HOW DO SILICIC MAGMA RESERVOIRS EVOLVE? A “CRYSTAL’S-EYE” VIEW

There is broad agreement that building large plutonic complexes involves multiple magmatic injections, although vigorous debate continues about the scale and timing of these inputs and the degree to which magma bodies are homogenized between eruptions. Volcanic eruptions provide periodic samples of the mobile parts of reservoirs, and crystal records can span 10’s to 100’s of kyr prior to each eruption, providing a window into reservoir processes operating both during quiescent periods and during periods of rejuvenation. We have studied silicic magma systems of a range of sizes, with eruptive volumes from <1 km³ (South Sister) to 10’s or 100’s of km³ (Okataina Volcanic Center, New Zealand; Yellowstone Caldera, WY), in order to examine the commonalities and differences between reservoir processes. Common features include evidence from zircon and feldspar trace-element and isotopic compositions for: 1) diverse magma compositions present within each reservoir system at a given time, 2) inputs of isotopically-juvenile magmas to the reservoirs throughout their history, and 3) consistently younger average ages of feldspar than of co-erupted zircon. These observations suggest a mechanism where erupted liquids carrying zircon are extracted from a crystal network, followed by crystallization of feldspar shortly prior to eruption. In turn, this would predict that plutonic rocks would contain more chemically-diverse major phases than corresponding volcanic rocks, and that the crystal cargo of only the most crystal-rich eruptions (i.e. remobilized mush) would be comparable to that in plutonic records. Another implication is that subsurface magma bodies spend much of their time as a locked mush or largely-solid body. At the same time, detailed study of Yellowstone eruptions shows more diverse compositions in zircon interiors than in zircon surfaces during a given time period, yet both surfaces and interiors show evidence for coherent changes in chemistry over time. This requires at least two regions of magma storage, each of which is evolving over time, and that some homogenization of the liquid is possible even within a crystal mush. Such coherent evolution of liquid compositions is not observed at Okataina, thus more work is necessary in order to determine what conditions are required for efficient homogenization.