Dear authors,

many thanks for providing an updated manuscript, in which you have addressed most of the concerns the reviewers raised. However, based on my own reading, I have a range of comments on the revised manuscript, which I kindly ask you to address/correct before this manuscript may be publishable in Biogeosciences.

Please add from your responses to reviewer #1 the response to the major Methods comments to the manuscript.

I wonder why you did not test whether the regression slopes statistically different across the years - or whether despite meteorological differences, the relative placing of the treatments was conservative across years? This would seem to make your study more robust.

In the beginning you state your hypothesis that you would find priming. But this is not discussed at all later in the manuscript. Instead, you only discuss the effects on plant production - why? presumably because you did not find any evidence for priming? Isn’t that worth noting? And if it’s not - why do you mention priming then in the introduction? Please rework the manuscript to be consistent between the hypotheses you state and your answers/results in Discussion and Conclusion

The use of the word resilience is somewhat inadequate. Your manipulation had low effects, which were statistically insignificant (probably partly because of the high variance in the measurements?). So is the lack of finding a response not more a sign for a low signal to noise ratio rather than an insensitivity of the ecosystem? After all, there is a systematic trend with litter addition for most variables - which does not suggest insensitivity to litter addition (or resilience). You need to define what you exactly mean by resilience. That of the N:P ratio, biomass production, or what else?

Minor comments:

Abstract:

“Global climate change has generally increased net primary production (NPP) which leads to increasing litter inputs.”

I don’t think that there is evidence for this (globally and in the past), and you do not provide any suitable reference. Future litter inputs may increase in dry climates because of CO2 fertilisation - but this is a hypothesis, not a fact. This sentence does also not link to projections of the ESMs you refer to later in the abstract. Please revise (or remove) this first sentence.

L 35 ff: It seems more logical to start with the “more realistic” result, than the outlier

L37: remove “quite”. Either it is resilient or it is not.

L38: Why does this highlight complexity? If you add litter within a certain range of ambient litter fall, nothing statistically significant happens (though the trends go into similar directions than with higher added litter), and you only see a statistically significant results once the perturbation becomes large. This is more an expression of the inevitable variability of ecosystem measurements (inducing noise) than the complexity of the ecosystem. Please revise or remove.

L61: I cannot see how Hungate 2003 makes any statement about the importance of litter dynamics in terms of feedbacks between plant growth and climate - the word litter does not appear in this paper.

L63: Please update Houghton et al. 2001 to something more up-to-date. It is inconsistent to quote IPCC 2013 for the impacts, but rely on 12 year older scenarios, which have not been used in IPCC 2013

L66: here and in the following: CO2 has a subscript “2”.

L67: remove “therefore” - this does not lead logically from the previous sentence.

L81: Why is here a quotation mark (and where is its end?)

L105: Here or somewhere else in the introduction you need to make the link between increased litter input from elevated CO2 (which generally does not increase the input of N and P), which you expect for your ecosystem, and the fact that when adding litter you add also new N and P to the ecosystem. Particularly, because in L 110 you expect to observe a priming effect - but how do you separate this from the (trivial) increase in the pool sizes because you have added N and P to the system - essentially providing fertiliser (even if in an undecomposed form).

L110 better use hypothesise?

L159: I would not consider the predictions of Earth system models “very realistic” to make statements about the changes in a particular ecosystem. Use “plausible” instead.

L162: Please be consistent between representing ecosystem stoichiometry either as C:N:P (L 138) or C:N, C:P, N:P

L178: Clarify: was this the end of the growing season? - use of the May-September precipitation earlier suggests otherwise. If it was not the end of the growing season - would this affect your conclusions about plant growth?

Section 3.3. They may not be significant, but it’s probably worth mentioning that there was still an notable trend with litter addition in most variables.

L246: decreased.

L 281ff: I do not understand why you single out temperature and precipitation here, because surely these ESM simulations do account for changes in CO2, temperature and precipitation, so the predicted 10-60% increase already includes climate change. If this was not the case, I would expect a more elaborate discussion as to where your estimate of 10-60% comes from.

L 287: Not really. There was also an increase in the other treatments - it was only not significant.

L292: here and in the following remove “quite”

L292: I do not understand the use of “only” here. Please reword.

L293: remove “indeed”.

L 294: remove “but”.

L296-L300: repeats more or less word-by-word the preceding sentences. Remove.

L300f: This would not be surprising at all

L307: Given that plants do not decompose organic matter - is this really a surprise?

L 311: is there any evidence for this? I do not follow why low increases in microbial biomass are necessarily a sign for changed carbon-use efficiency. The time-scale of your experiments does not allow for such a statement, because for instance, a large fraction of the microbial biomass that decomposed the litter was already dead when you took your measurements, a long time after litter addition. Since you do not provide any data on soil respiration and remaining litter, how can you be certain that the microbial carbon-use efficiency changed. What about changes in composition?

L320f: On that note, you should discuss that rain fall in 2009 and 2011 was only 2/3 of the average - a low change in biomass may be associated with this.

L323: Seems to belong to the following section. Remove “indeed”.

L325: This sentence is repeated word-by-word in the following section. Remove here.

L336: This sentence repeats the content of what was said in L330. Merge or remove.

L339: competition for what? why is this relevant?

L340 A higher photosynthate concentration leads to a higher C:N and C:P not to a decline. A decrease in C:N cannot explain increased photosynthate concentration - rather the opposite is true. What are you trying to say here?

L360: Your study does not support this statement. The litter you added is qualitatively not the same as the litter to be expected under elevated CO2. Furthermore, the addition of litter has different consequences than the increase in litter fall under elevated CO2, which would mostly result in increased C:N:P (as long as other N and P inputs do not change) and therefore further reduce nutrient availability.

L365: This is not a conclusion from your study and should thus not appear here, but in the introduction (if it was necessary at all, which I don’t think it does as your manuscript does not talk about N and P fertilisers as such).

L368: Why surprisingly?

L377: Note that these models do largely not account for either N nor P. I’d therefore refrain from calling them “realistic”.

L379: I’m confused here: why do you state increased in “NPP with nutrient imbalance”, when the increased input had the C:N:P of litter from this ecosystem. This does not seem to be an imbalance by itself.

L380: Again, I don’t follow why you make a distinction between the effects of temperature, precipitation and CO2 here. See my comment above. Note that your experiment does not mimic the effects of elevated CO2, because it assumed that the litter stoichiometry was unaltered.

L603 (and the following figure captions): “ in a steppe community of northern China”? So these results are not from your experiment, but some other place? Please simply remove (unless you really do not show the experimental data)

L 604: Please be more precise what you mean by year-effect? I assume that this means that for all litter treatments, there were consistent differences between the years?

L614 and in the following: refer to figure 2 for this information - do not repeat the same information four times.

Figure 1: It is difficult to spot the differences amongst the years. Consider representing smoothed temperatures and precipitation in a super-imposed way.

Figure 2-5: please add letters to identify panels. Note that the text suggests that these labels exist.

Figure 4 and 5 are illegible. Both axes and statistical relationships are too small. Note that there are also considerable errors in the labelling of X and Y axes, as well as in the placement of the significance letters. I also do not follow why in some panels of Figure, the letters have been omitted (as well as in Figure 5 also some regressions). I’d advise to remove the regression statistics from the Figure and provide these in a different form.

please correct the units of the X-axis label in Figure 3/Phosphorus column.