Soil weathering and accumulation rates of oxalate-extractable phases derived from alpine chronosequences of up to 1 Ma in age

In this study we compare newly-developed chemical weathering data with previously published data from soils developed along two chronosequences of glacial deposits in the European Alps and the Rocky Mountains (Wind River Range, USA). By combining these chronosequences, we are able to present a comprehensive dataset that represents a time period of > 1 Ma. We describe weathering trends of important elements using a number of weathering indices (e.g., K + Ca/Ti ratio, the weathering ‘index B’ of Kronberg and Nesbitt (1981) and the open mass transport function). Further, we describe the accumulation of Al, Fe, Si and Mn oxyhydroxides (including partially organic phases) as a function of time, and derive the corresponding accumulation rates. We calculated pedogenetically formed oxyhydroxides using an approach based on immobile elements. Our study represents one of only a few studies that describe rates of soil chemical weathering over a period as long as ~ 1 Ma. Results show that rates of chemical weathering clearly decrease along the chronosequences with increasing age of the soils. We find weathering rates are nearly four orders of magnitude lower in the 1 Ma-old soils than in the young soils. Our results suggest that the older soils may be reaching a steady state for these chemical properties in their present environments. A power function best explains the measured time-trends of the ‘index B’ and (K + Ca)/Ti) ratios in the soils. The best time-trend model for pedogenic weakly- to poorly crystalline phases and the relative losses/gains (based on the open-system mass transport function) were obtained with an exponential decay model function. In terms of the soil system, the decreases in the accumulation rate of the oxyhydroxides appears to be influenced not only by the factor of time but by climate as well (increased precipitation at higher altitudes slows the decrease in weathering rate over time). Thus, our ~ 1 Ma chronosequences also become pedogenic gradients since we describe variations in soil properties along biogenic gradients.