The review is provided as supplementary material.

Thank you for the opportunity to review the paper by Adame et al., titled 'Soil disturbance in wetlands by feral pigs increases greenhouse gas emissions”, submitted to European Geosciences Union - biogeosciences.

This study investigates whether soil disturbance caused by invasive hoofed mammals in tropical wetlands of northern Australia enhances greenhouse gas emissions, specifically carbon dioxide, methane, and nitrous oxide. The authors combined field measurements of greenhouse gas fluxes with some laboratory incubations of soil and root respiration, alongside measurements of various environmental parameters, to determine the drivers of changes in fluxes. The study found that the highest emissions occurred near disturbed sites, suggesting that feral pigs have a measurable impact on wetland greenhouse gas cycling.

Overall, this is an interesting and timely study addressing an important and underexplored topic. The manuscript is generally well written and presents novel findings with potential implications for wetland management and greenhouse gas accounting. However, several methodological details require clarification, and some interpretations should be presented more cautiously.

Specific comments to address:

Line 69 – “Old disturbed” reads awkwardly and should be rephrased.

Figure 1 – Please indicate what the grey shaded areas represent.

Line 120 – It is not clear at what depth soil physicochemical characteristics were measured. Please provide more details on how these measurements were conducted. For example, were soil cores collected and redox measured along a depth profile? Soil redox potential changes substantially with depth, particularly in wetlands where soils transition from aerobic surface layers to increasingly anaerobic subsurface conditions. This redox gradient strongly influences the production and distribution of microbial greenhouse gases, meaning hotspots for CO2, CH4, and N2O may differ with depth. Please clarify whether soil redox potential values were adjusted relative to the Standard Hydrogen Electrode (SHE) and specify the reference electrode used. This information is important for interpreting the redox conditions associated with greenhouse gas production. Understanding where and how soil redox was measured is important for interpreting these results.

Line 134 – More methodological detail regarding greenhouse gas sampling is required eg: How much gas was removed from the chamber headspace at each sampling interval? Were vials over-pressurised? Please clarify whether ambient/background atmospheric samples were collected alongside chamber measurements as including atmospheric reference samples is useful for quality assurance of GC measurements, identifying potential contamination or instrument drift, and contextualising chamber concentrations relative to ambient greenhouse gas concentrations. This is particularly important for low-flux measurements and GHG ‘uptake’ measurements to see if your starting concentrations are near atmospheric (start measurement) and final concentrations (end measurement) ~below atmospheric levels. What were the precision and detection limits of the GC system used? Were any samples outside the calibration range? What is meant by “linearity of fluxes within each chamber was checked”? eg was an r² threshold used to reject poorly linear fluxes? If so, what threshold was applied? How are airtight seals with soil confirmed?  When were chambers/collars installed? Eg were they pre-installed and revisited, or inserted immediately before measurements? Both approaches have important caveats that should be acknowledged. Long-term collar installation can alter soil biogeochemistry, whereas installing chambers immediately prior to measurements may disturb soils and release naturally accumulated gases, potentially altering measured fluxes. How was temperature measured for use in Flux Equation 1?

Line 157 – The manuscript extrapolates emissions to fluxes per hectare, but it may also be useful to cautiously estimate the impact per feral animal, or cautiously use reference satellite data, or published estimates to try upscale to the NT region and see what effect this disturbance could be having regionally or nationally. This could help contextualise the broader regional impact of feral pig disturbance.

Table 1 – There is a substantial soil temperature difference between reference and disturbed sites. I would have expected the opposite trend, where disturbed sites lacking vegetation cover would exhibit higher soil temperatures due to increased solar exposure and radiation heating bare/darker soil surfaces. Could this reflect a time-of-day sampling artefact? For example, were disturbed sites sampled in the morning and reference sites later in the day? Please consider and discuss.

Figures 3 and 4 – Nitrous oxide fluxes in Figure 3 range from approximately 200–800 µg/m2/h in disturbed sites, whereas Figure 4 appears mostly within ±5 µg/m2/h. This represents a difference of roughly two orders of magnitude, while methane and carbon dioxide fluxes appear comparable between figures. Please check whether this discrepancy is correct or due to an error.

Figure 6 – As mentioned above, were soil redox potential values adjusted to a standard hydrogen electrode/reference electrode? Please also specify the depth at which redox was measured and indicate what the lightly shaded area represents.

Line 291 – Disturbed plots are colder? Are these soil temperatures or air temperatures? This is currently unclear. Was this time-of-day artefact? Seems counterintuitive.

Line 306 – “13δC” is incorrectly formatted and should be written as δ13C throughout. I also do not think the statement that “disturbed soils showed slightly reduced δ13C values” is fully supported by the data. Four out of six sites showed essentially the same/ non-significantly different δ13C values between treatments, while only two disturbed sites were statistically more negative, and only by a small fractionation (~2‰). Therefore, the broader claim appears overstated. Furthermore, bulk soil δ13C is only an indirect indicator of methanogenic processing and integrates multiple carbon pools and decomposition pathways. Small differences in δ13C could simply reflect pig disturbance-driven mixing of soil organic matter, altered litter inputs, or differences in carbon turnover rather than enhanced methanogenesis directly

Line 303 – The discussion regarding oxygen limitation in soils is also somewhat speculative because oxygen concentrations were not directly measured; only soil redox was assessed, and respiration rates and possibly only at the surface. This does not necessarily indicate that subsurface soils were aerobic too. Wetlands occupy low landscape positions and commonly feature shallow water tables, so deeper anaerobic conditions may still occur irrespective of surface measurements.

Figure 7 – Consider adding an arrow indicating 'time' progression from left to right along the x-axis. Also consider using different colours or line styles for CO2 and N2O, as they are difficult to distinguish in black-and-white (and in colour) printing. It may also be worth discussing the concept of ecosystem recovery/re-establishment after disturbance as inset ‘e’. While this was not directly investigated here, it could represent an interesting direction for future studies examining post-disturbance dynamics once feral pigs move on.

Line 331 – I recognise the body of Australian literature may be limited but are there additional recent studies in tropical wetlands beyond Iram et al. (2022) that support this claim that seasonal differences in GHG fluxes are low which could be discussed here? A couple of other relevant papers that could be considered on seasonal Australian wetland GHGs are:

https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2023JG007556

https://aslopubs.onlinelibrary.wiley.com/doi/abs/10.1002/lno.11158

https://link.springer.com/article/10.1007/s13157-014-0522-5

and this undulate research on Australian alpine wetland GHGs: https://www.sciencedirect.com/science/article/pii/S0301479723018224

Conclusion: The introduction discusses national greenhouse gas accounting schemes or conservation programs to support feral pig exclusion as a mechanism to reduce greenhouse gas emissions. I suggest reconnecting this concept in the conclusion, albeit cautiously given this is the first study of its kind, nevertheless, it represents an important potential policy implication of this work.

It may also be worthwhile to provide a preliminary estimate of the broader magnitude of these processes. As noted above, this could potentially be achieved by combining measured disturbance impacts with satellite imagery, aerial mapping, or existing estimates of feral pig disturbance extent across the region.

End of review.