Tracing the source of nitrate in a forested stream showing elevated concentrations during storm events

To clarify the source of nitrate increased during storm events in a temperate forested stream, we monitored temporal variation in the concentrations and stable isotopic compositions including Delta O-17 of stream nitrate in a forested catchment (KJ catchment, Japan) during three storm events: I, II, and III (summer). The stream showed a significant increase in nitrate concentration, from 24.7 to 122.6 mu M, from 28.7 to 134.1 mu M, and from 46.6 to 114.5 mu M, during the storm events I, II, and III, respectively. On the other hand, the isotopic compositions (delta N-15, delta O-18, and Delta O-17) of stream nitrate showed a decrease in accordance with the increase in the stream nitrate concentration, from +2.5 parts per thousand to -0.1 parts per thousand, from +3.0 parts per thousand to -0.5 parts per thousand, and from +3.5 parts per thousand to -0.1 parts per thousand, for delta N-15; from +3.1 parts per thousand to 3.4 parts per thousand, from +2.9 parts per thousand to -2.5 parts per thousand, and from +2.1 parts per thousand to -2.3 parts per thousand for delta O-18; and from +1.6 parts per thousand to +0.3 parts per thousand, from +1.4 parts per thousand to +0.3 parts per thousand, and from +1.2 parts per thousand to +0.5 parts per thousand, for delta O-17 during the storm events I, II, and III, respectively. Besides, we found strong linear relationships between the isotopic compositions of stream nitrate and the reciprocal of stream nitrate concentrations during each storm event, implying that the temporal variation in the stream nitrate can be explained by simple mixing between two distinctive endmembers of nitrate having different isotopic compositions. Furthermore, we found that both concentrations and the isotopic compositions of soil nitrate obtained in the riparian zone of the stream were plotted on the nitrate-enriched extension of the linear relationship. We concluded that the soil nitrate in the riparian zone was primarily responsible for the increase in stream nitrate during the storm events. In addition, we found that the concentration of unprocessed atmospheric nitrate in the stream was stable at 1.6 +/- 0.4, 1.8 +/- 0.4, and 2.1 +/- 0.4 mu M during the storm events I, II, and III, respectively, irrespective of the significant variations in the total nitrate concentration. We concluded that the storm events have fews impacts on the concentration of unprocessed atmospheric nitrate in the stream, and thus the annual export flux of unprocessed atmospheric nitrate relative to the annual deposition flux can be a robust index to evaluate nitrogen saturation in forested catchments, irrespective to the variation in the number of storm events and/or the variation in the elapsed time from storm events to sampling.