the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 22 Jun 2026
Implementation of regenerative ditch borders in Dutch peat meadows: effects on soil CO2 fluxes and potential carbon trade-offs
Abstract. Drained peatlands account for 3 % of the Dutch national greenhouse gas (GHG) emissions. Topsoil removal (TSR; which leads to removal of built-up nutrients and labile soil carbon) in combination with rewetting (groundwater table management) in drained peatlands is often proposed as a restoration measure effective in reducing carbon dioxide (CO2) emissions, yet often politically challenging to implement. TSR is however currently applied at a smaller scale during the implementation of regenerative ditch borders (RDBs), but its effectiveness in reducing CO2 remains uncertain. We investigated the effects of RDBs on soil CO2 emissions in comparison to conventional borders (CDBs) in a year-round field survey (June 2023 – March 2024) in a lowland peat agroecosystem. Soil respiration was measured at four distances from the water’s edge (40, 80, 360 and 640 cm) over a ten-month period and used to fit statistical models with the predictors soil temperature, soil moisture content and exposed carbon (a variable integrating profile soil carbon density and groundwater level). The resulting model was used to calculate annual soil respiration, and to estimate the payback time (no. of years for net negative effects on cumulative CO2 emissions) of the removed carbon. Spatiotemporal variation in soil respiration was mostly explained by exposed carbon and soil temperature (32 % and 30 %, respectively). Soil respiration peaked at 65 % soil moisture content. At distances 40 and 80 cm reduced soil respiration in RDBs in spring and summer was driven by lower amounts of exposed carbon, while at 360 and 640 cm in RDBs higher soil temperatures and soil moisture content mostly counteracted this effect. Model-based annual soil respiration was 17 % lower in RDBs in comparison to CDBs. If it is assumed that all soil carbon removed during RDB construction is mineralized, the payback time can exceed 100 years. While RDBs can promote reductions in soil CO2 emissions and therefore more sustainable peatland-adapted agriculture, potential emissions from excavated carbon should be accounted for.
- Preprint
(1886 KB) - Metadata XML
-
Supplement
(1153 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2026-3155', Anonymous Referee #1, 14 Aug 2026
- AC1: 'Reply on RC1', Sanne Bethe, 18 Sep 2026
-
RC2: 'Comment on egusphere-2026-3155', Anonymous Referee #2, 07 Sep 2026
Review of „Implementation of regenerative ditch borders in Dutch peat meadows: effects on soil CO2 fluxes and potential carbon trade-offs“
Here Bethe et al. works a relevant knowledge gap about the effects of the topsoil removal (TSR) without rewetting. The previous literature has primarily considered couples of TSR with rewetting, so the decoupling these two management interventions is an interesting contribution to peatland management literature. The model formulation appears suitable, as it includes an exponential temperature response and the parabolic soil moisture term. But, in my point of view, the most original idea of the manuscript is the approach used to collapse the carbon profile and groundwater level into a single metric (section 2.5). This is a well done calculation to integrate these two variables. Besides, the combination of the counterfactual decomposition with the SHAP to partition the differences between regenerative ditch borders (RDBs) and conventional borders (CDBs) and to assess their contributions from a spatiotemporal analysis. I found very strength part of the study. I also appreciate that the manuscript is cautious in their analysis of trade offs, by not oversell the 17% of reduction by RBDs compared to the CBDs. So I hope my comments below can help the authors to improve the manuscript.
There is a problem in the experimental design, and consequently the level of the general conclusion. There are only 4 CDBs and 16 RBDs and all of them located within a single minerotrophic peat polder. So, talk about the Dutch peat meadows appear to be extended considerably beyond the site replication, so how do the authors can justify this broader inference based on a single ecosystem site?
I am also concerned about the apparent use the same data both to fit the model and generate the observed respiration, so it is feeding and payback the model, please consider it in the analysis as circularity validation. Please try to clarify wheter an independent validation of the data, cross validation, or another type of validation.
I am surprised that the daytime chamber campaigns missed April-May months, these months represent an important transition period in temperature dynamics, where ambient temperature increase rapidly and respiration rates change substantially, what is the reason for omitting them? But also, you need to acknowledge that your study is made from daytime and it is missing the diel effects, so in this context 17% of differences between RDBs and CDBs should be interpreted with caution.
Regarding the carbon-debt calculation and propagation of uncertainty, I can see that extrapolations came from a single point and carbon density were measured at only one site nd then extrapolated for other sites (40, 80 ad 640 cm), so the extrapolation errors are absent for the reported uncertainties. This is particularly important because the excavated carbon term appears to be a major factor controlling the estimated carbon payback time. If the underlying c density estimates are derived from a single sampling point, their extrapolation across the entire excavated volume may introduce substantial uncertainty and potential bias.Citation: https://doi.org/10.5194/egusphere-2026-3155-RC2 - AC2: 'Reply on RC2', Sanne Bethe, 18 Sep 2026
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 140 | 38 | 28 | 206 | 29 | 15 | 14 |
- HTML: 140
- PDF: 38
- XML: 28
- Total: 206
- Supplement: 29
- BibTeX: 15
- EndNote: 14
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The manuscript by Bethe et. al., investigates the effect of soil temperature, soil moisture and exposed carbon on CO2 emission from drained peatlands in proximity to ditches where topsoil removal has been implemented. They create 20 transects from ditch edges and into drained peatland fields and measure soil temperature, soil moisture and CO2, 5 to 8 times during the period. To estimate the exposed carbon, they measure carbon content from transect cores above daily groundwater levels. Furthermore they calculate the time it will take to reach climate neutrality using modelled soil respiration and carbon density removed as topsoil. They find that implementing TSR in ditch proximity of drained peatlands reduce annual soil CO2 emissions by 17%. Furthermore they estimate the payback time for peat removed as topsoil may exceed 100 years.
My only major concern regarding the manuscript is regarding the calculation of exposed carbon, as it is stated in the manuscript that only one core is sampled inside each transect which is used to estimate carbon content of complete soil profile inside the transect. While this method is valid, it is stated in the manuscript that the soil profiles, based on the cores collected, varied greatly and that the transects are in close proximity to one another. If possible, I would like an uncertainty added to carbon content estimated inside the transect and the matter mentioned in the discussion.
Besides this, I don’t have any major concerns.
Abstract
L27 – 30: TSR is proposed to reduce CH4 emisssion in rewetted peatlands more so than CO2.
L30: How is it politically challenging? I suggest either explaining it or removing it.
L37 -38: As the exposed carbon is not a common known variable, I would like some more explanation to it.
Introduction
L58: Last decades – 20? 50? 80? Be more specific
L 81-83: appears somewhat redundant with line 65-68 as the role of oxygen in stimulating CO2 emissions has already been explained. Consider combining these two statements.
L 103-105: Something appears to be missing from the sentence please check it.
L112: there is a discrepancy is the explanation of rewetting inside the parenthesis with the one in the abstract.
L126: Regenerative ditch border is first mentioned here. Please check abbreviation through the manuscript
Methods
307-336: The exposed carbon calculation relies on the assumption that the carbon densities measured at 360cm is representative of all other distances in each transect. As the exposed carbon variable is a major predictor for CO2 emissions, I find this assumption “brash” as it is known that peat and soil layers can vary much spatially. Please acknowledge this limitation and discuss its implication.
Can the uncertainty be quantified?
404: missing word – years?
Results
509-512: How is the soil moisture only measured to 83% if groundwater levels are at or exceed surface level? Please check it.
519-524: please add trend lines and r2 values to the relationships.
Discussion
705-707: This is important as it implies the importance of peat depth for CO2 emissions
729-740: Here I miss a discussion on the exposed carbon assumption. As stated in L735-736: “the DB sites in this study were in close proximity to each other” and 513-516: “ There was also considerable variation in the relative distribution and thickness of the different soil layer types in particular DB 13 was rich in clay, which could suggest a previous riverine history, and DB 7 had quite a large, disturbed gravel layer in the top 1 m.”. This may indicate that the transect soil layer may vary to some extent. “
I think the manuscript is well written and follows a logic trajectory throughout. The manuscript is relevant to the peatland scientific community, as it thoroughly investigates CO2 emissions in drained peatland and the specific method of including exposed carbon is, to my knowledge, still poorly investigated.