Paul F. Hoffman, Harvard University, Cambridge, MA 02138

Consider a snowball earth if it occurred today. Would the tills and associated sediments deposited along the grounding lines of the African, South American, Indonesian or Australian ice sheets have a high or low index of chemical weathering?

Prior to glaciation, all those continents possessed deeply-weathered (<100 m) soils over large areas developed over tens of millions of years. It would take much time for this heavily-weathered material to be removed by glacial action. It has been hypothesized that the "mid-Pleistocene" (~0.9 Ma) transition represents the time (1.8 Myr after the onset of glaciation) when the relatively thin soils of northern Laurentia were removed by the temperate Laurentide ice sheet (Clark & Pollard, Origin of the middle Pleistocene transition by ice sheet erosion of regolith, Paleoceanography 13, 1-9, 1998). In Antarctica, Plio-Pleistocene glaciomarine strata in the Amery Oasis have anomalous weathering indices attributed to derivation from pre-glacial weathered material (Passchier & Whitehead, Anomalous geochemical provenance and weathering history of Plio-Pleistocene glaciomarine fjord strata, Bardin Bluffs Formation, East Antarctica, Sedimentology 53, 929-942, 2006). The Late Mississippian-Early Permian glaciation of Gondwanaland failed to remove all pre-glacial regoliths from granitic substrates in the Transantarctic Mountains (Isbell et al., Reevaluation of the timing and extent of late Paleozoic glaciation in Gondwana: Role of the Transantarctic Mountains, Geology 31, 977-980, 2003). Until pre-glacial soils have been completely removed, glacial deposits could be weathered or unweathered, depending on their source materials, independent of climatic conditions during glaciation.

Would the progressive removal of pre-glacial regolith result in a steady decline in weathering index over time, unlike the ups and downs observed in the Fiq Formation in Oman (Rieu et al., Climatic cycles during a Neoproterozoic "snowball" glacial epoch, Geology 35, 299-302, 2007)? Maybe not. During the Ghaub glaciation (635 Ma) in Namibia, the Otavi carbonate bank presented a layer-cake stratigraphy in which a single limestone unit sandwiched between dolostone strata supplied the glacial debris. The till deltas deposited on the lower flanks of the bank during the glaciation do not display a simple "inverse" stratigraphy. Instead, till packets form a complex lateral and vertical mosaic of limestone-dominated, dolostone-dominated, and mixed compositions (Hoffman, S. Afr. J. Geol. 108, 557-576, 2005, Fig. 12c; Hoffman and Domack, in preparation). Provenance reversals are common. The ice-dynamic processes responsible are not known. One possibility is lateral shifting of the confluences of glacial tributaries that transport debris derived from different strata. In this scenario, changes in weathering index could result from changing contributions from tributary glaciers transporting weathered and unweathered material.

Changes in weathering index of Neoproterozoic glacigenic strata may relate to processes other than synglacial weathering.

Acknowledgement. Thanks to Frank Corsetti for pointing out Passchier & Whitehead (2006).