Search

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.earscirev.2009.03.004 Byline: Nathan D. Sheldon (a), Neil J. Tabor (b) Keywords: paleosols; paleoclimate; paleoenvironments; isotopes; geochemistry; pedogenesis Abstract: Paleosols (fossil soils) are preserved throughout the geologic record in depositional settings ranging from alluvial systems to between basalt flows. Until recently, paleosols were studied using primarily qualitative methods. In recent years, paleopedology has shifted from a largely qualitative field based on comparisons with modern analogues to an increasingly quantitative endeavor. Some of this change has been a result of applying existing techniques to new materials, but many of the innovations have been the result of applying new techniques to new materials, including thermodynamic modeling of soil formation, isotope geochemistry, and applications of empirical relationships derived from modern soils. A variety of semi-quantitative and quantitative tools has been developed to examine past weathering and pedogenesis, and to reconstruct both paleoenvironmental and paleoclimatic conditions at the time that the paleosols formed. Though it is often not possible to achieve the same temporal resolution as with marine records for paleoclimatic reconstructions, proxies based on paleosols are potentially a much more direct means of making paleoclimatic reconstructions because soils form at the Earth's surface, in direct contact with the atmospheric and climatic conditions at the time of their formation. Paleoclimatic and environmental properties that may be reconstructed using the new proxies include provenance, weathering intensity, mean annual precipitation and temperature during pedogenesis, nutrient fluxes into and out of the paleosols, the atmospheric composition of important gases including CO.sub.2 and O.sub.2, the moisture balance during pedogenesis, the soil gas composition, reconstructed vegetative covering, and paleo-altitude. Author Affiliation: (a) Department of Geological Sciences, University of Michigan, 2534 CC Little, 1100 N. University Ave., Ann Arbor, MI 48109, USA (b) Department of Earth Sciences, Southern Methodist University, P.O. Box 750395, Dallas, TX 75275, USA Article History: Received 8 April 2008; Accepted 15 March 2009

Fossil soils are present within mid-Cretaceous fluvial sediments of the Fossil Bluff Group, Alexander Island, Antarctica. The palaeosols contain in situ fossil trees and rooted plants. These palaeosols are typically composed of two dominant horizons: an upper organic-rich horizon with identifiable plant remains, and a lower horizon with well-developed blocky ped structures, clay cutans and mottling. These features are typical of modern soils in seasonally dry climates. Although evidence of flooding is apparent from interbedded fluvial sandstones and mudstones, the palaeosols indicate that this high-latitude mid-Cretaceous environment was predominantly seasonally dry. The presence of the zeolites laumontite, prehnite and pumpellyite indicates a later metamorphic overprint on the palaeosols, resulting from burial diagenesis of this region at depths of about 3.8 5.2 km and temperatures of 150-170[degrees]C. Keywords: Cretaceous, Antarctica, palaeosols, fossil forests, zeolites.

Neogene sediments belonging to the Siwalik Group crop out in the Himalayan foothills along the length of southern Nepal. Carbon and oxygen isotopic analyses of Siwalik paleosols from four long Siwalik sections record major ecological changes over the past [approximately]11 m.y. The carbon isotopic composition of both soil carbonate and organic matter shifts dramatically starting ca. 7.0 Ma, marking the displacement of largely [C.sub.3] vegetation, probably semi-deciduous forest, by [C.sub.4] grasslands. By the beginning of the Pliocene, all the flood plains of major rivers in this region were dominated by monsoonal grasslands. The floral shift away from woody plants is also reflected by the decline and final disappearance of fossil leaves and the decrease in coal logs in the latest Miocene. A similar carbon isotopic shift has been documented in the paleosol and fossil tooth record of Pakistan, and in terrigenons organic matter from the Bengal Fan, showing that the floral shift was probably continentwide. The latest Miocene also witnessed an average change of [approximately]4% in the oxygen isotopic composition of soil carbonate, as observed previously in Pakistan. The reasons for this are unclear; if not diagenetic, a major environmental change is indicated, perhaps related to that driving the carbon isotopic shift. Recently described pollen and leaf fossils from the Surai Khola section show that evergreen forest was gradually displaced by semi-deciduous and dry deciduous forest between 11 and 6 Ma. The gradual nature of this floral shift, which culminated in the rapid expansion of [C.sub.4] grasses starting [approximately]7.0 m.y. ago, is difficult to explain by a decrease in atmospheric pC[O.sub.2] alone (Cerling et al., 1993) but fits well with a gradual onset of monsoonal conditions in the late Miocene in the northern Indian subcontinent. Himalayan uplift, driving both monsoonal intensification and consumption of C[O.sub.2] through weathering, may be the common cause behind major late Miocene environmental change globally. However, the decline of effective moisture associated with monsoon development has probably slowed, not increased, the rate of consumption of C[O.sub.2] by chemical weathering of Himalayan sediments.

The Miocene Kiahera Formation at its type locality on Rusinga Island in the Kenyan part of Lake Victoria consists of a sequence of alluvial sedimentary and pyroclastic strata and calcareous palaeosols. The 80 m thick formation was formed by braidplain aggradation and pyroclastic airfall in response to volcanism from the nearby Kisingiri volcano. Numerous fining-upward sequences approximately 5 m to 15 m thick make up the bulk of the Kiahera Formation. A complete sequence contains sandy and conglomeratic beds at the base which fine upward to sandy siltstones, calcareous palaeosols and airfall tuffs. The fining-upward sequences may have been produced by aggradation during phases of volcanic eruption. During major eruptive episodes, scoriaceous nephelinite detritus flooded the alluvial system to create the lower, coarse-grained beds. As eruptive activity waned, airfall ash and overbank flood deposits accumulated on the landscape forming the upper, fine-grained beds of the sequence. Calcareous soils formed on these highly alkaline ashes between large ash fall events. The calcareous palaeosols are grouped broadly into two main types: tuffaceous clayey Inceptisols with distinct Ck horizons (Chido palaeosols) and sandy Inceptisols consisting of nephelinitic rock fragments and subordinate amounts of tuff (Okoto palaeosols). Both types show only weak to moderate soil development interpreted to be a result of episodic rejuvenation by fresh airfall ash and flood deposits. Comparison of depth of the calcareous subsurface horizon from these palaeosols with those in calcareous modern soils from the Serengeti Plain that also developed on alkaline volcanic ash allow for rainfall estimates of between 550 mm a to the -1 to 750 mm a to the -1 for the Kiahera palaeosols. This together with root traces, soil structure, and major element chemistry of the two palaeosol types are evidence of a landscape with sandy, stream-side woodland soils (Okoto palaeosols) and clayey, floodplain grassy woodland soils (Chido palaeosols).

Interpretations of paleosol records reveal that pre 2.2 Ga paleosols were highly oxidized with minimal or zero iron loss because they were formed under an oxic atmosphere or were positively affected by algal mats which contributed to iron remobilization. Critics, however, dispute such a hypothesis because aluminum is always reduced during iron depletion as suggested by high Ti/Al ratios. Aluminum loss was not demonstrated during the iron-reduced late Archean and early Proterozoic paleosols.