FROM ERUPTED BASALTS TO RHYOLITE: EVIDENCE THAT PHENOCRYST GROWTH OFTEN OCCURS RAPIDLY DURING ASCENT ALONG FRACTURES AND NOT NECESSARILY IN CRUSTAL RESERVOIRS

Among the subset of magmas that successfully erupt on Earth’s surface, there is growing evidence that their phenocrysts grew rapidly (over days) during ascent along fractures, in contrast to prevailing models that favor slow growth of large crystals (over ≥10³ years) in crustal reservoirs. The first case to examine are basalts that contain mantle xenoliths; these basalts typically erupt from relatively small-volume cinder and/or lava cones and invariably contain an olivine phenocryst population that is geochemically and texturally distinct from the olivine in the entrained mantle xenoliths. These basalts could not have stalled in a crustal reservoir without dropping their relatively dense mantle xenoliths; therefore, their phenocrysts must have grown during ascent. Additional evidence to support this hypothesis comes from phosphorous element maps of olivine phenocrysts; the pattern of P enrichment and depletion is consistent with initial dendritic growth (e.g., Welsch et al., 2014). The preservation of compositional zonation within olivine phenocrysts, despite relatively rapid rates of Mg-Fe diffusive exchange, further supports rapid growth of olivine phenocrysts in erupted basalts. Similar conclusions can be drawn from an examination of small-volume eruptions of basaltic andesite, andesite, dacite and rhyolite. For example, in eruptions of phenocryst-poor (<6%) andesite and dacite from small-volume cones (i.e., not large stratovolcanoes), the sparse plagioclase population often spans a remarkably wide range of composition (30-40 mol% An) and diffusion-limited growth textures (e.g., swallow tails) are not uncommon (e.g., Crabtree and Lange, 2011). Comparison with experimental results confirms that the variable plagioclase composition can only be attributed to changing dissolved water concentrations in the melt during degassing upon ascent. A similar pattern of rapid phenocryst growth during ascent is found in low-SiO₂ rhyolite obsidians (e.g., Waters and Lange, 2013; Waters et al., 2015) as well as large-volume eruptions of high-SiO₂ rhyolite (e.g., Jolles and Lange, 2018). If most erupted magmas grew their phenocrysts during ascent, what does this imply about the eruptibility of magmas that stall in the crust?