Networking 13 Berry Minerals to Sustain a High Yield of Firm Cranberry Fruits
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThis study introduces a novel weighted log ratio approach for cranberry mineral analysis, which represents a meaningful advancement in understanding nutrient interactions. The extensive field data collection and machine learning application show promise for practical agricultural applications.
Experimental design description incomplete: Key details are missing regarding sample collection protocols, randomization methods, and analytical quality control. The selection criteria for the 393 observations require clarification. How were potential confounding factors like soil variability and seasonal effects controlled?
Mathematical framework needs better contextualization: While the wlr methodology is mathematically sound, readers would benefit from clearer explanation of why compositional data analysis is necessary for this problem. The connection between statistical transformations and biological processes could be strengthened.
Results lack practical interpretation: Statistical significance doesn't always translate to field relevance. What do the observed wlr differences mean for actual grower recommendations? The cutoff values (42.615 Mg ha⁻¹ for yield, 6.132 N mm⁻¹ for firmness) need agronomic justification.
Writing quality impacts clarity: Multiple grammatical issues and awkward phrasing throughout the text hamper understanding. Professional editing would improve accessibility. Some technical terms are used inconsistently (e.g., "berry firmness" vs "fruit firmness").
Author Response
This study introduces a novel weighted log ratio approach for cranberry mineral analysis, which represents a meaningful advancement in understanding nutrient interactions. The extensive field data collection and machine learning application show promise for practical agricultural applications.
Reply
Indeed, freeing each element from noise caused by interactions with other elements has been a challenge.
Experimental design description incomplete: Key details are missing regarding sample collection protocols, randomization methods, and analytical quality control. The selection criteria for the 393 observations require clarification. How were potential confounding factors like soil variability and seasonal effects controlled?
Reply
See l. 82-95 in the revised version.
Mathematical framework needs better contextualization: While the wlr methodology is mathematically sound, readers would benefit from clearer explanation of why compositional data analysis is necessary for this problem.
Reply
See l. 110-113 in the revised version: The sample space for the D-parts composition under study was defined as follows:
See l. 123-132 for more explanations concerning wlr.
The connection between statistical transformations and biological processes could be strengthened.
Reply
See l. 62-70 in the revised version.
Results lack practical interpretation: Statistical significance doesn't always translate to field relevance. What do the observed wlr differences mean for actual grower recommendations? The cutoff values (42.615 Mg ha⁻¹ for yield, 6.132 N mm⁻¹ for firmness) need agronomic justification.
Reply
- 146-152 in the revised version for the realism of cutoff values.
I agree that statistical results is a proximate diagnosis and theoretical recommendations although they are supported by literature. There is indeed no recommendation possible from that exploratory research. Only corrective measures are suggested (l. 289-305). Specific recommendations will require conducting gypsum and silicate trials and calibrate soil and tissue tests against crop performance. There are limitations in this study (l. 306-311).
Writing quality impacts clarity: Multiple grammatical issues and awkward phrasing throughout the text hamper understanding. Professional editing would improve accessibility. Some technical terms are used inconsistently (e.g., "berry firmness" vs "fruit firmness").
Reply
I made several editorial corrections. Fruit and berry are interchangeable terms. I also had to reinforce the introduction by moving text from the Discussion section to the Introduction section as required by another reviewer. I added some references on plant breeding for berry firmness. I think the text is more balanced and fluid now.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe author presents an interesting study to achieve high yields of firm berries by addressing the elemental network. The novelty of this study lies in the use of compositional data analysis to understand how mineral nutrition affects berry yield and firmness, particularly in the use of weighted log ratios. This approach allows the author to propose other nutritional corrections in crop management to ensure the objective, rather than using the most common correction, N fertilization, which results in high proportions of soft berries. Although, the author provided very valuable information, I miss important information on methodology, on results and discussion. Nevertheless, I have identified some important issues that require the author' attention before publication in the Horticulturae Journal:
- Berries collection. It is not clear on the methodology where observations came from, and what means observations? Observations are plots? Replicates are plot-level or quadrat-level? What year was the harvest?
- The relationship between berry elemental composition and berry yield and firmness is not well supported on the introduction. Notably, I think that the cause-effect of these relationships are not clearly demonstrated, while there are supposed to be observations from multiple sites. There could be confounding factors that could alter the results and that are not included either in the experimental design or in the subsequent statistical analyses.
- Dual ratios used. The author proposes to use a dual logarithmic ratio, recently published by the author. However, there are some unresolved issues in this paper. How did the author calculate the coefficient 𝜑 ? And how was it decided? I miss information that tells us how it led to the choice of the coefficient. What is the biological significance of this coefficient 𝜑? I may err on the side of ignorance, and I think it is because of lack of detail in the introduction and clarification of why. But what do we need to correct the clr transformation, and what it brings apart from better model accuracy (which has no biological significance)?
- Some results are not well described or demonstrated.
- Line 158-162. These results do not seem to correspond to Table 1. How can the author relate high levels of an element with high yield or high firmness? The author needs to add statistical information on these relationships. Table 1 is only a summary of the minimum, maximum, and median of each element, berry firmness, and berry yield. Standard errors should be added to each value, depending on the level of repetition.
- Gain ratios. Author should add standard errors to each value, according to the level of replication.
- Figure 5 is the key to the objective of this article, but it is not clear. Which gain ratios are compared? All together? What is the purpose of mixing all gain ratios? Would it be better to separate between elements, between log ratio types, etc.?
Author Response
The author presents an interesting study to achieve high yields of firm berries by addressing the elemental network. The novelty of this study lies in the use of compositional data analysis to understand how mineral nutrition affects berry yield and firmness, particularly in the use of weighted log ratios. This approach allows the author to propose other nutritional corrections in crop management to ensure the objective, rather than using the most common correction, N fertilization, which results in high proportions of soft berries. Although, the author provided very valuable information, I miss important information on methodology, on results and discussion. Nevertheless, I have identified some important issues that require the author' attention before publication in the Horticulturae Journal:
- Berries collection. It is not clear on the methodology where observations came from, and what means observations? Observations are plots? Replicates are plot-level or quadrat-level? What year was the harvest?
Reply
See l. 82-95 in the revised version.
- The relationship between berry elemental composition and berry yield and firmness is not well supported on the introduction. Notably, I think that the cause-effect of these relationships are not clearly demonstrated, while there are supposed to be observations from multiple sites. There could be confounding factors that could alter the results and that are not included either in the experimental design or in the subsequent statistical analyses.
Reply
Introduction was rewritten (l. 24-75) after moving text from the Discussion section to the Introduction section and making editorial adjustments. I added some references on plant breeding for berry firmness. I think the text is more balanced and fluid now. A cause-effect relationship cannot be demonstrated, just statistical results on berry and firmness classes associated with weighted “balances” among elements. There were imposed gradients of N, P, K, Mg, Cu and B. Variations in Al and Si supply depended essentially on soil conditions. Nevertheless, machine learning classification models were quite accurate using wlr variables only.
- Dual ratios used. The author proposes to use a dual logarithmic ratio, recently published by the author. However, there are some unresolved issues in this paper. How did the author calculate the coefficient ? ? And how was it decided? I miss information that tells us how it led to the choice of the coefficient. What is the biological significance of this coefficient ?? I may err on the side of ignorance, and I think it is because of lack of detail in the introduction and clarification of why. But what do we need to correct the clr transformation, and what it brings apart from better model accuracy (which has no biological significance)?
Reply
See l. 123-132. The gain ratio measures the effectiveness of an attribute to classify data in decision-tree binary machine learning classification models. At the agronomic level, gain ratio assigns a coefficient of importance to an attribute regarding the split between high- and low-performing crops and derive nutrient standards. Gain ratio resembles the gini coefficient. The higher the gain ratio, the more important is the dual log ratio regarding the target variable in an agronomic database. Interactions (dual log ratios) are weighted differentially rather than giving them the same importance. Gain ratio depends on the database and the cutoff selected for the target variable for classification purposes. Dual ratios are meaningful physiologically while gain ratios are meaningful agronomically at the step of deriving nutrient standards for diagnostic purposes. The gain ratio is a means to alleviate the noise caused by interactions among nutrients.
- Some results are not well described or demonstrated.
- Line 158-162. These results do not seem to correspond to Table 1. How can the author relate high levels of an element with high yield or high firmness? The author needs to add statistical information on these relationships. Table 1 is only a summary of the minimum, maximum, and median of each element, berry firmness, and berry yield. Standard errors should be added to each value, depending on the level of repetition.
Reply
- 158-162 deleted (the importance of Al and Si is shown in Figs 3-4-6).
- Gain ratios. Author should add standard errors to each value, according to the level of replication.
Reply
Gain ratio is a metric measuring the effectiveness of the selected split in classification models. There is no associated standard error.
- Figure 5 is the key to the objective of this article, but it is not clear. Which gain ratios are compared? All together? What is the purpose of mixing all gain ratios? Would it be better to separate between elements, between log ratio types, etc.?
Reply
The gain ratios assigned to each dual log ratio and related to berry yield differed from those assigned to berry firmness as shown in the scatter diagram. This only demonstrates the differential importance of each dual ratio regarding the target variable. Gain ratios by element are shown in Figures 3 and 4.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsMinor revision should be undertaken to address terminological consistency, formatting issues, and mathematical expressions before publication.
Terminological Consistency:
Line 11: Change "XGBoosting" to "XGBoost" (standard terminology)
Line 9: Replace "elemental network" with "optimal mineral composition" or "mineral nutrient balance" for clarity.
Lines 128-132: The wlr formula formatting needs improvement: "i ≠ j" or "i=j" ?
Figure References:
"Figure 3" and "Figure 4" captions could be more descriptive
Language and Grammar Issues:
"accounted for interactions" → "accounting for nutrient interactions"
"Fruit firmness is affected by cell wall structure, turgor, cuticle properties and biochemical constitution is of utmost importance" - run-on sentence needs restructuring
These corrections will significantly improve the manuscript's clarity and professional presentation while maintaining the scientific rigor of the content.
Author Response
Minor revision should be undertaken to address terminological consistency, formatting issues, and mathematical expressions before publication.
Terminological Consistency:
Line 11: Change "XGBoosting" to "XGBoost" (standard terminology)
Done
Line 9: Replace "elemental network" with "optimal mineral composition" or "mineral nutrient balance" for clarity.
Done
Lines 128-132: The wlr formula formatting needs improvement: "i ≠ j" or "i=j" ?
Changed to:
Figure References:
"Figure 3" and "Figure 4" captions could be more descriptive
Figure 3. Gain ratios of elemental expressions regarding berry yield. Raw concentrations that do not consider nutrient interactions return distorted nutrient ranking. The clr and wlr expressions account for nutrient interactions as dual log ratios. The wlr expression also accounts for the importance of each dual ratio as gain ratio, modifying nutrient ranking compared to clr. The Cu ranked first for clr and wlr.
Figure 4. Gain ratios of elemental expressions regarding berry firmness. Raw concentrations that do not consider nutrient interactions return distorted nutrient ranking. The clr and wlr expressions account for nutrient interactions as dual log ratios. The wlr expression also accounts for the importance of each dual ratio as gain ratio, modifying nutrient ranking compared to clr. The Si ranked first for the three expressions.
Language and Grammar Issues:
"accounted for interactions" → "accounting for nutrient interactions"
Changed.
"Fruit firmness is affected by cell wall structure, turgor, cuticle properties and biochemical constitution is of utmost importance" - run-on sentence needs restructuring
Changed to: “Cell wall structure, turgor, cuticle properties and biochemical constitution are of utmost importance for fruit firmness, resistance to pests and diseases, handling, storage and shipping”
These corrections will significantly improve the manuscript's clarity and professional presentation while maintaining the scientific rigor of the content.
Reviewer 2 Report
Comments and Suggestions for AuthorsNo comments
Author Response
Thank you for reviewing.