|
The relative mobility of Fe in Paleoproterozoic paleosols has been used to constrain the timing of proposed increases in the pO2 level, where net Fe losses are interpreted to reflect Fe mobilization during weathering under an anoxic atmosphere, and Fe retention is interpreted to reflect oxidation "in place" during weathering under an O2-bearing atmosphere (e.g., Rye and Holland, 1998; [Amer.J.Sci.]). The 2.2 Ga paleosols that developed on the Hekpoort Basalt of the Pretoria Group, South Africa, are particularly important because they have been considered to be one of the youngest Fe-depleted paleosols, potentially placing constraints on the pO2 level and the timing of the inferred Great Oxidation Event (GOE). We investigate Fe isotope variations in the complete section of lateritic weathering profile developed in the Hekpoort paleosol from Botswana that were studied by Beukes et al. (2002) and Yang and Holland (2003) [Amer.J.Sci.]. Beukes et al. (2002) [Geology] suggested that the Hekpoort laterite profiles were developed through processes that were similar to those that occur in modern laterites. Isotopic mass-balance constraints provide a sensitive means for evaluating the extent of Fe mobilization and open- or closed-system behavior. In addition, the sensitivity of Fe isotope fractionation factors to redox state may constrain the mechanisms and oxidation state of Fe mobilization in paleosols, which ultimately may place constraints on ancient atmospheric O2 levels. Our results are not compatible with a model of simple internal redistribution of Fe within the paleosol column as claimed Yang and Holland (2003), but are well explained through fluid-rock interaction involving transport of aqueous Fe2+ through the system during paleosol evolution. We demonstrate that Fe isotopes can provide important constraints on open- vs. closed-system behavior of Fe in ancient weathering profiles, a distinction that is critical when calculating Fe mobility as a means for inferring surface redox conditions in the past (Yamaguchi et al., 2007; EPSL).
|