The manuscript presents results from a large-scale field experiment showing how leaf and plant eco-physiological processes scale with leaf nitrogen concentrations, and that this scaling is similar for both symbiotic nitrogen fixing trees and non-fixing tree species. The results are based on measurements in four field sites, in young trees growing under four nutrient manipulation treatments, therefore representing an impressive work effort and a solid outcome when analyzing all species together (although patterns for specific species differ but barely discussed).
The work is important and interesting, connecting together several pieces of the puzzle of the effect of leaf N and N-fixation on plant performance. The manuscript is long and contains many analyses which seem to be of different levels of importance. The length of the manuscript and complexity of analyses and figures sometime obscure its main findings. This could be improved to highlight important and unique results and avoid "fishing" for more results by over-analyzing the data. Below are some major and minor comments to the authors:
- Results:
- First, the per mass concentration of nitrogen in the leaves of leguminous fixers seems to be more variable (probably in response to the different fertilization treatments and growing conditions) compared to a narrower range of N concentrations in the leaves of the actinorrhizal N-fixers. This also affects the relationships with the physiological parameters.
The positive correlation between leaf N concentration per unit area and LMA (figure 2a) is trivial, it is as saying that the larger the leaf mass the more N it contains (as leaf area is in the denominator in both x- and y- axes). Furthermore, supplementary figure 3a shows that this is not a linear function, which also makes sense. The heavier leaves (per unit area) contain more support tissues (e.g. collenchima and sclerenchima) and therefore the increase in leaf mass does not result in a 1:1 increase in leaf N mass.
The result in figure 2b is not very different in this respect. The negative correlation between leaf %N and LMA is like saying that the larger the leaf area is it will contain more N - not very surprising. In fact, any alternative to these two findings would be surprising.
These results are also discussed in the discussion (starting around line 380), but I do not agree with some of the claims, which ignore the simple facts here. Also, it is important to take into account the method of measuring LMA for the species. For example, the compound leaves of acacia and Robinia may result in an especially high value of LMA depending on whether the central rachis of the leaf is weighed or not.

• The methods mention that within the pairs of trees they were 5 meters apart, and also that fertilization was applied at a 5 m radius around each tree. This brings about a few questions like: Did the fertilization areas of two adjacent trees overlap? Did you apply twice as much fertilizer per area in such overlaps? How do you control for competition between the two adjacent trees for the applied nitrogen (and P)?

• Line 165: An artificially strong light (1800 or 2000) level was applied for the gas exchange. As far as I understand these light levels did not match the measured light levels the trees were exposed to during the time of measurement (which I assume varied along the morning and afternoon hours). Did you allow time for light-acclimation before you conducted the measurements?

• Throughout the results and discussion the authors repeat the phrase that "leaf-level N limitation did not..." change or affect the physiological parameters measured. However, in most of the results (e.g. figure 3a,c,d) there is a different model for plants that are not N limited at the leaf level (including both fixers and non-fixers) and for the N-limited non-fixers. I could not resolve this contradiction regardless of the numerous times it appeared, indicating that the phrasing or terminology are either wrong or confusing. For example, in line 304: I could not understand the sentence "Both Asat and gsw were higher in trees that were not N limited at the leaf-level, leading to no difference in WUEi and δ13C due to leaf-level N limitation." This sentence contradicts the previous one, and the results in figure 3, which show higher Asat and gsw for leaves that are N limited (dashed line above solid black line).

Minor comments:
- The term "N fixer status" is not clear (intuitively it relates to something of N-fixers and not to the distinction between fixers and non-fixers), especially in the abstract, before it is defined.
- Line 57: The authors suggest that N fixation rates exceed 100 kg N per ha per year, but we know this value is at the high end of current estimates of actual N fixation in most studied plants to date.
- Line 157: Please explain how intrinsic WUE was calculated taking into account the four fertilization treatments per species and all the replicates for each species and treatment combination.
- Line 263: p-values in parenthesis do not match those in supplementary figure 2a,c.
- Figure legends: check and correct confusing sentences (e.g. line 656).
- Line 280: it is not clear what is "N fixer mixed-model mean regressions..." - do you mean the slope of the regression? or the intercept? or both?
- Lines 282-290: I could not understand the importance or meaning of the relationships described in these lines.
- Lines 392-3: Something in this sentence does not make sense.

The revised manuscript has greatly improved, and the removal of large parts of the analyses served it well. Still, the manuscript is based on a lot of data and a complex set of analyses which support the claims of the authors, but are not easy to read and follow.
I have one major comment regarding the re-analysis of the data from Adams et al. 2016 - One of the most surprising and problematic elements remaining in the revised version, was that Supplementary note 3 mentions that “We combined our experimental dataset with a portion of the global dataset from Adams et al.” but there is no clear explanation about which part of the global dataset was used and which part was omitted and why. This sentence also made me question whether the entire dataset was analyzed (for figure 4) or also only a subset of it. I did not find any mention of this in the methods section of the main text, nor a mention of the size of the dataset (how many species, replicates, etc). The only explanation given in supplementary note 3 is “we extracted the woody data from the global dataset, by omitting data from graminoids and forbs” which is based on wrong assumptions for several reasons, for example, non-woody legumes are probably included in the data of fixers and in your analysis.

Here are a few additional minor suggestions and comments:
Line 36 in the abstract says that stomatal conductance does not scale with leaf N but then gsw is listed in line 43, which is confusing.
Line 70: Citing a relatively new paper by Wurzburger seems appropriate in the context of this sentence: Wurzburger, N., Motes, J.I. and Miniat, C.F., 2022. A framework for scaling symbiotic nitrogen fixation using the most widespread nitrogen fixer in eastern deciduous forests of the United States. Journal of Ecology, 110(3), pp.569-581.
Line 88: Wolf et al. 2017 show plasticity in plant nitrogen concentration, not at the level of leaves specifically. But see Dovrat et al. 2020 for an example of such plasticity in the leaf N concentration.
Line 115: “relationship between N and photosynthesis”, I think “leaf N” should be added.
Line 164: The authors claim that their analyses can provide answers regarding the effect of the source of tree N – but, I suggest to be more cautious about such claim. Can you determine the source of N in each tree? N-fixing trees that were fertilized with N can contain N from both sources (soil and fixation), but they can also be using primarily N from the soil, and I did not find any results resolving this (only at the end of the discussion this is mentioned citing another publication from this experiment). Furthermore, the sentence is summing-up the research questions, so such a general claim might not be needed and a more accurate wording could be better in this case.
Line 576: the sentence is not clear – why N fixation is expected to be sustained under excess N supply? The opposite is more likely
Line 835 – figure captions: “Colored symbols represent different species” – it would be useful to explain if dots represent individual data points for all replicates and to note the number of replicates in the caption.
Line 841: caption of figure 2 – there are no panels in this figure.
Caption of figure 4 – it would be useful to note the number of species and number of replications in each group in this analysis (same for supplementary figure 4).
Line 880: caption of figure 5 – I think b and d have linear-log10 scales and therefore slopes.