Provenance of the Neoproterozoic rocks of the Gifberg Group (Western South Africa)

New petrographic and lithogeochemical data are presented for siliciclastic units of the Gifberg Group (western South Africa), a succession rarely studied in the past. The main detrital source for the oldest succession of the Gifberg Group, the Karoetjes Kop Formation, has been identified as sourced by local metamorphic rocks with a felsic geochemical composition. Nearly unweathered detritus (Chemical index of alteration: 50 to 60; K/Cs >>5000) occurs in the metadiamictite of this formation and is interpreted as palaeovalley infill, deposited only a short distance from its sources. The provenance signature of the overlying Aties Formation indicates significant mafic and/or intermediate sources. Possible sources are rocks of the Bushmanland Group and an unknown unmetamorphosed mafic to intermediate rock succession, which might have been exposed during the deposition of the Aties Formation. The overlying Bloupoort Formation is characterised by a geochemically homogenous metadiamictite (Swartleikrans Bed) with only slightly weathered detritus (K/Cs >10000) and overlying mature silica-rich metarenites and metaconglomerates. It is proposed that the Gifberg basin formed as a small-scale rift basin and then widened through time with subsequent longer sedimentary transport of the detritus which have caused the higher compositional maturity. The predominance of detrital input from local sources in the Karoetjes Kop Formation hinders correlation based on its mineralogical and geochemical content alone. A glacial origin for the Swartleikrans Bed was proposed previously based on the occurrence of possible dropstones. Certain criteria for a glacial depositional environment as outlined by Arnaud and Etienne (2011) were met in our study and include poor sorting, rapid lateral facies changes, crude stratification, variable thicknesses, compositional immaturity, and a gradational contact plus a chaotic texture with large boulders. While these observations point to a ‘rainout’ deposit, there is still a lack of evidence for glacial transport on grain surfaces or in the form of faceted clasts and pebbles.