the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 13 Jan 2026
Understanding Soil Organic Carbon Dynamics in Integrated Crop-Pasture Systems: Insights into Deep Carbon
Abstract. In soils from a long-term agricultural experiment (LTE) installed in Uruguay, integrated crop-pasture rotational systems promote greater soil organic carbon (SOC) accumulation than continuous cropping systems at the soil surface (0–20 cm) by avoiding losses of centennial C. Here, we explore whether old C losses due to continuous cropping extend to deeper layers or whether other factors control deep carbon dynamics in the same LTE in Uruguay. To answer this question, we analyzed the vertical profile of SOC, Δ14C and δ13C in fractions representing compartments of different stability at two points in time of two contrasting agricultural treatments (continuous cropping and integrated crop-pasture rotational system). Additionally, using site-specific data we fit dynamic compartmental models to describe the temporal trajectory of C stocks and Δ14C signature in the stable SOC fractions at depth on the integrated crop-pasture system. Based on these models, we used the estimated age distribution for soil C to assess whether it is possible to sequester C in these compartments on time scales relevant in terms of climate change mitigation. We found that the SOC fractions associated with the mineral phase (MAOM - HF) are highly isolated with respect to inputs of recent atmospheric CO2 and that this isolation increases substantially with depth in the soil profile. This degree of isolation was characterized by a progressive increase in the difference between Δ14C profile between MAOM-HF and LF-POM (more labile) fractions with depth. The differences found in the C stock between management systems could be explained by different losses of old legacy C and the high stability and isolation of the compartments associated with the mineral phase in the crop-pasture rotational system. The high stability of MAOM-HF in this agricultural system was reflected in the very old ages of these C pools, from approximately 700 years at surface to a value of several thousand years at depth. The inclusion of a vertical transfer mechanism for previously stabilized material was not necessary to explain the general age structure and the capacity of these systems to sequester new C inputs in deep stable layers. The results of this work imply that integrated crop-pasture rotational systems have not been able to sequester significant amounts of new C along the profile (and in particular at depth), although they are relevant to preserve the natural C legacy of these soils.
- Preprint
(1545 KB) - Metadata XML
-
Supplement
(30 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-6451', Sonja Keel, 12 Feb 2026
- AC1: 'Reply on RC1', Maximiliano González Sosa, 28 Jun 2026
-
RC2: 'Comment on egusphere-2025-6451', Anonymous Referee #2, 01 Jun 2026
The paper by González-Sosa et al. is an interesting study that combines soil fraction measurements, modelling and isotope analysis. The authors conducted their research as part of a long-term experiment in Uruguay, comparing the effects of different land management systems, including continuous cropping and crop-pasture rotation. They observed that the differences in the topsoil due to land management are generally not observed in deeper soils.
I found the paper interesting, and the data are important since deep soil information is generally lacking. The modelling approach also makes sense. I think this study is within the scope of the journal and could be suitable for publication with a few modifications.
First, I think using a paired t-test is not the most appropriate statistical approach for testing differences between soil depths and years, because a t-test assumes the data are independent, which is not the case here. Moreover, by multiplying the tests, you increase the likelihood of accepting H0 by mistake. I suggest performing a multivariate analysis that considers the lack of independence between your data.
Some important details are missing to enable a full understanding of what you did. In particular, it is unclear how you parameterised the model for the modelling.
In the introduction, it would also be interesting to explain that the concepts of POM and MAOM can involve very different protocols (see Leuthold et al., 2024, for example) and therefore leading to different conclusions.
In the introduction, I also suggest adding a few lines on the radiocarbon dead effect and how it can affect the results (see Copard et al., 2025).
On page 9, all the β are βi. I guess they should be β₁, β₂, etc.? ?
When β = 0, there is no transfer from the slow pool to the fast pool, but when β > 0, some of the slow pool is recycled into the fast pool in the layer below. What is the rationale for allowing the slow pool to be recycled in the layer below and not than within the same layer?
In Section 2.4, I understand that this has largely been developed in previous studies, but please add more details to ensure that readers can understand the basics without needing to read other studies.
The bulk density was measured only once. Did you measure it for all the layers? The fact that it has only been measured once introduces some uncertainty into the calculation of the 2008 stocks, and this should be discussed.
In Fig. 5, the MAOM goes up to -17‰, which is high for natural ecosystems. Do you know if any C4 plants have been growing on the site?
Regarding Fig. 7, I don't understand how you fit the profile. Why doesn't the line fit the points?
Copard, Y., Hatté, C., Cécillon, L., Colin, Y., Barré, P., Chenu, C., Cornu, S., 2025. Soil Carbon Dynamics Reshaped by Ancient Carbon Quantification. Glob. Chang. Biol. 31. https://doi.org/10.1111/gcb.70482
Leuthold, S., Lavallee, J.M., Haddix, M.L., Cotrufo, M.F., 2024. Contrasting properties of soil organic matter fractions isolated by different physical separation methodologies. Geoderma 445, 116870. https://doi.org/10.1016/J.GEODERMA.2024.116870
Citation: https://doi.org/10.5194/egusphere-2025-6451-RC2 - AC2: 'Reply on RC2', Maximiliano González Sosa, 28 Jun 2026
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 1,246 | 581 | 93 | 1,920 | 126 | 140 | 137 |
- HTML: 1,246
- PDF: 581
- XML: 93
- Total: 1,920
- Supplement: 126
- BibTeX: 140
- EndNote: 137
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
"general comments"
In this study González-Sosa et al. measured the effect of an agricultural management change on a large set of variables related to soil organic carbon in a long-term experiment in Uruguay established in 1963. At two points in time (2008 and 2021), samples were collected at 10 to 20 cm intervals down to 80 cm depth. Two types of fractionations were used (physical and density), and for each fraction the C content, the stable C isotope ratio and radiocarbon was measured. The results were used to test whether vertical transfer of C explains the age distribution in the profile using a model. The main conclusion of this study is that introducing grassland in a crop rotation leads to lower C losses at depth.
A lot of hard work has been invested in this study and the results presented are interesting and novel. Subsoil organic C and 14C analysis are still rare although they both provide important information to understand the soil C cycle. However, I have three main points to consider. Overall, the article is well written. Some sentences are a bit long and you may want to consider splitting them.
"specific comments"
"technical corrections"
L12: Based on your SOIL article this is not true. There is no accumulation of C, but a prevention of C losses.
L63: “aggregate formation”
L74: please double check if the name is correctly spelled.
L78: “agricultural sequences” are usually referred to as crop rotations.
L79: I think it would be good to add a reference here.
L81: If pastures are in a rotation they are typically referred to as leys. Since your grassland includes clover, you could refer to it as “grass clover ley”.
L105: please replace by “conservation management”
L155: I understand that this has been described elsewhere but it would be helpful to add just a bit more (everything that is relevant for C inputs): what type of fertilizer (organic/synthetic), residues left on the field or not.
L168: Add latin names of crops.
L172: Could the historic changes affect C inputs? If yes, then I think it would be important to mention this
L295 : I would rephrase this sentence as follows: “the SOC stock of the R system in the 20-80 cm layer (80.66 Mg ha-1) was 46% higher than that of the CC system (p value < 0.010)”
Fig. 3a: This might be a detail, but I honestly had difficulty understanding the figure(s) at first. I suggest changing the order of the dark and light grey label explanation (first the older one, then the newer one when reading from left to right). Additionally, please consider changing the order of the treatments so it aligns with the figure. Furthermore, the letters indicating the statistical results are barely visible and the tests are not mentioned in the legend. The same holds for Fig. 3b (and others).
Fig. 5: Text on the right is too small.