Julia Gensel and colleagues have presented a paper that characterizes organic matter and its recent spatial distribution in the Mkhuze Wetland System and its catchment. Overall, the paper is sound, but there are some major issues to clarify, especially concerning the n-alkane data.

• We thank anonymous Reviewer #1 very much for her/his very valuable comments. These will certainly be helpful in improving the quality of the manuscript. We appreciate very much to get a view from the outside and thereby to reflect at which points we have not yet expressed ourselves clearly enough. In addition, we are very grateful for the reviewer's attention, even to the small details that we didn't even recognize anymore. We appreciate very much that she/he took the time to improve our work. Thank you very much for that! Below are the individual responses to each comment.

1. Overall, I suggest major minor revisions to this paper. Two broad comments I have: The separation in three end-member, i.e. aquatic (C₂₅), woody (C₂₉) and grassy (C₃₃) is not visible in the n-alkane pattern of modern plants presented in this study. Please clarify why these end-member were chosen, because the vegetation in the studied area likely shows a different pattern for aquatic and grassy vegetation while woody plants were not investigated.

• a) We thank the reviewer very much for this very valuable comment. We will revise section 4.1 in the discussion section to make it clearer why we used the respective n-alkanes for each identified cluster. Also, we will replace the word "endmember" with "marker compound".
• b) We fully agree that the n-alkane distribution patterns of the plants studied show dominances of n-alkanes that are inconsistent with those selected to represent the identified clusters. Briefly, our approach involves identifying specific groups of marker compounds, i.e., C₂₃/C₂₅ and C₃₃/C₃₅, that are exclusive to certain plant types such as aquatic plants and grasses, respectively. These n-alkanes are not present (or only in negligible concentrations) in the other plant types studied and can therefore be used as marker compounds for the respective clusters. We selected one of the marker compounds in each case (the higher concentrated one across all samples) to represent the respective trends. The n-alkanes we selected to represent the cluster labeled "woody" were C₂₉ and C₃₁. We absolutely agree that these n-alkanes are not produced exclusively by woody plants but are found in almost all plants. We mention this limitation in line 347ff. For this reason, we will delete the term "woody" as a cluster name and instead explain that these particular n-alkanes are rather an integrated signal of all plants to avoid confusion. We will restrict the term "woody" to subsection 4.2.1.

2. Additionally, it is demonstrated that organic matter in lake surface sediments is dominated by local vegetation. I suggest a more clear statement there: What does this imply for paleoenvironmental and –hydrological studies at the end of the respective section and in the conclusions, where it is stated that this study is of great importance for future studies.

• We thank reviewer #1 very much for her/his comment. We will revise the relevant sections to provide a more concise and clear statement of implications for future paleoenvironmental studies.

Specific comments:

3. 3ff.: Please rephrase this sentence.

• We will.

4. 7f.: Please indicate which signal is present in the upstream area.

• We will add the information that the upper reaches have a much more degraded OM signature indicated by the R- and I-indices.

5. 9f.: […] higher dD values. Compared to what?

• We will rephrase the sentence to indicate that the comparison is with δD values from samples from upstream subareas.

6. 10ff.: I cannot follow these two sentences. First you state that lake surface sediments are dominated by local vegetation incorporating a local hydrological signal. Afterwards, you state that those sediments integrate hydrological conditions of the whole watershed. Please clarify these contradicting statements.

• We thank reviewer #1 for her/his comment. We will reword the sentences to better convey our main message to the reader. (Meaning that our study provides evidence that (paleo)environmental studies which assume watershed-integrated signals in sedimentary archives/surface sediments can be constrained under certain conditions.)

7. 27: The Mkhuze Wetland System, […], is […].

• We will edit the sentence as suggested.

8. 53ff.: Move this paragraph to line 23-28 to discuss risks and benefits of wetlands in one paragraph.

• We will act as suggested and are very grateful for this suggestion.

9. 60: I suggest a short section introducing to n-alkanes and their compound-specific (CS) d¹³C and dD isotopic composition including their interpretation. Here you could introduce why you choose C₂₅, C₂₉, C₃₃ and you can refer to recent studies analyzing CS d¹³C and CS dD in topsoils (e.g. Carr et al., 2014, org. geochem., Herrmann et al., 2016, QSR, 2017, org. geochem., Strobel et al. 2020, STOTEN).

• We will proceed as suggested. We will include a section on our general approach, including an introduction to each method and the rationale for its use. We will also reference CSIA in topsoils, as proposed.

10. 69: Please avoid one-sentence paragraphs in the whole manuscript.

• We will.

11. 71ff.: It seems like a word is missing?

• We thank the reviewer for her/his awareness and will add the missing word.

12. 73: Can you please characterize the river a bit more in detail, e.g. as episodic/periodic/… river system?

• We will provide the requested information.

13. 110: Please provide a reference for this numbers and also for % precip. in the next line.

• We will.

14. 119ff.: Please provide references for all these information.

• We will.

15. 131: Can you please provide information about the potential natural vegetation in your studied area?

• To the best of our knowledge, the vegetation cover shown in Figure 3 (left subfigure) is the oldest available information on the vegetation of the system, which is also consistent with typical wetland vegetation. Therefore, we are unable to provide any additional information.

16. 229ff.: Repetition of “Here” at the beginning of the sentence, please modify.

• We will do that, and we thank reviewer #1 again for her/his attention.

17. 270ff.: Is there a table to which you can refer, that the reader can follow this numbers? If not, please provide or add these data to table 1.

• We will add a table with the requested information.

18. 290ff.: Can you please refer to a figure? Figure 8?

• For both wetland grasses, the reference to the respective sub-figure of Figure 8 is already given after the common name in parentheses.

19. 321ff.: Please check if a minus is missing prior to the dD values.

• It is definitely missing and we will add it. We are very grateful for the reviewer's attention.

20. 339ff.: In figure 8A, B both show a distinct (C₂₇) C₂₉ dominance. Please clarify.

• This is absolutely correct. Despite the predominant presence of (C₂₇) or C₂₉n-alkanes, both species have elevated relative concentrations of the medium-chain n-alkanes C₂₃ and C₂₅, which is not the case in the other plant species. Thus, the presence of these medium-chain n-alkanes makes aquatic plants distinguishable from the others, allowing these specific alkanes to serve as marker compounds. As indicated in response "b" to comment #1, we will make this point more clearly in a revised version.

21. 345ff.: Okay, and how can you distinguish the two woody and grassy end-member when modern grassy plant samples in your study show distinct contribution of C₂₉ (figure 8C, E, F)?

• We distinguish the grass plant samples based on the presence of the identified marker compounds C₃₃ in combination with C₃₅, which have elevated concentrations only in grass plants, while other plants have no or negligible relative concentrations of these very long-chain n-alkanes. As mentioned earlier, we will clarify our approach and replace the word "end-member" with "marker compound" to avoid confusion. We will also avoid the term "woody" because the C₂₉ alkane represents an integrated signal from all plants, as all plants contribute to the long-chain alkanes in comparable relative concentrations. We will retain the term "woody" only in subsection 4.2.1. to explain that elevated measured concentrations are most likely due to woody vegetation, which is much more prevalent in the upper reaches of the Mkhuze River compared to the other sub-areas.

22. 353ff.: When there are differences in photosynthetic pathway, what does this mean for the dD signal? Does the photosynthetic pathway has an influence on this signal – see e.g. Sachse et al., 2012. If the abundance of C4 vegetation in the studied area changed during the past, what does this mean for the interpretation of dD in sedimentary record?

• a) Yes, the photosynthetic pathway most likely has an impact on the δD of the n-alkanes, as summarized in the review by Sachse et al., 2012. With respect to our study, the δD becomes important when discussing its highly enriched values in the lake area (Figure 9 and subsection 4.2.4). The differences in δD in the lake area are not due to a higher contribution from C4 plants, such as grasses. The data show that both the relative contribution of marker n-alkanes and their respective δ¹³C isotopic compositions are comparable with the upstream subareas, suggesting a similar or even the same vegetation community (see 455ff). Since the contributing vegetation does not differ, the difference in δD values must be due to hydrological reasons (differences in source water, evapotranspiration, etc.). We will add a sentence to clarify.
• b) For sedimentary records, vegetation change must be taken into account when interpreting the δD of plant wax-derived n-alkanes. Since we have not presented any sedimentary records, this answer remains speculative, but we agree with the statement of Sachse et al. that changes in vegetation cover most likely result in changes in sediment δD. If it is possible to track changes in vegetation cover through other proxies (e.g., relative alkane contribution, n-alkane-δ¹³C), we believe it should still be possible to attribute changes in δD to environmental/climate controls.

23. 358ff.: This section is more or less a data description, and a discussion of the data is very limited. I suggest to provide a statement if your findings match the expected environmental conditions in each sub-environment.

• These paragraphs were intended to provide an introduction and explanation of the approach for the following subsections as they apply to all four. However, we agree that the placement was confusing and will move this information to the newly planned "Approach" section.

24. 378: Would the odd-over-even predominance (OEP) of the n-alkanes may also provide useful information concerning the state of degradation?

• No, the observed OEP values do not provide useful additional information, except that the analyzed n-alkanes can be used as conservative tracers since no contamination or extremely advanced degradation processes can be observed.

25. 379ff.: Closing bracket is missing. Does this approach really work out at your study site? I agree that this approach is widely applied and I also noticed your statement in lines 332-339, but does this really work in your setting? The data are great, but could you may make this vegetation statement a little more cautious?

• a) We add the closing parenthesis (thank you for your attention even on these minor details).
• b) We are convinced that the approach works in our environment, but we fully agree that in addition to the mentioned statement in line 332-339, we should reword the sentence to tone it down. In general, we believe that such approaches should be used with caution in any case.

26. 387: Please see comments above (C₂₉ vs C₃₃). Additionally, I suggest to provide a figure showing the n-alkane patterns in each sub-environment.

• a) Please refer to the answers to the comments above.
• b) Unfortunately, because the subareas gradually merge into each other rather than being sharply separated (e.g., the contributing vegetation in floodplain, swamp, and lake areas is quite similar), a visual representation of the n-alkane patterns of the subareas is not helpful. The overall similarity would mask the much finer differences, making it difficult for the reader to see and understand our results and interpretations. It would simply weaken our argument, as the human eye would instinctively assume that they all look almost the same, rather than recognizing the differences that exist. Therefore, we chose to present the data as boxplots for direct comparison of individual subsections. The boxplot also have the advantage that some statistical information is directly apparent to the reader, such as the median, the dispersion of the data set, etc.

27. 394: Please provide a short paragraph in the introduction how the isotopic signals are interpreted at your site for both d13C (C3/CAM/C4) and dD (amount/source/continentality…). Recent calibration studies provide a nice overview for South Africa. Are there any plants using CAM-metabolism in your studied area?

• a) We will provide this information in the new section where we will summarize information about our general approach.
• b) The presence of CAM plants in the region is not mentioned in any of the previous reports and was not observed during the field trip.

28. 413: Please provide a figure showing the n-alkane patterns in each sub-environment, as mentioned above.

• See answer “b” to comment #27.

29. 426: Can I see this sub-deviation in any plot?

• We will change the sentence accordingly, since the partial deviations mentioned are no longer shown in the figures submitted.

30. 430: Please highlight these cluster in the respective figure.

• We will do as suggested.

31. 435: Please refer to the respective figure(s).

• We will do as suggested.

32. 442f.: What is the base of this argumentation? Please introduce to the site-specific dD interpretation earlier.

• a) We will modify the sentence to indicate that although the δD indicates slightly wetter growing conditions for the contributing plants, the n-alkanes cannot be derived exclusively from local vegetation because the extent is only moderate and shows more of a mixture of deposition of upstream transported material and, in addition, a slight overprinting with local sources.
• b) For the structure, we would like to maintain the order in which the bulk parameters (Rock-Eval interpretation of degradation and classic bulk parameters) are discussed first, then the relative concentration of n-alkanes, the n-alkane carbon isotope composition, and finally the n-alkane hydrogen isotope composition in each subsection (upper reach to delta and lake).

33. 453: One sentence paragraph which is a result and no discussion. Please modify.

• We will make sure that no one-sentence paragraph occurs in a revised manuscript. Therefore, we will include this information (and, moreover, repetitions) of results in the previous Rock-Eval interpretation, which is simply confirmed by the reference to the input of fresh material.

34. 465: […] which use the lake’s water as dominant water source? How about the effects of lake water evaporation, salinity to the dD and d¹³C signal, and emergent/submerged plants contributing to the C₂₅ to C₃₃n-alkane pool?

• a) We will include salinity in the characteristics of the lake. It is well known that the lake is subject to high rates of evaporation, which in turn increases salinity (in extreme periods, the lake can reach three times the salinity of the ocean). This process leads to isotopically enriched lake water. Indeed, this process is the main argument that leads us to conclude that although the vegetation type contributing to the lake surface sediments is very similar in terms of relative alkane concentration and carbon isotope composition compared to the upstream swamp, the corresponding offset in δD is actually a consequence of the contributing plants using the lake water as a water source.
• b) Evaporation can also affect the carbon isotopic composition of alkanes. C₃ plants, for example, are likely to adjust the opening time of their stomata during "water stress" to limit water loss through transpiration. This could affect the isotopic composition of plant n-alkanes, which should be reflected in more enriched δ¹³C values. Comparing the δ¹³C values of the alkanes in the lake area with those in the swamp area (wetland), there is a slight enrichment, which is not statistically significant, but could be explained by the "water stress" of the plants.
• c) One of the most important plant species in both the upstream swamp area and the lake area is the emerged wetland plant Phragmites australis, which is probably the main producer of medium-chain alkanes. In the lake area, submerged aquatic plants were neither mentioned in previous studies nor observed during the field campaign. The characteristics of the lake (salinity, strong winds, turbidity) are probably not particularly favorable for these plant species.

35. 489: Please highlight these clusters in the respective figure. For example, use colored/shaded circles in the background of the data points

• We will do as suggested.

36. 496: Does this statement implies that organic (bio)markers in the wetland are of local origin and thus reflect local eco-hydrological conditions instead of an integrated signal including the wetlands catchment? Please clarify, because it is an essential finding of your study and of great importance for future studies at the site.

• No, therefore we will clarify. In all areas except the (delta)/lake area, the characteristics of transported hinterland signals predominate, although in the floodplain and somewhat more pronounced in the swamp sub-area, a slight overprinting by local signals can be observed. The original statement is therefore intended to imply that the organic (bio)markers in the lake area are local in origin and thus reflect local ecohydrologic conditions rather than an integrated signal encompassing the river catchment.

37. 517: Remove space before comma.

• We will.

38. 535: A trap for local organic material – see comment above.

• See answer to comment #37.

39. 549: Isn’t this also the case at your site now – see comments above? Please clarify.

• See answer to comment #37.

40. 565. Generally, I agree with the conclusions, but doesn’t the results demonstrate that the lake sediments are dominated by local organic matter representing very local ecohydrological condition instead of the lakes catchment. This absolutely limits the usage of dD and d¹³C as paleoecological markers because local effects might overprint the environmental signal, which is a very important finding for future records – although it might be unexpected.

• a) We are in complete agreement!
• b) In interpreting δ¹³C and δD in sediment records derived from downstream areas of such "traps," we would expect to observe the change from an integrated signal to a locally received signal (if the "trap" formed within the sampled time period). This is the reason that the interpretation of sediment records used for paleoecological reconstruction should take into account the geomorphological setting of the sampling area.

Figures:

40. Figure 2: The grey shaded area is very hard to identify. Numbers at the precipitation and evaporation isolines are also very hard to read, please enlarge them.

• We will adjust the figure as suggested.

41. Figure 3: Please place the figure to the section where it is mentioned in the text. Are there any more recent data than 1996, which is already 25 years ago? Or is this landcover/-usage still present? It seems very important for your study to have the most recent land-cover map for comparison with your data and correct interpretations and implications for future paleo-studies.

• a) We will.
• b) To the best of our knowledge, this is the most current information on vegetation cover. Data availability for the region is rather sparse.

42. Figure 4: Overall, this figure provides a very nice overview of the analyses you did.However, I suggest to remove all the lab-steps, e.g. lipid extraction, and provide a little more details in the respective text and therefore reduce the size of the figure. Just keep the sample (e.g. plant samples [g]), used machines (e.g. GC-FID) and results (e.g. quantity n-alkanes).

• We will remove the figure and give the information as plain text.

43. Figure 5: Please name the figures in the text first and show them thereafter (Please check for all figures and tables)

• We will.

44. Figure 8: Why is CS d¹³C of C₂₃ and C₃₅ distinctly more positive compared to the other chain-length? Is there an amount-dependency in the IrMS? Please check to note if you did an amount and/or drift correction of your data for both d¹³C and dD. Additionally, none of the aquatic plant samples show a C₂₅-dominance, which is used as aquatic end-member later, but a distinct C₂₉-dominance which is interpreted as woody end-member later. Moreover, except for 8F, none of the grasses shows a C₃₃-dominance, which is used as grassy end-member in the following. Please clarify these issue, because it is very important for your manuscript.

• There is no amount dependency in the IrMS observable. A specific intra-laboratory cut-off threshold has been established, and only large enough peaks are integrated. Therefore, we are confident that the more positive values of alkanes C₂₃ and C₃₃ (for Nymphaceae, Figure 8A) and C₃₅ (for Vossia cuspidate, Figure 8E) are real. The isotopic composition of n-alkane C₃₅ (for Cynodon dactylon, Figure 8F) shows no such trend toward higher enriched values compared to the other alkanes, supporting that these values are not artifacts because they were all measured by the same person during the same time period and processed by the same person.

45. Figure 9: How can C₂₅ be the aquatic end-member when your plants show a (C₂₇) C₂₉ predominance? The same applies for C₂₉ and C₃₃ for woody and grassy vegetation, respectively. Is there a local study showing that C₂₉ is a woody end-member in ZA? There are respective end-member based on you modern aquatic and grassy plant samples, but they are ignored in this figure. Overall, I have to note that the presentation of the data is very nice! However, please consider valid end-member.

• We will replace the word "end member" with "marker compound" as mentioned in previous answers. In short, the identification of certain alkanes and their use as indicators of certain plant types is based on the exclusive occurrence of certain n-alkanes in certain plants, while they are not found in others.

46. Figure 10: I really like this figure! Maybe also indicate the pastures, which distinctly contribute to the C4 signal in the floodplains.

• We are pleased to read this, and we will emphasize the already present sugar cane plants as we have to agree they are not recognizable as such yet. Most likely we will add a fence to indicate an agricultural field.

Tables:

47. Table 1: I suggest to move this table to the Results section.

• We agree.

 Overall, we thank anonymous reviewer #2 for taking the time to comment on our manuscript and provide feedback. We hope that we can address his/her concerns with the detailed responses below. General

1. The manuscript is generally very well written and focusses on an interesting and important topic, the characterisation of the organic matter input from Mkhuze Wetland System in South African to Lake St. Lucia. The first problem that struck me when reading the introduction was, that aims and questions are very vague., It is not clear what is meant by assessing the status of the wetland (see below). Is the hydroglogical status meant (drained, undrained) or the soil degradation status or the degradation status of OM? Also using δ¹³C and δ²H of n-alkanes to characterize sinks and sources is no doubt a fore front method, but for sure not matured enough, to draw conclusions on plant type communities and degradation status of wetlands.

• a) We will describe the objective of the study in more detail for clarity. Our study is the first to investigate OM and thus carbon storage within the Mkhuze wetland system and aims to identify the carbon cycling within the Mkhuze Wetland System by identifying sources and sinks and ultimately the transport pathways of particulate OM transported by the Mkhuze River.
• b) Plant wax-derived n-alkanes are refractory tracers that are well preserved in sediments, even over geologic time. By analyzing their carbon and hydrogen isotopic composition, additional information can be derived about the hydrologic growth conditions exhibited by the plants, as well as the photosynthetic pathway (distinguishing them into C₃ and C₄ plants, respectively). In our study, the degradation status of OM is mainly inferred from the results of Rock-Eval analyses. The combination of conservative molecular tracers (n-alkanes) and analyses of OM degradation (Rock-Eval) shows that n-alkanes, although representing a small fraction of total OM, follow the same trends as bulk OM properties determined by Rock-Eval. The information on OM degradation derived from the results of the n-alkane analyses only consolidates Rock-Eval analyses.
• c) For clarity and better understanding, we will include a new section in a revised version of the manuscript to explain our approach and the methods used in more detail.

2. As such, compared to such ambitious aims, the introduction is very general. I totally miss discussion of state of the art on δ¹³C and δ²H organic matter tracing and what it can tell us about sinks and sources. The same holds true for stability of n-alkane concentrations when used as indicators for sediment or organic source attribution.

• a) Interpretation of n-alkane concentration in terms of degradation is used only to support the results from Rock-Eval analyses, which are the main method for determining OM degradation in our study.
• b) As mentioned earlier, we will introduce a new section to further explain the approach and methods used.

3. The description of the sampling concept is totally missing. All it states is, that “ten samples where collected.” However, Figure 1 displays around 30 sampling sites, so I assume that ten samples for each sub-environment was taken? This is totally unclear. A detailed map of vegetation communities is presented (Figure 3) but it is not at all clear, if all these communities were sampled as possible sources and if so, how many samples, which plant species etc....If the aim is, to track OM in the lake back to these communities, the detailed sampling scheme has to be described.

• a) The manuscript contains a subsection "2.2 Sampling" (132ff), which describes, among other things, how many samples were collected and also which plant species were collected. The names of the collected plant species are additionally mentioned in the manuscript in the corresponding results section (subsection "3.3.1 Plant Samples" and the corresponding figure (Figure 8 and its caption).
• b) The sentence quoted by reviewer #2 originally reads "Ten plant samples were collected." (line 134), while in line 140 the information is given that "a total of 41 surface sediment samples [...] were collected [...]".
• c) We will add a reference to Figure 3 in line 136.

4. The results are mainly a listing of all measurements done with differences in numbers and sizes. There is no real information gain for the reader, as none of these results are set into perspective and the discussion does not give a clear link back to these data descriptions. Not even the indices and parameters used are in any way explained in the results section (and only very briefly in the discussion). Variability and differences are hard to assess, as sampling numbers and possible errors are not described. It is not clear if error bars indicated in-field heterogeneity or analytical uncertainty. Figure 8 states that error bars might be intra-laboratory long-term errors.

• a) We agree that we have presented the results in the "Results" section without interpreting them, as it is our understanding that interpretation is part of the "Discussion" section. It is important for us to distinguish between the measured values and our interpretation of these values.
• b) The indices and parameters used are explained in the "Materials and Methods" section of the manuscript and literature references are provided for all indices and parameters mentioned.
• c) The equations used to calculate the carbon preference index (CPI) and average chain length (ACL) are given in the subsection "2.4 Distributional parameters of n-alkanes".
• d) The hydrogen index (HI) is introduced in line 167ff in subsubsection "2.3.1 Bulk organic matter analyses".
• e) R- and I-indices are introduced in line 170f within subsubsection "2.3.1 Bulk organic matter analyses".
• f) The meaning of the indices and parameters is given in the respective parts of the "Discussion" section and, in addition, a link back to the results is provided (also in numbers, if applicable); see e.g. lines 383f, 401ff, 427ff, 447ff. Otherwise, a cross-reference to the respective figure is given (see, e.g., lines 340, 350, 353, 384, 389, 400, 406, 412, etc.).
• g) However, for better understanding of our approach and the methods used, we will additionally include information on the parameters and indices used in the newly planned "Approach" section.
• h) The respective subsections of the "Materials and Methods" section contain information on replicate measurements and sample replicates. We have chosen to always present the largest of all possible errors (natural heterogeneity exceeds analytical errors). If the analytical errors are smaller than the long-term error within the laboratory or a replicate measurement was not possible due to limited sample size, the long-term error was assigned to more adequately reflect reality.
• i) When multiple errors can be represented, such as in the analysis of plant samples, we refine the corresponding caption, i.e., Figure 8.

5. The discussion is more a descriptive qualitative narrative of differences found in parameters within and between different sub-ecosystems. Indices for evaluation are not adequately introduced and partly interpreted in a wrong way (e.g., that CN ration of OM would be a general indicator of chemical stability). As such, I can not follow conclusions drawn and can not judge if these conclusions adequately assess the results. One example would be the conclusion “Sedimentary OM in the floodplain and swamp exhibit high variability in their source signatures and degradation status reflecting environmental diversity, with samples from the floodplain characterized by a mixture of degraded OM from the hinterland and fresh OM.”(line 575-577). With clearly high ongoing and very variable degradation of OM in these systems, concentrations of organic substances can not be used as a conservative tracer.

• a) To the best of our knowledge, the present study is the first to examine OM properties in the Mkhuze Wetland System. Therefore, no quantitative comparison with previous studies is possible. Intra-system comparison of different sub-areas leads to a qualitative assessment of whether certain characteristics of OM are found to a greater or lesser extent in other sub-areas.
• b) We will introduce a new "Approach" section to improve understanding of our approach by explaining our general approach and methods used, as well as the indices provided and their meaning (see also responses "b", "c" to comment #4).
• c) We will delete the sentence about the C/N ratio providing information about chemical stability, since in this case we did not intend to evaluate the stability of chemical bonds. We agree with Referee #2 that the wording is unfortunate and thank him/her for the comment. Nonetheless, the results of the C/N measurements serve as supporting information, since the main method for deriving information about OM degradation in our study is the Rock-Eval analyses.
• d) We will split and edit the mentioned sentence (line 575-577) into two sentences to improve readability and comprehensibility.
• e) n-Alkanes are refractory molecules that are well preserved over geologic time. Therefore, they can be used as conservative tracers, as is commonly done. In addition, we refer almost exclusively to relative concentrations of n-alkanes rather than absolute concentrations. Absolute concentrations of n-alkanes occur only in statements used to support the results of Rock-Eval analyses. n-Alkanes accumulate in sediments because of their chemical stability (now actually defined as stability of chemical bonds) when OM is degraded, and labile parts of OM are removed and utilized by microbes. Therefore, comparison of absolute n-alkane concentration normalized to dry weight or organic carbon can serve as an indicator of OM degradation (performed in our study only as additional information to our main Rock-Eval method).

6. Regarding the isotope tracers used, no un-mixing was done and the values were interpreted in a qualitative way, which is, from my perspective not leading to meaningful conclusions.

• In the early stages of the manuscript development, we attempted to apply a Bayesian un-mixing model, but at a certain point had to accept that the quality parameters of the statistical model were insufficient, as convergence could not be achieved. The reason for this is most likely that these models require either a larger sample size or endmembers with larger differences to be statistically reliable. Therefore, we opted for a qualitative approach by identifying marker compounds (groups of specific n-alkanes) that are present in certain plants and absent (or only present in negligible amounts) in others. These marker compounds define clusters of plant types (e.g., aquatic, grassy) that are segregated from each other and thus can be used to identify predominant sources of vegetation. This approach will be explained in more detail in a revised version of the manuscript in a new section "Approach".

7. All in all, I would judge this work as containing highly valuable and interesting data and results. But description of sampling concept is inadequate and interpretation of data is qualitive with numerous assumptions I am not sure can be hold.

• a) We thank Reviewer #2 for his/her comments attesting that the data presented are of high interest and value.
• b) We believe that our study contributes to a better understanding of the carbon cycle and carbon storage in the Mkhuze Wetland System. As discussed in detail in the "Discussion" section, our finding that OM is sequestered under current conditions in the swamp area of the wetland system studied may also be found in other wetland systems, suggesting that carbon sequestration in such systems is primarily hydrologically controlled.

Introduction

8. Generally, well written and interesting to read about the Mkhuze Wetland System. However the aim of “assessing the current status” is very vague to me (line 36). Which status do you mean? Hydrological? Soil degradation? Nutrient status?

• We will refine the description of the objectives of our study (see answer "a" to comment #1)

9. I totally miss discussion of state of the art on δ¹³C and δ²H organic matter tracing and what it can tell us about sinks and sources. Also, what about the stability of n-alkane concentrations in these systems? Are you sure you can use these as conservative tracers?

• a) We will include the requested information and references in the newly planned "Approach" section.
• b) Yes, we are sure that n-alkanes can be used as conservative tracers. Because of their chemical stability and resistance to microbial attack, they are well preserved (e.g., Eglinton and Hamilton, 1963), which is why they are commonly used as refractory tracers over geologic time (e.g., Eglinton and Eglinton, 2008).
• c) In addition, the high CPI values indicate that n-alkanes weren’t subject to advanced degradation processes.

10. 40-42 this assumes that you have species specific tracers

• We agree with reviewer #2 that plant wax-derived n-alkanes and their stable carbon and hydrogen isotope compositions are not capable of identifying specific plant species, but rather provide information about vegetation types. Combined with knowledge of the occurrence of (dominant) plant species in a given area, inferences about the most likely source can be made even at the plant species level.

11. 42-45 how can you assess the vegetation type (do you mean plant community?) with δ¹³C of nalkanes?

• The newly planned "Approach" section will also include an explanation of the use of n-alkane-δ¹³C to distinguish plant types with different photosynthetic pathways.

12. 45 hydrological conditions of what? Of the regime under which the plants grew? Of the soil? E.g., wetland, upland, drained? Not sure you can achieve this with δ²H?

• a) We will include the requested information in the new "Approach" section.
• b) Briefly, δD of n-alkanes preserves information about hydrological conditions experienced by plants during their growth phase. There are several possible interpretations, as several factors can affect the δD of plant wax (will be explained in the "Approach" section.

13. 51 this is a big aim, to the assess the status of the wetland systems in terms of its filter function and influence on Lake St. Lucia!

• a) We agree.
• b) See also response "a" to comment #1.

Methods

14. 133 what do you mean by “Ten samples were collected.” Of what? Ten repetitions within a site?

• a) The original sentence reads, "Ten plant samples were collected." (line 133). This means that we sampled a total of ten plants. The number of replicates of each species is given in lines 136ff.
• b) Further, line 140 indicates that a total of 41 surface sediment samples were collected.

15. 137 – 139 these plants were not collected? But all others were? Or these are the ones which you did collect?

• We absolutely agree that the English word "this" at the beginning of the sentence in line 137 refers to the subordinate clause mentioned earlier. We thank Referee #2 very much for pointing out this incorrect reference and will, of course, reword the sentence accordingly. However, we sincerely hope that most readers of the discussion paper have concluded from the context that we sampled the number of plants indicated in each case, rather than sampling all but one or two plants in the area.

16. Figure 4 is basically describing standard analysis and could be moved to supporting information.

• Figure 4 will be removed and the sample preparation information will be provided as plain text.

Results

17. 252 what is HI value?

• a) The Hydrogen Index (HI in mg HC/g TOC-1) is one standard parameter determined by Rock-Eval analyses. It is calculated by integrating the amounts of hydrocarbons released during thermal cracking of OM between defined temperature limits (line 167ff). Depending on the application of Rock-Eval analyses, different information can be derived. When examining sedimentary rocks, the degree of thermal maturity and kerogen type can be inferred. When examining organic matter in soils or sediments (as in our study), the hydrogen index is used to infer the origin of organic matter based on the difference in predominant biomacromolecules in aquatic organisms and algae (rich in lipids and proteins) compared to terrestrial plants (rich in carbohydrates).
• b) We will include a statement in the new "Approach" section that provides a more detailed introduction to the methods used and the information they provide, rather than just a literature reference.

18. 256 what is R-index?

• a) The R-index calculates the relative contribution of the most thermally stable HC pools (line 170f) and the I-index calculates the ratio between thermally labile and resistant HC pools (line 171f). Details can be found in the given reference at line 172, Sebag et al., 2016. Briefly, both indices use the integration of certain areas (A1-A5) under the S2 curve between certain temperature boundaries (for the exact boundary temperatures, the reader is referred to the given reference). These integrated areas relate to the differences in thermal stability of the biopolymers in the OM. Basically, one could say that the higher the R-index (R = refractory), the more pronounced are the thermally refractory pools within the OM and the higher the I-index (I = immature), the more pronounced are the thermally labile parts of the OM.
• b) In addition, as derived from their mathematical construct, these two indices (R-index and I-index) are inversely correlated when OM stabilization (R-index increasing) results from progressive decomposition of labile organic compounds, and relative enrichment in refractory ones. Then, in the I/R diagram, a “decomposition regression line” describe the decreasing labile pools and concomitant increase in more thermally stable pools, was observed in compost by Albrecht et al. (2015) and soils by Malou et al., (2020), Masseroli et al. (2021), Matteodo et al. (2018), Sebag et al. (2016), and Thoumazeau et al. (2020). However, situations with OM mixture from different sources may generate a distribution in the I/R diagram aside the “decomposition line”, i.e. a poorly related I–R indices.
• c) A brief explanation of the two indices will also be included in the new "Approach" section to improve understanding of the methods we use.

19. 260 what is I index?

• See answer to comment #18.

20. For all errors it is not clear from how many reps they are produced, if repetition at all or if this is analytical error.

• See answer „d“ and „e“ to comment #4.

Discussion and Conclusions

21. Paragraphs 343 -349 versus 332 – 338: I am not sure I understand you correctly, but this makes not much sense to me. First, you describe the differences in n-alkane patterns within plants, within sites and between different ecosystems, but then you assume that you can take literature values from generally well studied plants, such as trees as source values to be characteristic for your sites?

• We will refine subsection 4.1 in the discussion section to make it clearer why we used the respective n-alkanes for each identified cluster. Also, we will replace the word "end-member" with "marker compound". In short, we identified marker compounds, i.e., specific n-alkanes that are present in certain plant clusters while absent (or present only in negligible amounts) in others. We selected one of the marker compounds per cluster (the higher concentrated one across all samples) to represent the respective trends. The n-alkanes we selected to represent the cluster labeled "woody" were C₂₉ and C₃₁. We agree that these n-alkanes are not produced exclusively by woody plants but are present in almost all plants. We mention this limitation in line 347ff. We will delete the term "woody" as a cluster name and instead explain that these particular n-alkanes are more of an integrated signal of all plants to avoid confusion. We will restrict the term "woody" to subsection 4.2.1.

22. The discussion in 4.1. mainly compares the n-alkane concentrations determined in this study in comparison to literature values. But what is the message behind this paragraph?

• a) Comparison of the relative n-alkane concentrations determined in our study with the literature provides the rationale for our choice of specific n-alkanes to serve as marker compounds. Furthermore, limitations of the approach known from the literature (line 333ff) are mentioned and why we are nevertheless convinced that the approach remains valid despite these limitations.
• b) As mentioned in answer "a" to comment #25, the subsection will be refined.
• c) In addition, although more as a side effect, the placement of our results in the previously published literature should help the reader assess the credibility of the data presented.

23. Paragraph 4.2.: what is is this telling me regarding the aim of your study, the status of your system, what is the aim of this paragraph? What is the connection to your results?

• a) The first paragraphs of subsection 4.2 were originally intended to provide information applicable to all subsequent subsections (4.2.1 - 4.2.4).
• b) We thank Referee #2 for bringing to our attention that this has caused confusion. We will embed the information provided here in the new "Approach" section.

24. 369 the C/N ratio is an indicator of chemical stability? I do not think so?

• See answer ”c” to comment 5.

25. Paragraphs 4.2.1. to 4.2.4. describe the variation of the different measured parameters in each of the sub-ecosystem types and tries to induce state of degradation of OM or plant origin. This discussion is qualitative and not really set in perspective to literature values.

• a) The mentioned subsubsections provide the interpretation of the measured variables in terms of OM properties and forms the basis for the following inferences about OM transport pathways within the system. We are convinced, as mentioned above, that the methods used are fully capable of providing the necessary information and not just "trying" to do so.
• b) Internal comparison of subareas does not provide quantification, but rather an assessment of "more", "less", "similar", or "different" compared to upstream areas.
• c) To our knowledge, there are no previous studies of OM in the system, so a comparison to previously observed characteristics is not possible, nor is an assessment of whether OM is more or less degraded than in another wetland plausible.

26. Section 4.3. (line 467 – 480) starts with a general description of Mkhuze Wetland System which might be transferred to the methods or the introduction. Or are these statements conclusion from your data? If so, please make the link to your results.

• a) The first sentence refers to the previous subsection (4.2), where interpretations of OM properties from measured parameters are given, and introduces the following conclusion on OM transport pathways.
• b) The second sentence extends the interpretation of OM to sediments in general and justifies this with literature references.
• c) The following sentences (lines 472ff) are the essence of the interpretation of OM properties and inferred transport pathways. These are related to the available literature and any discrepancies that arise are discussed.

27. 494 – 499 is this general knowledge of literature or is this a conclusion from your results? Please make the link to your data

• It is a conclusion from our results, as shown by the respective sentence beginnings "Therefore, it can be concluded [...]" (line 494) and "The identification of the Mkhuze Swamps as [...]" (line 494).

28. Some conclusions might be considered speculative.... e.g., from the result that “....the higher hydrogen isotope signature of the sedimentary n-alkanes in the lake probably resulted from a dominant contribution of lakeshore vegetation” (line 464 – 465) the general conclusion is drawn, that “OM in the surface sediments of Lake St. Lucia originates primarily from lakeshore vegetation” (line 501). There is no unmixing of possible source signatures, no quantitative evaluation. This is just one example, which leaves the impression, that conclusions drawn are based on rather qualitative assumptions and might even be speculative.

• a) The Rock-Eval analyses clearly show that in the lake area, "These results differ drastically from the OM results of the upstream sub-environments, but they do not reflect aquatic autochthonous contributions (as indicated by the low HI). Although the sources of this OM are probably terrestrial, it is not a detrital (allochthonous) OM reworked from the catchment area, but rather a proximal (para-autochthonous) contribution." (line 449ff).
• b) Analyses of plant wax-derived n-alkanes and their stable carbon and hydrogen isotopic compositions indicate that the vegetation source in the lake area is quite similar to the upstream environments (similar δ¹³C and relative alkane concentrations), while the significant offset in δD indicates a different water source that was available to plants for growth (see lines 455-463).
• c) To point out that our study, like most if not all scientific studies, cannot provide 100% certainty in interpretations, we used the word "probably", "mainly", etc.
• d) However, the combined interpretation of both methods suggests that the shoreline of the lake as the origin of the OM is the most reasonable explanation (line 464f).
• e) We agree with reviewer #2 that our study is mainly a qualitative study comparing different characteristics of OM obtained by a combination of methods. But to put it simply: If we do not find the upstream OM characteristics in the lake area, but instead find completely different characteristics, then in all likelihood the OM must have come from somewhere other than upstream.