High Resolution Carbon and Oxygen Isotope Measurements of Laminations in Pedogenic Carbonate

Stable carbon and oxygen isotope ratios in pedogenic carbonate from buried soils provide a proxy for low-resolution Quaternary climate and environmental conditions. Samples of carbonate are taken from clast rinds, nodules or filaments in calcic soils. Most clast rinds exhibit micro-laminations that may preserve isotopic ratios of formation. The techniques typically utilized to sample pedogenic carbonate, however, are too coarse to sample individual laminations and likely result in time averages and therefore limit temporal resolution. We investigated the heterogeneity of both carbon and oxygen isotopes ratios in clast rinds at a 100 μm scale using a rapid CO2 laser extraction technique (Sharp and Cerling, 1996). A single 20 msec burst at low power releases CO2 from polished carbonate slabs and the CO2 is then analyzed using continuous flow GC-IRMS. Analyses take less than 5 minutes with a reproducibility of better than ±0.3‰ (δ13C) and ±0.4‰ (δ18O). We have made a two dimensional map of a thick carbonate rind on a limestone clast from a stage V soil to explore the potential for preservation of isotopic ratios in well developed soils and plan to analyze additional rinds from less well developed calcic horizons for comparison. The isotopic map reveals heterogeneities in δ13C of up to 4‰ at a sub-millimeter scale, possibly corresponding to 30% changes in the fraction of C4 plants. Also imaged are abrupt changes in δ13C of approximately 2‰ across sub-100 μm scale boundaries. One well-defined carbon isotope boundary is sub parallel to, but crosses, the lamination boundaries. Oxygen isotope compositions do not change systematically across the same boundary and generally appear more random. These observations are most easily explained by alteration of initial isotopic compositions. Alteration may preferentially affect oxygen isotope ratios leaving carbon isotope distributions relatively intact. It is also possible that both carbon and oxygen isotopes distributions are mostly primary but that variations in δ18O occur at a much smaller scale than variations in δ13C. The apparently random oxygen isotope zoning could then be explained as the result of sampling multiple homogeneous zones with 100 μm laser pits. Detailed analyses of clast rinds from less well-developed soils may help resolve these questions.