Zircon crystallization and recycling in the magma chamber of the rhyolitic Kos Plateau Tuff (Aegean arc)

INTRODUCTION Approximately 160,000 yr ago, one of the largest Quaternary explosive volcanic eruptions of the Mediterranean basin occurred just offshore of the island of Kos (Greece), resulting in ~60 km of volcanic ash and pumice emplaced as pyroclastic fall and fl ow deposits over an area >2000 km (Keller, 1969; Allen, 2001; Fig. 1). In comparison to the 3500 yr B.P. Minoan eruption, which created a 12-km-diameter caldera on the nearby island of Santorini, the Kos Plateau Tuff is a factor of two larger in volume (Pyle, 1997). These geologically nearly instantaneous events have devastating effects on human infrastructures and populations. Understanding how (and how fast) these highly explosive magmas bodies are generated is of obvious importance. In the last decades, the vast majority of petrogenetic studies focusing on these evolved (non-basaltic) magma bodies has stressed the complexities that are involved in the generation of such magmas. Although crystal fractionation typically plays a role, open-system behavior, with an important mass contribution from preexisting crust, is a well-established process in generating large-volume (>1 km) silicic magmas, particularly in continental arcs (e.g., DePaolo, 1981; Halliday et al., 1984; Hildreth and Moorbath, 1988; Reagan et al., 2003). Earlier work on the Kos Plateau Tuff (Keller, 1969) has suggested that the magma was derived by partial remelting of a pre–Kos Plateau Tuff granitic intrusion in the upper crust. Since such a hypothesis poses mass and thermal balance problems (e.g., Bachmann et al., 2002) and has rarely been suggested for other units, it should be tested further. In contrast to these previous interpretations, our new fi ndings argue that the Kos Plateau Tuff may represent a case of dominantly closed-system fractionation from mafi c parents. This chemically closed-system evolution can be tied to the geodynamics of the Aegean arc, and may shed light on the long-standing issue of whether silicic magmas remain for several hundreds of thousands of years in the upper crust above their solidii or whether they rapidly crystallize to solid plutons (e.g., Halliday et al., 1989; Mahood, 1990; Sparks et al., 1990; Reid et al., 1997; Vazquez and Reid, 2004; Bacon and Lowenstern, 2005; Charlier et al., 2005).