Final response - RC 1:

General comment:

Authors present a comprehensive study about soil respiration and its components across different ecosystems in Switzerland. Radiocarbon (14C) in respired CO2 was used to separate autotrophic and heterotrophic contributions and provided information about the age of the respired CO2 from both sources. Based on the fluxes, the age of respired C and the decomposability of SOC, the different ecosystems were categorized into five (soil) C cycling types. The manuscript is very well written and the study is highly valuable as it involved a whole set of different ecosystems in a replicated design. The interpretation of the results is straightforward. Maybe sometimes it’s a bit too straightforward.

Response: We thank the referee a lot for the overall positive evaluation and constructive comments.

Specific comments:

1.1 Table with site information

Comment: The descriptions of the sites could be improved. Figure 1 looks nice, but apart from showing the areal distribution of the sites, it does not contain any relevant information. I suggest moving Figure 1 to the supplementary material and replacing it in the main text with Table S1, which contains all the relevant site information. If basic soil data is available, such as C and N concentrations or stocks, this could also be presented in the table.

1.2 Land-use history

Comment: With regard to the cropland and disturbed peatland more information about land-use history needed to be introduced (and discussed). It is often referred that cropland on former grassland loses C, but it remains unclear when the three cropland sites were established and what was there before. The same for the disturbed peatland sites.

Response: We agree with the reviewer and present the land-use history more in depth as follows: The long-term trials at the cropland sites were established in 1989 (Reckenholz and Altwi; Hirte et al., 2021) and in 1976 (Changins; Maltas et al., 2018). All sites, however, had already been under cropland use prior to the establishment of these trials, amounting to > 35 years (Reckenholz, Altwi) and > 50 years (Changins) of continuous cropland management. The managed peatlands were drained in the second half of the 19th century and have been cultivated as croplands since then (Leifeld et al., 2011).

Accordingly, we propose to add the following clarifications in the methods (section 2.1):

• Line 120: “… are part of long-term field trials of the Swiss Federal Research Institute Agroscope, Switzerland, established 35 and 49 years ago (for Changins: e.g., Maltas et al., 2018; for Altwi, Reckenholz: e.g., Hirte et al., 2021). As part of the Swiss Plateau near settlements, the sites had been used as grassland, orchards and vineyards until the 19^th century and had then been converted to croplands before they were used for the long-term cropping trials (Oberholzer et al., 2014).
• Line 121: “The managed peatlands were drained in the second half of the 19th century and have been used for crop production ever since (Leifeld et al., 2011).”

To include land-use history in the discussion, we addressed this aspect by:

• Adding the following sentence in section 4.4, line 421: “The depletion of labile SOC is likely particularly pronounced in the investigated croplands, as they have been continuously managed as croplands for more than 35 years, likely even more than 100 years.”

1.3 Site selection (related to comment 1.10)

Comment: The choice of the three forest sites needed to be explained. As there are “only” three or four replicates in each land-use class, their choice (site climate, soil type...) has quite an effect on the results. Seems that some forest sites experienced very dry conditions during summer-sampling. Is this just by chance, or are these forests generally particularly dry in summer (Pfrynwald is very dry)? This could be mentioned in th emethods and discussed in the discussion section. The drought effect is for instance impressivel in the CO2 fluxes, which were smaller during summer in two forests than during March though soil temperatures had been much higher in summer.

Response: We thank the referee for this important remark. Our replication of specific land-use types was confined by time and funding resources to determine ¹⁴CO₂. Generally, in Switzerland, croplands and peatlands are found under relatively similar environmental conditions, whereas grasslands and forests occur across a broader range of climatic conditions. In selecting the three forest sites, we deliberately aimed to capture variability in climate and soil properties in order to test the robustness of land-use type effects on rates, ages, and source contribution of respired CO₂.

As the referee correctly noted, Pfynwald is characterized by very dry summer conditions, typical for inner-alpine valleys, whereas Beatenberg, located at ~1500 m a.s.l., experiences considerably lower MAT compared to the other forest sites. While these climatic differences and soil conditions lead to variation of respiration rates among forest sites, particularly under summer drought, the ages and source contributions of respired CO₂ were relatively consistent across forest sites. This highlights that land-use type exerts a stronger control on C cycling pathways than short-term climatic variation, even though the latter can strongly affect instantaneous respiration fluxes.

To more explicitly address our rationale in the methods, we add the following sentence in section 2.1, line 116:

“Generally, croplands and managed peatlands are found under relatively similar environmental conditions of the Swiss Plateau, whereas grasslands and forests occur across a broader range of climate conditions. We deliberately chose sites from Swiss monitoring programs to capture variability in climate and soil properties allowing us to test the robustness of land-use type effects on rates, ages, and source contribution of respired CO₂.”

To the discussion, we add the following sentence to section 4.3, line 409 (this suggestion is related to comment 1.10):

1.4 Replace season by months

Comment: The use of 'winter' and 'summer' as sampling times is generally a bit strange to me. While July–August is clearly summer, sampling in March does not represent typical winter conditions (20 March onwards is astronomical spring, and the whole of March is spring under the meteorological definition!). I therefore suggest referring to the sampling dates as 'July/August' and 'March' sampling throughout the text, rather than 'summer' and 'winter'. CO₂ fluxes were quite high in March, which is not typical of real winter, and all the snow had melted, which is also not typical of a Swiss winter. This issue has been discussed to some extent (much warmer air temperatures than soil temperatures in March, etc.). So why not drop the term 'winter' as a definition?

Response: We thank the referee for this thoughtful and well-taken comment. We agree that referring to the sampling periods as "summer" and "winter" can be misleading – particularly in the case of March, which does not accurately represent phenological or meteorological winter conditions. As suggested, we will revise the manuscript to refer to the sampling periods by month (i.e., "March" and "July/August") rather than by season.

1.5 Rationale for clipping vegetation

Comment: Another methodological issues is the clipping of all vegetation in the chambers already 1-2 weeks prior to measurement. Do you believe that this has no effect on autotrophic soil respiration, especially in the grasslands? I am not so sure. The reasoning for this design needed to be worked out here and explained why this approach was chosen.

Response: Thanks for the comment. While we briefly noted our reasoning and the associated drawbacks in the manuscript, we agree that the rationale behind clipping vegetation prior to in situ gas sampling should be explained more clearly.

Our primary reason for clipping the vegetation 1–2 weeks prior to sampling was to minimize the contribution of aboveground plant respiration to the measured CO₂ efflux and its isotopic composition and to reduce short-term disturbances, following other isotope-based studies (e.g., Wunderlich & Borken, 2012). This approach is briefly mentioned in the methods section (lines 134–135): “To exclude the contribution of above-ground plant respiration, vegetation inside the chamber was clipped at installation and before sampling if plants had regrown.”

We acknowledge, however, that clipping might still have affected the contribution of autotrophic respiration which we now discuss in the study limitations more extensively (section 4.7, lines 480–483):

“To exclude aboveground plant respiration from the in situ CO₂ flux, vegetation inside the chamber was clipped 1-2 weeks prior to sampling and directly before if there was plant regrowth. While this approach helped to isolate soil-respired CO₂, clipping of vegetation inside the chamber shortly before sampling could initiate a pulse of rhizosphere respiration (Wunderlich & Borken, 2012), potentially leading to an overestimation of autotrophic contributions. However, given the time allowed between clipping and sampling, we assume that the overall impact on autotrophic respiration in our study was limited. In addition, other studies suggest that such effects are minor or transient (Barneze et al., 2024; Bahn et al., 2006; Zhou et al., 2007).”

1.6 Sieving

Comment: Though I am not specialized in 14C modelling, the distinction between autotrophic and heterotrophic components is reasonable and the methodoligical attempts to account for CO₂ from carbonates are impressive.  There still remain some general methodological issues, such as e.g. that sieving and physical disruption couold make protected SOC accessigle to decomposers and affect the heterotrophic CO₂ efflux (and age of respired C).

Response: We agree that sieving can disrupt soil structure and expose SOC occluded within aggregates, making the incubation conditions less representative for field conditions. However, this cannot be avoided as roots have to be removed as they would affect the isotopic signature of heterotrophic respiration toward younger CO₂.

To reduce aggregate disruption, we sieved the soil at 4 mm, which is likely less disruptive than the more common 2 mm sieving.

1.7 Picking out roots

Comment: In the cropland soil, picking all the roots out might cause some bias too. Since the cropland sites are harvested two times a year, there should be some dead decomposing roots in the soil. I would consider the CO₂ from this source heterotrophic. Decomposing fresh roots would have a very recent 14C signature, similar to that of “autotrophic respiration”. By removing all roots from the soil sample only “older” C is left for the assessment of heterotrophically respired C. This might be a reason why the 14C data suggests such extremely high contribution of autotrophic respiration (>90%) from cropland (even during winter). However, this is just an idea. If you think it makes sense, you could include a discussion of such methodological constraints.

Response: We thank the referee for this thoughtful comment and agree that decomposition of fresh roots can influence the ¹⁴C signature of in situ soil-respired CO₂ and may lead to an overestimation of autotrophic respiration. We are confident that we captured this contribution during summer, as soil samples for incubations were collected on the same day as the in situ CO₂ sampling. In croplands and managed peatlands, living roots could be well distinguished from recent root litter, as white, water-filled live roots were concentrated beneath the crop rows, and we avoided damaging intact root systems when collecting soil cores.

However, we acknowledge that we may have missed the contribution of heterotrophic respiration from recent roots in winter, as we did not repeat soil sampling and instead used heterotrophic respiration signatures from summer-incubated soils. As the referee points out, this is particularly relevant for croplands and managed peatlands, where seasonal crop rotation regularly introduces recent root material.

We briefly addressed this limitation in the discussion section of seasonal dynamics (section 4.6, lines 467-470) and in the study limitations (discussion section 4.7, lines 483–484).

We agree that this limitation should be stated more explicitly and therefore added the following sentence:

1.8 ¹⁴C data of soil organic carbon (related to comment 1.12)

Comment: In the methods 2.6 a whole chapter describes the analysis of 14C in soil organic carbon. However, these data are nowhere presented or used in the manuscript, except for estimating the contribution of carbonate C to CO2. The SOC 14C values would be very interesting to the reader of this work – C transit times could be calculated and used to back up the theoretical framework.

1.9 Depletion of atmospheric ¹⁴CO₂ in winter

Comment: Maybe this is normal, but is there any reason why the delta 14C in atmospheric CO2 was much more depleted in in winter than in summer? (Page 13 L290)

1.10 SOC decomposability (related to comment 1.3)

Comment: Seems decomposability of SOC in the different soil and humus layers was analyzed by incubating at field soil moisture at 22°C (Page 7 L180onwards). This might be problematic because very dry soil might not respond similarly to the change in temperature to 22° as soil with sufficient water supply - hence one would end up with lower “decomposability” in dry soil (this maybe the reason for the very low decomposability of the organic layer at Pfrynwald, decomposability was high in the other two forests, Figure 5).

Response: Thanks for the comment. We agree that incubating soils at field moisture rather than under standardized moisture conditions can affect the comparability of SOC decomposability across sites. Our decision to use field-moist conditions was driven by the aim to more closely reflect in situ conditions and thereby better capturing the source contribution of heterotrophic respiration to the mixed CO₂ sampled in the field, as is common in isotope-based source partitioning studies (e.g., Schuur et al., 2006).

In addition to reflecting typical July conditions for the sites, we are confident that the SOC decomposability results from our study still remain suitable for comparison across sites and depths for several reasons.

• First, relative differences across land-use types were clear and consistent despite site-specific variability within each type.
• Second, (as referred to by the referee) for forest organic layers, moisture limitation likely led to an underestimation of decomposability. To better address this in the revised manuscript, we add the following sentence to the results (section 3.4, line 315):

“In Pfynwald, SOC decomposability in the organic layer was restricted due to low soil moisture conditions (Figure 1).”

• Third, in other land-use types – particularly croplands – SOC decomposability showed little variability despite differences in soil moisture across sites. Finally, most sites were incubated at moisture levels above 20%, where microbial activity is generally not strongly limited by water availability.

1.11 Source contribution in Reckenholz

Comment: Figure 6 croplands – why is the summer measurement at Reckenholz completely different when compared to all the other cropland measurements. Is there any explanation?

Response: We attribute this striking discrepancy mainly to two factors:

(1) Soil moisture conditions. In summer, soil moisture was substantially higher in Reckenholz (VWC = 27 ± 3 %) than in Changins (VWC = 13 ± 3 %) and Altwi (VWC = 14 ± 1 %; Fig. 1). The low soil moisture at Changins and Altwi likely restricted heterotrophic respiration, leading to lower relative contributions compared to Reckenholz. We address this point in the discussion (section 4.4, lines 430–435):

“We further assume that high relative autotrophic contributions are also related to hampered heterotrophic respiration due to adverse soil environmental conditions. The two sites, Altwi and Changins, with exceptionally high autotrophic contributions (96 %), exhibited very low water contents (<15 vol-%). In contrast, the Reckenholz site, with moderate moisture levels (27 vol-%), showed a much lower autotrophic contribution of only 15 % (Figs. 1, 5). The low water contents in cropland surface soils are probably due to the sparse vegetation cover, which enhances evaporative water losses.”

1.12 Confusion about the term destabilization (related to comment 1.8)

Comment: In general, I sometimes found the use of the terms 'destabilisation' and 'degradation' unclear. For example, on page 17, line 360, it says that 'destabilised C-depleted systems, where reduced C inputs result in the depletion of recent C material in SOC stocks', but no SOC-14C data is presented here, so how can this conclusion be drawn?

Response: We thank the referee for this comment. Referring to the given example (see response to comment 1.8): ∆¹⁴C values of bulk SOC are presented in Fig. S6 and used for discussing the depletion of labile C in croplands (methods section 4.4, lines 412-414).

1.13 Vulnerability of SOC

Comment: I very much like the grafical abstract, but I am not sure if the other conceptual figure 7 is fully conclusive. E.g. Increases "Vulnerability"  really if the age of the respired heterotrophic C in CO2 increases? What if all the young C is decomposed? - would then the soil be less "vulnerable"?

Response: We thank the referee for this constructive comment. In our conceptual framework, we use flux rates and ages of in situ and heterotrophically respired CO₂ to infer on the status of C cycling across land-use types. We consider unprotected SOC fraction in natural and semi-natural ecosystems as vulnerable to disturbances induced by environmental change (e.g., climate or land-use change). We propose that SOC vulnerability increases with the age of respired heterotrophic C because older SOC stocks have accumulated over long time periods and cannot be readily replenished if lost through enhanced decomposition. The referee is correct that if most of the younger, more labile SOC has already been decomposed and lost (as observed in croplands indicated by low SOC decomposability and old SOC and respired CO₂), the remaining SOC pool is comparatively more stable and thus less vulnerable to future disturbances.

Given the referee’s concerns about the conclusiveness of the conceptual framework, we propose to remove Figure 7 from the manuscript. Importantly, the discussion does not rely on this figure and remains valid without it.

Final response – RC 2:

Overview

The authors are investigating how different land-use types affect soil carbon cycling by comparing respiration rates and carbon isotopic signatures (¹⁴C and ¹³C) of bulk soil and soil-respired CO₂ across a wide range of landscapes representing grasslands, forests, croplands, and peatlands. By analyzing both the quantity and age of CO₂ released from autotrophic and heterotrophic sources during summer and winter, they identify distinct carbon cycling regimes ranging from rapid, recent carbon turnover in temperate grasslands to the release of millennia-old carbon in disturbed peatlands. This study is highly relevant to understand how land use and management impact carbon retention and loss at the ecosystem level and deepens on the important question of carbon stabilization timescales. This research has direct relevance for carbon cycling timescales modeling, climate change mitigation, and land-use policy development.

Overall comments

I found the manuscript enjoyable to read and well-written, with a clear presentation of ideas. It offers a valuable contribution by advancing the application of radiocarbon measurements to investigate soil carbon cycling—an approach highly relevant to the broader biogeosciences community.

2.1 Integration of literature

Comment: Nonetheless, to more effectively position this study within the broader context of current research, the manuscript would benefit from a deeper integration with the existing literature on the age of soil organic carbon (SOC) and respired CO₂. Specifically, comparing the findings with those of previous studies would strengthen the interpretation. A more explicit discussion of how bulk SOC and respired CO₂ radiocarbon measurements align or differ across ecosystems would enhance the relevance and impact of the conclusions. Additionally, incorporating more recent and thematically aligned references would clarify how this study extends, supports, or challenges established understanding in the field.

For example, you might integrate the findings of the following recently published articles into your introduction and discussion:

https://doi.org/10.1029/2024JG008191



https://doi.org/10.5194/bg-21-1277-2024

https://doi.org/10.5194/soil-10-467-2024

https://www.nature.com/articles/s41598-021-85506-w

https://doi.org/10.5194/bg-10-7999-2013

https://doi.org/10.1111/gcb.17320

https://www.nature.com/articles/s41561-020-0596-z

Response: We thank the referee for these literature suggestions. We will integrate several of the proposed references into the revised manuscript where they are most relevant to our study. As the manuscript already includes 88 references, we will focus on those that most directly complement our study to ensure clarity and conciseness.

Abstract

2.2 Clarification of “retarding system”

Comment: Pretty clear and succinct. Thanks! However, the rationale behind the “retarding system” seems to be counterintuitive. The interpretation that more bomb-derived Δ¹⁴CO₂ in heterotrophic respiration at higher elevations reflects a slowing of C cycling appears counterintuitive. Enrichment in bomb ¹⁴C generally indicates faster cycling of relatively recent carbon. To robustly support the claim of retarded carbon turnover at higher elevation, it would be helpful to refer to bulk SOC Δ¹⁴C values, which are expected to become more ¹⁴C-depleted (older) with elevation. Clarifying this distinction between respired and bulk signatures would strengthen the argument.

Response: We thank the reviewer for this thoughtful comment. As noted in our response to comment 1.8, we refer to bulk SOC Δ¹⁴C values of alpine grasslands (Fig. S2) in the discussion (section 4.2, lines 391–393), supporting our interpretation of slower C cycling at higher elevation.

We acknowledge, however, that categorizing alpine grasslands as retarding systems based on higher (bomb-derived) Δ¹⁴CO₂ values in heterotrophic respiration may not be intuitive and requires further explanation. The key lies in the interpretation of Δ¹⁴C signatures across soil depth and in the context of the atmospheric bomb spike. Since atmospheric Δ¹⁴CO₂ levels have been declining for decades, higher (positive) Δ¹⁴C values do not necessarily indicate faster turnover of very recent C; rather, they reflect the decomposition of decadal-old, bomb-derived C. Conversely, more depleted Δ¹⁴C values can indicate either pre-bomb C (centuries old) or very young, modern C fixed only a few years ago.

Bulk SOC Δ¹⁴C values across our elevation gradient (Fig. S2) support this interpretation. In topsoils, Δ¹⁴C values first increase from near-modern levels, peak around ~1500 m, and then decline at higher elevations. At the highest sites, bulk SOC Δ¹⁴C values in 0–5 cm soil layers remain > 0 ‰ (bomb-derived), whereas in 5–10 cm they decline to pre-bomb values (as correctly expected by the referee), showing the depth-dependent imprint of the bomb spike. Thus, bulk soil Δ¹⁴C values follow a polynomial distribution with elevation, reflecting the bomb spike across the elevational gradient. With increasing soil depth, the bomb spike is less reflected, as pre-bomb levels dominate Δ¹⁴C values of bulk soil SOC.

Our Bayesian mixing model revealed that about 70% of heterotrophic respiration originates from topsoil layers, aligning Δ¹⁴CO₂ values of soil-respired CO₂ with Δ¹⁴C of topsoil bulk SOC. Thus, the elevational pattern of bulk soil Δ¹⁴C values is mirrored in the respiration data. Increasing Δ¹⁴CO₂ of in situ soil respiration and heterotrophic respiration from topsoil layers with elevation indicates enhanced decomposition of bomb-derived, decadal-old C at high elevations. In contrast, subsoil Δ¹⁴CO₂ remains older than modern and shows no clear elevational trend. As Δ¹⁴CO₂ values of total soil respiration and heterotrophic respiration from topsoil layers are continuously increasing with elevation, topsoil layers contribute relatively stronger to heterotrophic respiration at higher elevation as subsoil layers which is also confirmed by Bayesian mixing model outputs.

Also, from a mechanistic point of view SOC cycling rates have to slow in these alpine grasslands as MAT’s are around 0°C (Table 1) and both productivity as well as respiratory activity are low.  

Taken together, these findings show that enrichment in bomb-derived Δ¹⁴CO₂ at higher elevations reflect an overall slower C cycling and longer C retention in alpine grasslands.

For a more coherent argumentation of alpine grasslands as retarding systems based on Δ¹⁴CO₂ and bulk SOC Δ¹⁴C in the manuscript but still keep the discussion concise, we modify the following paragraph in the discussion section 4.2, lines 379-385:

Graphical abstract

2.3 SOC decomposability and C retention

Comment: The graphical abstract looks to me a bit counterintuitive when observing that the direction in which SOC decomposability increases, is where there is more C retention. A high SOC decomposability indicates “a large fraction of readily available organic matter” (L86-88), with subsequent short persistence and low C accumulation, right? Maybe some clarification is needed.  The graphical abstract should be very explicit without room for misinterpretation. Also, why is C stabilization and C retention in opposite extremes of the Y axis? Don’t they point out to actually the same effect? Or what does belong to the X and Y axis?

Response: We sincerely thank the referee for this valuable comment. Following the suggestions of the two referees, we have revised the terminology (SOC decomposability, C retention, C stabilization, etc.) and provide a clearer definition.

On SOC decomposability and C retention:

The referee is correct that, in general, a high SOC decomposability implies a larger fraction of readily available organic matter. Readily available organic matter accumulates in systems, where decomposition and transformation of plant residues is “retarded”, leading to an accumulation of labile C often as POM (Duborgel et al., 2025; GCB). In our study, this happens in forest soils where the chemical recalcitrance slows litter processing, and in alpine grasslands by low temperatures. To avoid confusion, we remove the term “C retention” from the graphical abstract.

On C stabilization:

In the graphical abstract, “C stabilization” was intended to capture the effect of ecosystem disturbance. In less disturbed systems (e.g., natural grasslands), more SOC remains stabilized, whereas in highly disturbed systems (e.g., croplands or managed peatlands), labile SOC is rapidly lost through enhanced microbial decomposition. However, as also discussed in response 1.12, we recognize that the term “C stabilization” is conceptually overlapping with physico-chemical “stabilization” and therefore potentially confusing. For this reason, we now omit the terms “C (de)stabilization” in the manuscript and use “depleted” and “degraded” instead.

Introduction

2.4 Rhizosphere respiration

Comment: L45: The definition of "root respiration" in the manuscript is conceptually unclear and potentially misleading. By including heterotrophic respiration driven by root exudates under this term, the distinction between autotrophic and heterotrophic processes is blurred. I recommend replacing "root respiration" with "rhizosphere respiration" and clearly distinguishing the contributions of plant and microbial respiration to avoid misinterpretation.

2.5 Literature suggestion

Comment: L63: This article might be of interest to introduce the allocation of C in the organic layer and other pools in a forest, analyzed from the radiocarbon perspective  https://doi.org/10.1029/2024JG008191.

2.6 Formality

Comment: L71: “e.g.” can be deleted in this and the subsequent reference parenthesis?

2.7 Concept of transit time

Comment: L76: The concept of transit time, while seemingly central to this study, is not new. Its original formulation can be traced back to Eriksson (1971) and Bolin and Rodhe (1973). It would be appropriate to acknowledge these foundational works to provide historical context. That said, it is also scientifically valuable (as you did) to see this concept being re-applied to current questions in terrestrial carbon cycling. That said, its re-application to contemporary questions in terrestrial carbon cycling, as done here, is both timely and valuable. However, although the concept is introduced in this paragraph, it is not revisited or further developed in the methods, results, or discussion. It would be helpful to understand whether there was a specific reason for not integrating this concept more explicitly throughout the analysis.

2.8 Literature suggestions

Comment: L74-100: The references cited in this section are important early contributions; however, a more comprehensive and updated literature review would strengthen the background, especially considering the significant methodological and conceptual advances in the field over the past three decades and the diversity of researchers emerging from different groups. Including more recent work would better reflect the dynamic evolution of approaches used to study carbon cycling and persistence using radiocarbon measurements. This would not only highlight the continued relevance of the field but also situate the present study more clearly within ongoing scientific discourse.

L91: This article might be of interest https://doi.org/10.5194/soil-10-467-2024 for comparing the effects of land-use in SOC stability in croplands and managed ecosystems.

2.9 Hypotheses

Comment: L101-107: Please define what is young and what is old in terms of timescales. Younger than X or older than X? Also, I wonder if the five hypotheses could be summarized within an underlying mechanism that explains the transit time (age of respired CO₂) across land-use types? Just a thought/suggestion that can be useful for better synthesis.  

Response: Thanks for this comment. In the revised manuscript, we added the age ranges we expect. We think there is not one underlying mechanism, but rather an interplay of different factors determining the observed ages of respired CO₂. In the revised manuscript, we now begin our hypothesis section as follows:

Materials and methods

2.10 Study sites

Comment: Please consider providing more detailed information on the study sites, especially for the managed peatlands. Since key parts of the discussion rely on mechanisms such as seasonal changes in water table depth, SOC content, and site-specific management practices, a clearer description of these variables is essential for the reader to fully understand the interpretation of seasonal Δ¹⁴CO₂ variability.

For instance, site-level context would also assist in evaluating how representative the findings are, and how much variability exists among peatland responses, at least for the sampled seasons. This is particularly relevant since the manuscript proposes generalizable categories of carbon cycling (stabilizing vs. destabilized systems).

Response: Thanks, we show the site information including SOC contents more prominently in the new Table 1 and provide additional information on the land-use history of croplands (section 2.1, lines 116-120:

“[Croplands]…are part of long-term field trials of the Swiss Federal Research Institute Agroscope, Switzerland, established 35 and 49 years ago (for Changins: e.g., Maltas et al., 2018; for Altwi, Reckenholz: e.g., Hirte et al., 2021). As part of the Swiss Plateau near settlements, the sites had been used as grassland, orchards and vineyards until the 19^th century and had then been converted to croplands before they were used for the long-term cropping trials (Oberholzer et al., 2014).”

and managed peatlands:

“The managed peatlands were drained in the second half of the 19th century and have been used for crop production ever since (Leifeld et al., 2011). Crop types differ between each peatland site and season, with most of the cultivated crops being vegetables (Table 1).”

2.11 LULUCF abbreviation

Comment: L112: Introduce LULUCF abbreviation by writing it complete.

2.12 Figure caption correction

Comment: Figure 1: In the caption: “Five” instead of “six”?

2.13 Headspace volume of chambers

Comment: L132-151:  As the height of the PVC frames might be different when inserted into the ground. Please clarify if the respiration rates where normalized by the actual headspace volume of the chamber. Also, indicate how many replicates of CO₂ fluxes for each chamber were taken.

2.14 Soil volume estimation in alpine grasslands

Comment: L159: Please be more specific on the volume estimation method.

2.15 Weighted flux estimation with Q10 adjustment (related to comment 2.25)

Comment: L185: The use of a uniform Q₁₀ value of 2.4 to adjust respiration rates across all sites assumes a relatively high and consistent temperature sensitivity of microbial respiration. Given the diverse ecosystems studied (alpine grasslands, forests, peatlands), this assumption may not hold universally. Temperature sensitivity of SOC decomposition is known to vary with substrate quality, moisture availability, microbial community composition, etc. Could you elaborate on the justification for applying a single Q₁₀ value?

Response: We thank the referee for this comment. We agree that the temperature sensitivity of SOC decomposition varies across ecosystems, which cannot be fully captured by a uniform Q₁₀. However, since site- and depth-specific Q₁₀ values were not available for our study, we applied a standardized value of 2.4, as suggested for soil respiration by Raich & Schlesinger (1992). This provided a consistent basis for adjusting incubation-derived heterotrophic respiration rates (measured at 22 °C) to field temperatures measured during in situ CO₂ sampling. We further accounted for depth-related temperature changes following Bourletsikas et al. (2023). These adjusted rates were only used to weight the relative contributions of each depth layer to the total heterotrophic flux during in situ field measurements. We used these weighted fluxes in the Bayesian mixing model to improve the model performance.

Importantly, the SOC decomposability data are always presented as directly measured during incubations and were not temperature-adjusted.

For improved clarity, we propose to move the relevant paragraph from section 2.4 (lines 183–186) to section 2.8 (line 248) and rephrase it as follows:

2.16 Basal respiration as a proxy for SOC decomposability

Comment: Moreover, are all sites assumed to be primarily temperature-limited in their respiration response, or could other limiting factors (substrate or water availability) affect the comparability of basal respiration as a proxy for SOC decomposability across sites? The uncertainty of this assumption deserves further discussion.

Response: We thank the referee for raising this point. In our study, all soil samples were incubated under the same temperature (22 °C), so differences in basal respiration and SOC decomposability cannot be attributed to temperature sensitivity but instead reflect other factors. Basal respiration under these standardized conditions primarily serves as a proxy for substrate availability, meaning that differences in SOC decomposability across land-use types and sites reflect variation in the amount of readily available SOC.

2.17 Device specifications

Comment: L209: Maybe not necessary to specify the characteristics of the device once again if it was already introduced in the previous paragraph.

Results

2.18 Statistical significance of Δ¹⁴CO₂ values

Comment: L281: The statement regarding differences in Δ¹⁴C of total soil respiration between summer and winter across temperate grassland, alpine grassland, and forest ecosystems appears somewhat strong given the actual magnitude of those differences. While the p-value may indicate statistical significance, the practical or ecological significance is less clear. To aid interpretation, it would be helpful to include mean values directly on the box plots. Additionally, the phrasing of this statement could be revised to avoid overgeneralization and better reflect the nuance in the data.

Response: We thank the referee for this comment. We would like to emphasize that in the current version, we report the statistical significance of the overall seasonal effect on Δ¹⁴CO₂ values of total soil respiration, but also acknowledge that this effect is differently pronounced across land-use types (lines 281-284):

“The Δ¹⁴CO₂ values of total soil respiration differed significantly between summer and winter (p = 0.001) with Δ¹⁴CO₂ values being generally closer to atmospheric CO₂ in winter. However, the effect of seasonality varied significantly across land-use types (p < 0.001). While managed peatlands exhibited the largest seasonal changes in soil respired CO₂, temperate grasslands and forests showed the lowest seasonality (Fig. 3).”

2.19 Δ¹⁴CO₂ depth pattern in croplands and managed peatlands

Comment: L295: This is a particularly interesting observation: “values in croplands and managed peatlands increased from very depleted values at 0–5 cm to less depleted values at 5–10 cm soil depth.” Such pattern suggests a counterintuitive vertical trend in ¹⁴C values, potentially pointing to surface inputs of older carbon or distinct microbial processing dynamics across layers. It would be valuable to elaborate more on the possible mechanisms behind this in the Discussion.

Response: We thank the referee for this thoughtful comment. In croplands and managed peatlands, soil management practices are the most likely explanation for the observed vertical trend in Δ¹⁴CO₂ values. Ploughing can redistribute older SOC toward the surface, thereby altering the expected depth gradient. In addition, at the two cropland sites (Altwi and Changins), the highly depleted Δ¹⁴CO₂ values observed in the 0–5 cm layer are most likely explained by CO₂ release from carbonate weathering, which contributes depleted radiocarbon signatures to the soil-respired flux.

Because the effects of ploughing on the vertical Δ¹⁴CO₂ profile are relatively minor and do not affect the overall classification of croplands and managed peatlands into specific C cycling systems, we propose to streamline the text by focusing on the carbonate effect at the two cropland sites. Specifically, we will revise the results (section 3.3, lines 295–296) as follows:

“However, Δ¹⁴CO₂ values from two cropland sites increased from very depleted values at 0–5 cm to less depleted values at 5–10 cm soil depth.”

Further, we will modify the discussion (section 4.4, lines 422) as follows to clarify the role of carbonate weathering:

2.20 Autotrophic respiration in forests

Comment: L327: The finding that forests exhibited the lowest average autotrophic contribution (~40%) and were the only land-use type with dominant heterotrophic respiration is intriguing. Given that forests are generally characterized by high productivity, substantial root biomass, and continuous carbon allocation belowground, one might expect a relatively high autotrophic contribution to total soil respiration. Could you please elaborate on the underlying mechanisms that might explain this pattern? For instance, could this be related to deeper rooting systems with less activity near the soil surface (where respiration was measured), slower root turnover, seasonal carbon allocation dynamics, or differences in microbial decomposition relative to root respiration? Clarifying this could help resolve what appears to be a counterintuitive result.

Response: We thank the referee for this thoughtful comment and for highlighting that our explanation of this result requires further elaboration. While we touched on this in sections 4.1 (L371–375) and 4.3 (L393–396), we agree that the underlying mechanisms behind the relatively low autotrophic contribution in forests may deserve a more explicit discussion.

Several factors predominantly contribute to the pattern of relatively high heterotrophic contributions in forests:

• C allocation dynamics: Forests allocate a larger proportion of assimilates to aboveground biomass, and the transfer of recently fixed C belowground occurs with a longer lag compared to grasslands (Schaufler et al., 2010; Gao et al., 2021). This slower and more diffuse carbon transfer may reduce the apparent contribution of autotrophic respiration.
• Rooting system: Compared to grasslands, forests typically have a less dense but deeper rooting system as compared to grasslands (e.g., Jackson et al. 1996). Together, this reduced the relative contribution of autotrophic respiration to total soil respiration and shifts the source partitioning toward a relatively higher heterotrophic contribution.
• Role of the organic layer: Forest soils are characterized by a substantial organic layer, which alone contributed 21–34% of total soil respiration in our study and contributes to an overall high relative contribution of heterotrophic respiration.

Taken together, these mechanisms suggest that the relatively low autotrophic contribution observed in forests as compared to especially grasslands reflects differences in rooting density and depth, C allocation strategies, and the strong contribution of heterotrophic respiration from the organic layer. To clarify this point, we will modify the discussion (section 4.3, lines 394-396) as:

Discussion

2.21 Synthesis

Comment: L343-350: This synthesis provides a strong and concise summary of the main findings. However, this section would benefit from a clearer contrast with existing studies to better position its contributions within the wider body of research. Several of your findings are in line with, or complementary to, previous work, especially regarding carbon age dynamics in grasslands, peatlands, and forests.

2.22 Conceptual figure 7

Comment: Figure 7: In the figure, SOC decomposability and in situ soil CO₂ efflux are positioned at opposite ends of the conceptual spectrum. Could the authors clarify why these two variables, while related, are treated as contrasting rather than complementary indicators of carbon cycling dynamics? This distinction may be counterintuitive, as one might expect systems with high SOC decomposability to also exhibit elevated CO₂ efflux, depending on environmental constraints. This distinction would help readers better interpret the position of each land-use system within the diagram.

Response: We thank the referee for this insightful comment, which indeed points out a key limitation of the current presentation of Figure 7. Our intention was not to suggest that SOC decomposability and in situ soil CO₂ efflux are opposing processes; as correctly noted, they are in fact complementary indicators of C cycling. The figure was designed to highlight that these two variables can behave differently across ecosystems due to environmental constraints.

Specifically, while SOC decomposability reflects the intrinsic stability of SOC (measured under standardized incubation conditions), in situ soil CO₂ efflux represents microbial activity under prevailing field conditions. Environmental factors such as low temperatures in alpine grasslands can strongly suppress microbial activity and thus in situ soil respiration (resulting in a low annual CO₂ flux), letting accumulate readily available SOC. This explains, for example, why alpine grasslands exhibit high SOC decomposability in laboratory incubations but low annual soil CO₂ efflux in the field. Conversely, managed peatlands show relatively low SOC decomposability (indicative of advanced degradation of the organic matter) yet still exhibiting a high in situ CO₂ efflux due to their large SOC stocks and favorable conditions for microbial decomposition.

Thus, the apparent “contrast” between SOC decomposability and in situ soil CO₂ efflux in the figure is not meant to imply that they are conceptually opposed, but rather to illustrate that their relationship depends on site-specific environmental constraints. This interplay provides important insights into whether systems are characterized by C retention (labile SOC not decomposed due to constraints) or C loss (high fluxes despite degradation).

2.23 Parentheses in citations

Comment: L385 and 424: Omit parenthesis when citing the author directly inside the sentence. Apply this change consistently along the manuscript.

2.24 Advanced degradation

Comment: L441-442: Does it need to be an “advanced” degradation? Or could it be that the water table decrease enhances immediate mineralization of previously-locked old C stocks? Would you please add some lines on this idea?

Response: We thank the referee for this comment. We fully agree that a lowering of the water table can immediately enhance the mineralization of previously protected, older C stocks, thereby contributing to the release of old CO₂. This process is certainly part of what we observe. However, our interpretation goes beyond the ¹⁴C signal alone. By additionally assessing SOC decomposability across soil depths, we find that decomposability is particularly low in topsoil layers that have been exposed to aeration the longest. This suggests that much of the labile SOC has already been processed and lost.

Therefore, the combination of (i) Δ¹⁴CO₂ signatures indicating the release of old CO₂ and reflecting the mobilization of long-stored C, with (ii) low SOC decomposability, reflecting the depletion of readily available SOC, points to an advanced stage of peat degradation rather than only a short-term response to water table decline. We will clarify this reasoning in the text around lines 441–442 as follows:

2.25 Seasonal patterns (related to response 2.15)

Comment: Section 4.6. The discussion of seasonal controls on respiration appears internally inconsistent. In line 459 states that “autotrophic respiration is rather driven by plant phenology than by temperature,” yet phenology itself is strongly modulated by temperature and photoperiod. A few sentences later the text mentions the 9 °C air–soil temperature offset in March to explain enhanced autotrophic activity, implicitly re‑introducing temperature as a key driver.

In addition, the Q₁₀ value of 2.4 across all land-use types assumes that soil respiration is always strongly controlled by temperature. However, if autotrophic respiration (from roots and associated organisms) is not mainly driven by temperature—while heterotrophic respiration (from microbes decomposing organic matter) is—then adjusting all respiration rates with the same temperature response could be misleading. This approach might distort the seasonal balance between autotrophic and heterotrophic respiration, which in turn could affect the interpretation of the Δ¹⁴CO₂ values, since these depend on correctly estimating the contribution of each respiration source. I suggest clarifying:

1. Mechanistic rationale – reconcile why phenology can override temperature for roots in winter, yet a temperature‑based Q₁₀ correction is still appropriate for the overall flux. For example, is the assumption that heterotrophic respiration is temperature‑limited while autotrophic respiration is controlled by recent photosynthate supply?
2. Implications for Δ¹⁴CO₂ patterns – discuss whether the Q₁₀ scaling might underestimate or overestimate winter heterotrophic contributions, and how that affects the conclusion that autotrophic respiration dominates in March.

Addressing these points would strengthen the argument and resolve the apparent contradiction between phenology‑driven and temperature‑driven controls on seasonal soil respiration.

Response: We thank the referee for this detailed comment, which allows us to clarify the interpretation of seasonal controls on soil respiration in Section 4.6.

First, regarding autotrophic controls: our intention was not to present phenology and temperature as mutually exclusive drivers, but rather to highlight that in our study autotrophic respiration appeared to be primarily influenced by plant phenology, with temperature still contributing. As the referee correctly points out, phenology itself is modulated by temperature, and both drivers likely interacted to explain the relatively high autotrophic contributions we observed in March. To avoid any ambiguity, we will revise the text in Section 4.6 (lines 459–461) as follows:

“We attribute these findings to a combined effect of plant phenology and temperature. Generally, autotrophic respiration is more strongly linked to phenology than to soil temperature (Atarashi-Andoh et al., 2012), and we likely captured an active state of the phenology in March for grassland and forest sites.”

Second, regarding the Q₁₀ correction (related to response 2.15): We would like to clarify that the Q₁₀ adjustment was solely used to estimate relative contributions of heterotrophic respiration of different depth layers to total in situ heterotrophic respiration. These heterotrophic respiration rates adjusted to field temperature were only used as weighted inputs to the Bayesian mixing modeling to improve model accuracy.

Importantly, this correction was neither applied to the total in situ soil respiration flux nor to the presented SOC decomposability data.

Conclusions

2.26 Forests as preserving systems

Comment: L497 and L28: The classification of forests as "preserving systems" that stabilize carbon could be reconsidered or further nuanced. While the presence of decadal-old CO₂ and dominant heterotrophic respiration may suggest delayed C release, the low SOC stocks across the soil profile and few data points evidencing old Δ¹⁴C signatures raise questions about whether forests in this study actively stabilize carbon, or rather reflect low belowground C allocation with partial stabilization of what is retained. It may be more accurate to describe these systems as exhibiting limited input and slow turnover, rather than implying strong carbon preservation capacity. A brief clarification in the conclusions could help avoid overgeneralization.

Response: We thank the referee for this attentive comment and agree that the term “preserving” can be misleading in the context of C cycling in forest ecosystems – especially in consideration of SOC stocks and stabilization, as pointed out by the referee. In both alpine grasslands and forests, C cycling is slowed down, but because of different reasons. We therefore suggest, to rename the C cycling categorizations of these to land-use types into “retarding”:

• Alpine grassland: Temperature-constrained retarding system --> emphasizes that the slower cycling is mainly due to colder temperatures in these systems.

• Forests: Input-constrained retarding system --> indicates that the slower cycling in forests is mainly due to the older C inputs with lower quality.

Supplementary material

2.27 Labels in figures

Comment: The readers would benefit from having a label with mean values for S3, S4 and S5, since the differences can be very subtle. For example, between total soil respiration (filled circles), weighted heterotrophic respiration (open squares) and autotrophic respiration (open triangles) in S5.

2.28 Atmospheric background at cropland sites

Comment: Figure S3: Please add some explanation on why the atmospheric background in croplands is so negative compared to the other land-use types. This is specially intriguing.

Response: Thanks for this comment. The more pronounced depletion of atmospheric background Δ¹⁴CO₂ in croplands is most likely because of a larger contribution of fossil emissions to the atmospheric signal as cropland sites are located in proximity to highways and larger roads. We therefore add the following sentence to the revised manuscript (result section 3.3, line 291:

“Due to the proximity of cropland sites to fossil CO₂ sources (i.e., highways, larger roads), their atmospheric Δ¹⁴CO₂ values were more depleted than for other land-use types (Fig. S3).”