Controls (or drivers) of the exchange of methane between the atmosphere and the surface of various wetlands have always been complicated, because of the nature of the processes and pathways and the large spatial variability across short distances, combined with temporal variability. Initially, simple variables such as temperature, water table and plant functional type were useful in ‘explaining’ patterns of methane exchange across landscapes. Since then, attention has been given to a variety of properties and processes which have led to a better ‘explanation’ and perhaps ‘prediction’ of the methane exchange. For belowground activities, important processes are provision of carbon for methanogenesis and the ability of plants to transport methane directly (avoiding methanotrophy in the profile) as well allowing oxygen to create a potential zone of methanotrophy in the rhizosphere. Thus, linking belowground root/rhizome traits and methane emission can be ‘complicated’ but a necessary advance.

This manuscript is an example of the latter, focusing on the role of root and rhizome traits in the exchange (emission) of methane, based along a gradient created by the thawing of permafrost palsa in northern Sweden. Three ‘thaw stages’ have been recognized at this site, with changes in thermal regime, hydrology and vegetation characteristics creating a very large variation in methane emission rates. The research focuses on determining root and rhizome traits across triplicates of each thaw phase, though this is not done in the automatic flux chambers, but from sites located nearby with similar properties. As noted by the authors, there is not complete species similarity in mean coverage between the triplicate chamber and core data but when combined into herbaceous and shrub categories there is general agreement (Table 1). Daily median flux of methane was calculated, to avoid the usual ‘burp’ of methane occurring occasionally which would be unlikely to be derived from root processes, and chamber carbon dioxide flux was used to define seasonal patterns. Root and rhizome data were collected carefully from depth increments and biogeochemical data were used to provide evidence for some of the processes leading to methane production, and how they might be related to root activity.

The manuscript is well structured and written and provides valuable evidence for the role of belowground plant activities in controlling methane emissions. It is ‘digestible’ in that the information in the manuscript is restricted to two Tables and five Figures, with voluminous supporting data in Appendices. The strong Introduction ends with several expectations/hypotheses based on previous studies, but in this case applied to the specific Stordalen site, in which palsa are thawing and thus perhaps similar to features elsewhere in the Arctic. Based on methane fluxes from the nine chambers and root characteristics from the ‘similar’ sites, the study presents strong evidence for correlations between methane emission rate and the herbaceous and shrub root and rhizome properties, and whether patterns changed with the season, based on carbon dioxide exchange. In general, it appears that correlations were stronger for root properties compared to rhizomes, though there are mixed relationships shown in Figures 3 and 4. The reasons for positive, negative or no correlations were examined in the Discussion and the relevance of these relationships may vary with the specific features of Stordalen, which are somewhat unusual. The relationships between the traits and methane emission were examined in terms of processes related to changes from shrubs to herbaceous plants and the root trait structure (e.g. density, specific length and diameter), and drew upon results presented elsewhere which could be applicable at Stordalen. The overall conclusion was that belowground root/rhizome activities were an important component of methane emission, that change from shrubs to herbaceous would lead to increased emissions, but there is a bit of ‘ying-yang’: shrubs may supply more carbon substrates, whereas herbaceous plants may be important conduits of methane to the atmosphere. Such is life in the real world of Nature and ‘teams’ like these authors are needed to see at least part of the whole.

The study is ‘well referenced’, with 145 in the References, perhaps too many?

Specific comments:

In Table 1, it seems that mosses formed a significant plant cover in the chambers (e.g. Sphagnum balticum, capillifolium and riparium in the three thaw stages). Would methane emission rates be affected by this coverage and some differences between chambers and cores and different wetness? This is addressed in 4.2 line 486 onwards …..

A tremendous amount of work went into this study, particularly in the measurements of root and rhizome properties collected from the cores. If I understand correctly, a 12 cm diameter peat core was dived into 10 cm depth increments and these were then divided into four vertical quarters, which was used for trait measurements, resulting in subsampling of the core sections (and subsampling of that in some cases). This are measurements based on an initial volume of 1130 cm3, and it must take considerable time and patience to extract roots and rhizomes, weighing in some cases to 0.00001 g. pH in 0.01M Cacl2 is much easier and quicker …….

While I accept that the belowground contribution is important to processes such as methane emission, are there any ways of making it ‘easier’ to define and measure these properties over a wider range of sites? If so the properties could be included and tested within models which are grappling with the high spatial and temporal variability of methane emission rates.

Technical issues:

I found few. I thought that the first two sentences of 2.6 (lines 214-215) could be combined into one: they are repetitive. Minor typos in places.

Fig. B13. I was confused the right hand diagram. I think the dashed lines have circles whereas the solid lines have triangles …… not what is written in the legend?

I did find that the total root length of shrubs in the intact sites was about 180 km per square meter (to 30 cm depth). Jeez, as it is 10 km from Stordalen to Abisko, it means that one square meter of intact permafrost has enough shrub root to go back and forth 18 times or perhaps all the way to Norway or the Norwegian Sea? Might get the attention of readers that permafrost is not ‘dead’ but rich in roots and though root length decreases in thawed sites, the impact of roots and rhizomes and change from shrub to herbaceous plants allows for larger methane emission rates.

General

This study is an exploratory in situ experiment in finding soil biomass trades that can explain CH4 emission. Specifically, these biomass trades are linked to plant-mediated transport of CH4 in different thawing stages in a permafrost peatland in north Sweden.

The topic of this study is relevant and needed to get a better understanding in the processes behind measured CH4 fluxes, which is relevant for upscaling and modelling.

It is a very well written paper (with some minor suggestions for improvement below) and the gathered data seems to be analysed with precision.

However, due to the limited amount of samples and not having a reference (no biomass) in the different thawing conditions, it is very hard to draw conclusions out of the results. The authors should therefore be careful in their statements and conclusions. I will elaborate on this more below.

Further, for persons who do not work in permafrost regions (like me) it would be very helpful to get a better explanation what kind of processes are relevant in this system and specifically what is changing when thawing occurs. Also explain what a palsa exactly is.

Overall, I think the study is a relevant stepping stone to a wider study with or more sites, or more samples from this site. I would recommend it for publishing after revisions, where 1) statements and conclusions are nuanced, 2) other explanatory variables are explored (if possible), and 3) where authors give suggestions on what is needed to link below-ground biomass trades to plant-mediated CH4 transport based on what they have learned from their study.

Concerns

If I look at Figure 3 and Figure B6, I see mainly a trend between thaw stage and CH4 fluxes.  As the authors state in the discussion, other drivers for CH4 fluxes are soil moisture, water table depth, substrate availability and (not mentioned) temperature. There is a clear increase in CH4 fluxes with the increase in thawing of permafrost. How do the above-mentioned drivers change with the thawing gradient? It would be very relevant to know how the below-ground biomass characteristics are related to the drivers mentioned above. If it’s possible then a multi variate regression or a mixed effect model could be done to exclude the effect of other drivers than below ground biomass characteristics. Or add these drivers to you PCA analyses. Authors state that ‘we studied the permafrost thaw-driven CH4 flux variation with a plant trait-based approach with a focus on belowground traits’ (L392), I think you can only do that if you exclude the other factors.

Further, I think the authors could have done some more investigation in looking into relationships between GPP-CH4 and diurnal cycles in CH4 fluxes to give more evidence for plant mediated transport. 

The following sentence is due to the above mentioned uncertainty too speculative ‘SRL was highest and root diameter lowest at the partly thawed stage, possibly reflecting increased resource acquisition particularly for shrubs with high SRL and low root TD’ .

dC13 and ac are measured and it seems that the results are interesting (Fig. B13). Maybe it’s worth adding that to the results, but also explain a bit better the mechanism behind the results and why it is relevant. It would help to make two who dC13 and ac in separate plots. Be careful with statements about CH4 oxidation being higher or lower. It’s not clear for the data isn’t it. Or if it is, then please explain a bit better why it is.

Details

L110-117: This part could be in the introduction, which helps to understand the system.

L134: -2 is in superscript

L155, 166: Peat and soil are used in a similar way (peat temperature or soil temperature?)

L222: Filtering on R2 only could lead to removing very low fluxes. Adding the actual slope to the filter (allowing a higher R2 with lower slopes would help)

L243-245: I don’t understand what that means. And for which results is this used?

L293: This is a very long title

L298: 1.5x lower

L301: 22.2x  22x (remove decimal to be consistent)

L312: Introduce the abbreviation CV, and add for between mean and intact

L323-328: Could you show all results in a graph? That would be easier to follow

L382-383: why is it relevant to know if CH4 is produced by hydrogenotropic or acetoclastic methanogenesis?

L416: How heterogeneous is your plant cover? Can differences not be explained by that or the sample size?

L488-490: It is not clear what is the result of this study and what not. It’s also not clear why this is relevant.

L498: ‘…with belowground traits for the first time…’, what does that mean?

L500-501: ‘…differed only slightly in strength.’ What does this mean, that there is a linear trend in every season and only slightly differ? Are they significantly different? If not, don’t write it.

L504: In figure 5 I mainly see a relation between the three thawing stages which can be related to many different factors. It would be better to replace Figure 5 with figure B6, that one is clearer.

L504: ‘strongly associated’ what does associated mean? (same in next sentence). Be specific about that

L505: ‘…root porosities not responding strongly to changes…’ this statement is too big for the results you have.