Determining the Provenance of Late Paleozoic Loess Using Radiogenic Isotopes

Abstract:

Loess deposits in the sedimentary record of the Late Paleozoic tropics are anomalous because Quaternary loess is commonly glaciogenic, and thus confined to mid-high latitudes. The only loess in the tropics today is associated with mountain glaciation. Given the icehouse conditions of the Late Paleozoic, tropical loess could perhaps reflect sourcing from glaciogenic weathering in the Central Pangaean Mountains. Our goals are to constrain spatial and temporal variations in atmospheric “dustiness” and loess/dust provenance. To test the glacial weathering hypothesis and evaluate loess provenance, we collected preliminary data from two intervals (Moscovian or middle Pennsylvanian) of the Copacabana Formation, Madre de Dios Basin (Bolivia). Two 15-m sections representing carbonate inner platform environments were processed to separate the fine-grained silicate mineral fraction (SMF). Given the proximity to arc volcanism, ash fall could complicate the wind-blown weathering signal; initial work is aimed at assessing the volcanic versus continental inputs to the basin. Visual inspection and initial physical - chemical measurements of the dust-loess size fraction indicates discrete differences between the two sections, as well as between ash-rich and ash-poor intervals. The lower section has intervals with up to 72% suspected ash and SMF (non-ash) of 1-7%. The suspected ash layers in the lower section have higher grain sizes (20-40 µm) and distinct ¹⁴³Nd/¹⁴⁴Nd isotopic compositions (ε_Nd = -3.5, -3.8) compared to the ash-poor intervals (8 µm, ε_Nd = -5.0 to -8.8). The upper section, free of visible ash layers, has high SMF (mean 4-7% up to 41%) with similar size distributions (6-11 µm) and ε_Nd (-7.4 to -7.7) to the lower section, but more radiogenic ⁸⁷Sr/⁸⁶Sr compositions (0.797 versus 0.71 - 0.75). The distinct physical and isotopic characteristics of the ash and SMF in the sections and sub-intervals suggests that deconvolving ash and loess/dust inputs will be possible using isotopic means and that we will be able to assess variability in loess/dust provenance that might ultimately help address the genesis of this voluminous loess.