The article “Controls on heterotrophic soil respiration and carbon cycling in geochemically distinct African tropical forests soils” investigates the role of soil chemistry, fertility and geochemical composition as drivers of soil respiration under laboratory conditions in soils collected along slope gradients in tropical Africa. The article fits the scope of the journal and it will be of great interest for the journal readers.

The introduction is to large extents well structured, though I am not sure if the sub-headers are really needed. The introduction contains a lot of information on geochemical (e.g. Al Fe SiO2) parameters influencing soil C dynamics, but not so much on available P or N, which turn out in this study to be strong determinants of soil respiration. Moreover, little information is given in the introduction on the role of aggregation, or microbial biomass as C sequestration ‘pump’ (or also about microbial enzymatic control). I would recommend to extend each of the topics a bit more to make the introduction to make the link even stronger towards the research questions and to the results presented. In addition, I think the hypotheses could be more specifically state, e.g. in line 93 to 96 it would be helpful to mention which change in geochemical properties would cause which response by microbial decomposers more ‘explicitly’, by stating the expected mechanism, or by hypothesizing under which conditions faster or slower soil C turnover could happen, and how this has been influencing soil C stocks in the long term.

The material and methods provide a detailed characterization of the study sites and the respective soil properties including many references to articles that are currently in review, which is a bit difficult to trace. The incubation experiment set-up is very clearly described and sound. Also the statistical analysis is provided in detail, which is great. One minor point that I could suggest to improve the role of soil depth would be to explore linear models and include soil depth ‘nested’ into topographic position and geochemical region (or nested per sample location), as the different depths are not independent of each other.

The results are well described, some specific suggestions are given below, the figures and tables are adequate. The discussion section is relative to the other parts quite long and reads a bit lengthy. It could maybe be a bit shorter or more streamlined towards the initial research questions and hypotheses.

For me surprising was that in the discussion section further analyses were presented that appear rather as an extension of the results, and may in my opinion therefore also maybe be rather moved to the results section. In addition, it includes the analysis of much larger set of parameter, which have not really been introduced before (e.g. dissolved organic C, bioavailable P or enzymes), and I thought, which would already be included in the rPCA analysis? Maybe it would be better to also move the graphs in the appendix section, as they rather support already stated emerging patterns. In addition, terminology changes also a bit (e.g. mixed region vs. sediment region) – double check please.

Moreover, large parts of the discussion are rather discussion the results of the study (see in the technical comments), but there could be more discussion relating the findings to other results found in other tropical/subtropical/montane forest sites and put the results more in to a larger context.

Finally, in the discussion a large part is about microbial nutrient limitation, it would be great to already introduce this as a possible control in the introduction in more detail.

Technical comments:

Line 77: this could need a reference.

Line 222: could you provide a reference for subsoil conditions?

Line 255:  I guess there were no real differences between the plateau, mid-slope, slope positions – still it would be great to mention, why these different locations were not considered anymore.

Line 281: The sentence ‘Within non-valley positions…’ is redundant.

Line 299-301: Does this describe exactly the same as is stated in Line 286-288?

Line 315: I don’t understand the x indicanting no significant difference between depth intervals within geochemical regions, I am not sure which differences the letters demonstrate – within regions or across all regions and depths. Please can you clarify this?

Line 324: should there not be two different results? Or should this indicate the D14C of bulk soil and of respired CO₂ were highly correlated.

Line 393: delete ‘from it’.

Line 429. Maybe introduce indicators for N & P limitation of microbial decomposers a bit earlier already.

Line 429: Could you repeat what is considered as poor quality? (e.g. CN ratios of soil organic matter or any other parameters?)

Line 434: check sentence – lower compared to what – and check tenses – fossil C content ‘was’ low.

Line 439: is there maybe also another study that shows that organo-mineral complexation could be saturated depending on which organo mineral complexes are present in soil (e.g. Quesada 2020, Dötterl 2018).

Line 443: add after the brackets: ), in our study aggregation…

Line 460: I would recommend to put this entire section (The role of mineral related C stabilization mechanisms) more into relation with other studies, at the moment, it is rather focusing on either studies from the same data set and reads a bit as an extended results section and could be shortened a bit.

Line 506 the same suggestion as above, I think this section can be shortened too, screen for repeated results.

Line 538: check tenses – there is sometimes a switch between present and past tense within sentences.

Line 570: namely twice in the same paragraph  

In this manuscript, the authors present the results of a 120 day soil incubation study of three soil types (mafic, felsic, and mixed sediment) sampled across topographic positions and investigate the biological and chemical controls on respiration rates and respired 14CO2. This manuscript is a significant contribution to the soil organic carbon research because the soils are from tropical Africa, a region where few studies have been carried out. Furthermore, this manuscript looks at the important C stabilization mechanisms in highly weathered soils, which are also poorly sampled and understood relative to less weathered soils often found in temperate regions. Another positive is that the researchers examined 3 depths within the soil profile. I found no glaring issues with the methods used in this study. I did have some questions about the way some of the data were presented. On one hand, this paper could be shorter and more streamlined, but on the other hand the lack of data from Tropical Africa does greatly increase the value of all the parameters reported here. I offer some suggestions on how to shorten the paper below. 

Line comments:

107-125: This whole paragraph would be better as a Table intruducing the sites and their location and chemistry. A table would be easier to read than a paragraph and make it easier to compare the sites. 

154: 12 mm seive seems like a rather large size when the usual is 2mm. It seems that the authors did not want to disrupt aggregates. that reasoning should be given here. 

174: An average of the respiration rates over 120 days, when the rates usually decrease exponentially, seems like an odd metric. Why was this parameter chosen instead of say, cumulative C loss over 120 days?

185: How was the 14C collected from three replicate jars into one evacuated container? Wasn't the vacuum in the container a different strength for each replicate so that they may not have been sampled equally?

254: How did you evaluate the distinctness of the RCs based on F-values?

Section 3.1: To help streamline the manuscript, I recommend getting rid of the discussion of TPR in the results, since SPR is the focus of the manuscript. Perhaps the TPR graphs and language could be in the supplement? I am not sure what additional understanding the TPR variable really adds here. 

280-281: I think this sentence is basically a repeat of the first sentence

Fig 1. Are the standard errors based on the replicates or the measurement times since all were averaged to get these values.

Fig 2. I think these graphs could better show the differences between the bulk and respired 14C based on how you discuss the results in section 3.2. It would be easier to comapare bulk and respired 14C if they were put on the same graph. The way they are now it is hard to see when they are similar and when they are not. 

345-349: I am not sure what the extrapolation of the respiration rates of the fossil organic C add here and in Table 1. Given the caveats, which you mention in the discussion, it would be better to leave these numbers to the discussion only. 

Fig 4. After all the data that is presented in the results, it is odd that teh discussion starts off with yet more data! I find figure 4  overwhelming. It has 8 graphs, each with three correlations, with a total of 24 to examine! Many of these are not signifcant. I suggest saving the whole figure for the supplement and choosing 1-3 graphs to highlight in the discussion. Furthermore, something should indicate which relationships are significant here, maybe make the r and p values bold where they are significant?

474: I am confused by the attribution of mineral stabilization mechanisms to controlling SPR here as amorphous and crystallized oxides had no relationship to SPR and pyrophosphate-extractable had a positive relationship indicating it was not stabilizing the carbon. 

Fig 5. Can you bold the p values for what is significant here? Same for the similar graphs in the Appendix.

502: specify high C:N here