General comments

This manuscript addresses the effect of low-cost gully interventions in farmer communities in Southern Ethiopia. Based on the monitoring of upward expansion of gully heads, the author show that severe gully erosion can be minimized with riprap, grass planting and check-dam integrating. Over a period of 26 month these rehabilitation activities ceased gully head expansion, whiles untreated gully heads expanded 671 cm on average. Informant interviews, group discussions and household surveys show that farmer´s knowledge and perception of their capacity to act against gully erosion increases after on-farm trials and field day demonstrations. Soil degradation due to gully erosion causes persistent problems regarding agricultural and pasture activities especially in Sub-Saharan rural communities. Thus, the topic is relevant to a larger community of readers since it provides insights on how farming communities can be empowered via education and field training to reduce severe soil erosion with low-cost measurements. While the authors are well presenting the relevance of their study, the main issue is that the discussion and presentation of the results are not well-structured and remains too descriptive. I strongly suggest a complete rewrite of the result and discussion section of the manuscript to enhance the main take away message of this study, which is indeed important. The result and discussion section also contains many parts, which belong to the method section. Another flaw of the study is the lack of clear and testable hypothesis to which the discussion section could tight back again. In addition, there are parts in the study design and methodology which needs more clarification. Especially the statistical analysis needs to be described in more detail to comprehend each step in the interview, survey and group discussion evaluation. Even though this study focuses on qualitative analysis, parts of the study design regarding soil physicochemical properties need more clarification. Finally, the conclusion section now is only a short summary of the study but should focus on what the reader have learned after reading this study. Please see my comments on that below. At this stage, the manuscript needs a major revision before considered to be published in EGUsphere. 

Below are the edits and comments on the manuscript

Abstract

In general, the abstract is well structured and presents the main take away messages from the study. There are minor points which need to be addressed:

Line 15: Can you please explain buy-in and input in this context?

Line 18: Do you mean with “in-context changes” the field day training of the farmers?

Line 18 and 23: You first mentioned that the measurements are low-cost but then farmers are mentioning the high cost of input material at least for the gabion check-dams. What does low-cost mean in this context? Is the input material cheap or the labor time? And for whom is it low-cost (for the individual farmer, the community)?

Introduction

The introduction is well-structured and written. It presents the significance of the land degradation problem caused by gully erosion in general and shows why this issue is even more severe in the study region. The author discusses soil erosion measurements and presents the problem, that farmers may perceive such actions as requiring too much labor and high cost to implement them on their own land. While the authors are well presenting the problem and how to potentially mitigate this issue with on-farm experiments combined with qualitative analysis (interviews, group discussions, household surveys), the introduction lacks clear hypothesis based on theoretical reasoning. Please see my further comments below:

Line 30: Give examples for key environmental services.

Line 33: Give examples for those environmental and anthropogenic factors.

Line 36: Please provide more reference for this statement (global problem of gully erosion).

Line 37: Provide more insights why Africa is the worst-hit continent regarding gully erosion and what the drivers are.

Line 39 – 41: More references needed for this statement.

Line 53 – 55: Explain more why the problem is bigger in poorer countries instead of repeating that Ethiopia is challenging gully erosion.

Line 79 – 80: Elaborate this sentence a bit more. What do you mean with biophysical evidence in this context?

Line 86: Hard to follow. Gullies as a simple indicator for gully erosion? What do you mean with simple indicators in this context?

Line 88: Give reasoning why your study region and research has a wider applicability to other regions. You need first present you results to make this statement. This statement fits better to the conclusion section.

Methods

In general, the method section needs a more detailed description of the soil characteristics and management in the study area and how this affect gully erosion. More information is needed how the land cover, climate gradient and strong rainfall events were treated in the analysis since those factors will most likely impact gully erosion. Also, what initializes gully erosion in the study area (bare soils without any vegetation cover, tillage on steep slopes, cattle)?  See my comments below:

Figure 1: Where are the study sites located at panel c? The panel a, b and c need more description since it is not clear from the figure alone what they are show.  What about parent material mineralogy and texture? Does differences in soil parent material affect gully erosion?

Table 1: I would present the last four parameters in a separate table since it shows social-economic information. I also would suggest using the unit ha for presenting the landholding area since it is more common for readers. How is the climate gradient treated in the analysis? More information needed about the land cover, land use and management and how this impacts gully erosion in the study area.

Line 123: What is the reason for only choosing gullies with depths less than 3 m in the analysis?

Line 131: Better references are needed to describe the soils and their erosion history in the study area since this is important information for the reader and the overall study.

Table 2: Is this table showing the results after 26 months? This would rather belong to the result section.  

Table 3: Are you sure this is a Luvisol? Based on the interlayering of horizons with different clay content, it could also be a Fluvisol. I would also suggest using a soil classification considering tropical soils. What is a kebele?

Line 138: What is woreda?

Figure 2: You need to describe treatments in more detail so that the reader can reproduce the method. How do you define a knowledgeable farmer? 

Line 167: Where can I see the questionnaire used in this study?

Line 171: Did you also monitor the gully heads during and after strong rainfall events? And how was changing vegetation cover considered due to crop growth, harvesting etc.?

Line 172 – 173: How exactly was the permanent reference point established? This information is very important. How did you ensure that the reference point was not changed during the monitoring?

Line 176: Why was erosion further down the gully not considered in the study?

Foot note 1: Please include the information from the footnote in the text body since it distracts the reader.

Line 185 – 186: How exactly does the coding scheme work?

Line 183: What is deductive coding?

195 ­ 198: Please explain this approach in more detail.

Results and discussion

This section is the main flaw of this manuscript. It is poorly structured and contains many parts, which belong to the method section. This section needs to be divided in an individual result and discussion section.  In general, this section is too descriptive and does not discuss the results in a bigger context. More references are needed. The discussion should then focus what the reader can learn from this study.  In addition, the author needs to show that their field measurements are the dominant factor influencing gully erosion since the study area also has a gradient in climate and soil properties and most probably changes in crop cover over the seasons. The comments below:

Line 206: Please present the also the standard deviation from the mean value.

Line 206 – 207: This needs to be mentioned in the method section first. Please also report standard deviation. How exactly was soil erosion calculated?  

Line 207 – 209: What about other factors which could impact gully erosion besides the interventions (like climate, land use management, crop cover, differences in soil properties etc.)? You need to show that the effect of halting upward expansion of gully heads is mainly driven by the treatments.

Figure 3: How do you explain the sudden increase of gully erosion after 26 months in all control plots? How do you explain the differences in gully erosion with time across control plots? What do you mean with “(…) the gullies were spatially distributed, and a single control cannot work”? Are they different in soil properties, rainfall intensity, amount of cattle etc., which could affect gully erosion?

Line 216 – 218: Do you mean that the amount and the depth of gullies are different across slope positions?

Figure 4: What is the meaning of the yellow arrows? What are SWC measures? What are the differences between drivers and pressures?

Line 252 – 253: This sentence is hard to follow. What do you mean by that statement?

Line 254: How can land affected by gully erosion converted into productive land? The measurements used in this study only stop the upward extension of already existing gully systems.

Line 256 – 258: This is a circular argument which needs revision.

Line 259 – 260: Again, how can land already affected by gully erosion converted into productive land? How is accessibility related to soil degradation in this context?

Line 263: If there is sediment accumulation happening, after implementing the rehabilitation measures, then there is still soil erosion happening. Thus, this argument needs revision. And how does this go together with your results, showing no gully erosion anymore after the implementation of the measurements?

Line 270: How are gullies restored into productive land?

Figure 5: What does “Promoted changes in cultivated and grazing land management” mean?

Line 277 – 278: Does this mean that the farmers are still dependent on external help after the implementation? How does this come together with your argument that farmers capacity to act by their own after attending to your field demonstrations?

Line 279 ­– 280: This is an important statement from the respondents, even though it is a small fraction of 10 %. You should elaborate on this why this is the case.

Line 301 – 302: Even though this is an important point, how does this relate to gully rehabilitation in detail? What do you mean with incentive mechanisms?

Line 310 – 313: This is important information and need to be discussed in more detail. What can we learn from this? Based on this outcome, how could the overall gully rehabilitation measures be improved?

Line 318 –322: What are the drivers behind? Have such trends been observed in similar studies before?

Line 326 – 334: What are the reasons and what can we learn from that. This is too descriptive and need to be discussed in more detail.

Figure 6 and Figure 7: What is the difference between those figures? What are the takeaway messages? Overall, these figures need to more explanation in the caption.

Line 369 – 371: This is not a strong argument but you mention this in the abstract as a main message of this study (empowering farmers to act).   

Table 5: What does “Treated-Non-Treated” mean? In general, this table needs more explanation. Do positive values mean yes-answers and negative numbers no-answers in the interview? What is the reason for a p-value <0.1? What is the difference to table 6?

Line 380 – 382: This belongs to the method section.

 Line 383: Please specify what you mean with a “general cue” in this context.

Line 383 – 386: This belongs to the method section.

Line 387 – 388: I cannot follow this sentence. Please revise.

Line 388: With a value of p = 0.12 there is no statistical evidence at all.

Line 390 – 396: This belongs to the method section.

Line 397 – 403: Again, this belongs to the method section. You need to explain your statistical analysis step by step first. Otherwise, it is hard for the reader to comprehend your results and interpretations. What do you mean with “a single underlying characteristic”? What is the alpha-value? What is an averaged aggregate store?

Line 404 – 409: See comment above. This belongs to the method section. How does a simple multinomial logit model work?

Footnote 2: This belongs to the method section. No footnotes please since it interrupts the reading flow.

Line 409 – 410: How does the evidence look like? What can we learn from that? What are similar studies say about that?

Line 412: To which sample do you refer to?

Line 414: How does this clear evidence look like? In this information shown in table 6?

Line 416: This is not clear. Are the farmers feel empowered after the field day demonstration or are they still pessimistic? And how does this fit to the above-mentioned clear evidence?

 Line 418 – 424: Explain why your applications are practical for any region. Based on what data is your application most likely to be deployed and maintained? You mentioned before that some farmers are still pessimistic and depend on external help to implement the erosion measurements.  You conclude that your study is in good agreement with similar studies. However, this needs to be discussed in more detail using more references than just one sentence at the end of the section.

Conclusion

The conclusion needs to be rewritten. Now the conclusion is just a short summary of the study.

Competing interests

Which authors exactly are members of the editorial board? Be transparent with this.

Thank you for your submission. This was an interesting manuscript that fits within the scope of SOIL. Greater attention to detail in certain areas of the manuscript will greatly improve its readability. It seems like a lot of generalities are stated in the intro, “global problem”, “worsening effects” but not a lot of details is provided to back up these claims. Some additional measures of the environmental and economic impacts of gully erosion would really be helpful.

The results and discussion section was difficult to follow, I tend to find combined sections to be this way. The results of the gully remediation methods don’t seem to be the real takeaway but the impacts and perceptions of local communities? Could you restructure the results/discussion to make these points more clearly? Finally, since you are working with human subjects, did go through institutional review boards? I think that some statement to this effect is needed.

Intro

Line 36: Could you provide some more context for gully erosion as a global problem? Contributions to marine/water pollution? Land degradation?

Lines 38-40: are these on the ground measurements or from model estimates? Could you specify?

Line 40: What are the economic costs? Could you provide an actual value to drive home your point about the high cost?

Line 48: change to “Across different countries” and delete “of the world”. And what initiatives and where?

Line 50: change big to large. And how much money and labor are required? It would provide context for the scale of the problem.

Lines 53-55: These two sentences seem out of place and almost read like they are the start of another paragraph/thought. Should they be added to the paragraph below?

Lines 61-62: How much worse? Again, are there measures or statistics from these references to bolster your sentence?

Line 63: What do you mean by context specific? Could you clarify?

Line 80: Add “on” after the word “demonstrating”

Line 86: change “big” to “large”

Methods

Table 1. Demographic information would seem more appropriate to present in a separate table from the geophysical parameters.

Table 2. Can you change z to “elevation” or “altitude” and add units. Can you also add units to x and y values? Are these lat or long?

Line 132-134: I think you mean higher clay content. Also, it is interesting that this soil has alternating Bt and E horizons? I’m assuming this is a loess-derived soil and that these are likely different loess packets?

Line 140: you mean “comprising”.

Lines 151-169: I had a very difficult time understanding who was surveyed, where, and why. Especially once I started reading the results and discussion section. You had focus groups and a community-wide survey? In the community-wide survey? Some of these were in kebele with rehabilitated gullies and others weren’t? Or were they all around gully rehabilitation projects and you were surveying them pre- and post-intervention?

Line 162: change “big” to “large”.

Line 161-169: the description of the respondents is not clear as written.

Line 178: Can you not repeat key twice in one sentence? I know the discussion points and informants are important but you have already established this.

Results and discussion

Line 206: Why assume a bulk density? You could use a pedotransfer function from your soil data to estimate a bulk density. See Rawls 1983 or similar papers for this.

Line 301: What do you mean by free riders?

Line 337-338: I’m not sure I follow this sentence? So benefits were mainly environmental and costs were mainly social?

Figure 6: What does the number refer to? Is this the sum of the ratings? I also think that since the mean rating can only between 1-3, you should have these on separate figures. Or a two part figure.

Table 5. Should you say treated-non-treated or pre and post intervention? Or are you comparing responses to areas from which you did not before any gully rehabilitation to areas that you did? Also how did you test significance?

Section 3.4 The description of the survey questions presented in Table 6 read more like methods and should be moved there. Also was this the same group of people that were surveyed in the community wide survey that results were presented in Table 5?

Lines 418-424: This paragraph seemed likes a better conclusion than the current concluding paragraph.

Acknowledgements. Who was the grant awarded to? Don’t you also need to include some statements about IRB for using human subjects?