Effect of Genotype × Environment Interaction on Yield of Maize Hybrids in Greece Using AMMI Analysis

Round 1

Reviewer 1 Report

In their paper, Katsenios et al, present the results of employing AMMI analysis to analyze variance in maize grain yield. They use five commercial maize hybrids and six experimental locations in Greece during a two-year experiment, to figure out the adaptation of genotypes in these environments and which genotypes are more suitable for cultivation in the mega-environments. The article is concise and to the point but there are many weaknesses that need to be addressed. The main problem is in Results and Discussion section which is weak; authors present their results without discussing them citing relevant literature and the two paragraphs they dedicate for this (lines 186-194 and lines 214-220) are too generic with no specific reference to their results. Authors dedicate lots of space for describing their Materials and Methods but nearly the same space for citing and discussing the results. According to my opinion they should expand the Discussion of results – enhancing it with some handful insights of their work and possible future implications. Also conclusions section is too long; maybe they can move some of the info they present there into the previous section. English should be checked and corrected throughout the text as in several occasions the syntax is not ideal.  

More specific comments:

Line 42; Zea mays L. should be in italics.

Line 125; what CEC stands for?

The authors use maize hybrids that belong to different maturity groups, e.g., Gen3 needs 116-123 days to maturity while GEN5 135-140 days. I wonder whether this affects the result of the evaluation? I would like to hear authors opinion on this matter.  

Author Response

Dear reviewer, thank you for your time reading and reviewing our paper. 

The commends were really useful and definately have helped the quality of the paper. Please find attached the point to point corrections. 

Author Response File: Author Response.docx

Reviewer 2 Report

The manuscript after revision still lacks a clear focus. AMMI is applied to a small number of hybrids (5) and environments (6 locations x 2 years). The statement of the objectives is quite general detecting GxE and narrow adaptation.

Technically this can and has been done. However, the relevance of the results obtained remains largely unexplored and is subject to still missing pieces of information.

The authors did not specify, if the experiment has been done under rain-fed or irrigated conditions and if the latter is the case, how irrigation has been performed. Instead, the reader is still left with the general statement, that irrigation is applied for maize cultivation in Greece.They explain in their response letter that in recent years summer rainfall occured in the region of interest. If this were a motivation to study options for rainfed production this would be fine but needs to be clearly indicated. If irrigation was applied, the details are important, as irrigation will define an important element of the environmental conditions and in addition referring to rainfall sums as a potentially distinctive charateristics of the environments is not convincing any more. Furthermore, two locations share the same climate information, although 3 (lon) and 6 (lat) minutes apart. Thus, on-site information can not have been obtained for all locations.

At the end narrow adaptation for 3 mega-environments are described. First it would be good to explain why the recommendation to fuse mega-environments with a small number of constituting environments was not applied. Second, the contrast between mega-environments 1 and 3 can be interpreted as characterised by warmer more westerly sites with basic soils versus cooler more easterly sites with acidic soils. If this result should indicate a geographical pattern beyond the specific experiment, more information on the representativity of these contrasts in space and time were needed. If the focus shall be on the difference between hybrids 2 versus 4 and 5, a more in depth discussion of this contrast would be needed.

Beyond these considerations further improvements in the description of the AMMI results and a profound language correction are needed.

Author Response

Dear reviewer, thank you for your time and effort to make out paper better. 

I trully believe it is better tham before 

Author Response File: Author Response.docx

Reviewer 3 Report

In this study, the authors analyzed Genotype x environment interaction on yield of maize hybrids in Greece. Overall, the research is interesting and has the potential to be published after the authors work on the following revisions.

In lines 116 and 117.  22.2 and 22.4 oC What is It?

In line 125 and 126. Could you explain more about fertilizer?

What soil types were in your locations?

In total this work is very interesting and very valuable!

Author Response

Dear reviewer, thank you for your time spending on our paper. 

please find attached the point to point answers. 

Hopefully you will find it ok to be published

Author Response File: Author Response.docx

Reviewer 4 Report

The manuscript by Katsenios et al. “Genotype x environment interaction on yield of 2 maize hybrids in Greece” reports on anAMMI analysis of yield for five maize genotypes across six environments.

The paper is written correctly and appropriate methodology was used. Yet the manuscript lacks generally details and more in depth analyses, together with more extensive reference list. The reference list presented is limited and does not include many publications on maize so it should be integrated. Presentation of results also should be improved and more detailed (see the attached file for some details).

Although the authors have analysed 12 genotype x year x location combinations, which potentially are an excellent basis for G x E analysis, they included only grain yield which is a final product of interactions of different yield components. Therefore, not much can be concluded on the underlining factors that determine a genotype’s better adaptability to a certain environment. In fact, besides concluding on which genotype is more suitable to a group of environments, the authors do not discuss any of e.g. possible genetic/physiological reasons. By limiting to comment in a rather simple and incomplete way, different pedo-meteorological data, the importance of the results remains rather local without offering new insights applicable to wider maize cultivation or even breeding.

The authors are encouraged to integrate other traits and repeat their analysis.

Some suggestions to improve the content are given in the attached file.

Comments for author File: Comments.pdf

Author Response

Dear reviewer, 

thank you for your help. It looks better now. 

Find attaced the point to point answers

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The authors have made a significant progress to improve their manuscript and their efforts are much appreciated. However, the Results and Discussion section is still weak, in some cases too generic and vague, and the references added are not always discussed in relation to the results. For example, references added in lines 212-220.

In lines 272-274, authors refer to a publication by Cinta Romay et al (Crop Science, 2010) and they write: “climatic conditions, especially high temperatures (daily mean temperature over 10 °C), in a grain yield experiment of maize in Spain were found to be important”. In their paper authors Cinta Romay et al, say that (and I quote the original paper): “The main climatic covariates for yield were related to days with mean temperature over 15°C and maximum temperature in September, but they were not consistent across locations. Therefore, if yield under stress conditions is a breeding goal, several climatic variables, especially those related to high temperatures, and genotypic traits such as kernel depth and ear length should be considered”. Could the authors discuss this point a bit more as the meaning here is not quite clear.  

Again, right after (lines 274-276), authors write a sentence that is very generic and does not add something valuable to the discussion. Authors here refer to a paper from BRANKOVIĆ-RADOJĈIĆ et al (Genetika, 2018) that deals with the evaluation of maize grain yield and yield stability by AMMI analysis yet they choose to only mention here that “mean temperature during the cultivation period (April-–October), as well as precipitation (rainfed cultivation), were considered as abiotic stress factors”. Could the authors elaborate on this point a bit more? 

Always in the Results and Discussion section, lines 186-190, authors are citing references that state the main environment effects account for more than 75% of the variation. Yet their results show that environment effects account for only 6.85% of the variation. How do they explain such a difference in the reported results and the results cited in literature? 

In total, the Discussion section should be improved. 

Author Response

Dear reviewer, thank you for your useful commends and the time spended on our paper. 

please have a look again and hopefully let us publish it. 

Reviewer 2 Report

The manuscript has been improved by adding information on the choice of hybrids, irrigation and the corrections of presentation. Yet, still it does not allow to fully appreciate the relevance of the results reported.

Plants use soil water. Thus, it is not the amount of rain but the availability of water in the soil that determines water status. Some minor effects apart, rain is influencing plant performance indirectly via filling the soil water resource. In rainfed environments, very often rainfall is the only source of water and thus precipitation amounts are a good estimator of water availability. Under irrigation, the amount of water applied can have varying weight in determining plant available water. Thus, it is an important step having added the information that surfce irrigation was applied. However, not yet sufficient. If possible please add a short description of the way an equal water availability at the different sites was assured. Did you e.g. refill measured soil water content to a prescribed level, or compensate for measured or estimated evaporation losses or else? Did the treatment lead to optimal water availability at all sites or may drought still have played a role? Given these considerations, it is not evident why it is „important“ to note the differences in precipitation between sites (L235/236, as well as respective statement in the abstract). May be this is the case, than the putative cause will need to be named. If it is just a curiosity to note, the text needs to be reworded, here.

Weather data were not measured at the site at least in one site. Thus, at least a short phrase needs to be added „or were representative for the site xxx at a distance lower than 5 km from the site yyy.

The relevance of the identified mega-environments is still not evident. From your response in the cover letter, it seems that you think there is no identifiable geographical pattern in the climatology of the sites. If they are not representative for climatic differences, i.e. if in the long-term (many years) average a site with high maximum temperatures in the two experimental years may not be characterised by high maximum temperatures, then they do not have any predictive value. Remark, with the observation in my previous review I did not want to claim that this was the correct description of the geographical gradient but just to state that it is important to describe it. If the mega environments are not representative for whatever regions, then which are the conclusions a farmer can draw?

Furthermore, if high maximum temperatures are a distinctive feature of the mega-environments, shifts in the climatic regime may have an impact on the suitability of hybrids therein.

In conclusion, I suggest to further move ahead in transforming the paper from one that seems to be based on the assumption that the application of a statistical method (AMMI) by itself defines the significance of outcomes and improve the contextualization for the pedo-climatic features of your target regions.

The correct technical terms for what you address as monthly high and low temperatures are maximum and minimum temperatures.

L221: many other cultivars → cultivars of many other species

Author Response

Dear reviewer , 

thank you for the time you spended on our paper. 

please have a look and let us know what you think. 

Author Response File: Author Response.docx

Reviewer 4 Report

Dear Authors and Editors,

the effort to improve the manuscript is evident and appreciable, however, I still believe that including only the final yield is not sufficient, and that other parameters should be added to the analysis e.g. days to maturity. As I wrote in my previous review, final yield is a product of multiple traits, which should be taken into consideration (at least some of them). As they stand, I find the results of limiting impact. For example, in the description of hybrids used, parameters such days to maturity or number of kernel rows have been indicated to illustrate good characteristics of the hybrids, but these parameters have not been analysed across the environments, to establish if they are stable under different growing conditions, and if the breeders can identify a particular trait to target. To target only final yield is of unpredictable outcome as it is a complex quantitative trait.

Discussion part has been enlarged by adding new references, but they were not correlated with the obtained results. One example regards justification of application of the AMMI analysis (lines 206-220 in the resubmitted manuscript). There is no doubt that the AMMI analysis is widely used and valid approach, and there is no doubt that the authors have conducted it well. However, instead of listing other studies on maize that also used it, the authors should have compared the published results with their own (if possible), to be able to say e.g. “…similar results were obtained in the study by…” or “…in contrast to what observed in the study by…”.  This is also valid for the discussion on pedo-meteorological conditions and the suitability of a given genotype to a certain environment. The discussion should include the comparison with other studies that have included same hybrids (or some of them) and/or similar environments tested. Otherwise, the results remain informative for Greek farmers and breeders only and the paper contains little novelty . In my view, an international scientific journal should provide new knowledge that can be in some measure more transferrable across the community. I understand that the hybrids chosen are widely cultivated in Greece, but if they are cultivated also in other countries, please give this information, so that your results may be useful to other cultivation areas.

Author Response

Dear reviewer thank you for the time you spended on our paper. 

please have a look again 

thank you 

Author Response File: Author Response.docx

Round 3

Reviewer 1 Report

Authors efforts to improve the paper according to previous reviews are substantial. The quality of presentation has been improved and so has the english language used which is now fine. The authors have addressed my previous comments in both review rounds - except the comment about the variation explained by environment effects (6.85% vs 75% reported in the literature they cite) however this is only of minor significance. 

Author Response

thank you for your time 

Author Response File: Author Response.docx

Reviewer 2 Report

General remarks:

The mere fact that a statistical method can technically be applied does not provide a proof of it's appropriateness for the analyses of a given data set. As you think that there is no way to establish the representativity of the macro-regions identified with AMMI for larger circumscribed regions and if you think that any of the studied sites could equally well be the hottest or the coolest in coming years, the validity of the macro regions is questionable. In addition, two years are to few, to allow for a statistical inference concerning inter-annually varying climatic conditions. For me the most convincing results are the identification of differences between soils of different acidity and hybrids more or less suited to cope with high temperatures – results that could have been obtained without applying AMMI. As the editor decided in an earlier round to go on improvind the paper, I do not base my recommendation on these considarations.

Specific suggestions:

I suggest to add the description of the irrigation method that was given with the last letter of response to the paper.

Author Response

thank you for your time and useful commends 

Author Response File: Author Response.docx

Reviewer 4 Report

Dear authors, I appreciate your effort to improve the paper and thank you for your extensive reply.

Author Response

thank you for your time and useful commends 

Author Response File: Author Response.docx

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

The manuscript needs to be amended with some essential information on experimental conditions and environmental data. Furthermore, it needs to be better focused on the objectives and errors in the interpretation of the results need to be corrected.

In detail:

Aims and scope: At the end of the introduction section it is stated realistically that the paper aims at identifying mega-environments. However in the results and discussion section relatively long passages are dedicated to saying that AMMI is the adequate method. There are no specific tests applied to test the adequacy or efficiency of AMMI with respect to alternative methods. The paper does not contribute to evaluate AMMI but simply applies it. Which is fine, Thus, these before mentioned sections should be removed from the discussion section. Part of them can be integrated into the introduction. Instead, a more in depth exploration of the mega-environments would serve to strengthen the message of the paper.

How and where were the weather data recorded (distance from the experimental sites). What do “mean monthly high temperature” and “mean monthly low temperature” stand for? Are these the averages of the mean monthly maximal and minimal temperatures (march through september) or the highest and lowest mean monthly temperatures in one of the months?

Add the fertilizer quantities.

L99/100 State explicitly that grain yield is studied, already here.

From the description of the experimental conditions it seems that the experiment has been conducted under rain-fed conditions. If this is correct, state this fact explicitly. Then it is mentioned that maize is cultivated with irrigation in Greece. Thus, it would be good to explain the relevance of performing the experiment under rain-fed conditions.

L199-213 need to be re-evaluated. It is the distance from the origin that depicts a high strength of interaction. Environments 6, 8 and 12 as well as genotype 4 do not show a big distance to the origin. The evaluation of correlated or uncorrelated interactions need to be critically revised. See for reference Gauch and Moran 2019 who explicitly state: “Two markers of the same kind (both genotypes or else both environments) in the same direction from the origin have similar interaction patterns; markers in opposite directions have opposite interactions; and markers nearly at right angles have uncorrelated interaction patterns.”

Note the hottest environment leads to low yield and average yields in the environments are negatively correlated with all three temperature variables.

Reviewer 2 Report

- The authors talk about genotype x environment interaction on the yield of maise hybrids in Greece (five hybrids and 12 environments). The manuscript has serious flaws. Authors did not conduct the research correctly.

- The experiment is set up incorrectly. The authors opted for five genotypes of different FAO groups, which prevent them from comparing genotypes with each other because different characteristics characterise such genotypes. It can be seen from what was written in section 2.1.2.

- The authors should have used the genotypes of the same maturing group or three or more genotypes of the same FAO group each to determine which FAO group was most suitable for breeding in a particular environment.

- In addition to the above, the significant disadvantage is that authors sowed all genotypes on the same number of plants per hectare (85000 plants/ha). Sowing should have been carried out according to the manufacturer's recommendations.

- Authors should study writing instructions more carefully.

- The conclusion in section the Abstract is missing. 

- On page 2, the sentences say the same thing (lines 48-51).

- After the goal, the authors should state the working hypothesis.

- Page 2, line 91 - authors should have additionally marked the listed genotypes with numbers as they appear in further text.

- Authors should specify that for environment labels, one should refer to Tabel 2.

- A more detailed description of section Statistical analyses is needed.

- Similar results from other research are lacking in section the Results and Discussion.

- Some sentences from section the Introduction are repeated in section the Results and discussion.

- The written Conclusions are incorrect due to the shortcomings of the experiment.

Reviewer 3 Report

In their paper, Katsenios et al, describe the use of AMMI analysis to detect Genotype x Environment interaction effects in maize cultivation in Greece. They use five commercial maize hybrids and six experimental locations in a two-year experiment in order to figure out the adaptation of genotypes in these environments and which genotypes are more suitable for cultivation in the mega-environments. Although the authors present some interesting results, there are several drawbacks in the present manuscript and weaknesses that should be addressed and reviewed. More specifically, :

Paragraph starting line 225 ending line 236 is problematic. My general impression by going through the results and discussion section is that for those not familiar with this kind of analysis it is really difficult to follow the conclusions the authors make for example in this case by looking and studying the two tables (table 5 and table 6). The authors should explain better what the analysis produces and how it is interpreted; for example, what does the term AMMI model family means, how it is concluded that “the first mega environment is consisted of five environments (ENV1, ENV2, ENV4, ENV5 and 227 ENV7) in which GEN2 was the best adapted genotype”, what does the F stands for in Table 5 etc. 

Throughout the text, authors repeat information in the form of citations they have already used in the introduction or in other parts. They should avoid this kind of repetition. For example, citation [29] Gurmu et al, is mentioned in line 193 and then it is repeated in lines 216-217. 

The manuscript needs extensive editing of english language. Sometimes the wrong syntax makes the text uncomprehensible. Some examples:

Line 76; evaluate or appraise is the right verb here? They can’t be both.

Line 81; “…as a result of (not from)”

Line 94; maybe produced is a better verb here instead of gave.

Line 97; “had a size of 60m2” and it consisted

Line 110; overcome? Maybe the right verb is surpass?

Line 163; “measurement” is redundant here. Grain yield is a characteristic by itself measured in kg/ha.

Lines 202-203; “It is important to note that at these environments, it has been recorded the highest rain precipitation during the cultivation period”. This is odd syntax. I think it should written as “It is important to note that the highest rain precipitation during the cultivation period has been recorded at these environments”.

Line 210; missing reference [30].

Line 211-212; This sentence does not make sense.

Line 242; cultivations is inappropriate word here. Maybe authors should replace it with crops? Or something similar?