Shale weathering across a latitudinal climosequence

Open Access
Author:
Dere, Ashlee Laura
Graduate Program:
Geosciences
Degree:
Doctor of Philosophy
Document Type:
Dissertation
Date of Defense:
June 16, 2014
Committee Members:
  • Susan Louise Brantley, Dissertation Advisor
  • Timothy Stapler White, Dissertation Advisor
  • Lee Kump, Committee Member
  • Jason Philip Kaye, Committee Member
Keywords:
  • geochemistry
  • soils
  • Critical Zone
  • erosion
  • mineralogy
  • weathering rates
Abstract:
The thin layer of weathered rock, or regolith, at the Earth’s surface supports most terrestrial life, yet the rates and mechanisms of regolith formation are poorly quantified. Furthermore, the role of climate in controlling regolith formation is unclear, limiting our ability to predict the future availability of soil. To investigate the influence of climate on weathering rates and regolith depth, a transect of study sites across a range of climates in the northern hemisphere was established on a single lithology – iron-rich, organic-poor Silurian shale – as part of the Susquehanna Shale Hills Critical Zone Observatory, Pennsylvania, USA. Sites increase in mean annual temperature (MAT) from Wales to New York, Pennsylvania, Virginia, Tennessee, Alabama and Puerto Rico. Mean annual precipitation (MAP) is high in Wales and Puerto Rico and increases slightly from north to south through the Appalachian Mountain sites. Across these sites, ridgetop regolith depth increases with temperature from north to south (35 – 632 cm). Regolith depth on slopes, however, vary less than ridgetop sites and range from 52–86 cm across the Appalachian sites. Based on meteoric 10Be inventories in the augerable regolith, soil residence times, an approximation of the weathering duration, increase from 10 ky at the northern sites to 120 ky in Puerto Rico but erosion rates remain constant within error across the climosequence (~40 m My-1). Thus, climate appears to have a larger effect on the thickness and residence time of regolith on the ridgetop than on the slope. Using Na as a proxy for plagioclase feldspar weathering, the extent of Na depletion at the soil surface (estimated using the mass transfer coefficient) increases from 20% in Wales and Pennsylvania to 100% Na depletion at the surface in Puerto Rico. These observations are consistent with a transition from kinetically-limited weathering in the north, where weathering rates are limited by mineral dissolution kinetics, to transport-limited weathering in Puerto Rico, where weathering rates are limited by the removal of weathered material. Na loss is the deepest reaction observed in the augerable regolith. Na release during plagioclase dissolution may therefore represent the chemical initiation of regolith formation as fractured bedrock is transformed to finer-grained material. The time-integrated Na release rates increase exponentially with MAT and linearly with MAP: calculating an Arrhenius-type temperature dependence yields an apparent activation energy for feldspar dissolution across the climosequence of 115 ± 25.0 kJ mol-1 (excluding New York). This value is equal to the sum of the activation energy and enthalpy of the feldspar dissolution reaction, as expected for kinetically-limited sites. Presumably, if we had more sites along the climosequence, the Arrhenius plot would show a break in slope related to the decrease in apparent activation energy as transect sites are increasingly transport-limited to the south and the apparent activation energy decreases to equal the enthalpy of feldspar dissolution. Although feldspar may initiate the chemical changes responsible for regolith formation, this mineral constitutes < 12% of the bulk shale. Mg depletion can be used as a proxy for the dissolution of the more abundant chlorite-like minerals in the shale, including vermiculite and hydroxy-interlayered vermiculite (HIV). The extent of Mg depletion at the land surface also increases toward the south but in contrast to Na, Mg is also re-precipitated in secondary clay minerals, maintaining Mg concentrations in the soil. Therefore, although the loss of Mg from regolith along the climosequence can be described with an apparent activation energy of 64.0 ± 14.4 kJ mol-1 (excluding New York), the calculated temperature dependence is not interpreted with respect to an activation energy of dissolution (although the value is within the range of laboratory estimates of chlorite activation energy). New York, the only site underlain by locally-derived shale till, provides a test of the kinetic limitation of regolith formation in the north. At this site, Na release rates are faster than Pennsylvania, a site with similar MAT and MAP. Enhanced Na release in New York shows that the interpretation of kinetic limitation at the northern sites is valid, i.e., an increase in mineral surface area created by glacial grinding causes an increase in weathering rate. Thus, glaciers affect shale weathering rates by increasing mineral surface area and accelerating weathering rates for kinetically limited systems. These observations are consistent with simple reactive transport models for weathering of rock on a convex hillslope. The regolith data can also be considered using a quantitative pedogenic energy model, i.e. the Effective Energy and Mass Transfer (EEMT) model. EEMT combines the effects of precipitation, temperature and biota as a predictor of soil development. According to the model, the biota do not drive the observed differences in regolith depth and weathering. In fact, the variable that is the best predictor of ridgetop regolith depth or weathering extent is MAT. The effect of temperature on the weathering of feldspar and chlorite minerals is observable because variables other than climate have been held constant (e.g. erosion and lithology). Results from this investigation will inform quantitative models of soil formation and hillslope evolution as a function of climate and lithology.