Present-day elemental and mineral weathering rates based on solute fluxes are compared quantitatively to past long-term rates determined from solid-state elemental fractionation in a saprolitic granite regolith at Panola, Georgia, USA. Saturated fluid flow across a low-permeability kaolin duripan controls the rate of steady-state unsaturated flow in the underlying saprolite. Water and Cl mass balances and experimental conductivities produce a minimum fluid flux density of 8x10^-2 m yr^-1 and a fluid residence time of 12 years. Solute Si flux, based on pore water concentrations and infiltration rates, is 27 mmoles yr^-1, compared to a long-term flux rate of 17 mmoles yr^-1, based on regolith Si loss and reported ³⁶Cl dating of the regolith surface. Similarities in short- and long-term fluxes imply that parameters influencing silicate weathering, including precipitation, temperature, and vegetative cover, while not necessarily constant, have not significantly impacted Si leaching rates during the last several hundred thousand years.

Linear decreases in solid-state Mg with decreasing regolith depth permit the calculation of the long-term biotite weathering rate under isovolumetric steady-state weathering conditions. A rate constant of 3x10^-17 moles m^-2 s^-1 is up to 5 orders of magnitude slower than that reported for experimental dissolution of biotite, implying very different reaction kinetics during natural weathering. Short-term biotite weathering fails to produce expected increases in solute Mg and K concentrations with increasing depth and fluid residence times in the regolith. This discrepancy indicates that ion exchange disequilibrium and open-system biologic uptake in an aggrading forest ecosystem are of sufficient magnitudes to overwhelm solute fluxes resulting from biotite weathering.