the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 28 Feb 2024
Carbon degradation and mobilisation potentials of thawing permafrost peatlands in Northern Norway
Abstract. Permafrost soils are undergoing rapid thawing due to climate change and global warming. Permafrost peatlands are especially vulnerable since they are located near the southern margin of the permafrost domain in the discontinuous and sporadic permafrost zones. They store large quantities of carbon (C) which, upon thawing, are decomposed and released as carbon dioxide (CO2), methane (CH4) or dissolved organic carbon (DOC). This study compares carbon degradation in three permafrost peatland ecosystems in Finnmark, Norway, which represent a well-documented chronosequence of permafrost formation. Peat cores from active layer, transition zone and permafrost zone were thawed under controlled conditions and incubated for up until 350 days under initially-oxic or anoxic conditions while measuring CO2, CH4 and DOC production. Carbon degradation varied among the three peat plateaus but showed a similar trend over depth with largest CO2 production rates in the top of the active layer and in the permafrost. Despite marked differences in peat chemistry, post-thaw CO2 losses from permafrost peat throughout the first 350 days in the presence of oxygen reached 67–125 % of those observed from the top of the active layer. CH4 production was only measured after a prolonged anoxic lag phase in samples from transition zone and permafrost, but not in active layer samples. CH4 production was largest in thermokarst peat sampled next to decaying peat plateaus. DOC production by active layer samples throughout 350 days incubation exceeded gaseous C loss (up to 23-fold anoxically), whereas little DOC production or uptake was observed for permafrost peat after thawing. Taken together, permafrost peat in decaying Norwegian peat plateaus degrades at rates similar to active layer peat, while highest CH4 production can be expected after inundation of thawed permafrost material in thermokarst ponds.
- Preprint
(1203 KB) - Metadata XML
-
Supplement
(1009 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2024-562', Anonymous Referee #1, 26 Mar 2024
Summary:
Kjær et al. presents an incubation study using soils from permafrost peatlands in northern Norway. The authors collected material from active to mineral layer in a permafrost habitat, and a comparison core from thawed thermokarst soil. The authors used aerobic and anaerobic headspaces in the incubations, at different depths from an inland-to-ocean gradient of permafrost peatland habitats. The first 20 days were sampled multiple times a day and give insights to the microbial activity directly after permafrost material thaws. Each depth group was divided in four with two vials being kept at field moisture, and the other two vials having water added for a slurry mixture. The replicates were further divided so that one vial begins in an aerobic headspace and one vial remains anaerobic for the duration of the incubation. The authors provide rates and cumulative values for the O2, CO2, N2, CH4 and N2O and geochemical data for the soil cores. Stable isotope signature analysis and dissolved organic carbon samples were also taken to supplement the gas production dataset. The main findings of the study were that the three sites had similar patterns by depth in the various parameters measured. Notably, the active layer samples were highly productive in DOC production, a pattern not observed in the post-thaw samples.
I am enthusiastic about the content of the study, and especially of the high resolution gas measurements. The paper is pertinent and could be of interest to the scientific community with its high resolution gas measurements, and strategic selection of interesting study sites. However, multiple major comments should be considered.
MAJOR COMMENTS:
The authors clearly state their objectives with the paper, though a hypothesis was not provided. This was reflected further in the paper when the results were more reports of the data and did not prove / disprove what the authors had hypothesized how the various depths would respond to the treatment groups. Some measurements (ie. stable isotope signatures) are reported in the results but not further integrated into the rest of the paper. Further analysis could also be done with the cumulative gas production between coring sites. The novelty of this study lies with its high-resolution measurements, especially the gas production from the first ~20 days; the community could benefit from the quantification of this fast carbon pool that is largely not captured with ex-situ incubations to date.
I am not entirely convinced that the paper in its current form presents a robust statistical analysis to meet the authors objectives. The largest issue being that there were many treatment groups but no technical replicates. A replicate of n=1 is statistically weak, though the high resolution of the measurements provides some compensation for this. Overall, I think that the site / depth replicates could be binned differently to explore stronger hypotheses that are developed in response to the comment above. T-test’s were performed to determine difference between the geochemical parameters of the sites but further analysis could be done on the gas fluxes than determining if the production rate is exponential or not.
Important parameters such as the general vegetation of field sites, soil type, and soil bulk density were not reported. These parameters are essential for the comparison of these results to similar experiments, and usability for those looking to use the data for reviews or inputs for earth system models.
The authors reference the cores being collected on an inland-to-sea gradient. This was mentioned in the introduction and site description but not further discussed. The natural gradient a transect introduces a range of variables (ie. climate, vegetation, etc.) that likely influence the data and could give greater relevance to the patterns seen in this study. It would be of value for the authors to shape their discussion of what research questions they were looking to answer with this gradient when they selected their soil coring locations.
MINOR COMMENTS:
Line 34: Some grammatical errors, some rephrasing for clarity needed.
Line 43: Here I would suggest expanding on what has been introduced here. Anaerobic respiration in peatlands is just starting to be understood, and this experiment's high resolution measurements afford a unique opportunity to capture processes that happen right as the material starts to thaw from its frozen state.
Line 45: Although it could be considered “common knowledge”, I think it's worth briefly elaborating on why CO2, CH4 and carbon cycles are important on a global scale in the introduction.
Line 50: This is the first mention of the sea / inland dynamic. A paragraph on the variables this natural gradient introduces would improve the paper and better emphasize the importance of the data. Clearly stating which sea or body of water is referenced would make for a more easily readable paper.
Line 51-52: A brief discussion of why active layer / permafrost layer vary in oxygen availability and potential for degradation would improve paper. Suggested readings, citations.
Clymo, R. S., & Hayward, P. M. (1982). The Ecology of Sphagnum. In A. J. E. Smith (Ed.), Bryophyte Ecology (pp. 229–289). Springer Netherlands. https://doi.org/10.1007/978-94-009-5891-3_8
Hodgkins, S. B., Tfaily, M. M., McCalley, C. K., Logan, T. A., Crill, P. M., Saleska, S. R., Rich, V. I., & Chanton, J. P. (2014). Changes in peat chemistry associated with permafrost thaw increase greenhouse gas production. Proceedings of the National Academy of Sciences, 111(16), 5819–5824. https://doi.org/10.1073/pnas.1314641111
Schädel, C., Beem-Miller, J., Aziz Rad, M., Crow, S. E., Hicks Pries, C. E., Ernakovich, J., Hoyt, A. M., Plante, A., Stoner, S., Treat, C. C., & Sierra, C. A. (2020). Decomposability of soil organic matter over time: The Soil Incubation Database (SIDb, version 1.0) and guidance for incubation procedures. Earth System Science Data, 1511–1524. https://doi.org/10.5194/essd-12-1511-2020
Line 56: A definition and introduction of thermokarst earlier in the introduction would improve the flow and better emphasize the importance of the data
Line 60: Consider rephrasing for clarity
Line 80: While the site history on a geologic scale is informative, a description of the modern vegetation at the time of sampling, soil types at each depth, and a broader site introduction is necessary.
Line 104: How were the layers decided? Visual inspection of the soil horizons, or was there a metric used to delineate a change in the soil core? Example photos of vegetation or soil of sites would be useful in the supplement, if possible.
Line 179: A brief description of how the authors went from the GC output to their reported units of umol CO2/CH4 g-1 dw-1 and µg g dw-1 225 d-1 using the cited Molstad et al method would increase the reproducibility of the values used in this study and readability of paper.
Line 190: For the statistics, it is unclear if the peat characteristics applies to the geochemical analysis or the gas chromatography measurements. If the t-test, etc was performed in a statistics program (ie. R, python, etc) it should be stated and the package credited.
Line 216: Whether or not the gas production is exponential / linear at different timepoints has some value, but the larger question is if the rate of production is different between the treatment groups (slurried or field moisture, aerobic or anaerobic, the inland-to-sea gradient between field sites, and depth groups). Here the study's largest weakness of having no replicates becomes an issue. Its difficult to make assertions about these treatments inducing higher or lower rates of GHG production with one technical replicate per treatment
Line 254: Were the incubations “constantly agitated”? The methods state that the slurry treatment groups were mixed for an hour to disperse the peat, then the peat settled
Line 257: Samples completely under the water line are mostly in an anaerobic environment. I would suspect that the inhibited rate of O2 diffusing into the water would be the limitation, and not the reduced headspace volume.
Line 265: Table could be moved to Supplement, and values of interest simply discussed in the results.
Line 275: It is likely not that the thawing permafrost was necessarily “stimulated” but rather that the authors are observing a lag phase common in ex-situ incubation studies with permafrost soils.
Knoblauch, C., Beer, C., Liebner, S., Grigoriev, M. N., & Pfeiffer, E.-M. (2018). Methane production as key to the greenhouse gas budget of thawing permafrost. Nature Climate Change (4), 309–312. https://doi.org/10.1038/s41558-018-0095-z
Line 279: Same comment as Line 254
Line 305: Same comment as Line 265
Line 314-317: Needs Figure references and more specific language
Line 324-328: This paragraph would benefit from references to Figures and/or specific references to the data.
Line 339: The authors imply that the CO2 production from their study is higher than that of another permafrost site using a similar methodology. While this is useful, I would recommend a reassessment of this interpretation. Incubations are useful in their ability to isolate individual ecosystem-scale controls on GHG production; however, they are limited in direct comparisons of the production values as (the authors state in Line 350) there are many discrepancies between incubation studies.
Line 373: Bringing in the stable isotope results and expanding on what the significance of the different patterns of depletion / enrichment in each site and by depth here would add value to the discussion
Line 381: I suggest moving Figure S4 into the main body of the manuscript and discussing this dataset more. I would also suggest expanding on how these were analyzed in the methods.
Line 393: For this study (Panneer Salvem et al 2017) and the others used in this discussion, some explanation is needed as to why the authors chose these particular studies to compare their results so closely to the comparison papers results.
Line 405: There has been significant focus on exponential / non-exponential production rates, but no reason given as to why this is significant to answering a hypothesis or adds value to the study.
Line 408: It is unclear why this high-altitude / high-latitude comparison is being made
Line 419-420: Same comment as Line 324-328
Line 443: For the conclusion of this paper, the authors should consider framing the study more robustly into the significance of these high resolution measurements into the current state of knowledge on permafrost C dynamics.
Line 453: Perhaps an earlier mention of these elements (DOC runoff and down-stream ecosystems) if it is the main recommendation of the study. For example, expanding on the statement in Line 400 would achieve this.
Citation: https://doi.org/10.5194/egusphere-2024-562-RC1 -
CC1: 'Comment on egusphere-2024-562', Claire C. Treat, 05 Apr 2024
Kjaer et al. use a high depth-resolution of core sections from three separate palsas in northern Fennoscandia to look at potential C losses with permafrost thaw. They measure peat properties, potential CO2 and CH4 production under aerobic and anoxic conditions as well as the DOC released during the incubation. The strengths and unique features of this experiment include the high resolution incubation data, the combination of both gaseous and dissolved pathways of C loss, and the inclusion of a thawed replicate for comparison and the use of three individual sites. Additionally, the paper is relatively easy to read although the discussion section needs reference to the figures in the main text. Most importantly, the results support the conclusions.
There have been several incubation experiments previously of permafrost peats; these are appropriately referred to in the study and the discussion. What I’m missing is what this study adds to the earlier ones. I’m also missing what from this study can be generalized because the interpretations of the results are very specific to each site and also quite qualitative. The conclusions section redeems this but this is the only place where the findings are presented so clearly. A revision of the paper, particularly the results and discussion, keeping in mind the key points from the conclusion is warranted in order to streamline.
In the overall justification of the study, the explanation (hypotheses) of why the differences in C production (CO2, CH4, DOC) might occur is missing, beyond stating that post-thaw degradation kinetics might depend on peat quality and formation history (line 49). I think there is an earlier study that might help with the interpretation here. This study is using a chronosequence of timing of permafrost formation as well as variable timing of peatland initiation: Treat et al. (2021) explores the influence of the “residence time in the active layer” on the potential carbon losses using modeling. I think that would be a really useful theoretical framework to adopt here for the development of the hypotheses, data analysis, and interpretation of the results if the goal is to link peat quality and formation history with potential C production, which I think is appropriate and interesting.
There are a few further challenges in this study, both with the experimental design and with the interpretation. The main issue with the experimental design is the lack of replicates, at least in the current analysis. The analysis uses different (categorical) depths and use different treatments, one replicate of each treatment per depth. The replicates are the sites. However, sites aren’t used as replicates; instead, they are discussed individually but always quantitatively. The main problem that I have with this approach is that it really limits the interpretation and application of these results beyond these specific study site (as well as not allowing for statistics). The authors could explore different methods of binning which could increase the number of replicates and might make the trends clearer and potentially allow statistical analyses. Some more creative analyses may help in linking some of the various measures, such as PCAs or NMDS.
Additional comments
Table 1 could include info on timing of peat formation and permafrost formation, particularly since it’s used as a justification for the study design.
Tables 2 and 3 are not very intuitive. I think this point is important, is there a more effective way to get this across? Maybe as additional lines in Figure 6 and 7?
Not everything in the thesis needs to be included in the paper as the thesis counts as a reference, for example the slurry vs loose peat experiment does not show up in the main conclusion and could be referenced in the methods section.
This study does a nice job summarizing earlier aerobic:anaerobic CO2 produciton ratios in section 4.1; perhaps a table might be a nice way to summarize this.
389-390: I think it’s quite unlikely that oxic samples would produce methane as long as the headspace remained oxic.
Conclusions: currently doesn’t discuss the aerobic:anaerobic production ratios that are a major point in the discussion
Citation: https://doi.org/10.5194/egusphere-2024-562-CC1 -
RC2: 'Comment on egusphere-2024-562', Anonymous Referee #2, 12 Apr 2024
This manuscript presents interesting findings regarding carbon degradation in permafrost peat cores, stratified by depth and sampled from various regions. The key findings indicate comparable CO2 losses between the active layer and permafrost/transition zone, with DOC production surpassing gaseous C losses, and CH4 emissions occurring following inundation of thawed peat material in thermokarst ponds. These results offer significant and new insights into carbon dynamics in permafrost environments.
The manuscript is well-written, logically structured, and engaging to read. One issue is is the lack of technical replication which have been thoroughly addressed already by other reviewers. Another one is the presentation of the data as gram per incubation duration. In this manner, the data cannot be readily compared to other studies or generalized. It would be advantageous to present the data so that they can be used fo modeling carbon dynamics in different permafrost soils. Therefore, I recommend reporting the data on a gram (of mol) of dry soil or soil carbon basis (e.g., mg CO2-C g-1 d-1 or mg CO2-C g-1 C-1). Ideally, decay rates should be computed by fitting them into a first-order models with one or more pools, as it has been practiced by others. Please look at the paper by Schädel et al. (2020) Earth Sys. Sci. Data, 12, on how to report data from incubation studies to include them into a larger database on decomposability of SOC.
Regarding specific comments, Line 264 starting with "CO2 production in thermokarst cores....". is not supported by the data (referencing Table 2), and later the discussion highlights decreasing CO2 fluxes in thermokarst cores, likely attributed to water inundation. By the way, can you also provide soil moisture data from the cores, to substantiate that conclusion? It is important to get information about future CO2 fluxes in these ecosystems, and the MS makes a significant contribution here.
In Figure 8, it would be beneficial to clearly indicate that the bars are stacked to facilitate comparison between the two components (CO2/DOC). I would maybe even present them next to each other, to make the comparison easier.
Finally, while the data and discussion on dissolved organic carbon (DOC) and lateral runoff are really interesting, the significance of extracting DOC before and after one year of incubation warrants consideration. I am wondering what a DOC extract before and after one year of incubation really tells us. DOC is permanently produced and consumed, the net change must be very small as compared to the gross changes. Which makes it even more difficult to believe that cumulative CO2 emissions are lower than net DOC production in these soils, since DOC is very labile and thus heavily mineralized. Can it be that this enrichment in DOC is due to die-off of microbes after such a long incubation? I encourage the authors to double-check their calculations and report the proportion of carbon lost as CO2/CH4-C from the initial carbon pool after 350 days, along with the proportion that would have been produced as DOC and is thus susceptible to leaching (in percentage), together with some critical evaluation of the DOC data presented.
Citation: https://doi.org/10.5194/egusphere-2024-562-RC2
Data sets
Permafrost peatlands: gas emission and elemental analysis S. T. Kjær et al. https://doi.org/10.5281/zenodo.10696561
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 194 | 58 | 15 | 267 | 31 | 6 | 10 |
- HTML: 194
- PDF: 58
- XML: 15
- Total: 267
- Supplement: 31
- BibTeX: 6
- EndNote: 10
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1