Nitrous oxide (N2O) is a potent greenhouse gas and ozone-depleting substance. Global streams and rivers contribute substantially to N2O emissions, with most studies focusing on human-impacted lowland systems. Data are scarce for pristine regions, particularly permafrost areas undergoing rapid change. Northern Hemisphere permafrost soils contain a vast reservoir of nitrogen, and as permafrost thaws, this nitrogen can be released into fluvial networks, potentially leading to increased N2O production. Alaskan studies show elevated inorganic nitrogen and N2O in streams from thawing permafrost, but direct measurements of N2O emissions from permafrost-affected rivers are lacking. The Qinghai-Tibet Plateau (QTP), the largest cryosphere outside the poles, contains significant permafrost and nitrogen stores and feeds major Asian rivers. Human activities also add to potential N2O increases. This study directly measures N2O concentrations and fluxes in four QTP watersheds with varying permafrost characteristics to understand N2O dynamics in these alpine environments and to evaluate their contribution to the global N2O budget.

Existing research highlights the significant role of streams and rivers in the global nitrogen cycle and N2O emissions. Studies primarily focus on human-impacted lowland areas, showing a strong link between anthropogenic nitrogen enrichment and elevated N2O fluxes. However, data from pristine, high-altitude, permafrost regions are limited, creating uncertainty in global N2O emission estimates. While studies in Alaska have indicated increased nitrogen delivery from thawing permafrost into streams, leading to higher inorganic nitrogen and N2O concentrations, the extent to which these findings translate to elevated N2O emissions from permafrost-affected rivers remains unclear. The QTP, a critical water source for Asia, presents a unique system with its vast permafrost and nitrogen reserves, coupled with increasing human influence, making it crucial to understand the N2O dynamics in its fluvial networks.

The study was conducted in four headwater catchments on the East Qinghai-Tibet Plateau (EQTP), covering a wide altitudinal range (1650–4600 m). Sampling was done across different seasons (spring, summer, fall) from 2016 to 2018. The study area encompassed various permafrost zones: continuous, discontinuous, sporadic, and isolated. Dissolved N2O concentrations were measured using the headspace equilibration method on a gas chromatograph. N2O fluxes were measured using floating chambers, differentiating between diffusive and ebullitive fluxes. Simultaneously, water and sediment samples were collected for physicochemical and microbial analyses (including dissolved inorganic nitrogen (DIN), dissolved oxygen (DO), pH, total phosphorus, and microbial gene abundances related to nitrogen transformations). Vegetation coverage and normalized difference vegetation index (NDVI) were determined using GIS analysis. Statistical analyses included Pearson correlations, stepwise regression, and regression tree analysis to identify relationships between environmental variables and N2O concentrations and fluxes. Finally, a Monte Carlo simulation was used to upscale N2O emissions from the study area to the entire QTP.

All sampled streams and rivers showed N2O supersaturation. However, dissolved N2O concentrations (average 12.4 ± 1.7 nmol L⁻¹) were one-third of the global average. N2O fluxes (average 9.4 ± 6.2 µmol m⁻² d⁻¹) were an order of magnitude lower than the global average. Ebullition contributed minimally to total N2O fluxes. Riverine DIN concentrations were low and negatively correlated with vegetation cover, suggesting plant uptake of terrestrial nitrogen. A strong positive relationship was found between NO₃⁻ and N2O concentrations under undersaturated DO conditions. The ratio of nitrite reductase to nitrous oxide reductase genes in riverbed sediments was low compared to other lotic systems, suggesting a molecular basis for low N2O yield. Stepwise regression analysis showed weak relationships between N2O fluxes and various environmental variables. N2O fluxes were highest in 3rd-order streams, declined in 4th and 5th-order rivers, and were slightly elevated in 6th and 7th-order rivers. Upscaling estimates suggest that QTP streams and rivers emit 0.432–0.463 Gg N2O-N yr⁻¹, which is a minor contribution to global fluvial emissions (~0.15%). The percentage of N2O in total GHG emissions for the EQTP was also low compared to human-impacted rivers.

The low N2O emissions from EQTP rivers, despite the potential for high emissions from thawing permafrost, are primarily attributed to low riverine DIN due to plant uptake, unfavorable conditions for denitrification (undersaturated DO), and a low N2O yield due to microbial community composition. The findings highlight the unique N2O dynamics in high-altitude environments and the importance of considering plant uptake and microbial processes when assessing N2O emissions from permafrost regions. The weak relationships between N2O fluxes and environmental variables indicate complex interactions governing N2O dynamics in these systems. The longitudinal trends in N2O fluxes likely reflect changes in hyporheic exchange and suspended sediment loads with stream order. The study's limitations regarding the lack of 1st and 2nd-order stream data introduce uncertainty in the upscaling estimates, but the overall conclusion remains that current N2O emissions from QTP rivers are relatively minor.

This study provides the first comprehensive assessment of N2O emissions from permafrost-affected rivers on the EQTP. The unexpectedly low emissions are explained by the interplay of terrestrial plant uptake, limited denitrification under undersaturated dissolved oxygen conditions, and a low N2O production yield, highlighting the unique biogeochemical dynamics of these high-altitude systems. While currently a minor source, future warming and increased human impacts may drastically increase N2O emissions from these rivers, amplifying climate warming. Further research should focus on higher temporal and spatial resolution data, especially from low-order streams, to refine emission estimates and better understand future changes.

The study's upscaling estimates have uncertainties due to the limited data from 1st and 2nd-order streams, which may contribute disproportionately to overall N2O emissions. The extrapolation of N2O fluxes from higher-order streams to these low-order streams introduces uncertainty into the total emission calculations for the entire QTP. While the study provides valuable insights into N2O dynamics in permafrost-affected rivers, additional research is needed to fully capture the variability and potential future changes in these systems.