The oldest part of the Barberton granitoid-greenstone terrain, South Africa: evidence for crust formation between 3.5 and 3.7 Ga

Abstract

Many stratigraphic and age relationships in the southern part of the Barberton greenstone belt (BGB) remain unresolved due to strong deformation including thrusting, nappe stacking and strike-slip faulting. A relatively undisturbed sequence from the lower Onverwacht Group (∼ 3.48-3.45 Ga), followed by the upper Onverwacht (∼ 3.42-3.3 Ga), the Fig Tree Group (∼ 3.26-3.23 Ga) and the Moodies Group (> 3.22 Ga) was established by single zircon dating using various techniques, but the position of the Theespruit Formation is still uncertain. Using the single zircon evaporation and the vapour digestion techniques we obtained remarkably uniform 207Pb/ 206Pb and UPb ages of 3544 ± 3 to 3547 ± 3 Ma for felsic rocks mapped as Theespruit in the Steynsdorp Anticline of the southeastern BGB, some 100 Ma older than all other dated greenstone units. These rocks were intruded by the 3502–3511 Ma old Steynsdorp TTG pluton containing zircon xenocrysts as old as 3553 ± 4 Ma. A 3.5 Ga granodiorite plug intrusive into metavolcanics of the Komati Formation (lower Onverwacht Group) contained two 3702 ± 2 Ma zircon xenocrysts, the oldest so far measured in the Barberton-Swaziland area and testifying to the presence of very ancient crust in the region. The Theespruit felsic metavolcanics have ϵ Nd(t) values between + 1.1 and − 1.1 and Nd T DM model ages between 3.5 and 3.7 Ga, suggesting variable contamination of their protoliths with older continental crust. The above zircon ages extend the history of the BGB back to ∼ 3.55 Ga. We suggest that the area of the Steynsdorp Anticline constitutes the oldest nucleus of the BGB onto which successively younger units were tectonically and magmatically accreted. We also speculate that the mafic-felsic volcanic units of the southern BGB may perhaps represent distinct oceanic plateaux, rather than ocean floor material, which amalgamated between 3.55 and 3.42 Ga ago. Our data support the concept that the BGB consists of a number of discrete, fault-bounded terranes, and that large-scale lithological correlations are therefore not justified.