Petrological cannibalism: the chemical and textural consequences of incremental magma body growth

The textures of minerals in volcanic and plutonic rocks testify to a complexity of processes in their formation that is at odds with simple geochemical models of igneous differentiation. Zoning in plagioclase feldspar is a case in point. Very slow diffusion of the major components in plagioclase means that textural evidence for complex magmatic evolution is preserved, almost without modification. Consequently, plagioclase affords considerable insight into the processes by which magmas accumulate in the crust prior to their eventual eruption or solidification. Here, we use the example of the 1980–1986 eruptions of Mount St. Helens to explore the causes of textural complexity in plagioclase and associated trapped melt inclusions. Textures of individual crystals are consistent with multiple heating and cooling events; changes in total pressure (P) or volatile pressure ( $P_{{{text{H}}_{ 2} {text{O}}}}$ P H 2 O ) are less easy to assess from textures alone. We show that by allying textural and chemical analyses of plagioclase and melt inclusions, including volatiles (H₂O, CO₂) and slow-diffusing trace elements (Sr, Ba), to published experimental studies of Mount St. Helens magmas, it is possible to disambiguate the roles of pressure and temperature to reconstruct magmatic evolutionary pathways through temperature–pressure–melt fraction (T– $P_{{{text{H}}_{ 2} {text{O}}}}$ P H 2 O –F) space. Our modeled crystals indicate that (1) crystallization starts at $P_{{{text{H}}_{ 2} {text{O}}}}$ P H 2 O  > 300 MPa, consistent with prior estimates from melt inclusion volatile contents, (2) crystal cores grow at $P_{{{text{H}}_{ 2} {text{O}}}}$ P H 2 O  = 200–280 MPa at F = 0.65–0.7, (3) crystals are transferred to $P_{{{text{H}}_{ 2} {text{O}}}}$ P H 2 O  = 100–130 MPa (often accompanied by 10–20 °C of heating), where they grow albitic rims of varying thicknesses, and (4) the last stage of crystallization occurs after minor heating at $P_{{{text{H}}_{ 2} {text{O}}}}$ P H 2 O  ~ 100 MPa to produce characteristic rim compositions of An₅₀. We hypothesize that modeled $P_{{{text{H}}_{ 2} {text{O}}}}$ P H 2 O decreases in excess of ~50 MPa most likely represent upward transport through the magmatic system. Small variations in modeled $P_{{{text{H}}_{ 2} {text{O}}}}$ P H 2 O , in contrast, can be effected by fluxing the reservoir with CO₂-rich vapors that are either released from deeper in the system or transported with the recharge magma. Temperature fluctuations of 20–40 °C, on the other hand, are an inevitable consequence of incremental, or pulsed, assembly of crustal magma bodies wherein each pulse interacts with ancestral, stored magmas. We venture that this “petrological cannibalism” accounts for much of the plagioclase zoning and textural complexity seen not only at Mount St. Helens but also at arc magmas generally. More broadly we suggest that the magma reservoir below Mount St. Helens is dominated by crystal mush and fed by frequent inputs of hotter, but compositionally similar, magma, coupled with episodes of magma ascent from one storage region to another. This view both accords with other independent constraints on the subvolcanic system at Mount St. Helens and supports an emerging view of many active magmatic systems as dominantly super-solidus, rather than subliquidus, bodies.