1. This study investigates the effects of whole soil warming on the decomposition of root litter by depth in the soil profile in Blodgett Experimental Forest. The authors found, after 3 years of +4 degree C warming, distinct depth-dependent decomposition of labeled root litter where warming accelerated the topsoil root litter, but not the subsoil root litter. This is an interesting and novel contribution to the field. 

Thanks for your constructive, critical comments, and thorough review of our manuscript.

2. The introduction and results are clear and well-structured, but the discussion is very long. 

We agree and will condense and simplify the discussion.

3. Please acknowledge the limitations in study design regarding the large spatial heterogeneity of soil properties (plus microbial community differences and conditions at depth) at this site and the low sample size of n=3. C inputs (amount and type) and the effects of warming are expected to differ by depth in the soil profile, yet the study design does not account for this. Since there was increasing variability in microbial communities with depth, this should be more directly addressed since the main conclusion of this work is the depth-dependent responses. 

We agree, soils in general have a large spatial heterogeneity, in our case especially in subsoils. This high spatial heterogeneity is shown in our study with large standard errors and further reflected by the large standard errors of subsoil PLFA analysis reported by our sibling paper (Pegoraro et al., 2025, in review in Soil Biology and Biochemistry). We mentioned this problem for example in lines 605 and 606, but we will explicitly emphasize the limitation of our study design in the Method & Material and in the Discussion section.

4. The use of a single root litter type from an annual grass is also a limitation, as this coniferous forest can be expected to have root contributions from fine roots of conifers which should have different chemical composition like lignin and lipids. Please justify the use of wild oat roots instead of conifer roots.

We completely agree that coniferous tree fine roots could have been a better substrate for such an in-situ decomposition experiment. Uniform ¹³C-labelling of slowly growing trees is technically extremely challenging, but can be achieved with fast growing wild oat. Therefore we used wild oat as a model substrate, also in several fine root decomposition experiments (Castanha et al., 2018; Hicks Pries et al., 2018), and can compare reuslts.

5. If the natural temperature gradient was maintained, there may still be artifacts of the heating coil that are not accounted for, like differentials in soil drying by depth and consequent influence on decomposition dynamics at different depths. There could also be differences in microbial community distribution in close proximity to the heating coils vs. further away.

This argument is absolutely right. The soil moisture was monitored throughout the whole period of experiment across soil profile and its natural gradient was kept (Pegoraro et al. (2025, in review, Hicks Pries et al. (2017). The incubation experiment was conducted with a consistent distance to the heating rods in all the plots, and the warming magnitude was maintained.

6. Measuring after 3 years does not capture short-term priming effects, which would be expected more immediately than 3 years later. Please take this into account when addressing priming in the results and discussion sections. For example, the statement on L439, “Thus, positive priming occurred,” from my perspective, cannot be so definitive. 

Agreed. We will modify this argument based on this comment.

7. Regarding statistical analysis and model selection, LME and AIC were used to assess the best fit models but it is not directly stated which models were compared. This could be added in a supplementary section. It is also unclear whether the depth, temperature, and their interaction were modeled as fixed or random effects in all cases.

We will clarify this and report different model structures in the supplementary material.

8. The study suggests that microbial activity was higher in ambient subsoil compared to warmed subsoil, based on the accumulation of mid-length fatty acids, but if microbial activity was higher in ambient conditions, one would expect greater decomposition of SOM and added root material. Instead, results suggest greater root-litter preservation in ambient plots. Perhaps there are alternative explanations for the accumulation of fatty acids, such as selective preservation, microbial necromass accumulation, or sorption to mineral surfaces.

Agreed and thanks for the suggestions. We wanted to emphasize that the accumulation of mid-length fatty acids was a sign of added labelled root-litter being more incorporated in microbial biomass. And the mass change of labelled-root derived hydrolysable lipids is lower in warmed plots compared to ambient plots (Figure 3) except for fatty acids. We will avoid using the term “microbial activity” because this was not directly measured by our study.

9. While bulk root-litter decomposition was not significantly different between warmed and ambient plots, lipid composition changed, and fatty acid accumulation occurred under ambient conditions. The authors state that subsoil decomposition was unaffected by warming, but this contradicts their molecular marker results showing that microbial metabolism and decomposition pathways did shift.

Indeed, the composition of hydrolysable lipids changed with warming. We speculated that the accumulation of mid-chain fatty acids occured because of root litter being incorporated in microbial biomass as suggested by PLFA data (Pegoraro et al. 2025, in review). However, we cannot exclude that our analysis also measured processed or microbial necromass which could bias our interpretation, and studying microbial metabolism is beyond the scope of this study.

Specific Comments: 

10. L76-77: Unclear what is meant by “harnessing roots”

We will briefly explain the definition of “harnessing roots”.

11. L113-114: It is unclear if soil depths were heated the same amount or not from this sentence. I recommend explaining what the natural temperature gradient is rather than refer to another paper. 

We will explain the temperature gradient briefly in this section.

12. L121-146: The coring system, excavation and root additions are a little confusing. For each depth, was a soil core extracted, then the labelled roots added to the hole, then the soil replaced for that depth? Or was the soil inside the core, plus the core itself, left in the hole for the duration of the experiment?

Sorry for the confusion. Briefly, the soil core was added back to the drilled hole for the duration of the experiment. The customized soil core (polycarbonate) was hammered together with an aluminum tube outside of the soil core. The soil core contained 4 parts (0-10, 10-45, 45-85, and 85-95 cm). The soils in the latter three soil core parts were removed and mixed with labelled root and then added back to the corresponding part. Then all the parts were threaded on each other and added back to the soil.

13. L148-150: Why were those specific depths chosen? Is it because of the known rooting depths of conifers in that forest? If so, this should be described in the site description section at the top of methods.

We will add this information to the Method & Material sections.

14. L222: Specify what is meant by region-specific in this context.

We will do as suggested.

15. L322: The figure text is very small and hard to read. Instead of separating the ambient and warmed graphs, it would be better to have the sets of bars next to each other for direct comparison (e.g., ambient and warmed 10-14 cm, ambient and warmed 45-49, etc.). 

We will enlarge the text and group the bars by depth.

16. L339: Line about the error bars is not needed in the text since it’s in the figure captions.

We will delete the sentence.

17. L345: Was this difference statistically significant? Specify either way. 

We will specify it.

18. L467-468: For clarity, instead of “This argues for co-metabolic decomposition of the added root litter,” “this indicates…”

We will clarify this in more detail.

19. L642-644: What is meant by “the warming was heterogeneous” in this sentence? 

We wanted to say the warming effects on root-litter decomposition were heterogeous in subsoil mainly due to its spatial heterogeneity, much less substrate, and much lower microbial abundance. We will clarify.

Technical Corrections: 

20. L46: missing word: “…biotic factors THAT could change…”

We will add the missing word.

21. L50: grammatical errors: “Moreover, roots impact on SOM dynamics in subsoil in two way:” 

We will modify the error.

22. L50-52: What is meant by “They are more likely to form stable SOM to aboveground plant biomass”? 

It means compared to aboveground plant residues, roots or belowground plant residues are more likely to be associated with minerals.

23. L54-55: Grammar revisions needed.

We will modify it.

24. L68-69: “Besides” is an awkward way to start a sentence.

We will delete it.

25. L 70: grammatical errors

We will revise the sentence.

26. L71: Missing the word “the”

We will add it.

27. L96-97: Revise second hypothesis for clarity and maintain consistency in tense used. Relative accumulation to what?

We will revise to keep consistency and be more precise.

28. L176: Write out the word dichloromethane for clarity and consistency with other acronyms.

We will do it.

29. L188: Remove extra space after min

We will remove it.

Reference

Castanha, C., Zhu, B., Hicks Pries, C. E., Georgiou, K., and Torn, M. S.: The effects of heating, rhizosphere, and depth on root litter decomposition are mediated by soil moisture, Biogeochemistry, 137, 267–279, https://doi.org/10.1007/s10533-017-0418-6, 2018.

Hicks Pries, C. E., Castanha, C., Porras, R. C., and Torn, M. S.: The whole-soil carbon flux in response to warming, Science, 355, 1420–1423, https://doi.org/10.1126/science.aal1319, 2017.

Hicks Pries, C. E., Sulman, B. N., West, C., O’Neill, C., Poppleton, E., Porras, R. C., Castanha, C., Zhu, B., Wiedemeier, D. B., and Torn, M. S.: Root litter decomposition slows with soil depth, Soil Biology and Biochemistry, 125, 103–114, https://doi.org/10.1016/j.soilbio.2018.07.002, 2018.

Pegoraro, E., Zosso, C. U., Wiesenberg, G. L. B., Castanha, C., Hicks Pries, C. E., Porras, R., Soong, J. L., Schmidt, M. W. I.,  Torn, M. S.: The depth-dependent microbial response to root litter input in an experimental whole-soil warming study, 2025, under review.

This paper leveraged the Blodgett Forest whole-soil-profile warming experiment in a mixed coniferous temperate forest to examine warming effect on the decomposition of root-derived carbon, which serves as the primary organic inputs to soils. The study employed two approaches for this purpose by examining the lipid biomarkers of roots (suberin) in soils at three depths (10-14, 45-49, and 85-89 cm) after three years of warming treatment and via three-year in-situ incubation of 13C-labelled grass root-litter at each depth. The authors found that the decomposition of added root-litter was only accelerated in the topsoil (10-14 cm) but not in the subsoil (45-49 and 85-89 cm) with warming. Hence, the impact of warming on the decomposition of root-litter in a temperate forest is depth-dependent. Overall, the authors employed novel and complementary approaches to examine how root carbon decomposition respond to warming, which has significant implications for understanding how soil carbon cycling may be altered by climate change. The application of root-specific biomarkers and compound-specific 13C analysis in investigating root carbon turnover deserves applause. I have a few relatively minor comments/suggestions to improve the readability and to strengthen the conclusions of the paper.

We are grateful for your insightful comments and comprehensive review, which have greatly helped us improve the manuscript.

First, quite some of the results were not statistically significant (including the PE—I would say that no significant PE was induced). Please avoid overstating the results, which can be confusing. Some of the discussions are pure speculations without much data support. Please reduce these as well. I find some of the results repetitive, which can be made more succinct.

We will report statistical significance power and avoid overinterpretation, reduce speculations, and simplify the discussion in the re-submitted version.

Second, as the authors mentioned, the dose of added roots was different on SOC basis for different soil layers. How would you expect it to influence the results? Can you specify? For instance, subsoil root decomposition may be underestimated due to the high dose of carbon added?

It’s a very critical point. We will specify this and discuss the impact of the same amount of roots added to different soil depths on our results. It could also be the duration of the experiment masking the difference of root litter decomposition in the short-term.

Third, the application of root-specific biomarkers and compound-specific 13C analysis in investigating root carbon turnover deserves applause. How do you expect this approach to be used in the future? How would you recommend to improve its application? I would love to see the authors comment on this, which is a novel aspect of the study and worthy of further application.

Thanks for the comment. We will discuss this in the re-submitted version.

Additional detailed comments below:

1. Line 51: “than”, not “to”.

We will modify.

2. 54: …debate continues, on how…

We will modify.

3. 96: The working hypotheses can be better refined. The second one is not really a hypothesis (it’s known, right?).

We will re-phrase  the second hypothesis.

4. How much did the examined lipids contribute to the added OC (with 13C labels)? Did the percentage change under warming?

That’s a very good point. We will add this.

5. 505: more slowly.

We will modify the grammar mistake.

1. The low sample size of only n=3 and the large variation of the subsoil data (such as the extremely wide ¹³C-excess error bar of 85-89 cmin Figure 2) may mask the true effect of warming. It is suggested to explain the statistical power (such as post-hoc power analysis), or discuss the influence of small samples on the conclusion.

Thank you for the detailed review and valuable feedback on our manuscript. We will add detailed information of linear mixed effect models in the supplementary material and discuss the limitation of small sample size.

1. This study used the root systems of annual grasses (wild oats) instead of those of local dominant coniferous trees (such as pine trees). The lignin content of wild oats is low and the C/N ratio is low. The decomposition rate may be faster than that of woody roots, which may overestimate the effect of warming on the carbon loss of topsoil. It is suggested to discuss this limitation or supplement the control experiments on coniferous tree roots.

We are fully aware of and will discuss the limitation using wild oat as substrate in our study. To facilitate cross-study comparisons of fine root decomposition, we selected wild oat as a standardized model substrate (Castanha et al., 2018; Hicks Pries et al., 2018). This approach also circumvents the technical difficulties associated with achieving uniform ¹³C labeling in coniferous species, which are more challenging than labeling the fast-growing wild oat roots and will in theory take several years.

1. Heating cables may cause soil moisture gradients (such as subsoil drying), but the influence of temperature increase on the moisture content of each soil layer is not quantified in the text (Line 116). It is suggested to supplement the monitoring data of soil temperature and humidity or discuss the possible impact of heating on the habitat of microorganisms.

This is a very constructive comment. The soil moisture was monitored in the duration of the experiment. The data is available in our sibling paper (Pegoraro et al., 2025, under review in Soil Biology and Biochemistry) and the impact of changing soil moisture on microbial biomass (PLFA) was also being discussed in that paper.

1. The paper mentions the selection using the Linear Mixed Effects Model (LME) and the AIC model, but does not clearly state the specific structures of fixed effects (such as warming and depth) and random effects.

We will add the structure of different models in the supplementary material.

1. It is claimed that "there is no significant primingeffect" (Line 439), but Figure 4a shows that there is negative excitation in the subsoil (inhibiting the decomposition of primary carbon). It is necessary to clarify the statistical test results (pvalue), or modify the expression.

Sorry for the confusion. We will add the statistical power in the results.

1. Warming in the subsoil did not change the total carbon content of the root system (¹³C recovery rate), but molecular markers indicated changes in microbial metabolism (such as fatty acid accumulation). It might be due to the increased input of microbial residues (PLFA contribution), or the enhanced physical protection of subsoil carbon (such as mineral binding) caused by warming. It is suggested to discuss the impact of changes in community structure in combination with the existing microbial data.

We will discuss this based on the results of Pegoraro et al. (2025) in the resubmitted version.

1. The enrichment of C16-C18 fatty acids in the subsoil (>100% initial amount) may result from the input of microbial membrane lipids, but the interference from plant sources has not been ruled out. It is suggested to distinguish the contributions of microorganisms and plants through δ¹³C-PLFA analysis.

In Pegoraro et al. (2025): 1) root addition increased microbial biomass in ambient but not in warmed subsoils. 2) the very high ¹³C-excess of PLFA in all microbial groups identified (actinobacteria, fungi, Gram–, and Garm+) indicated that labelled root litter had been incorporated in microbial biomass. However, it is very difficult to distinguish the contribution of plant and microorganisms to mid-length fatty acids in our study, and this is why usually we don not consider these compounds as specific biomarkers both for suberin and PLFA.

1. Figure 3 cannot visually compare the differences between ambientand warmed. It is suggested to change the presentation form of the chart.

We will group the bars by each depth for better comparison.

1. The results of the primingeffect in Figure 4 need to be marked with statistical significance (e.g. * p < 05).

We will add statistical significance with asterisks when there is.

1. Avoid overinterpretation (e.g. Line 439 "Thus, positive priming occurred").

We will delete this sentence to avoid confusion.

1. The 3-year experiment may have failed to capture the short-term excitation effect or the delayed response of the bottom carbon pool. It is suggested to discuss the necessity of long-term observation.

We will discuss this.



Reference

Castanha, C., Zhu, B., Hicks Pries, C. E., Georgiou, K., and Torn, M. S.: The effects of heating, rhizosphere, and depth on root litter decomposition are mediated by soil moisture, Biogeochemistry, 137, 267–279, https://doi.org/10.1007/s10533-017-0418-6, 2018.

Hicks Pries, C. E., Sulman, B. N., West, C., O’Neill, C., Poppleton, E., Porras, R. C., Castanha, C., Zhu, B., Wiedemeier, D. B., and Torn, M. S.: Root litter decomposition slows with soil depth, Soil Biology and Biochemistry, 125, 103–114, https://doi.org/10.1016/j.soilbio.2018.07.002, 2018.

Pegoraro, E., Zosso, C. U., Wiesenberg, G. L. B., Castanha, C., Hicks Pries, C. E., Porras, R., Soong, J. L., Schmidt, M. W. I.,  Torn, M. S.: The depth-dependent microbial response to root litter input in an experimental whole-soil warming study, 2025, under review.