Watershed- and plot-scale tests of the mobile anion concept

Anion fluxes from a forest soil are usually correlated with those of base cations (BC). Declines in base cation deposition or long-term depletion from the soil may change these relationships. We used multiple regression to identify biogeochemical variables predicting annual volume-weighted concentrations of BC in streamwater draining a forested watershed, and analysis of variance to compare the effects of Ca and Cl inputs on BC fluxes out of soil horizons in irrigated plots. For the watershed, anion concentrations in streamwater predicted BC export most precisely (R² = 0.84). The best two-variable model (adjusted R² = 0.91) also included BC concentration in bulk deposition. Consistent with predictions from equations governing exchange chemistry, the proportion of charge contributed by Ca²⁺ increased with increasing total anion concentration, while that of Na⁺ decreased. At the plot scale, Cl^- concentrations in treatment solutions had a stronger effect (p = 0.06) on BC concentration in Oa-horizon solutions than did Ca²⁺ concentrations (p = 0.33). In individual horizons of individual plots, BC and total ion concentrations were correlated, but cation composition was not consistent within horizons from different plots. This study detected no evidence of long-term cation depletion in the soils controlling streamwater, but did detect extremely base-poor plots. Because acid deposition affects surface horizons first, streamwater chemistry may not be an adequate way to assess nutrient supply of forest soils.