Zircon xenocrysts obscured the zircon date for the lower Koras Group, southern Africa

Abstract

The Koras Group is a bimodal volcanosedimentary group located in post-tectonic grabens in a foreland thrust complex in the Kaaien Terrane of the Mesoproterozoic Namaqua-Natal Province of southern Africa. It contains two sequences of mafic and felsic volcanic rocks with an unconformity between them, only the lower sequence being slightly folded. The Koras Group was long regarded as having formed at the end of the 1 210 to 1 000 Ma Namaqua Orogeny, because it lacks the severe deformation and metamorphism of the underlying rocks, with igneous minerals preserved in many samples. Following years of unsuccessful attempts to precisely date the volcanic rocks, the first two ion probe U-Pb zircon studies both reported ages of ~1 172 Ma for the Swartkopsleegte Formation felsic lava in the slightly folded lower sequence (based on relatively few dated zircons) and ~1 100 Ma for the Leeuwdraai Formation rhyolite in the undeformed upper sequence. Thus a major 70 m.y. hiatus seemed apparent between the lower and upper sequences despite their similar geochemistry and rift-related setting. This gave rise to models which envisaged the Kaaien Terrane being unaffected by the syn- to late-tectonic deformation, migmatisation and granite intrusions, documented between 1 200 and 1 150 Ma in the adjoining Namaqua-Natal terranes to the west. A high-pressure (10 kbar) metamorphic event, recognised in the Kaaien Terrane basement just south of hardly-deformed Koras Group exposures and dated at 1 150 Ma, is inconsistent with such models. A re-investigation and microbeam dating campaign on the Koras Group confirms the 1 101 ± 2 Ma (n = 6) age for felsic volcanic rocks of the upper sequence, but establishes a new reliable age of 1 114 ± 4 Ma for the lower one (n = 2). The 1 170 ages obtained in the earlier two studies were revisited and are now considered to reflect the age of zircon xenocrysts from the source rocks, which dominate the zircon population of some Swartkopsleegte Formation samples. Several criteria to distinguish autocrystic (magmatic) from antecrystic (age-overlappping xenocrystic) data points were investigated. One sample had high Th levels in only the younger zircons, but histograms of sufficiently precise 207Pb/206Pb ages provided the main criterion. Calculations of zircon crystallisation temperature intervals were not useful in predicting the abundance or proportions of magmatic and antecryst zircons. A multi-episode model of magmatic generation and crystallisation events is probably appropriate. In cases when felsic volcanic samples yield few zircons, care must be taken to avoid the problem exposed in this study. The Koras Group sediments have similar detrital zircon U/Pb age distributions to those of the Rehoboth Basement Inlier. This supports the concept that the Kaaien Terrane originated as the southern part of the Rehoboth Province.