Dear Dr. Duku,
I had close look at manuscript, the review and your responses (which propose mostly the right way for improving the study). In line with both reviewers, I think the submission needs to be thoroughly revised with respect to a) structure and clearness of the presentation, b) calibration of the SWAT model c) proposed measures for ecosystem services and c) an even more thorough reflection of the reliability of the proposed approach to quantify ecosystem services.
Allow me to explain points b) and c).

b) The use of the NSE in case of strong seasonal discharge data is tricky, as parts of the signal reflects simply the seasonality of rainfall input (Schaefli and Gupta HP, Do Nash Values have value). This is particularly difficult when working with monthly data. The null model is thus not the overall average, but the average annual cycle of monthly/ or better daily discharge values. A comparison of SWAT against the NSE of these simple to calculate null model is indispensable to judge how much SWAT explains on top of this.

c). It might be not an uncommon practice to use unverified models for such purposes, but it is a bad practice, and not acceptable from my point of view. As far as is see, you have a 2 years data set for nitrogen and sediments, hence you could go for a validation either using a common split sampling (calibrate on 1 and validate on the other) or you can leave out very second data point in the calibration and using the remaining points in a scatter plot for verification. In this respect I would be interesting to compare also concentrations and not just the loads
I agree with reviewer two that the average soil water content in the growth period, is not conclusive whether a plant will survive. A duration curve (days in the period were soil moisture is below a threshold), and the duration of these periods, where soil water is below critical thresholds seems much more appropriate and straight forward to calculate.
Allow me finally to acknowledge your attempt to address an extremely difficult issue, for which nobody holds the philosophers stone. Such an attempt is thus insufficient per default and provides therefore many opportunities for scientific learning, about the underlying limitations we face today. Your paper will have impact if you particularly also address the last point. Scientific learning is not where the good success stories are, scientific learning starts if we also clearly share the deficiencies.

Best regards,

Erwin Zehe

Dear Dr. Duku,

I am delighted to tell you that both reviewer recommend publication of your study in HESS. Reviewer 1 suggested a few minor changes to furhter optimize the already very nice presentation of your results for the reader, which are worth to be considered.

Best regards,

Erwin Zehe

Dear Dr. Duku,

I am very much pleased with the final version of your manuscript, which is herewith accepted for publication in HESS.

Best regards,

Erwin Zehe