the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 27 Jun 2025
Rates of soil organic carbon loss from rainforest to pasture conversion at a deforestation hotspot in the Amazon basin
Abstract. Deforestation in the Amazon basin drives critical losses of soil organic carbon stocks (SOCs), but the impacts of pasture conversion in tropical hotspots of deforestation remain poorly quantified in soil depths and time. Using the chronosequence approach, we sampled mature forests and pastures from 1 to 30 years old to address variations in soil bulk density (SBD), soil organic carbon concentration (SOCc) and stocks (SOCs). Mass spectrometry of the isotope ratio (IRMS) was used to obtain δ13C values and estimate the input of forest and pasture SOC to the total pool. The rainforest soil (0–30 cm) stored 83.328 Mg C ha-1, after 30 years of pasture conversion, SOCs declined by 21 % (66.061 Mg C ha-1), while SBD increased 15 % (1.215 to 1.397 g cm-3). Soil carbon turnover was depth dependent: forest-derived SOCs loss rate at 10–20 cm (0.112 year-1) was twice as fast in the upper horizon (0–10 cm; 0.063 year-1); simultaneously, pasture-derived SOCs gain was faster in the upper 10 cm (0.055 % year-1) reflecting the change from forests deep roots to a pastures shallower system. After 30 years of pasture conversion, forest-derived SOC still represented 19.6 % of the total pool in the 0–10 cm soil horizon, highlighting the important effect of previous vegetation.
- Preprint
(927 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-2959', Anonymous Referee #1, 30 Jun 2025
- AC1: 'Reply on RC1', Valentina Lara, 14 Nov 2025
-
RC2: 'Comment on egusphere-2025-2959', Anonymous Referee #2, 19 Nov 2025
The manuscript comprises the results of an extensive soil sampling in the Colombian Amazon, which aims to understand SOC dynamics after deforestation in a time-for-space substitution approach. The topic is timely and aligns well with the journal’s scope. The dataset is robust, and the analytical methods are generally sound. Yet, the manuscript does not formulate true hypotheses nor conceptually grounded aims (l. 58ff.). Instead, the questions posed remain largely descriptive and are not connected to the processes or mechanisms that would explain the observed patterns. As a result, the study describes what is being examined but not why these patterns should occur. Without clearly articulated hypotheses, it is not possible to evaluate whether the results support or contradict the authors’ intended framework. Hence, we suggest to the editor very major revision, but see high potential for improvements.
Overall, the dataset has strong potential to contribute to the understanding of land-use change and SOC dynamics in tropical systems. Yet, in its current form, the manuscript remains underdeveloped and does not reach the conceptual depth needed for publication. It gives the impression of being somewhat premature, and parts of the analysis read not yet sufficiently elaborated. There appears to be substantial expertise within the author team that, if more fully integrated, could help enhance the conceptual framing and deepen the interpretation of the results.
The introduction and conclusion briefly mention SOC changes through time, but the temporal trajectory is not sufficiently analysed or discussed. We have some reservations about whether the sampled depths are sufficient to capture or explain effects of deforestation on SOC dynamics with depths, given that trees can root several meters deep while the sampling was limited to 30 cm, with detailed analyses conducted on 10 cm increments. Although the authors have clearly invested considerable effort in data collection and analysis, the results and discussion do not provide sufficient interest, some interpretations appear speculative and the line of argument is not yet fully convincing. The language is vague in many places. Also, a careful refinement of the title could help guide readers more effectively and add to the overall clarity of the manuscript. Overall, the study has promising elements, and we encourage the authors to undertake a substantial revision to enhance the clarity, depth, and interpretive strength of the manuscript before it should be considered for publication.
Specific comments
The abstract is generally well written. Our main recommendation would be to rephrase the first sentence: as currently written, it oversimplifies the complex and sometimes paradoxical evidence on SOC dynamics following forest loss in rainforests (Fujisaki et al., 2015), as well as the substantial body of existing research on this topic. We encourage the authors to include more on the background in which this study is framed, ensuring it reflects both the current state of knowledge in this area and the general aim of this study.
L. 22 “fragile”: The fragility framing of the Amazon ecosystem may not fully capture its dynamics (Flores et al., 2024; Longo et al., 2018; Sakschewski et al., 2016). The system is broadly resilient, yet certain disturbances, particularly extensive forest conversion, can increase its vulnerability to critical thresholds
L. 22f. “[they] play a key role in regulating the global carbon cycle”: Specify how. The link between local soil processes and global carbon dynamics should be described more clearly to avoid sounding overly general.
L. 27f. “it breaks the equilibrium of C balances in this important ecosystem”: Please rephrase to more accurately reflect the processes that occur following forest loss. The current phrasing oversimplifies the issue. Don’t hesitate to include more detailed information where appropriate.
In Material & Methods: Please specify the climate classification of the study region to provide clearer environmental context. Also, is potentially small- and large-scale agriculture an important driver for land-use change or is it only cattle ranching? Do you have information on the pastures’ management or grazing intensity? Are they still being actively used?
L. 69 “the hottest hotspot of deforestation” – why is this important in the context of this study? It also is stated in the manuscript’s title but we don’t fully get why this information is important in the context of this study
The Results section should emphasize stronger key findings rather than describe data presented in figures and tables; latter should support the results...
L.182ff: It would be helpful to clarify if the loss of the litter layer after slash-and-burn affects the comparability of soil depth measurements, rather than active compaction in 20-30 cm?
L. 216ff: the argumentation that forest SOC is greater lost at 10-20 cm due to deeper roots being lost is quite contradicting, considering that tress can root up to several meter but this study only has data up to 30 cm. Can you explain why specifically in 10-20 cm (and not 20-30 cm) this effect is more pronounced in the context of root losses?
L. 229ff: What about potential lag effects? Or generally slower processes? (Stahl et al. 2016)
Very well written conclusion. We particularly appreciate the flow, coherence, and the way the summary of findings connects to future directions in research.
Technical corrections:
- Throughout the manuscript one site is differently named: “mature forest”, “forest”, “rainforest”. Please stick to one to avoid confusion.
- We would suggest avoiding vague and subjective qualifiers like “important” (e.g. l. 20, l. 27, l. 239, l. 267) providing more concise explanations of the arguments.
- Please be more consistent with the use of abbreviations (e.g. C, SOC, … ) throughout all the text.
- L. 12 “pastures from 1 to 30 years old” – Please rephrase that this is the “time since conversion from forest” or “time since pastures have been established in this specific area”.
- L. 14f. “The rainforest soil ….” – this sentence is quite chaotic. Maybe wrong punctuation?
- L. 18ff. We really liked this conclusion! Please provide some further perspectives on the meaning and contribution of these findings in a broader context.
- L. 28ff: “For example…”: Please rephrase to “Between 1990 and 2000, it was recorded that the carbon sink decreased by 30% in the Amazon forest (references), and could have even turned into a net C source in the SE Amazon between the years 2010 and 2020 (references).” – also is this value of 30% related to the Amazon forest soil, the Amazon forest ecosystem?
- L. 30ff “However, ...”: Please include something in the direction of “Carbon balances are mostly based on ecological approaches including allometric equations to estimate C storage for e.g. policy decisions (REDD+), but that they are also oversimplifying processes above- and belowground in regards to the soil-atmosphere interface (e.g. also C loss as CO2 or CH4)”.
- L. 31ff “… , which can even exceed the total amount of C stored in the AG counterpart”: Upon reviewing Malhi et al. (2009), it appears that Fig. 2 may not fully support the point being made, so another reference might strengthen this section. We also noted a slight inconsistency in discussing the role of carbon stored at 2–3 m depth while basing the analysis in this study on samples collected only to 30 cm.
- L. 36. Please change to “In the tropics of Asia, Africa and South America, …” – please use parentheses more conservatively. Also, the claim is quite general (all soils in all the tropics change by 14%?) and the content of this sentence is quite similar to l. 40ff. “In the Amazon basin…”. Avoid these repetitions and redundancies.
- L. 42f. “These findings…” – again, too general
- L. 45 “high variability” – in space or time? Or both?
- L. 46ff. The explanation of the methodology of isotopic signatures is redundant here – it should be added to the Material & methods part – and the second part of this paragraph (l. 53ff) reads more like a discussion.
- L. 75: Ferralsols, Oxisols, Acrisols and Ultisols start with a capital letter.
- Fig1: Please translate the legend into English. What the yellow, beige or green areas in the map?
- L. 81f “Natural terra firme forests were established as the …”. Replace “were established” by “were set”.
- L. 82f. “personal communication...”: Is there any chance you can confirm this information via satellite data (e.g. historical imagery)?
- In Chapter 2.3: Was inorganic carbon present or not? If not, please add that total C equals organic C, if yes, did you remove it before or how did you handle inorganic C?
- L. 95: change “sede” to “campus” – if you agree with the translation.
- L. 101ff. “Soil samples…”: This steps’ goal is unclear to us.
- L. 110f. “… d is the distance from the surface of the soil surface (in m)” please change to “d is is the soil thickness [m]”?
- In Chapter 2.4: If applicable, please add references of equations 1 to 4.
- L. 114ff: Did you meet all assumptions of normality and homogeneity of variance for the ANOVA with your dataset? If yes, state so in text.
- L. 136: delete “during the 30-year post-deforestation period”.
- L. 138: change “SBD at” to “SBD to” & “fell” to “decreased”.
- 139: add a comma before “respectively”.
- Fig 3: Make error bars as solid lines and slightly dodged, and lines between the years dashed. Or do you assume a linear trend between the years? Improve image quality to 600 dpi, in case it is not specified otherwise by the journal.
- L. 162: change “was higher on” to “was lower in”. Please check results with Tab.2 – they mismatch.
- L. 163: change “were key in estimating” to “allowed the estimation of”
- L. 163ff: Tab 2 shows that actually is 10-20cm the depth with the greatest forest-derived C loss and not 20-30 cm (!) - please check this!
- Table 2: Equations of fitted models are not equivalent to Equation 4. Please check the operators between summands - also if models were then calculated correctly. Also, for better general understanding, please add what k and r describe in the tables’ description and metadata for reproducibility in e.g. the supplements.
References
Flores, B. M., Montoya, E., Sakschewski, B., Nascimento, N., Staal, A., Betts, R. A., Levis, C., Lapola, D. M., Esquível-Muelbert, A., Jakovac, C., Nobre, C. A., Oliveira, R. S., Borma, L. S., Nian, D., Boers, N., Hecht, S. B., ter Steege, H., Arieira, J., Lucas, I. L., … Hirota, M. (2024). Critical transitions in the Amazon forest system. Nature, 626(7999), 555–564. https://doi.org/10.1038/s41586-023-06970-0
Fujisaki, K., Perrin, A.-S., Desjardins, T., Bernoux, M., Balbino, L. C., & Brossard, M. (2015). From forest to cropland and pasture systems: A critical review of soil organic carbon stocks changes in Amazonia. Global Change Biology, 21(7), 2773–2786. https://doi.org/10.1111/gcb.12906
Longo, M., Knox, R. G., Levine, N. M., Alves, L. F., Bonal, D., Camargo, P. B., Fitzjarrald, D. R., Hayek, M. N., Restrepo-Coupe, N., Saleska, S. R., da Silva, R., Stark, S. C., Tapajós, R. P., Wiedemann, K. T., Zhang, K., Wofsy, S. C., & Moorcroft, P. R. (2018). Ecosystem heterogeneity and diversity mitigate Amazon forest resilience to frequent extreme droughts. New Phytologist, 219(3), 914–931. https://doi.org/10.1111/nph.15185
Sakschewski, B., von Bloh, W., Boit, A., Poorter, L., Peña-Claros, M., Heinke, J., Joshi, J., & Thonicke, K. (2016). Resilience of Amazon forests emerges from plant trait diversity. Nature Climate Change, 6(11), 1032–1036. https://doi.org/10.1038/nclimate3109
Stahl, C., Freycon, V., Fontaine, S., Dezécache, C., Ponchant, L., Picon-Cochard, C., Klumpp, K., Soussana, J.-F., & Blanfort, V. (2016). Soil carbon stocks after conversion of Amazonian tropical forest to grazed pasture: Importance of deep soil layers. Regional Environmental Change, 16(7), 2059–2069. https://doi.org/10.1007/s10113-016-0936-0
Citation: https://doi.org/10.5194/egusphere-2025-2959-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 637 | 91 | 15 | 743 | 20 | 26 |
- HTML: 637
- PDF: 91
- XML: 15
- Total: 743
- BibTeX: 20
- EndNote: 26
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
In this manuscript the authors present a relative straightforward study conducted in South East Colombia in which they use the chronosequence approach to assess changes in soil bulk density, soil C concentration, soil C stocks and origin of soil C based on 13C isotope analysis. They show that the soil C stocks declined over time, but surprisingly forest-derived soil C stocks losses were faster in the 10-20 cm depth interval than in the 0-10cm depth interval whereas the pasture derived soil C stocks increased faster in the 0-10 cm depth interval, which they explain with the change from deep roots in the forest to shallow roots in the pastures.
The strength of this study is that it was conducted in the field, based on a careful site selection with sufficient replication. It was conducted in an understudied area of the Amazon basin and as such it is important that these data become available to the scientific community. However, there are some critical points, that I think the authors should address before I can recommend publication.
-As is mentioned by the authors in the discussion (l. 202) soil texture can have a major influence on decomposition rates. Unfortunately, soil texture is never mentioned and may have had a strong influence on the results. The authors mention that soil were Ferralsols or Acricols (l. 75). Whereas these soils all have in common that they are highly weathered with a low CEC they may have substantial difference in soil texture, especially since Acrisols have undergone clay translocation. The best solution for this problem is if you can analyze back-up sample for soil texture (sand-silt-clay). At a minimum you should convince the reviewers that there were no significant differences in soil texture among the sampling sites.
-To assess forest- and pasture-derived C normally you also need an 13C values of the ‘endmembers’ (forest litter and pasture litter). If I am not mistaken, this is not mentioned. Please give the ‘pure’ signatures of C3 and C4 that you used and explain how you assessed them.
-In your assessment you assume that the forest was replaced with a 100% C4 pasture. However, in my experience, pure C4 pastures do not exist, especially in frontier areas like where you worked. So the C-input in the pasture, will be at least partly be from C3 plants and this contribution may even increase with pasture age if a pasture degrades and C3 bushes start to grow. I think this may be the main reason why you actually found slower forest-derived SOCc loss in the top 10 cm compared to the 10-20 cm depth interval: you assume that all C3 carbon in the top 10cm was forest-derived, but part of it is probably from C3 herbs or bushes in the pastures. There is no simply way to fix this, but you could do an analysis in which you make assumptions about the C3 carbon input and how this may have affected your results.
-I noticed in Figure 1 that there are irregular yellow areas that have no pasture age. Could you explain what these areas are?
-I did not see the original data (maybe I missed them). However, I think it is very important that these data will become available without restrictions and I encourage you to include them in an appendix or in a data repository. I know that you write that they will be made available upon request, however, it is possible that the first author cannot be reached in some years which would make the data unavailable.