Atmospheric nitrogen (N) deposition affects soil nitrate () dynamics in forests, yet mechanisms driving accumulation in semi‐arid regions remain unclear. Here, we quantified atmospheric and microbial contributions to soil by analyzing 17 O anomalies in 26 semi‐arid forests under high N deposition in northern China. Heavier N‐polluted temperate forests showed lower 17 O anomalies in soil (1.0‰ ± 0.8‰), indicating lower atmospheric contributions (4.2% ± 3.2%), but higher atmospheric concentrations (0.3 ± 0.3 mg N kg −1 ) and microbial nitrification rates (459.7 ± 225.4 kg N ha −1 yr −1 ). Both contribution and concentration of atmospheric in soils increased with inputs, whereas nitrification rate positively correlated with N deposition independent of precipitation, suggesting deposited N pollutants rather than water as a main driver of soil production and accumulation. These findings are important for understanding sources and accumulation of in temperate forests under N deposition.
Graphical abstract summarizing the integrated assessment of water quality and biodiversity responses to microhabitat construction in Beijing's five river basins. The figure shows water quality improvement (reduced BOD5, COD) but increased phytoplankton at basin scale, contrasted with a demonstration site where restoration led to 80.5% phytoplankton reduction, increased water transparency, and expanded submerged vegetation. Urban river restoration requires understanding water-ecosystem coupling. This study investigated microhabitat construction effects across five Beijing river basins and a demonstration site. At 47 basin-wide monitoring sites (2024–2025), BOD5 and CODmn decreased (p < 0.01), while phytoplankton increased (p < 0.001), reflecting persistent eutrophication. TN was the strongest predictor of biodiversity (r = −0.773, p < 0.0001). PCA showed PC1 and PC2 explained 73.88% of total variance (PC1: 40.24%, PC2: 33.64%). At the Jinghe River demonstration site, integrated restoration was associated with 80.5% phytoplankton reduction (t = −4.690, p = 0.0001), transparency improved from 40–50 cm to 70–100 cm, and submerged vegetation exceeded 40%. Due to the observational design without control sites, causal attribution cannot be established; the results demonstrate associations between integrated restoration (∼200 yuan/m2) and improved ecological conditions, offering a cost-effective nature-based solution for water-scarce megacities.