Anorthositic Rocks of the Duluth Complex: Examples of Rocks Formed from Plagioclase Crystal Mush

Abstract

Anorthositic rocks compose 35–40% of the Middle Proterozoic (Keweenawan; 1·1 Ga) Duluth Complex—a large, composite mafic body in northeastern Minnesota that was intruded beneath a comagmatic volcanic edifice during the formation of the Midcontinent rift system. Anorthositic rocks, of which six general lithologic types occur in one area of the complex, are common in an early series of intrusions. They are characterized on a local scale (meters to kilometers) by nonstratiform distribution of rock types, variably oriented plagioclase lamination, and composite intrusive relationships. Variably zoned, subhedral plagioclase of nearly constant average An (∼60) makes up 82–98% of the anorthositic rocks. Other phases include granular to poikilitic olivine (Fo66–38), poikilitic clinopyrox-ene (En′73–37), subpoikilitic Fe-Ti oxides, and various late-stage and secondary minerals. Whole-rock compositions of anorthositic rocks are modelled by mass balance to consist of three components: cumulus plagioclase (70–95 wt.%), minor cumulus olivine (0–5%), and a gabbroic postcumulus assemblage (5–27%) representing a trapped liquid. The postcumulus assemblage has textural and compositional characteristics which are consistent with crystallization from basaltic magma ranging from moderately evolved olivine tholeiite to highly evolved tholeiite (mg=60–25). Sympathetic variations of mg in plagioclase and in mafic minerals suggest that cumulus plagioclase, though constant in An, was in approximate equilibrium with the variety of basaltic magma compositions which produced the postcumulus assemblages. Standard models of mafic cumulate formation by fractional crystallization of basaltic magmas in Duluth Complex chambers, although able to explain the petrogenesis of younger stratiform troctolitic to gabbroic intrusions, are inadequate to account for the field, petrographic, and geochemical characteristics of the anorthositic rocks. Rather, we suggest an origin by multiple intrusions of plagioclase crystal mushes—basaltic magmas charged with as much as 60% intratelluric plagioclase. The high concentrations of cumulus plagioclase (70–95%) estimated to compose the anorthositic rocks may reflect expulsion of some of the transporting magma during emplacement or early postcumulus crystallization of only plagioclase from evolved hyperfeldspathic magma. Although the evolved compositions of anorthositic rocks require significant fractionation of mafic minerals, geophysical evidence indicates that ultramafic rocks are, as exposure implies, rare in the Duluth Complex and implies that plagioclase crystal mushes were derived from deeper staging chambers. This is consistent with interpretations of olivine habit and plagioclase zoning. Moreover, plagioclase could have been segregated from coprecipitating mafic phases in such lower crustal chambers because of the buoyancy of plagioclase in basaltic magmas at high pressure. The geochemical effects of plagioclase suspension in basaltic magmas are consistent with observed compositions of cumulus plagioclase in the anorthositic rocks and with the geochemical characteristics of many comagmatic basalts. The petrogenesis of the anorthositic rocks and the overall evolution of Keweenawan magmas can be related to the dynamics of intracontinental rift formation.