Correlations on the southern Kaapvaal Craton Margin, 2: Ventersdorp strata at T’Kuip date the end of Klipriviersberg Volcanism

Abstract The basal sedimentary unit, which overlies 2 905 Ma basement granite at T’Kuip, in the southernmost exposures of the Ventersdorp Supergroup, is part of the Kameeldoorns Formation of the Platberg Group. The T’Kuip exposures contain mafic tuff interlayered with conglomerate and wacke. These are stratigraphically similar to the Kameeldoorns Formation section previously described in borehole KFN1 from the Free State goldfields and at Taung (local Mohle Fm.), where an interbedded tuff, containing accretionary lapilli, was dated at 2 733 ± 3 Ma. At all three of these localities, the Kameeldoorns Formation is overlain by 2 720 ± 2 Ma Makwassie Formation rhyolite porphyries. The tuff samples from T’Kuip and borehole KFN1 plot in the Klipriviersberg field on Ti-Zr-P and other geochemical diagrams, defined by literature analyses of Ventersdorp Supergroup volcanic rocks. They have similar spidergrams to the Klipriviersberg Group lavas, but generally higher heavy rare earth element (REE) levels, related to crustal contamination and fractionation. Accretionary lapilli in the tuff samples, formed in fluidised gravity flows, show that their andesitic compositions were attained before explosive extrusion. The tuffs in the Kameeldoorns Formation are considered to represent the waning phase of Klipriviersberg volcanism. This interpretation is strengthened by the presence of Klipriviersberg-type lavas interbedded with sediments in the lower Kameeldoorns Formation in borehole KFN1. The tuffs from T’Kuip contain three young, concordant zircon grains, with insignificant common lead, which have a combined age of 2 728 ± 7 Ma (16 points in two samples). Including another 14 discordant grains, corrected for significant common lead, confirms this age group, with discordia intercepts 2 729 ± 6 and 121 ± 71 Ma. This date is considered to represent the final phase of Klipriviersberg-type volcanism. Older zircon xenocrysts have a 2 850 to 2 925 Ma main group and a few ages up to 3 100 Ma, inherited from the Kaapvaal Craton basement. The previous ion probe U-Pb zircon date on Klipriviersberg lavas (Orkney Formation from a quarry) yielded 2 714 ± 16 or 2 691 ± 14 Ma 207Pb/206Pb ages, using the 208Pb or 204Pb common lead correction method respectively. However, the five low-common lead points yield a concordia age of 2 711 ± 22 Ma, independent of the problematic common lead correction. This agrees with our 2 728 ± 7 Ma date within error. Klipriviersberg volcanism probably only lasted 5 Myr, as is the case for most large igneous provinces (LIPs). Our work shows that Klipriviersberg volcanism was followed within 0 to 17 Myr by the extensive 2 720 ± 2 Ma Goedgenoeg and Makwassie formation bimodal volcanism with geochemistry clearly distinct from that of the Klipriviersberg Group. Klipriviersberg volcanism began after the 2 799 ± 9 Ma age of Ventersdorp Contact Reef youngest zircons but is probably significantly younger. Two possible Klipriviersberg correlates are the ~2 781 Ma Derdepoort Formation mafic lavas and ~2 788 Ma dykes in Witwatersrand strata, which both plot in the Klipriviersberg geochemical field and could represent the beginning of Klipriviersberg volcanism. However, any large-volume mantle basaltic melt should correspond geochemically to the Klipriviersberg Ti-Zr-P field, as do several younger dykes. Geochemistry needs to be combined with stratigraphic context to attain confident correlation, as is the case with the tuffs from T’Kuip. Precise ages on ultramafic lavas from the base of the Klipriviersberg Group are still lacking, due to their low-grade metamorphism and zircon-poor modes. Palaeomagnetic inferences of a ~2 780 Ma age for the Klipriviersberg Group are considered to be ambiguous. The presently available data is compatible with Klipriviersberg magmatism either having had a geologically short duration between 2 730 and 2 720 Ma, or a much longer ~70 Ma duration from ~2 780 to 2 720 Ma. Ventersdorp magmatism gave rise to one or more LIPs on the Kaapvaal Craton, possibly during the assembly of the Superia supercontinent. The LIP printing geochemical model indicates that Klipriviersberg volcanism began with a plume or enriched mantle source and progressed to a subduction zone-modified lithospheric mantle (SZLM) source. The overlying Goedgenoeg Formation samples indicate an initial mid-ocean ridge basalt (MORB)-like depleted mantle source. However, granitoids and lavas ranging from 2.9 to 2.05 Ga also have similar LIP print pattern spidergrams. This suggests the presence of a long-lived subduction-enriched mantle source beneath the Kaapvaal Craton, reactivated by short-lived plume or subduction events.