Disentangling in-stream nitrate uptake pathways based on two-station high-frequency monitoring in high-order streams

Stream and river systems are an important compartment of nitrogen (N) transport and retention from terrestrial landscape to marine ecosystems. In-stream nitrate uptake in rivers involves complex assimilatory and dissimilatory pathways, which often exhibit spatiotemporal variability due to stream hydraulic, biotic (e.g., phytoplankton and periphyton) and abiotic (e.g., temperature and light availability) variations. Two-station based multi-parameter high-frequency monitoring allows quantitative disentangling of multi-path nitrate uptake dynamics at the reach scale. However, such monitoring and analysis are still limited to few small river types (e.g., headwaters and spring-fed rivers) and have not been well explored in higher order streams with varying hydro-morphological and biogeochemical conditions. Here, we conducted the two-station high-frequency monitoring in five high-order stream reaches in central Germany (i.e., two in the 4^th order Weiße Elster River and three in the 6^th order Bode River). Two-station 15-min time series of nitrate-N and dissolved oxygen were used to calculate the N mass balance and whole-stream metabolism, respectively. The mass-balance based net nitrate uptake rates (U_NET) differed between reaches with contrasting morphology (e.g., 13.8±3.85 mg N h^-1 in the more natural Weiße Elster compared to 2.05±0.83 mg N h^-1 in the modified reach of Bode), as well as between different periods in the same reach (e.g., higher in post-wet period than in dry period). The measured GPP and the related autotrophic nitrate assimilation (U_A) were determined by seasonal-varying radiation and riparian-canopy shading conditions. Heterotrophic N uptake (U_D), including denitrification and heterotrophic assimilation, was further disentangled as the difference between U_NET and U_A. This rarely reported uptake pathway showed relatively higher values than U_A, especially during late spring periods; moreover, it exhibited obvious diel signals that are significantly negatively correlated with DO. We further summarized difficulties and cautious considerations in conducting such two-station monitoring campaign at larger reach scales. In conclusion, benefiting from the less labor-consuming and high-frequency sensor monitoring, the two-station methods for N mass balance and stream metabolism can be applied at larger reach scales, and can well disentangle the multiple N uptake pathways that often exhibit high spatiotemporal heterogeneity.