A NEW PEDOTRANSFER FUNCTION FOR ESTIMATING SOIL PH IN PALEOSOLS

Soil pH is a master variable that governs the types and abundance of cations in solution and, ultimately, nutrient availability. Because it is a solution property normally measured using a soil paste it is impossible to directly measure pH in lithified paleosols. A new pedotransfer function for estimating soil pH in paleosols is presented based on bulk geochemistry. A large and diverse range of soil B horizons (n = 619) were used to perform regression analysis on pH versus a variety of molecular concentrations and ratios. The molecular ratio of aluminum to aluminum plus calcium (AlCa) is found to hold the highest capacity to predict pH (R² = 0.67, SE = 0.84). An independent testing dataset (n = 304) of soils not included in the training dataset was compiled from published literature and used to successfully cross-validate pH predictions using AlCa. Two deep-time paleosol case studies are used to justify the utility of AlCa. Short-term drying events recorded in a succession of paleosols across the Paleocene-Eocene Thermal Maximum (PETM) in the Bighorn Basin, WY, are accompanied by resolvable shifts from acidic (5.5-6) to slightly alkaline (~7.5) pH in paleosol B horizons. Such changes are not revealed using a previously published Vertisol-specific pH pedotransfer function. The Baltic Paleosol—a Neoproterozoic (560-600 Ma) oxisolic weathering surface formed on Baltica basement—preserves a vertical acid to moderately alkaline pH gradient previously reconstructed using aluminum-phosphate-sulfate (APS) mineral systematics. The AlCa pedotransfer function predicts pH values of ~5.5 in the kaolinitic B horizon of the Baltic Paleosol, which closely fits pH values predicted by the APS system. Taken together, these case studies demonstrate that AlCa is useful for reconstructing soil pH within individual paleosol profiles, even those formed in primitive Precambrian terrestrial ecosystems, and across critical paleoclimatic transitions.