| 04 Feb 2026
Nitrous oxide dynamics across nitrogen and pH gradients in headwater streams
Abstract. Headwater streams in agricultural landscapes can contribute substantially to nitrous oxide (N₂O) emissions, yet the environmental controls on stream N₂O dynamics remain poorly resolved, particularly in systems with low pH. We investigated 72 Danish headwater streams spanning broad gradients in pH (5.0 - 8.8), land use, and soil type to identify the main drivers of N₂O variability. Nitrate (NO₃⁻) was the strongest predictor of N₂O saturation, and its positive association with N₂O intensified under acidic conditions according to linear mixed models. Ammonium, dissolved organic carbon, and stream depth also showed significant but weaker positive relationships with N₂O. Spatial differences among streams explained considerably more variation than seasonal or regional patterns, underscoring the dominance of local factors. Streams with pH < 6 consistently exhibited higher N₂O saturation, and generalized additive modelling indicated a marked decline in N₂O levels beginning near pH 6. Despite generally high N₂O saturation, approximately 9 % of observations displayed undersaturation, which occurred mainly in streams with low NO₃⁻ concentrations and across all seasons. Our results indicate that acidic, weakly buffered catchments may enhance in‑stream N₂O accumulation even at moderate nitrogen levels. These findings highlight the need to consider pH‑related controls when assessing N₂O dynamics in freshwater networks and when designing mitigation strategies for agricultural landscapes.
Review of “Nitrous oxide dynamics across nitrogen and pH gradients in headwater streams”
Here, Carstensen et al. studied N2O dynamics in headwater streams thoroughly, in this case with streams with important differences among pH and nitrogen gradients. N2O is probably the least studied greenhouse gas in freshwater ecosystems, because its concentrations are lower than CH4 and CO2, but also due to the limited and reliable devices for its measurements compared to the other greenhouse gases. So, this work makes a relevant contribution to understanding of greenhouse gas budget and biogeochemical processes in freshwater ecosystems, specifically in streams environments. However, the manuscript contains several methodological issues that must clarify before going one step forward to the acceptance. Below see my main comments and other comments that I hope can help to improve the manuscript.
The most critical issue I see in the methodology concerns gas sample collection. Because, you introduce ambient air during the equilibration step. Therefore, if the dissolved gas concentration in the sample is already close to atmospheric equilibrium (or below it), this approach is very inaccurate for determining N2O concentrations and could strongly bias the results. Could you quantify the uncertainties associated with this procedure and include with the controls (ambient air samples collected at the site)? Also, the manuscript does not clearly specify the volume of gas injected into the vials. Moreover, because you are working near atmospheric concentrations, did you verify the vacuum in the glass vials? Was the vacuum uniform among vials, and was it measured before and/or after sample filling? Any variability in the initial vacuum would increase uncertainty and could introduce systematic bias in the reported concentrations (because you are in the atmospheric ranges).
Continuing with the N2O measurements, the analytical device you used appears to have limited sensitivity for detecting small concentration changes. Although you report a detection limit of 0.15 ppm (which is close to half of the current atmospheric concentration), the sensitivity is not specified and for low values is extreme relevant. This information is essential, particularly because your measurements are close to atmospheric levels (ca. 0.336 ppm) or below. Please report the analytical sensitivity and precision in the Methods section. The concentration ranges reported in Lines 131, 132 can only be considered reliable if the sensitivity and associated uncertainties of the analysis are clearly stated. Given the reported detection limit, your theoretical lower quantification range would be approximately 55% below atmospheric concentration. Without explicitly accounting for analytical uncertainty, the reported undersaturation values may not be robust or even detectable within the methodological constraints of the instrument.
I performed a brief calculation using the values reported in Table 1, considering your sampling setup (40 mL water sample and 10 mL headspace) and assuming no initial N2O in the headspace. Using water concentrations of 0.1 (low), 2.4 (mean), and 22.2 (max) micrograms per L, and applying the temperature dependence of Henry’s law at 5, 10, and 20 °C, the expected equilibrium N2O concentrations in the headspace for the lowest concentration scenario would fall below the detection limit at all temperatures. Given this, could you expand the Methods section to clarify (i) the exact gas volume injected for analysis and the calibration curve, (ii) the effective detection and quantification limit under your measurement configuration, and (iii) how you corrected headspace concentrations to N2O dissolved concentrations (including temperature corrections and blank removal)? All of this is for the 9% of N2O sinks reported.
Other comments