The manuscript describes greenhouse gas flux measurements from a constructed domestic waste water treatment system (using percolation through soil along trenches). The study reasonably includes all stages of effluent, from septic tanks to trenches (where effluent is introduced below-ground through pipes), and vents to the atmosphere. Collecting this kind of data is challenging, given the seasonal and diurnal fluctuations, and events such as ebullition and flushing through vents. The combination of spot measurements and use of continuous chambers is hence appropriate.

My main issue with the analysis is the framing of fluxes in comparison to control soils. This may seem reasonable, as any changes from undisturbed (“natural”) conditions wold be attributed to the use of effluent treatment. However, the results indicate significantly reduced fluxes of CO₂, which the authors interpret as a net sink of CO₂ through effluent treatment. It is grossly simplistic, as the fate of C introduced from domestic effluent is not explained or even considered. To become a sink of CO₂, the soils have to assimilate C somehow, but no such mechanism is proposed. Measurements use dark chambers, so potential fertilisation effects on vegetation resulting in increased photosynthesis can not account for this finding. It is more likely that the soil disturbance, including replacement of soil by gravel, has reduced CO₂ flux artificially in the treatment beds. A true control treatment would have been required, with trench construction and hence disturbance identical to that of the effluent treatment beds.

To be publishable, the authors have to re-evaluate the flux calculation (which is never presented or explained in the methods). The assumed “net CO₂ sink” has a major influence on the net GHG balance, which is not robust, and can not be presented as such.

There is relatively little discussion of results. References are included to compare individual aspects to literature, but I missed a comprehensive evaluation. Maybe a separation of ‘Results’ and ‘Discussion’ would work better.

On balance, I struggle to justify the publication in Biogeosciences. Microbial processing and GHG implications of domestic effluent is included, which is good, but I missed a reliable treatment of interaction with soil biota, and consideration of sources and sinks beyond a simplistic flux comparison (which would make this paper relevant to BGS). It’s a shame, as the data set is impressive, and surely useful for assessing GHG impacts such domestic schemes. Unfortunately, the designed is flawed by lack of a true control, making it difficult to provide robust GHG budgets.

Detailed comments

76: Can you give at least some detail of what a ‘rotating biological contactor’ is?

103: I’m unsure what “each of the two ST chambers” means. There seems to be no descriptions of “chambers” of STs, and it becomes confusing when you describe flux chamber measurements.

167-184: Section 3.2 reports results that are not part of this study, and don’t relate to methods presented earlier. Please either present methods of measurement to obtain these values in the methods, or integrate the information provided here into the description of sites and STUs.

232/233: This should be Figure 2B (?)

269/270: Reported fluxes of 0.00, and a range of [0.00; 0.00] are not meaningful. If fluxes were measured, they should be reported with three significant figures, whatever the magnitude. The same applies in line 274.

284: You state “clearly seem to increase”, but later state that results are not statistically significant, which is contradictory.

311-314: This largely repeats information already given in the Methods.

325: If the net flux of CH4 is negative, the STU as a whole (comprising effluent for STs, soil and vegetation) is a net sink of CH4. The ‘natural’ CH4 sink (measured on control soils) has been diminished due to gross CH4 emissions increasing, but the net term remains negative.

340: The ‘median net uptake’ is not well explained, and I think even misleading. The data show positive CO₂ fluxes for all treatments, and as I assume the chamber was obscure, there is no realistic prospect of CO₂ uptake (but possibly for methane). Here. You seem to refer to the difference between effluent treatments in the STU to control soils, where a lower flux in treatments is deemed a ‘net uptake’. If at all, this would be a gross uptake, as the net effect of all processes is evidently still a source of CO₂. I suggest moving away from the terminology of ‘net uptake’ when discussing these fluxes, and provide interpretation of possible mechanisms and processes in the discussion.

Figure 4: Extremely high CO₂ fluxs are not really plausible (especially in control plots). Have flux regressions been quality-checked?

Figure 4: Colours mean different things between panels, which is confusing. For A, panels already separate location, so no need to use different colours. I assume that the locations are the same in Panel B (i.e. PE on left, Se in middle, Control on right)?

353: “in the west of Ireland”, or “in western Ireland”.

354-356: Was there a notable flux response to the drought period, and did it differ between STU and control plots?

362- 365: This sentence on relative contributions is hard to interpret. Can you explain your distinction of absolute vs. relative fluxes better?

424-427: Avoid repeating methods here.

459: The same problem with presenting significant figures. Please don’t present fixed number of decimal places, but instead always the same number of significant figures for all fluxes.

479: “higher lower”?

509/510: This is not correct. Units should be kg CO₂-eq cap-1 yr-1. Likewise in lines 622 and 623. (Please check throughout text!)

Figure S2: Please provide more information in the figure caption to allow readers to follow what’s shown without reading the text in detail. What are the two colours, and what are “chamber 1” and “chamber 2”?

This manuscript describes differences in greenhouse gas fluxes measured continuously or discretely from two onsite wastewater treatment systems that include secondary treatment as part of the treatment train: one with a rotating biological contactor, the other with a coconut husk media filter. The treated water is dispersed to a soil treatment unit and, in both cases, untreated septic tank effluent is also dispersed to the STU. Comparisons of flux values obtained using continuous and discrete measurements are made for the septic tank, the soil above the STU, and the vents at the end of the pipes that deliver effluent to the STU. GHG fluxes from the STU are compared to those from a Control area.

There are a number of issues that I think need to be addressed:

1. The difference in CO₂ flux between Control and STUs is often negative, that is, the STU is somehow acting as a sink for CO₂. The possible mechanism(s) by which this takes place are not really discussed in the paper. Very few microbial processes assimilate CO₂ in wastewater (e.g., autotrophic ammonia oxidation), and these would likely be minimized by both secondary treatment processes, which promote ammonia oxidation before it reaches the STU. One large difference between the Control and STU soils is the absence of subsurface horizons in the latter, which would have been removed to install the effluent delivery system. The removed soil would contribute to CO₂ flux at the soil surface which, when compared to Control soil, would have a lower CO₂ The authors should, then, reconsider comparisons with Control soil, not only for CO₂, but for all three gases (assuming they don’t have data for an STU that did not receive effluent), since gross consumption and production of CH₄ and N₂O can take place in the “missing” soil.
2. There are several published studies on GHG emissions from secondary treatment units that show that these can be considerable. The treatment units used in this study both rely heavily on microbial processes to remove and transform C and N, which produces CO₂ and N₂O. In addition, mechanical mixing and/or turbulent flow in these units tends to result in loss of CH₄ and N₂O form effluent to the atmosphere. In the absence of values for these emissions, the flux values that were measured lack context. Differences in flux between secondary treated effluent and tank effluent could help provide some context.
3. There is, in general, very little discussion of biogeochemical processes that could explain results in this paper, and limited discussion of results in the context of the current published literature. For the most part flux values are reported and compared within the study, without getting into the biogeochemical and/or abiotic processes that may that drive these in the soil or the effluent. It may be that Biogeosciences is not a good match for this work.
4. Most researchers working in this area will not have access to the equipment needed for continuous measurements of GHG fluxes; rather, discreet flux measurements are more likely to be made by most. As such, the results of this study could be made more useful by developing a minimum data set (spatially and temporally) required to approximate the accuracy of flux estimates made using continuous measurements. Although I understand this has clear limitations related to climate, treatment train, etc., it would be a good start, and a meaningful contribution to the field.