The manuscript “Elevated CO2 and N gradually weaken the influence of
diversity on ecosystem stability” shows how two global change drivers interfere with stability in experimental grasslands. The paper is well-structured, clearly written, and the analyses seem correct. I have the following questions and comments:

a/The methods say that species that were not initially seeded were removed. There is also mention of realised richness, implying that species went extinct in some subplots. But then, species ricness is also controlled for, as it is an indepenent variable in this kind of studies. Does this mean that species were reseeded when lost from a plot at the end of a growing season? If so, this may be an important point to discuss. Ideas around temporal stability typically consider fluctuations of coexisting species, either because of internal dynamics, because of environmental fluctuations, or both. How to position the experimental approach followed in the CC experiment within this framework?

b/There exist quite some papers showing how global change drivers affect richness effects on productivity. It could be interesting to check to what extent the presented results confirm/refute some of their results. I am thinking of Hong et al. 2022 Ecology Letters, Steudel et al. 2012 Ecology Letters, Baert et al. 2018 Ecology Letters. Please only do this if you find it interesting.

c/There exists a meta-analysis on global change drivers on temporal stability in grasslands, which the paper cites (Su et al. JEcol.). It would be nice to see a comparison between the results in that paper with the results in this study. Of course, one could probably not say much about the weaking with time, but maybe at a more general level a comparison would be useful (e.g. if the temporal mean effects of richness on stability change with N and CO2 as found for the other global change drivers in Su et al.)?

d/There also is a preprint that sits very close to the submitted manuscript (https://www.authorea.com/users/754665/articles/724523-short-term-versus-multi-decadal-responses-of-community-synchrony-biodiversity-and-stability-to-multiple-global-change-drivers). Would be good to explain in what sense the presented analyses go beyond / complement this other manuscript.

Specific comments:

P8, L34-36: “The dependent variable and the measure of plant diversity were natural-log transformed for the analysis.” What is the justification for this transformation and wasn’t using glmm an option instead? This might also avoid having to add 0.1 to the transformed richness values.
P8, L53: Has temporal autocorrelation of residuals been checked?

P10: Isn’t the comparison between planted and realised richness interesting for more than methodological reasons? What does it mean when both give the same results?

P12, L16: I don’t think May 1974 is relevant here, as this paper talks about asymptotic stability, while the submitted manuscript talks about temporal stability. These two are not necessarily related (see Arnoldi et al. 2016, Journal of Theoretical Biology 389). Also the citation to Tilman 1996 is not relevant, since this is not a theoretical paper.

Thank you for the very thorough revision of the manuscript.