Underclay paleosols are common throughout Pennsylvanian successions in the Appalachian basin region, but there is great disagreement concerning their genesis. Through careful examination of the morphology, micromorphology and whole-rock geochemistry of a suite of underclay paleosols from the Lower Pennsylvanian Crooked Fork Group in eastern Tennessee, we demonstrate a three stage pedogenic development that includes an early well-drained stage followed by saturated soil conditions associated with rising sea-level, and concluding with burial diagenesis. The underclays contain features common to well-drained paleoVertisols and paleoAlfisols, as well as poorly drained paleoHistosols. Macro- and micro-morphologic features such as pedogenic slickensides, angular blocky ped structure, illuviated clay pore-fillings and sepic-plasmic micro-fabrics, indicate a well-drained period early in paleosol development. Gley overprinting of vertic features, including drab, low-chroma paleosol colors, siderite-replaced rhizocretions, sphaerosiderite and pyrite nodules, extensive leaching and translocation of alkali and alkaline earth elements, and kaolinitization of smectites and hydroxy-interlayer vermiculite (HIV), indicate a second stage of paleosol development during which the ancient soil was saturated and experienced moderate to low Eh and low pH conditions. Finally, burial and coalification of organic material and the stratigraphic juxtaposition of marine shales and sandstones on top of terrestrial deposits indicate a third stage of underclay development. This three-stage model of underclay genesis was driven by high-order glacio-eustatic sea-level fluctuations, as evidenced by stratigraphic relationships within the Crooked Fork Group.

Paleosols can be used as proxies for determining paleoclimate. Previous studies proposed a shift in climate in North America at the Mississippian-Pennsylvanian boundary, from semi-arid to tropical or sub-tropical rainy. Using whole-rock geochemical data an estimate of 1300 mm/yr of rainfall was obtained using the chemical index of alteration without potash (CIA-K) as a proxy for paleo-mean annual precipitation (MAP). This value is consistent with modern sub-tropical humid environments and supports the hypothesis for climate change at the end of the Mississippian.