenocrysts from the most evolved rhyolite to the least evolved dacite pumices have variable and low δ¹⁸O values. The maximum range of calculated δ¹⁸O melt values are 1.5‰ from 4.28‰ to 5.76‰. Quantitative crystallinity data on pumices (and fallout) from the rhyolite magmas reveal a span of 3 to 19% with the dacite pumice having a crystallinity of 22%. However, size distribution of phenocrysts among the rhyolites change progressively, with narrow and overlapping feldspar and pyroxene compositions across all rhyolites (sanidine with Ab: 54-59, Or: 46-41; cpx with En: 57-58 Wo: 41-41.5) and slightly wider range of minerals from dacite.

Clustered mineral compositions, banded pumices composed of different rhyolites, and low δ¹⁸O values strongly suggest different rhyolites erupted from a contiguous and shallow reservoir rather than separated reservoirs. The addition of hydrothermally altered material is evident from the low δ¹⁸O values and we suggest the addition of this material occurred prior to the trace element zonation of the highly fractionated rhyolites (Eu/Eu* 0.08-0.16). Intermediate rhyolites cannot be solely generated through the mixing of the end-member rhyolite melts but require, as minimum, some subsequent modification. Mixing of a more mafic magma appears to have generated the dacitic magma and our data do not suggest it to be a remobilized crystal mush. All this suggests magmatic evolution occurring within melt rich rhyolite magma(s) and nearby evolution of intermediate magmas rather than punctuated extraction of melt from intermediate crystal mush.