Exploring the sources of sulfur ion deposition and runoff in forest watersheds on the northern side of Lake Biwa, central Japan

We aimed to determine the sources of sulfur ions in precipitation in forest watersheds and runoff from the watersheds by collecting and analyzing bulk deposition and streamwater samples from the small Kutsuki and Surumi forest watersheds, which are on the northern side of the Lake Biwa basin in central Japan. The major ions and the delta sup(34)S values in the samples were analyzed. Samples were continuously collected from 1990 to 2010, and the average +- standard deviation (minimum, maximum) SO4sup(2-) concentrations were 1.60 +- 0.31 (0.76, 3.58) mg /L in the streamwater samples from Kutsuki and 6.58 +- 1.54 (3.68, 16.1) mg /L in the stream water samples from Surumi. The concentrations in the stream water samples from the different watersheds were significantly different (t-test, p < 0.01). However, the average SO4sup(2-) concentrations in the bulk deposition samples from the two watersheds were similar, being 2.16 +- 1.30 (0.20, 10.2) mg /L in Kutsuki and 2.27 +- 1.27 (0.35, 6.07) mg /L in Surumi (t-test, p > 0.05). The average non-sea-salt (nss-) delta sup(34)S values in the bulk deposition samples were +5.28 permille in Kutsuki and +5.89 permille in Surumi. The average nss-delta sup(34)S values in the streamwater samples from Kutsuki and Surumi were significantly different (t-test, p < 0.01), at +2.14 and -3.97 permille, respectively. It has been found that the delta sup(34)S values in coal products from northern China vary between -3 and -1 permille and that the delta sup(34)S values in Japanese oil vary between +5 and + 18 permille. Our findings suggest that the significant differences between the nss-delta sup(34)S values in the streamwater samples from the different forests were caused by differences in the abilities of the forest canopies to catch dry SO4sup(2-) deposition. The Kutsuki watershed forest consists of young conifers and deciduous broadleaf trees and has a relatively open canopy, whereas the Surumi watershed forest consists of mature conifer trees packed tightly together and has a less open canopy. There are no domestic sources of air pollution to the north of the Kutsuki watershed, and it appears that the Kutsuki watershed was only affected by SO4sup(2-) deposition from the Asian continent. In contrast, the dense canopy of the Surumi watershed was affected by local SO4sup(2-) pollution from national roads (which are 1 km northwest of Kutsuki) rather than continental sources. Overall, our results suggest that the differences in the SO4sup(2-) concentrations in the streamwater from the two watersheds are caused by the differences in the forest canopies in the two watersheds.