Lying in wait: deep and shallow evolution of dacite beneath Volca ´n de Santa Marı ´a, Guatemala B. S. SINGER 1 *, B. R. JICHA 1 , J. H. FOURNELLE 1 , B. L. BEARD 1 , C. M. JOHNSON 1 , K. E. SMITH 1 , S. E. GREENE 1 , N. T. KITA 1 , J. W. VALLEY 1 , M. J. SPICUZZA 1 & N. W. ROGERS 2 1 Department of Geoscience, University of Wisconsin-Madison, Madison, USA 2 Department of Earth and Environmental Science, The Open University, Milton Keynes, UK *Corresponding author (e-mail: bsinger@geology.wisc.edu) Abstract: The Plinian eruption in October 1902 of 8.5 km 3 of dacitic pumice and minor basaltic andesite scoria and ash at Volca ´n de Santa Marı ´a, Guatemala violently interrupted a 25 kyr period of repose that had followed ≏75 kyr of cone-growth via extrusion of 8 km 3 of basaltic ande- site lava. Two-oxide and pyroxene thermometry reveal an oxidized (Ni-NiO + 2 log units) and thermally-zoned magma body in which basaltic andesite with 54 wt% SiO 2 at 1020 8C and dacite with 65 wt% SiO 2 at 870 8C coexisted. Plagioclase in dacite pumice and basaltic andesite scoria shows remarkably similar zoning characterized by repeated excursions toward high anorthite and increases in Mg, Fe, and Sr associated with resorption surfaces along which dacitic to rhyolitic melt inclusions are trapped. The melt inclusions increase slightly in K 2 O as SiO 2 increases from 69 to 77 wt%, whereas H 2 O contents between 5.2 and 1.4 wt% drop with increasing K 2 O. These observations suggest that crystallization of the plagioclase, and evolution of a high-silica rhyolitic residual melt, occurred mainly in the conduit as the compositionally- zoned magma body decompressed and degassed from .180 MPa, or .5 km depth, toward the surface. The similarity of plagioclase composition, zoning, and melt inclusion compositions in pumice and scoria suggests that crystals which grew initially in the cooler dacite, were exchanged between dacitic and basaltic andesite magma as the two magmas mingled and partially mixed en route to the surface. Since 1922 . 1 km 3 of dacitic magma similar to the 1902 pumice has erupted effusively to form the Santiaguito dome complex in the 1902 eruption crater. Trace element and Sr–Nd–Pb–O and U–Th isotope data indicate that cone-forming basaltic andesite lavas record processes operating in the deep crust in which wallrock heating sufficient to induce partial melting and assimilation involved several pulses of recharging mantle-derived basalt over at least 50 kyr. A fundamental shift in process coincides with the termination of cone-building at 25 ka: the 1902 dacite reflects .40% fractional crystallization of plagioclase + amphibole + clinopyroxene + magnetite from ≏20 km 3 of basaltic andesite magma left-over following cone-building that cooled slowly without assimilating additional crust. Small contrasts in Sr–Nd–Pb ratios, a modest contrast in d 18 O(WR), and a large difference in the ( 238 U/ 230 Th) activity ratio between the 1902 scoria and dacite indicate that these two magmas are not consan- guineous, rather this basaltic andesite is likely a recent arrival in the system. A glass – whole rock – magnetite– amphibole 238 U– 230 Th isochron of 9.5 + 2.5 ka for a 1972 Santiaguito dacite lava suggests that deeper, occluded portions of the silicic magma body, not erupted in 1902, incubated in the crust for at least 10 kyr prior to the 1902 eruption. Basaltic andesite inclusions in the Santia- guito dacite lava domes are interpreted to be modified remnants of the cone-forming magma par- ental to the 1902 dacite. Supplementary material: Electron probe analyses of glass standards, and SIMS data from stan- dards and melt inclusions for the hydrogen measurements are available at http://www.geolsoc.org. uk/SUP18606 The origin of andesitic to dacitic magma in subduc- tion zones remains a major focus of igneous petrol- ogy owing to the central role it plays in crustal evolution and the hazards posed by explosive, often deadly, eruptions. Many eruptions of dacite also contain lesser amounts of more mafic com- ponents, basalt to basaltic andesite in composition, that beg the question of whether the dacite is genetically related to the mafic magma, or if the ascent of mafic magma into the dacite may have pro- moted eruption (e.g. Sparks et al. 1977; Pallister et al. 1992). Whereas geochemical observations (e.g. Hildreth & Moorbath 1988) and theoretical models (e.g. Dufek & Bergantz 2005; Annen et al. 2006) provide a conceptual framework for under- standing how intermediate magmas result from the From:Go ´mez-Tuena, A., Straub, S. M. & Zellmer, G. F. (eds) 2014. Orogenic Andesites and Crustal Growth. Geological Society, London, Special Publications, 385, 209–234. First published online June 11, 2013, http://dx.doi.org/10.1144/SP385.2 # The Geological Society of London 2014. Publishing disclaimer: www.geolsoc.org.uk/pub_ethics by guest on January 16, 2014 http://sp.lyellcollection.org/ Downloaded from