CONSTRAINING THE SOURCE OF PROTEROZOIC ANORTHOSITES

Identifying the ultimate source of Proterozoic anorthosites – mantle vs. crust – is the primary goal of all petrogenetic studies of these enigmatic bodies. The mineralogy of anorthosite (plag±olivine±opx±cpx) is consistent with that crystallized from mantle-derived basaltic melts. Early key experiments demonstrated that An_40-60 plagioclase is not stable at pressures above about 1.5 GPa and that with increasing pressure the boundary curve in the system Di-An-Ab shifts towards the An-Ab join, permitting enhanced plagioclase production from basaltic melts at depth (Lindsley, 1968). Recent experimental work on opx-normative gabbroic/dioritic rocks from Harp Lake and Rogaland appears to show that some proposed anorthosite parental liquids lie across the trace of the plag+2-px cotectic from 1-1.3 GPa and that they straddle the thermal divide on the plag+px liquidus surface, thus apparently requiring a mafic source region (i.e. lower continental crust) (Longhi, 1999). It is however unclear that small amounts of dry partial melting of lower crustal granulite will produce melt compositions that are strongly plag-saturated nor will it yield the large quantities of melt (and corresponding cumulates) required by mass balance constraints. More importantly, noritic-gabbronoritic lower crust is opx-normative and cannot be responsible for producing the olivine-bearing anorthosites or troctolites typical of the largest Proterozoic anorthosites. Radiogenic isotopic studies (Pb, Nd, Sr, Os) are particularly useful for constraining crustal input to anorthosite and have successfully traced out different-aged crustal reservoirs beneath them. These studies reveal that while depleted asthenospheric mantle may be a major source component for some anorthosites, others contain a contribution from chemically “enriched” continental lithospheric mantle that partially melts at relatively shallow depths beneath extended crust. Finally, it is likely that no magma associated with Proterozoic anorthosites escapes contamination during ascent through the lower crust, which may act as a highly effective near-solidus “reactive filter” capable of stabilizing plagioclase as a liquidus phase for the duration of these long-lived (tens of millions of years), low magma flux, magmatic systems.