Trace elements in zircon record changing magmatic processes and the multi-stage build-up of Archean proto-continental crust

Abstract

Zircon trace element geochemistry has become an increasingly popular tool to track crustal evolution through time. This has been especially important in early-Earth settings where most of the crust has been lost, but in some fortuitous instances detrital zircons derived from that lost crust have been preserved in younger sediments. To study the formation and geochemical evolution of continental crust from the Hadean to the Paleoarchean, the 3.6 to 3.2 Ga Barberton Greenstone Belt in southern Africa is an excellent target due to its outstanding preservation and presence of detrital zircons that span almost a billion years. Here, we use trace elements, in combination with hafnium and oxygen isotopes, of 3.65 to 3.22 Ga detrital and tuffaceous zircons of the Moodies and Fig Tree groups and compare their geochemistry to previously studied 4.2 to 3.3 Ga detrital zircons from the Green Sandstone Bed of the Onverwacht Group. The major detrital zircon age clusters in the former at 3.55 Ga, 3.46 Ga, and 3.26–3.23 Ga overlap with episodes of TTG emplacement and felsic volcanism in the Barberton area, suggesting a local provenance. In contrast, age clusters at 3.65 Ga and 3.29 Ga of the Moodies and Fig Tree groups as well as 4.2 to 3.3 Ga detrital zircons from the Green Sandstone Bed do not have known intrusive sources and were likely derived from outside the present-day Barberton belt. This indicates that more than half of the felsic igneous events in the detrital zircon record do not have a whole-rock representation that can be directly studied. The similar compositions and inferred crustal evolution histories recorded in zircons from the Fig Tree and Moodies groups, as well as from the Green Sandstone Bed, suggest that they were derived from connected terranes experiencing similar crustal processes diachronously. Together, they show three phases of felsic continent formation, reflecting different crustal processes: (1) long-lived protocrust formed in the Hadean from undepleted mantle sources. These zircons are vastly different from younger zircons and, hence, Barberton TTGs are not good analogues of Hadean crust formation. (2) At 3.8 Ga, onset of significant crustal growth though cyclic juvenile additions and hydrous melting, possibly within a volcanic plateau setting but an arc-like setting cannot be excluded based on this data. (3) Between 3.4 and 3.3 Ga, felsic crust is generated through a previously unrecognized episode of crustal growth by shallow melting of mafic, mantle-derived sources. This is immediately followed by the onset of crustal thickening through the transport of surface-altered, hydrated materials to deep crustal levels. Since there is geological evidence for extension and shortening at that time this may reflect the onset of horizontal movement. Whether this last geodynamic setting reflects modern-style plate tectonics or not, continent formation and the onset of plate tectonics in the Barberton area occurred through complex multi-stage processes spanning almost a billion years, most of which is only accessible through the detrital zircon record.