KOMATIITIC TUFFS OF THE 3.5-3.2 GA ONVERWACHT GROUP OF THE SOUTHERN BARBERTON GREENSTONE BELT, SOUTH AFRICA

Komatiitic rocks are formed by the eruption and solidification of very low viscosity lavas (0.1-10 Pa s). Ultramafic pyroclastic debris is rare and most recorded komatiitic tuffs are thin, spotty accumulations of ash or locally derived breccias. The southern domain of the 3.5-3.2 Ga Onverwacht Group, however, contains at least six laterally extensive tuff beds with petrological and geochemical signatures consistent with reported komatiitic lavas. These range from a few centimeters to over 15 m thick, are nearly aphyric, and only rarely include chromites or silica pseudomorphs of olivine phenocrysts. Unlike associated dacitic ashes, these lack evidence of magmatic quartz or feldspar. Bulk compositions are dominated by SiO₂ and Al₂O₃, a consequence of pervasive early silicification. However, immobile element ratios show that the komatiitic ashes have higher Ti/Zr (60-117) ratios than dacitic ashes (Ti/Zr = 13-51) and are more like local komatiitic lavas (Ti/Zr = 60-133). Like many komatiites, the best preserved ashes also have nearly chondritic rare earth element abundances and ratios.

These ash layers are regionally widespread, some traceable over 10 km. Volcanic vents are unknown but the absence of breccia, scarcity of phenocrysts, fine grain size, and homogeneity in bed thicknesses imply these were distal ash deposits. This suggests these were not restricted phreatic or littoral explosions but large eruptions that generated plumes which dispersed ash over a broad area. The low vesicularity of the glass shards and the lack of hydrous mineral phases also argue against these ashes as a product of explosive exsolution of magmatic volatiles. Rather, these ashes were likely a result of intense fragmentation during the interaction of magma with external water.

The common association of komatiitc tuff beds with sediments deposited during local volcanic quiescence and the observation that every ash bed conformably overlies a komatiitic flow sequence suggest that local komatiitic episodes were terminated by phreatomagmatism. However, Al₂O₃/TiO₂ ratios in the ash and lavas are different implying that the pyroclastic debris may not be petrogenetically related to the flows. Instead, the ashes may represent sources unrelated to those that produced the Barberton komatiitic lavas.