Research advances in riverine nitrate source identification using stable isotopes of nitrogen and oxygen

Nitrate is a major water pollutant, and identifying its sources and quantifying their relative contributions provide an essential scientific basis for watershed water pollution control and management. Stable nitrogen and oxygen isotopes of nitrate, expressed as δ¹⁵N and δ¹⁸O, exhibit relatively consistent and distinguishable characteristics among different nitrogen sources, thereby providing critical information for source identification in aquatic environments. In recent years, with continuous innovations in nitrate isotope pretreatment methods and advances in analytical techniques, nitrate source apportionment in water environments has progressed from early qualitative discrimination based on a single isotope to quantitative analysis based on dual isotopes. This development has been further strengthened by the integration of multi-isotope approaches and Bayesian mixing models, such as SIAR and MixSIAR, which enable quantitative partitioning of multiple nitrogen sources under complex environmental conditions. Against this background, the present review synthesizes the characteristic ranges and interpretive significance of nitrate nitrogen and oxygen stable isotopes in aquatic systems, and, on the basis of bibliometric statistics, examines the major research hotspots concerning the application of these isotopes in riverine systems. Current studies are found to focus primarily on source identification and apportionment. Accordingly, this review summarizes the development of analytical methods for nitrate isotope measurement, discusses in detail the effects of land use and rainfall-runoff processes on nitrate δ¹⁵N and δ¹⁸O values, evaluates the methodological evolution of nitrate source analysis from qualitative identification to quantitative apportionment, and finally proposes priorities for future research. Existing studies indicate that the principal nitrate sources in rivers include chemical fertilizers, soil nitrogen, domestic sewage, livestock manure, atmospheric deposition, and industrial wastewater. The relative contributions of these sources are strongly regulated by land-use patterns, rainfall processes, and hydrological conditions. In agricultural areas, nitrate is commonly derived from fertilizer application, soil nitrification, and manure inputs, and its transport is closely associated with storm runoff, irrigation return flow, and leaching through the vadose zone. In forested catchments, nitrate is generally dominated by soil nitrogen transformation and atmospheric deposition, while strong vegetation uptake and soil retention often reduce overall nitrogen export. In wetlands, alternating aerobic and anaerobic conditions favor denitrification, which not only removes nitrate but also enriches the residual nitrate pool in δ¹⁵N and δ¹⁸O. In urban rivers, nitrate is frequently associated with domestic sewage, industrial discharges, and runoff from impervious surfaces, and its isotopic composition often reflects the combined effects of point-source inputs and in-stream biogeochemical processing. These findings demonstrate that nitrate isotopes can serve not only as tracers of pollution sources but also as indicators of nitrogen transformation processes such as nitrification and denitrification. Despite these distinct advantages, several limitations remain. End-member isotope databases are still inadequate, especially at regional scales where source signatures may vary substantially with local climate, geology, hydrology, and human activities. In addition, dynamic process studies remain relatively weak, particularly with respect to high-frequency observations and time-series analysis, which restricts the ability to capture rapid responses of nitrate sources and transformation pathways during storm events, seasonal transitions, and other hydrological disturbances. Uncertainty arising from source overlap, spatiotemporal variability in end-member compositions, and process-related isotope fractionation also continues to challenge robust source apportionment. Future studies should therefore prioritize the construction of regional end-member databases, strengthen high-frequency and process-oriented monitoring, and promote the integration of multi-isotope approaches with other tracers, including microbially derived indicators. Greater efforts should also be directed toward coupling multi-factor tracing systems with optimized quantitative models, so as to improve both the accuracy and the applicability of nitrate source apportionment. Such advances will enhance the capacity of stable isotope techniques to support precise diagnosis of nitrogen pollution and provide more reliable scientific support for watershed water quality protection and management.