Interactive comment on “ Higher response of terrestrial plant growth to ammonium than nitrate addition ” by Liming

This manuscript addresses an important and interesting question whether the effect of N addition on plant growth is different for different chemical forms of N. The authors conducted a meta-analysis to compare the effect of NH4 and NO3 addition on plant growth. The authors found that the ratio of the effect of NH4 to NO3 addition depends on plant species, while the effect on below-ground and above-ground growth is different with respect to NH4 to NO3. Further, the authors had assessed the trends in the ratio of NH4 to NO3 in global total wet N deposition. As a conclusion, the authors suggest that future assessments and predictions on the vegetation response to atmospheric N enrichment could benefit from a better understanding of plant strategies for acquiring different forms of N. Overall, the results are useful. However, some key points are not supported in the paper, which should be considered before the paper can be accepted


Q2-
The dose of N added.The references listed include studies using an N addition well above the realistic range of atmospheric deposition, which at a global scale might be 0-50 kg ha -1 yr -1 .It is not reasonable to assume that plant growth will respond linearly (as I understand the author's analysis assumes) across a wide range of doses, and it would be informative to constrain an analysis informing deposition responses to the range of likely deposition.

Response:
We agree with the reviewer that the dose of N addition in most experiments is much higher than the realistic range of atmospheric deposition.As shown by the following figure, we further separated the field N experiments into three groups in terms of the dose of N addition (i.e., 0-5, 5-10 and >10 g m -2 yr -1 ).We found the highest response of plant growth to NH4 + -N addition was under the dose of 0-5 g m -2 yr -1 , while that to NO3 --N addition occurred with the N amount of 5-10 g m -2 yr -1 .Thus, considering the reality of N deposition rate (i.e., 0-5 g m -2 yr -1 ), plant could be more sensitive to NH4 + -N than NO3 --N addition.In this revised version, we have added this figure as Fig. 7, and added sentences to discuss this important point on Lines 364-370 as: "6.The dose of N addition in most experiments in this study is above the realistic range of atmospheric deposition (e.g., 0-5 g m -2 yr -1 ).When separated the field N experiments into three groups in terms of the dose of N addition (i.e., 0-5, 5-10 and >10 g m -2 yr -1 ), we found that the highest response of plant growth to NH4 + -N addition was under the dose of 0-5 g m -2 yr -1 , while that to NO3 --N addition occurred with the dose of 5-10 g m -2 yr -1 (Fig. 7).Thus, considering the reality of N deposition rate (e.g., 0-5 g m -2 yr -1 ), plant could be more sensitive to NH4 + -N than NO3 --N addition."

Q3-
The counter-ion used.Studies with ammonium, for example, might use sulphate, chloride, phosphate etc. and the study design needs to ensure that any growth responses are only due to the ammonium itself.However, having checked references cited as studying ammonium responses in the paper's reference list, I am not convinced this is the case.I expected, for example that all the studies labelled as NH4 would apply this with an inorganic anion, but they seem to include studies applying fertiliser with a mix of nutrients with N in the form of NH4.The authors need to include full details of the chemical form of the applied N in all cases, and to ensure this is included as a factor in met-analysis.

Response:
The reviewer is right that neither NH4 + -N nor NO3 --N could be added without an inorganic anion or cation.This is also one major reason to use NH4NO3 in previous experimental studies.In this revised version, we have done two improvements to highlight this issue.First, as suggested by the reviewer, we have added the detailed information of the chemical form of applied N in all studies in the supplementary Table S1.Second, we analyzed the impacts of different inorganic anions or cations.As shown by the following figure, there was no significant difference between the anions or cations (both P > 0.05).This figure has also been added as the supplementary Fig. S3 in this revised version.We also added a few sentences to raise this point as one limitation of this study on Lines 386-389: "Lastly, neither NH4 + -N nor NO3 --N could be added in the experiments without an inorganic anion or cation.Although no difference was detected between anions or cations (Supporting information Figure S3), their effects could be included in the NH4 + -N or NO3 --N addition in this study."Response: The reviewer raised an important issue which has been neglected by the previous syntheses.As shown by a recent work by Poorter et al. (2016;New Phytol. 212: 838-855), there are many differences between plants growing in greenhouse and the field.As suggested by the reviewer, we divided the experiments into two groups, including "field" and "greenhouse".As shown by the following figure, the response patterns in the "field" experiments are similar with those presented in Fig. 4.That is, plant growth was more enhanced by significantly enhanced by NH4 + -N than NO3 --N addition (panel a).However, greater responses were found for plants growing in greenhouse (panel b).These results are important to explain the general patterns shown by Fig. 4, so we added this figure as supporting Fig. S1 in this revised version.We also have added sentences in the "Results" and "Discussion" sections to highlight the issue raised by the reviewer as: Line 226-231: "It should be noted that these response patterns were calculated on all available studies, including experiments which were conducted in both greenhouses and the field.As shown by the supporting information Figure S1, the general patterns of Fig. 4a were similar with that across the field experiments.However, greater responses of plant growth to both NH4 + -N and NO3 --N addition were found across the greenhouse experiments." Line 353-357: "4, The different response patterns between greenhouse and field experiments (Supporting information Figure S1) suggests that the general plant responses to N addition in this meta-analysis (i.e., Fig. 4) could be larger than the natural ecosystems.More research efforts exploring the differences and similarities between plants growing in the greenhouse and field are needed (e.g., Poorter et al., 2016).

Q5-As clearly argued by Referee 1, plant growth responses to N can be constrained by a range of other environmental factors, such as climate and water availability. Soil characteristics are also likely to critical in determining responses to added nitrate and ammonium. These factors need to be considered in the authors' analysis.
Response: Based on the field experiments, we have analyzed the dependences of plant response upon environmental factors, including mean annual temperature (MAT), mean annual precipitation (MAP) and latitude.As shown by the following figure, MAT and latitude only affected plant growth response to NO3 --N addition, while MAP negatively influenced plant growth responses to NH4 + -N and NO3 --N addition.These patterns collectively suggest that plants growing in colder and drier regions are more sensitive to NH4 + -N and NO3 --N addition than that growing in warmer and wetter regions.This finding is inconsistent with the response patterns of net primary productivity to N addition at the global scale (LeBauer & Treseder, 2008).We have added this figure as Fig. 6 in this revised version, and added a few sentences to discuss the difference between these patterns and those in the previous studies on Lines 358-363 as: "5. Mean annual temperature (MAT) and latitude only affected plant growth response to NO3 --N addition, while mean annual precipitation (MAP) negatively influenced plant growth responses to NH4 + -N and NO3 --N addition (Fig. 6).These patterns collectively suggest that plants growing in colder and drier regions are more sensitive to NH4 + -N and NO3 --N addition than that growing in warmer and wetter regions.This finding is inconsistent with the response patterns of net primary productivity to N addition at the global scale (LeBauer & Treseder, 2008)." Q6-It is also surprising to me that several relatively recent long-term field experiments designed to compare the effects of ammonium and nitrate directly are not included in the papers used in the analysis.These include studies of an ombrotropic mire (Sheppard et al., 2014, mentioned in the intro but not included in the meta-analysis), of a Tibetan alpine meadow (Song et al. (2015) and maquis vegetation (Dias et al., 2017).Since these studies directly compare the growth and ecological effects of the two forms of nitrogen, they seem a more important source of information than a comparison of independent studies of the two forms which may be biased by the many differences in experimental design which are not accounted for in the authors' analysis.
Response: As pointed out by the reviewer, long-term studies with inter-comparison between ammonium and nitrate addition are more direct to answer the question in our study.We have checked the references which suggested by the reviewer.Below are the reasons for why they were excluded in our analysis: • Song & Yu (2015) conducted an 8-yr N fertilization experiment to examine the effect of long-term N addition on ecosystem stability its driving mechanisms.They found the form of N (ammonium and nitrate) did not affect the stability.In their Fig.1c, no difference in the responses of community biomass to ammonium and nitrate addition were detected.We didn't include the data from this experiment because our study only included species-level data.However, their results have been discussed in this revised version on Lines 301-306 as: "Furthermore, NH4 + -N addition show a higher stimulation on plant growth but a lower variability among plant functional types than NO3 --N addition (Fig. 4), contrasting the result of Song & Yu (2015), which observed the synchrony among functional groups were not affected by N form after 8-year fertilization in an alpine meadow on the Tibetan Plateau.Our result suggests that future N deposition with more NH4 + -N input will more evenly enhance the growth of different plant functional types." • Sheppard et al. ( 2014) reported the interesting responses of relative cover changes among species to ammonium and nitrate addition in an ombrotrophic bog.They found some species (e.g., Calluna vulgaris) preferred for ammonium N while others (e.g., Hypnum jutlandicum) for nitrate N. We didn't include this study in our analysis because the data were reported with coverage rather than dry mass (or biomass).However, we also cited this work as the evidence for the species-specific preferences for ammonium and nitrate on Lines 309-312 as: "However, within an ecosystem, the preference for ammonium or nitrate could varies among species.For example, Sheppard et al. ( 2014) reported the relative cover changes of some species in an ombrotrophic bog (e.g., Calluna vulgaris) preferred for ammonium N while others (e.g., Hypnum jutlandicum) for nitrate N." • Dias et al. ( 2017) used a 7-yr N fertilization experiment to study the effect of N deposition on a Mediterranean maquis community.In their Table 1, they reported the responses of leaf traits to different N treatments.There are two reasons we didn't include these data.First, this study includes three N treatments, including 40A (i.e., 40 kg ha -1 yr -1 NH4 + -N as a 1:1 mixture of NH4Cl and (NH4)2SO4), 40AN (i.e., 20 kg ha -1 yr -1 NH4 + -N and 20 kg ha -1 yr -1 NO3 --N as NH4NO3) and 80AN (i.e., 40 kg ha -1 yr -1 NH4 + -N and 40 kg ha -1 yr -1 NO3 --N as NH4NO3).Their results reflect the difference between NH4 + -N and NH4NO3 rather between NH4 + -N and NO3 --N.Their finding of no difference between NH4 + -N and NH4NO3 is consistent with the finding in our study.Second, only leaf trait (i.e., mg leaf -1 ) data were reported in the Table 1, and the total leaf biomass were not provided in this study.

KNO 3 Ca(NO 3 ) 2 NaNO 3 (NH 4 ) 2 SO 4 NH 4 Cl
established plants, to provide a clear guide to likely responses under field conditions.If other studies under controlled environments are to be included, this needs to included as a separate factor in the meta-analysis.