Review 2023 Tangarife-Escobar et al., Moisture and temperature effects on the radiocarbon signature of respired carbon dioxide to assess stability of soil carbon in the Tibetan Plateau

Tnagarife-Escobar et al. present a well-executed and thoughtful experiment carried out in an important geographic region. I enjoyed reading the work, and find it well argued and well composed for the most part. I would suggest that some additional data be included in the main text (which is now not given at all, or is relegated to the supplementary/appendix materials) to help clarify and strengthen arguments in the discussion. The modeling and empirical components are a bit disconnected from one another as currently presented. If the experimental results cannot be used directly in the soilR simulations, perhaps some of the data may be presented in merged figures to help the reader more directly understand the connections between these two components of the work (see detailed comments below).

Detailed comments:

Abstract line 17: The statement that temperature is a significant variable contradicts the results stated earlier in the abstract.

Lines 56-64: This is a very clear and succinct explanation of C pools and turnover. Nicely done!

Line 110: How did the presence of inorganic C potentially affect the Delta 14CO2 values?

Methods: Soil incubation times… the duration of the incubations is cited many times in the discussion as a potential confounding variable in the interpretation of the Delta 14CO2 data. However, the length of the incubations is not given in the methods. The length of the incubations should be added to the manuscript along with a discussion of how variable lengths of incubations for the individual treatments might have influenced the Delta 14CO2 data. I’m assuming different treatments were incubated for different lengths of time since the methods indicate that they were incubated until a certain amount of CO2 was produced, and given the differences in respiration rates given in the appendix the length of incubation time might have varied by an order of magnitude? Could this have an influence on the age of C being respired (i.e. longer incubation times allowed for decomposition of more structurally and/or chemically "stabilized” substrates)?

Methods: What is the reasoning behind the choice of WFPS values? Ninety-five percent is very high. Doesn’t this value inhibit evolution of gases from the soil matrix? How/why were 65% and 95% chosen?

Table 2: Unclear what is being compared here. Is the anova between grasslands and peatlands at each treatment level of temp/moisture? Or is it comparing different levels of temp within each soil category? It's confusing because the soils weren't radiocarbon dated *after* incubation, correct?

Table 3: This is really a lot of different conditions... and on top of that you discuss the type I, type II or type III systems. How do these three things relate to one another ("fast/slow", "parallel/series", "type I/II/II")? Also, please add to the "System" column "grassland" and "peatland" in addition to "fast" and "slow". I know it's in the text directly below, but it would help the reader keep on top of all the modeling approaches.

Figure 5: I feel that it is important to have an additional two panels in this figure showing the total amount of C respired by each of the treatments for a given length of time or the respiration rates. This information is referenced in the discussion, but I don't see it anywhere. In the discussion, the manuscript makes a point about the relative importance of the age vs. the amount of respired C, so the amounts should be shown. See additional comment regarding appendix table A1 below.

Figures 4 and 5: The 10 deg C thing… something unique seems to be happening at this temperature in the peatland soils. Do you have some explanatory hypotheses? This temperature also has strong outliers in both soil types (Figure 4), would you please comment on this?

Figures 6-9: Please label all these panels of figures as "grass vs peat" and "fast vs slow". Preferably in the figure itself, but at least in the caption. This will help the reader more easily keep track of what they're looking at.

Model/data fusion: Can soilR not use the Delta 14C of the respired C to constrain alpha/gamma and k1/k2 values? Or is that too computationally intensive at this point? I find the paper to be well written, but there is not a lot of integration of the incubation data with the modeling exercise. What would help me understand the connections would be plotting (some? All?) the data from figure 4 onto figures 6-9. This would directly show me how the experimental results map onto the different modeled scenarios. This would really the reader more quickly understand the connections between the type I/II/III systems and the model parameters (gamma/alpha and decomposition rate constants).

Table A1: Ok. Here is where the significant temperature effect is... this looks like a pretty linear response of respiration rate to increasing temperatures. Why not include this result in the main text? I think it's alluded to in the abstract, but there's no actual evidence of a temperature effect included in the main text at this point (maybe add to figure 5).

Conclusions: The conclusions lack punch. What are the broader implications of this work? How do the experimental treatments relate to current climate projects for the QTP? Will the peatlands dry out and change this from a type X to a type X system? Will the grasslands get hotter and therefore respire X more gigatons of C on an annual basis? The introduction states that these soils are being studied because of they hold vast stores of C. What do your experiments suggest for the fate of these C stores under future climate scenarios?

In this study, Tangarife-Escobar and co-authors incubated peatland and grassland soils at four different temperatures and two water-filled pore spaces to better understand temperature and moisture effects on the 14C signature of bulk and respired soil carbon. They also used a mathematical model to analyze how decomposition rates and other soil parameters affect the ∆14C of bulk and respired C and their relationship. Papers that investigate the relationship between carbon age and transit time are crucial in order to better understand carbon destabilization with changing environmental conditions. Although this paper has the potential to be impactful, it needs to be revised to address important flaws in the interpretation of the incubation results and to improve the link between the incubation and mathematical model.

Overall, there are major issues with the interpretation of the incubation results and this is my main concern about this paper. The authors stated that temperature and soil manipulations caused a response in the direction of the ∆14C (Abstract and Lines 233-234); however, these statements are not supported by the statistical analysis, which show that temperature had no effect on the ∆14C of bulk soil or respiration. The authors do not present any information on the statistical model that was used to interpret these results in the methods section, and as such, it becomes difficult to understand how they arrive at some conclusions related to the ∆14C and CO2 flux (e.g., was site included in the stats model? What was the stats model for the CO2 flux?). Additionally, the text in the results section rarely includes wording on whether a main effect was significant, and sometimes the authors will discuss the interactive effect of moisture and temperature on ∆14C, even though the interaction was not significant.

It is also unclear how the incubation informed the SOC decomposition model, for example did the CO2 flux response to temperature and moisture get used in the model to influence the decomposition constants, or were these based on Manzoni et al. 2009 (Lines 182-185)? If the incubation did not influence the model, then what was the reason to include the incubation results in the paper? The authors should expand on the link between the incubation and model and how they influenced each other.

Finally, the writing can be improved and the authors should check the manuscript for typos and inconsistency in terminology. 

Detailed comments:

Abstract

The abstract does not currently satisfactorily connect the results to the big takeaway statements and implications for the net carbon balance of the area.

Line 10: Is this bulk soil 14C?

Line 14: Consider not using terms like k in the discussion of results, but instead refer to the decomposition rates. To a general audience saying what the effect of a ‘low k value’ may not carry a lot of meaning; however, discussing the impacts of ‘low or high decomposition rates’ will aid in the interpretability of the results.

Lines 13-16: It is not clear how these two sentences are linked, or how the first statement leads to the second. It seems like they should be switched: The correspondence between ∆14C and age and transit time strongly depended on the internal dynamics of the soil (k, α, γ and number of pools) as well as on model structure. When decomposition rates were low (low k values), the (replace “modified ∆14C” by the direction in which the ∆14C changed, did it increased or decreased the age of bulk/respired CO2?) due to the incorporation of 14C-bomb in soil (does the incorporation of 14C-bomb mean anything to the reader at this point in the abstract? Consider writing what this means (e.g., the proportion of C cycling on decadal timescales increased). What does this result mean for carbon cycling in wet/dry or cold/wet systems? The abstract is missing the implication of the results.

Lines 16-18: Why are the authors concluding that the stability of carbon depends strongly on changes in temperature if this did not affect the ∆14C of respired CO2 in either soil?

Lines 18-19: It is not clear how modeling improved predictions on interactions between terrestrial and atmospheric carbon.

Introduction

The introduction is a bit choppy. Some of the paragraphs can be combined and the order should be reconsidered to improve the readability of the paper. For example, paragraphs 3 and 6 are both about carbon stabilization/de-stabilization, yet they are broken up by paragraphs 4 and 5.

Please be consistent with word choices, e.g., de-stabilization, destabilization, (de)-stabilization, and the use of carbon or ‘C.’

Please add text in the introduction on how the nuclear weapons test enriched atmospheric 14C to aid in the interpretation of positive vs negative values. 

Line 4: add comma for non-essential clause: “, and in consequence,”

Line 34: Comma before non-essential clause “, which store”

Line 37: Driving the net carbon balance to become a C source? Is it already a source of C?

Line 80-81: Add Pegoraro et al. 2021 to the list of citations for release of old C from deep soil layer after drainage.

Lines 81-82: Is this a perennial or seasonal frozen layer?

Line 86: What is the impetus for the older C increase with increasing moisture? One would think that higher soil moisture would decrease the decomposition of SOC, and thus preserve older C, based on the citations in the previous paragraph.

Line 93: replace ‘their’ by ‘the’

Line 94: The part about the model seems to have been thrown in at the end without much explanation. Why does the model help the interpretation? What are the challenges of understanding age and transit time and how does the model tackle that? The model results are a big part of the results section.

Methods

Please provide coordinates for sites.

How was the soil collected, with an auger, or another method?

The site and soil parameters would be nicer if presented on a table to so both sites can be easily compared.

Line 129: These are analytical replicates, not field replicates?

Table 1 has an exclamation point that seems out of place

The methods are missing information on statistical analyses to discuss the differences in the respired and bulk 14CO2 in the results. This is a big missing component that seriously concerns me, especially since there are issues in the interpretation of the results.

Results

Line 217: Remove duplicate ‘to’ and add comma before ‘which’

Line 219: Was the difference, or maybe lack thereof, statistically significant?

Figure 3: Are these results for all moisture and temperature levels? Why combine them? Having a graph that shows the effect of moisture would aid in the interpretation of the results. Figure 5 is not appropriate to show the significant result of the moisture effect on the 14CO2 signature since it splits it by temperature levels, and there were no significant temperature and moisture interactions.

Line 227: Please add the p-value for the CO2 model and discuss the results as significant or not. The statistical model needs to be added to the methods.

Line 233: It is not clear from the stats results or the graph that temperature and moisture manipulations caused a response in the vertical and horizontal direction in the ∆14C of bulk versus CO2 space. Ecosystem type seems to be the driver of the clustering, please explain how this conclusion was made.

Discussion

Line 291: Please discuss the moisture results as whether they are significant or not for each ecosystem. Additionally, since there was not a significant moisture x temperature interaction, it is not accurate to discuss the moisture effect on different temperature treatments. It is also unclear whether the sites were added in the model structure; therefore, I’m not sure if the moisture effect can even be discussed separately for each site. All of this needs to be addressed in the methods and results sections.