The tectonic evolution of the Belingwe Greenstone Belt in Zimbabwe has been a matter of great controversy. In the most recent contribution to this debate, Bolhar et al. (2003) reported major element, trace element, and Nd isotope data for mafic volcanic rocks of different stratigraphic units of the greenstone belt. Three stratigraphic units of the ca. 2.9 Ga Lower Greenstones and two units of the ca. 2.7 Ga Upper Greenstones were sampled. Two groups of geochemically distinct volcanic rocks were observed in each stratigraphic unit, a group of unfractionated rocks and a group of light rare earth element–enriched rocks with negative anomalies for Nb, Ta, Ti, and P, and low ϵNd values. The geochemical signatures of the second group of rocks have been attributed to processes of assimilation of continental crust, followed by fractional crystallization.
These observations are important because of the ongoing debate about whether mafic volcanic rock units in greenstone belts, and the Belingwe belt in particular, represent ophiolite-like sequences that originated in oceanic environments and were obducted onto continental crust or whether they represent continental flood basalt–like, autochthonous sequences that extruded onto continental crust. Based on the geochemical evidence, Bolhar et al. favored an ensialic origin for the Belingwe volcanic rocks and an autochthonous origin for the Upper Greenstones.
If we accept that the geochemical signatures are indeed a result of contamination by continental crust rather than related to other processes, such as the presence of an enriched mantle source, source contamination, or subduction, we think that an ensialic and autochthonous origin, i.e., deposition on the gneissic basement now surrounding the greenstone belt, cannot be inferred from the geochemical evidence presented.
1. Contacts between basement rocks and mafic volcanic sequences in greenstone belts are sheared. In the Belingwe belt, the basal contact of the Upper Greenstone volcanic sequence is a major thrust fault (Kusky and Winsky, 1995; Hofmann et al., 2003a), negating an autochthonous or paraautochthonous origin. All contacts between stratigraphic units of the Lower Greenstones are sheared.
2. Mafic volcanic greenstone sequences formed in submarine settings well below wave-base, and this is also the case for the Upper Greenstone volcanics of the Belingwe belt, which contain pillow basalts and intercalations of turbidites (Hofmann et al., 2003b). Time-equivalent continental flood basalts (Ventersdorp Group, South Africa; Nelson et al., 1992), and continental flood basalts in general, are subaerial lava flows. In addition, Archean continental flood basalts, exemplified by the Ventersdorp and Dominion Groups in South Africa and the Fortescue Group in Australia, although geochemically very similar to the enriched Belingwe volcanic rocks, completely lack unfractionated basalts. The geochemical signatures are attributed to melting of an enriched mantle source (subcontinental lithospheric mantle) rather than crustal contamination processes (references in Nelson et al., 1992).
3. Many volcanic greenstone sequences show evidence that they formed in a near-continental setting, as exemplified, for example, by gneiss bodies engulfed in mafic lava or quartzose sediments overlying pillow basalt in the Mafic Formation of the Midlands Greenstone Belt, Zimbabwe (Dirks et al., 2002). However, as in many modern tectonic settings, no clear-cut distinction can be made between a continental or oceanic origin, because modern-day oceans develop by thinning and rifting of continental crust. As such, contamination of basalts produced in infant oceans and backarc basins by continental crust can be expected. Backarc basins flanked or underlain by thinned continental crust have commonly been cited as analogues for Archean greenstone belts.
4. Bolhar et al. state that their data are inconsistent with an oceanic plateau and mid-ocean-ridge setting. However, some oceanic plateau basalts show evidence for crustal contamination (Kerguelen Plateau; Ingle et al., 2002). In addition, many ophiolites show geochemical signatures different from mid-ocean-ridge basalts; these suprasubduction zone geochemical signatures are in contrast to the lack of structural-stratigraphic evidence for subduction-related activity (Moores et al., 2000), a similar case of controversy between geochemists and field-oriented researchers.
5. An important observation by Bolhar et al. is that the volcanic rocks of each stratigraphic unit show remarkably similar geochemical signatures, so that essentially identical petrogenetic processes are implied. This is in contrast to the lithological attributes of individual stratigraphic units. For example, the 2.90 Ga Hokonui Formation of the Lower Greenstones is lithologically similar to an island-arc sequence, because it is characterized by submarine to possibly subaerial andesitic to dacitic lavas, pyroclastic and epiclastic rocks, and only minor amounts of mafic volcanic rocks. The 2.69 Ga Zeederbergs Formation is a submarine lava plain sequence of basalts and basaltic andesites, similar to the extrusive sequence of modern oceanic crust or oceanic plateaus. It thus appears that trace element geochemistry is unable to distinguish between different extrusive settings in the Belingwe belt, or that geochemically similar magmas extruded in different settings, the latter indicating a relatively strong mantle imprint on magma chemistry.