Quantification of chemical weathering rates across an actively eroding hillslope

Chemical weathering rates for landscapes are difficult to quantify because the timescales over which weathering occurs are often unknown. in this study, we use timescales defined by prior cosmogenic nuclide analyses and a suite of geochemical measurements to calculate weathering rates in saprolite and soil and to determine how these rates vary across an eroding hillslope. We also estimate the relative contributions of solute and erosional mass loss to landscape lowering. Analyses were conducted oil a soil-mantled hillslope developed on a granodiorite pluton in southern NSW, Australia. Mass loss in solution accounts for 35% to 55% of the total mass loss from the hillslope. Saprolite close to the soil-saprolite boundary is less weathered at the ridge than at distance from the ridge. The calculated flux of silica from the saprolite is 5 tons km(-2) yr(-1) and does not vary with distance from the ridge or overlying soil thickness. The assumption of steady state, but spatially variable erosion rates with a temporally constant thickness of soil and saprolite allows us to calculate the downslope variation in the rate of solute losses. Soil silica weathering rates initially increase from similar to 12 tons km(-2) yr(-1) at the ridge crest to similar to 20 tons km(-2) yr(-1) on the convex region of the hillslope, and then decrease to 13 tons km(-2) yr(-1) by the time soils have been transported 45-50 m. Soils closer to the surface are more weathered than deeper soils. This geochemical stratification suggests that the soil column does not become thoroughly vertically mixed as soils move down slope. Dependence of solute loss on transport distance may explain the previously noted discrepancy between the observed soil thickness and that predicted based on curvature. Variation of solute loss with transport distance may also contribute to the linear dependence of soil thickness on distance from the ridge. (c) 2005 Elsevier B.V. All rights reserved.