The mafic rocks of the Bushveld complex: a review of emplacement and crystallization history, and mineralization, in the light of recent data

Abstract

Palaeomagnetic, isotopic and chemical evidence confirm that the emplacement of the complex was achieved by multiple injections of magma into the chamber. Gravity, magnetic, resistivity and regional seismic data favour the emplacement of the separate limbs as inclined, wedge-shaped bodies without continuity through the geographic contre of the complex. Pre-Bushveld faulting, and the folding of floor rocks, may have exerted significant control on the distribution, thickness and lithology of the Rustenburg Layered Suite (RLS). Palaeomagnetic data imply that the layering was initially horizontal, at least down to the Curie temperature of magnetite. The concept of an Integration Stage (Wager and Brown, 1968) should be extended to include the last major episode of magma addition, and mixing with liquid residua, that gave rise to the upper part of the Main Zone and the Upper Zone. Superimposed on smaller-scale fluctuations of composition within individual units of cyclically banded cumulates is a broad pattern of mega-cyclicity that is traceable through the lowermost 2000 m of the RLS. Normal fractionation trends are here traceable through intervals 200–400 m thick, with intervening sequences of progressive reversal 50–300 m thick. Such major reversals culminate in olivine-rich cumulates. Isotopic inhomogeneity within individual cyclic units, and even within single samples, as well as distinctive textures indicative of resorption of earlier crystalline phases, support a model of periodic mixing of residual liquid-plus-crystal mushes with batches of fresh, primitive liquid of essentially the same lineage during growth of the Lower and Critical Zones. The Main Zone constitutes a discrete lineage. The northwestern sector is identified as a proximal facies of the Western limb, where its main feeder zone was located. Over a strike distance of ca. 170 km, both Upper and Lower Critical Zone cumulates grade to more feldspathic and chemically evolved sequences of the distal facies. Profiles showing the increase of Ti and Al in orthopyroxenes, with stratigraphic height, are inflected at the level where cumulus plagioclase enters the paragenesis. On this criterion it is also possible to distinguish proximal and distal norites. The close association of PGF with chromitites is emphasized, and progressive increase in the metal ratios (Pt+Pd+Rh)/(Ru+Os+Ir) is correlated with progressive evolution of Cr-spinel compositions with stratigraphic height. The close lithological, chemical and isotopic resemblance between the UG1 and overlying UG2 units poses the problem of explaining why the former is sub-economically mineralized, whereas the latter is a major orebody. Current genetic models for PGE mineralization, based on the control of liquid densities by plagioclase crystallization, may explain Merensky Reef-type orebodies, but are not of general application in exploration.