Riverine response of sulfate to declining atmospheric sulfur deposition in agricultural watersheds

Sulfur received extensive study as an input to terrestrial ecosystems from acidic deposition during the 1980s. With declining S deposition inputs across the eastern United States, there have been many studies evaluating ecosystem response, with the exception of agricultural watersheds. We used longterm (22 and 18 yr) sulfate concentration data from two rivers and recent (6 yr) data from a third river to better understand cycling and transport of S in agricultural, tile-drained watersheds. Sulfate concentrations and yields steadily declined in the Embarras (from ~10 to 6 mg S L^-1) and Kaskaskia rivers (from 7 to 3.5 mg S L^-1) during the sampling period, with an overall -23.1 and -12.8 kg S ha^-1 yr^-1 balance for the two watersheds. There was evidence of deep groundwater inputs of sulfate in the Salt Fork watershed, with a much smaller input to the Embarras and none to the Kaskaskia. Tiles in the watersheds had low sulfate concentrations (< 10 mg S L^-1), similar to the Kaskaskia River, unless the field had received some form of S fertilizer. A multiple regression model of runoff(cm) and S deposition explained much of the variation in Embarras River sulfate (R² = 0.86 and 0.80 for concentrations and yields; n = 46). Although atmospheric deposition was much less than outputs (grain harvest + stream export of sulfate), riverine transport of sulfate reflected the decline in inputs. Watershed S balances suggest a small annual depletion of soil organic S pools, and S fertilization will likely be needed at some future date to maintain crop yields.