Petrogenetic relationship between Palaeoproterozoic tholeiitic dykes and associated high-Mg noritic dykes, Labrador, Canada

Abstract

Trace-element and SrNd isotopic studies of the Palaeoproterozoic (2235 Ma) Kikkertavak dyke swarm in Labrador provide information on early sub-continental lithosphere development. There are three distinct chemical dyke types: two are tholeiitic, the third of high-Mg norite composition. Petrogenetically, it is not possible to relate any of the chemical groups one to another by simple fractional crystallisation processes; nor can they be derived by contaminating asthenospheric magmas with local crustal components. From their SrNd isotopic characteristics in relation to their rare-earth and multielement patterns, it is clear that significant heterogeneity existed in the source regions of the dyke magmas at the time they were generated, and that the mantle processes responsible can be linked with the phase of crustal growth in the region at ∼3.0 Ga. Modelling suggests that the observed compositions can be accounted for by lower lithosphere-asthenosphere mantle sources which have been variably depleted by previous episodes of basalt magma extraction, and then re-enriched by a lamproitic melt component and possibly by progressive infiltration of melts from a subducted sediment or subduction component. High-Mg noritic dykes are relatively common worldwide in the Palaeoproterozoic (and are often broadly coeval with tholeiitic dykes), but are much less common in younger dyke suites. Most have compositions which require a refractory major-element composition coupled with significant re-enrichment in incompatible lithophile elements. Recent models for cyclic mantle overturns with related deep mantle plumes begin to provide the right thermal and compositional environment beneath immature lithosphere which can generate the dyke-types observed.