Dear authors:

Thank you for making open discussion. Now your manuscript received two referee comments and you provided your responses. Both referees evaluated that your manuscript has good scientific significance, good presentation quality, and fair scientific quality. The relatively low evaluation of scientific quality might come from the limited ecological and physiological measurements, as I pointed out at the initial stage. Nevertheless, the referees found adequate merits in your study as one of a few experiments on seasonal drought impacts. Referee #1 recommended using “recovery” instead of “resilience. This contains not only terminological but also conceptual issue, and I agree with you to use the latter term. Referee #2 also provided insightful comments (even with a figure), and I’m convinced that considering these comments would improve scientific quality of this manuscript. Finally, I conclude that the manuscript should be reconsidered after major revisions.

Dear authors:

Thank you providing your comments and a revised manuscript.
I asked to reviewers to examine your replies and got the second (follow-up) comments as you see below.
The referee recommends you to make major revisions, considering your replies to another referee.
Therefore, I conclude that your manuscript should be reconsidered after major revision.

# second comments from referee #1

RECOMMENDATION: MAJOR REVISION

GENERAL COMMENTS
The authors made many changes in an already good manuscript, based on earlier reviewer comments. To me, the story reads well and was based on a sufficiently substantial amount of data, so I agree with the authors to not report on the ecophysiological data here. However, at this stage I feel obliged to recommend major revisions again because two out of three of my main previous comments do not seem sufficiently or correctly addressed. At first sight, this also applies to an important comment by Reviewer #2:
1) Analyses of drought stress index “Is” were performed and reported on, but it seems to me that “Is” was not calculated correctly. Please see the specific comments.
2) According to your replies, the dataset and R script would be given, but I cannot find any link to these files in the manuscript. If I somehow overlooked the link, please refer to where it is shown, or otherwise insert the link. Also, in general, I highly recommend to refer in your author replies to where in the manuscript changes have been made, e.g. Lines X -Y, and if relevant enough (not simply a technical correction, ...), also quote these lines.
3) Reviewer #2 made an excellent remark on variation in inter-annual ANPP vs intra-annual variation in ANPP across treatments. I at first sight more or less agree with your replies, but it seems you did not really incorporate your answers in the manuscript. Or if you did, please refer to where this information has been inserted in the manuscript, see also my previous comment. While Rev#2’s remark may not be about the primary research questions, I agree on its importance and it should be briefly addressed in at least a few lines or a paragraph of Discussion.

SPECIFIC COMMENTS
Line 145 – I see that you calculated REW as a function of soil water potential (MPa) instead of soil water content (no unit, or %). While I applaud the measurements of SWP, I do think that Eq. 2 and the threshold of 0.4 you refer to can only be used in that form based on SWC, as was done in the publications you cite. Using SWP instead of SWC is not equivalent because their relationship is not linear, cf. pF curves – and definitely the threshold of 0.4 would change. So could you please convert the water potentials to SWC, then calculate REW per day, and sum daily shortages below 0.4 over time? In case I misunderstood something here, or you know an alternative stress measure based on SWP, feel free to calculate that instead but explain well in your reply and in the manuscript M&M.
Table 2 – Why are the “Is” values integers? This, in combination with your previous reply to former Line 341 makes me think that you misinterpreted the meaning of “Is”: it is not the number of days with stress (when REW < 0.4), so not simply a “yes/no response”, but the sum of REW’s below 0.4. For example, if we have a series of REW: 0.6 0.7 0.5 0.4 0.3 0.1 0.2 0.3 0.7, then Is = 0 + 0 + 0 + 0 + 0.1 + 0.3 + 0.2 + 0.1 + 0 = 0.7. To me it seems as if you would come to a wrong value of Is = 4 here = number of days with soil moisture stress, but not Is. Please comment or clarify, and correct in the Table and text if needed.

TECHNICAL CORRECTIONS
Line 145 – “Eg” should be “Eq.”
Line 147 – Symbols for field capacity vs wilting point were confused.
Line 355 – Suggestion to remove “the” from “the Is”.
Line 526 – The reference list was given twice in this manuscript version.
Table 2 – Some “Is” values were highlighted in violet. Please remove.
######

Dear authors:

Thank you for providing a revised manuscript and for making meaningful discussions, including a follow-up response to mail-based comments. I studied the materials and concluded that the manuscript was adequately revised on the basis of the referee suggestions. Both referees gave positive evaluations in terms of scientific significance of the drought experiment, and the amendments made by the authors improved scientific quality. The outcomes of this study may carry implications concerning the impact assessment of extreme weather events on ecosystems. Finally, I recommend that the manuscript would be acceptable for publication.