Research on the Main Influencing Factors and Variation Patterns of Basal Area Increment (BAI) of Pinus massoniana

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Line 109, add more details about the Jigong Mountain Nature Reserve, for example:
“…is located in [province/specific coordinates], characterized by a subtropical monsoonal climate with an average annual temperature of X °C and average annual precipitation of Y mm. Elevation ranges from A to B meters, with dominant soil types including [soil type], which may…”

Insertion before line 110 (immediately before “…were”): “The candidate variables were selected based on”

Insertion after line 117 (immediately after “…and the enhancement”): “However,”

Insertion point: After line 118 (following the end of the introductory sections)

Insert the following text in English/for example:

“Nevertheless, the methodology has potential limitations. The use of multiple feature selection algorithms introduces risks of multicollinearity, and the application of GAMs, while effective for capturing nonlinearity, may be sensitive to outliers and boundary data. Furthermore, limited geographic replication could constrain the generalizability of the findings to broader subtropical forest systems.”

Lines 130–132 Recommendation: include a bibliographic note or explicit mention of the methods used to obtain these data (e.g., time period, measurement instruments, satellite sources, modeling).

In Figure 1, also specify the photo source.

Section 2.2 is well organized and carefully follows the steps of a dendrochronological study. However, this section needs:

• complete references for software and formulas,
• justification for the exclusion of cores.

Lines 175–183 Observation: The significant elevation difference between the station (114 m) and the site (Jigong Mountain) may introduce bias in interpretation. Recommendation: Justify the choice of the station and discuss potential limitations regarding the accuracy and representativeness of the data in a mountainous context.

Lines 168–174 Calculation of the Competition Index – Hegyi Index  
The choice of a distance‐dependent index is appropriate for the ecological study. The formula is clearly presented and the variables are explicit.  
Observation: Details are missing on the competitor selection algorithm and practical implementation.  
Recommendation: Include information on the software used, the distance‐cutoff criteria, the method version, and provide a full citation for reference [39].

At section 2.3.1 “Principal Component Analysis (PCA)”:

Observation: The text notes that variables were evaluated, but does not specify which variables scored highest or how those results were interpreted in an ecological context.

Recommendation: Add a brief discussion identifying which types of variables (e.g., diameter at breast height (DBH), tree height, key climatic parameters) emerged as most influential in the PCA, and explain how these findings support your ecological hypotheses.

Also in this section, terminological clarity is needed: terms like PCAr and PCAl are defined, but it would be useful to include a conceptual diagram or an explanatory paragraph clearly distinguishing the two approaches for readers less familiar with PCA.

Line 209: “canonical correlation analysis (CCA).”  
“RDA was implemented using a first-order canonical correlation analysis (CCA) model.”  
Revision: Mention and justify the choice of the axis. Mention and justify selecting axis 1 as the benchmark for importance — add clarifications regarding the proportion of variance explained.

Line 211: “The absolute value of each variable’s loading on the first canonical axis …”  
Revision: Specify whether the included variables were ecological, climatic, dendrometric, etc.

Line 216–218: Description of the OOB permutation
Revision: Add the number of trees used in the RF and the impurity criterion.

Line 221: “…implemented using sklearn.inspection.permutation_importance…”
Revision: Indicate whether a fixed random_state was used and whether the stability of the results was tested.

Line 224–225: “…calculating the importance based on average reduction in MSE…”
Revision: Specify the hyperparameters: number of trees, maximum depth, learning rate.

Line 233: “…with all hyperparameters set to default values…”
Revision: Justify the choice of default values or describe any tests of their impact on model performance.

Line 243–244: General GAM formula  
Revision: Specify the software used (e.g., R – mgcv, Python – pyGAM) and the smoothing method (e.g., splines, penalty, k).  

Required revision (lines 253–254): Figure 3 is mentioned but lacks an analytical discussion.  
Recommendation: Expand the interpretation of the heatmap to highlight method-specific discrepancies or convergences.

Ecological contextualization (lines 255–256): The ecological rationale behind the dominance of tree age and the competition index is briefly mentioned but not elaborated.  
Recommendation: Integrate references and discussions from tree physiology and forest ecology literature to explain how ontogenetic processes and density‐dependent effects regulate radial growth.

Lines 273–289: The choice of k values (3, 5, 7) is not explained. Are they based on previous studies, computational constraints, or the characteristics of the observed data?

The manuscript does not mention the software package or programming environment used for the GAM implementation (for example, in R or in Python: mgcv, pyGAM).  
Recommendation: Justify the selected range for k and the choice of basis functions.

Specify the computational framework (including version and package) and describe any smoothing penalty criteria used to avoid overfitting.  
Justify the selected range for k and the choice of basis functions. Specify the computational framework (including version and package) and describe any smoothing penalty criteria used to avoid overfitting.

Line 351–353  
“This pattern aligns with general tree growth dynamics…”  
Revision: Missing bibliographic citations for growth; add a valid reference.

Line 376–378  
“…suggesting that this temperature may represent a physiological optimum…”  
Revision: The 17 °C threshold is presented as an “optimum” — confidence intervals or significance tests are missing.

Line 401–402  
“…the response curve remains relatively flat across the entire observed range.”  
Review: This statement needs graphical or statistical support — add standard deviations or the rate of change of BAI per unit precipitation.

Line 436–438: “…soil depth has already exceeded the physiological threshold…”  
Reason: Specify what “threshold” means—whether it is drawn from the literature or observed empirically.  

Line 443: “…potentially limiting the detection of late-stage growth dynamics…”  
Reason: It would be useful to state the maximum age of the trees and how that relates to published senescence data.  

Line 448: “…weakened the generalizability of the findings…”  
Reason: Propose strategies to expand the sample—e.g., include a wider altitudinal range.  

Lines 449–455: “…climate variables… aggregated at the annual scale…”  
Reason: Well noted—but specify which seasonal data would be most informative (e.g., March–April temperatures for cambial activity).  

Line 461: “…provide a scientific basis for ecological suitability assessments…”  
Reason: A practical example would be desirable—explain how the model could inform silvicultural planning or adaptive management.

Line 471: “…does not exceed 70 years…” “When the tree age does not exceed 70 years…”  
Review: It is unclear whether 70 years represents a methodological cutoff (imposed by sampling) or a biological inflection point in growth.  
Suggestion: Clarify whether this threshold is derived from empirical data or from the literature, and whether any post-70-year data are available.

Line 483–484: “…potential and general productivity…” “…the potential for carbon sequestration and general productivity…”  
Review: The statement is valuable, but it would be useful to link this potential to the concrete indicators used in the study (e.g., BAI as a proxy for productivity).

Author Response

Comments 1:Line 109, add more details about the Jigong Mountain Nature Reserve, for example:

“…is located in [province/specific coordinates], characterized by a subtropical monsoonal climate with an average annual temperature of X °C and average annual precipitation of Y mm. Elevation ranges from A to B meters, with dominant soil types including [soil type], which may…”

 

Insertion before line 110 (immediately before “…were”): “The candidate variables were selected based on”

 

Insertion after line 117 (immediately after “…and the enhancement”): “However,”

 

Insertion point: After line 118 (following the end of the introductory sections)

 

Insert the following text in English/for example:

 

“Nevertheless, the methodology has potential limitations. The use of multiple feature selection algorithms introduces risks of multicollinearity, and the application of GAMs, while effective for capturing nonlinearity, may be sensitive to outliers and boundary data. Furthermore, limited geographic replication could constrain the generalizability of the findings to broader subtropical forest systems.”
Responses 1:We sincerely thank the reviewer for the suggestion regarding the description of the study area. After careful consideration, we would like to clarify that a detailed introduction to the Jigong Mountain Nature Reserve, including its geographic location, climate characteristics, elevation range, dominant soil types, and vegetation composition, is already provided in Section 2.1 titled "Study Area." This section offers a comprehensive overview intended to help readers understand the ecological context of our research.

To maintain conciseness and avoid unnecessary repetition, we decided not to duplicate this information in the Introduction section around Line 109. Instead, we have ensured that Section 2.1 contains all relevant details in a clear and accessible manner. We truly appreciate the reviewer’s concern for completeness and have rechecked this section to confirm that it adequately informs readers who are not familiar with the region.

In addition, we have accepted the reviewer’s other helpful suggestions for improving the Introduction. We have added a sentence to clarify the rationale behind the variable selection process, and we have included a new paragraph to acknowledge potential methodological limitations, as requested.

Comments 2:Lines 130–132 Recommendation: include a bibliographic note or explicit mention of the methods used to obtain these data (e.g., time period, measurement instruments, satellite sources, modeling).
Response 2:We thank the reviewer for this helpful suggestion. The climate data used in this study were obtained from existing literature and are cited accordingly in the manuscript. We appreciate the reviewer’s attention to this issue and have ensured that the citation is clear and traceable.

Comments 3:In Figure 1, also specify the photo source.
Response 3:We thank the reviewer for highlighting the need to specify the image source in Figure 1. The map was created by the authors based on field survey data, and the basemap was obtained from the publicly available satellite imagery service provided by ESRI (https://services.arcgisonline.com /arcgis/rest/services/World_Imagery/MapServer). This information has now been added to the manuscript to ensure proper attribution and enhance transparency.

Comments 4:Section 2.2 is well organized and carefully follows the steps of a dendrochronological study. However, this section needs:

complete references for software and formulas,

justification for the exclusion of cores.
Response 4:We thank the reviewer for this valuable comment. As noted, we have already included the criteria for core exclusion in Section 2.2, indicating that 27 out of 87 cores were removed due to failure to reach the pith or low correlation with the master chronology as verified by COFECHA. In addition, we have added citations for the key formulas used in basal area increment (BAI) calculations. These revisions improve the clarity, traceability, and reproducibility of the analytical procedures. While WINDENDRO does not have a formal citable reference.

Comments 5:Lines 175–183 Observation: The significant elevation difference between the station (114 m) and the site (Jigong Mountain) may introduce bias in interpretation. Recommendation: Justify the choice of the station and discuss potential limitations regarding the accuracy and representativeness of the data in a mountainous context.
Response 5:We feel great thanks for your professional review work on our article. As you are concerned, there are several problems that need to be addressed. According to your valuable suggestions, we have revised the manuscript to add a justification for the selection of the Xinyang Meteorological Station. Specifically, we explained in the text that this station was chosen because it is the closest to the study area and provides the most complete and continuous long-term climate records. Although there is an elevation difference between the station and the actual sampling sites in Jigong Mountain, high-elevation climate data with comparable temporal coverage are not available. Furthermore, previous studies conducted in the same region have also used data from the Xinyang Meteorological Station for tree growth and climate analysis, such as:

1. Wei, X.-X. et al. (2024). Response of Radial Growth of Pinus tabuliformis to Climatic Factors in the North–South Transition Zone of China—A Case Study of Jigong Mountain National Nature Reserve. Geographical Science, 44, 1643–1652.
2. Yang, L. (2020). Response of Pinus massoniana and Pinus tabuliformis Growth to Climatic Factors in Jigong Mountain. Master’s thesis, Henan University, Kaifeng, China.

Thank you again for your thoughtful comments and helpful suggestions to improve the scientific quality and clarity of our manuscript.

Comments 6:Lines 168–174 Calculation of the Competition Index – Hegyi Index

The choice of a distance‐dependent index is appropriate for the ecological study. The formula is clearly presented and the variables are explicit.

Observation: Details are missing on the competitor selection algorithm and practical implementation.

Recommendation: Include information on the software used, the distance‐cutoff criteria, the method version, and provide a full citation for reference [39].
Response 6: We feel great thanks for your professional review work on our article. As you are concerned, our original manuscript lacked sufficient detail regarding the implementation of the Hegyi competition index. In the revised version, we have now provided a comprehensive explanation. Specifically, the index was calculated using the ForestStatTool package in R (version 4.4.3), based on the three nearest neighboring trees (k = 3) within a 10 m × 10 m plot. A buffer zone correction was applied to exclude trees within 4.9 m of the plot edge. All spatial coordinates were standardized prior to calculation. These methodological details have now been explicitly stated, and we have also provided a full citation for Reference [40] to support this approach.

 

Comments 7:At section 2.3.1 “Principal Component Analysis (PCA)”:

 

Observation: The text notes that variables were evaluated, but does not specify which variables scored highest or how those results were interpreted in an ecological context.

 

Recommendation: Add a brief discussion identifying which types of variables (e.g., diameter at breast height (DBH), tree height, key climatic parameters) emerged as most influential in the PCA, and explain how these findings support your ecological hypotheses.

 

Also in this section, terminological clarity is needed: terms like PCAr and PCAl are defined, but it would be useful to include a conceptual diagram or an explanatory paragraph clearly distinguishing the two approaches for readers less familiar with PCA.
Response 7:We feel great thanks for your professional review work on our article. As you pointed out, additional clarification was needed regarding the terminological distinction between PCAr and PCAl. In the revised manuscript, we have added a paragraph in Section 2.3.1 to clearly explain the conceptual differences between these two approaches. Specifically, PCAr incorporates the response variable (BAI) through regression modeling, providing a supervised assessment of variable importance, while PCAl evaluates variable importance based solely on the variance structure of the predictor variables in an unsupervised manner.

Regarding the concern about identifying key variables and interpreting the results in an ecological context, we would like to clarify that the main findings from PCAr and PCAl have already been presented and discussed in Section 3 of the manuscript. Since Section 2.3.1 is dedicated to methodological description, we intentionally avoided repeating the interpretation there to avoid redundancy.

Comments 8:Line 209: “canonical correlation analysis (CCA).”

“RDA was implemented using a first-order canonical correlation analysis (CCA) model.”  

Revision: Mention and justify the choice of the axis. Mention and justify selecting axis 1 as the benchmark for importance — add clarifications regarding the proportion of variance explained.
Response 8:We thank the reviewer for this valuable suggestion. In the revised manuscript, we have clarified that only one constrained axis (RDA1) was available due to the univariate nature of the response variable (BAI). We further justify the selection of RDA1 as the benchmark for interpreting variable importance, as it accounted for 26.8% of the total variance. This explanation has been incorporated to enhance clarity and methodological transparency.

Comments 9:Line 211: “The absolute value of each variable’s loading on the first canonical axis …”  

Revision: Specify whether the included variables were ecological, climatic, dendrometric, etc.
Response 9:Thank you for your helpful comment. In the revised manuscript, we have clarified the types of variables involved in the RDA analysis, explicitly stating that they include ecological (e.g., elevation, slope), climatic (e.g., temperature, precipitation), and dendrometric factors (e.g., age, competition index).

Comments 10:Line 216–218: Description of the OOB permutation

Revision: Add the number of trees used in the RF and the impurity criterion.
Response 10:We sincerely thank the reviewer for the helpful suggestion. In response, we have revised the methodological description of the Random Forest (RF) model to explicitly include the number of trees (500) and the impurity criterion (squared error). Additionally, we specified the use of a fixed random state (42). These updates improve transparency and ensure the methodological clarity of the RF analysis.

Comments 11:Line 221: “…implemented using sklearn.inspection.permutation_importance…”

Revision: Indicate whether a fixed random_state was used and whether the stability of the results was tested.
Response 11:We appreciate the reviewer’s insightful comment. In response, we have added clarification in the manuscript indicating that a fixed random_state (set to 42) was used to ensure reproducibility. We also conducted a sensitivity test by varying the number of decision trees from 100 to 1000. The results showed minimal variation in the relative importance rankings of variables, which demonstrates the stability of the permutation-based importance estimates. Therefore, we selected 500 trees as a balanced choice.

Furthermore, we emphasize that our focus in this study was to identify and compare the relative importance of variables rather than to build a predictive model. Hence, extensive parameter tuning was not the primary objective.

Comments 12:Line 224–225: “…calculating the importance based on average reduction in MSE…”

Revision: Specify the hyperparameters: number of trees, maximum depth, learning rate.
Response 12:We appreciate the reviewer’s suggestion. In response, we have added a detailed description of the hyperparameter settings for the BRT model used to calculate variable importance. Specifically, the number of trees was set to 100, the maximum depth of each tree was set to 3, and the learning rate was fixed at 0.1. These parameters were selected based on conventional settings for interpretative purposes, as our primary objective was to assess variable importance rather than optimize predictive performance.

Comments 13:Line 233: “…with all hyperparameters set to default values…”

Revision: Justify the choice of default values or describe any tests of their impact on model performance.
Response 13:Thank you for this helpful suggestion. To assess the sensitivity of model performance to the n_estimators parameter, we tested a range of values from 50 to 1000 (with increments of 50) and evaluated mean squared error (MSE) on a validation set. The lowest MSE was observed when n_estimators=450. However, the differences in variable importance compared to those obtained using the default setting (n_estimators=100) were relatively small. Given that our primary aim was to assess relative variable importance rather than optimize predictive performance, and considering the minor impact observed, we decided to retain the default values for simplicity and reproducibility.

Comments 14:Line 243–244: General GAM formula  

Revision: Specify the software used (e.g., R – mgcv, Python – pyGAM) and the smoothing method (e.g., splines, penalty, k). 
Response 14:We sincerely appreciate the reviewer’s insightful comment. In response, we have clarified the software environment and methodological details in the revised manuscript. Specifically, the GAMs were implemented using the mgcv package (version 1.9-1) in R (version 4.4.3), employing penalized regression splines including thin plate ("tp") and cubic ("cr") spline bases. Furthermore, the detailed selection process of the basis dimension k and spline types has been thoroughly described in the subsequent experimental section to enhance reproducibility and methodological transparency.

Comments 15:Required revision (lines 253–254): Figure 3 is mentioned but lacks an analytical discussion.

Recommendation: Expand the interpretation of the heatmap to highlight method-specific discrepancies or convergences.
Response 15:We thank the reviewer for pointing out the need for a more analytical interpretation of Figure 3. In the revised manuscript, we have expanded the description of the heatmap by highlighting both the consistent patterns and the method-specific discrepancies in variable importance. For example, tree age and competition index were consistently ranked as top contributors, while climatic and topographic variables showed more variability across approaches. A brief explanation of these discrepancies has been added to the results section for clarity. The deeper causes underlying these differences, such as the linearity assumptions of PCA and RDA versus the nonlinear modeling capacity of machine learning methods are addressed in detail in the discussion section (Section 4.1).

Comments 16:Ecological contextualization (lines 255–256): The ecological rationale behind the dominance of tree age and the competition index is briefly mentioned but not elaborated.  

Recommendation: Integrate references and discussions from tree physiology and forest ecology literature to explain how ontogenetic processes and density‐dependent effects regulate radial growth.
Response 16:We thank the reviewer for this insightful suggestion. We agree that explaining the ecological mechanisms underlying the high importance of tree age and competition index is essential. To maintain structural clarity, we have kept the Results section focused on the presentation of quantitative outcomes, while the ecological interpretation has been integrated into the Discussion section, along with relevant literature in tree physiology and forest ecology.

Comments 17:Lines 273–289: The choice of k values (3, 5, 7) is not explained. Are they based on previous studies, computational constraints, or the characteristics of the observed data?

 

The manuscript does not mention the software package or programming environment used for the GAM implementation (for example, in R or in Python: mgcv, pyGAM).  

Recommendation: Justify the selected range for k and the choice of basis functions.

 

Specify the computational framework (including version and package) and describe any smoothing penalty criteria used to avoid overfitting.  

Justify the selected range for k and the choice of basis functions. Specify the computational framework (including version and package) and describe any smoothing penalty criteria used to avoid overfitting.
Response 17:We appreciate the reviewer’s thoughtful comments. In the revised manuscript (Section 2.4), we now justify the selection of basis dimensions (k = 3, 5, 7) based on the structure of the dataset. Because only eight plots were sampled, certain variables (e.g., soil thickness) had only eight distinct values, making larger k values inappropriate due to risks of overfitting or poor convergence. Similarly, very small k values (e.g., k = 2) may lead to under-smoothing. To balance model flexibility and data constraints, we selected this moderate range of k values. We also specify that GAM modeling was implemented in R (version 4.4.3) using the mgcv package, which incorporates penalized regression splines to avoid overfitting. These clarifications have been added to the manuscript.

Comments 18:Line 351–353  

“This pattern aligns with general tree growth dynamics…”  

Revision: Missing bibliographic citations for growth; add a valid reference.
Response 18:We thank the reviewer for this helpful suggestion. In the revised manuscript, we have added appropriate references to support the general pattern of declining tree growth with age and its underlying mechanisms.

Comments 19:Line 376–378  

“…suggesting that this temperature may represent a physiological optimum…”  

Revision: The 17 °C threshold is presented as an “optimum” — confidence intervals or significance tests are missing.
Response 19:We thank the reviewer for this valuable comment. We acknowledge that our previous wording may have overstated the interpretation of the temperature threshold. In the revised manuscript, we have clarified that the observed peak at approximately 16.96 °C represents a statistical optimum derived from the fitted GAM curve within the observed temperature range, and not necessarily a physiological threshold. We have accordingly revised the relevant sentence to avoid biological over-interpretation while preserving the core finding. This change improves the precision of our interpretation and aligns with the reviewer’s concern.

Comments 20:Line 401–402  

“…the response curve remains relatively flat across the entire observed range.”  

Review: This statement needs graphical or statistical support — add standard deviations or the rate of change of BAI per unit precipitation.
Response 20:We thank the reviewer for pointing this out. Our original intention was to emphasize that, within the observed range, the BAI–precipitation response curve is relatively flat in comparison to those of other climatic variables. To clarify this point and avoid potential overgeneralization, we have revised the sentence accordingly in the manuscript.

Comments 21:Line 436–438: “…soil depth has already exceeded the physiological threshold…”  

Reason: Specify what “threshold” means—whether it is drawn from the literature or observed empirically. 
Response 21:We appreciate the reviewer’s insightful comment. In this context, the term “physiological threshold” was intended to convey that soil thickness in most plots was sufficient to meet the basic water and nutrient demands for Pinus massoniana growth. Once this requirement is fulfilled, additional increases in soil depth may no longer translate into enhanced radial increment. We have revised the sentence to clarify that this threshold is conceptual and based on ecological reasoning rather than a specific value drawn from the literature or field measurements.

Comments 22:Line 443: “…potentially limiting the detection of late-stage growth dynamics…”  

Reason: It would be useful to state the maximum age of the trees and how that relates to published senescence data.
Response 22:In South China, according to the regulations of forest resources investigation, when the age of Pinus massoniana is more than 61 years it is called over matured tree or forest stand, in our study the maximum age was 69, but there is almost lack of the published senescence data more than 70 years. Only in some natural reserves few little populations of Pinus massoniana were found.

Comments 23:Line 448: “…weakened the generalizability of the findings…”  

Reason: Propose strategies to expand the sample—e.g., include a wider altitudinal range.
Response 23:Thank you for this insightful suggestion. We have revised the manuscript to include specific strategies to improve representativeness in future studies. These include sampling trees with a wider age range, expanding altitudinal coverage, and selecting plots with diverse slope aspects and slope.

Comments 24:Lines 449–455: “…climate variables… aggregated at the annual scale…”  

Reason: Well noted—but specify which seasonal data would be most informative (e.g., March–April temperatures for cambial activity).
Response 24:We appreciate the reviewer’s constructive suggestion. In the revised manuscript, we have elaborated on the seasonal sensitivity of tree growth to climate variability by specifying that April–May temperatures are particularly critical for cambial activity and xylem formation in conifer species, as supported by Rossi et al. (2016). This addition helps clarify the ecological rationale for incorporating finer-scale climate data in future analyses.

Comments 25:Line 461: “…provide a scientific basis for ecological suitability assessments…”  

Reason: A practical example would be desirable—explain how the model could inform silvicultural planning or adaptive management.
Response 25:We thank the reviewer for the constructive suggestion. In response, we have expanded the conclusion section by adding a practical example of how the GAM results could inform forest management. Specifically, we now mention that the findings can guide site selection and stand density adjustment for Pinus massoniana under different environmental conditions, thus enhancing the real-world applicability of our results.

Comments 26:Line 471: “…does not exceed 70 years…” “When the tree age does not exceed 70 years…”

Review: It is unclear whether 70 years represents a methodological cutoff (imposed by sampling) or a biological inflection point in growth.

Suggestion: Clarify whether this threshold is derived from empirical data or from the literature, and whether any post-70-year data are available.
Response 26:Thank you for your insightful comment. The 70-year threshold does not represent a biological inflection point, nor is it derived from the literature. Rather, it corresponds to the maximum age of trees within our current dataset (i.e., 69 years). We currently do not have access to data from older individuals. In future studies, we aim to incorporate older age classes to more comprehensively assess long-term growth dynamics.

Comments 27:Line 483–484: “…potential and general productivity…” “…the potential for carbon sequestration and general productivity…”  

Review: The statement is valuable, but it would be useful to link this potential to the concrete indicators used in the study (e.g., BAI as a proxy for productivity).
Response 27:We sincerely appreciate the reviewer’s constructive suggestion. In the revised manuscript, we have added a clarification that BAI (basal area increment) is commonly used as a proxy for stand productivity and biomass accumulation in forest ecology. We hope this clarification helps to better contextualize the discussion on the carbon sequestration potential and overall productivity of Pinus massoniana.

Reviewer 2 Report

Comments and Suggestions for Authors

1.Line 19: Change "commonly" to "common".

2.The Introduction section needs to point out the defects or deficiencies of past research, and what scientific question this study aims to solve compared to previous studies, in order to strengthen the importance of this research.

3.Although the paper uses a variety of methods to assess the importance of variables, including traditional multivariate statistical methods and modern machine learning algorithms, the differences between these methods can be further refined in the presentation and interpretation of the results. It is recommended to add a specific discussion on the applicable scenarios of different methods. For example, it can be elaborated in what situations machine learning algorithms (such as RF, BRT, and XGBoost) are more advantageous than traditional statistical methods (such as PCA and RDA), and how these differences affect the interpretation of variable importance.

4.For the Latin names of plants mentioned in the paper, please use the full name when they first appear in each section (including abstract, main text, figures, tables, etc.), and use the abbreviated genus name for subsequent mentions.

5.Please adjust the size of each sub - figure in Figure 5 to be consistent.

6.Lines 341 - 344: Please discuss the reliability of the averaged importance evaluation method. You can also cite other similar evaluation methods or use the results obtained from the study to discuss its reliability.

7.The overall discussion in "4.2 Influencing variables of BAI in Pinus massoniana" lacks an ecological theoretical perspective. Please appropriately combine ecological theories to explain the observed phenomena or results.

8.The paper notes that climate variables significantly impact the radial growth of Masson pine and uses generalized additive models (GAM) to reveal the nonlinear relationships of these variables. However, the ecological significance behind these nonlinear relationships is somewhat under - discussed. It is suggested to conduct an in - depth analysis of the ecological mechanisms of these nonlinear responses. For instance, regarding the nonlinear responses to temperature and precipitation, you can combine the physiological characteristics of Masson pine to explore its physiological adaptation mechanisms under different temperature and precipitation conditions, as well as the potential responses of such nonlinear relationships to future climate change.

9.The research conclusions of the paper are primarily based on data from a specific location (Jigong Mountain Nature Reserve). Although this site is representative, the generalizability of the conclusions may be limited. It is recommended to add a detailed discussion on the scope of applicability of the research conclusions in the discussion section. You can combine research findings from other regions on Masson pine to analyze the similarities and differences in the responses of Masson pine growth to environmental factors across different regions, thereby more accurately assessing the applicability of the conclusions of this study in a broader area.

10.Lines 478 - 481: This part is the focus of the Conclusion. It is necessary to specifically explain what kind of management strategies and ecological protection can be provided for forests based on the results of this study.

11.The reference format is somewhat messy. Please standardize it according to the journal requirements, including the abbreviation of journal names, the italicization of Latin names in article titles, etc.

Author Response

Comments 1:1.Line 19: Change "commonly" to "common".
Response 1:Thank you for pointing out the grammatical inconsistency. We sincerely apologize for the oversight. The phrase has been revised in line 19 to ensure proper usage.

Comments 2:2.The Introduction section needs to point out the defects or deficiencies of past research, and what scientific question this study aims to solve compared to previous studies, in order to strengthen the importance of this research.
Response 2:We thank the reviewer for this insightful suggestion. As advised, we have revised the Introduction section to more explicitly address the limitations of previous research and to clarify the specific scientific gap this study aims to fill. We now emphasize:

1. That ring width, commonly used in past studies, may inadequately reflect actual stem volume increment, making BAI a more ecologically meaningful growth indicator.
2. That previous studies often relied on linear statistical methods, which may overlook nonlinear interactions among environmental factors.
3. That there is a need to integrate multivariate statistical approaches with machine learning algorithms to better capture complex growth–environment relationships.

These revisions help highlight the significance and novelty of our approach, which combines BAI with multiple variable importance algorithms and GAM-based modeling to understand Pinus massoniana radial growth under subtropical conditions.

Comments 3:3.Although the paper uses a variety of methods to assess the importance of variables, including traditional multivariate statistical methods and modern machine learning algorithms, the differences between these methods can be further refined in the presentation and interpretation of the results. It is recommended to add a specific discussion on the applicable scenarios of different methods. For example, it can be elaborated in what situations machine learning algorithms (such as RF, BRT, and XGBoost) are more advantageous than traditional statistical methods (such as PCA and RDA), and how these differences affect the interpretation of variable importance.
Response 3:As your suggestion, we have added more discussion about the different methods, according to our goal, we would not like the screen out a more fitting method, but to employ six variable importance assessment which have strong explanatory power for the GAMs. So, the situations or scenarios are not the only answers. See Figures 3 the heatmap shows the importance of each method.

Comments 4:4.For the Latin names of plants mentioned in the paper, please use the full name when they first appear in each section (including abstract, main text, figures, tables, etc.), and use the abbreviated genus name for subsequent mentions.
Response 4:Thank you for your careful review. According to your suggestion, we have checked all sections of the manuscript, including the abstract, main text, figure and table captions. For each section, we now use the full Latin name (e.g., Pinus massoniana) upon first mention and apply the abbreviated genus name (e.g., P. massoniana) for subsequent occurrences. These revisions have been made to ensure taxonomic clarity and consistency.

Comments 5:5.Please adjust the size of each sub - figure in Figure 5 to be consistent.
Response 5:Thank you for your helpful comment. In the revised manuscript, we have adjusted the dimensions of all sub-figures in Figure 5 to ensure uniform size and visual consistency. The updated figure now maintains a consistent layout for improved readability and presentation quality.

Comments 6:6.Lines 341 - 344: Please discuss the reliability of the averaged importance evaluation method. You can also cite other similar evaluation methods or use the results obtained from the study to discuss its reliability.
Response 6:We thank the reviewer for raising the important issue of reliability in variable importance estimation. In the revised text, we have elaborated on the rationale for using averaged importance values, highlighting how differences among methods reflect distinct assumptions and modeling capacities. We further explained that averaging helps reduce methodological bias and improve robustness, thus providing a scientifically sound approach for identifying key growth drivers.

Comments 7:7.The overall discussion in "4.2 Influencing variables of BAI in Pinus massoniana" lacks an ecological theoretical perspective. Please appropriately combine ecological theories to explain the observed phenomena or results.
Response 7:We appreciate the reviewer’s insightful suggestion. In the revised version of Section 4.2, we have incorporated relevant ecological theories to enhance the interpretation of our results. These additions provide a deeper ecological context for our findings and help elucidate the mechanisms underlying the observed patterns in BAI variation of P. massoniana.

Comments 8:8.The paper notes that climate variables significantly impact the radial growth of Masson pine and uses generalized additive models (GAM) to reveal the nonlinear relationships of these variables. However, the ecological significance behind these nonlinear relationships is somewhat under - discussed. It is suggested to conduct an in - depth analysis of the ecological mechanisms of these nonlinear responses. For instance, regarding the nonlinear responses to temperature and precipitation, you can combine the physiological characteristics of Masson pine to explore its physiological adaptation mechanisms under different temperature and precipitation conditions, as well as the potential responses of such nonlinear relationships to future climate change.
Response 8:We sincerely thank the reviewer for the insightful suggestion. We fully agree that exploring the underlying ecological mechanisms is essential for interpreting the observed nonlinear responses of Masson pine to climate variables. In the revised version, we have expanded Section 4.2 by integrating ecological and physiological interpretations to explain the nonlinear patterns of radial growth in response to temperature and precipitation. We elaborated on the drought-tolerant nature of P. massoniana and its conservative water-use strategy, which explains its weak response to precipitation variability in the study area. These additions enhance the ecological relevance of our findings and strengthen the interpretation of climate–growth interactions.

Comments 9:9.The research conclusions of the paper are primarily based on data from a specific location (Jigong Mountain Nature Reserve). Although this site is representative, the generalizability of the conclusions may be limited. It is recommended to add a detailed discussion on the scope of applicability of the research conclusions in the discussion section. You can combine research findings from other regions on Masson pine to analyze the similarities and differences in the responses of Masson pine growth to environmental factors across different regions, thereby more accurately assessing the applicability of the conclusions of this study in a broader area.
Response 9:We sincerely thank the reviewer for this insightful suggestion. We fully agree that discussing the broader applicability of our findings is essential, given the site-specific nature of the study. In response, we have added a paragraph at the end of the Discussion section that compares our results with findings from other regions where P. massoniana has been studied. These comparisons illustrate that while certain growth drivers—particularly temperature—are consistently important, the sensitivity to precipitation varies according to regional climate regimes. Based on this synthesis, we clarify that our conclusions are most applicable to regions with climatic and ecological conditions similar to those of Jigong Mountain, particularly in the transitional belt from subtropical to warm-temperate zones of central China.

Comments 10:10.Lines 478 - 481: This part is the focus of the Conclusion. It is necessary to specifically explain what kind of management strategies and ecological protection can be provided for forests based on the results of this study.
Response 10:We thank the reviewer for encouraging us to specify concrete conservation and management strategies. In the revised conclusion, we have added a sentence emphasizing the ecological significance of P. massoniana within the Jigong Mountain Nature Reserve and the value of using BAI to forecast long-term growth dynamics. This addition helps clarify how the findings may support conservation planning and ecological forecasting in protected areas.

Comments 11:11.The reference format is somewhat messy. Please standardize it according to the journal requirements, including the abbreviation of journal names, the italicization of Latin names in article titles, etc.
Response 11:Thank you for your careful observation. In the revised manuscript, we have thoroughly standardized the reference list according to the journal's formatting guidelines. This includes applying journal name abbreviations, italicizing Latin species names in article titles, and ensuring consistency in punctuation and structure.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors has responded actively to the first round of revision comments. The paper has been significantly improved after the first revision, yet some minor formatting issues still remain. The main points are as follows:

1.On Line 56, please confirm whether "oak" should be italicized.

2.All the sub - figures in Figure 4 should be combined into one complete figure instead of being spread across two pages.

3.Throughout the paper, the use of the letter "k" in terms of italics and case (upper or lower) should be unified.

Author Response

Comments 1:1.On Line 56, please confirm whether "oak" should be italicized.

Response 1:We appreciate the reviewer’s careful attention to terminology. In Line 56, the word "oak" is used as a common name rather than a scientific genus or species name. Therefore, according to scientific writing conventions, it does not require italicization. We have retained the word in regular font.

Comments 2:2.All the sub - figures in Figure 4 should be combined into one complete figure instead of being spread across two pages.

Response 2: Thank you for the suggestion. In the revised version, we have integrated all sub-figures in Figure 4 into a single composite figure to ensure a coherent layout and avoid splitting across two pages. This adjustment improves the clarity and readability of the figure.

Comments 3:3.Throughout the paper, the use of the letter "k" in terms of italics and case (upper or lower) should be unified.

Response 3: We appreciate the reviewer’s observation. We have carefully reviewed all instances of the letter "k" throughout the manuscript and revised them to ensure consistent formatting in terms of italics and case. Specifically, italicization is applied when "k" denotes a variable or parameter, and regular font is used otherwise, in accordance with standard scientific notation.