the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 20 Mar 2025
Organic soil carbon balance in drained and undrained hemiboreal forests
Abstract. Drainage of organic soils is associated with increasing soil carbon (C) efflux, which is typically linked to losses in soil C stock. In previous studies, drained organic forest soils have been reported as both C sinks and sources depending on, e.g., soil nutrient and moisture regime. However, most of the earlier research was done in boreal region, and both the magnitude of C efflux and the impact of soil moisture regime on soil C stock are likely to vary across different climatic conditions and ecosystems, depending further on vegetation. A two-year study was conducted in hemiboreal forest stands with nutrient-rich organic soil (including current and former peatlands) and a range of dominant tree species (black alder, birch, Norway spruce, Scots pine) in the Baltic states (Estonia, Latvia, Lithuania). In this study, we analysed the C balance of organic soil in drained (19) and undrained (7) sites. To assess the C balance, soil respiration was measured along with evaluation of C influx into the soil through aboveground and belowground litter. To characterize the sites and factors influencing the C fluxes, we analysed soil temperature, water table level, physical and chemical parameters of soil and soil water. On average, no changes in soil C stocks (0.45±0.50 t C ha⁻¹ year⁻¹) were observed in drained sites dominated by black alder, birch, or Norway spruce, while drained Scots pine sites showed soil C removals with a mean rate of 2.77±0.36 t C ha⁻¹ year⁻¹. In undrained birch- and spruce-dominated sites, soil functioned as mean C sink at 1.33±0.72 t C ha⁻¹ year⁻¹, while the undrained black alder stands showed an uncertain C balance of 1.12±2.47 t C ha⁻¹ year⁻¹. The variability in C balances were influenced by the nutrient-rich soil exhibiting a wide range of nutrient conditions and organic matter quality. Thus, indicating that soil macronutrient concentrations and pH can determine whether the soil functions as a C source or sink.
-
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
(1834 KB)
-
Supplement
(1621 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1834 KB) - Metadata XML
-
Supplement
(1621 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-1032', Anonymous Referee #1, 09 Apr 2025
This manuscript addresses the issue of the extent to which peatland soils change their C stock in response to drainage. This is a somewhat contentious issue because of its importance in developing reliable estimates of the C budget of entities, such as political states, in countries where peatland drainage, for forestry and other uses form an important part of the estimation, such as the use of Emission Factors. Contentious because the literature shows very variable results on C balance because of the variability in environment (e.g. climate), extent of drainage (lowering of water table), organic soil thickness, properties of the soil, duration of drainage, vegetation cover and methods employed. The value of the manuscript is in its adoption of similar methods across a wide range of sites, driven by the variability above, involving an area in which peatland drainage for forestry is significant, and including three countries.
The methods involve a variety of processes leading to C import and export of the soil, so a large number of measurements, and estimations, need to be made. Here, the same methods were applied (more or less) across 26 sites, of which 7 were undrained and 19 were drained and over 2 years. This follows a literature review in 2023 (Jauhianen et al. Biogeosciences 20, 4819–4839; the Reference is incomplete on p. 21) for greenhouse gas fluxes in boreal and cool temperate regions with the sites in this manuscript being in the hemi-boreal region of Estonia, Latvia and Lithuania. The distinction of hemi-boreal is a bit confusing with other terms, such as cool temperate, Cool Temperate Mist climate region etc. and could be clarified (says on line 50 ‘between the temperate and boreal zones’). It appears that there was no attempt (or success) to include a comparison of drained and undrained sites, based on the latitude and longitude data in Table S1, though there appear to be two pairs in Latvia (Fig. 1). Please clarify.
The results are combined into a series of graphs and tables (many in Supplementary Information) for individual measurements which are combined into estimates of the annual soil C balance combination which forms the focus of the initial part of the Discussion. The authors recognize there is great variability with some unexpected patterns emerging, though the limited replication of site types means that categorization of type is unwarranted (lines 536 and following). There is also a recognition that the period of drainage at the sites is not clearly known, but could be many decades, so that the early effect of drainage may be muted in the results that were obtained. There is a detailed discussion of the merits and limitations of the methods employed to contribute to the estimate of the soil C balance.
The conclusions of the study are in an assessment of the overall C balance of several forested peatlands, drained or undrained, within a fairly narrow range and that a variety of properties, ranging from environmental to tree type, can influence the results. In the C balance, the expectation is that this measure (C tons balance etc.) converts into CO2. This maybe the case, but what about other C forms in the C cycle? Methane would play an insignificant role in the C balance for most of the sites, given the low water table in most sites, including the undrained ones: probably up to 0.05 t C/ha/yr in the wetter sites and maybe CH4 uptake in the drier sites. Loss of dissolved organic carbon (DOC) would result from leaching of the soil, and may account for up to 0.10 t C/ha/yr additional loss, but also small to most of the soil C balance estimates that have been made.
The manuscript started with a comment on the use of Emission Factors by the IPCC and states, though no EF values were given. If the objective of the study, beyond the science of the forested systems, was to contribute to a better estimate of the variability and magnitude of EF, it would be useful to see how the authors think these study would contribute to that objective. What ‘better’ estimate of EF could have been made using the results assembled in the manuscript, with a lot of good, hard work over two years and standardized methods, compared to the ‘guesswork’ of the past?
Specific comments:
24 It seems that the estimated changes in C do not involve + and – signs. Such as soil C removal from drained Scots pine sites was 2.77 units while C sink occurred in undrained black alder sites there was an average sink of 1.33 units. Throughout the manuscript could ‘loss’ estimates be given a negative sign (e.g. -2.77 +/- 0.36 units) and ‘gain’ estimates be given a positive sign (1.33 +/- 0.72 units. The graphs showing the ‘C balance’ (Fig 9 and 10) include negative values, please be consistent. The notation used in Figures also varies: for example Fig 8 has ‘Carbon flux’ and 9 and 10 have ‘C balance’ with the same units and meaning. Please standardize.
15 the boreal region
35 why not use ‘faster’ rather than ‘higher’ to describe a rate?
46 One of the studies was on a drained peatland used for horticultural crop production, so is not representative of the types used in the EF estimates.
83 Jauhianen et al. was incompletely cited in the References.
136 data were (plural)
164 how ‘small’ was insignificant?
460 Basal area had the strongest correlation with C balance, yet in Fig. 10, the R2 of 0.14 was the smallest in the 6 graphs, several with p values < 0.01. Please check. It would be good to include the slope of the regression to indicate how much change in C balance was created by a change in the independent variable. For example, a reduction in pH from 6 to 2.5 (!) would result in a C balance (gain) of about 3 units. An increase in bGV of 0.1 to 3.1 units would result in a C balance of -1 to 4 units.
Citation: https://doi.org/10.5194/egusphere-2025-1032-RC1 - AC1: 'Reply on RC1', Aldis Butlers, 16 Jun 2025
-
RC2: 'Comment on egusphere-2025-1032', Jens-Arne Subke, 08 May 2025
General comments:
The manuscript looks improved from a previous version. Deriving C balances of peatland under contrasting drainage and vegetation is challenging, and requires a significant number of sites to enable a meaningful separation of potentially confounding factors. The presented study has a somewhat imbalanced representation of tree species, drainage and soil types so that results have to be considered with some caution.
What the study does provide is a general overview of peatland C balances across the Baltic states. There are remaining issues of some confounding elements (pine forests on undrained sites are absent, yet pine sites drive several relationships discussed – see specific comments). Overall, however, I think that the discussion contains sufficient detail and awareness of limitations. The amount of work that went into this study is considerable, and there is useful data that is worth publishing to broaden our understanding of peatland C balances in this region.
Emission Factors form a significant part of the rationale of the study, taking up several paragraphs in the introduction. This is not reflected in the discussion which focusses much more on fundamental understanding of C balances, not emissions reporting. It would be better to address this by setting the scope of the study and motivation for study differently in the introduction.
Soil pH is cited throughout the manuscript as an important correlator of C balance. It is generally presented as a causal link of lower pH and C stocks. However, the cause of pH differences are not considered meaningfully, where conifer plantations are likely to have reduced pH due to acidic litterfall. The correlation between C stocks and pH are hence linked to vegetation more than pH being an independent driver of C stocks. This should be much clearer in the discussion (e.g. 520-525).
There is also an apparent mismatch between opening arguments of conducting this study across the three Baltic states, as they share an ecoclimatic region. I agree, but found the partial focus in the analysis to separate results by country unhelpful. This is strongly biased by the distribution of vegetation and drainage across the study, resulting in limited insights, The presentation of data can be streamlined significantly by removing the “country” aspect throughout.
There are a number of specific issues that I address below.
Specific comments:
26-28: The past two sentence should be merged. What you say seems to contradict the previous statements where source and sink behaviours are presented as functions of drainage and tree species. Make it clear how different parameters influence carbon balances without causing contradictions.
105: Why does the analysis distinguish sites by country? The argument presented is that this is one ecoclimatic zone with site replication across the three Baltic states. From that rationale, national boundaries are arbitrary and the analysis should focus on environmental drivers of biogeochemical patterns. This focusses analysis and removes part of detailed results/discussion.
300 (Fig. 2): I am worried by the confounding effect of drainage and tree species. Table 1 indicates that undrained spruce forests were sampled, but this figure shows only deciduous forest on undrained sites. Looking at pH in particular, the observed difference ascribed to drainage status is caused by undrained sites not including coniferous forest with more acidic litter input. Comparing birch and alder forests only, there is no evident difference by drainage. Looking at Fig. 10, the pH distribution by species and drainage seem to be different to what is shown in Fig. 2 (e.g., drained/Birch has values below 4 in Fig 10, not in Fig. 10; undrained Spruce pH values shown in Fig. 10, not Fig. 2). This has to be clarified, as the discussion has to take account of these patterns and potential confounding influences.
398: This is not evident from Table 3. aGV is c. 39% of total GV (sum of aGV and bGV) in drained, and c. 49% of total in undrained.
Citation: https://doi.org/10.5194/egusphere-2025-1032-RC2 -
AC2: 'Reply on RC2', Aldis Butlers, 16 Jun 2025
Thank you for your comments and the generally positive assessment of the revised manuscript. We appreciate your thorough cross-checking and the suggested refinements, which we will implement. Please find our responses to the questions and observations attached.
-
AC2: 'Reply on RC2', Aldis Butlers, 16 Jun 2025
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-1032', Anonymous Referee #1, 09 Apr 2025
This manuscript addresses the issue of the extent to which peatland soils change their C stock in response to drainage. This is a somewhat contentious issue because of its importance in developing reliable estimates of the C budget of entities, such as political states, in countries where peatland drainage, for forestry and other uses form an important part of the estimation, such as the use of Emission Factors. Contentious because the literature shows very variable results on C balance because of the variability in environment (e.g. climate), extent of drainage (lowering of water table), organic soil thickness, properties of the soil, duration of drainage, vegetation cover and methods employed. The value of the manuscript is in its adoption of similar methods across a wide range of sites, driven by the variability above, involving an area in which peatland drainage for forestry is significant, and including three countries.
The methods involve a variety of processes leading to C import and export of the soil, so a large number of measurements, and estimations, need to be made. Here, the same methods were applied (more or less) across 26 sites, of which 7 were undrained and 19 were drained and over 2 years. This follows a literature review in 2023 (Jauhianen et al. Biogeosciences 20, 4819–4839; the Reference is incomplete on p. 21) for greenhouse gas fluxes in boreal and cool temperate regions with the sites in this manuscript being in the hemi-boreal region of Estonia, Latvia and Lithuania. The distinction of hemi-boreal is a bit confusing with other terms, such as cool temperate, Cool Temperate Mist climate region etc. and could be clarified (says on line 50 ‘between the temperate and boreal zones’). It appears that there was no attempt (or success) to include a comparison of drained and undrained sites, based on the latitude and longitude data in Table S1, though there appear to be two pairs in Latvia (Fig. 1). Please clarify.
The results are combined into a series of graphs and tables (many in Supplementary Information) for individual measurements which are combined into estimates of the annual soil C balance combination which forms the focus of the initial part of the Discussion. The authors recognize there is great variability with some unexpected patterns emerging, though the limited replication of site types means that categorization of type is unwarranted (lines 536 and following). There is also a recognition that the period of drainage at the sites is not clearly known, but could be many decades, so that the early effect of drainage may be muted in the results that were obtained. There is a detailed discussion of the merits and limitations of the methods employed to contribute to the estimate of the soil C balance.
The conclusions of the study are in an assessment of the overall C balance of several forested peatlands, drained or undrained, within a fairly narrow range and that a variety of properties, ranging from environmental to tree type, can influence the results. In the C balance, the expectation is that this measure (C tons balance etc.) converts into CO2. This maybe the case, but what about other C forms in the C cycle? Methane would play an insignificant role in the C balance for most of the sites, given the low water table in most sites, including the undrained ones: probably up to 0.05 t C/ha/yr in the wetter sites and maybe CH4 uptake in the drier sites. Loss of dissolved organic carbon (DOC) would result from leaching of the soil, and may account for up to 0.10 t C/ha/yr additional loss, but also small to most of the soil C balance estimates that have been made.
The manuscript started with a comment on the use of Emission Factors by the IPCC and states, though no EF values were given. If the objective of the study, beyond the science of the forested systems, was to contribute to a better estimate of the variability and magnitude of EF, it would be useful to see how the authors think these study would contribute to that objective. What ‘better’ estimate of EF could have been made using the results assembled in the manuscript, with a lot of good, hard work over two years and standardized methods, compared to the ‘guesswork’ of the past?
Specific comments:
24 It seems that the estimated changes in C do not involve + and – signs. Such as soil C removal from drained Scots pine sites was 2.77 units while C sink occurred in undrained black alder sites there was an average sink of 1.33 units. Throughout the manuscript could ‘loss’ estimates be given a negative sign (e.g. -2.77 +/- 0.36 units) and ‘gain’ estimates be given a positive sign (1.33 +/- 0.72 units. The graphs showing the ‘C balance’ (Fig 9 and 10) include negative values, please be consistent. The notation used in Figures also varies: for example Fig 8 has ‘Carbon flux’ and 9 and 10 have ‘C balance’ with the same units and meaning. Please standardize.
15 the boreal region
35 why not use ‘faster’ rather than ‘higher’ to describe a rate?
46 One of the studies was on a drained peatland used for horticultural crop production, so is not representative of the types used in the EF estimates.
83 Jauhianen et al. was incompletely cited in the References.
136 data were (plural)
164 how ‘small’ was insignificant?
460 Basal area had the strongest correlation with C balance, yet in Fig. 10, the R2 of 0.14 was the smallest in the 6 graphs, several with p values < 0.01. Please check. It would be good to include the slope of the regression to indicate how much change in C balance was created by a change in the independent variable. For example, a reduction in pH from 6 to 2.5 (!) would result in a C balance (gain) of about 3 units. An increase in bGV of 0.1 to 3.1 units would result in a C balance of -1 to 4 units.
Citation: https://doi.org/10.5194/egusphere-2025-1032-RC1 - AC1: 'Reply on RC1', Aldis Butlers, 16 Jun 2025
-
RC2: 'Comment on egusphere-2025-1032', Jens-Arne Subke, 08 May 2025
General comments:
The manuscript looks improved from a previous version. Deriving C balances of peatland under contrasting drainage and vegetation is challenging, and requires a significant number of sites to enable a meaningful separation of potentially confounding factors. The presented study has a somewhat imbalanced representation of tree species, drainage and soil types so that results have to be considered with some caution.
What the study does provide is a general overview of peatland C balances across the Baltic states. There are remaining issues of some confounding elements (pine forests on undrained sites are absent, yet pine sites drive several relationships discussed – see specific comments). Overall, however, I think that the discussion contains sufficient detail and awareness of limitations. The amount of work that went into this study is considerable, and there is useful data that is worth publishing to broaden our understanding of peatland C balances in this region.
Emission Factors form a significant part of the rationale of the study, taking up several paragraphs in the introduction. This is not reflected in the discussion which focusses much more on fundamental understanding of C balances, not emissions reporting. It would be better to address this by setting the scope of the study and motivation for study differently in the introduction.
Soil pH is cited throughout the manuscript as an important correlator of C balance. It is generally presented as a causal link of lower pH and C stocks. However, the cause of pH differences are not considered meaningfully, where conifer plantations are likely to have reduced pH due to acidic litterfall. The correlation between C stocks and pH are hence linked to vegetation more than pH being an independent driver of C stocks. This should be much clearer in the discussion (e.g. 520-525).
There is also an apparent mismatch between opening arguments of conducting this study across the three Baltic states, as they share an ecoclimatic region. I agree, but found the partial focus in the analysis to separate results by country unhelpful. This is strongly biased by the distribution of vegetation and drainage across the study, resulting in limited insights, The presentation of data can be streamlined significantly by removing the “country” aspect throughout.
There are a number of specific issues that I address below.
Specific comments:
26-28: The past two sentence should be merged. What you say seems to contradict the previous statements where source and sink behaviours are presented as functions of drainage and tree species. Make it clear how different parameters influence carbon balances without causing contradictions.
105: Why does the analysis distinguish sites by country? The argument presented is that this is one ecoclimatic zone with site replication across the three Baltic states. From that rationale, national boundaries are arbitrary and the analysis should focus on environmental drivers of biogeochemical patterns. This focusses analysis and removes part of detailed results/discussion.
300 (Fig. 2): I am worried by the confounding effect of drainage and tree species. Table 1 indicates that undrained spruce forests were sampled, but this figure shows only deciduous forest on undrained sites. Looking at pH in particular, the observed difference ascribed to drainage status is caused by undrained sites not including coniferous forest with more acidic litter input. Comparing birch and alder forests only, there is no evident difference by drainage. Looking at Fig. 10, the pH distribution by species and drainage seem to be different to what is shown in Fig. 2 (e.g., drained/Birch has values below 4 in Fig 10, not in Fig. 10; undrained Spruce pH values shown in Fig. 10, not Fig. 2). This has to be clarified, as the discussion has to take account of these patterns and potential confounding influences.
398: This is not evident from Table 3. aGV is c. 39% of total GV (sum of aGV and bGV) in drained, and c. 49% of total in undrained.
Citation: https://doi.org/10.5194/egusphere-2025-1032-RC2 -
AC2: 'Reply on RC2', Aldis Butlers, 16 Jun 2025
Thank you for your comments and the generally positive assessment of the revised manuscript. We appreciate your thorough cross-checking and the suggested refinements, which we will implement. Please find our responses to the questions and observations attached.
-
AC2: 'Reply on RC2', Aldis Butlers, 16 Jun 2025
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 541 | 111 | 18 | 670 | 44 | 19 | 25 |
- HTML: 541
- PDF: 111
- XML: 18
- Total: 670
- Supplement: 44
- BibTeX: 19
- EndNote: 25
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Raija Laiho
Andis Lazdiņš
Thomas Schindler
Kaido Soosaar
Jyrki Jauhiainen
Arta Bārdule
Muhammad Kamil-Sardar
Ieva Līcīte
Valters Samariks
Andreas Haberl
Hanna Vahter
Dovilė Čiuldienė
Jani Anttila
Kęstutis Armolaitis
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1834 KB) - Metadata XML
-
Supplement
(1621 KB) - BibTeX
- EndNote
- Final revised paper