Exogenous Gibberellic Acid Advances Reproductive Phenology and Increases Early-Season Yield in Subtropical Blackberry Production

Round 1

Reviewer 1 Report

Overall, I felt this paper clearly reflected sound science, and it was generally well-described. I also think the paper will be of significant interest--it was to me. I  think there are a number of areas where the clarity of the article could be improved or the story tightened, and also a few places where the authors could unpack some additional possibilities for the reader to consider (within bounds). In particular, I think a greater weight to alternative interpretations in the Discussion might be warranted, so long as it does not expand too far beyond the scope of the study's inquiry. 

Specific comments (line numbers listed if applicable):

1. Some of the material in the Intro feels like it could be moved to the Methods, as it is background biology on the study species, but this could be my personal preference. As an example, info on primo- vs. florocane reproduction (lines 38-55). 
2. 59-61: The past tense here feels odd, even though you are describing results from specific past studies. Maybe introduce the studies in basic terms, or else generalize (if appropriate) and bring into present tense.
3. 70: One or more words seems to be missing here.
4. 56-72: GA3 applications to try to stimulate bud break and also produce parthenocarpic fruits when pollination success has been low have been explored in several commercial blueberries. The results have been quite mixed--see for example https://pdfs.semanticscholar.org/fa79/29475046b771fd182d4a74e687a53a788181.pdf. I tried using a GA3 treatment and ended up scorching my plants to death with it. A cursory lit search will turn up a number of other papers. I think readers may find discussion of this body of literature interesting, and it may allow you to more thoroughly discuss reasons why this approach may work better in your context than in others.
5. 80: Grammar is a little off here. Rephrase.
6. 81-88: This paragraph is very confusing. It is talking about primocane varieties and their challenges, but your study focuses on floricanes, so the material in this paragraph feels irrelevant. Maybe carve off most of this into another stand-alone body paragraph that fleshes out pluses/minus of the different varieties and why your study focuses on floricane ones. Then, the last paragraph can state objectives plus hypotheses and predictions (which are currently omitted). 
7. 107: In re: management practices--such as? Are these those described above this line?
8. 112: How was this amount determined? Is it the recommended concentration from the manufacturer? I'd be curious to know why a range of applications weren't considered--sample size considerations? Logistical limitations? Was this just a first-try sort of study?
9. 123: "Fruit set" has been very inconsistently used in the literature. IMO, it should refer only to the ratio of fruits to flowers, not to raw fruit number, which should be referred to simply as such. I think you mean fruit number here (or, if you prefer, "fruits set"). I appreciate that you clarified what you meant here though. 
10. 124: Why the largest ones? Also, by what measurement? Volume? Mass? Length? Why not just test a sample of marketable fruits blended together?
11. 128-135: I'm having a hard time following this. My read on this is that you had 4 plots per cultivar = 12 plots total, each of 6 plants (n=72 plants). Then, 3 of each plot got one GA treatment and the other 3 got the other. The random effects were then the plot (since plants in the same plot are psuedoreplicates) and also plot x cultivar (which assumes a given cultivar might respond differently to each plot's unique conditions, and this effect would be "noise"). Do I have this correct? If not, a schematic or clarification may help. My concern here is that generally the rule of thumb is that, to estimate variances for random effects effectively, it's helpful to have 5+ data points per "group." 6 plants per plot is fine, but just 3 plots per cultivar is iffy. That random factor may not be as powerful as a result. However, given that it seems like that random factor is present more to be "safe rather than sorry" I would say it might be ok. Still, the authors should check to see if their conclusions change much if this random factor is removed. 
12. 135-148: Very nice! Clean and clear, and all seems appropriate. Any model assumptions checked (these are admittedly hard for non-linear mixed models, but nonetheless)? Or any attempt to consider potentially leveraging points?
13. 159: Wow, out there on Christmas Eve! Now that's dedication to science. 
14. 161-167: Much of this is in the Table. Journals vary in the amount of repetition they want/permit between text and tables. Personally, I think the dates that are redundantly presented could be removed but the derived data (e.g. days after budbreak) should remain in the text. This applies to the rest of the paragraphs in the Results--see if you can thin the repetition between text and tables.
15. 179: How is the date of first harvest determined? This should be mentioned in the methods (unless I just missed it). Is there an objective rule of thumb or is it more subjective?
16. 185: "Early" vs. "late" season should be clarified somewhere in the Methods. When are these periods, and what sort of biophysical correlate do they have? Or are they more ad hoc harvesting terms?
17. 197: I am not personally a stickler for the distinction between alphas of 0.05 and 0.1, but I don't think we should be subtle about it either. In the methods, you note "significant" is less that 0.05, but Table 2 does not follow this rule. Personally, I think that's ok--I can decide for myself as a reader how I feel about these results using a less conservative alpha. But some may find it deceptive. I would formally add a "marginal significance" category and indicate which p values fall into which zone in this table, or else state in the Methods that you will be using alpha = 0.1 in some tests but 0.05 in others.
18. 202-213: I am not sure that this adds to your story. I am not surprised to learn that fruit number and yield are positively correlated. Since you never formally stated your hypotheses, I can't say if you went in with this being an open question or not. Consider justifying presenting these results more clearly or else move to the Supplemental Materials. You revisit this idea in lines 241-242. Here it seems the message is that berry #, not berry weight, is what primarily drives yield. If that's the intended read, copying language from 241-242 to 202-213 might be helpful. But, nonetheless, your results are long and sometimes feel repetitive and you may consider tightening your story in these areas. 
19. 216: It's common practice to remove non-significant higher-order terms (whether or not it ought to be). It doesn't look like you did this here...any particular reason why this choice was made?
20. 225: What criteria are used to determine marketability? I assume a citation somewhere speaks to this but it could be highlighted for the reader's convenience. Also, letters are missing in the late column for the GA3 treatment. 
21. 307: I think this section tells an interesting story, but I don't personally feel it's more interesting than the GA3 story. I would consider reversing the order of these first two sections in the discussion to lead with the punchier story.
22. 316: Feels like a source speaking to typical temperate phenologies would be appropriate here. 
23. 328-329: Grammar needs a little work here. 
24. 335: A stronger topic sentence here would be good--what is the main point of this paragraph? Feels too general. 
25. 349-358: A very nice paragraph. 
26. 357-358: Are there pictures we could see to judge for ourselves? Maybe at least include in the SM if so. 
27. 361: Why would advanced onset of flowering make a difference? Better pollination? More time for seed/fruit development? Feels like incomplete logic. 
28. 366-367: If you saw fit, you could speculate more as to why. Could it just be that the one-time application has a residence time and the GA3 is broken down or becomes inactive over time, such that its effect becomes muted? In other words, if you applied a second time (or later), would the drop-off become less pronounced, do you think? I know you didn't look at this, specifically, but some cautious speculation feels warranted. Or speak to the need to study this?
29. 397-404: In blueberries, GA3 can actually damage floral tissues but, at the same time, stimulate some parthenocarpic fruit development. If I remember, the effects can be positive or negative based on timing and rate of application. Because some fruits may develop without (as many) seeds, there is a drop in fruit weight as a direct result of the loss of seed mass. Also, I think it would be likely the plant senses the lack of seeds and then invests less in the fruit as it develops also. Blueberry and blackberry are pretty different, but not less so than blackberry and cherry/peach--the relevant lit on blueberry might really offer some alternative explanations. 
30. 405-415. Is it slowing ripening, or is it maybe decoupling the processes of pigment development vs. sugar accumulation (and/or organic acid degradation)? Maybe speak to these possibilities unless there is reason to think they are less valid (which might also be interesting). Also, would a drop to 9% SSC be noticeable to end consumers? Is this likely to be a drop in sugars, organic acids, maybe dissolved starches, or something else/a combination?
31. 432-438: Good point!
32. 464: Paper looks to be reasonably and thoroughly sourced. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Lin and Agehara evaluated the effects of gibberellic acid application during dormancy on phenology and yield parameters of blackberry cultivated in a subtropical climate. The manuscript is well written and can provide useful results to the field. The methods are adequate in general although some details need to be better explained. Similar comments can be derived from the discussion section. I think the manuscript is suitable for publication after minor corrections are implemented. Below are my general and detailed comments on the manuscript.

Major/general comments

I think the abstract is lacking a bit of information about the methods. Mentioning the duration of the experiment, as well as the moment of application of GA₃ could help complement this section.

Methods

Mentioning the method used for determining the beginning and the end of the winter season is required when computing winter chill accumulation. I realized this is stated in the supplementary materials but it needs to be mentioned in the main manuscript.

How did you define the moment of application of GAs? Usually, dormancy braking agents are applied when a portion of chill requirement has been achieved. In the case of other dormancy breakers (i.e. HCN), this timing is crucial to have positive rather than negative or undesirable effects. Please explain how you decide that 25 December 2015 and 27 January 2017 are the dates for spraying GAs. Is the difference of about one month in the second season related to chill accumulation?

I think the section on phenology measurements needs to describe a bit more about the bud break stage. Was this done according to a global or internal scale? What does it mean ‘flowering’ for the authors? Adding reference pictures about the stage you consider as bud break of flowering could help readers understand.

What does the grading standards mean for the manuscript? I am not sure where this is shown in the results. If these grades are not shown in the results section, consider removing this statement. Otherwise, explain what these grading standards are.

The explanation/description for analyzing the parameters dates of bud break, flowering and harvest is missing. If, in the case this data was not statistically analyzed and only used as raw data, it could be mentioned.

Perhaps, the linear relationships established in figure 1 and figure 2, should be mentioned somewhere in the methods.

Results

When showing the results on phenology, I am not sure if it does make sense to present the data in terms of date. Plants do not realize about the dates, so I would suggest at least complementing with a more biological parameter such as chill and heat accumulation. Additionally, and related to methods, you show dates of bud break as for example 3/11/2016, however, it is not clear to me if this is grouped data or all plants you used as replications showed bud break on the same date.

The results showed in section 3.2, when you define early and late harvest contradicts the results of harvest date showed in section 3.1, where you show only one harvest date. This needs to be clarified somewhere.

I think the structure of the results is not very clear. In sections 3.2 and 3.3, you show the results on marketable yield and stablish correlations with average berry weight and fruit set. However, the results for these specific variables are shown in sections 3.4 and 3.5. Consider re-organizing the results section in a way that makes sense to see results and figures in the order they appear in the manuscript.

Minor/detailed comments

Consider using ‘Chilling Hours’ instead of ‘Chill Hours’ since the first one is the most common term for the units of the Chilling Hours model. Similar with Growing Degree Days instead of Grow Degree Days.

L12. Consider merging the information presented in parenthesis. For example, ‘… exogenous gibberellic acid (GA₃; 0 and 49 g ha^-1).’ Same in other parts of the manuscript.

L11-14. Consider being more specific in this sentence by mentioning the moment of application. Perhaps, in terms of biological processes could help understanding the specific goal of this study.

L14-16. Although the information presented here is easy to understand, I would suggest showing the numbers right after the variable mentioned. Otherwise, this can be a bit overwhelming.

L48-50. The effects of day length during eco-dormancy are at least controversial up to my knowledge. It is not clear if bud scales can sense the change in photoperiod in winter and if so, how the process is. Consider adding a proper reference for this and add the words ‘… and day length changes in some species (The proper reference).’

L69. Above sea level?

L178. By 5 days…

L186-188. I think the 499% is wrong. Thirty-four (the increase in yield) divided by 7 produces 4.857 or 485.7%. Please check this and other computations in the same way.

L303. This is the first time you use TSS please explain it.

Table 2. I think the footnote ‘w’ should be placed in the ‘Marketable yield (g/plant)’ title.

L314. Add a space after Table.

L320-321. I am not sure how did you come to this statement. It seems that your experiment only considered one level of chilling and therefore one moment of application of GAs. Therefore, there is no way you can test with this experiment the influence of winter chill on phenology. The earliness of variety ‘Natchez’ may be given by the low chill requirement, which might be achieved in both the subtropical and temperate climate conditions. After minimum chill requirement is achieved, bud burst dates are mainly driven by growing temperatures, which are less frequent in temperate climates. This could explain the advance of 2 to 4 weeks you mention in line 315. In this regard, stating that the earliness can be expressed regardless winter chill may be wrong to me.

L368. Respectively is repeated

L385-386. It is not clear to me where these estimates come from. Are you using the results on yield from your study mixed with planting density from another study? Don´t you think this could lead to erroneous estimates since the panting distance in your study is about 8.958 plants per hectare (0.61 x 1.83)? Moreover, I think you need to be a bit more specific about the yield you used (i.e. to mention the variety and treatment to which this yield belongs).

L391-393. I am not sure I fully understood the meaning of this sentence. By now, I think the concept you want to explain is that the plants are bigger in a commercial orchards compared to your design. However, you do not say this explicitly in the sentence. One could think that less distance within rows can lead to a greater number of plants, which could compensate for the lost in plant size. Please clarify this point.

L411-412. This is confusing to me. Your results demonstrate that GA application advanced fruit ripening in two out of three varieties across seasons. Please explain it better.

L424-426. Where does this point come from? In methods, you do not mention that the application of GAs was defined according to chill accumulation. Usually, the application of dormancy breakers is performed once a specific amount of chill accumulation is achieved. Did this study evaluate the application of GAs when maximum chill accumulation occurred? If not, do you think the results would be applicable when GAs are sprayed at maximum chill accumulation? To me, ‘maximum chill accumulation’ is too vague and can be very late in the season. Perhaps, buds already break naturally when maximum chill accumulation is not even achieved yet. Please clarify.

I think the information/table presented as supplementary materials is relevant for the study and should be included in the main manuscript.

This is a comment about style. Consider using more active sentences. For example: ‘We recorded dates of bub break…’ instead of ‘Dates of bud break were recorded…’

Author Response

Please see the attachment.

Author Response File: Author Response.pdf