Elevation Alone Alters Leaf N and Leaf C to N Ratio of Picea crassifolia Kom. in China’s Qilian Mountains

Round 1

Reviewer 1 Report

The authors have made many improvements to this manuscript.  I like the addition of the map.  The authors should add more information about the sample analysis.  How were the soil samples sieved?  How were the depths combined? The added text needs to be read through for proper sentence structure.  The sections are too small in the discussion.  They should be combined to discuss similar factors affecting leaf stoichiometry. 

Lines 63-64: No acronyms are needed for precipitation and temperature if they are not used again in the abstract.

Lines 181-184:  This sentence needs to be rewritten to be clearer.

Figure 1.  I appreciate the addition of the map.  Can you add what the colors on the map mean?  Is the maximum elevation 3300 or 3200 m?

Line 344: Shorten the section title.

Lines 384-385:  “Soil pH decreased with increasing elevation”.

Table 4: Are these values R or R²?

Lines 557-561:  This sentence should be broken into a sentence on each topic.

Author Response

Dear Reviewer,

Thank you for your positive feedback on our work before. These comments incredibly improved the quality of our manuscript (forests-1363822). We have carefully addressed all the comments. In this letter, we provided a point-by-point explanation of the revisions and noted them in red. In addition, many minor edits were also done throughout the text. Since the minor edits are too many to list, we sincerely invite you to check it again. We provide the revised version with track, and number of pages and lines in the present response letter are all based on it. We hope that our revisions now could meet your requirements.

Sincerely

Niu and co-authors

 

Comments

The authors have made many improvements to this manuscript. I like the addition of the map. The authors should add more information about the sample analysis. How were the soil samples sieved?  How were the depths combined? The added text needs to be read through for proper sentence structure. The sections are too small in the discussion. They should be combined to discuss similar factors affecting leaf stoichiometry. 

Response

Thank you very much for your valuable suggestions. We have added more information about sample analysis according to your suggestion. Soil samples were air-dried and ground through a 2 mm mesh sieve prior to laboratory analysis (L109-110, P3). The soil nutrient values of 0-0.4 m in the study area were the average of the measured values in each layer (L124-125, P4).

We have read through the text added to the manuscript to ensure correct sentence structure, and modified the inappropriate sentences.

In the discussion section, we integrated the effect of climatic factors (MAP, MAT) and soil properties (soil pH, SOC : STN) on the leaf stoichiometry of P. crassifolia, respectively (4.3.1, 4.3.2) (L274-334, P9-10).

Comments

Lines 63-64: No acronyms are needed for precipitation and temperature if they are not used again in the abstract.

Response

Thanks for your valuable suggestion. We have deleted the acronyms for mean annual precipitation, mean annual temperature and soil organic carbon to soil total nitrogen ratio in the abstract (L29-30, P1).

Comments

Lines 181-184:  This sentence needs to be rewritten to be clearer.

Response

Thank you very much for your suggestion. In order to make it clear and ensure the structure of this manuscript, we deleted this sentence (L55-59, P2).

Comments

Figure 1.  I appreciate the addition of the map.  Can you add what the colors on the map mean?  Is the maximum elevation 3300 or 3200 m？

Response

Thanks for your valuable advice and kindness. The colors from green to red on the map mean elevation from low to high. And the maximum elevation is 3200 m. For better visualization, we adjusted the map and added a color-elevation legend as following (L119-121, P3):

 

Fig 1. The five sample sites at 2400 m, 2600 m, 2800 m, 3000 m and 3200 m in the study area. 

Comments

Line 344: Shorten the section title.

Response

Thanks for your comments. We have shortened the section title to “Variation in soil properties and P. crassifolia leaf stoichiometry across elevation” (L139, P4).

Comments

Lines 384-385:  “Soil pH decreased with increasing elevation”.

Response

Thanks for your kindness. We have improved this sentence according to your suggestion (L159, P4).

Comments

Table 4: Are these values R or R²?

Response

Thank you for your comments. These values are R.

Comments

Lines 557-561:  This sentence should be broken into a sentence on each topic.

Response

Thank you for your valuable comments. After consideration, we revised the section “Effect of elevation on soil properties” (4.1) (L218-236, P9), and deleted this sentence.

 

Author Response File: Author Response.doc

Reviewer 2 Report

This study is a solid addition to the field of stoichiometry and the more information we have about how nutrient concentrations and ratios change over gradients of potential drivers the better we will understand the dynamics of vegetation nutrient strategies.  My major concern about the study is that soil nutrients were presented as concentrations rather than pools.  Although some have argued that concentrations may be more relevant to microbial communities, most literature presents the soil nutrients as pools (concentration times bulk density).  Given the wide variation in soil types in this study bulk densities may vary widely and might change the relationships between leaf and soil nutrients.  I realize at this point it may be too late to go back and measure bulk density but in any case I feel the authors should at least address the issue and justify their decision.

Also given the very low MAP, there is a case to be made that tree growth at all these sites is primarily water-limited and nutrient limitation plays at most a secondary role.  Again I think the paper would be stronger if the author's addressed this in the conclusions. 

With respect to the nutrient ratios presented please state clearly if the ratios presented are mass based or mole based ratios probably in the methods section.  Also with respect to 14 and 16 being critical ratios and useful in determining whether N or P is limiting to productivity, this idea was based on an excellent study of managed wetland herbaceous ecosystems in Europe (Korselman and Mueleman) and the similarity of the values they found to the Redfield ratio.  I have great respect for that set of studies but feel it is inappropriate to apply those values widely to all terrestrial ecosystems as is unfortunately often done.  I would strongly recommend at least rewriting the section of the discussion on N versus P limitation based on the leaf N:P to be much more speculative and follow up with this would need to be confirmed with fertilization studies.  Several studies have shown that high P concentrations are essential for fast growth but in these high elevation systems I suspect rapid growth to compete for a spot in the canopy and access to light is not essential for survival.

Author Response

Dear Reviewer,

Thank you for your positive feedback on our work before. These comments incredibly improved the quality of our manuscript (forests-1363822). We have carefully addressed all the comments. In this letter, we provided a point-by-point explanation of the revisions and noted them in red. In addition, many minor edits were also done throughout the text. Since the minor edits are too many to list, we sincerely invite you to check it again. We provide the revised version with track, and number of pages and lines in the present response letter are all based on it. We hope that our revisions now could meet your requirements.

Sincerely

Niu and co-authors

 

Comments

This study is a solid addition to the field of stoichiometry and the more information we have about how nutrient concentrations and ratios change over gradients of potential drivers the better we will understand the dynamics of vegetation nutrient strategies.  My major concern about the study is that soil nutrients were presented as concentrations rather than pools. Although some have argued that concentrations may be more relevant to microbial communities, most literature presents the soil nutrients as pools (concentration times bulk density). Given the wide variation in soil types in this study bulk densities may vary widely and might change the relationships between leaf and soil nutrients. I realize at this point it may be too late to go back and measure bulk density but in any case I feel the authors should at least address the issue and justify their decision.

Response

Thank you very much for your valuable comments. We realized this issue and addressed it in the text as following:

Notably, nutrient concentrations are not only relevant to microbial communities, but are also related to bulk density when they are presented as nutrient pools (concentration times bulk density) [59-61]. In this case, soil bulk density might change the relationship between soil nutrients and leaf stoichiometry, because it varies widely with elevation in the study area [62,63], but this needs further exploration. LC : LN increased with SOC : STN, suggesting high SOC : STN can limit growth of P. crassifolia as available soil N is mainly used by soil microbes [64] (L328-334, P11).  

Comments

Also given the very low MAP, there is a case to be made that tree growth at all these sites is primarily water-limited and nutrient limitation plays at most a secondary role.  Again I think the paper would be stronger if the author's addressed this in the conclusions. 

Response

Thank you very much for your valuable comments and kindness. We revised the conclusions section (L375-398, P12-13) and addressed that the leaf stoichiometry of P. crassifolia in the study area was more sensitive to MAP and MAT than soil properties (L386-388, P13).

Comments

With respect to the nutrient ratios presented please state clearly if the ratios presented are mass based or mole based ratios probably in the methods section.  Also with respect to 14 and 16 being critical ratios and useful in determining whether N or P is limiting to productivity, this idea was based on an excellent study of managed wetland herbaceous ecosystems in Europe (Korselman and Mueleman) and the similarity of the values they found to the Redfield ratio.  I have great respect for that set of studies but feel it is inappropriate to apply those values widely to all terrestrial ecosystems as is unfortunately often done. I would strongly recommend at least rewriting the section of the discussion on N versus P limitation based on the leaf N:P to be much more speculative and follow up with this would need to be confirmed with fertilization studies.  Several studies have shown that high P concentrations are essential for fast growth but in these high elevation systems I suspect rapid growth to compete for a spot in the canopy and access to light is not essential for survival.

Response

Thank you very much for your valuable suggestions. The nutrient ratios in this manuscript are mass based, and this has been clearly stated in the method Section (2.3) (L125, P4).

We fully agree with your opinion that it is inappropriate to apply those values widely to all terrestrial ecosystems. Besides, given that nutrient limitations based on the leaf N : P are to be much more speculative, we believe that it is more reasonable from the actual soil stoichiometry. So we rewrote the nutrient limitation of P. crassifolia growth in the revised version in Section 4.4 (L348-372, P12) as below: 

Previous studies have shown that the LN : LP might be an important indicator of nutrient limitation [33]. Different researchers have judged these elements limiting plant growth on the basis of different LN : LP thresholds. For example, a study by Koerselman et al. [70] on plant communities in wetland systems showed that plant growth was limited by N when LN : LP was less than 14, by P when it was greater than 16, and by both N and P when it was between 14 and 16. However, Gusewell [71] found that plants were limited by N when the LN : LP was less than 10, by P when it was greater than 20, and that fertilization had no significant effect on the LN : LP when it was between 10 and 20. Therefore, nutrient limitations based on LN : LP are more speculative, and actual soil stoichiometry may be more useful [72]. Previous studies found that an SOC : STN below 25 indicated sufficient N for plant growth [64]. Since the SOC : STN in the study area was 14.06, it was inferred that the growth of P. crassifolia may not be limited by N. Coupled with global warming and N deposition, the rate of net N mineralization would be accelerated [72], and N available for P. crassifolia growth will further increase. In the study area, SOC : STP was 106.27, significantly higher than 99.0 in severely P deficient soils [66]. In addition, STP (0.56 g·kg-1) was lower than the global average (2.8 g·kg-1) [73]. Therefore, we speculated that P. crassifolia growth may be susceptible to P limitation in the present study area.

Several studies have shown that high P concentrations are essential for fast growth to compete for a spot in the canopy and access to light (e.g., [66]). However, in high elevation systems, plants take up P not for rapid growth, but for resistance to harsh conditions such as low temperatures and drought [1,34,39]. Given low P concentration in the topsoil, P. crassifolia may have to absorb P in deep soil through roots [74,75], or by resorption from the litter to ensure its growth requirements [14].

Author Response File: Author Response.doc

Round 2

Reviewer 1 Report

I appreciate the changes made by the authors to this manuscript.  I believe that this manuscript needs minor revisions.

All of the references are showing as an error.

Figure 1. Consider having the elevation legends in the same direction.  They are opposite in this figure.

Figure 3: Why is elevation not included in the PCA?

Table 4. What kind of relationship are you presenting?  R?

Sections 4.3 and 4.3.1 could be combined.

Author Response

Thank you very much for your positive feedback on our work before. These comments incredibly improved the quality of our manuscript ‘Elevation alone alters leaf N and leaf C to N ratio of Picea crassifolia’ (forests-1363822). We have carefully addressed all the comments. In this letter, we provided a point-by-point explanation of the revisions and noted them in red. We provide the revised version with track and the number of pages and lines in the present response letter are all based on it. We hope that our revisions now could meet your requirements.

Sincerely,

Niu and co-authors.

 

Comments

I appreciate the changes made by the authors to this manuscript. I believe that this manuscript needs minor revisions.

All of the references are showing as an error.

Response

Thanks for your suggestion. We have checked and corrected the references.

Comments

Figure 1. Consider having the elevation legends in the same direction. They are opposite in this figure.

Response

Thank you very much for your suggestions and kindness. We adjusted the direction of the elevation legends in the map as below (L113-114, P3):

 

Fig 1. The five sample sites at 2400 m, 2600 m, 2800 m, 3000 m and 3200 m in the study area.

Comments

Figure 3: Why is elevation not included in the PCA?

Thanks for your comments. The Figure 3 is a Redundancy analysis (RDA) for the leaf stoichiometric indices of P. crassifolia and environmental indices, including soil nutrients and their ratios, for elevation. As described in manuscript (L199-200, P8), the ‘elevation’, an important topographic factor, is a comprehensive indicator, which could lead to the variation of temperature and precipitation, plant composition, geological substrates, light resources, and even the disturbance regime [1,3,14]. In this study, elevation affected P.crassifolia leaf stoichiometry mainly by MAP, MAT, soil pH and SOC : STN. The response of P.crassifolia leaf stoichiometry to elevation, is actually to these four environmental factors. Therefore, the ‘elevation’ needs to be excluded from the RDA. If ‘elevation’ is used as a factor for RDA, there will be a significant self-correlation between elevation and factors such as temperature, precipitation, and soil properties. Besides, in terms of this study, repeated analyses will make it difficult to correctly reflect the response of P.crassifolia leaf stoichiometry to the change of elevation (comprehensive environmental factors).

Comments

Table 4. What kind of relationship are you presenting? R?

Response

Thanks for your comments. Firstly, we have adjusted the title of Table 4 to ‘The Pearson correlations between the dominant factors from the RDA and leaf stoichiometry of P. crassifolia’ (L193, P8). Then the values in Table 4 are correlation coefficients, ranging from -1 to 1. It reflects the linear relationship between the independent and dependent variables. The closer to 0, the smaller the correlation, the closer to -1 or 1, the larger the correlation. The ‘＋’ and ‘－’ indicate the direction of correlation, with the former being positive and the latter being negative.  

Comments

Sections 4.3 and 4.3.1 could be combined.

Response

Thank you very much for your valuable suggestions. We have integrated 4.3 and 4.3.1 as following (L243-275, P9-10):

4.3 Effects of MAP and MAT on leaf stoichiometry of P. crassifolia 

From the RDA results, about 40% of the total variation in leaf stoichiometry of P. crassifolia could be explained by the measured parameters (Fig 2), Of these environmental factors, MAP and MAT were the main explanatory variables, accounting for 30% of the total variation (Table 3), thereby supporting our hypothesis.

In the present study, MAP was negatively correlated with LN, positively correlated with LC : LN, but not correlated with other indicators (Table 4). The LN of P. crassifolia decreased with an increase in MAP, as reported in other studies [20]. This may be because increased precipitation alleviates drought stress and protects the plant’s photosynthetic apparatus [44]. Accordingly, as the photosynthetic metabolism rises, N, which is the main component of the enzyme, is consumed in greater quantities [44]. However, other studies found a positive correlation [39] or no correlation [44] between LN and MAP. The reason for this disparity may be related to different plant responses to water avaliabilities and different adaptation strategies for survival [6,7]. With a constant LC and a negative relationship between MAP and LN, the LC : LN showed a significantly positive relationship with MAP, suggesting that the greater the precipitation, the slower the growth rate of P. crassifolia [40]. P. crassfolia growth was sensitive to high MAP; plants may consume greater amounts of N under high MAP than under lower MAP as mentioned above.

In contrast to MAP, MAT was positively correlated with LN and negatively correlated with LC : LN, and had no correlation with other indicators (Table 4). This was mainly related to the limitation of low temperatures on the rate of nutrient turnover [35]. Generally, low temperatures can decrease soil microbial activity and in turn, reduce the decomposition rate of organic materials and the release of soil nutrients that could subsequently be absorbed by plants [46]. However, some studies found MAT was negatively correlated with LN [37,38], because plants growing at low temperatures require more nutrients to maintain normal ecophysiological processes [35,47]. In addition, MAT may also affect LN by regulating SOC : STN; that is, the decreased temperature inhibits the microbial activity and reduces the rate of N mineralization, resulting in a high SOC : STN [46,48] (details of SOC : STN influencing LN in Section 4.3.2). The LC : LN was significantly and negatively related to MAT, indicating that P. crassifolia grew faster at low elevations than at high elevations due to the moderate temperature and soil nutrients such as STP at 2600 m (Table 1).

 

Author Response File: Author Response.docx

Reviewer 2 Report

The authors have done an excellent job of addressing my concerns as a reviewer.  The careful revisions they have made greatly improve the manuscript.

Author Response

Dear Reviewer:

Thank you very much for your positive feedback on our work before. These comments incredibly improved the quality of our manuscript ‘Elevation alone alters leaf N and leaf C to N ratio of Picea crassifolia’ (forests-1363822).

Sincerely,

Niu and co-authors.

 

Comments

The authors have done an excellent job of addressing my concerns as a reviewer. The careful revisions they have made greatly improve the manuscript.

Response

Thank you very much for your positive feedback on our manuscript.

Author Response File: Author Response.doc

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

This paper reports of soil and foliar chemistry of Pinus crassifolia sampled at 5 sites in China that differ in elevation. The authors conclude that “Except for [N]leaf and [C]leaf : [N]leaf, P. crassifolia’s other leaf stoichiometries remained relatively stable across elevations, thereby partly supporting the homeostasis hypothesis. This suggests that climate change may have little effect on P. crassifolia’s leaf carbon assimilation capacity; however, it appears to have significant effects on its growth and biomass-stored carbon”. The paper is generally well written and appropriate for the journal. However, I do not feel that the experimental design [5 sites] and the results [only a significant difference in foliar N] support the broad conclusions stated above. The data simply show that at the five sites only foliar N exhibited a significant difference [ratios are a consequence of this]. While the 3 highest sites have the lowest foliar N the two highest sites have a significantly higher foliar N than the middle site – so is elevation really the driver? There is also insufficient information presented on the soil data or the statistical analysis for me to fully evaluate the paper. I do not understand how and why parameters that exhibited no significant difference among sites were evaluated using RDA – there are more variables than sites [only 5 sites]. The discussion is far too long and makes broad generalizations that are not supported by the data.

Main comments.

1. The authors report foliar C [e.g. lines 18 and 19] and infer that higher C means greater C assimilation when really doesn’t. Carbon assimilation is driven by biomass production; C concentrations are a result of various differences in nutrient concentrations.
2. The last sentence of the abstract refers to parameters not assessed in this paper. This paper simply has foliar chemistry at five sites.
3. The presentation of results is confusing in that the authors report maxima and minima concentrations [e.g. abstract line 20], but there is no significant difference among the five sites in foliar P?
4. Annual temperature and precipitation at the sites are modelled and do not take into account local factors such as aspect that may differ from the modelled relationship. It also provides no information on the climate during the sampling year – which is most relevant for foliar chemistry.
5. I have many methodological questions: 1) line 103 states that 3 soil depths were taken so which horizon is shown in Table 1?; 2) line 105, leaves were ground but were they washed before this as it will affect foliar chemistry; 3) how big were the trees? 4) What was the associated forest composition?; 5) at what height and aspect was foliage taken? [I am not going to make any editorial comments except to note on line 114 – derminted should be determined]; 5) why were other macronutrients [Ca, Mg, K, S, etc] not measured as these are needed for full interpretation of leaf chemistry?. All these factors will influence the foliar chemistry and foliar stoichiometry and should be documented in a study such as this.
6. Similarly, the data analysis needs further clarification. Based on the sampling there are 3 plots and 5 elevations [n=3 per elevation?], and each plot has 9 soil samples, that I assume were composited somehow? The authors state a T-test was used, which does not seem to be an appropriate test. It seems an ANOVA would be more appropriate with 3 replicates? Did the authors check for normality? It appears that a linear correlation was performed, but I am not sure how this is possible when there are only 5 “treatments” and I do not know how the RDA was performed. I expect the authors are treating the 3 plots at each elevation as independent samples?
7. I think the authors overstate the results and the role that elevation is playing. Elevation is just one factor that differs among sites. Only foliar N [and C:N] is actually significantly different among sites. I agree that the two lowest sites have higher N, but foliar N increases from 2800 to 3200m [significantly so] – so how is elevation actually affecting foliar chemistry?
8. Line 143-152 – authors discuss “differences” in foliar C and P – but there are no significant differences.
9. Table 2 – I am not sure what test this is showing [correlation?], but if one looks at the data the problem with this analysis is quite clear. For example, the authors show in Table 2 that SOC and DTN differs significantly with elevation, but these are driven by the data at 3200m [Table 1]. There is no difference among the other 4 elevations. Are the authors suggesting a “tipping point”. I’m not saying that differences with elevation do not occur, I am saying that these data are not clear evidence of this. I would have liked to have seen more sites at different elevations.
10. Figure 1 – a similar issue is evident here. There is quite clearly a large drop in foliar N between 2600 and 2800m, but after that foliar N increases – so what mechanism is driving this? An abrupt change within a narrow window followed by an increase [not discussed]. A simpler explanation is that it is just site-to-site variability and other factors are at play [see comments on methods].
11. Table 3 – to be honest I do not know how the analysis in Table 3 were done or why. This data set does not seem appropriate for this type of analysis and given the fact that only foliar N was different among sites I do not know how you can explain differences in P for example that did not differ? It makes no sense to me. This leads to table 4 in which the authors report soil pH [I do not know which horizon] is negatively related to foliar C – but foliar C does not differ significantly among sites. Further, what possible mechanism is at play here?
12. Discussion – because of the large number of issues with the methods and results – the discussion outlines differences that do not really exist and in some cases do not adequately explain the data that is shown. For example, in line 234 the authors report that precipitation is negatively correlated with foliar N. This really is an artifact driven by the fact that modelled precipitation increases in elevation and the 3 higher sites have lower foliar N. However, differences in foliar N are clearly not linear [there is a step change at 2600m] followed by an increase after 2800m, when “modelled” MAP increases in a linear fashion. Precipitation cannot explain the step decrease and increase thereafter even though a linear correlation would suggest a relationship. Likewise, section 4.2.3 that discusses in influence of soil pH on foliar chemistry is very dubious given that we are talking about very small differences in pH [>0.1 pH units difference between 4 of the 5 sites, which is sampling error] and there was no difference in foliar C. The whole discussion is similarly flawed.
13. Finally I disagree with the conclusion that this is detailed study and the authors continue to talk about differences [e.g. P] that do not exist among sites. The authors do not appropriately evaluate or interpret the data presented in this study resulting is many misleading statements.

Reviewer 2 Report

 This manuscript describes the stoichiometry of P. crassifolia at five elevations in China.  The use of [C]leaf, [P]leaf, [C]leaf : [P]leaf, [N]leaf : [P]leaf, etc. is very difficult to read throughout the manuscript.  Since you mostly discuss leaf nutrients, I think you can leave leaf out of these.  However, I think you are doing this manuscript a disservice if you do not discuss soil nutrients more.  I think discussing how elevation, soils, and tree nutrition work together in this region would be a huge improvement to this manuscript.  I think that this manuscript should be reconsidered after major revisions.

Lines 92-96:  Can you state the elevation ranges for these unique vegetation zones and soils?

Table 1: Please include a map of the sample areas for readers with no knowledge of this area.

Lines 140-152: These two paragraphs should be combined.  Or have one paragraph to describe soil and one for foliage.

Line 170:  Where is Figure 2? I think that this figure would be valuable to explain the factors affecting leaf stoichiometry.

Lines 200-202: Weren’t temperature and precipitation estimated using elevation?

Lines 202-203: I don’t think you would expect a large change in leaf carbon across sites.  Maybe you would see a difference in carbon pools, but not in concentration.

Lines 204-205: Again, you would need leaf C pools to talk about carbon sequestration.

Lines 212-213: This sentence is confusing.  Do you mean that lower temperatures lowered leaf N concentrations?  Your highest elevation had greater leaf N than the two lower elevations.

Lines 224-227: Do you have any data on the tree productivity at these sites? Also, this sentence is written such that you found that elevation… You need to reword it so that we know that other studies have found…

Lines 253-262 and 286-291: I think the biggest effect of temperature will be on forest floor and soil C:N ratio, which then affects foliar N concentration.  This relationship was covered for Douglas-fir in Littke et al., 2014 Forest Science.

Lines 312-315:  Maybe I am reading this wrong, but it looks like you could indicate deficiency of N and/or P at any N:P ratio.