Drought Stress Response of Turf-Type Perennial Ryegrass Genotypes in a Mediterranean Environment

Round 1

Reviewer 1 Report

The manuscripts details the response of perennial ryegrass cultivars (and one tall fescue) to drought in a two year field study. The author’s stated objectives were to assess difference among perennial ryegrass varieties and understand their drought responses. I have some reservations about if this study, as written, answers these questions about differences among cultivars and their drought responses. I do think understanding drought responses among cultivars is important and feel the authors may be able rewrite sections of the manuscript to be acceptable, allowing readers to draw more meaningful conclusions. 

L41 – Drought escape is considered distinct from drought avoidance. [ Furthermore, there is some debate if drought avoidance is an appropriate term, as drought is a meteorological condition that cannot be avoided, and those who prescribe to that idea prefer the term desiccation avoidance. However, I feel that drought avoidance is used widely enough that its meaning is not misunderstood.]

L46 – I don’t feel like the sentence accurately captures the meaning of drought tolerance. It is not that drought tolerant plants are desiccating and losing color, but that they are able to maintain metabolic functions at reduced water potentials.

L49 – I think osmotic water stress is a strange term to use, usually this is used in the context of artificial osmotic stresses, used as those induced by an osmoticum like PEG or mannitol.

L54 – I would remove the word “prolonged” as that seems to be the real distinction between acute and chronic, the length of time/intensity. Also, in both scenarios described, the soil will become increasingly dry, so this does not seem like an important distinction.

L59 – Missing a period at end of sentence.

L61 – Mention specifically this is a screen for drought tolerance, but is drought tolerance and avoidance really separated in this kind of study?

L62 – Is a comparison against a single tall fescue genotype really a meaningful comparison?

L63 – Length of time without irrigation is very environment /ET dependent so there needs to be real caution in how this is described.

L64 – A proper hypothesis is supposed to contain a explanation of proposed phenomena, not just a prediction of result

L80- More background on cultivars could be beneficial. Are any of these older cultivars predicted to be less drought resistance?

L108 – I am never sure the real value that can be gleaned from these kinds of recovery measurements, as plots are recovering from very different points.

L118 – I would have thought to move the information on collecting environmental data all together (i.e. previous paragraph), and measurements of plant parameters all together.

L116- how was soil moisture measured (e.g. one spot per plot? measurement depth? Etc.; more details needed)

Figure 1 has extra lines, probably just a graphical error?

For the statistical model, still slightly unclear how dates were compared across years, even if it was centered on the end of the drought, because there were still differences in the number of days (Were the ends of one trial trimmed off? I am not saying this needs to be changed, but I was still slightly unclear on this analysis so clarification may be beneficial). However, I think the authors make an excellent point that basing analysis on a fixed number of days is not always appropriate (since two droughts are not always identical) and should be applauded for taking an approach that addresses this. Additionally, the authors may want to consider if year should really be a random effect, or not. It is conceivable that the year had a meaningful effect and data could be potentially analyzed separately by year?

L151 – So does this mean that green cover was the only measurement that was able to detect difference in drought performance since it was the only measurement with an interaction term? Can this can be extrapolated to say that on most measurements, there was no difference in drought performance among cultivars? And if so why were no differences detected?

L168 – The discussion of why stolons may be beneficial is very thin. Although I don’t disagree that stolons may have been beneficial, stolons were not measured in this study, so the appropriateness of this conclusion could be debated. [perhaps it would be more appropriate to discuss stolons during recovery at any rate?]

L179 – I think there are many papers that have reported % green cover and soil moisture content, however those previous authors probably do not put great emphasis on this due to different soils having difference amounts of available soil water/soil water potential. The next sentence may be making this point but it is unclear. 

L180 – Statement on drought tolerance here unclear, and how it relates to saturation.

L184- What is the hysteresis between soil moisture and percent green cover? Maybe the soil moisture requirements are even greater, and it takes a week for effects to be detected? Also if this is an important conclusion of the study presenting results as soil water potential may make it more easily translatable (although I understand this data may not be available).

L193 – While this is true, it will be highly dependent on a number of factors (plant, soil, environment), and this needs to be emphasized more.

Table 3 – should clarify this is cumulative ET

For the recovery period if soil moisture never even returns close to what was seen at the beginning of the trial, but stays at the levels when stress symptoms first appeared (~20%) then is this really a true recovery?

L206 – I am not sure if you should say “the species” since you study technically has two species, I assume you mean the perennial ryegrass, but this should be explicitly stated

Figure 3 – Again does this tell us anything about actual drought performance, or just genetic differences inherent among cultivars?

L231- Mention models using different parameters gave better results, but this data is not presented. This seems like a major short coming since it seems to be an important findings. This again comes up in the conclusions (GDD vs ET) and I think data supporting this would be important.

L245 - Is the greater quality in 'New Orleans' due to drought tolerance? This point is not made clear, similarly differences during recovery due to improved drought resistance not made clear.

L238 – I don’t know if humidity reduced the effect of ET on drought stress, or just effected the severity and rate of onset of drought symptoms?

L248 – have you thought about presenting data as % of controls [week zero], to help normalize for inherent differences among cultivars?

Could differences in plot coverage have potentially effected differences in water use rates?

Author Response

Thank you to the anonymous reviewers for your time and efforts in reviewing our manuscript. We have made changes to the manuscript and feel that most/all comments and requests for changes have been adequately addressed. Below please find our rebuttal statements (listed in bold) to referees’ comments and suggestion for changes.

Reviewer: 1

The manuscripts details the response of perennial ryegrass cultivars (and one tall fescue) to drought in a two year field study. The author’s stated objectives were to assess difference among perennial ryegrass varieties and understand their drought responses. I have some reservations about if this study, as written, answers these questions about differences among cultivars and their drought responses. I do think understanding drought responses among cultivars is important and feel the authors may be able rewrite sections of the manuscript to be acceptable, allowing readers to draw more meaningful conclusions. 

L41 – Drought escape is considered distinct from drought avoidance. [ Furthermore, there is some debate if drought avoidance is an appropriate term, as drought is a meteorological condition that cannot be avoided, and those who prescribe to that idea prefer the term desiccation avoidance. However, I feel that drought avoidance is used widely enough that its meaning is not misunderstood.]

Paragraph corrected accordingly.

L46 – I don’t feel like the sentence accurately captures the meaning of drought tolerance. It is not that drought tolerant plants are desiccating and losing color, but that they are able to maintain metabolic functions at reduced water potentials.

Sentences modified as per suggested

L49 – I think osmotic water stress is a strange term to use, usually this is used in the context of artificial osmotic stresses, used as those induced by an osmoticum like PEG or mannitol.

“Osmotic” removed to avoid confusion

L54 – I would remove the word “prolonged” as that seems to be the real distinction between acute and chronic, the length of time/intensity. Also, in both scenarios described, the soil will become increasingly dry, so this does not seem like an important distinction.

Removed

L59 – Missing a period at end of sentence.

Checked

L61 – Mention specifically this is a screen for drought tolerance, but is drought tolerance and avoidance really separated in this kind of study?

Correct, the study is an acute drought study, the difference among drought avoidance and tolerance is species characteristic. The term tolerance have been replaced with general term of drought resistance.

L62 – Is a comparison against a single tall fescue genotype really a meaningful comparison?

It is general consent that tall fescue is more drought tolerant than ryegrass. Although we are aware of differences among tall fescue varieties in term of drought resistance, we feel that for the purpose of this study one tall fescue variety was sufficient.

L63 – Length of time without irrigation is very environment /ET dependent so there needs to be real caution in how this is described.

Correct, it is also soil and relative humidity dependent, hence the need of testing in multiple climate with different soil characteristics.

L64 – A proper hypothesis is supposed to contain a explanation of proposed phenomena, not just a prediction of result

Changed

L80- More background on cultivars could be beneficial. Are any of these older cultivars predicted to be less drought resistance?

More information has been added in the materials and methods. The ancienty of a cultivar is not necessarily linked to its tolerance to drought. A cultivar may have been bred primarily for other features, such as quality or color.

L108 – I am never sure the real value that can be gleaned from these kinds of recovery measurements, as plots are recovering from very different points.

The experiment was conducted in a rain-out structure, if we decided to follow every cultivar until a fix green cover percentage to recovery them from the same point, we would run the risk of comparing different recovery periods with different temperatures and/or relative humidity. For this reason, we decided to end of the drought phase at the same time for all cultivars.

L118 – I would have thought to move the information on collecting environmental data all together (i.e. previous paragraph), and measurements of plant parameters all together.

The paragraph from line 108 to 125 (old version line numbers) has been split and environmental parameter collection and calculation have been joined all together, as well as measurements on plants.

L116- how was soil moisture measured (e.g. one spot per plot? measurement depth? Etc.; more details needed)

Information has been added (line 116).

Figure 1 has extra lines, probably just a graphical error?

Sometimes converting word files in pdf could cause graphical errors, but the figure on word file is correct.

For the statistical model, still slightly unclear how dates were compared across years, even if it was centered on the end of the drought, because there were still differences in the number of days (Were the ends of one trial trimmed off? I am not saying this needs to be changed, but I was still slightly unclear on this analysis so clarification may be beneficial).

How dates were compared across years has been clarified in the text (lines 163-165).

However, I think the authors make an excellent point that basing analysis on a fixed number of days is not always appropriate (since two droughts are not always identical) and should be applauded for taking an approach that addresses this. Additionally, the authors may want to consider if year should really be a random effect, or not. It is conceivable that the year had a meaningful effect and data could be potentially analyzed separately by year?

In this kind of field experiment where environmental conditions are not totally controlled (e.g.  greenhouse) but at the same time also not completely uncontrolled (e.g. field), year could be included in the model as random or fixed effect depending on the aim of the study. Year was included in the model as random effect because our goal was to study the response to drought stress beyond climatic changes of years. The incorporation of year as random term permits to control non-independence in the data, arising from grouped observations.

L151 – So does this mean that green cover was the only measurement that was able to detect difference in drought performance since it was the only measurement with an interaction term? Can this can be extrapolated to say that on most measurements, there was no difference in drought performance among cultivars? And if so why were no differences detected?

This is a good point. Visual quality and color are less sensitive to changes because are integer numbers with a small range (from 3 to 9 in this study). Instead, we expected significant interaction between cultivars and dates for NDVI. However, similar to other unpublished studies, green cover seems to detect differences in stress better than NDVI. Our study does not provide an answer for such an interesting result.

L168 – The discussion of why stolons may be beneficial is very thin. Although I don’t disagree that stolons may have been beneficial, stolons were not measured in this study, so the appropriateness of this conclusion could be debated. [perhaps it would be more appropriate to discuss stolons during recovery at any rate?]

The discussion has been improved.

L179 – I think there are many papers that have reported % green cover and soil moisture content, however those previous authors probably do not put great emphasis on this due to different soils having difference amounts of available soil water/soil water potential. The next sentence may be making this point but it is unclear. 

Some studies have been entered at lines 209-213 and the sentence has been improved.

L180 – Statement on drought tolerance here unclear, and how it relates to saturation.

“after saturation” has been deleted.

L184- What is the hysteresis between soil moisture and percent green cover? Maybe the soil moisture requirements are even greater, and it takes a week for effects to be detected? Also if this is an important conclusion of the study presenting results as soil water potential may make it more easily translatable (although I understand this data may not be available).

Soil moisture was recorded once a week before weekly irrigation, so it was influenced by the whole previous week. The soil water potential may be very helpful for such a discussion, unfortunately we do not have this information.

L193 – While this is true, it will be highly dependent on a number of factors (plant, soil, environment), and this needs to be emphasized more.

The sentence has been rewritten as “Our results suggested that an average visual quality rating of 7 can be maintained in turfs constituted by perennial ryegrass for about two weeks without irrigation in a coarse-silty soil”.

Table 3 – should clarify this is cumulative ET

Clarified in the table.

For the recovery period if soil moisture never even returns close to what was seen at the beginning of the trial, but stays at the levels when stress symptoms first appeared (~20%) then is this really a true recovery?

The irrigation amount used during the recovery phase is one that is considered appropriate for most cool-season turfgrasses. For our region such a value would be 80% of ET.

L206 – I am not sure if you should say “the species” since you study technically has two species, I assume you mean the perennial ryegrass, but this should be explicitly stated

“Species” has been changed with “perennial ryegrass”.

Figure 3 – Again does this tell us anything about actual drought performance, or just genetic differences inherent among cultivars?

The figure state differences among cultivars and average drought and recovery period, thus it is mainly related to genetic differences. We added it for completeness of information as the main variable “cultivar” was significant for reported parameters.

L231- Mention models using different parameters gave better results, but this data is not presented. This seems like a major short coming since it seems to be an important findings. This again comes up in the conclusions (GDD vs ET) and I think data supporting this would be important.

Data for model comparison have been added.

L245 - Is the greater quality in 'New Orleans' due to drought tolerance? This point is not made clear, similarly differences during recovery due to improved drought resistance not made clear.

The sentence has been clarified.

L238 – I don’t know if humidity reduced the effect of ET on drought stress, or just effected the severity and rate of onset of drought symptoms?

Thanks for the clarification. The sentence has been improved.

L248 – have you thought about presenting data as % of controls [week zero], to help normalize for inherent differences among cultivars?

Presenting data differently can result in misleading conclusions or can be challenging to discuss. We prefer to present data as is which makes it easier to understand.

Could differences in plot coverage have potentially effected differences in water use rates?

Yes, it could. However, this aspect is outside the scope of our study.

Author Response File: Author Response.pdf

Reviewer 2 Report

The work concerns an important problem as water deficit is increasing and not only in the Mediterranean region. Progressing climate change increases the importance of this research. The research was well planned and carried out with the use of adequate methods. The research results were presented in a clear and understandable way. The practical and theoretical background of the research was also well outlined.
The average assessment of the originality and scientific importance of the work results from the possibility of further extending and deepening the analyzes carried out.
The work, however, requires some explanation. One of the hypotheses adopted (see lines 64 and 65) seems to have no continuation in the work. Otherwise, the "old" and "new" varieties should be identified and analyzed for resistance to drought stress.

Author Response

Reviewer: 2

The work concerns an important problem as water deficit is increasing and not only in the Mediterranean region. Progressing climate change increases the importance of this research. The research was well planned and carried out with the use of adequate methods. The research results were presented in a clear and understandable way. The practical and theoretical background of the research was also well outlined.
The average assessment of the originality and scientific importance of the work results from the possibility of further extending and deepening the analyzes carried out.
The work, however, requires some explanation. One of the hypotheses adopted (see lines 64 and 65) seems to have no continuation in the work. Otherwise, the "old" and "new" varieties should be identified and analyzed for resistance to drought stress.

The concepts “old” and “new varieties” has been deleted (Lines 78-79). Drought tolerance is only one aspect for breeding, and newer cultivars could be selected for different characteristics such as quality, colour, but not necessarily drought tolerance.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The results from these experiments may not have a major impact on their own, but I feel the research fits into the broader context of the important issue of drought and does provide some valuable information. I have some reservation about the statement "the first date of 2017 and the last date of 2018 were used 2 times." in the description of analysis, but this likely has no impact on the results or their interpretation. 

I feel the authors have sufficiently answered my initial questions and concerns. 

Author Response

Both Reviewer 1 and the Academic Editor expressed concerns/issues with our methodology of using the first date of 2017 and the last date of 2018 twice (see comments below). Therefore, we ran the models again but without using the time periods in question 2 times, treating the trial as unbalanced. The results of analysis of variances do not change and values in Figure 2 and Table 3 change only slightly for the first and last date. This has very little to no effect on results and discussions.  The following changes were made to the text:

• At lines 163-164 the sentence “In order to have the same number of dates for the drought and recovery periods of both years, the first date of 2017 and the last date of 2018 were used 2 times.” has been deleted.
• Figure 2, Table 3, and Table 4 have been corrected accordingly.
• At lines 186, 244, 250, and 251 results have been changed according to the new Figure 2 and Table 3

We hope this addresses the reviewers concerns adequately and the reviewers will accept our revisions.