MAGMA EMPLACEMENT AS A CRYSTAL MUSH:, TUNNEL DIKE, CHIEF JOSEPH DIKE SWARM

The thick feeder dikes of the Columbia River flood basalts provide substantial insights to magma transport. Continental flood basalts have had major impacts on the Earth, causing global climate fluctuations and mass extinctions. Understanding magma migration from source to surface is an important goal when studying these systems. In this study we focus on the Tunnel Dike, an excellent exposure of the Chief Joseph Dike Swarm. This dike has previously been interpreted to have fed a post main-phase lava flow. Samples were collected at regular spacing across the dike. Concurrent research investigating textures across this dike indicates the presence of numerous large crystals, with complex textural variations, such that these crystals were not grown in situ from local cooling effects. This implies that crystals were carried in as cargo, having crystallized deeper and mobilized to shallower depths. Indeed, plagioclase crystals appear to be aligned in discrete bands separated by plagioclase poor regions, with bent and broken plagioclase grains present, which are consistent with shearing of a dilatant mush. Thin sections were analyzed with an electron microprobe and published whole-rock geochemical data were utilized in MELTS modeling. Chemical maps reveal the formation of anti-rapakivi feldspars, though inconsistently developed. Anti-rapakivi is more pronounced where plagioclase was in contact with the liquid phase, and absent where plagioclase grains are in contact. Given the whole-rock composition, MELTS predicts the anti-rapakivi development at T = 910 ºC, and critical crystallinity at T = 980 ºC, thus providing a constraint on emplacement temperature. Our modeling is consistent with sufficiently low temperatures to allow for a thick crystal mush, though hot enough to be both mobile and to have not yet developed anti-rapakivi texture. We thus interpret these observations to indicate that the dike was emplaced as a mobile crystal mush. While this dike may have fed extensive lava flows of low-crystallinity magma, this mush represents the final magma to be emplaced in the dike, and may well mark a rheological shut-off of the magmatic plumbing system. These findings are in stark contradiction to the general assumption that dikes are emplaced as near-liquidus magma and have widespread implications for thermal modeling in dike systems and magmatic flux rates.