UAV NDVI-Based Vigor Zoning Predicts PR-Protein Accumulation and Protein Instability in Chardonnay and Sauvignon Blanc Wines

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The present manuscript integrates UAV-based multispectral NDVI vigor zoning with RP-HPLC quantification of pathogenesis-related (PR) proteins (thaumatin-like proteins, TLPs, and chitinases), together with heat-test turbidity (ΔNTU) and bentonite fining requirements. The work spans two vintages (2023–2024) and two cultivars (Chardonnay and Sauvignon Blanc) in Casablanca Valley, Chile.

Overall, the manuscript is interesting and well written. Nevertheless, several points require clarification to improve readability and to better justify methodological choices.

1. Introduction

The Introduction is well structured, linking PR proteins (TLPs/chitinases) to vine vigor and stress. It cites foundational literature on protein haze formation and relevant precision viticulture research. However, the manuscript would benefit from a brief comparison of NDVI versus NDRE, to justify the choice of NDVI for vigor zoning (particularly considering NDVI saturation under high canopy density).

The primary objective—to establish a quantitative relationship between NDVI-derived vigor zones and PR-protein accumulation, and to evaluate implications for protein instability and bentonite demand—is achieved, with coherent and internally consistent results. However, given key design choices (e.g., pooled must per zone; technical versus biological replication; and possible fining prior to stability testing), the evidence is predominantly associational, which limits causal inference and may constrain generalizability. These design decisions should be explicitly described and justified.

2. Materials and methods

The methodology is generally adequate for an exploratory, vineyard-scale association study, but several issues require clarification to strengthen validity and reproducibility:

2.1. Vigor zoning inconsistency

There is an inconsistency in the zoning approach: k-means clustering is described in the Methods, whereas tertiles are reported in the Results. Please clarify which method was ultimately applied and ensure that all figures/tables consistently reflect the chosen approach.

2.2. Replication and independence (risk of pseudo-replication)

Pooling grapes from 30 vines per zone into a single must, followed by three fermentations, yields technical replicates of one pooled batch per zone rather than independent biological replicates. This can inflate the effective degrees of freedom and may overstate precision. A stronger design would include multiple independent pooled musts per zone (e.g., three separately pooled batches) or, alternatively, statistical models that reflect the hierarchical structure (e.g., mixed-effects models with vines nested within zones, zone as fixed effect, and year as random effect). At minimum, the current limitations should be clearly acknowledged.

2.3. Potential contradiction regarding fining

The Methods state that bentonite fining was performed prior to bottling. If so, ΔNTU and bentonite trials conducted afterward would reflect post-fining wines and may underestimate native instability and required doses. However, given the very high ΔNTU values reported for Sauvignon Blanc (e.g., 48–73 NTU in 2024), it appears likely that the wines were not pre-fined prior to heat testing. Please clarify the workflow and correct the description accordingly.

2.4. Heat-test protocol and external standards

The study uses a heat test of 2 h at 80 °C and a stability criterion of ΔNTU ≤ 2.0. This aligns with common winery protocols, but the manuscript should explicitly cite relevant external standards or industry references (e.g., AWRI/ETS/Laffort or equivalent) to support the methodological choice and threshold.

2.5. Confounders and covariates

Basic maturity and composition metrics (°Brix, pH, TA, GAE) are measured, but additional variables that can influence protein haze and bentonite demand—such as YAN (yeast assimilable nitrogen) and the use of proteases or alternative fining agents during vinification—are not reported. These factors should be included where possible or discussed as limitations.

3. Results and discussion

The multivariate analyses (LDA/PLS-DA) show separation among vigor classes. The reported 5-fold cross-validation error (~39–44%) suggests moderate classification performance, which is consistent with physiological overlap between zones. However, to reduce the risk of optimistic bias, the authors should consider reporting confusion matrices, per-class performance metrics, and permutation testing (or equivalent validation).

Overall, the objectives are largely achieved within an exploratory precision-enology framework.

4. Conclusions

The main conclusion, that UAV-based NDVI vigor zoning can help identify vineyard zones with higher risk of protein instability and support site-specific bentonite fining, appears supported by the data and aligns with winery practice regarding heat stability testing and bentonite trials. However, causal language (e.g., “drives”, “mechanistically linked”) should be avoided. Terms such as “predicts”, “is associated with”, or “identifies” are more appropriate.

Summary of key points to address

• Discuss NDVI saturation later in the season and the potential advantages of NDRE; if possible, include a brief comparison.
• Resolve the zoning inconsistency (k-means vs. tertiles) to ensure reproducibility.
• Clarify that fermentation replicates from a single pooled must are technical replicates; discuss implications for inference and pseudo-replication.
• Clarify whether wines were pre-fined before ΔNTU testing; correct the workflow description accordingly.
• Discuss potential confounding variables (e.g., YAN, canopy temperature, radiation exposure) and justify why they were not measured.

Minor writing and formatting issues

The manuscript should also be carefully proofread to correct typographical and formatting errors (e.g., spacing, punctuation, missing brackets, equation readability, and significant digits consistency), including:

• Line 171: missing space in “wine quality[30]”, and similar instances elsewhere.
• Lines 250–251: missing bracket in “Agisoft Metashape Professional (version 1.6.3, Agisoft LLC, St. Petersburg, Russia [31]”.
• Line 273: missing period after “improving vegetation segmentation accuracy”.
• Lines 281–282 (Table 1): MSAVI equation—square root formatting is not clearly visible.
• Line 343: missing period after “maturity and productivity at harvest”.
• Lines 561–562 (Table 2) and 623–624 (Table 3): ensure consistent significant digits.

Finally, improving figure quality would strengthen the manuscript.

Comments on the Quality of English Language

No comments on the quality of the English language are provided. Some recommendations are included within the suggestions.

Author Response

REVIEWER 1#

 

Comments General

The present manuscript integrates UAV-based multispectral NDVI vigor zoning with RP-HPLC quantification of pathogenesis-related (PR) proteins (thaumatin-like proteins, TLPs, and chitinases), together with heat-test turbidity (ΔNTU) and bentonite fining requirements. The work spans two vintages (2023–2024) and two cultivars (Chardonnay and Sauvignon Blanc) in Casablanca Valley, Chile.

Overall, the manuscript is interesting and well written. Nevertheless, several points require clarification to improve readability and to better justify methodological choices.

 Response General

Dear Reviewer,

We thank the reviewer for the positive assessment of the manuscript and for the constructive suggestions. In response, we have revised the manuscript to improve clarity and methodological transparency. Specifically, we clarified the vigor zoning approach (k-means clustering), justified the use of NDVI in comparison with NDRE, and explicitly described the replication structure, distinguishing technical from biological replicates. We also clarified the vinification workflow, acknowledged potential confounding factors (e.g., YAN and microclimatic variables), and revised the statistical and multivariate analyses to emphasize their exploratory and associative nature.

These revisions improve readability and ensure that the interpretations are consistent with the experimental design and scope of the study. We appreciate the reviewer’s comments, which have strengthened the manuscript.

Comment 1#

 

1. Introduction

The Introduction is well structured, linking PR proteins (TLPs/chitinases) to vine vigor and stress. It cites foundational literature on protein haze formation and relevant precision viticulture research. However, the manuscript would benefit from a brief comparison of NDVI versus NDRE, to justify the choice of NDVI for vigor zoning (particularly considering NDVI saturation under high canopy density).

The primary objective—to establish a quantitative relationship between NDVI-derived vigor zones and PR-protein accumulation, and to evaluate implications for protein instability and bentonite demand—is achieved, with coherent and internally consistent results. However, given key design choices (e.g., pooled must per zone; technical versus biological replication; and possible fining prior to stability testing), the evidence is predominantly associational, which limits causal inference and may constrain generalizability. These design decisions should be explicitly described and justified.

 

Response 1 #

Dear Reviewer,

Thank you for this constructive comment. We have revised the Introduction to include a brief comparison between NDVI and NDRE, explicitly acknowledging the potential saturation of NDVI under very dense canopies and discussing the conditions under which NDRE may offer higher sensitivity. We also justify the selection of NDVI based on its extensive validation in viticulture, its suitability for VSP-trained canopies, and its stronger and more consistent association with grape maturity (°Brix) observed in this study. These clarifications have been added to the Introduction (Lines 147–160).

In addition, the associational nature of the evidence is now explicitly stated, given the pooled must design and the distinction between biological and technical replication, thereby avoiding causal overinterpretation.

Comments 2#

2. Materials and methods

The methodology is generally adequate for an exploratory, vineyard-scale association study, but several issues require clarification to strengthen validity and reproducibility:

 

Response 2#

Dear Reviewer,

We thank the reviewer for this assessment. The Materials and Methods section has been revised to improve clarity, validity, and reproducibility in line with the exploratory vineyard-scale nature of the study. We now explicitly describe the vigor zoning approach, clarify the experimental unit and replication structure, distinguish technical from biological replicates, and detail the vinification and protein stability workflow. In addition, we have clarified the scope of the statistical analyses and acknowledged methodological limitations where appropriate.

These revisions ensure that the experimental design and analytical framework are clearly defined and reproducible.

 

Comment 3#

2.1. Vigor zoning inconsistency

There is an inconsistency in the zoning approach: k-means clustering is described in the Methods, whereas tertiles are reported in the Results. Please clarify which method was ultimately applied and ensure that all figures/tables consistently reflect the chosen approach.

 Response 3#

Dear Reviewer,

Thank you for this observation. We confirm that vigor classes were delineated using k-means clustering applied to vine-level NDVI data, performed independently for each cultivar and growing season.

References to tertile-based classification were the result of a transcription error when transferring the methodology into the manuscript and have now been removed.

The revised text (Lines 269–273) now consistently describes the k-means approach and clarifies that unequal class sizes reflect true spatial heterogeneity rather than methodological artifacts.

Comment 4 #

2.2. Replication and independence (risk of pseudo-replication)

Pooling grapes from 30 vines per zone into a single must, followed by three fermentations, yields technical replicates of one pooled batch per zone rather than independent biological replicates. This can inflate the effective degrees of freedom and may overstate precision. A stronger design would include multiple independent pooled musts per zone (e.g., three separately pooled batches) or, alternatively, statistical models that reflect the hierarchical structure (e.g., mixed-effects models with vines nested within zones, zone as fixed effect, and year as random effect). At minimum, the current limitations should be clearly acknowledged.

Response 4#

Dear Reviewer,

Thank you for this important comment. We agree that the pooled must, rather than the individual fermentations, represents the true biological experimental unit.

The manuscript has been revised to explicitly clarify that, for each cultivar, vintage, and vigor zone, grapes from 30 vines were pooled to generate a single composite must per zone, which constitutes the biological replicate. The three micro-vinifications conducted from each pooled must are now clearly defined as technical replicates intended to assess fermentation-related variability (Section 3.4.2, Lines 709–714).

The Statistical Analyses section has also been revised to reflect this structure and to emphasize that LDA and PLS-DA were used as exploratory tools to identify associative patterns rather than to support inferential or causal claims (Section 2.5.6, Lines 816–820; 829–831).

Comments 5#

 2.3. Potential contradiction regarding fining

The Methods state that bentonite fining was performed prior to bottling. If so, ΔNTU and bentonite trials conducted afterward would reflect post-fining wines and may underestimate native instability and required doses. However, given the very high ΔNTU values reported for Sauvignon Blanc (e.g., 48–73 NTU in 2024), it appears likely that the wines were not pre-fined prior to heat testing. Please clarify the workflow and correct the description accordingly.

 

Response 5#

Dear Reviewer,

Thank you for highlighting this point. We confirm that all ΔNTU measurements and bentonite dose trials were conducted on unfined wines. Bentonite was applied only after protein instability had been assessed and the minimum stabilizing dose determined. The Materials and Methods section has been revised to clarify this workflow (Section 2.5.5, lines 799–803).

Comments 6#

 2.4. Heat-test protocol and external standards

The study uses a heat test of 2 h at 80 °C and a stability criterion of ΔNTU ≤ 2.0. This aligns with common winery protocols, but the manuscript should explicitly cite relevant external standards or industry references (e.g., AWRI/ETS/Laffort or equivalent) to support the methodological choice and threshold.

 

Response 6#

Dear Reviewer,

Thank you for this valuable suggestion. The heat stability test applied in this study (2 h at 80 °C, ΔNTU ≤ 2.0) follows the validated protocol described by Salazar et al. (2012) for white wines from the Casablanca Valley and is consistent with routine winery practice.

In addition, this methodological choice is supported by comprehensive reviews on white wine protein instability, which identify the heat test as the most widely used and reliable approach for predicting haze formation and defining bentonite requirements under both research and industrial conditions (Cosme et al., 2020). These authors report that heating at 80 °C for 1–3 h is among the most commonly applied conditions and provides good correlation with natural protein precipitation.

The manuscript has been revised to explicitly cite Cosme et al. (2020) in the Methodology section and to include this reference in the reference list, together with Salazar et al. (2012). The rationale for the selected temperature, duration, and stability threshold is now clearly described in Section 3.5.4 (Lines 777–797), thereby strengthening methodological transparency and reproducibility.

Comments 7#

 2.5. Confounders and covariates

Basic maturity and composition metrics (°Brix, pH, TA, GAE) are measured, but additional variables that can influence protein haze and bentonite demand—such as YAN (yeast assimilable nitrogen) and the use of proteases or alternative fining agents during vinification—are not reported. These factors should be included where possible or discussed as limitations.

 Response 7#

Dear Reviewer,

Thank you for this valuable comment. Yeast assimilable nitrogen (YAN) was measured during vinification using the o-phthaldialdehyde assay (NOPA), following the methodology described by Salazar et al. (2017) (Eur Food Res Technol, 243:2043–2054, which was added in Material and methods and references).

Based on these measurements, diammonium phosphate (DAP) was added at a dose of 20 g/hL to all fermentations to ensure YAN levels above 200 mg/L. This nitrogen adjustment was applied uniformly across all vigor zones and cultivars, thereby avoiding differential nitrogen effects among treatments.

No proteolytic enzymes or alternative fining agents were used during vinification; protein stabilization relied exclusively on bentonite fining.

These clarifications have been incorporated into the Materials and Methods section (Lines 725–732), and the implications of nitrogen management for the interpretation of protein instability are now explicitly discussed (Lines 386–404).

Comments 8#

 Results and discussion

The multivariate analyses (LDA/PLS-DA) show separation among vigor classes. The reported 5-fold cross-validation error (~39–44%) suggests moderate classification performance, which is consistent with physiological overlap between zones. However, to reduce the risk of optimistic bias, the authors should consider reporting confusion matrices, per-class performance metrics, and permutation testing (or equivalent validation).

Overall, the objectives are largely achieved within an exploratory precision-enology framework.

Response 8# 

Dear Reviewer,

Thank you for this constructive comment. We agree that the classification performance of the multivariate analyses should be interpreted cautiously within an exploratory framework. Accordingly, we have clarified the exploratory purpose of the multivariate analyses in the Statistical analyses section, explicitly stating that LDA and PLS-DA were applied to evaluate separation patterns among vigor classes rather than to develop fully predictive classification models (Lines 816–820).

In addition, the cross-validation results are now explicitly reported and interpreted in the Results and Discussion. The 5-fold cross-validation error rates (38.9–44.4%) are discussed as expected for physiological classifications in commercial vineyards, where vigor gradients are continuous rather than discrete, resulting in partial class overlap and preventing unrealistically high classification accuracy (Lines 530–535). This interpretation directly addresses the risk of optimistic bias and places the observed performance within a realistic biological context.

Given the exploratory precision-enology objective of the study and the limited sample size, additional validation tools such as confusion matrices or permutation testing were not pursued. Nevertheless, the observed discrimination achieved using a limited set of biochemical predictors supports the robustness of PR proteins as indicators of vine stress intensity and enological instability.

Comments 9#

 Conclusions

The main conclusion, that UAV-based NDVI vigor zoning can help identify vineyard zones with higher risk of protein instability and support site-specific bentonite fining, appears supported by the data and aligns with winery practice regarding heat stability testing and bentonite trials. However, causal language (e.g., “drives”, “mechanistically linked”) should be avoided. Terms such as “predicts”, “is associated with”, or “identifies” are more appropriate.

 Response 9#

 Dear Reviewer,

We thank the reviewer for this important clarification. The Conclusions section has been carefully revised to remove causal language and to emphasize the associative and predictive nature of the observed relationships. Terms implying direct causality have been replaced with expressions such as “is associated with”, “identifies”, and “predicts”, in line with the exploratory design of the study. These modifications ensure that the conclusions accurately reflect the scope, limitations, and practical intent of the experimental approach. See the changes on Lines 851-853.

Comments 10#

 Summary of key points to address

• Discuss NDVI saturation later in the season and the potential advantages of NDRE; if possible, include a brief comparison.
• Resolve the zoning inconsistency (k-means vs. tertiles) to ensure reproducibility.
• Clarify that fermentation replicates from a single pooled must are technical replicates; discuss implications for inference and pseudo-replication.
• Clarify whether wines were pre-fined before ΔNTU testing; correct the workflow description accordingly.
• Discuss potential confounding variables (e.g., YAN, canopy temperature, radiation exposure) and justify why they were not measured.

 Response 10#

Dear Reviewer,

Thank you for this helpful summary. All the points raised have been addressed individually throughout the revised manuscript. Briefly: (i) the potential saturation of NDVI under dense canopies and the comparative advantages of NDRE are now discussed, including a brief NDVI–NDRE comparison; (ii) the vigor zoning procedure has been clarified to ensure reproducibility; (iii) the distinction between biological and technical replication in the microvinification design has been explicitly stated, together with its implications for inference and pseudo-replication; (iv) the experimental workflow has been corrected to clearly indicate that ΔNTU measurements were conducted on unfined wines; and (v) potential confounding variables (e.g., YAN and microclimatic factors) are now discussed, with justification for their exclusion under the uniform commercial management conditions of the vineyard.

 Comments 11#

 Minor writing and formatting issues

The manuscript should also be carefully proofread to correct typographical and formatting errors (e.g., spacing, punctuation, missing brackets, equation readability, and significant digits consistency), including:

Response 11#

 Dear Reviewer,

Thank you for pointing out these minor writing and formatting issues. The manuscript has been carefully proofread and revised to correct typographical and formatting errors, including spacing, punctuation, missing brackets, equation readability, and consistency of significant digits throughout the text, tables, and figures. The specific issues raised have been addressed as detailed below.

 Comments 12#

• Line 171: missing space in “wine quality[30]”, and similar instances elsewhere.

Response 12#

Dear Reviewer, the missing space in “wine quality[30]” has been corrected to “wine quality [30]”. Similar spacing issues have been reviewed and corrected throughout the manuscript. See revised Line 185.

Comments 13#

• Lines 250–251: missing bracket in “Agisoft Metashape Professional (version 1.6.3, Agisoft LLC, St. Petersburg, Russia [31]”.

Response 13#

Dear Reviewer, the missing closing bracket in “Agisoft Metashape Professional (version 1.6.3, Agisoft LLC, St. Petersburg, Russia [31]” has been added and the reference formatting has been corrected accordingly. See revised Lines 608-610

Comments 14#

• Line 273: missing period after “improving vegetation segmentation accuracy”.

Response 14#

Dear Reviewer, the missing period after “improving vegetation segmentation accuracy” has been added. See revised Line 632

 Comments 15#

• Lines 281–282 (Table 1): MSAVI equation—square root formatting is not clearly visible.

Response 15#

Dear Reviewer, the MSAVI equation in Table 4 has been reformatted to improve readability, and the square root notation is now clearly visible. See revised Lines 640-641

 Comments 16#

• Line 343: missing period after “maturity and productivity at harvest”.

Response 16#

Dear Reviewer, the missing period after “maturity and productivity at harvest” has been added. See revised Line 702

 Comments 17#

• Lines 561–562 (Table 2) and 623–624 (Table 3): ensure consistent significant digits.

Response 17#

Dear Reviewer, significant digits have been standardized in Tables 1 and 2 to ensure numerical consistency throughout the manuscript. See revised Lines 335–336 (Table 1) and Lines 416–417(Table 2).

Comments 18#

Finally, improving figure quality would strengthen the manuscript.

Response 18#

 Dear reviewer, the quality and resolution of all figures have been improved to enhance clarity and readability.

 

 

Reviewer 2 Report

Comments and Suggestions for Authors

This study established a quantitative link between vineyard vigor and protein instability in white wine by using UAV multispectral imagery and NDVI-based vigor zoning. However, the paper suffers from several shortcomings in methodological detail, statistical analysis, and discussion depth that must be addressed to improve scientific rigor.

The micro-vinification scale (2 L) is too small and may fail to fully replicate commercial winemaking conditions, limiting the external validity of the results.

The bentonite dosage was determined solely with the threshold ΔNTU ≤ 2; the justification for this criterion and its alignment with industry standards are not provided.

The possible presence of spatial autocorrelation (e.g., mutual influence among neighboring vines) was not examined.

Misclassified cases should be analyzed for shared traits (soil type, microclimate) to delineate the model’s applicability boundary.

No explanation is offered for why Sauvignon Blanc responds to vigor changes more markedly than Chardonnay (e.g., cultivar-specific traits, differences in rooting depth).

The study lacks direct analysis linking the findings to established PR-protein regulatory mechanisms such as ABA or ethylene pathways.

PR-protein peaks in the chromatograms were identified only by retention-time matching without mass-spectrometric confirmation.

The text alternates between “vigor” and “vigour”; choose either American or British spelling and use it consistently.

Author Response

Comments General

This study established a quantitative link between vineyard vigor and protein instability in white wine by using UAV multispectral imagery and NDVI-based vigor zoning. However, the paper suffers from several shortcomings in methodological detail, statistical analysis, and discussion depth that must be addressed to improve scientific rigor.

 

Response General

Dear Reviewer,

We sincerely thank you for the time and effort devoted to evaluating our manuscript and for the constructive comments aimed at strengthening its scientific rigor. We have carefully considered each point raised and revised the manuscript accordingly to improve methodological clarity, analytical transparency, and interpretation. Our detailed responses are provided below.

 

Comment1#

The micro-vinification scale (2 L) is too small and may fail to fully replicate commercial winemaking conditions, limiting the external validity of the results.

 

Response 1#

 

Dear Reviewer,

We acknowledge this limitation. The 2 L micro-vinification scale was intentionally selected to enable controlled and replicated fermentations under standardized cellar conditions, while preserving variability derived from vineyard vigor. This scale is commonly used in enological research when the objective is to explore vineyard-driven effects rather than to reproduce full commercial winemaking processes. Accordingly, the manuscript now more clearly frames the study as an exploratory, vineyard-scale association analysis aimed at risk assessment, rather than at simulating industrial cellar kinetics. This clarification has been reinforced in the Materials and Methods and Discussion sections.

 

Comments 2#

The bentonite dosage was determined solely with the threshold ΔNTU ≤ 2; the justification for this criterion and its alignment with industry standards are not provided.

 

Response 2#

Dear Reviewer,

We have now clarified and justified the ΔNTU ≤ 2 threshold in Section 2.5.4. The heat stability test (2 h at 80 °C) and the selected turbidity threshold follow the protocol validated by Salazar et al. (2012) and Cosme et. al (2020) and are consistent with widely adopted industry and research standards. The relevant reference has been explicitly cited to ensure methodological transparency (new lines 790–797).

 

Comments 3#

The possible presence of spatial autocorrelation (e.g., mutual influence among neighboring vines) was not examined.

 

 

Response 3#

Dear Reviewer,

We recognize that spatial autocorrelation may influence vine-level observations. However, the analytical focus of this study is on NDVI-defined vigor zones, not on individual vine inference. By aggregating vines into spatially coherent vigor zones, the analysis intentionally captures integrated physiological patterns rather than fine-scale spatial dependence. This scope and its implications are now clarified in the Discussion

 

Comments 4#

Misclassified cases should be analyzed for shared traits (soil type, microclimate) to delineate the model’s applicability boundary.

 

Response 4#

Dear Reviewer,

We acknowledge that further analysis of misclassified samples (e.g., soil or microclimatic traits) could provide additional insight into model boundaries. Given the exploratory objective of identifying zone-level biochemical patterns rather than optimizing classification accuracy, this analysis was beyond the scope of the present study. This limitation, together with its relevance for future research, is now explicitly acknowledged in the revised manuscript.

 

Comments 5#

No explanation is offered for why Sauvignon Blanc responds to vigor changes more markedly than Chardonnay (e.g., cultivar-specific traits, differences in rooting depth).

 

Response 5#

The stronger response observed in Sauvignon Blanc is now discussed in the manuscript. We note that Sauvignon Blanc is known to exhibit higher PR-protein accumulation and greater sensitivity to abiotic stress compared with Chardonnay, which may contribute to the more pronounced vigor-related gradients observed. This cultivar-specific behavior has been incorporated into the Discussion to better contextualize the results.

These modifications are included in Section 3.3, lines 421–455

 

 

Comments 6#

The study lacks direct analysis linking the findings to established PR-protein regulatory mechanisms such as ABA or ethylene pathways.

 

Response 6#

Dear Reviewer,

We acknowledge this limitation. No direct molecular or hormonal analyses (e.g., ABA or ethylene measurements) were performed in this study. The experimental design was intentionally focused on vineyard-scale physiological variability and enological outcomes, rather than on elucidating PR-protein regulation at the molecular or signaling-pathway level.

Accordingly, the manuscript does not claim mechanistic evidence linking NDVI-derived vigor differences to specific hormonal pathways. Any physiological interpretation is framed strictly in the context of previously established literature and is presented as contextual rather than demonstrative. We believe this clarification appropriately aligns the interpretation with the scope and scale of the study.

 

Comments 7#

PR-protein peaks in the chromatograms were identified only by retention-time matching without mass-spectrometric confirmation.

 

Response 7#

Dear Reviewer,

PR-protein peaks were identified based on retention time and chromatographic behavior consistent with extensively validated RP-HPLC protocols for wine PR proteins. While mass spectrometric confirmation would provide additional molecular resolution, the applied approach is standard in enological protein studies and adequate for comparative, zone-level analysis. This methodological choice and its limitations are now explicitly stated.

 

Comments 8#

The text alternates between “vigor” and “vigour”; choose either American or British spelling and use it consistently.

 

Response 8#

Dear Reviewer,

Thank you for noting this inconsistency. The manuscript has been fully standardized to American English spelling (“vigor”) throughout.

 

Reviewer 3 Report

Comments and Suggestions for Authors

The experimental work considered two white grape cultivars, in two years and two vineyards. The vineyards were mapped for vigor using unmanned aerial vehicle (UAV), 19 multispectral imagery, and NDVI-based vigor indices. The vigor indices were successfully correlated with the presence of PR proteins in musts and wines.

The work was properly designed and conducted, and clearly presented. The conclusions are correct and robust.

Author Response

Comments 1#

The experimental work considered two white grape cultivars, in two years and two vineyards. The vineyards were mapped for vigor using unmanned aerial vehicle (UAV), 19 multispectral imagery, and NDVI-based vigor indices. The vigor indices were successfully correlated with the presence of PR proteins in musts and wines.

The work was properly designed and conducted, and clearly presented. The conclusions are correct and robust.

 

Response 1#

Dear Reviewer, we sincerely thank you for your positive evaluation and for recognizing the merits of our experimental design, methodology, and presentation. Your acknowledgment that the work was properly conducted and that the conclusions are robust is greatly appreciated.