Zoning patterns of mineral, rock, and isotopic compositions in the Hall Canyon pluton in southeastern California are used to determine and discuss the petrologic processes that operate in magma chambers that solidify to form peraluminous plutons. The pluton is biotite-muscovite granodiorite at lower levels and grades to a 150-m-thick upper zone of muscovite granite, capped by a 10- to 20-m-thick zone of pegmatite, aplite, and quartz veins at the roof of the pluton. Mineral and whole-rock compositions are essentially homogeneous in the >600 m of exposed lower zone, whereas the thinner upper zone/roof zone is systematically zoned to more evolved compositions roofward. Compositions of garnet, plagioclase, and muscovite in the upper zone/roof zone track bulk-rock compositions and, in the case of muscovite, range beyond those that have generally been considered magmatic, suggesting that compositional criteria for distinguishing magmatic from secondary muscovite cannot be drawn without consideration of bulk-rock composition.
Initial ϵNd values of −18 to −19 for the pluton indicate that the magma sources were entirely intracrustal and lower Proterozoic in age. The lower zone could represent melt formed by dehydration melting of a biotite-rich metaigneous or metapsammitic source rock. Upper zone/roof zone rocks are too evolved to have been unmodified crustal melts. Even the least evolved was a product of fractional crystallization, and further in situ fractionation of the upper zone/roof zone produced compositional zoning that defines whole-rock Sr and Nd isochron ages of ca. 72 Ma. Slight differences in initial isotopic ratios indicate that lower zone magma and magma parental to the upper zone/roof zone were derived from different source rocks and remained separate during most of the subsequent fractionation of the upper zone/roof zone. Melt separation was inefficient, leaving behind melt-rich mushes that formed rocks that do not bear strong chemical signatures of being cumulate rocks. We suggest that the upper zone/roof zone was static while it fractionated by loss of residual liquid upward, whereas the thicker lower zone may have been capable of convection that stirred the magma as it crystallized, preventing it from becoming zoned. Late in the crystallization of the pluton, meter-scale layering of the granite formed when biotite muscovite granitic magma was injected as sills into the partly crystalline margin of the chamber near the upper zone–lower zone boundary.
We suggest that peraluminous plutons are more heterogeneous isotopically and usually have less regular zoning patterns than metaluminous plutons, because the small effective heights of the magma bodies inhibit convection. This results, in part, from the sill-like form of many peraluminous plutons, but also from melt production rates in zones of purely intracrustal melting that are so low that there is significant cooling of magma in a chamber between arrival of successive melt batches. If resident magma has crystallized to the point of “lock-up,” it cannot mix with new arrivals and does not contribute to the height of magma capable of convection. Systems in which melting is largely the result of advection of heat by mantle-derived basalt might be expected to have larger magma supply rates, because the volume of partial melt produced is not so strictly limited by the abundance of hydrous phases and by the approach of quartz and feldspar to cotectic proportions in the source, and because some of the mafic magma is incorporated into the chamber. Convection would be favored by the resulting greater thickness of magma capable of flow.