Messing up the mush model? Single crystal constraints from the Miocene Jarbidge Rhyolite (USA)

Much attention has been recently focused on understanding the petrogenesis and geologic significance of voluminous crystal-rich dacitic ignimbrites and their relationship(s) to crystal-poor high-Si rhyolite magmas, which in the mush-model, are volumetrically minor relative to their less evolved counterparts in a given magmatic system. Most of these studies have focused on voluminous magma bodies that erupted explosively, partially due to the fact that such eruptions are characterized by far-ranging and often catastrophic effects. However, numerous examples of fairly voluminous, crystal-poor effusive rhyolites exist in the geologic record. Well documented examples of these (e.g. Taylor Creek Rhyolite, Bearhead Rhyolite, rhyolites of the Coso volcanic field, lava flows from the Bruneau-Jarbidge eruptive center, etc; Bonnichsen, 1982; Bacon et al., 1981; Duffield and Dalrymple, 1990; Justet et al., 2001) all share in common a scenario where high-silica (and relatively crystal-poor) rhyolite lava flows were derived from an underlying, temporally and spatially extensive magma body. Implicit in these examples is that the underlying magma body was a compositionally and physically uniform liquid-crystal mush that was infrequently (or not at all) sampled by the effusive eruptions (Bachmann and Bergantz, 2008). As a result, direct or indirect sampling of the underlying mush is generally dependant on explosive eruptions that have evacuated more voluminous portions of these types of magma bodies, thus also the focus on explosive eruptions mentioned earlier. Examples of voluminous, crystal-rich silicic effusive products are sparse in the published geologic literature. Thus we present geologic constraints on this style of magmatism, including single crystal plagioclase ⁸⁷Sr/⁸⁶Sr isotope data, from samples of the mid-Miocene Jarbidge Rhyolite (JR; NV-USA). Sampled JR lava flows erupted from ~16.6-15.0 Ma and represent the effusive eruptions of ~500 km³ of ferroan calc-alkalic rhyolite (71-77 wt% SiO₂ whole rock). JR lava flows are crystal rich (~20-40% modally) and are dominated by an assemblage of smoky quartz (up to 6 mm), plagioclase (An_10-40), sanidine, and pyroxene. The mineralogy/mineral textures, major, trace, and REE (e.g. "hot and dry") compositions of JR samples are consistent with fractional crystallization and assimilation playing a major role in their petrogenesis. ⁸⁷Sr/⁸⁶Sr isotope values from plagioclase from six JR samples (obtained by LA-MC-ICPMS) range from ~0.709 to 0.721. Significant range in ⁸⁷Sr/⁸⁶Sr exists between and in some cases, within samples. The single grain complexity in the plagioclase cargo in terms of Sr isotopes is also observed in sanidine crystals that have polymodal ⁴⁰Ar/³⁹Ar results (Callicoat et al., in review) suggesting that JR lava flows erupted from a long-lived magmatic system. These data, coupled with JR field relationships, and its high crystallinity, suggest that the JR in our study area represents a noteworthy example of an effusively erupted rhyolite mush body. Furthermore, based on this study, the application of crystal-mush models to explain the generation of crystal-poor high-Si magma generation must also consider the existence of "source" mushes with broadly similar compositions.