dear authors,

Your manuscript has now been seen by two reviewers. Both of them provided some excellent suggestions to improve your manuscript. While I fully agree with both reviewers that a quantitative synthesis of the various meta-analyses would be a worthwhile effort, this would amount to an enormous amount of extra work. It would require access to the raw data of all meta-analyses (to remove overlap between datasets; calculating an average across meta-analyses without removing such overlap would amount to pseudo-replication). About 10 years ago I was involved in such a "meta-meta-analysis" for just one variable (soil C stocks under elevated CO2), synthesising results for only 4 meta-analyses (see Hungate et al. 2009, GCB). That analysis alone resulted in an enormous amount of work; doing the same thing for 16 meta-analyses and 42 variables is not realistic.

However, I agree with both reviewers that "synthesis" in the title implies that this is exactly the kind of work you would be doing. A proper synthesis would also address points like the one brought up in comment 4 by Dr. Dijkstra. Perhaps the authors could choose a phrase that more accurately describes their approach? "Comparison" or "assessment" could both work.

Both reviewers provide good suggestions to add depth to your discussion, and you indicated you would be willing to incorporate these. In the absence of a true quantitative synthesis of the meta-analysis, addressing comment 2 by Dr. Dijkstra and comment 21 by Dr. Baker seem especially important.

dear authors,

Thank you for submitting this revised manuscript. The concerns that were raised by the reviewers have been addressed satisfactorily. Having read this revised version, I would like to suggest a few minor edits. Once these are incorporated, the manuscript can be accepted (line numbers refer to the revised version):

- L192: ".. diagrams illustrate..." (not "..diagrams illustrates..")
- L293. what exactly are "direct systems"? and what conditions are you referring to here? soil moisture? please clarify.
- L308-310. This statement should be backed up by references (e.g. primary studies that show the impact of the factors you mention).
- L398-401. This is an important point that shows up too late in the manuscript. I suggest moving this to the end of the introduction (e.g. somewhere between L73-79), or to the methods section (e.g. at the end of section 2.2).

dear authors,

Thank your for addressing the remaining issues. I am happy to inform you that your manuscript is now accepted for publication.

all the best,
Kees Jan van Groenigen