Effect of Tree Size Heterogeneity on the Overall Growth Trend of Trees in Coniferous Forests of the Tibetan Plateau

Round 1

Reviewer 1 Report (New Reviewer)

This manuscript “Effect of Tree Size Heterogeneity on the Overall 2 Growth Trend of Trees in Coniferous Forests of 3 the Tibetan Plateau” examines applicability of iterative growth model (IGM) to community scale tree growth trend and effects of tree size heterogeneity under different climatic conditions and elevation regions. The study found a unimodal pattern constraining radial growth of trees at community scale and suggested an indicator for overall variation in tree radial growth. Below are my minor comments and suggestions.  

            The main concern of this paper is about the presentation of resulting figures. The authors introduced new results in the discussion section using one main figure (Fig. 5) and several supporting figures in Appendix B. I don’t see a clear reason for them not to be addressed in the results section.

For readers who may not be familiar with tree growth theory or trend, I suggest elaborating “metabolic growth theory” in more detail in the introduction. A supporting figure indicating growth trajectories over time in different hypothetical scenarios would also be helpful to understand the main findings of this paper.  

 

Minor comments:

L16: tree heterogeneity in tree size à the heterogeneity in tree size

L66: elaborate “metabolic growth theory”

L153: “See supplementary information” à “See Appendix A”

L236-237: The sentence seemed to be not completed

L237: Fig. S1 à Fig. A1

L249: It is not clear what this sentence means: The curve in (B) is consistent with (C). Does it mean that the curve in (B) is the same as the curve in (C). Or, they were fitted with the same method?

Figure 4: Where is the height separating the high and low elevation regions (3600 m) from? Is this a tree line? It needs to be mentioned either in the method section or result section (3.3).

L340-357: Fig. S2, S3, S4 à Fig. A2, A3, A4

L387: history best à historical best

L423: Eq. 1 or Eq. S1?

Author Response

Response to reviewer 1 comments

 

Dear Editors

Thank you and the reviewers very much for taking your time to read/edit our manuscript carefully, and we really appreciate your helpful comments/suggestions. According to the comments of the reviewer, here we have made a major revision to the previous manuscript. Bellow you will find our responses to comments made. To make clearly and exactly, our responses directly followed the comments made by the reviewers. A revised version of the manuscript is also included at the end.

Sincerely yours,

Shumiao Shu

 

Reviewer: 1

This manuscript “Effect of Tree Size Heterogeneity on the Overall Growth Trend of Trees in Coniferous Forests of the Tibetan Plateau” examines applicability of iterative growth model (IGM) to community scale tree growth trend and effects of tree size heterogeneity under different climatic conditions and elevation regions. The study found a unimodal pattern constraining radial growth of trees at community scale and suggested an indicator for overall variation in tree radial growth. Below are my minor comments and suggestions.

The main concern of this paper is about the presentation of resulting figures. The authors introduced new results in the discussion section using one main figure (Fig. 5) and several supporting figures in Appendix B. I don’t see a clear reason for them not to be addressed in the results section.

For readers who may not be familiar with tree growth theory or trend, I suggest elaborating “metabolic growth theory” in more detail in the introduction. A supporting figure indicating growth trajectories over time in different hypothetical scenarios would also be helpful to understand the main findings of this paper.

 

Response: We thank the reviewer for his/her suggestions regarding the presentation of the results and the incompleteness of the related concepts. These recommendations are fully accepted.

 

In the Introduction section, we have added a brief description of metabolic growth theory:

Metabolic growth theory, also known as the ontogenetic growth model (OGM)  (Hou et al. 2008; Zuo et al. 2012), provides a promising basis for understanding tree growth. This theory dates back to observations of animal metabolism in the early 19th century (Huxley & Teissier 1936). Its core can be summarized as follows: the growth rate of an organism is proportional to the difference between its total metabolic and maintenance metabolic rates, where the two metabolic terms are proportional to the 0.75 and 1 power of the organism's size, respectively (see Appendix B).

 

It is described in further in Appendix B:

Mathematically, classical metabolic growth theory, also known as the ontogenetic growth model (OGM), is a special case of IGM at b = 0.75 and T → 0.

 

In the Results section, we have moved Fig. 5 into Fig. 3 and added the appropriate descriptions. In addition, we believe that Fig. S1 to S3 would be more appropriately placed in the Supplementary Information to make the main results clearer and the text more concise.

 

Not only that, the coefficient of variation of the radial growth rate for the last 5 years is also lower than the average normalized growth trend (i.e., half of the normalized curve). Note that the reason for using the average curve here is that the coefficient of variation is determined based on the mean value. From these results, it is evident that unimodal patterns limit tree radial growth, following the IGMR-U.

 

In the methods section, we present the differences between the IGM and the classical metabolic growth model (OGM) in explaining tree radial growth

 

 

Fig. 2 Classical metabolic growth theory (OGM) vs. generalized metabolic growth theory (IGM)

f(r) and r represent the actual tree ring growth rate and tree radius, respectively. Here, we show the best growth trajectory of trees within the same forest, so we assume that the actual f(r) (gray dots) is below the curve. It is worth noting that the IGMR-U is more common, indicating that the gray dots tend to appear more often below this curve.

 

Minor comments:

L16: tree heterogeneity in tree size à the heterogeneity in tree size

Response: Thanks, we have revised this error

 

L66: elaborate “metabolic growth theory”

Response: Thanks, we have added the relevant descriptions in the introduction and Supplementary information.

 

L153: “See supplementary information” à “See Appendix A”

Response: Thanks, we have revised this error.

 

L236-237: The sentence seemed to be not completed

Response: Thanks, we have revised this error.

In addition, we used the Wilcoxon Signed-Rank test to analyze the differences between the best historical growth rates and the current growth rates of trees at different elevation ranges. The Wilcoxon test is the non-parametric version of the paired t-test, which is appropriate for any distribution of data, and especially for small sample sizes.

 

L237: Fig. S1 à Fig. A1

Response: Thanks, we have corrected this typo.

 

L249: It is not clear what this sentence means: The curve in (B) is consistent with (C). Does it mean that the curve in (B) is the same as the curve in (C). Or, they were fitted with the same method?

Response: Thanks, we have revised this inaccurate description.

The green curve in (B) is obtained from the 0.95 quantile fit of the data by Eq. 2, where b = k = 0.736 ± 0.10. This curve, along with its half, is also plotted in (C) and (D). B, C and D: Dot density is represented by different colors, with warmer colors representing higher density and cooler colors representing lower density.

 

Figure 4: Where is the height separating the high and low elevation regions (3600 m) from? Is this a tree line? It needs to be mentioned either in the method section or result section (3.3).

Response: Thanks, we have added relevant notes and references. Conversely, the current growth rate of trees situated above the treeline (uppermost elevation of an individual tree, > 2 m height, typically > 3600 m elevation for TB) (Wang et al. 2022)

 

 

L340-357: Fig. S2, S3, S4 à Fig. A2, A3, A4

Response: Thanks we have corrected this typo.

 

L387: history best à historical best

Response: Thanks, we have corrected this typo.

 

L423: Eq. 1 or Eq. S1?

Response: Thanks, we have corrected this typo.

 

Author Response File: Author Response.docx

Reviewer 2 Report (New Reviewer)

General comments:

1.      The major contribution of the manuscript is to extend the iterative growth model to the tree-ring scale. Authors focused on the effect of tree size heterogeneity in communities under different climatic conditions on this variability across elevations.

2.      The topic is very interesting and indeed fills the gaps in forestry research on tree rings patterns.

3.      It should be noted that the authors use a sufficiently detailed database.

4.      However, as a research paper needs to be improved. Below is the list of some issues which need to be fixed.

 

Specific comments:

Line 160: ... The value of T is ... It's not T.

Line 180: ... based on the maximum radius and age of trees ... Formula (2) does not show a direct dependence on T, it should be additionally defined by TGT.

Line 212: ... a unimodal growth trajectory test for trees ... What is this mystified test?

Line 236: ...  We found that the estimated r/R mostly ranged from 0 to 1. ... Could it be otherwise?

... Not only that r/R was normally distributed under CVR < 0.375... Bad sentence.

Lines 244 and 246: Probably, relative radius.

Line 248: The curve shown in green corresponds to the single value of the constants, so you cannot specify b and k from a certain interval.

Lines 253 and 254: You talk about the correlation coefficient, but you write R².

Line 393: Throughout the text, it is appropriate to change f(r) to a single letter.

 

The article is written in acceptable language.

Author Response

Response to reviewer 2 comments

 

Dear Editors

Thank you and the reviewers very much for taking your time to read/edit our manuscript carefully, and we really appreciate your helpful comments/suggestions. According to the comments of the reviewer, here we have made a major revision to the previous manuscript. Bellow you will find our responses to comments made. To make clearly and exactly, our responses directly followed the comments made by the reviewers. A revised version of the manuscript is also included at the end.

Sincerely yours,

Shumiao Shu

 

Reviewer: 2

General comments:

1. The major contribution of the manuscript is to extend the iterative growth model to the tree-ring scale. Authors focused on the effect of tree size heterogeneity in communities under different climatic conditions on this variability across elevations.
2. The topic is very interesting and indeed fills the gaps in forestry research on tree rings patterns.
3. It should be noted that the authors use a sufficiently detailed database.
4. However, as a research paper needs to be improved. Below is the list of some issues which need to be fixed.

 

Response: We thank the reviewer for his/her suggestions regarding the presentation of the results and the incompleteness of the related concepts. We fully accept these suggestions and clarify two of the reviewers' concerns.

 

 

 

 

 

 

 

 

Specific comments:

Line 160: ... The value of T is ... It's not T.

Response: Thanks, we have corrected this typo.

 

Line 180: ... based on the maximum radius and age of trees ... Formula (2) does not show a direct dependence on T, it should be additionally defined by TGT.

Response: Here we need to clarify.

Eq.2 is not independent of T. In fact, k is determined by both b and T. We use a mathematical trick here. We first derive two boundary equations (Eqs. 1b and 1c) for T as it converges to its upper and lower limits. Then, based on these results, we further generalize Eq. 2.

 

Line 212: ... a unimodal growth trajectory test for trees ... What is this mystified test?

Response: Thanks, we have revised this inaccurate description.

Afterwards, we calculated the coefficient of variation of tree radius (CVR) at each site and tested whether tree radial growth trajectories conformed to Eq. 2 (where k = 0.75). Based on test results, we estimated the HBGT and CVR of each site using Eqs. 2 and 3.

 

Line 236: ...  We found that the estimated r/R mostly ranged from 0 to 1. ... Could it be otherwise?

Response: Thanks, we've removed the inappropriate description.

 

... Not only that r/R was normally distributed under CVR < 0.375... Bad sentence.

Response: Thanks, we've corrected the inappropriate description.

 On the other hand, the normalized Eq. 1-c still constrains these growth rates when the CVR is larger (CVR > 0.375, white bars in Fig. 3A).

 

 

Lines 244 and 246: Probably, relative radius.

Response: Thanks, we've corrected the inappropriate description.

 

Line 248: The curve shown in green corresponds to the single value of the constants, so you cannot specify b and k from a certain interval.

Response: Here we need to clarify.

The IGM takes into account the thermodynamic significance of respiration and gives a range of values for T.

It is this range that leads to Eqs. 1-b and 1-c (assuming b = 0.75).

With this step, we summarize Eq. 2 (This is a mathematical trick)

Interestingly, Eq. 2 also holds when b takes on other values (general formula)

Using Eq.2 to fit the data, we find that k = 0.736.

Mathematically, Eq. 2 and Eq. 1-c (where b = 0.75) are equivalent when k = 0.75.

Therefore, this again supports our assumption that b = 0.75.

 

Lines 253 and 254: You talk about the correlation coefficient, but you write R2.

Response: Thanks, we've corrected this

 

Line 393: Throughout the text, it is appropriate to change f(r) to a single letter.

Response: Thanks, We seriously considered this suggestion, but since this representation is rooted in previously published results, and since some of our current papers use the same notation, we thought it would be better to use function notation.

 

Round 2

Reviewer 2 Report (New Reviewer)

The authors answered all the questions that arose. Writing is acceptable. Paper can be accepted.  

 

Writing is acceptable.

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

 

The research topic is exciting where the authors determined whether tree radial growth in natural subalpine forests follows a unimodal growth pattern and its variability due to tree size heterogeneity using a robust method. The results are therefore beneficial for understanding the changes in the extent of subalpine forests. Before accepting the manuscript for publication, it would require some editing and additions to certain sections, especially introduction and discussion.

 

Major points

Motivation of extension of the IGM to the tree-ring scale (IGMR) is not clear.

In the analysis, the authors emphasize much on tree size but need to discuss more on age of the trees that influence the growth rates.

Radial growth also influenced by habitat factors. In discussion, the authors should emphasize how forest structure might influence growth.

 

Minor points…

L 17: Largest and longest-lived is not clear.

L23: growing season temperature? Is it mean temperature or min and max?

L39-41: this portion needs to rewrite.

L42: add ‘tree’ after on

L44-47: this line needs to rewrite to

L53: add ‘factors’ rather influences

L130: ‘We conducted statistics on every tree core in all chronologies.’  what are those?

L187 & 191: use ‘could’ instead of can

L188: use , after' forest age

L189: use were instead of are

L200: use R² instead of R2.

Though previous results were written in past tense, I suggest to write this paragraph also in past tense. 3.3. Tree radial growth assessment

L222: It’s need to clarify high and low frequency signals?

L258-260: need to rewrite because the meaning is not clear

L261-62: size structure is related to stand dynamics rather adaptation

L287: cross-elevating is not clear and suggest to rewriting the sentence

In some sections English need to be improved, especially in writing results. see the minor points. 

Author Response

Response to reviewer 1 comments

Dear Editors

Thank you and the reviewers very much for taking your time to read/edit our manuscript carefully, and we really appreciate your helpful comments/suggestions. According to the comments of the reviewer, here we have made a major revision to the previous manuscript. Bellow you will find our responses to comments made. To make clearly and exactly, our responses directly followed the comments made by the reviewers. Revised version of the manuscript is also included at the end.

Sincerely yours,

Shumiao Shu

 

Reviewer: 1

 

Comments and Suggestions for Authors

 

 

The research topic is exciting where the authors determined whether tree radial growth in natural subalpine forests follows a unimodal growth pattern and its variability due to tree size heterogeneity using a robust method. The results are therefore beneficial for understanding the changes in the extent of subalpine forests. Before accepting the manuscript for publication, it would require some editing and additions to certain sections, especially introduction and discussion.

 

 

 

Major points

 

Motivation of extension of the IGM to the tree-ring scale (IGMR) is not clear.

――Thanks for your very good suggestion. We highlight in the introduction section that the model is expected to explain the radial growth of trees in natural forests.

Tree radial growth is an important indicator of tree growth that captures both climate change and intrinsic growth trends. However, it is challenging to determine the overall radial growth trends followed by trees in natural forests. On the one hand, climate, competition, disturbance, and functional traits regulate the unimodal growth trajectory and fluctuate or change over time (Hérault et al. 2011). On the other hand, trees with different sizes may respond differently to these influences, and shape different growth trajectories (Yao et al. 2023). The iterative growth model (IGM) highlights that tree growth conforms to a single-peaked pattern, where tree (radial) growth rate is mediated by tree size (or radius), the rate of maintenance respiration per unit biomass, and the maximum tree size (or radius) allowed by the environment (Shu et al. 2021). Evidence from plantations with similar tree size and stand age supports the model and suggests that climate can induce changes in the height and length of unimodal radial growth curves (Yao et al. 2023). We hypothesize that tree growth trajectories in natural forests are still constrained by the single-peaked model and can be described by the IGM.

 

 

In the analysis, the authors emphasize much on tree size but need to discuss more on age of the trees that influence the growth rates.

 

――Thank you. We considered this suggestion carefully. In our results, size was the main concern, because age may have a weaker explanation for growth. For example, trees suppressed by the crown of a large tree grow slowly. They have experienced the same time, but the growth trajectory of the small tree is much different from the large tree. Thus, although age and size are positively correlated, size has a more direct effect on resource allocation. Our model is also size dependent.

 

 

Radial growth also influenced by habitat factors. In discussion, the authors should emphasize how forest structure might influence growth.

――Thanks for your very good suggestion. We have rewritten the discussion

 

 

 

 

Minor points…

 

L 17: Largest and longest-lived is not clear.

―Thanks. We have rephrased this

Using the IGMR, we reconstructed the historical best growth trajectory (HBGT) of trees within the same community based on the tree with the largest radius and/or longest age in the community

 

 

L23: growing season temperature? Is it mean temperature or min and max?

―Thanks. We have rephrased this

growing season average temperature

 

L39-41: this portion needs to rewrite.

L44-47: this line needs to rewrite to

L53: add ‘factors’ rather influences

 

―Thanks. We have rewritten this paragraph.

Tree radial growth is an important indicator of tree growth that captures both climate change and intrinsic growth trends. However, it is challenging to determine the overall radial growth trends followed by trees in natural forests. On the one hand, climate, competition, disturbance, and functional traits regulate the unimodal growth trajectory and fluctuate or change over time (Hérault et al. 2011). On the other hand, trees with different sizes may respond differently to these influence factors, and shape different growth trajectories (Yao et al. 2023). The iterative growth model (IGM) highlights that tree growth conforms to a single-peaked pattern, where tree (radial) growth rate is mediated by tree size (or radius), the rate of maintenance respiration per unit biomass, and the maximum tree size (or radius) allowed by the environment (Shu et al. 2021). Evidence from plantations with similar tree size and stand age supports the model and suggests that climate can induce changes in the height and length of unimodal radial growth curves (Yao et al. 2023). We hypothesize that tree growth trajectories in natural forests are still constrained by the single-peaked model and can be described by the IGM.

 

 

L42: add ‘tree’ after on

―Thanks. We have rephrased this.

 

 

L130: ‘We conducted statistics on every tree core in all chronologies.’  what are those?

 

―Thanks. We have rephrased this. We conducted statistics on growth information for each chronology, including the current diameter (rc), age (L), and average growth rate over the past five years (f(r)c) for each tree core.

 

L187 & 191: use ‘could’ instead of can

―Thanks. We have rephrased these.

 

L188: use , after' forest age

L189: use were instead of are

L200: use R2 instead of R2.

―Thanks. We have revised these minor errors

 

 

Though previous results were written in past tense, I suggest to write this paragraph also in past tense. 3.3. Tree radial growth assessment

―Thanks. We have revised this section

 

L222: It’s need to clarify high and low frequency signals?

―Thanks. We have revised this section

Radial growth of trees is usually considered to be influenced by age (or size) and climatic factors, showing an age-dependent low frequency (with a stable trend) and a climate-sensitive high frequency (rapid change) signals (Wilmking et al. 2017; Peltier & Ogle 2020; Yao et al. 2023). After removing the modulations of the age effect trend, this signal typically exhibits stable climate sensitivity, termed stationarity assumptions or uniformity principles (Wilmking et al. 2017; Peltier & Ogle 2020). 

 

L258-260: need to rewrite because the meaning is not clear

L261-262: size structure is related to stand dynamics rather adaptation

―Thanks. We have revised this section. Overall, warm season precipitation contributed to size convergence of trees in subalpine forests (Fig 3C). Concurrently, forest growth or development can also spontaneously reduce tree size heterogeneity. At lower elevations, increased precipitation can promote radial growth by reducing water stress and accelerating xylem activity during the growing season (Vieira et al. 2021). This boost will be more pronounced for smaller trees that grow faster, thus reducing CVR. However, at higher elevations with lower temperatures, increased precipitation may cause trees to experience more snowfall events and physical disturbances, resulting in increased CVR. Thus, significant negative (r = -0.79, p < 0.01) and positive (r = 0.44, p < 0.01) correlations were observed between CVR and the precipitation of coldest quarter at low and high elevations, respectively (see Appendix B, Fig. S1). However, the role of precipitation and forest growth or development in regulating size heterogeneity is relatively limited, suggesting the repeated effects of disturbance and competition on forests. These results suggest that precipitation plays a key role in shaping the location and structure of the treeline (Lloyd & Graumlich 1997; Sigdel et al. 2018).

 

 

 

 

 

L287: cross-elevating is not clear and suggest to rewriting the sentence

―Thanks. We have revised this section.

.....These results imply that global warming affects tree growth variability differently in the high and low elevation ranges of subalpine.

 

Author Response File: Author Response.docx

Reviewer 2 Report

Reviewer Report – Forests

Yuelin Wang et al.

Effect of tree size heterogeneity on the overall growth trend of trees in coniferous forests of the Tibetan plateau

This ms attempts to analyse the factors affecting tree growth in forests of the Tibetan plateau, and separate the abiotic and biotic influences. It is centred on a model that aims to combine thermodynamic principles, a ‘generalized metabolic growth theory’ and tree growth data. I am an ecologist, not a modeller, and found it very hard to follow the rationale behind the development of the models. I cannot comment on whether the model is sound. However, I can say that although the writing is grammatically correct, it is poor at explaining methods and results and the implications of the findings. There is also very little ecological contextualisation, and I the plant physiological aspects are either poorly understood or poorly explained (or both). There is also a plethora of abbreviations, which makes reading the paper difficult and disrupts the flow of ideas as I was constantly having to go back and search for their meaning. At the very least a table of abbreviations is needed. In addition, the figure captions lack the necessary detail to understand what is presented. Some of the figures are split between pages, which makes it difficult to find the correct caption.

Overall, the paper reads as though it was written by modellers with limited understanding of ecology. From what I can deduce from the figures, there are some potentially interesting results here, but I would like to see the underlying models evaluated by someone with appropriate expertise.

L32-34. Plant growth patterns depend on the measure that is used – radial growth rates peak earlier than biomass increments, which typically start very small and increase with plant size until near the end of the lifespan (see Bowman DMJS, Brienen RJW, Gloor E, Phillips OL, Prior LD (2013) Detecting trends in tree growth: not so simple. Trends in Plant Science 18, 11-17.)

L43-47 ‘rate of energy dissipation per unit biomass’ – not sure what is meant by this

L53-54 – I am not convinced by the argument here that tree size heterogeneity I key to influencing the overall radial growth trend of trees in natural forests. It could be put as a motivating hypothesis for this work.

L58-59 – rewrite this sentence ‘Warming-driven range shifts, particularly those of treelines, ….’

L60-63 – unclear sentence

L68 – Why are subalpine forests ‘atypical representatives of mountain ecosystems’ ?

L74-77 – sentence hard to understand . In fact, all these aims should be expressed more clearly and compellingly. Why should forests with heterogeneous sizes (as opposed to more uniform sizes) be different? (there could be reasons, but these are not explained properly).

Study area – are the forests at these sites natural or planted? Are they harvested – and if so, selectively or clear-felled? Are they mostly single or mixed ages? If the former, what disturbance initiates stand replacement? And are they mixed or single species forests?

Fig. 1 – explain the boundaries in the inset (provinces?) It looks like your study was conducted in 3 provinces? But they don’t seem to correspond to the four names in the main part of the figure. Should one of these names be ‘Tibetan Plateau’, and presumably this isn’t a province?

L113 – I am not convinced that this is a ‘historical best growth trajectory’

L119 – there are no Eq 4a or Eq 4b

L133 – replace ‘preciseness’ with ‘robustness’ or ‘reliability’

Fig. 2A and 3A. x-axis label – what is ‘site sacle’

Fig. 2 – you need to briefly explain what Normalised growth rate is. You need to explain the different colours of the data points. And in the final line of the caption you mean C and B (not D and C)

Fig. 3. What effect might elevation have other than on climate? Could it be subsuming some of the precipitation effects? I might have missed something here, but why are you using average/maximum tree age, whereas elsewhere you have looked at average/maximum size (radius).

L207-213 – this paragraph was unclear. The references to Fig 4C and 4D should probably be to 4B and 4C.

Fig. 4A is not properly explained in the caption, which makes it difficult to interpret. What does ‘Number’ refer to – trees? What does each point represent? (figure captions should be self-contained). 4D and 4E – these are not directly comparing the historical best and current radial growth rates, but rather are comparing the difference in the two with the two meteorological variables. (It is very important to be precise.) Units of y axis are not given. And don’t split figures across pages

L222-224 – references for these statements?

The first paragraph of the discussion was quite difficult to follow, and it was sometimes unclear what is already known from the findings of this study (e.g. lines 226-229).

L248 – and whether the forests are single- or multi-aged!

L253 – frost damage is usually reduced by increased precipitation. However, decreased solar radiation associated with increased precipitation may constrain growth

Fig. 5 – need key to colours

 

L268 – ‘Subalpine forests … have a normal distribution of tree sizes … because trees of different sizes experience different ecological pressures’ – I am not convinced by the logic here – nor in the rest of this paragraph.

L275 – I am not convinced by the argument that size inequality should be negatively correlated with overall growth variability – wouldn’t the relationship be positive, all else being equal?

I found the entire discussion confusing - it needs to be more carefully structured and clearly argued.

 

 

The English language itself is quite good, but ideas are poorly explained and the logical arguments are poorly constructed.

Author Response

Response to reviewer 2 comments

Dear Editors

Thank you and the reviewers very much for taking your time to read/edit our manuscript carefully, and we really appreciate your helpful comments/suggestions. According to the comments of the reviewer, here we have made a major revision to the previous manuscript. Bellow you will find our responses to comments made. To make clearly and exactly, our responses directly followed the comments made by the reviewers. A revised version of the manuscript is also included at the end.

Sincerely yours,

Shumiao Shu

 

Reviewer: 2

 

Reviewer Report – Forests

Yuelin Wang et al.

Effect of tree size heterogeneity on the overall growth trend of trees in coniferous forests of the Tibetan plateau

This ms attempts to analyse the factors affecting tree growth in forests of the Tibetan plateau, and separate the abiotic and biotic influences. It is centred on a model that aims to combine thermodynamic principles, a ‘generalized metabolic growth theory’ and tree growth data. I am an ecologist, not a modeller, and found it very hard to follow the rationale behind the development of the models. I cannot comment on whether the model is sound. However, I can say that although the writing is grammatically correct, it is poor at explaining methods and results and the implications of the findings. There is also very little ecological contextualisation, and I the plant physiological aspects are either poorly understood or poorly explained (or both). There is also a plethora of abbreviations, which makes reading the paper difficult and disrupts the flow of ideas as I was constantly having to go back and search for their meaning. At the very least a table of abbreviations is needed. In addition, the figure captions lack the necessary detail to understand what is presented. Some of the figures are split between pages, which makes it difficult to find the correct caption.

Overall, the paper reads as though it was written by modellers with limited understanding of ecology. From what I can deduce from the figures, there are some potentially interesting results here, but I would like to see the underlying models evaluated by someone with appropriate expertise. 

The reviewers' concerns regarding the methodology, results, and implications of the findings exposed some problems in our paper. We greatly appreciate these insights. After careful consideration, we have revised the content. First, we have explained the model used here in detail and included it in the supplementary material. Second, we have responded point-by-point to the reviewers' reports (see below). Finally, we have revised all content that we believe is misunderstood and incorrect.

In response to reviewers' concerns about the model, we give details of the model here.

 

Iterative growth model (IGM)

Tree respiration involves the transport, release, and use of energy stored in photosynthetic carbohydrate products. This supports tree growth, maintenance, and longevity. These energy-demanding processes also follow the first and second laws of thermodynamics and the allometric scaling laws of metabolism. Based on these rules, we constructed a general kinetic framework for organism growth (Shu et al. 2021)

(S.1)

where f(m) is the total biomass of new tissue created during the formation time T of a unit of tissue and hence f(m)/T represents the average growth rate over this formation time, as well as the growth rate; b is the metabolic scaling exponent, related to a space-filling fractal (self-similar)-like network; g_r is the cost of respiration needed to produce a new unit of tissue; and m_r is a unit of tissue's rate of maintenance respiration (per unit of time). Generally, g_r is stable, and m_r is sensitive to the environment and is driven by temperature, with its trend following the Arrhenius equation. Mathematically, Eq. S.1 highlights a growth iterative mechanism. Namely, growth can be described as a series of spontaneously iterated feedbacks - each of length T. At each iteration the organism moves from the initial biomass m₀ (slightly larger than the threshold biomass for growth, o) and approaches the final mass M. Thus, we refer to Eq. 1 as an iterative growth model (IGM).

Moreover, the IGM contains two implicit thermodynamic and mathematical constraints. The first constraint is that T < g_r/m_r. We derive this from the thermodynamic significance of respiration. To counteract natural degradation (entropy increase), organisms must continuously use negative entropy to maintain the complexity, variety, and order of their components. Usually, organisms obtain useful energy (e.g., chemical energy stored in photosynthetic products or food) from the environment and return equivalent amounts of energy to the environment in less useful forms, such as dissipated energy or heat. In this process, energy provides negative entropy or the required order to organisms. Thus, from an entropy perspective, during time T, the growth energy proportional to g_r decreases the entropy of a new unit of tissue relative to that of their free precursor monomers (Clarke 2019), causing free monomers to achieve an appropriate ordered state. At the same time, the maintenance energy proportional to Tm_r contributes the negative entropy to maintain the low entropy state of a unit of old tissue, and m_r and Tm_r are also proportional to the entropy increase rate and entropy accumulation of a unit of old tissue during time T, respectively. Assuming there is no difference between the new and old units of tissue, the new-unit tissue can be synthesized only when Tm_r must be less than g_r, i.e., T < g_r/m_r. When T → 0 and g_r/m_r, integrating or iterating Eq. 1 will produce two smooth functions driven by time (t), i.e., the Richards and Gompertz equations (Shu et al. 2021).

(S.2-a)

(S.2-b)

where L = 1-M^b-1×m₀^1-b, r = m_r/g_r(1-b), m₀ is the first biomass observed, and n is the number of iterations and is equal to t•m_r/g_r. These results indicate that actual growth dynamics lie somewhere between these equations (Eqs. S.2-a and S.2-b) and may not be an explicit analytic solution in most cases.

Second, from a mathematical perspective, M maintains a strict relationship with other parameters.

(S.3)

where D is the average f(m), mainly determined by the ability of plants to absorb resources and the supply of resources, and TM/D or g_r/m_r×(2b+2)/(1-b) represents the total growth time. The basis for this equation is an integral transform from f(m) to M (9).

 

IGM-based tree radial growth model

Based on the allometric scaling relationship between tree radius (r) and m, i.e., r ∝ m^2/b(West et al. 1999), the IGM can be extended to the tree ring scale (IGMR). Assuming that f(r) is the tree-ring growth rate, the relationships between f(r), r, and potential maximum radius (R) can be expressed as follows:

(S.4-a)

The value of T is related to the thermodynamic significance of respiration and ranges between 0 and g_r/m_r (Shu et al. 2021). The limits T → 0 and g_r/m_r provide us with upper and lower boundaries for f(r). Under b = 0.75 (Mori et al. 2010), these boundaries are:

  (S.4-b)

(S.4-c)

Theoretically, the more rapidly a unit of tissue grows (T → 0), the closer f(r) approaches Eq. 1-b. Conversely, as T → g_r/m_r, f(r) approaches Eq. 1-c (Shu et al. 2021). We referred to Eqs. 1-b and 1-c as the thermodynamic lower (IGMR-L) and upper (IGMR-U) limits of the IGMR. In addition, according to the IGM, the total growth time (TGT) of an organism is g_r/m_r × (2b + 2)/(1 - b). This agrees with the TGT of Eq. S.4-b. However, the TGT of Eq. S.4-c is 32/3 × g_r/m_r. Thus, we hypothesized that the pattern of radial growth in trees shifts from Eq. S.4-b to Eq. S.4-c over time to ensure that radial growth keeps pace with biomass growth. To further simplify the radial growth model, we can combine Eqs. 1-b and 1-c, resulting in a more concise expression:

(S.5)

where k is a parameter associated with T and b. Specially, when b = 0. 75, 1/3 < k < 3/4.

 

References

1.

Clarke, A. (2019). Energy Flow in Growth and Production. Trends in Ecology & Evolution, 34, 502-509.

2.

Mori, S., Yamaji, K., Ishida, A., Prokushkin, S.G., Masyagina, O.V., Hagihara, A. et al. (2010). Mixed-power scaling of whole-plant respiration from seedlings to giant trees. Proceedings of the National Academy of Sciences of the United States of America, 107, 1447-1451.

3.

Shu, S.-m., Zhu, W.-z., Kontsevich, G., Zhao, Y.-y., Wang, W.-z., Zhao, X.-x. et al. (2021). A discrete model of ontogenetic growth. Ecological Modelling, 460, 109752.

4.

West, G.B., Brown, J.H. & Enquist, B.J. (1999). A general model for the structure and allometry of plant vascular systems. Nature, 400, 664-667.

 

 

L32-34. Plant growth patterns depend on the measure that is used – radial growth rates peak earlier than biomass increments, which typically start very small and increase with plant size until near the end of the lifespan (see Bowman DMJS, Brienen RJW, Gloor E, Phillips OL, Prior LD (2013) Detecting trends in tree growth: not so simple. Trends in Plant Science 18, 11-17.)

――Thanks. We have carefully considered the reviewers' comments, but we do not believe that the comments here conflict with our research. In fact, we only emphasized the single-peaked growth pattern of the trees, whether radial growth or biomass growth. Here, we plotted theoretically predicted tree radial and biomass growth trends, which are consistent with the findings of the references listed by the reviewer. Note that the biomass data used here are a set of stochastic simulations, and the radial growth data are expanded from r ∝ m^(3/8)

 

 

 

L43-47 ‘rate of energy dissipation per unit biomass’ – not sure what is meant by this

――Thanks.We have replaced this unclear description with “the rate of maintenance respiration per unit biomass”. L63-64

 

L53-54 – I am not convinced by the argument here that tree size heterogeneity I key to influencing the overall radial growth trend of trees in natural forests. It could be put as a motivating hypothesis for this work.

 

――Thank you for your very good suggestion. We have added a paragraph: Compared to planted forests, natural forests tend to have greater tree size heterogeneity. This means that the overall radial growth trend of trees in natural forests cannot be considered as a simple scale-up of individual growth. In multi-aged stands, the causes of size heterogeneity are related to repeated disturbances or silvicultural interventions that regenerate these new age classes (Nepstad et al. 2007; O'Hara & Ramage 2013; Shu et al. 2019). In single-aged stands, competition directly drives tree size differentiation (Forrester 2019). Obviously, to a large extent, tree size heterogeneity may mediate the effects of competition and disturbance on tree growth. Under different climatic conditions, how tree size heterogeneity influences the overall radial growth trend of trees is still unknown. L70-79

 

L58-59 – rewrite this sentence ‘Warming-driven range shifts, particularly those of treelines, ….’

――Thanks. Range shifts caused by warming have already been observed in some biomes, particularly those subalpine forests close to the treeline (Pepin et al. 2015). L82 -83.

 

 

L60-63 – unclear sentence

――Thanks. We have reworded. Tree radial growth consists of age-dependent low-frequency and climate-sensitive high-frequency signals. Although the high-frequency signal is a result of the rapid response of tree radial growth to climate, the age effects and sampling strategies still affect the accuracy of tree growth assessments and climate responses (Wang et al. 2017). L84-L88

 

L68 – Why are subalpine forests ‘atypical representatives of mountain ecosystems’ ?

――Thanks. We have modified the inappropriate description. The subalpine ecosystem occupies elevations just below tree-line between 2,700 and 3,500 m. They are not only widely distributed but also sensitive to global climate change (Shu et al. 2019; Shi et al. 2021). L94-96

 

 

L74-77 – sentence hard to understand . In fact, all these aims should be expressed more clearly and compellingly. Why should forests with heterogeneous sizes (as opposed to more uniform sizes) be different? (there could be reasons, but these are not explained properly).

――Thanks for your very good suggestion.We have modified this and added a description (L100-103). The aim of this study was to determine the effect of tree size heterogeneity on the overall radial growth trend of trees in natural forests. The experiment consisted of three steps. First, we extended the IGM to the tree-ring scale (IGMR) and evaluated its constraints on community-scale tree growth trends.

 

Study area – are the forests at these sites natural or planted? Are they harvested – and if so, selectively or clear-felled? Are they mostly single or mixed ages? If the former, what disturbance initiates stand replacement? And are they mixed or single species forests?

――Thanks, we have added the relevant description in Tab.1. There is some information that we did not find, but we believe that this has a very limited impact on our results. This is because the data collection generally follows a consistent specification.

 

Fig. 1 – explain the boundaries in the inset (provinces?) It looks like your study was conducted in 3 provinces? But they don’t seem to correspond to the four names in the main part of the figure. Should one of these names be ‘Tibetan Plateau’, and presumably this isn’t a province?

――Thanks, we have modified Fig. 1. L 122.

 

 

L113 – I am not convinced that this is a ‘historical best growth trajectory’

――Thanks. We need to clarify this point. Both our model and our data show that the radial growth of trees is constrained by a single-peak pattern (Eq.2 and Figs. 2B and 2C). Thus, we can then infer this constrained trajectory based on the maximum radius and lifespan, i.e., historical best growth trajectory. Note that in Fig. 2B, the x-axis and y-axis are r/R and f(r)_c/f(r)_m, respectively, where f(r)_m = R/maximum TGT. Thus this best trajectory can be determined with the knowledge that R, maximum TGT and k (0.736).

r―tree radius

R―tree maximum radius

f(r)_c―average growth rate over the past five years

f(r)_m―average tree ring width

TGT―total growth time

 

L119 – there are no Eq 4a or Eq 4b

――Thanks, we have corrected this error. L151.

 

L133 – replace ‘preciseness’ with ‘robustness’ or ‘reliability’

――Thanks, we have modified this inappropriate wording.

 

Fig. 2A and 3A. x-axis label – what is ‘site sacle’

――Thanks. Here we need to clarify this point. The sample for calculating the coefficient of variation of the tree radius can be limited to a particular site or cover the same species in different sites.

 

Fig. 2 – you need to briefly explain what Normalised growth rate is. You need to explain the different colours of the data points. And in the final line of the caption you mean C and B (not D and C)

――Thanks. We have added the necessary notes and fixed the errors. Normalized growth rate: ratio of the average growth rate of tree cores over the past five years to the average growth rate of all tree cores (f(r)_c/f(r)_m). The red and green lines indicate negative and positive effects, respectively. L236-237. L 245-246.

 

Fig. 3. What effect might elevation have other than on climate? Could it be subsuming some of the precipitation effects? I might have missed something here, but why are you using average/maximum tree age, whereas elsewhere you have looked at average/maximum size (radius).

――Thanks for your advice. It is undisputed that elevation is correlated with temperature (precipitation). Here (Fig. 3B), we use the hierarchical partitioning method (Lai et al., 2022) to estimate the individual contribution of elevation and temperature (precipitation). Other environmental factors, such as soil and radiation, are also related to elevation. The analysis here emphasizes the interpretation of the dependent variable by other elevation-dependent environmental factors.

The average/maximum tree age indicator reflects the growth stage of the forest community. The average/maximum size (radius) indicator is the independent variable that drives equation 2. Although these two indicators are highly correlated, they are used for different purposes. In Fig. 3C, mean/maximum tree age is used to represent normalized stand age, considering that CVR and OVG are also normalized variables.

 

Lai, J., Zou, Y., Zhang, J. & Peres-Neto, P.R. (2022). Generalizing hierarchical and variation partitioning in multiple regression and canonical analyses using the rdacca.hp R package. Methods in Ecology and Evolution, 13, 782-788.

 

 

L207-213 – this paragraph was unclear. The references to Fig 4C and 4D should probably be to 4B and 4C.

――Thanks you for your advice. We modified this unclear text to read:Conversely, the current growth rate of trees situated above the treeline is generally lower than the historical best value. If we assumed that the average radial growth rate is only half of the ideal, then we can calculate the historical average growth rate. We defined the difference between the current average growth rate and the historical average growth rate as the net increase in radial growth rate. In the low elevation region, there is a strong negative correlation between this net increase rate and the mean temperature of the wettest quarter (Fig. 4D). On the other hand, at the upper treeline, the minimum temperature of the coldest month is positively correlated with this net change (Fig. 4E). L 253-261.

 

Fig. 4A is not properly explained in the caption, which makes it difficult to interpret. What does ‘Number’ refer to – trees? What does each point represent? (figure captions should be self-contained). 4D and 4E – these are not directly comparing the historical best and current radial growth rates, but rather are comparing the difference in the two with the two meteorological variables. (It is very important to be precise.) Units of y axis are not given. And don’t split figures across pages

――Thanks you for your advice. We have modified this. Fig. 4 Current average radial growth rate of trees at different radius vs. model-estimated historical best radial growth rate at these radius (A), comparison of their differences at different elevations (B and C) (W-test, ***P<0.001) and the correlation of this difference with climatic factors (D and E). L265-268.

 

 

L222-224 – references for these statements?

――Thanks. We have added relevant references. L277

Wilmking, M., Scharnweber, T., van der Maaten-Theunissen, M. & van der Maaten, E. (2017). Reconciling the community with a concept—The uniformitarian principle in the dendro-sciences. Dendrochronologia, 44, 211-214.

Peltier, D.M.P. & Ogle, K. (2020). Tree growth sensitivity to climate is temporally variable. Ecology Letters, 23, 1561-1572.

 

The first paragraph of the discussion was quite difficult to follow, and it was sometimes unclear what is already known from the findings of this study (e.g. lines 226-229).

――Thanks. We have rewritten this section.

Tree radial growth is typically believed to be influenced by age (or size) and climatic factors, manifested as age-dependent low-frequency and climate-sensitive high-frequency signals (Wilmking et al. 2017; Peltier & Ogle 2020; Yao et al. 2023). After removing the modulations of the age effect trend, this signal typically exhibits stable climate sensitivity, termed stationarity assumptions or uniformity principles (Wilmking et al. 2017; Peltier & Ogle 2020). Based on this principle, it is possible to obtain high-resolution global information on tree species’ responses to global change, forest carbon and water dynamics, and past climate variability and extremes from tree ring dynamics (Wilmking et al. 2020). However, our research indicated that there is no essential difference between high-frequency and low-frequency growth signals. Mathematically, the high-frequency signal is the limited fluctuation of the low-frequency signal, and both are mediated by tree size or radius, mr/gr, and potential maximum size or radius, where environmental and resource intake could significantly affect mr and the maximum size or radius.. Some evidence supports this conclusion. For example, age effects and sampling strategy affect the accuracy of tree growth assessment and its climate response (Wu et al. 2013; Sun & Liu 2015; Wang et al. 2017). Moreover, changes in tree physiological status (Peltier & Ogle 2020) result in different climate-growth relationships (Wu et al. 2013; Sun & Liu 2015) and inevitably feed back to the size-to-growth constraint (Coomes et al. 2012; Pillet et al. 2018; Shu et al. 2019). In fact, size has a greater effect than cellular senescence on age-related declines in relative growth and net assimilation rates (Mencuccini et al. 2005). Our model highlights that tree size, specifically the radius, determines the radial growth trend and climate sensitivity (Fig. 5). That is, the variation coefficient of radial growth rate still shows a single-peaked pattern on the radius gradient.

L 272-295.

 

L248 – and whether the forests are single- or multi-aged!

――Thanks. We added the reasons for the formation of tree size heterogeneity to the introduction

In multi-aged stands, the causes of size heterogeneity are related to repeated disturbances or silvicultural interventions that regenerate these new age classes (Nepstad et al. 2007; O'Hara & Ramage 2013; Shu et al. 2019). In single-aged stands, competition directly drives tree size differentiation (Forrester 2019). Obviously, to a large extent, tree size heterogeneity may mediate the effects of competition and disturbance on tree growth.

This paragraph has also been revised. L303-317

 

L253 – frost damage is usually reduced by increased precipitation. However, decreased solar radiation associated with increased precipitation may constrain growth

――Thanks. We have modified the view.

CVR is more related to competition and disturbance, so we considered disturbance events.

However, at higher elevations with lower temperatures, increased precipitation may cause trees to experience more snowfall events and physical disturbances, resulting in increased CVR. L308-311

 

Fig. 5 – need key to colours

―Thanks. We have added the relevant descriptions

Our model highlights that tree size, specifically the radius, determines the radial growth trend and climate sensitivity (Fig. 5). That is, the variation coefficient of radial growth rate still shows a single-peaked pattern on the radius gradient.

 

 

L268 – ‘Subalpine forests … have a normal distribution of tree sizes … because trees of different sizes experience different ecological pressures’ – I am not convinced by the logic here – nor in the rest of this paragraph.

―Thanks. We have rewritten this paragraph.

Our research suggests that treeline expansion may be related to both tree size heterogeneity and temperature in different seasons. On the one hand, size inequality may cause an overall decrease in tree growth. We found that tree CVR was significantly lower in the low elevation region (< 3600 m) than in the high elevation region (> 3600 m) (see Appendix B, Fig. S2), while the radial growth rate of low elevation trees was also overall higher than the estimated historical best value (Figs. 4B and 4C). This difference can be attributed to the negative effect of CVR on OVG (Fig. 3). We speculated that competition and disturbance may be the main causes of increased CVR and decreased OVG. Usually, in natural forests, smaller trees are more vulnerable to asymmetric competition, whereas larger trees are prone to being affected by disturbance (Coomes et al. 2012; Pillet et al. 2018; Shu et al. 2019). For subalpine forests, this pattern also broadly applies, but the proportion of large individuals decreases significantly with increasing elevation (see Appendix B, Fig. S3), implying that disturbance may be related to treeline formation. On the other hand, net changes in tree radial growth rates at low and high elevations may be differentially affected by temperature in different seasons. In the subalpine forest belts, precipitation tends to be more abundant during the growing season. Nevertheless, rising temperatures may lead to an increase in respiration rates, which can lead to a decrease in the allocation of photosynthetic product to growth (Peng et al. 2013). Consequently, we can observe a significant negative correlation between net changes in tree radial growth rates and the mean temperature of the wettest quarter (Fig. 4D). At higher elevations, low temperatures limited tree growth, showing a positive correlation between minimum temperature and this change (Fig. 4E). These results imply that global warming affects tree growth variability differently in the high and low elevation ranges of subalpine. L320-344

 

 

L275 – I am not convinced by the argument that size inequality should be negatively correlated with overall growth variability – wouldn’t the relationship be positive, all else being equal?

―Thanks. We need to clarify this. We can quantify the overall tree growth variability (OVG) relative to HBGT for the same community based on Eq. 2 as follows:

(3)

where c(r_i) and f(r_i)_HBGT denote the current diameter and average growth rate over the past five years, respectively, for tree i. Our subsequent results give the specific form of f(r)_HBGT and demonstrate that it follows a single-peaked trajectory and constrains the tree radial growth. When tree size heterogeneity is high, disturbance and competition may be more frequent, then c(r_i) may be generally lower, resulting in lower OVG. Therefore, tree size heterogeneity (CVR) has a negative effect on OVG.

 

 

I found the entire discussion confusing - it needs to be more carefully structured and clearly argued.

―Thanks. We have rewritten the discussion

Your helpful comments helped to improve the quality of the manuscript significantly. Thank you again!

Author Response File: Author Response.docx

Reviewer 3 Report

Journal: Forests (ISSN 1999-4907)

Manuscript ID: forests-2383562

Title: Effect of tree size heterogeneity on the overall growth trend of trees in coniferous forests of the Tibetan Plateau

 

Overall  Comments and Suggestions for Authors

Dear authors,

Regarding the tree growth and the effect of tree size heterogeneity, this manuscript might be interesting to the relevant researchers who deals with similar issues such as tree biometrics and forest growth, especially for Tibetan Plateau. Although the research question and ideas were interesting, several issues were unclear and needed with more explanation from my point of view. Those were mainly ascribed from the data samples, data description, the reasons of selecting IGM model, preset parameters, and results presentations.

I hope that this manuscript can be improved based on peer-review’s comments. My specific comments were provided in detail as follows.

 

Kind regards,

 

Reviewer

 

 

Point 1.

In Materials,

Since the authors analyzed the tree growth, it is important to note the number of sample trees used. In Table 1, it does not describe the tree growth information at all. An additional table with summary statistics (Mean, S.D., Min, and Max) should clearly present the age, dbh, and height information with the number of samples. Currently, it is hard to be persuasive.

 

 

Point 2.

In Methods,

The authors immediately dived into the Iterative Growth Model with very short text about hypothesis in Introduction. I consider that the supporting ideas on selecting this model on the basis is not enough to convince. In the study field about forest and tree growth, there are other types of nonlinear growth functions that were originated from ideas by hypothesizing the relationship between catabolism and anabolism. I judged that authors should provide any persuasive statements firstly before diving into the designated IGM model. I cannot find any rationales regarding this point. Therefore, the logic and methodology were not convincing to me.

 

Point 3.

Regarding the selected models, furthermore, the authors noted that the applied metabolic exponent was preset to 0.75 and there was lack of information about why it has to be like this. Many parts were blind, and explanations were not provided enough. These kinds of parts must be carefully dealt with and make it convincing for the rest of manuscripts.

 

Point 4.

I consider that the number of sites were not sufficient to explain or consider the climatic or site-related variables in this study. In order to consider the effects additionally, I believe that the number of site samples should be much more. In this context, Figure 4 and the related results were not convincing that much although it seems to show some trends. Also, In Figures S2 and S3, we can see that it does not describe the sufficient number of samples.

 

Point 5.

I think that other figures should be able to provide some information about the data characteristics and trends which authors want to present. To do so, before showing the coefficient of variation directly, the distribution or dispersion about each of the variables can be presented with the specific radius as well as r/R.

 

Point 6.

I think the issues mentioned above should be solved for clarity. Other than these issues, I appreciate the authors’ study and efforts on this topic.

 

 

Minor comments.

The figures should clearly present the information all including legends. For examples, the different colors were not described properly in Figures 2 and 5.

Many parts of Appendix A were overlapped with the Methods. It needs changes more concisely.