Biogenic Volatiles Emitted from Four Cold-Hardy Grape Cultivars During Ripening

Round  1

Reviewer 1 Report

Line 18 Justification for producing the data set was weak/inappropriate. Performance has not been linked to VOC’s. Research has shown that numerous variables can influence the VOC’s in wine, but there is only value to the data set if fresh VOC’s relate to wine VOC’s.

Line 26 Doesn’t make sense.

Line 37 Question starting a sentence with an abbreviation. Also need single quotes around cultivar names if these aren’t identified as cultivars in the sentence.

Line 48 Need single quotes for cultivar names.

Line 49 How do you harvest wines?

Line 52 Doesn’t make sense.

Line 52 Past tense should be used: were not are.

Line 62 Table included in the text does not explain duplicate or multiple sampled grapes identified with the same compound and why this number varies so much for the various compounds.

 Line 70 When were cultivars randomly placed in the vineyard?

Line 76 Past tense should be used: were not are.

Line 90 Past tense should be used: were not are.

Line 93 Don’t understand relevance of 3.2. Non-destructive sampling of biogenic volatiles as feasibility was never mentioned again. Feasibility is singular and should be was not were.

 Line 116 Reference to Figure 2 seems more appropriate to this sentence.

Line 122 Sentence doesn’t make sense: preconditioned to what?

Line 122 Past tense should be used: were not are.

Line 137 Supplier, city and state is still needed.

Line 160 Past tense should be used: was not is.

Line 164 Past tense should be used: was not is.

Author Response

Comments and Suggestions for Authors

Line 18: Justification for producing the data set was weak/inappropriate. Performance has not been linked to VOC’s. Research has shown that numerous variables can influence the VOC’s in wine, but there is only value to the data set if fresh VOC’s relate to wine VOC’s.

Authors’ response: We appreciate Reviewer’s comment about the link between in-vivo VOCs to wine aromas. The statement in Line 18 holds true as the performance of cold-hardy grapes is still little known compared with established vinifera grapes. Finding that link is a long-term goal. The justification for the data set is not limited to wine-making. We summarize, for the first time, VOCs emitted from growing grapes. Besides wine-making, VOC emissions from plants (biogenic emissions) in general are of interest to scientists working on atmospheric air quality, land use, plants’ interactions with ecosystem (e.g., response to environmental and biological stress). 

Line 26 Doesn’t make sense.

Authors’ response: We corrected grammar in this sentence.

Line 37 Question starting a sentence with an abbreviation. Also need single quotes around cultivar names if these aren’t identified as cultivars in the sentence.

Authors’ response: Edited to: “Volatile organic compounds (VOCs) emitted in-vivo from cold-hardy grape cultivars Frontenac, Marquette, St. Croix, and La Crescent grape berries from veraison to harvest were captured by two novel methods using SPME and characterized by GC-MS.”

Line 48 Need single quotes for cultivar names.

Authors’ response: Added single quotes around cultivar names.

Line 49 How do you harvest wines?

Authors’ response: Edited the sentence: “An automated SPME method for aroma analysis in wines has been described elsewhere [1], and used to characterize aromas in ‘Frontenac’ and ‘Marquette’ wines [2], and ‘Brianna’ and ‘Frontenac gris’ wines [3], and even concentrations of pigments and tannins in ‘Frontenac’ and ‘Marquette’ berries [4] harvested at different stages of berry ripening.”

Line 52 Doesn’t make sense.

Authors’ response: We corrected this sentence to “….and this dataset serves as the starting point to help growers to farm for flavor.”

Line 52 Past tense should be used: were not are.

Authors’ response: The data set (if published) will continue to ‘serve’.  Thus, the use of a present tense is more appropriate then past tense. 

Line 62 Table included in the text does not explain duplicate or multiple sampled grapes identified with the same compound and why this number varies so much for the various compounds.

Authors’ response: We added sentence right before Table 1: “Data analysis including multivariate analyses of the entire data set in Table S1 is provided elsewhere [7].” We also added new reference [7], i.e. a companion manuscript where the data analysis pertaining to Table S1 is provided.  We provided the information about soon-to-be submitted companion manuscript (now submitted and cited as [7] in the cover letter to the original submission. 

Line 70 When were cultivars randomly placed in the vineyard?

Authors’ response: Edited to: “Cultivars were randomly placed in both test vineyards in 2008 as part of the NE-1020 Cold Hardy Wine Grape Cultivar Trial.”

Line 76 Past tense should be used: were not are.

Authors’ response: Edited per Reviewer’s comment.

Line 90 Past tense should be used: were not are.

Authors’ response: Edited per reviewer’s comment.

Line 93 Don’t understand relevance of 3.2. Non-destructive sampling of biogenic volatiles as feasibility was never mentioned again. Feasibility is singular and should be was not were.

Authors’ response: We corrected the grammar to ‘was’. In addition, we added the results of preliminary tests to develop in-vivo method in Appendix A as suggested by Reviewer 3.

Line 116 Reference to Figure 2 seems more appropriate to this sentence.

Authors’ response: Moved reference to Figure 2 to the first sentence in section 3.4.

Line 122 Sentence doesn’t make sense: preconditioned to what?

Authors’ response: Edited sentence to: “These chambers with attached sampling ports for SPME insertion were preconditioned by thermal desorption before first use at 107 °C for 48 h.”

Line 122 Past tense should be used: were not are.

Authors’ response: Edited sentence to: “The PVF chambers were also cleaned prior to each sampling by rinsing in deionized water and oven baked overnight at 107 °C.”

Line 137 Supplier, city and state is still needed.

Authors’ response: Thank you!  Edited to: “Frozen berries were hand-crushed in the lab, placed into 20 mL amber screw top vials (P/N: 16-6000, Wheaton, Millville, NJ USA) with PTFE/silicone septa.”

Line 160 Past tense should be used: was not is.

Authors’ response: Edited sentence to: “Flow to the MS and sniff port was determined by fixed restrictor columns, 1 part to MS and 3 to sniff port.”

Line 164 Past tense should be used: was not is.

Authors’ response: Edited sentence to: “Carrier gas was ultra-high purity (UHP) helium (99.999%, Airgas, Des Moines, IA, USA).”

Reviewer 2 Report

The primary use for this material is as a future reference for those interested in wine composition for these varieties.  Headspace volatiles collected via the SPMI apparatus during ripening may be of interest to chemists and enologists, so this dataset will be useful to them.  A potential weakness is that many of the compounds that produce aromas and flavors in wine are products of fermentation, and their precursers (eg. bound terpenes) are odorless until after fermentation.  So samples from intact (or even crushed) berries might not offer much insight into wine flavor components - although it may indentify precursers. 

It may be beyond the scope of this paper, but a bit of interpretation of the data (eg. trends in appearance of specific compounds, relationship to other datasets or classes of flavor/aroma compounds from grapes) would be useful. This paper stops at methods, and makes no commentary about the findings and what they might mean - even the statement in the abstract about crushed vs intact berries (more VOCs from crushed) is not reiterated in a 'results' section.  How does this dataset fit within established literature on wine (or wine grape) aroma compounds (or classes of them?).  There must be a review article somewhere that would be relevant and citeable.  

Line 186,  instead of 'Special Crops', should read 'Specialty Crops'

Author Response

Comments and Suggestions for Authors

The primary use for this material is as a future reference for those interested in wine composition for these varieties.  Headspace volatiles collected via the SPMI apparatus during ripening may be of interest to chemists and enologists, so this dataset will be useful to them.  A potential weakness is that many of the compounds that produce aromas and flavors in wine are products of fermentation, and their precursers (eg. bound terpenes) are odorless until after fermentation.  So samples from intact (or even crushed) berries might not offer much insight into wine flavor components - although it may indentify precursers.

It may be beyond the scope of this paper, but a bit of interpretation of the data (eg. trends in appearance of specific compounds, relationship to other datasets or classes of flavor/aroma compounds from grapes) would be useful. This paper stops at methods, and makes no commentary about the findings and what they might mean - even the statement in the abstract about crushed vs intact berries (more VOCs from crushed) is not reiterated in a 'results' section.  How does this dataset fit within established literature on wine (or wine grape) aroma compounds (or classes of them?).  There must be a review article somewhere that would be relevant and citeable. 

Authors’ response: Excellent point. We added new reference [7], i.e. a companion manuscript where the data analysis pertaining to Table S1 is provided.  We provided the information about soon-to-be submitted companion manuscript (now submitted and cited as [7] in the cover letter to the original submission. We added sentence right before Table 1: “Data analysis including multivariate analyses of the entire data set in Table S1 is provided elsewhere [7].”

Line 186,  instead of 'Special Crops', should read 'Specialty Crops'

Authors’ response: Edited sentence to: “This research was funded by the United States Department of Agriculture’s Specialty Crops Research Initiative Program of the National Institute for Food and Agriculture, grant number 2011-51181-30850, titled “Northern grapes: integrating viticulture, winemaking, and marketing of new cold-hardy cultivars supporting new and growing rural wineries.”

Reviewer 3 Report

The manuscript of Rice et al. describes the volatiles analysis of 4 grape cultivars important in Midwestern US. Volatiles from in vivo grape skin and crushed grapes were collected using SPME and vacuum assisted SPME. The volatiles composition was characterized through GC-MS analysis. 

I generally like the idea of this paper and its novelty. Indeed, data about the analyzed grape cultivars is lacking in literature and the proposed sampling methods are very interesting. But some changes are necessary to make the paper more readable and scientifically robust. 
Just my impression: honestly, I would split the paper in 2 parts: the first, to be published in this journal, is the volatiles composition of the grape berries using the already published extraction methods (classic SPME extraction) followed by a simple PCA analysis. While a second paper with a better description of the vacuum assisted sampling and the bag assisted sampling should be described and published in a more chemistry oriented journal (Molecules or Metabolites from the same editor for example).

If you want to keep the paper in this format, I suggest you to follow these four major suggestions:
1) in the methods section, insert a table to resume the experimental design and the collected samples. This will improve paper readability.
2) Supplementary table 1 reports several compounds found in the same sample at the same retention time with the same name. If they are the same compound, their area should be summed and reported as a unique entity. Some compounds are reported with the same name but have different retention time; authors should evaluate whether the difference is due to an analytical issue or they are different compounds (for example Acetophenone has been found at 17 and 18 minutes in all the samples; difference of 1 minute in GC is very high; the same for Isovaleraldehyde, Caryophyllene and many others). Retention index should be used instead (if available). It looks like the difference is given by the year of analysis, but it is just my deduction. Authors should verify the origin of such RT differences and improve their identifications.
3) Statistical analysis of the data should be performed to demonstrate the reliability and replicability of the sampling methods. While the methods seem to be very interesting, their replicability should be verified. I would suggest a simple Principal Component Analysis and evaluation of compounds' relative standard deviation (RSD). The resulting scatter plot should highlight the overall replicability of the methods. Moreover, a comparison between methods would be appreciated.
4) lines 92-96: It seems like the authors performed already a method development for the extraction methods. This is very interesting and these results should be reported.

Minor comments:
lines 156-161: why did you split the flow to the olfattometry after GC separation?
lines 173-174: compounds identified through injection of the standards should be indicated in the supplementary table 1. Their identification is more reliable.

Author Response

Comments and Suggestions for Authors

The manuscript of Rice et al. describes the volatiles analysis of 4 grape cultivars important in Midwestern US. Volatiles from in vivo grape skin and crushed grapes were collected using SPME and vacuum assisted SPME. The volatiles composition was characterized through GC-MS analysis. 

I generally like the idea of this paper and its novelty. Indeed, data about the analyzed grape cultivars is lacking in literature and the proposed sampling methods are very interesting. But some changes are necessary to make the paper more readable and scientifically robust. 
Just my impression: honestly, I would split the paper in 2 parts: the first, to be published in this journal, is the volatiles composition of the grape berries using the already published extraction methods (classic SPME extraction) followed by a simple PCA analysis. While a second paper with a better description of the vacuum assisted sampling and the bag assisted sampling should be described and published in a more chemistry oriented journal (Molecules or Metabolites from the same editor for example).

If you want to keep the paper in this format, I suggest you to follow these four major suggestions:
1) in the methods section, insert a table to resume the experimental design and the collected samples. This will improve paper readability.

Authors’ response: Excellent point. We added new reference [7], i.e. a companion manuscript where the data analysis pertaining to Table S1 is provided.  We provided the information about soon-to-be submitted companion manuscript (now submitted to Molecules and cited as [7] in the cover letter to the original submission. We added sentence right before Table 1: “Data analysis including multivariate analyses of the entire data set in Table S1 is provided elsewhere [7].” In [7] we report the PCA analyses and the key aspects of in-vivo sampling with both vacuum-assisted extraction from a single berry and the whole cluster.

2) Supplementary table 1 reports several compounds found in the same sample at the same retention time with the same name. If they are the same compound, their area should be summed and reported as a unique entity. Some compounds are reported with the same name but have different retention time; authors should evaluate whether the difference is due to an analytical issue or they are different compounds (for example Acetophenone has been found at 17 and 18 minutes in all the samples; difference of 1 minute in GC is very high; the same for Isovaleraldehyde, Caryophyllene and many others). Retention index should be used instead (if available). It looks like the difference is given by the year of analysis, but it is just my deduction. Authors should verify the origin of such RT differences and improve their identifications.

Authors’ response: Good point. We added “Each row represents data for a single vine, i.e., three to four replicates were samples. Small variations in GC column retention time in Table S1 are typical with manual injection of SPME-based samples’ in the text commenting on the content of Table S1. We also added “The use of conventional retention indexes (e.g., Kovats retention index) is not appropriate for identification in this type of column configuration” in section 3.6 (data acquisition and analysis) to inform readers why conventional RT index would not be appropriate to use.

3) Statistical analysis of the data should be performed to demonstrate the reliability and replicability of the sampling methods. While the methods seem to be very interesting, their replicability should be verified. I would suggest a simple Principal Component Analysis and evaluation of compounds' relative standard deviation (RSD). The resulting scatter plot should highlight the overall replicability of the methods. Moreover, a comparison between methods would be appreciated.

Authors’ response: Excellent point. We added new reference [7], i.e. a companion manuscript where the data analysis pertaining to Table S1 is provided.  We provided the information about soon-to-be submitted companion manuscript (now submitted to Molecules and cited as [7] in the cover letter to the original submission. We added sentence right before Table 1: “Data analysis including multivariate analyses of the entire data set in Table S1 is provided elsewhere [7].” In [7] we report the PCA analyses and the key aspects of in-vivo sampling with both vacuum-assisted extraction from a single berry and the whole cluster.

4) lines 92-96: It seems like the authors performed already a method development for the extraction methods. This is very interesting and these results should be reported.

Authors’ response: Thank you for this comment.  We added the results of feasibility testing on Chilean grape cluster in three new figures in Appendix A (Figs. A1, A2, A3). We replaced ‘(data not shown)’ with ‘(Fig. A1, Fig. A2, Fig. A3). Based on the results of feasibility test, a 65 μm PDMS/DVB SPME fiber coating type and 30 min sampling time was selected as enabling collection of the largest number of VOCs, with the least variability in the least amount of time. Thus,..”

Minor comments:
lines 156-161: why did you split the flow to the olfattometry after GC separation?

Authors’ response: Section 3.6 (Data acquisition and analysis) has an explanation that ‘Olfactometry was not utilized in the research.’ The split is permanently installed and has been used in previous research of wine aroma, e.g., [1-3].

lines 173-174: compounds identified through injection of the standards should be indicated in the supplementary table 1. Their identification is more reliable.

Authors’ response: Good point.  In fact, compounds in Table S1 were verified with pure standards (i.e., by headspace sampling from a standard bottle using SPME and direct injection to GC-MS).  We edited the last line in section 3.6 to clarify this point. 

Round  2

Reviewer 1 Report

Line 53 keep the word point.

Reviewer 3 Report

I think that the changes to the paper are appropriate and it reached a good quality. I suggest to publish the manuscript in the present form.