The Petrogenetic Association of Carbonatite and Alkaline Magmatism: Constraints from the Spitskop Complex, South Africa.

The 1341 Ma old Spitskop Complex in South Africa is one of a series of intrusions of alkaline affinity, which were emplaced into the central Kaapvaal Craton over the time period 1·4-1·2 Ga. Spitskop contains calcite and dolomite carbonatite closely associated with pyroxenite, ijolite and nepheline syenite, and provides an ideal opportunity to study the petrogenetic relationships between alkaline silicate and carbonatite magmatism. The pyroxenites are not alkalic and are preserved as xenoliths within a plug-like intrusion of ijolite. Nepheline syenites are highly peralkaline, though not agpaiitic, and intrude the ijolites as a series of sheets. These units are cut by a plug of carbonatite composed of an incomplete marginal zone of calcite and dolomite-calcite carbonatite, and a larger central zone of ferroan dolomite carbonatite. Clinopyroxene compositions change systematically from diopside-rich compositions in the pyroxenites to aegirine-augite-hedenbergite in the ijolites to acmite-dominated compositions in the nepheline syenites. Whole-rock chemical data indicate, however, that the nepheline syenites and ijolites are unlikely to be related through fractional crystallization of any reasonable combination of their component minerals (clinopyroxene, nepheline, perthite) from a common parental magma. Low total rare earth element (REE) concentrations and flat to convex-up normalized patterns in the syenites contrast strongly with the steep, light REE (LREE)-enriched patterns in the ijolites. The silicate and carbonatite components differ markedly in their [epsilon]Sr-[epsilon]Nd compositions, the carbonatites having more depleted values ([epsilon]Sr -10 to +10; [epsilon]Nd -1 to -8) than the silicates ([epsilon]Sr 0 to +33; [epsilon]Nd -8 to -13). In addition, the calcite-rich carbonatites have more negative [epsilon]Nd (-6 to -8) than the dolomite carbonatites (-1 to -4). Contrasting isotopic compositions along with the geochemical variations within and between the silicates and carbonatites argue against them being derived from conjugate immiscible liquids. Instead, it is proposed that the carbonatites evolved from primitive carbonate liquids produced directly by low-degree melting of carbonated mantle peridotite. A preliminary model is presented to explain how mantle carbonatite melts can ascend through the mantle and into the crust. It is proposed that the silicate magmatic rocks associated with the carbonatite are produced by melting of enriched mantle lithosphere induced by the influx of deeper-sourced carbonatite melts. [ABSTRACT FROM PUBLISHER]