Origin of the UG2 chromitite layer, Bushveld Complex

Chromitite layers are common in large mafic layered intrusions. A widely accepted hypothesis holds that the chromitites formed as a consequence of injection and mixing of a chemically relatively primitive magma into a chamber occupied by more evolved magma. This forces supersaturation of the mixture in chromite, which upon crystallization accumulates on the magma chamber floor to form a nearly monomineralic layer. To evaluate this and other genetic hypotheses to explain the chromitite layers of the Bushveld Complex, we have conducted a detailed study of the silicate-rich layers immediately above and below the UG2 chromitite and another chromitite layer lower in the stratigraphic section, at the top of the Lower Critical Zone. The UG2 chromitite is well known because it is enriched in the platinum-group elements and extends for nearly the entire 400 km strike length of the eastern and western limbs of the Bushveld Complex. Where we have studied the sequence in the central sector of the eastern Bushveld, the UG2 chromitite is embedded in a massive, 25 m thick plagioclase pyroxenite consisting of 60-70 vol. % granular (cumulus) orthopyroxene with interstitial plagioclase, clinopyroxene, and accessory phases. Throughout the entire pyroxenite layer orthopyroxene exhibits no stratigraphic variations in major or minor elements (Mg-number = 79.3-81.1). However, the 6 m of pyroxenite below the chromitite (footwall pyroxenite) is petrographically distinct from the 17 m of hanging wall pyroxenite. Among the differences are (1) phlogopite, K-feldspar, and quartz are ubiquitous and locally abundant in the footwall pyroxenite but generally absent in the hanging wall pyroxenite, and (2) plagioclase in the footwall pyroxenite is distinctly more sodic and potassic than that in the hanging wall pyroxenite (An(45-60) vs An(70-75)). The Lower Critical Zone chromitite is also hosted by orthopyroxenite, but in this case the rocks above and below the chromitite are texturally and compositionally identical. For the UG2, we interpret the interstitial assemblage of the footwall pyroxenite to represent either interstitial melt that formed in situ by fractional crystallization or chemically evolved melt that infiltrated from below. In either case, the melt was trapped in the footwall pyroxenite because the overlying UG2 chromitite was less permeable. If this interpretation is correct, the footwall and hanging wall pyroxenites were essentially identical when they initially formed. However, all the models of chromitite formation that call on mixing of magmas of different compositions or on other processes that result in changes in the chemical or physical conditions attendant on the magma predict that the rocks immediately above and below the chromitite layers should be different. This leads us to propose that the Bushveld chromitites formed by injection of new batches of magma with a composition similar to the resident magma but carrying a suspended load of chromite crystals. The model is supported by the common observation of phenocrysts, including those of chromite, in lavas and hypabyssal rocks, and by chromite abundances in lavas and peridotite sills associated with the Bushveld Complex indicating that geologically reasonable amounts of magma can account for even the massive, 70 cm thick UG2 chromitite. The model requires some crystallization to have occurred in a deeper chamber, for which there is ample geochemical evidence.