Developing and Comparing Individual Tree Growth Models of Major Coniferous Species in South Korea Based on Stem Analysis Data
Round 1
Reviewer 1 Report
General remarks of the reviewer
Title:
The title of the article is accurate and directly relates to the purpose of the research.
Abstract:
In the abstract, it must be noted that the sample trees came from 129 sites, 1 representative tree for 1 plot, which in the context of regional research indicates a relatively small empirical material.
Keywords: The keywords are specific to the topic under study.
Introduction:
The literature review is relatively poor in the context of the wealth of research on modeling tree growth. In particular, please complete with the most important publications from the 21st century.
Materials and Methods:
2.1.2. Plot Characteristic with Stand Density
The stand density index is the quotient of actual and maximum density. It seems logical that numerical and not indicator relationships were used to determine this characteristic. Please specify where the maximum density was obtained from.
2.2.1. Sample Tree Selection and Wood Disc Collection
The selection of one representative sample tree, despite the fact that it comes from the most numerous classes in biosocial terms in the stand: dominant or co-dominant, free from damage and desirable in terms of trunk quality, is associated with a very high risk that it will not be representative in the context of breast height form factor (F1,3) and in consequences of volume, which is very important when modeling growth.
By definition, DBH is a diameter at a height of 1.3 meters, for research to have global and not only local significance, this level must be maintained.
2.2.2. Tree Ring Measurements and Stem Analysis
It should be clarified that the 5-year height increment determined in this way, due to the determination of the tree top at a given age in the stem analysis, is subject to an error from 0 to the annual height increment - on average 0.5 increment.
I suggest separating the Results and Discussion chapters, especially after the previously suggested completion of the literature review, from subchapter 3.3. Model Evaluation and Applicability, You can smoothly go to the discussion of Your own results with the research of other authors.
Results:
3.1. Parameters Estimates and Fit Statistics
Please eliminate errors in the analysis of tables 4-6.
Please justify the choice of the Chapman-Richard function for the volume of the stem, which shows little importance for the parameters.
3.2. Growth Simulation and Characteristics
"The MAI culminated for each variable at an older age than the age of the CAI". This statement can be omitted because it is a constant relationship between MAI and CAI.
3.3. Model Evaluation and Applicability
“These plots represent dense stands, where the growth may not be as pronounced as that in low-density stands”. This is at least a debatable view, depending on the analyzed biometric feature, e.g. taking into account fast-growing species, the growth of height rate due to vertical competition is more dynamic in stands with higher density.
Conclusions:
On a general level, the conclusions are rational and constructive. Due to the remark in subsection 3.1. Parameters Estimates and Fit Statistics, explain the choice of the Chapman-Richard function for volume despite the insignificance of the parameters of the equation.
Technical Notes
Please correct stylistic, logical and technical inaccuracies in the text.
Details in the attached manuscript.
Summary of the review:
Recommends the Forest Editorial Board publish article after major revision.
Comments for author File: 
Comments.pdf
Author Response
Our Responses to the Comments of Reviewer 1
Journal: Forests (ISSN 1999-4907)
Manuscript ID: forests-2102253
Title: Developing and Comparing Individual Tree Growth Models of Major Coniferous Species in South Korea based on Stem Analysis Data
Authors: Yeongwan Seo, Daesung Lee, Jungkee Choi*
All line numbers in the responses refer to those in the revised manuscript.
General comments: General remarks of the reviewer. Details in the attached manuscript.
Response: We thank you for reviewing our manuscript and for providing valuable comments. We have carefully addressed all issues indicated by you. We hope that the manuscript is suitable for further process. The detailed authors’ responses were as follows.
Kind regards,
Authors
Point 1: Title: The title of the article is accurate and directly relates to the purpose of the research.
Response 1: We thank you for your evaluation.
Point 2: Abstract: In the abstract, it must be noted that the sample trees came from 129 sites, 1 representative tree for 1 plot, which in the context of regional research indicates a relatively small empirical material.
Response 2: We thank you for the note. We are aware of this. We just mentioned the number of samples and material description briefly in Abstract. The relevant model evaluations and caution can be found throughout the manuscript, especially in Discussion.
Point 3: Keywords: The keywords are specific to the topic under study.
Response 3: We thank you for your evaluation.
Point 4: Introduction: The literature review is relatively poor in the context of the wealth of research on modeling tree growth. In particular, please complete with the most important publications from the 21st century.
Response 4: We thank you for your comment. First, although there are many precedent studies regarding tree growth models, the studies on tree growth models using stem analysis data cannot be sufficiently found and there is a lack of studies particularly on the same species and/or modelling functions in both regional and world-wide level.
Second, in our manuscript, we tried to hold references which convey some meaningful points in terms of stem analysis, sigmoid function, and biometric growth pattern especially if it’s comparable. However, we sorted out the literatures and discarded the other references if it did not suit the purpose to make literature review be meaningful, helpful, and informative rather than just bulky and redundant. We believe the references were clear and concise as those were listed in the bibliography.
Third, please note that we tried to make the Introduction clear and concise to avoid any verbose explanation. Still, if the reviewer specifies which part of introduction should be improved and suggest the available literatures for those, our authors will sincerely appreciate it and consider citing.
Point 5: Materials and Methods: 2.1.2. Plot Characteristic with Stand Density
Lines 97–98: The stand density index is the quotient of actual and maximum density. It seems logical that numerical and not indicator relationships were used to determine this characteristic. Please specify where the maximum density was obtained from.
Response 5: We thank you for your comment. It was originated from Lee and Choi (2019). Per your suggestion, we specified the source and revised the sentence accordingly in Lines 99–100.
Point 6: 2.2.1. Sample Tree Selection and Wood Disc Collection
Lines 104–106: The selection of one representative sample tree, despite the fact that it comes from the most numerous classes in biosocial terms in the stand: dominant or co-dominant, free from damage and desirable in terms of trunk quality, is associated with a very high risk that it will not be representative in the context of breast height form factor (F1,3) and in consequences of volume, which is very important when modeling growth.
Response 6: We thank you for your comment. We understand your concept with regards to the relative tree biometric characteristics. Still, we believe that our current models are adequate. It’s because sample tree size distributed numerous ranges and because our model was not targeted directly for volume equation, for example, as a function of DBH and/or height. All data were based on the stem analysis measurement and calculation.
Point 7: By definition, DBH is a diameter at a height of 1.3 meters, for research to have global and not only local significance, this level must be maintained.
Response 7: We thank you for your comment.
Even though the dbh is the most common variable, the definition of the measurement height is internationally unambitious, which means that it is not clearly defined. For example, if you come to Australia, New Zealand, Myanmar, or India, the dbh is measured in 4.7 feet. If you are measuring in South Korea, the dbh is measured at 1.2 m. Foresters in the United States are measuring in 4.5 feet. The rest of the world would measure the dbh in 1.3 meters with metric units.
In this context, we obviously provided the definition of the dbh as mentioning South Korea in Lines 110–111. Our description should be clear throughout the whole manuscript with this information.
Point 8: 2.2.2. Tree Ring Measurements and Stem Analysis
It should be clarified that the 5-year height increment determined in this way, due to the determination of the tree top at a given age in the stem analysis, is subject to an error from 0 to the annual height increment - on average 0.5 increment.
Response 8: We thank you for your comment. We agree that there is a small possibility in height increment when it comes to the stem analysis. It would be an innate characteristic of stem analysis as experts know. It should be noticeable and understandable with our description in Lines 132–134.
Point 9: I suggest separating the Results and Discussion chapters, especially after the previously suggested completion of the literature review, from subchapter 3.3. Model Evaluation and Applicability. You can smoothly go to the discussion of your own results with the research of other authors.
Response 9: We thank you for your comment. First, we considered separating the Results and Discussion before first submission. Since we have structured the logic with the comparison of previous studies which have similar approaches or materials, we decided to keep the original outline. By doing so, we still believe that one can understand our results and discussion conveniently and get straightforward ideas on each subsection of 3.1. Parameters Estimates and Fit Statistics, 3.2. Growth Simulation and Characteristics, and 3.3. Model Evaluation and Applicability. Otherwise, there could have been unnecessary repeats of explanation with separate chapters.
Second, as aforementioned, we tried to cite the literatures by focusing only on the necessity, importance, and comparability. Hence, the selected literatures were 43 references as listed in bibliography. we believe that the number of cited literatures shouldn’t be lacking considering our criteria in terms of stem analysis data, modeling approach, selected sigmoid function, and tree biometric features. If you would agree that stem analysis research works with the field and office are so demanding to collect the sample data, measure, and calculate dbh, height, basal area, and moreover volume (especially large trees), the lack of research paper could be understandable or possible.
Third, the references cited in Results and Discussion are 11, 13–17, 24–25, and 30–43 in the places where those are necessary and/or meaningful. The literatures cited include a rather wide temporal and spatial range in terms of publication date and country, respectively. We found some recent publications, but it was not included in this manuscript. It’s because those did not suit our criteria and purpose. Still, we would appreciate the reviewer if you can possibly offer any other literatures which fit our criteria as mentioned above.
Point 10: Results: 3.1. Parameters Estimates and Fit Statistics
Please eliminate errors in the analysis of tables 4-6.
Response 10: We thank you for your comment. We checked all once again and revised it correctly.
Point 11: In line 233, justify the choice of the Chapman-Richard function for the volume of the stem, which shows little importance for the parameters.
Response 11: We thank you for your comment.
First of all, the insignificant p-value we decided to evaluate was p>0.01 as written and additionally revised in Lines 208, 211, and 212. Second, the parameters of Chapman-Richards model for Lk volume were viable with our criteria.
Third, considering all these evaluations, namely parameter estimates, fit statistics, residuals, and prediction suitability, we selected the best growth function for each variable. It means that we also take into account the suitability of model behavior in terms of prediction.
Fourth, to make this explanation clearer, we have revised the sentence in Lines 233–235 as below.
“The prediction suitability, such as asymptote, growth rate, and inflection point, was also evaluated for the final model selection in addition to the statistical assessment (e.g., Supplementary files).”
Point 12: 3.2. Growth Simulation and Characteristics
Lines 287-288: "The MAI culminated for each variable at an older age than the age of the CAI". This statement can be omitted because it is a constant relationship between MAI and CAI.
Response 12: We thank you for your comment. We are already aware of the fact. However, if you see the context and flow, by referring the sentences in lines 286-293, we believe that the description, including before and after the sentence, will help some readers get ideas about the degree of age difference between MAI and CAI by species and variables.
Point 13: 3.3. Model Evaluation and Applicability
In lines 341–3422: “These plots represent dense stands, where the growth may not be as pronounced as that in low-density stands”. This is at least a debatable view, depending on the analyzed biometric feature, e.g. taking into account fast-growing species, the growth of height rate due to vertical competition is more dynamic in stands with higher density.
Response 13: We thank you for your comment. We agreed your comment, and thus, we have revised the sentence so that it means both ways of growth characteristics in Lines 344–346.
“These plots represent dense stands, where the growth of biometric feature, for example diameter vs height, may differ from that in low-density stands.”
Point 14: Conclusions.
On a general level, the conclusions are rational and constructive. Due to the remark in subsection 3.1. Parameters Estimates and Fit Statistics, explain the choice of the Chapman-Richard function for volume despite the insignificance of the parameters of the equation.
Response 14: We thank you for your comment. We have revised the manuscript in the subsection 3.1. Parameters Estimates and Fit Statistics and obviously explained the reasons why we chose the Chapman-Richard function for volume. Thus, we consider the present conclusions as rational and constructive.
Point 15: Technical Notes.
Please correct stylistic, logical and technical inaccuracies in the text.
Details in the attached manuscript.
Response 15: We thank you for your comment. We have checked and responded it adequately. Please review our revised manuscript. It should be displayed with Track Changes.
Point 16: Summary of the review:
Recommends the Forest Editorial Board publish article after major revision.
Response 16: We sincerely appreciate your expertise, effort, and time you spent on this review once again. It helped us a lot to revise the original manuscript and rethink about our research work. We hope that the manuscript is suitable for further process. Please feel free to make any additional comments on the authors’ responses.
Author Response File: 
Author Response.docx
Reviewer 2 Report
The manuscript deals with the crucial issue of creating reliable allometric equations for predicting important variables such as growth rates and volume increment for three common forestry species in South Korea. The authors collected an impressive dataset, given the extremely hard work to get such data, especially for larger trees. The analyses were well done. The text is also very clear and well-written, and the results – description, tables and plots – are properly presented.
I would just stress more that the logistic model was significantly worse than other models despite still having a good fit because this information would be helpful for others who model tree growth.
There is no doubt that the manuscript fits well into the scope of Forests in terms of quality and studied topic. And I am thankful to the authors for such well-prepared manuscripts; it was interesting reading for me, and I look forward to seeing it published.
I find very few minor issues, which are below:
36-37 sentences not logically linked, split them in two, i.e. remove and between them: ….since the 1970s, and in 2020 approximately 36.9% (2,319,832 ha) of forests…..
152 add citations about the different backgrounds for each model, it is not clear where such information was obtained from.
198 add “all” before “…the models”, it is important to stress that all the tested models fitted the data well.
211 I would also mention that the logistic model performed much worse than all other models
Author Response
Our Responses to the Comments of Reviewer 2
Journal: Forests (ISSN 1999-4907)
Manuscript ID: forests-2102253
Title: Developing and Comparing Individual Tree Growth Models of Major Coniferous Species in South Korea based on Stem Analysis Data
Authors: Yeongwan Seo, Daesung Lee, Jungkee Choi*
All line numbers in the responses refer to those in the revised manuscript.
General comments: The manuscript deals with the crucial issue of creating reliable allometric equations for predicting important variables such as growth rates and volume increment for three common forestry species in South Korea. The authors collected an impressive dataset, given the extremely hard work to get such data, especially for larger trees. The analyses were well done. The text is also very clear and well-written, and the results – description, tables and plots – are properly presented.
I would just stress more that the logistic model was significantly worse than other models despite still having a good fit because this information would be helpful for others who model tree growth.
There is no doubt that the manuscript fits well into the scope of Forests in terms of quality and studied topic. And I am thankful to the authors for such well-prepared manuscripts; it was interesting reading for me, and I look forward to seeing it published.
Response: We thank you for reviewing our manuscript and for providing valuable comments. Also, we are so pleased to receive your appreciation of our manuscript and data especially for the tree samples.
With regards to the logistic model, we totally agree your comment and wish that our examples will help other modelers get any ideas from our study. By reflecting your comment, we added a sentence to refer the supplementary file for this matter in Lines 213–216. It should give a clearer connection and information.
Lastly, it was our honor to get your complimentary remark. We have revised the manuscript and responded your comment properly. We hope that the revised manuscript will provide clearer information on this topic and serve practical forestry well. The detailed responses are as follows.
Kind regards,
Authors
Point 1: I find very few minor issues, which are below:
In lines 36-37, sentences not logically linked, split them in two, i.e. remove and between them: ….since the 1970s, and in 2020 approximately 36.9% (2,319,832 ha) of forests…...
Response 1: We thank you for your comment. We totally agreed your suggestion. The sentences have been revised accordingly in Lines 37–39.
Point 2: In line 152, add citations about the different backgrounds for each model, it is not clear where such information was obtained from.
Response 2: We thank you for your comment. To make it clear, we have added a citation at the end of the sentence in Line 154. Per your suggestion, moreover, we have revised the sentence to match the background clearly in Lines 151–153: ecology, height growth of the forest stand, age distribution of the human population, animal growth, and probability distribution for failure rate, respectively.
Point 3: In line 198, add “all” before “…the models”, it is important to stress that all the tested models fitted the data well.
Response 3: We thank you for your comment. The phrase has been revised per your suggestion in Line 200.
Point 4: In line 211, I would also mention that the logistic model performed much worse than all other models
Response 4: We thank you for your comment. We definitely agreed your comment. Accordingly, the sentence has been improved by revising and adding additional explanation in Lines 213–216 as below.
“The Logistic function performed much worse than all the other functions in terms of the predicted curve with an extreme asymptote. Therefore, these models with unstable parameter estimates and/or unreliable behavior of prediction were excluded from the best function (Tables 3–6, Supplementary files).”
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
The Authors responded to all comments. In some cases, they are not fully in line with the views of the reviewer, but that's what real scientific discourse is all about. Fundamental issues have been clarified and corrected, assuming that the research is of a regional nature, I request the Forests Editorial Board to accept the article in present form, and I wish the Authors further success in modeling the growth of forest stands.
Reviewer
