Review of hess-2024-250

Title: Hydrological Controls on Temporal Contributions of Three Nested Forested Subcatchments to DOC Export, by Blaurock et al.

Blaurock et al. present high frequency DOC and discharge data for a period of one year from three nested forest headwater subcatchments in the Bavarian Forest National Park (Germany). They aim to explore differences in DOC export between the three subcatchments, which have distinct vegetation, microclimate, soil types, and topographical characteristics, among different “hydrological periods” namely spring, summer, autumn, winter, and snowmelt. Precipitation inputs drive overall exports whereas differences in runoff contribution and hydrological connectivity between the different catchments drive seasonal differences in DOC exports between the subcatchments.

High-frequency sensors and increasingly used to better understand biogeochemical mechanisms and mobilization processes at the catchment scale, particularly in relation to the important constituent DOC. This study provides further insights into the topic and should be of general interest for the readers of Hydrology and Earth System Sciences. I do have a number of concerns and suggestions and a list of other relatively minor comments that the authors should address before the manuscript can be accepted for publication.

General comments

The specific comments I provide below are extensive enough for the authors to consider the revision of their manuscript. Here I summarize my main points of criticism:

1. The definition of “hydrological periods” needs a more rigorous explanation, including e.g. information on how dry or rainy periods are determined.
2. I miss a more compelling explanation/interpretation for the higher runoff generation at MG.
3. Likewise, I am not convinced about the explanations given for the low runoff generation and high flow-weighted DOC concentrations across all conditions observed at HSsub. The authors repeatedly mention that hydrological connectivity needs to be established at this site, but it appears that even during low hydrological connectivity HSsub provides water with high DOC concentrations. Where does it come from?

Specific comments

Abstract

12. Maybe "soil type" instead of "soils"?

12-14. In the abstract, you associate limited hydrological connectivity with snowmelt, summer, and winter. In general, snowmelt periods are associated with high hydrological connectivity instead, and I think it is similar in your study. Moreover, you do associate autumn with limited hydrological connectivity later on in the text (at least for subcatchments HSsub), so this part of the abstract is somewhat inconsistent with your interpretations.

1 Introduction

33. More explicitly, “because of in-stream metabolic processes”.

35. I would say "In addition" rather than "But".

44. The high groundwater levels also favour the build-up itself of organic matter in the soil (because of limited mineralization due to hypoxic conditions).

47. I would say concentrations "generally" increase.

2 Material and Methods

72. Do you perhaps mean "2.1 Study site"?

75 – Figure 1. If I understand correctly, the study is based on data only from the Hinterer Schachtenbach catchment (delineated in red in the figure), which is part of the bigger Grosse Ohe catchment, which I understand is also depicted in the figure outside the delineated area. I would suggest to only present the Schachtenbach catchment in the figure, as the rest of the illustration is more distracting than informative.

90. Are these rock outcrops or exposed bedrock, or how are these rocks “interspersed” in the soils?

91-95. Just out of curiosity: in the period 2018-2021 there was a drought followed by a large infestation of Norway spruce by bark beetles in large parts of Central Europe. Was the forest in your study are not affected by this disturbance?

104-111. The gap filling at HS and Q construction at KS are fine. However, I am confused about the range of values shown in Figure S1 and Figure S2 compared to the range of values that you present here and that I could see in your raw data in the Figshare file. Specifically, the upper values are much larger in Figures S1 and S2 compared to the upper values shown in the study for all three subcatchments (e.g. the highest Q at HS according to the data here is 0.75 m3/s whereas it appears to be as high as ca. 3 m3/s in Figure S1 for an antecedent period). Could you clarify this point?

120-125. Thank you for this detail explanation. Just out of curiosity: the fact that you move D3 from KS to MG following the failure of D2 makes me think that you prioritized having data from MG compared to having data from KS. Is this correct and if so, why?

125-135. Nice!

138-139. However, you have an additional period which you define as “snowmelt”, which in fact is the longest of all.

139. Perhaps “large extent” instead of “larger extent”.

136-150. I think this classification is fine, but I wonder whether more details can be provided so it appears less arbitrary. For example, you mention “starts” or “ends” of “dry” or “rainy” periods to delimitate your hydrological periods, but no information on how you define a dry or rainy period is given or in reference to what. Also, the snowmelt period appears to be excessively long (Feb to Apr 2021). Did snow cover take that long to melt?

168-170. Fair assumptions but what do you know about the in-stream processing of DOC in your system? I would be inclined to think that it is probably limited, but there is increasing evidence in the literature that in-stream DOC processing might be larger than previously thought, even in low order streams with non-labile DOC. What is the chemical character of the DOM? You can probably have a proxy for this with the absorbance data.

3 Results

177-178. The “leading to sharply rising DOC…” implies that precipitation events are responsible for the increase. This is of course true, but via hydrological activation of upper soil layers that have build-up DOC during summer. Anyway, these explanations belong to the discussion so I guess what I am trying to say here is that you could avoid using terms like “leading” and simply describe the observed patterns without further implications on the processes.

191. Typically, DOC export is reported in kg/ha or g/m2. Please, transform the 3931 kg/km2 into either of these other units for better comparison with other studies.

210. Do you mean “DOC export” instead of “DOC concentration”?

210-218. To me, the interesting thing about this kind of figure is to relatively compare the evolution of cumulative discharge and cumulative DOC export, with focus on when the lines deviate. For example, after both discharge and DOC evolved comparatively, DOC disproportionally increases at some point in mid-autumn, but this disproportionate increase is cancelled out during winter (with the exception of the mid-winter snowmelt event). Then again at the beginning of the snowmelt period DOC disproportionally increase relative to discharge, but as the snowmelt period advances, DOC decreases relative to the discharge, suggesting some kind of dilution effect or even production-limitation taking place then. You have some explanations in the discussion around this figure, but I would suggest to make these points more explicit, here in the results when you describe the figure, and later in the discussion.

229 – Figure 5. Again, I would rather present DOC exports in kg/ha or g/m2.

242. But KS contributed less than expected according to its area ratio, right?

4 Discussion

268. See my previous suggestions regarding the units of annual DOC export.

312-313. I don’t think I can agree here. The discharge time series show that flow is low towards the end of snowmelt, and therefore I can hardly imagine soils being saturated at this point. Also, according to Figure 4, DOC decreases with respect to discharge during this period, and it is only the activation of DOC source areas during spring rainfall events that can explain this pattern. Note as well that the way you define the different periods is very much influencing your findings in terms of DOC because you are using main DOC drivers in the definitions.

343-345. Also, potentially lower evapotranspiration as the deciduous trees drastically reduce transpiration during winter.

345. Rather than “after the snowmelt period”, it could be better to write “during spring” “or from spring on”.

351-362. To me, one of the most striking results in your study is the significantly higher runoff contribution of MG compared to the other two subcatchments, especially compared to KS which a priori are more comparable. In this part of the text you provide hypotheses that aim to explain this observation, but I am not convinced that they can fully account for it. What about the role of vegetation? After all, evapotranspiration is a main component of the water balance. I can see that MG has the highest percentage of forest in a state of rejuvenation. Could it be that the overall evapotranspiration in this subcatchments is relatively lower than at KS because the forest is not fully developed and this contributes to the higher runoff contribution from MG?

375-380. But the higher flow-weighted DOC export at HSsub occurs across all periods and not only during high wetness. And, as you mention before in the text, autumn is a particularly limited period in terms of hydrologic connectivity between HSsub catchment soils and the stream, which “inhibit DOC mobilization”. Therefore, I remain puzzled as to why HSsub (i) is so inefficient at generating stream runoff, especially during autumn (the runoff ratio of 0.13 is strikingly low), and (ii) can still provide water with high DOC concentration so that flow-weighted exports are high across all conditions. I think these points are the most critical to revise in a new version of the manuscript as they are also the most interesting in terms of catchment process understanding.

382. In Figure S6, how is it that the mean of daily flow-weighted DOC export of “all" periods is lower than each of the periods for all three subcatchments? Are you using the summer period to calculate the mean in “all”? Even if you do, I find it difficult to arrive to those values. I would expect to see something similar in relative terms to what is shown in Figure 6b (in fact, the values should be proportional so the relative differences should be the same).

5 Conclusions

425-431. Droughts could also lead to bark beetle infestation and death of trees, with important hydrological and biogeochemical consequences at the catchment scale.

In their manuscript, entitled “Hydrological Controls on Temporal Contributions of Three Nested Forested Subcatchments to DOC Export”, Blaurock et al. analyzed DOC export across different hydrological periods. By the use of high-frequency data, they could show clear differences between the catchments in terms of the timing and magnitude of DOC export, which the authors could explain by differences in soil, topography, vegetation, and microclimate. The manuscript is well-written and contributes interesting insights into the drivers of riverine DOC export at the headwater catchment scale. My comments might be abundant, but they are largely of minor character. I do agree with the main points raised by reviewer 1. Furthermore, I did not agree with the author's repeated argumentation that the most important factor for DOC export was precipitation and, with that, discharge (see my detailed comment to lines 272-274). Furthermore, the authors should be careful in stating that certain catchments characteristic “control” DOC export, while not having directly tested this, but "only" having provided a sound explanation. Overall, studies like this one that get to the bottom of the drivers of DOC export from forested headwaters are highly valuable to the readers of HESS.

General (but still minor) comments:

I suggest highlighting the importance of high-frequency data to identify “hot moments” of DOC export that would be covered if low-frequency data were looked at alone. A comparison of how many % of DOC were exported within only % of the time would be cool, from my perspective. Within your framework, you could highlight that X% DOC was exported during the autumn period, while this period only covers X% of the entire year.

I suggest you make it more explicit that increasing concentrations with increasing discharge make DOC export during high flow periods disproportionally high. Stagnant or even decreasing concentrations could still show higher exports during high flow, but not as pronounced.

What about stream density? Is it higher in the upstream catchments? I could imagine that a network of (temporary) streams largely increases hydrological connectivity. If, with your knowledge of the catchments, you agree – I suggest you add stream density and this argumentation.

Line-by-line comments:

L9: “paramount importance to understand all processes of the global carbon cycle” sounds overstated to me; could you please tone it down a little?

L28-29: This doesn’t read smoothly to me. If most of the carbon is inorganic, I miss the connection to organic carbon and the link to why organic carbon is also important. When you then, in L32-33, state that DOC is a major component of exported C, this appears contradictory to me.

Generally, I am missing at least a little information on the role of DOC for nutrient cycling (i.e., important electron donor for denitrification) and its impact on aquatic ecosystems.

L60: have you looked at the TWI? It has often been found to correlate with DOC concentrations and could be interesting here as well. However, this is more a note than a comment that I want to see addressed.

L65: this section should be concretized. Why did you specifically want to go beyond the event scale (or low-frequency data)? What were your expectations? It reads a little listless.

L78: I assume HSsub should be HS_sub?

L85: I would prefer you do not repeat exactly what is already given in the table. Either rephrase or take it out. And check this throughout the manuscript, please.

L132: I very much appreciate this sound method section. However, I would like to see the fit between grab samples and sensor measurement as a scatterplot in the SI.

Table 3: I would like to have average discharge and DOC concentration included here as well to be able to tell if high loads are mainly due to high C or Q.

Figure 3 & Figure 6: I am a little confused with your unit of (flow-weighted) DOC export. According to your equation, DOC export is C [mg/L] * Q [L/min] * t [min], so the unit should be mg, right? How do you get mg /L*d in Figure 3 then and in Figure 6b mg/L…? What is it exactly that you define as flow-weighted DOC export, and how does it differ from DOC export and from concentrations? Please clarify.

Figure 5: Would it get too messy to add lines for cumulative Q here as well? Maybe as thin or transparent lines? I really liked being able to compare that to the export in the previous figure. But I leave that up to your discretion.

Figure 6a, b: What about the green (a) and black (b) columns reaching the limit of the y-axis? Are the values beyond the y-axis limit? Could you please change the limit or indicate these specific values somewhere in the figure?

L51-52: The information in the last sentence could also very well be integrated into Table 4, which would be more consistent, from my point of view.

L269-270: I would appreciate it if you could name these ranges briefly.

L272-274: I partly disagree here. Precipitation is not simply equal to discharge… besides being driven by precipitation, discharge is further driven by catchment wetness that also relates to temperature controlling snowmelt and evapotranspiration, vegetation, and soil type – all of which control how much water is stored in the catchment and enable hydrological connectivity and transport. From your analysis and results, I would rather see a direct link to discharge than to precipitation alone.

Moreover, to me, “the most important factor” implies you have run some kind of statistic to rank the importance of factors.

Please rephrase your argument.

L275: In your figures, you show the average daily DOC export, where can I see the absolute solute export? If it’s a “new” result it should not appear here in the discussion for the first time.

L298: Again, some numbers would help me here. What is the ‘typical base flow concentration’?

L319: See my argument above. I agree that a lack of precipitation events can limit DOC flushing. However, especially in summer, there is not only a lack of precipitation but also higher evapotranspiration, reducing catchment wetness and connectivity and thus discharge.

L320-326: I might be mistaken here. But is freshly fallen leaf litter directly turned into DOC? Doesn’t it take some time to decay until it is DOC?

L370: Good point! If not the entire area is connected, catchment area can be misleading.

L329: P ≠ Q; see my argumentation above

L410-414: You discuss this, and it sounds very reasonable to me, but you do not directly prove this. Thus, you should be careful with phrases like “is controlled by”. Instead, “can be explained by”, or  “we argue that…” would be more appropriate.

L419-431: This is not really a conclusion, rather than an Outlook. Thus, I suggest calling this section “Conclusion and Outlook”