the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 05 Feb 2025
Plant community composition controls spatial variation in year-round methane fluxes in a boreal rich fen
Abstract. Climate change is expected to impact the methane budget of boreal peatlands, highlighting the need to understand the factors that influence methane cycling, including plant community structure. In northern peatlands, the majority of methane is transported through plants, and the magnitude of this process is strongly linked to plant community composition. Therefore, detailed information about the role of plants regulating year-round methane fluxes is highly valuable. This paper explores the causes of spatial variability in plot-scale methane fluxes in a northern boreal rich fen. Methane fluxes were measured using the manual chamber technique in the context of fine-scale biomass variations in plant community compositions from 36 study plots over 232 days throughout a full year. The mean methane flux rates for snow-free and snow seasons were 2.55 and 0.21 mg CH4/m2/h, respectively. We found a significant correlation between methane fluxes and a vascular plant cluster associated with the occurrence of the sedge Carex rostrata during year-round, snow-free and snow season periods. More precisely, C. rostrata grew at the point of flux measurement in 13 plots and 44–49 % of the measured methane fluxes originated from these plots during the three periods. The biomass of vascular plants, sedges, and C. rostrata, as well as the ratio of vascular plant to bryophyte biomass, also significantly correlated with methane fluxes in year-round and snow-free season. By identifying vegetation-driven emission hotspots, these results can enhance efforts to upscale emission predictions and improve ecosystem-scale methane modelling. Thus, our findings provide valuable insights for predicting realistic future changes in peatland methane emissions throughout the year.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(2144 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2144 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-217', Anonymous Referee #1, 11 Mar 2025
- AC1: 'Reply on RC1', Eeva Järvi-Laturi, 10 Apr 2025
-
RC2: 'Comment on egusphere-2025-217', Anonymous Referee #2, 12 Mar 2025
This manuscript presents results from a study of methane fluxes in a northern calcareous fen and investigates the role of plant community composition on the measured fluxes. Methane flux was measured throughout the annual cycle, allowing for an investigation of the role of vegetation in both snow-free and snow covered periods. As little data is available in the wintertime in northern peatlands, this adds to our understanding of winter and annual methane emissions in peatlands.
Overall, the study is careful conducted, and an impressive number of methane fluxes were measured and used in the analysis. Plots were then assigned to plant community types based on a cluster analysis of species composition and the presence and biomass of Carex rostrata was observed to result in higher methane fluxes in the snow-free period and annually. Bryophyte composition alone was not a good predictor of spatial variation in methane flux. Although the authors indicate that other environmental variable such at water table depth (WTD) were not strong predictors of methane flux, it is possible that these interacted with the plant community, although this was not fully explored in the present analysis. Currently, the role of environmental variables is largely assessed with a correlation analysis, but I provide some additional suggestions of how this could be further explored. Given the wet nature of the study site, it still may be that these environmental variables do not explain much variation. However, there is a large amount of unexplained variation in methane flux among plots where sedge biomass is low, and it may be that WTD or soil temperature (or some of the other variables measured) would explain some of this variation if these plots are investigated separately.
My other main suggestion is to improve clarity in some of the methods and reporting of results, with specific suggestions outlined below.
Specific comments:
Lines 21-23: It was not clear to me exactly what this sentence was aiming to convey. Can you reword to make this clearer. I have tried to interpret it and if the following suggestion captures the correct meaning, then you can use it to help with the update “Plant community dominated by Carex rostrata accounted for 13 of the measured plots with these plots contributing 44–49% of the measured methane flux during the three periods”.
Line 30: northern peatlands are not the main terrestrial source of methane. Wetlands may be, but northern peatlands only account for a small portion of the wetland total. Please update this sentence for clarity. e.g., the global methane budget indicates wetland emissions of 248 Tg CH4/yr https://www.globalcarbonproject.org/methanebudget/ , while the reference cited here estimates emissions of only 38 Tg CH4-C/yr from the whole northern region.
Line 50: oxidize instead of oxidate
Line 131: How did you cut them to estimate biomass? Did you include only green tissue, or some depth that you considered active? As am sure you know, it can be difficult to define living moss biomass, so a few more methological details would be useful here.
Lines 137-138: What was the extent of microtopography at the site? Did you consider correcting the WTD measurements for local elevation variation to better represent the actual WTD at the flux measurement collars? Do this effect the interpretation of the role of WTD for accounting for variation in methane flux?
Line 164: It’s not clear if this VP to bryophyte ratio is based on biomass or cover. I assume biomass, but it can be made clearer in the text.
Line 165: Please provide additional information about the correlation analysis. Was this Pearson correlation? It also isn’t really clear that this was done using average conditions across the sample periods and not instantaneous values (this is my interpretation after reviewing Table B2), which would likely give different results (so just be clear here in the methods what was done). Also, see my comments below about considering the variables together in a multiple regression analysis to assess whether there are interactions between the environmental variables and vegetation clusters.
Lines 193-194: Are the numbers in brackets averages across all the plots? Please specify in the text and include an estimate of variation (e.g., standard error or standard deviation)
Lines 199-200: I had a hard time following which species were with which cluster. Maybe add numbers in front of the clusters to clearly separate them.
Lines 264-266: I totally agree with this statement, but maybe it is also important to highlight here that fluxes were much less variable over the snow-period and did not vary significantly among the identified species clusters. This would suggest that even much lower sampling effort could effectively capture winter fluxes, helping to estimate annual emissions. Vargas and Le 2023, Biogeosciences also supports this conclusion https://doi.org/10.5194/bg-20-15-2023
Lines 275-280: What about interactions between the soil environment (WTD and temperature) and the plant communities? You didn’t find a significant correlation of CH4 with WTD alone, but it could be that there was a significant correlation in some plant communities and not other, resulting in no significant pattern across the whole dataset. Did you consider multiple regression models that included the plant community information alongside the environmental drivers?
Lines 289-291: This does not seem to align with the results in Figure 5a where there were no significant differences in CH4 flux among the VP clusters in the snow-period. Are you overstating here?
Line 325: I’m not sure how easy it will be to estimate BM ratio with remote sensing. The community identity or importance of sedges on an areal coverage basis would seem like a variable that is easier to measure with imagery.
Line 341-342: It’s not clear to me where you did this? The correlation analysis looks at each variable individually and does not consider if they interact in predicting flux. Based on Figure 7, there is a wide range of CH4 fluxes when C. rostrata biomass is low, so some other variable must be explaining this. Is it possible that the response of CH4 to the environmental variable differed among the clusters? Did you investigate this (e.g., something like an ANCOVA or multiple regression with the cluster type as a categorical variable that interacts with things like pH, WTD and soil temperature)?
Line 377: I highly encourage the authors to deposit the full datasets in an open access data repository to ensure availability of the data for future studies/meta analyses.
Table B3: Please add units to the columns.
Citation: https://doi.org/10.5194/egusphere-2025-217-RC2 -
AC2: 'Reply on RC2', Eeva Järvi-Laturi, 10 Apr 2025
This manuscript presents results from a study of methane fluxes in a northern calcareous fen and investigates the role of plant community composition on the measured fluxes. Methane flux was measured throughout the annual cycle, allowing for an investigation of the role of vegetation in both snow-free and snow covered periods. As little data is available in the wintertime in northern peatlands, this adds to our understanding of winter and annual methane emissions in peatlands.
Overall, the study is careful conducted, and an impressive number of methane fluxes were measured and used in the analysis. Plots were then assigned to plant community types based on a cluster analysis of species composition and the presence and biomass of Carex rostrata was observed to result in higher methane fluxes in the snow-free period and annually. Bryophyte composition alone was not a good predictor of spatial variation in methane flux. Although the authors indicate that other environmental variable such at water table depth (WTD) were not strong predictors of methane flux, it is possible that these interacted with the plant community, although this was not fully explored in the present analysis. Currently, the role of environmental variables is largely assessed with a correlation analysis, but I provide some additional suggestions of how this could be further explored. Given the wet nature of the study site, it still may be that these environmental variables do not explain much variation. However, there is a large amount of unexplained variation in methane flux among plots where sedge biomass is low, and it may be that WTD or soil temperature (or some of the other variables measured) would explain some of this variation if these plots are investigated separately.
My other main suggestion is to improve clarity in some of the methods and reporting of results, with specific suggestions outlined below.
- Response: We thank for the thorough feedback on our manuscript. We want to highlight that multivariate analyses were performed to the plant community data and that the correlation values are achieved both from Canonical and Detrended Correspondence Analyses (CCA and DCA), and pairwise comparisons with Pearsons’ correlations. We have clarified this throughout the manuscript to show which correlations are from which analyses. Further comments related to this topic can be found from the following responses.
Specific comments:
Lines 21-23: It was not clear to me exactly what this sentence was aiming to convey. Can you reword to make this clearer. I have tried to interpret it and if the following suggestion captures the correct meaning, then you can use it to help with the update “Plant community dominated by Carex rostrata accounted for 13 of the measured plots with these plots contributing 44–49% of the measured methane flux during the three periods”.
- Response: What we actually meant was that C. rostrata was present (as an individual species) at 13 plots. Only 10 plots were clustered as C. rostrata-community clusters, but the species was also present in other clusters. With this sentence we wanted to highlight the contribution of the species to high fluxes, not only the community clusters indicated by the species. We have edited the sentence as follows: “C. rostrata was present at 13 out of 36 plots, and these 13 plots contributed 44–49 % of the measured methane fluxes during the three periods.”
Line 30: northern peatlands are not the main terrestrial source of methane. Wetlands may be, but northern peatlands only account for a small portion of the wetland total. Please update this sentence for clarity. e.g., the global methane budget indicates wetland emissions of 248 Tg CH4/yr https://www.globalcarbonproject.org/methanebudget/ , while the reference cited here estimates emissions of only 38 Tg CH4-C/yr from the whole northern region.
- Response: Corrected.
Line 50: oxidize instead of oxidate
- Response: Corrected.
Line 131: How did you cut them to estimate biomass? Did you include only green tissue, or some depth that you considered active? As am sure you know, it can be difficult to define living moss biomass, so a few more methological details would be useful here.
- Response: We removed the non-living parts based on subjective assessment which was consistent between the samples (the first authors was solely responsible of this). We included only the colorful or leafy parts to represent the aboveground biomass. We have revised the manuscript accordingly.
Lines 137-138: What was the extent of microtopography at the site? Did you consider correcting the WTD measurements for local elevation variation to better represent the actual WTD at the flux measurement collars? Do this effect the interpretation of the role of WTD for accounting for variation in methane flux?
- Response: The experimental site is designed so that majority of the experimental plots (n=30/36) are located in lawn microhabitats that also generally depict our study site. The remaining six plots are located in mud microhabitat. Based on personal observations, hollows and hummocks do occur at the site. Small scale variations in microtopography do occurr but generally the variation is relatively minor. The suggestion to correct the WTD for local elevation could have increased the accuracy of the measurements but, unfortunately, this was not considered at the time of fieldwork. However, the WTD throughout the fen during snow-free season varied only approximately 15 cm (between -9.3 cm to 7 cm) and in plot-scale, the averages of minimum and maximum values were -4.7 cm and 1.8 cm, respectively. Therefore, we think it is reasonable to assume that the WTD did not vary extremely within the 1 m distance, i.e., the distance between the collar and the WTD tube.
Line 164: It’s not clear if this VP to bryophyte ratio is based on biomass or cover. I assume biomass, but it can be made clearer in the text.
- Response: The ratio is calculated as biomass of vascular plants divided by the biomass of bryophytes. We have clarified this in the text.
Line 165: Please provide additional information about the correlation analysis. Was this Pearson correlation? It also isn’t really clear that this was done using average conditions across the sample periods and not instantaneous values (this is my interpretation after reviewing Table B2), which would likely give different results (so just be clear here in the methods what was done). Also, see my comments below about considering the variables together in a multiple regression analysis to assess whether there are interactions between the environmental variables and vegetation clusters.
- Response: The correlation values in Table B2 are Pearson correlations. However, we gained correlation values also from multivariate analyses (DCA and CCA). In the revised manuscript we clearly state from which analyses the correlation values are derived from. In the analyses we used snow-free season averages to represent the overall environmental conditions since parameters were measured at different frequencies. We think that this instantaneous values approach provides a comprehensive view of long-term trends and reduces short-term variability, aligning with our study objectives and the multivariate analysis approach.
Lines 193-194: Are the numbers in brackets averages across all the plots? Please specify in the text and include an estimate of variation (e.g., standard error or standard deviation)
- Response: The numbers are total biomass estimates of all the plots, calculated with the biomasses of the collected samples. We have clarified this in the text but do not find it necessary to add standard deviation for the total biomass, as it was not taken into consideration in any of the analyses. Biomasses and the standard deviations of the collected samples can be found in Table A1.
Lines 199-200: I had a hard time following which species were with which cluster. Maybe add numbers in front of the clusters to clearly separate them.
- Response: The other reviewer also criticized this and therefore, we now use only the name of the strongest indicator species as a name for the cluster and revised the text accordingly throughout the revised manuscript.
Lines 264-266: I totally agree with this statement, but maybe it is also important to highlight here that fluxes were much less variable over the snow-period and did not vary significantly among the identified species clusters. This would suggest that even much lower sampling effort could effectively capture winter fluxes, helping to estimate annual emissions. Vargas and Le 2023, Biogeosciences also supports this conclusion https://doi.org/10.5194/bg-20-15-2023
- Response: As the reviewer is suggesting, our observations are in line with Vargas and Le (2023) that lower sampling efforts during the mid-winter could effectively capture the fluxes, aiding in the estimation of annual emissions. We have added this in the revised manuscript.
Lines 275-280: What about interactions between the soil environment (WTD and temperature) and the plant communities? You didn’t find a significant correlation of CH4 with WTD alone, but it could be that there was a significant correlation in some plant communities and not other, resulting in no significant pattern across the whole dataset. Did you consider multiple regression models that included the plant community information alongside the environmental drivers?
- Response: Please, note our response above where we describe the use of community analysis - both Detrended Correspondence Analysis (DCA) and Canonical Correspondence Analysis (CCA) - followed by Pearson correlation. CCA captures the complex interactions between the species, community clusters and all environmental variables used in the analysis. The plant community compositions we used were 1) All species (including both vascular plants and bryophytes), 2) Vascular plants alone, and 3) Bryophytes alone. In these analyses, pH and WTD were strongly connected to bryophyte clusters but not to vascular plant clusters. Soil temperature at 5 cm depth did not show significant correlation with any community cluster. Therefore, it is correct to state that some communities correlate significantly with edaphic factors and these relations are discussed in detail in sections 4.2 and 4.4.
Lines 289-291: This does not seem to align with the results in Figure 5a where there were no significant differences in CH4 flux among the VP clusters in the snow-period. Are you overstating here?
- Response: In figure 5a we show linear regression, which does not show significant differences between the communities, like the reviewer noted. However, in multivariate analyses, the first ordination axis of vascular plant data correlated significantly with snow season fluxes (r = 0.445 in DCA, -0.402 in CCA). The significant results during the snow season from DCA and CCA but not from linear regression models suggest that the relationship between vegetation and snow season methane fluxes is complex and non-linear and likely involves interactions with other environmental factors that linear regression did not capture. CCA’s ability to handle multivariate data provides a more reliable and accurate representation of the ecological dynamics during the snow season in comparison to linear regression models. This explanation has been added to the revised version.
Line 325: I’m not sure how easy it will be to estimate BM ratio with remote sensing. The community identity or importance of sedges on an areal coverage basis would seem like a variable that is easier to measure with imagery.
- Response: With remote sensing we mean mainly multispectral imagery, which has been shown to detect ratios of vascular plants and mosses reliably (Wolff et al. (2023), https://www.sciencedirect.com/science/article/pii/S1470160X23002820). We have clarified this in the revised version.
Line 341-342: It’s not clear to me where you did this? The correlation analysis looks at each variable individually and does not consider if they interact in predicting flux. Based on Figure 7, there is a wide range of CH4 fluxes when C. rostrata biomass is low, so some other variable must be explaining this. Is it possible that the response of CH4 to the environmental variable differed among the clusters? Did you investigate this (e.g., something like an ANCOVA or multiple regression with the cluster type as a categorical variable that interacts with things like pH, WTD and soil temperature)?
- Response: The values in 4.4 are again from both multivariate analyses and Pearson correlations. The reviewer is right that there is variation in the fluxes that cannot be explained by C. rostrata biomass – an observation that we have discussed shortly in the manuscript. However, it was an interesting idea to explore the difference in responses to environmental variables between the clusters still further. To address the reviewer's concern, we chose to examine the plots which did not have any C. rostrata growing in them (23/36 study plots). We analyzed the environmental variables together with snow-free season methane fluxes only on these plots to see which other factors, in addition to C. rostrata biomass, might show significant relationship with the fluxes. We tested this with linear model and ANOVA in R studio and discovered significant relationships between the fluxes and pH (p < 0.001) and litter coverage (p = 0.01). None of the other environmental or biomass variables resulted in a significant p-value. The Pearson correlation values between snow-free season methane fluxes and pH and litter were -0.66 and -0.29, respectively. These findings have been included in the revised manuscript in results and in discussion.
Line 377: I highly encourage the authors to deposit the full datasets in an open access data repository to ensure availability of the data for future studies/meta analyses.
- Response: We agree on the importance of open access data. However, data used for this study is part of a larger dataset of EcoClimate experiment at Oulanka Research Station. Full data repository is being developed, and all the data will be freely available at some point in the near future. You can familiarize yourself with the experiment via these links:
- https://www.oulu.fi/en/research/research-infrastructures/oulanka-research-station
- https://anaee.fi/facility/ecoclimate/#:~:text=Oulanka%20research%20station%20owns%20and%20operates%20the%20EcoClimate,instrumented%20and%20is%20designed%20to%20run%20for%20decades.
Table B3: Please add units to the columns.
- Response: Corrected.
Citation: https://doi.org/10.5194/egusphere-2025-217-AC2
-
AC2: 'Reply on RC2', Eeva Järvi-Laturi, 10 Apr 2025
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-217', Anonymous Referee #1, 11 Mar 2025
General comments:
This manuscript by Järvi-Laturi et al, looks into fine-scale spatial variation in methane flux in a boreal peatland over a full year. By measuring species-specific vascular plant and moss biomass together with chamber-based methane flux measurements, the authors found that increasing sedge biomass, in particular that of Carex rostrata, seem to increase methane emissions significantly especially in the snow-free season and full-year scales. Based on their analyses, vegetation composition and biomass seemed to be a stronger driver of methane fluxes than abiotic environmental variables at this site. The authors attributed these results to both enhanced provision of carbon substrates for methanogens and methane transport from soil to the atmosphere.
This is a study that provides the much-needed data and overview of both wintertime and full-year methane fluxes in a peatland, data that still to this day are quite scarce and thus valuable. As the authors mention in the manuscript, regional and global wetland methane budgets contain large uncertainties, some of which are related to the spatial variation in methane fluxes and vegetation composition. Therefore, this study, which looks at small scale (between-plot) methane flux variation, has potential in adding to our understanding of plant-mediated methane emissions in carbon-rich peatland ecosystems. While I see a lot of value and potential in this work, I recommend a list of improvements (major or minor, depending on how biomass measurements were conducted):
Specific comments:
- My main criticism is related to the plant biomass measurements and their representativeness of the chamber collars:
a) It is unclear how you scaled the vascular plant biomass measurements to the collars. You mention that you counted the number of shoots per species within the collar, but did you use this shoot number to scale the mean biomass of 10 samples to the actual collar (or did you actually take average of 20 sample plants if you ignored the fertile/sterile division? See comment 1 c)? If you did not scale these measurements to the collar, you cannot reliably estimate the collar species biomass, and so I recommend to do this and re-analyze your data and fix the results in 3.2, 3.3, 3.4, 3.5 with appropriately scaled biomass values. If you did do this, please explain this clearly and in more detail in the methods section.
b) It is also unclear how you took the moss biomass samples and scaled them to the collar. How did you determine which species were “most common” and which were not? What was the percentage cover limit (if there was one)? What was the spatial scale that you used for estimating the “most common” species- was it across the whole study site or within individual collar? If it was across the whole study site, I don’t quite understand the logic of taking a biomass sample equaling to 5 % of the collar area (i.e. 33.025 cm2) especially if the collar had an actual percentage cover <5%, in which case you may have overestimated the species biomass in the collar. Or did all of these “most common” species have >5% coverage in all collars?
If you looked at this at the scale of individual collars, did you take a sample that was equal to 5 % of the collar area per species for the five most common species within each collar and for the rest of the species found within the collar, you took a sample over an area equal to 1% area of the collar (i.e. 6.605 cm2)? What did you do if there were less than five species within the collar? Did every collar really have 10 species within them (now the text kind of makes it sound like there were but it doesn’t seem likely to me)? Please specify this in the methods.
And, most importantly, how did you scale the moss biomass samples to the percentage coverage within the individual collars? As with the vascular plant biomass, if no scaling was done, the moss biomass measurements do not represent the actual collar moss biomass and the data should be re-analyzed with appropriate scaling. Please specify this clearly in the methods.
c) How did you determine the locations for vascular plant and moss biomass sampling? Were the soil conditions (e.g. pH, soil moisture) similar to the collar? Did you look at and compare the general species composition in the collar vs the plots where you collected the biomass samples (between-species competition could affect some of the plant trait expression and thus biomass), for example by determining percentage cover? How did you decide which plants and moss patches to pick?
For mosses, did you look at e.g. moss stem density in some way to try to estimate the moss biomass in the collar and in the sampling points more reliably than just percentage coverage (the same percentage coverage can represent very different moss biomasses in different collars due to variation in moss stem density and other structural properties)? If not, the moss biomass estimates may be very uncertain and I would recommend discussing these uncertainties explicitly and in much more detail in the manuscript. Given these uncertainties, I would also recommend not to emphasize the ratio between vascular plants and bryophytes as an important methane flux predictor as much as you have so far in this manuscript, or at least combine it with adequate discussion about its uncertainties.
If you did not estimate the similarity (in terms of abiotic/biotic variables) between the collar and the plot where you collected the representative biomass samples, I would be careful making strong conclusions about collar-specific plant species biomass variation.
d) What do you mean by the “fertile” and “sterile” categories for the plant biomass? In my understanding fertile vs sterile categories are used in the context of evolutionary plant biology and plant reproduction (i.e. fertile vs sterile flowers). Or did you use it to somehow determine whether the species had vegetative culms (e.g. for Carex) from previous year? It is unclear to me how this classification is relevant to the topic of methane flux spatial variability, especially because you do not talk about these classes afterwards. If you used some kind of scaling for the collar biomass (see comment 1 a), did you take use of these fertile/sterile classes in that as well? Please add a clarification for this separation, what the rationale is behind it, and what you mean by the terms.
It is now also unclear whether the species-specific plant biomass is calculated as the mean of n=20 biomass samples (fertile + sterile) per species per plot, or are the species-specific plant biomasses actually still divided into the fertile (mean of n=10 biomass samples) and sterile (mean of n=10 biomass samples) classes. Based on your results, it seems that you took the mean of 20 samples by combining the fertile and sterile samples? Please add a clarification to your methods. - Since you are examining the spatial variation of methane fluxes within one study site and how plant biomass contributes to this variation (included in your research questions), I would suggest including additional measures of spatial methane flux variability (e.g., daily or seasonal coefficient of variation or other spatial variation metrics). This way you could quantify the spatial heterogeneity in methane fluxes which I think is currently lacking in this manuscript. Quantifying the spatial variation would be important background information for showing that there is indeed spatial methane flux variation in your peatland and then go to investigating the contribution of vegetation on it. You already touch on it a bit in 3.1 but only by talking about ranges in mean methane fluxes and visually showing plot-scale variability in Fig. 3 (which are good to show as well but do not really quantify the variation).
- Did you measure methane fluxes from one plot once a day or multiple times a day? How did you decide which plots to measure each day when half of the plots were measured per day (n=18 out of n=36)? Please add more detail about this in the methods.
- 116: The accumulated flux: it is based on a “24-hour” accumulated flux, but, based on your methods, you measured only between 8 am and 6 pm. It would be good to discuss that these accumulated fluxes are based on daytime fluxes and do not include nighttime fluxes, which may lead to the annual accumulated fluxes being quite uncertain. You do mention that you assumed that the fluxes did not vary significantly in the diurnal scale but some justification may be needed here (can you refer to, e.g., the EC data to show this?).
- 118: you mention that you used the value from a previous measurement for days without flux measurements. What was the maximum number of consecutive days where there were no measurements? Methane fluxes (and some plant-mediated methane transport proxies) have been found to vary a lot in daily and multiday scales (see e.g. Knox et al. 2021: https://doi.org/10.1111/gcb.15661).
- 161-162: How did you test the significance between BM variables and methane fluxes using LOESS? LOESS does not test hypotheses, it is used for exploring nonlinear trends (which I believe you did here). You could rephrase this to highlight that (exploring nonlinear trends between BM and methane fluxes between VP clusters).
- 185 (figure 3). It is hard to identify the individual plots based on color in this plot. If they just represent the different plots without considering the vegetation composition within the plots, I don’t think the coloring here is needed. You could just replace it with black lines and remove the legend (but mention that the black lines represent the different plots in the caption), for example.
Also, the plot numbers themselves do not really give any valuable information for the reader. If you want to show all the plot fluxes separately here, the plot numbers could be replaced with something simpler, such as 1-36, to improve the readability of this figure (the current numbering adds more complexity for the reader who is not familiar with your study site).
On the other hand, if you want to keep the coloring, could you do it based on e.g. vegetation composition grouping (for example based on your vegetation clusters)? If you do this, I would also recommend changing the red color of the soil temperature to black, because the red and green are difficult to separate visually for some readers with color-blindness.
The axis texts are also a bit small so please increase the font size. - 260-263: Could you add a sentence or two about how you would estimate the methane fluxes to change if you had measured them the same way as Alm et al 1999 or similar studies?
- 267-268 (and forward in this section): the plot numbers are not informative for the reader. You could instead describe the dominant vegetation of these plots (e.g. based on the vegetation clusters, which you show in B7).
- 289-291: I understand your reasoning for the uncertainties in the wintertime methane fluxes and plant contributions on them but I would like to see some more in-depth discussion about this based on other studies. How can you make this conclusion based on your data? Depending on the snowpack properties (e.g. porosity) you might have also measured lateral methane flux which did not originate from the actual collar, especially in windy conditions or due to pressure changes between the chamber enclosure and the surrounding atmosphere and snow. Also, I would like to see a better reasoning behind your statement of the plants contributing to the measured winter methane flux even through the snowpack. In theory, this might be possible if there were broken stems or culms that were exposed to the air above the snowpack, which could possibly contribute to the Venturi effect via pressure changes especially in windy conditions but otherwise I am not currently very convinced, especially since you did not find any significant differences between the vegetation clusters in snow cover seasons (Fig. 5).
- 292-307: I would move a majority of this part to the results section and discuss only the general aspects. Based on your research questions which are about the relationship between vegetation and methane flux, it doesn’t seem so relevant to me to discuss the species distribution in such length here. This information would also be more useful in the results section, because then the reader has more of an idea about what kind of vegetation the individual plots contain and where they were located (see my previous comments about plot numbering, possible grouping in figures and naming).
- 321: does it really provide labile carbon also in winter under the snowpack? Photosynthesis (and thus root exudation) is unlikely that efficient in those conditions, at least to the same extent as in the growing season, and especially so far up north as Puukkosuo where daylight hours are very few. Root decomposition could be one way too (but how efficient is microbial decomposition under the snow in cooler temperatures?) but I would like to see a bit more discussion based on more studies here.
- 342: it might be good to briefly discuss another explanation where vegetation would not be the main driver of the pH changes, and also add a bit more detail into how vegetation actually could have explained the pH variation.
- 344-346: This is an interesting finding and warrants a more detailed discussion. What do the other studies say and how could these theories apply to your study? Could the higher NO3- and NO2- concentrations contribute to pH or vegetation in some way that would enhance methanogenesis?
- 349: this could indeed be the case but, to support this argument, you could also add a number to represent the lack of strong temporal variation in WTD (e.g. standard deviation or coefficient of variation if you want to compare growing vs non-growing season variation for example).
- 350-353: Two points:
a) The correlation between peat depth and plant biomass makes sense in the biological sense that, when there is more peat, there is also more space for roots especially for more deeply-rooting vascular plants. Since you did not find significant correlations between peat depth and methane fluxes, I would be careful drawing strong conclusions about the influence of peat depth on methane fluxes via vegetation (but see my next point).
b) On the other hand, it is also possible that in the presence of deeply-rooted aerenchymatous vegetation, such as C. rostrata, the roots may provide labile carbon substrates in deep peat where methanogenesis increases despite the dominance of recalcitrant peat (i.e. indirect influences of peat depth on methane fluxes). The release of labile carbon compounds via root exudation could also trigger microbial carbon priming (see e.g. Waldo et al 2019: https://doi.org/10.1007/s10533-019-00600-6). However, be careful about your interpretations about the wintertime vegetation influences based on your data (see previous comment about wintertime fluxes), and keep in mind that the direct relationship between peat depth and methane flux was still nonsignificant. - 354: Please add more discussion about why soil temperature may not have correlated significantly with methane fluxes- soil temperature has been an important predictor of methane fluxes in multiple studies and discussing this opposing result would be warranted.
Technical comments:
- 30: Add “(CH4)” after “methane”. You could also replace the rest of the “methane”s with “CH4” if you want, especially since you use it in the flux units throughout the paper.
- The word ”dynamics” is used quite a lot throughout the introduction. I would recommend changing it to something more specific, as in some cases (e.g. “methane dynamics”) it may sound a bit vague.
- Generally through the whole manuscript: the term “year-round” doesn’t sound very good to my ear. How about “full-year”?
- 32: Saunois et al have a newer global methane budget paper (currently a preprint): https://doi.org/10.5194/essd-2024-115
- 34: instead of using the word “spatial and temporal dynamics”, maybe “spatiotemporal variation” or something similar would be better?
- 39: “ecosystem process” – maybe use another word, for example “These ecosystem-level processes…”
- 42: remove “layers” after topsoil, and add why rising temperatures lead to increased topsoil oxidation?
- 45: I would change the topic sentence to something shorter. Perhaps remove mention of hydrology and just start with “Vegetation type and its responses..”
- 47-48: maybe change the words “deeper” and “upper” to “anoxic” and “oxic” (this way it would focus on methane being transported from anoxic soil through the oxic soil and into the atmosphere)
- 49: I would be careful with the wording “better than any abiotic factor”- please add more references, or modify the sentence so that it doesn’t sound so definitive
- 59: the part “extensive, year-round, plot-scale flux data are, however, limited” sounds a bit complicated. Maybe something like “However, ... full-year methane flux data at the plot scale are limited”?
- Methods: the model numbers could be written in parentheses after the instrument, e.g. at row 99 you could put the LI-COR model number “LI-7810” in parentheses after mentioning the instrument. You already do this in the 2.5 section so it would be good to keep it consistent.
- 60: would “..spatial variability in methane fluxes..” work better?
- 71: what is “normal period”?
- 73-74: please add a detail saying where the pH was measured (peat I assume?)
- 76: was the variation standard deviation or other measure of variation? Or do you mean that 6.3 cm was the mean WTD during the study period? Please specify.
- 77: graminoids are herbaceous plants so this sentence should be changed accordingly (you could, for example just call them “vascular plants typical of rich fens” and then give the species examples)
- 91: The figure caption could be made even simpler, how about just starting from: “A map of Puukkosuo rich fen..”?
- 95: You could remove the mention of “manual” here since you introduce it later in this paragraph.
- 97: this sentence (“…, doing measurements from half (n=18) of the study plots per day”) could be made smoother, for example just: “.. from half (n=18) of the study plots per day”.
- 105: the end of the sentence starting with “making the possible dilution..” is a bit hard to understand, could you make this a bit clearer? Do you mean leakage? Good that you mention this though.
- 108: what exactly do you mean by “successful”? Visible linear increase in CH4 concentration?
- 115: it is a bit unclear now how you determined the snow cover- did you define it snow-free when there was snow but you were able to set the chamber on the collar? For transparency, it might be good to add this detail here.
- 145: do you have more details of the pH analyzer, other than the brand?
- 146-147: write the numbers in the molecules in subscript (e.g. NH4)
- 148: please add that you estimated the litter cover as a separate percentage cover, if this is the case. This could also be actually mentioned already in the plant community data where you talk about moss percentage cover.
- 151: “VP” abbreviation appears here for the first time but you don’t introduce it before this. Please add the abbreviation to the appropriate spot in the text (maybe introduction?) so you can then start using it: “vascular plants (VP)..”
- 181: put the “4” in “CH4” in subscript
- 204: add the name of the statistical test you used to obtain the F-values (“F=..”) for the first occurrence of the letter.
- 206 and forward: write the species names in italics and I would also write the complete names, e.g. “C. rostrata”. It would improve the readability if you wrote them in full form (the genus does not have to be written out since you have already discussed the species before).
- 215 (figure 4): please increase the axis text font size, and consider writing out the species names in the legend. Caption: replace “dot in the graph” with “data point”. Based on this plot, it also seems that there might be another plot group or cluster in T. ces where there are lower fluxes (the lower yellow point cloud which I would imagine could lead to a different smooth curve? Did you look into this? What might contribute to this trend? This is a bit extra but maybe worth discussing and/or looking into.
- 220 (figure 5): 1. increase the font size for axis texts. 2. Even though you list them in the caption, I would still write out the whole species names instead of the abbreviations in the plot. 3. It is very hard to see the median line in the dark blue boxplots so changing the color to something lighter might help readability. Also, even though the colors look nice, are they really needed here since you also give the same information on the x axis as cluster names? Or, if you would prefer keeping the colors, you could also consider removing the legend and in the caption write something along the lines of “the colors represent the clusters and are shown for clearer visualization”.
- 225: to remind the reader what these are, please add the term before “DCA” and “CCA” abbreviations: e.g. “detrended correspondence analysis (DCA)”.
- 226 forward: you could write the species names in complete forms here.
- 243: would something a bit more specific be better instead of calling the ratio “BM ratio”? For example, “VP:BRYO ratio”? The reader might forget what exactly “BM ratio” consists of and would need to come back to the definition of this term.
- 245: add “p” to the second p-value: “and p ≤ 0.01”.
- 250 (figure 6): increase the font size of axis texts and write out the species names in italics.
- 257: move the Jammet et al reference to the end of the sentence, and if possible, try to find another reference here since you mention multiple northern fens. Or you could just say “.. in a northern boreal rich fen (Jammet et al. 2017).”
- 256-260: I think these sentences should be in the results section and not in discussion. For example in the 3.1 section.
- 264: is this percentage based on your results? If yes, please indicate so in this sentence, and if not, add a reference.
- 274-275: plant traits are part of vegetation, so you could rephrase this by for example: “…could not be explained by aboveground plant biomass..”. Also, give examples of these plant traits, as well as the “ecohydrological aspects” and microbiota, and how they might contribute to the spatial variability in CH4 flux between the plots.
- 315-316: replace the "organic matter“ with "carbon substrates", and replace "and providing pathways” for example with: “.. for methanogenesis through deep root systems throughout the year”.
- 318: add “methane” in front of “transport”: “..may be due to the species’ high methane transport rate..”
- 318-319: why would C. rostrata have low oxidation potential in your study? As you say next in this sentence, this species has high root porosity (so it could also oxidize the rhizosphere), so why would the methane transport exceed the effect of methane oxidation in your study? Clarify briefly.
- 319: saying both “high porosity” and “large aerenchyma” is not needed as they refer to the same thing. You could instead just say “.. and high root porosity”.
- 325: this is a bit vague sentence. How about: “Thus, VP:Bryophyte ratio could be used as a parameter in peatland methane flux models together with remotely-sensed data products.” (But see comment 1 c)
- 330: add “gas” “high transport efficiency”: “high gas transport efficiency”
- 332-333: This sentence is a bit unclear. Do you mean that C. rostrata had more shoots and therefore plots with more C. rostrata shoots transported and emitted more methane?
- 334: add “methane” to “transport efficiency”: “methane transport efficiency”
- Figure 7: move this to the results? And increase the axis text font size and write out the species names in italics.
- 359: remove the mention of “causality” because you did not use methods for estimating causal relationships in this study.
- 361: remove “answer our first research question and”, and replace “affects the flux” with “affects methane flux”.
- 363-364: remove “answer our second research question and”
- 368: you didn’t really discuss plant traits in the discussion part, so I would remove the mention of plant traits here. Or, you could say for example: “Our findings suggest that, in addition to species-specific plant traits, the biomass ratio of vascular plants and bryophytes could potentially be used as a parameter for predicting peatland methane emissions” (but see comment 1 c).
- 373: I don’t think you have to show the reference at the end. The closing sentence would be stronger without it.
Citation: https://doi.org/10.5194/egusphere-2025-217-RC1 - AC1: 'Reply on RC1', Eeva Järvi-Laturi, 10 Apr 2025
- My main criticism is related to the plant biomass measurements and their representativeness of the chamber collars:
-
RC2: 'Comment on egusphere-2025-217', Anonymous Referee #2, 12 Mar 2025
This manuscript presents results from a study of methane fluxes in a northern calcareous fen and investigates the role of plant community composition on the measured fluxes. Methane flux was measured throughout the annual cycle, allowing for an investigation of the role of vegetation in both snow-free and snow covered periods. As little data is available in the wintertime in northern peatlands, this adds to our understanding of winter and annual methane emissions in peatlands.
Overall, the study is careful conducted, and an impressive number of methane fluxes were measured and used in the analysis. Plots were then assigned to plant community types based on a cluster analysis of species composition and the presence and biomass of Carex rostrata was observed to result in higher methane fluxes in the snow-free period and annually. Bryophyte composition alone was not a good predictor of spatial variation in methane flux. Although the authors indicate that other environmental variable such at water table depth (WTD) were not strong predictors of methane flux, it is possible that these interacted with the plant community, although this was not fully explored in the present analysis. Currently, the role of environmental variables is largely assessed with a correlation analysis, but I provide some additional suggestions of how this could be further explored. Given the wet nature of the study site, it still may be that these environmental variables do not explain much variation. However, there is a large amount of unexplained variation in methane flux among plots where sedge biomass is low, and it may be that WTD or soil temperature (or some of the other variables measured) would explain some of this variation if these plots are investigated separately.
My other main suggestion is to improve clarity in some of the methods and reporting of results, with specific suggestions outlined below.
Specific comments:
Lines 21-23: It was not clear to me exactly what this sentence was aiming to convey. Can you reword to make this clearer. I have tried to interpret it and if the following suggestion captures the correct meaning, then you can use it to help with the update “Plant community dominated by Carex rostrata accounted for 13 of the measured plots with these plots contributing 44–49% of the measured methane flux during the three periods”.
Line 30: northern peatlands are not the main terrestrial source of methane. Wetlands may be, but northern peatlands only account for a small portion of the wetland total. Please update this sentence for clarity. e.g., the global methane budget indicates wetland emissions of 248 Tg CH4/yr https://www.globalcarbonproject.org/methanebudget/ , while the reference cited here estimates emissions of only 38 Tg CH4-C/yr from the whole northern region.
Line 50: oxidize instead of oxidate
Line 131: How did you cut them to estimate biomass? Did you include only green tissue, or some depth that you considered active? As am sure you know, it can be difficult to define living moss biomass, so a few more methological details would be useful here.
Lines 137-138: What was the extent of microtopography at the site? Did you consider correcting the WTD measurements for local elevation variation to better represent the actual WTD at the flux measurement collars? Do this effect the interpretation of the role of WTD for accounting for variation in methane flux?
Line 164: It’s not clear if this VP to bryophyte ratio is based on biomass or cover. I assume biomass, but it can be made clearer in the text.
Line 165: Please provide additional information about the correlation analysis. Was this Pearson correlation? It also isn’t really clear that this was done using average conditions across the sample periods and not instantaneous values (this is my interpretation after reviewing Table B2), which would likely give different results (so just be clear here in the methods what was done). Also, see my comments below about considering the variables together in a multiple regression analysis to assess whether there are interactions between the environmental variables and vegetation clusters.
Lines 193-194: Are the numbers in brackets averages across all the plots? Please specify in the text and include an estimate of variation (e.g., standard error or standard deviation)
Lines 199-200: I had a hard time following which species were with which cluster. Maybe add numbers in front of the clusters to clearly separate them.
Lines 264-266: I totally agree with this statement, but maybe it is also important to highlight here that fluxes were much less variable over the snow-period and did not vary significantly among the identified species clusters. This would suggest that even much lower sampling effort could effectively capture winter fluxes, helping to estimate annual emissions. Vargas and Le 2023, Biogeosciences also supports this conclusion https://doi.org/10.5194/bg-20-15-2023
Lines 275-280: What about interactions between the soil environment (WTD and temperature) and the plant communities? You didn’t find a significant correlation of CH4 with WTD alone, but it could be that there was a significant correlation in some plant communities and not other, resulting in no significant pattern across the whole dataset. Did you consider multiple regression models that included the plant community information alongside the environmental drivers?
Lines 289-291: This does not seem to align with the results in Figure 5a where there were no significant differences in CH4 flux among the VP clusters in the snow-period. Are you overstating here?
Line 325: I’m not sure how easy it will be to estimate BM ratio with remote sensing. The community identity or importance of sedges on an areal coverage basis would seem like a variable that is easier to measure with imagery.
Line 341-342: It’s not clear to me where you did this? The correlation analysis looks at each variable individually and does not consider if they interact in predicting flux. Based on Figure 7, there is a wide range of CH4 fluxes when C. rostrata biomass is low, so some other variable must be explaining this. Is it possible that the response of CH4 to the environmental variable differed among the clusters? Did you investigate this (e.g., something like an ANCOVA or multiple regression with the cluster type as a categorical variable that interacts with things like pH, WTD and soil temperature)?
Line 377: I highly encourage the authors to deposit the full datasets in an open access data repository to ensure availability of the data for future studies/meta analyses.
Table B3: Please add units to the columns.
Citation: https://doi.org/10.5194/egusphere-2025-217-RC2 -
AC2: 'Reply on RC2', Eeva Järvi-Laturi, 10 Apr 2025
This manuscript presents results from a study of methane fluxes in a northern calcareous fen and investigates the role of plant community composition on the measured fluxes. Methane flux was measured throughout the annual cycle, allowing for an investigation of the role of vegetation in both snow-free and snow covered periods. As little data is available in the wintertime in northern peatlands, this adds to our understanding of winter and annual methane emissions in peatlands.
Overall, the study is careful conducted, and an impressive number of methane fluxes were measured and used in the analysis. Plots were then assigned to plant community types based on a cluster analysis of species composition and the presence and biomass of Carex rostrata was observed to result in higher methane fluxes in the snow-free period and annually. Bryophyte composition alone was not a good predictor of spatial variation in methane flux. Although the authors indicate that other environmental variable such at water table depth (WTD) were not strong predictors of methane flux, it is possible that these interacted with the plant community, although this was not fully explored in the present analysis. Currently, the role of environmental variables is largely assessed with a correlation analysis, but I provide some additional suggestions of how this could be further explored. Given the wet nature of the study site, it still may be that these environmental variables do not explain much variation. However, there is a large amount of unexplained variation in methane flux among plots where sedge biomass is low, and it may be that WTD or soil temperature (or some of the other variables measured) would explain some of this variation if these plots are investigated separately.
My other main suggestion is to improve clarity in some of the methods and reporting of results, with specific suggestions outlined below.
- Response: We thank for the thorough feedback on our manuscript. We want to highlight that multivariate analyses were performed to the plant community data and that the correlation values are achieved both from Canonical and Detrended Correspondence Analyses (CCA and DCA), and pairwise comparisons with Pearsons’ correlations. We have clarified this throughout the manuscript to show which correlations are from which analyses. Further comments related to this topic can be found from the following responses.
Specific comments:
Lines 21-23: It was not clear to me exactly what this sentence was aiming to convey. Can you reword to make this clearer. I have tried to interpret it and if the following suggestion captures the correct meaning, then you can use it to help with the update “Plant community dominated by Carex rostrata accounted for 13 of the measured plots with these plots contributing 44–49% of the measured methane flux during the three periods”.
- Response: What we actually meant was that C. rostrata was present (as an individual species) at 13 plots. Only 10 plots were clustered as C. rostrata-community clusters, but the species was also present in other clusters. With this sentence we wanted to highlight the contribution of the species to high fluxes, not only the community clusters indicated by the species. We have edited the sentence as follows: “C. rostrata was present at 13 out of 36 plots, and these 13 plots contributed 44–49 % of the measured methane fluxes during the three periods.”
Line 30: northern peatlands are not the main terrestrial source of methane. Wetlands may be, but northern peatlands only account for a small portion of the wetland total. Please update this sentence for clarity. e.g., the global methane budget indicates wetland emissions of 248 Tg CH4/yr https://www.globalcarbonproject.org/methanebudget/ , while the reference cited here estimates emissions of only 38 Tg CH4-C/yr from the whole northern region.
- Response: Corrected.
Line 50: oxidize instead of oxidate
- Response: Corrected.
Line 131: How did you cut them to estimate biomass? Did you include only green tissue, or some depth that you considered active? As am sure you know, it can be difficult to define living moss biomass, so a few more methological details would be useful here.
- Response: We removed the non-living parts based on subjective assessment which was consistent between the samples (the first authors was solely responsible of this). We included only the colorful or leafy parts to represent the aboveground biomass. We have revised the manuscript accordingly.
Lines 137-138: What was the extent of microtopography at the site? Did you consider correcting the WTD measurements for local elevation variation to better represent the actual WTD at the flux measurement collars? Do this effect the interpretation of the role of WTD for accounting for variation in methane flux?
- Response: The experimental site is designed so that majority of the experimental plots (n=30/36) are located in lawn microhabitats that also generally depict our study site. The remaining six plots are located in mud microhabitat. Based on personal observations, hollows and hummocks do occur at the site. Small scale variations in microtopography do occurr but generally the variation is relatively minor. The suggestion to correct the WTD for local elevation could have increased the accuracy of the measurements but, unfortunately, this was not considered at the time of fieldwork. However, the WTD throughout the fen during snow-free season varied only approximately 15 cm (between -9.3 cm to 7 cm) and in plot-scale, the averages of minimum and maximum values were -4.7 cm and 1.8 cm, respectively. Therefore, we think it is reasonable to assume that the WTD did not vary extremely within the 1 m distance, i.e., the distance between the collar and the WTD tube.
Line 164: It’s not clear if this VP to bryophyte ratio is based on biomass or cover. I assume biomass, but it can be made clearer in the text.
- Response: The ratio is calculated as biomass of vascular plants divided by the biomass of bryophytes. We have clarified this in the text.
Line 165: Please provide additional information about the correlation analysis. Was this Pearson correlation? It also isn’t really clear that this was done using average conditions across the sample periods and not instantaneous values (this is my interpretation after reviewing Table B2), which would likely give different results (so just be clear here in the methods what was done). Also, see my comments below about considering the variables together in a multiple regression analysis to assess whether there are interactions between the environmental variables and vegetation clusters.
- Response: The correlation values in Table B2 are Pearson correlations. However, we gained correlation values also from multivariate analyses (DCA and CCA). In the revised manuscript we clearly state from which analyses the correlation values are derived from. In the analyses we used snow-free season averages to represent the overall environmental conditions since parameters were measured at different frequencies. We think that this instantaneous values approach provides a comprehensive view of long-term trends and reduces short-term variability, aligning with our study objectives and the multivariate analysis approach.
Lines 193-194: Are the numbers in brackets averages across all the plots? Please specify in the text and include an estimate of variation (e.g., standard error or standard deviation)
- Response: The numbers are total biomass estimates of all the plots, calculated with the biomasses of the collected samples. We have clarified this in the text but do not find it necessary to add standard deviation for the total biomass, as it was not taken into consideration in any of the analyses. Biomasses and the standard deviations of the collected samples can be found in Table A1.
Lines 199-200: I had a hard time following which species were with which cluster. Maybe add numbers in front of the clusters to clearly separate them.
- Response: The other reviewer also criticized this and therefore, we now use only the name of the strongest indicator species as a name for the cluster and revised the text accordingly throughout the revised manuscript.
Lines 264-266: I totally agree with this statement, but maybe it is also important to highlight here that fluxes were much less variable over the snow-period and did not vary significantly among the identified species clusters. This would suggest that even much lower sampling effort could effectively capture winter fluxes, helping to estimate annual emissions. Vargas and Le 2023, Biogeosciences also supports this conclusion https://doi.org/10.5194/bg-20-15-2023
- Response: As the reviewer is suggesting, our observations are in line with Vargas and Le (2023) that lower sampling efforts during the mid-winter could effectively capture the fluxes, aiding in the estimation of annual emissions. We have added this in the revised manuscript.
Lines 275-280: What about interactions between the soil environment (WTD and temperature) and the plant communities? You didn’t find a significant correlation of CH4 with WTD alone, but it could be that there was a significant correlation in some plant communities and not other, resulting in no significant pattern across the whole dataset. Did you consider multiple regression models that included the plant community information alongside the environmental drivers?
- Response: Please, note our response above where we describe the use of community analysis - both Detrended Correspondence Analysis (DCA) and Canonical Correspondence Analysis (CCA) - followed by Pearson correlation. CCA captures the complex interactions between the species, community clusters and all environmental variables used in the analysis. The plant community compositions we used were 1) All species (including both vascular plants and bryophytes), 2) Vascular plants alone, and 3) Bryophytes alone. In these analyses, pH and WTD were strongly connected to bryophyte clusters but not to vascular plant clusters. Soil temperature at 5 cm depth did not show significant correlation with any community cluster. Therefore, it is correct to state that some communities correlate significantly with edaphic factors and these relations are discussed in detail in sections 4.2 and 4.4.
Lines 289-291: This does not seem to align with the results in Figure 5a where there were no significant differences in CH4 flux among the VP clusters in the snow-period. Are you overstating here?
- Response: In figure 5a we show linear regression, which does not show significant differences between the communities, like the reviewer noted. However, in multivariate analyses, the first ordination axis of vascular plant data correlated significantly with snow season fluxes (r = 0.445 in DCA, -0.402 in CCA). The significant results during the snow season from DCA and CCA but not from linear regression models suggest that the relationship between vegetation and snow season methane fluxes is complex and non-linear and likely involves interactions with other environmental factors that linear regression did not capture. CCA’s ability to handle multivariate data provides a more reliable and accurate representation of the ecological dynamics during the snow season in comparison to linear regression models. This explanation has been added to the revised version.
Line 325: I’m not sure how easy it will be to estimate BM ratio with remote sensing. The community identity or importance of sedges on an areal coverage basis would seem like a variable that is easier to measure with imagery.
- Response: With remote sensing we mean mainly multispectral imagery, which has been shown to detect ratios of vascular plants and mosses reliably (Wolff et al. (2023), https://www.sciencedirect.com/science/article/pii/S1470160X23002820). We have clarified this in the revised version.
Line 341-342: It’s not clear to me where you did this? The correlation analysis looks at each variable individually and does not consider if they interact in predicting flux. Based on Figure 7, there is a wide range of CH4 fluxes when C. rostrata biomass is low, so some other variable must be explaining this. Is it possible that the response of CH4 to the environmental variable differed among the clusters? Did you investigate this (e.g., something like an ANCOVA or multiple regression with the cluster type as a categorical variable that interacts with things like pH, WTD and soil temperature)?
- Response: The values in 4.4 are again from both multivariate analyses and Pearson correlations. The reviewer is right that there is variation in the fluxes that cannot be explained by C. rostrata biomass – an observation that we have discussed shortly in the manuscript. However, it was an interesting idea to explore the difference in responses to environmental variables between the clusters still further. To address the reviewer's concern, we chose to examine the plots which did not have any C. rostrata growing in them (23/36 study plots). We analyzed the environmental variables together with snow-free season methane fluxes only on these plots to see which other factors, in addition to C. rostrata biomass, might show significant relationship with the fluxes. We tested this with linear model and ANOVA in R studio and discovered significant relationships between the fluxes and pH (p < 0.001) and litter coverage (p = 0.01). None of the other environmental or biomass variables resulted in a significant p-value. The Pearson correlation values between snow-free season methane fluxes and pH and litter were -0.66 and -0.29, respectively. These findings have been included in the revised manuscript in results and in discussion.
Line 377: I highly encourage the authors to deposit the full datasets in an open access data repository to ensure availability of the data for future studies/meta analyses.
- Response: We agree on the importance of open access data. However, data used for this study is part of a larger dataset of EcoClimate experiment at Oulanka Research Station. Full data repository is being developed, and all the data will be freely available at some point in the near future. You can familiarize yourself with the experiment via these links:
- https://www.oulu.fi/en/research/research-infrastructures/oulanka-research-station
- https://anaee.fi/facility/ecoclimate/#:~:text=Oulanka%20research%20station%20owns%20and%20operates%20the%20EcoClimate,instrumented%20and%20is%20designed%20to%20run%20for%20decades.
Table B3: Please add units to the columns.
- Response: Corrected.
Citation: https://doi.org/10.5194/egusphere-2025-217-AC2
-
AC2: 'Reply on RC2', Eeva Järvi-Laturi, 10 Apr 2025
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 949 | 96 | 29 | 1,074 | 31 | 49 |
- HTML: 949
- PDF: 96
- XML: 29
- Total: 1,074
- BibTeX: 31
- EndNote: 49
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Cited
Eeva Järvi-Laturi
Teemu Tahvanainen
Eero Koskinen
Efrén López-Blanco
Juho Lämsä
Hannu Marttila
Mikhail Mastepanov
Riku Paavola
Maria Väisänen
Torben Røjle Christensen
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2144 KB) - Metadata XML
General comments:
This manuscript by Järvi-Laturi et al, looks into fine-scale spatial variation in methane flux in a boreal peatland over a full year. By measuring species-specific vascular plant and moss biomass together with chamber-based methane flux measurements, the authors found that increasing sedge biomass, in particular that of Carex rostrata, seem to increase methane emissions significantly especially in the snow-free season and full-year scales. Based on their analyses, vegetation composition and biomass seemed to be a stronger driver of methane fluxes than abiotic environmental variables at this site. The authors attributed these results to both enhanced provision of carbon substrates for methanogens and methane transport from soil to the atmosphere.
This is a study that provides the much-needed data and overview of both wintertime and full-year methane fluxes in a peatland, data that still to this day are quite scarce and thus valuable. As the authors mention in the manuscript, regional and global wetland methane budgets contain large uncertainties, some of which are related to the spatial variation in methane fluxes and vegetation composition. Therefore, this study, which looks at small scale (between-plot) methane flux variation, has potential in adding to our understanding of plant-mediated methane emissions in carbon-rich peatland ecosystems. While I see a lot of value and potential in this work, I recommend a list of improvements (major or minor, depending on how biomass measurements were conducted):
Specific comments:
a) It is unclear how you scaled the vascular plant biomass measurements to the collars. You mention that you counted the number of shoots per species within the collar, but did you use this shoot number to scale the mean biomass of 10 samples to the actual collar (or did you actually take average of 20 sample plants if you ignored the fertile/sterile division? See comment 1 c)? If you did not scale these measurements to the collar, you cannot reliably estimate the collar species biomass, and so I recommend to do this and re-analyze your data and fix the results in 3.2, 3.3, 3.4, 3.5 with appropriately scaled biomass values. If you did do this, please explain this clearly and in more detail in the methods section.
b) It is also unclear how you took the moss biomass samples and scaled them to the collar. How did you determine which species were “most common” and which were not? What was the percentage cover limit (if there was one)? What was the spatial scale that you used for estimating the “most common” species- was it across the whole study site or within individual collar? If it was across the whole study site, I don’t quite understand the logic of taking a biomass sample equaling to 5 % of the collar area (i.e. 33.025 cm2) especially if the collar had an actual percentage cover <5%, in which case you may have overestimated the species biomass in the collar. Or did all of these “most common” species have >5% coverage in all collars?
If you looked at this at the scale of individual collars, did you take a sample that was equal to 5 % of the collar area per species for the five most common species within each collar and for the rest of the species found within the collar, you took a sample over an area equal to 1% area of the collar (i.e. 6.605 cm2)? What did you do if there were less than five species within the collar? Did every collar really have 10 species within them (now the text kind of makes it sound like there were but it doesn’t seem likely to me)? Please specify this in the methods.
And, most importantly, how did you scale the moss biomass samples to the percentage coverage within the individual collars? As with the vascular plant biomass, if no scaling was done, the moss biomass measurements do not represent the actual collar moss biomass and the data should be re-analyzed with appropriate scaling. Please specify this clearly in the methods.
c) How did you determine the locations for vascular plant and moss biomass sampling? Were the soil conditions (e.g. pH, soil moisture) similar to the collar? Did you look at and compare the general species composition in the collar vs the plots where you collected the biomass samples (between-species competition could affect some of the plant trait expression and thus biomass), for example by determining percentage cover? How did you decide which plants and moss patches to pick?
For mosses, did you look at e.g. moss stem density in some way to try to estimate the moss biomass in the collar and in the sampling points more reliably than just percentage coverage (the same percentage coverage can represent very different moss biomasses in different collars due to variation in moss stem density and other structural properties)? If not, the moss biomass estimates may be very uncertain and I would recommend discussing these uncertainties explicitly and in much more detail in the manuscript. Given these uncertainties, I would also recommend not to emphasize the ratio between vascular plants and bryophytes as an important methane flux predictor as much as you have so far in this manuscript, or at least combine it with adequate discussion about its uncertainties.
If you did not estimate the similarity (in terms of abiotic/biotic variables) between the collar and the plot where you collected the representative biomass samples, I would be careful making strong conclusions about collar-specific plant species biomass variation.
d) What do you mean by the “fertile” and “sterile” categories for the plant biomass? In my understanding fertile vs sterile categories are used in the context of evolutionary plant biology and plant reproduction (i.e. fertile vs sterile flowers). Or did you use it to somehow determine whether the species had vegetative culms (e.g. for Carex) from previous year? It is unclear to me how this classification is relevant to the topic of methane flux spatial variability, especially because you do not talk about these classes afterwards. If you used some kind of scaling for the collar biomass (see comment 1 a), did you take use of these fertile/sterile classes in that as well? Please add a clarification for this separation, what the rationale is behind it, and what you mean by the terms.
It is now also unclear whether the species-specific plant biomass is calculated as the mean of n=20 biomass samples (fertile + sterile) per species per plot, or are the species-specific plant biomasses actually still divided into the fertile (mean of n=10 biomass samples) and sterile (mean of n=10 biomass samples) classes. Based on your results, it seems that you took the mean of 20 samples by combining the fertile and sterile samples? Please add a clarification to your methods.
Also, the plot numbers themselves do not really give any valuable information for the reader. If you want to show all the plot fluxes separately here, the plot numbers could be replaced with something simpler, such as 1-36, to improve the readability of this figure (the current numbering adds more complexity for the reader who is not familiar with your study site).
On the other hand, if you want to keep the coloring, could you do it based on e.g. vegetation composition grouping (for example based on your vegetation clusters)? If you do this, I would also recommend changing the red color of the soil temperature to black, because the red and green are difficult to separate visually for some readers with color-blindness.
The axis texts are also a bit small so please increase the font size.
a) The correlation between peat depth and plant biomass makes sense in the biological sense that, when there is more peat, there is also more space for roots especially for more deeply-rooting vascular plants. Since you did not find significant correlations between peat depth and methane fluxes, I would be careful drawing strong conclusions about the influence of peat depth on methane fluxes via vegetation (but see my next point).
b) On the other hand, it is also possible that in the presence of deeply-rooted aerenchymatous vegetation, such as C. rostrata, the roots may provide labile carbon substrates in deep peat where methanogenesis increases despite the dominance of recalcitrant peat (i.e. indirect influences of peat depth on methane fluxes). The release of labile carbon compounds via root exudation could also trigger microbial carbon priming (see e.g. Waldo et al 2019: https://doi.org/10.1007/s10533-019-00600-6). However, be careful about your interpretations about the wintertime vegetation influences based on your data (see previous comment about wintertime fluxes), and keep in mind that the direct relationship between peat depth and methane flux was still nonsignificant.
Technical comments: