Watershed-scale changes in terrestrial nitrogen cycling during a period of decreased atmospheric nitrate and sulfur deposition

AbstractRecent reports suggest that decreases in atmospheric nitrogen (N) deposition throughout Europe and North America may have resulted in declining nitrate export in surface waters in recent decades, yet it is unknown if and how terrestrial N cycling was affected. During a period of decreased atmospheric N deposition, we assessed changes in forest N cycling by evaluating trends in tree-ring δ15N values (between 1980 and 2010; n = 20 trees per watershed), stream nitrate yields (between 2000 and 2011), and retention of atmospherically-deposited N (between 2000 and 2011) in the North and South Tributaries (North and South, respectively) of Buck Creek in the Adirondack Mountains, USA. We hypothesized that tree-ring δ15N values would decline following decreases in atmospheric N deposition (after approximately 1995), and that trends in stream nitrate export and retention of atmospherically deposited N would mirror changes in tree-ring δ15N values. Three of the six sampled tree species and the majority of individual trees showed declining linear trends in δ15N for the period 1980–2010; only two individual trees showed increasing trends in δ15N values. From 1980 to 2010, trees in the watersheds of both tributaries displayed long-term declines in tree-ring δ15N values at the watershed scale (R = −0.35 and p = 0.001 in the North and R = −0.37 and p <0.001 in the South). The decreasing δ15N trend in the North was associated with declining stream nitrate concentrations (−0.009 mg N L−1 yr−1, p = 0.02), but no change in the retention of atmospherically deposited N was observed. In contrast, nitrate yields in the South did not exhibit a trend, and the watershed became less retentive of atmospherically deposited N (−7.3% yr−1, p < 0.001). Our δ15N results indicate a change in terrestrial N availability in both watersheds prior to decreases in atmospheric N deposition, suggesting that decreased atmospheric N deposition was not the sole driver of tree-ring δ15N values at these sites. Other factors, such as decreased sulfur deposition, disturbance, long-term successional trends, and/or increasing atmospheric CO2 concentrations, may also influence trends in tree-ring δ15N values. Furthermore, declines in terrestrial N availability inferred from tree-ring δ15N values do not always correspond with decreased stream nitrate export or increased retention of atmospherically deposited N