Deconvolving Flood Plain Dynamical Processes from Pedogenic Processes on Ancient Floodplains

Thursday, 18 December 2014: 5:30 PM
Nathan D Sheldon, Univ of MI-Earth and Environmental Sciences, Ann Arbor, MI, United States
Paleosols (fossil soils) preserved in ancient floodplain systems represent one of the most widely used and potentially powerful continental paleoclimatic archives. At the same time, to apply most of the quantitative paleoclimate proxies requires the deconvolution of floodplain dynamics from pedogenic processes. For example, a paleosol could be weakly developed because of low atmospheric CO2 levels, low amounts of precipitation, or because of short formation duration due to frequent channel avulsion. The interpretation of local floodplain dynamics in paleo-floodplain systems is often simplistic, assuming both straightforward uniformitarianism and also that a single vertical section represents that lateral diversity of environments, however, these assumptions have rarely, if ever, been put to the test. Herein, a variety of paleoclimatic and paleobiological proxies will be examined in well-preserved paleo-floodplain settings in Spain, Wyoming, and Montana to test those assumptions. Multi-proxy (phytolith, stable isotope) paleovegetation studies along paleo-floodplain transects in Montana (Miocene, Eocene) indicate substantial heterogeneity at the scale of tens to hundreds of meters, floodplain dynamics-driven succession, and cryptic paludal or everwet areas that are not recognizable purely on the basis of sedimentology. Similarly, rapidly aggrading floodplains in fluvial distributary systems (Spain, Miocene) or in dryland basins (Montana) often record significant mismatches between paleosol-based and paleobotanically based estimates of paleoprecipitation, likely due to variable sediment accumulation rates. Both of those sets of results indicate that single vertical sections are unlikely to represent the breadth floodplain environments and properties operating across paleo-floodplain systems. In contrast, newly described mineralogical proxies based on rock magnetics that can be used to reconstruct paleoclimatic/paleoenvironmental properties appear to be robust at the km-scale, and also to reflect the position of the paleo-water table, suggesting they are less subject to floodplain-scale variability in sediment supply or channel dynamics and may provide new tools for linking modern process-based floodplain observations and measurements to paleo-floodplain settings.