Sulfur isotope dynamics in two Central European watersheds affected by high atmospheric deposition of SOx

Sulfur fluxes and delta(34)S values were determined in two acidified small watersheds located near the Czech-German border, Central Europe. Sulfur of sulfate aerosol in the broader region (mean delta(34)S of 7.5 parts per thousand CDT) was isotopically heavier than sulfur of airborne SO2 (mean delta(34)S Of 4.7 parts per thousand). The annual atmospheric S deposition to the Jezeri watershed decreased markedly in 1993, 1994, and 1995 (40, 33, and 29 kg/ ha.yr), reflecting reductions in industrial S emissions. Sulfur export from Jezeri via surface discharge was twice atmospheric inputs, and increased from 52 to 58 to 85 kg/ha.yr over the same three-year period. The delta(34)S value of Jezeri streamflow was 4.5 +/- 0.3 parts per thousand, intermediate between the average atmospheric deposition (5.4 +/- 0.2 parts per thousand) and soil S (4.0 +/- 0.5 parts per thousand, suggesting that the excess sulfate in runoff comes from release of S from the soil. Bedrock is not a plausible source of the excess S, because its S concentration is very low (<0.003 wt.%) and because its delta(34)S value is too high (5.8 parts per thousand) to be consistent with the delta(34)S of runoff. A sulfur isotope mixing model indicated that release of soil S accounted for 64 +/- 33% of sulfate S in Jezeri discharge. Approximately 30% of total sulfate S in the discharge were organically cycled. At Nacetin, the same sequence of delta(34)S(IN) > delta(34)S(OUT) > delta(34)S(SOIL) was observed. The seasonality found in atmospheric input (higher delta(34)S in summer, lower delta(34)S in winter) was preserved in shallow (<10 cm) soil water, but not in deeper soil water. delta(34)S values of deeper (> 10 cm) soil water (4.8 +/- 0.2 parts per thousand) were intermediate between those of atmospheric input (5.9 +/- 0.3 parts per thousand) and Nacetin soils (2.4 +/- 0.1 parts per thousand), again suggesting that remobilization of soil S accounts for a significant fraction (roughly 40 +/- 10%) of the S in soil water at Nacetin. The inventories of soil S at both of these sites are legacies of more intense atmospheric pollution during previous decades, and are large enough (740 and 1500 kg S/ha at Jezeri and Nacetin, respectively) to supply significant sulfur fluxes to runoff for several more decades. The ongoing release of this stored soil S may significantly delay the recovery of water quality under declining atmospheric S deposition. Analysis of possible scenarios that would result in different S isotope composition of rainfall, runoff and soil suggested that biologic S isotope fractionation must be involved. Mineralization of organic soil S was recorded in two opposite but complementary vertical isotope trends: while soil water had lower delta(34)S values in deeper horizons, bulk soil had higher delta(34)S values in deeper horizons. Copyright (C) 2000 Elsevier Science Ltd.