Pea Breeding Lines Adapted to Autumn Sowings in Broomrape Prone Mediterranean Environments

Round 1

Reviewer 1 Report

See the attached file

Comments for author File: Comments.pdf

Author Response

WE APPREICATE THE DETAILY REVISION MADE. ALL COMMENTS WERE CONSIDERED USEFUL AND WERE INCORPORATED, AS WE DESCRIBE:

- The whole is a bit too long, I suggest that certain Tables are placed in supplementary Tables and keep Figures in the manuscript: Tables 4, 5, 6, 7 and 9, as for ascochyta blight. This would make it possible to lighten the text a little. Table 8 is interesting to maintain. OK, THESE TABLES HAVE BEEN MOVED TO SUPPLEMENTARY TABLES.

From the breeder's point of view, the use of a selection index is very efficient, but I am surprised that it is not possible to weight the different factors in the MTSI index. THIS WAS REVISED, SEE DETAILED RESPONSE BELOW

The discussion on the role of the environment in broomrape infection is very important. We also see that the environment also plays a role into yield and other production variables, which is not surprising. But I think we could have gone further in considering the environment. For example, soil characteristics (depth, water reserve) play an important role as well, also for diseases. It would have been possible to take them into account. THIS SUGGESTION IS VALUABLE. THESE AS OTHER VARIABLES COULD BE CONSIDERED. HOWEVER, AS REV1 ALREADY STATES, THE MS IS ALREADY TOO LONG AND WE SHOULD FOCUS. ALSO, IT IS INTERESTING TO NOTICE THAT WE RECENTLY PUBLISHED A NUMBER OF SIMILAR COMPARATIVE STUDIES IN OTHER CROPS (I.E. LATHRUS SATIVUS, VICIA ARTICULATA, VICIA FABA,  PEA…) PROVIDING LESS DETAILS ON THE SITES THAN IN THIS MS. HERE, FOR THE FIRST TIME, WE DID PARTICULAR EFFORTS TO ANALYZE SOIL SAMPLES PROVIDING SUCH DATA FOR THE SITES. INDEED, MORE (LIKE THE ONES SUGGESTED) COULD ALSO BE ADDED. BUT THIS, WE WILL CONSIDER THIS FOR COMING STUDIES.

Furthermore, the nitrogen status of the plants was not checked, when it has an influence on yield, and diseases like powdery mildew). You also could give an evaluation of the density of foliage, and of the intensity of branching. Nevertheless, the observed links are already very interesting. YES, INDEED WE COULD HAVE ALSO ACCESSED ALSO NITROGEN STATUS OF THE PLANTS, BUT WE DID NOT. BUT THIS, WE WILL CONSIDER THIS FOR COMING STUDIES.

I have a general remark on the use of the concept of stability. You use this concept in the dynamic meaning (behavior of a line in comparison with all other lines, or with the average behavior). But for certain variables as broomrape for instance, I think that the static meaning (environmental variance) would be more appropriate: CORRECT, WE WERE USING THE CONCEPT OF DYNAMIC STABILITY, INDICATED BY PROXIMITY TO TEA_a (=IDEAL ENVIRONMENT, CALCULATED AS AVERAGE OF ALL ENVIRONMENTS). WE NOW ADDED ENVIRONMENTAL VARIANCE TO ALL TABLES TO SHOW STATIC STABILITY.

Title: It could refer to the control of broomrape, which takes an important place in the article. OK, TITLE WAS MODIFIED ACCORDINGLY.

Abstract. l. 14. remove “However”. OK, DONE

would be useful to precise here the number of trials (= 9). OK, DONE

Broomrape infection shows little influence of powdery mildew and ascochyta blight. But the soil characteristics and the nitrogen status of the plants could complete the diagnosis. RIGHT, BUT UNFORTUNATLY WE DO NOT HAVE SUCH ASSESSMENTS

Introduction. l. 71. Is the scoring method correct? NOT SURE ABOUT THE MEANING OF THIS QUESTION. WE WERE NOT REFERRING TO ANY SPECIFIC SCORING METHOD IN THIS PARAGRAPH, BUT JUST TO THE TRAIT. ANY CASE, THE SENTENCE WAS REFORMULATED

Does the citation for references 14-16 include reference 15? YES, OTHERWISE IT WOULD BE [14,16]. THIS STYLE IS NOT MY CHOICE, BUT IS IMPOSED BY JOURNAL RULES

A reference for GGE biplot analysis would be useful. REFERENCE ADDED: Yan & Holland 2010

Same question as for line 72, for the references 20-23. SAME ANSWER: THIS STYLE OF CITING REFERENCES IS IMPOSED BY THE JOURNAL. [12, 20-23] = 12,20,21,22,23

They are 9 environments (3 locations x 3 years), and not 6. This is how it is presented and discussed in Supplementary Table 1. ABSOLUTELY RIGHT!. CORRECTED

Materials and Methods. 2.1. Plant material and experimental design. l. 91-92. The differences between both locations Cordoba-Almezos and Cordoba-Cortijo seem to be not so important, as they have the same soil type and composition, very close climatic conditions, and the same high level of broomrape presence (Table 2). Is the statement of contrasting environments well supported?. “CONTRASTING” WAS DELETED. ANYHOW, THESE TWO SITES ARE INDEED CLOSE, BUT WE FOUND THEM HELPFUL, BOTH HAVING HIGH BROOMRAPE INFECTION BUT AT DIFFERENT LEVELS. INDEED IT WOULD HAD BEEN NICER TESTING THE LINES IN DIFFERENT COUNTRIES AS WE DID IN THE PAST, BUT THIS WAS NOT POSSIBLE THIS TIME AND WE HAD TO STICK TO NEARBY SITES THAT WE CAN VISIT WITHIN THE DAY. THESE THREE WERE CONTRASTING IN LEVELS OF INFECTION

OK for the 3rd site which has different sol characteristics and a contrasting level of broomrape infestation. INDEED THIS SITE WAS A BIT MORE CONTRASTING, BUT WE WERE HAPPY HAVING THE OTHER TWO AS WELL, AND WE CONTINUE GIVING PRIORITY TO THESE TWO SITES FOR CURRENT SELECTIONS

In addition for Table 2, it would be useful to have de soil depth or water reserve, as water stress could be a major factor in site differentiation. RIGHT, UNFORTUNATELY WE DO NOT HAVE SUCH ASSESSMENTS

Table 1. In the elite cultivars are French winter cultivars, and they are relatively old. By construction, comparison with recent lines may favor these lines. WELL… WE DO NOT QUITE AGREE ON THIS, AS SHOWN BY THE FACT THAT IT WAS THE OLDER CULTIVAR (MESSIRE) THE ONE YIELDING BETTER. THEREFORE, BEING A BIT OLDER OR BEING THE MOST RECENTLY RELEASED IN NOT THE POINT HERE. ON THE CONTRARY, THE POINT IS THAT LINES DEVELOPED (SELECTED) NOT TARGETING THESE SPECIFIC CONDITIONS (AUTUMN SOWINGS UNDER DROUGHT AND BROOMRAPE PRONE CONDITIONS) UNLIKELY WILL RANK BETTER THAN LINES WITH BROOMRAPE RESISTANCE  AND SELECTED SPECIFICALLY FOR THESE CONDITIONS. WE INCLUDED OLD AND NEW CULTIVARS THAT WERE IN USE IN THE REGION, OR BEING TESTED FOR RECOMMENDATION LISTS. THE PURPOSE WAS JUST TO HAVE SOME QUITE DISTINCT ONES AND NOT SO MUCH TO BE VERY DETAILED IN THESE SPECIFIC COMPARISONS. THE MAJOR POINT IS ADAPTATION TO RAINFED MEDITERRANEAN CONDITIONS UNDER BROOMRAPE INFESTATION, CONDITIONS FOR WHICH NONE HAS BEEN BREED. IN FACT, AS SAID, THE “BEST” HIGHER YIELDING WAS THE OLDEST ONE (MESSIRE. TABLE 3). THIS REINFORCE THE VIEW THAT IT IS NOT A POINT OF OLD OR NEW CULTIVAR, BUT A POINT OF BEING BREED FOR THESE CONDITIONS OR NOT. IN ORDER TO AVOID CONFUSIONS, WE REMOVED “ELLITE” CULTIVARS, RAPLACING IT BY CHECK CULTIVARS. IN CURRENT STUDIES WE ARE COMPARING LINES TO BE RELEASED WITH A RANGE OF MODERN CULTIVARS. BUT HERE THE OBJECTIVE IS DIFFERENT.

Further in the text (for instance at line 284) you mention "normal leaf” accessions (I suppose in opposition to Afila ones). It would also be useful to indicate this characteristic in Table 1. AS INDICATED IN THE MENTIONED LINE 284, ONLY J4, J20 AND MESSIRE WERE NORMAL LEAF TYPES, ALL OTHERS BEING SEMI-AFILA. WE TRIED TO HAVE TABLE 1 SIMPLE. THE MS IS ALREADY TOO LONG. 

Precise that the more detailed data are presented in Supplementary Table 1. OK, DONE

2.2 Assessments. l. 113. About the crop stature: Why not keep the two components of this indicator separate? This would give a better indication of lodging in particular. THIS IS HOW ASSESMENTS WERE MADE. WE SCORED CROP STATURE AND CROP APPEARANCE, BUT NOT LODGING SEPARATELY. LODGING COULD BE STIMATED, BUT AS LONG AS WAS NOT DIRECTLY ASSESSED WE PREFER TO KEEP DATA AS COLLECTED. FROM THIS SEASON WE ARE ASSESSING THEM SEPARATELY

114-121. The “crop appearance” is a subjective appreciation, and the scale used is a discrete one. It is not a continuous variate, do you make it as continuous one in order to calculate correlations by using mean of several observations?. IT IS TRUE THAT IS SUBJECTIVE APPRECIATION AS ESPLAINED IN LINES 120-130. HOWEVER, WE DO NOT AGREE THAT IT CANNOT BE HANDLED AS A CONTINUOS VARIABLE. IN FACT IF IS A CONTINUOUS ONE, EVEN WHEN FOR SIMPLICITY WE SCORED FROM 1 TO 5, ANY INTERMEDIATE LEVELS ARE POSSIBLE. ANYCASE, NORMAN (2010) DIRECTLY ADDRESSED THE ISSUE OF ROBUSTNESS AND EXPLORED THE IMPACT OF THREE CHARACTERISTICS: THE SAMPLE SIZE, THE NON-NORMALITY, AND ORDINAL-LEVEL MEASUREMENT, ON THE USE OF PARAMETRIC METHODS. HE CONCLUDE THAT PARAMETRIC STATISTICS CAN BE USED WITH DISCRETE DATA, WITH SMALL SAMPLE SIZES, WITH UNEQUAL VARIANCES, AND WITH NON-NORMAL DISTRIBUTIONS, WITH NO FEAR OF ‘‘COMING TO THE WRONG CONCLUSION’’. THESE FINDINGS ARE CONSISTENT WITH EMPIRICAL LITERATURE DATING BACK NEARLY 80 YEARS.

Norman, G. 2010. LIkert scales, levels of measurement and the “laws” of statistics. Adv in Health Sci Educ., 15, 625-632. https://doi.org/10.1007/s10459-010-9222-y

2.3 Statistical analysis. l. 133. OK with the Arcsine transformation. But were the transformation satisfying?  WE TESTED SQRT(X), ARCSIN(SQRT(X/100)), ARCOSIN(X/100) AND X² TRANSFORMATIONS AND FOUND ARCSIN(SQRT(X/100)) THE MOST ADEQUATE ONE

l.149. Remove first expression “selection for”. OK, REMOVED

The index does not allow to introduce a ponderation for the different factors? It would have allowed you to discuss the relative importance you assign to each factor. No it does not. All we could do was to introduce only the 4 traits with more effects on yield, leaving aside the two with less. WE ONLY FOUND A WAY TO WEIGHT STABILITY AND PERFORMANCE OF EACH TRAIT (DOI: 10.13140/RG.2.2.13931.49448), BUT NOT TO WEIGHT TRAITS. NEW FIGURE IS PROVIDED

Results. l. 179. It is not the supplementary Table 1, but the suppl Table 2. OK, CORRECTED

Lines 180-182. I can't find the values indicated in the text when I look at table 2: For environmental effect on grain yield and biomass: 60 and 61% instead of 53-52% indicated in the text (l.180), and 3% for powdery mildew, not 4% (l. 181) % . For G*E interaction effect for broomrape: 33% in suppl Table 2 instead of 24% indicated in the text, 23% for grain yield instead of 19%, 22% for plant biomass instead of 19%. OK, CORRECTED

Figure 1. Please, can you clarify the meaning of the acronym Eta2% in the title? The figure shows the magnitude of the effects for each variable, but does not provide access to the exact values, which we can see in suppl Table 2. OK, THE TITLE HAS BEEN CHANGED TO “PROPORTION OF VARIANCE ASSOCIATED (ETA² %) OF GENOTYPE, ENVIRONMENT AND GENOTYPE BY ENVIRONMENT INTERACTION AND ERRORS IN ANOVA TO GENERAL PARAMETERS VARIABILITY IN PEA”

Lines 184-185. The values for (G+GE)/(E+G+GE) sum of squares are not exact. The lowest value is 39% for dry biomass and 97% for powdery mildew. RIGHT, THIS HAS BEEN CORRECTED IN FIGURES

Yang et al is not the ref 32, but ref 33. OK, CORRECTED

Supplementary Table 1. Attention to the label of the columns which is not cut at the right place: You should write Pre-Tmax instead of PreTm-ax and so on for the other headings. OK, CORRECTED

Supplementary Table 2. I would prefer having the complete table of variance analysis, with the mean square for every effect in the model, and for the residual, giving us the precision of each trial (via the R² value). We have only a rough idea with Figure 1. OK, NEW VERSION OF SUPPLEM. TABLE 2 PROVIDED

Grain yield l. 200. Please, make a referral to Table 11 for the correlation. Same at l. 255, l. 347. DONE.

Lines 208. “distinctive groups of environments were identified”. NOT SURE WHAT THE REVIEWER WANTED US TO DO HERE, AS IT WAS WRITTEN EXACTLY LIKE THIS ALREADY. PROBABLY THE REQUEST WAS TO DELETE “TWO” AT THE BEGINNING OF THE SENTENCE. WE ARE NOT SURE THIS IS NEEDED, BUT DID IT, JUST IN CASE, AS IT DOES NOT HARM.

Lines 215-216. Here you are talking about relative stability (one genotype compared to others), or dynamic stability according to Becker et Léon (1988). But it would have been interesting to look at static stability, showing whether a genotype maintains the same level of production across different environments, independently of other genotypes. For instance line NS21 shows an interesting static stability while being of a correct level of productivity. In comparison, the lines J20, J4 and NS16 are very stable in a static meaning, but with a very low productivity. a line like NS21 can be interesting because it is polyvalent and not specialized for one type of environment. Same remark for the other variables studied (biomass l. 242 and 245, where we can see that NS21 is also very stable for biomass, with a high level of production)…: see my remarks following. Thus a suggest to add a column in each Table 3 to 9, and also supplementary Table 3, with the environmental variance of the variable considered. OK, DONE. AS EXPLAINED ABOVE, WE CONSIDERED STABLE THE LINES CLOSE TO EXIS TEA_a (=IDEAL ENVIRONMENT, CALCULATED AS AVERAGE OF ALL ENVIRONMENTS), THEREFORE, WE WERE INDEED REFERRING TO DINAMIC STABILITY. NOW WE ALSO ADDED ENVIRONMENTAL VARIANCE TO ALL TABLES TO REFER TO STATIC STABILITY, AS SUGGESTED.

Crop biomass at harvest In Table 4, the second line mean is false: 8667 instead of 9016, but the grand mean is correct. OK, CORRECTED

Flowering date l. 255. Refer to Table 11. DONE

Crop stature l. 276. On Figure 5, the green circle includes the NS34 line, which is not consistent with the comment in the text. OK, CORRECTED

Crop appearance l. 284. You should give the information in Table 1 on the character “Normal leaf", “semi-leafless”… genotypes. IN TABLE 1 WE GIVE ONLY THE PEDIGREE OF THE LINES, DESCRIBING THEN LATER. ONLY THESE THREE (J4, J20 AND MESSIRE) ARE NORMAL LEAF, AS EXPLAINED HERE.

Lines 287. “Enduro” instead of “Endure”. OK, CORRECTED

Line 295. Again an evocation of stability. CLARIFIED NOW

Broomrape infection l. 306. Yes, the environmental conditions play a role on broomrape infection. You could refer to the results presented in figure 10. WE CANNOT YET INTRODUCE FIG10, OTHERWISE, WE SHOULD PLACE IT HERE

Line 319. Again you mention the stability here, in the dynamic meaning. However, aren’t the interesting lines the ones with a low level of broomrape regardless of environment? Thus static stability is to be considered here. CLARIFIED NOW

Powdery mildew l. 329. What does DS mean? Please, clarify. CLARIFIED IN THE TEXT NOW, DISEASE SEVERITY

Line 330. You should add NS34 too. OK, CORRECTED

In Table 9, certain values are false for Kayanne (997) and Bableca (927 or 997): I think a point is missing. And genotypes are not properly sorted by the increasing note of powdery mildew: line NS1 should be placed below the line NS34. OK, CORRECTED

Line 335. There is no line NS7 in the panel: NS7 must be removed from the text. OK, CORRECTED

Lines 336-337. Nevertheless, some lines show instability for powdery mildew, even if this is not visible on the graph: NS22, the most productive one, shows a significant instability, both in the static and dynamic meanings. THIS FIGURE HAS BEEN MODIFIED TO ALLOW BETTER VISUALIZATION

Ascochyta blight. l. 345. There is no line J40, I suppose you mean J4. OK, CORRECTED

Line 347. The correlation between ascochyta and crop stature is 0.74 in Table 11. Please refer to Table 11, so that we can check. INDICATED

Multi-trait stability Index (MTSI) l. 350. Selected genotypes don’t include NS22, which is the best for yield and biomass at harvest, but also for broomrape infection, with a good static stability across different environments. I think that a supplementary table with the results of the calculation of Waasby and MTSI would be useful (it would facilitate the understanding of lines 358 and followig). RECALCULATED AND CLARIFIED NOW.SUPPLEMENTATY TABLES 9 AND 10 HAVE BEEN ADDED SHOWING THIS.

Table 10. Grain yield, Crop appearance. CORRECTED. NOTICE THAT THIS IS NOW TABLE 5

Correlations between traits and Non-Metric Multi-Dimensional Scaling Ordination (NMDS). l. 381. Remove “of” (before powdery mildew). CORRECTED

Line 399. What about soil characteristics? THERE IS LITTLE TO CONCLUDE ON THIS

On Figure 10 and 11, the original axes would be useful. WE PROVIDE THE AXES AS GIVEN BY PAST PROGRAM

Line 416. “spring” rain. CORRECTED

Discussion l. 493. Please clarify the term MAS (marker assisted selection). THIS WAS CLARIFIED A FEW SENTENCES ABOVE, IN THE SAME PARAGRAPH

Line 512 and following. Powdery mildew is also favored by high level of N nutrition, which is related to soil characteristics. RIGHT, BUT AS INDICATED, INFECTION OCCURRED RATHER LATE, WITH NO GREAT EFFECT ON YIELD. N NUTRITION WOULD INDEED AFFECT A LITTLE THE LEVEL OF INFECTION IN THE VERY SUSCEPTIBLE LINES, BUT WE DID NOT PAY ATTENTION TO THESE SMALL VARIATIONS, AS IN FACT WHAT WE HAD WAS HIGHLY R VS HIGHLY S, AND ANYHOW, EVEN IN THESE HIGHLY S, INFECTION AFFECTED LITTLE AS IT OCCURRED LATE

lINE 521. A point should be removed in “worthy to notice”. CORRECTED

Lnies 543 to 547. OK, this is an important point for breeders. See remark on l. 157: I am surprised that no ponderation is proposed in the MTSI index. WE TOO, AND WE FIND A WEAKNESS AND WE COMMENT ON THIS. WE COULD NOT PONDERATE THE TRAITS AND ALL WE COULD DO WAS TO REMOVE FROM THE ANALYSIS THE TWO TRAITS LESS INFLUENCIAL ON YIELD. WHAT WE CAN IS TO PRIORITIZE THOSE GENOTYPES WITH THE HIGHEST YIELD IN DETRIMENT OF STABLE GENOTYPES FOR THIS TRAIT

Line 556. “marker” instead of maker”. CORRECTED

Bibliography. It is OK, but unless I forget, I did not find references 17 and 21 in the text. THIS WAS RIGHT FOR 17. CORRECTED NOW

There is a significant number of self-citations. THIS MIGHT BE RIGHT, BUT FOR SELF-DEFENSE WE CAN ARGUE THAT THERE ARE NOT MANY OTHER GROUPS WORKING (NONE PUBLISHING) ON BROOMRAPE RESISTANCE IN PEA, AND FEW ON BREEDING FOR ADAPTATION TO MEDITERRANEAN ENVIRONMENTS. SOME (self-)REFERENCES WERE REMOVED. STILL, WE REMAIN MOST OPEN TO REMOVE ANY REFERENCE CONSIDERED REDUNDANT, AND TO INCLUDE ANY ONE CONSIDERED MISSING.

Reviewer 2 Report

In this article, Rubiales et al present a multi-environment field testing to evaluate and compare the agronomic performances and disease resistance (in particular broomrape) of 19 pea breeding lines with elite cultivars. Genotypic stability and performances were evaluated by HA-GGE biplot analysis and genotypes were ranked according to a multi-trait stability index. The genetic gain was positive for all the traits examined.

Overall the paper is good, and the results showing genetic gain for the assessed traits are interesting and promising for pea breeding adapted for Mediterranean climate. However the format and structure of the paper must be improved, and the results better discussed. The discussion is currently too long considering that a substantial part of it consists of general considerations on pea breeding that have more their place in an introduction. In this context, the “past history” of the breeding lines evaluated in the paper should be described in the introduction (and/or materials and methods) rather than in the discussion (see paragraph starting on line 476). On the other hand, results should be better discussed and the findings described in the article should be better highlighted (and the corresponding figures/tables cited, e.g. lines 465, 473, 513, 517 etc…). I would also appreciate a more detailed discussion on the MTS index. I was surprised to see that genotype NS22 was ranked almost last (among the breeding lines), while it ranked first for yield and dry biomass… Also discuss how to use this genotype ranking for further selection/advice to farmer (this aspect is –too- briefly addressed on line 542-545) The lines were selected for broomrape resistance (Rubiales et al., 2021), and further crossed to elite cultivar to improved standing ability and plant vigour. Did the authors achieve their goal? Could they comment on this?

I have also several comments on the “materials and methods” and “results” sections (see below) and the whole manuscript should be checked for typing errors (I have listed a few of them below, but please reread carefully your manuscript). 

Abstract is too long (it should be a total of about 200 words maximum) and should be a single paragraph. Both background and results could be reduced. Explain the term “crop appearance” (which is not intuitive neither common) or remove.

Figures and tables should have a short explanatory and informative title. Abbreviations and/or extra explanations should be given in the caption and not in the title (example of Figure S1). Please check also the title of your figures and tables for consistency (e.g. number of environments and/or pea lines studied in Figures 7 and 8).

Introduction: Better introduce the breeding program that led to the breeding lines tested here (some of these information are currently in the discussion).  Alternatively

Introduce the WASSBY/MTSI index.

Materials and Methods:

- In “plant material and experimental design”: the 19 breeding lines should be better described (see above). Consider to give extra explanations here and/or in the introduction.

- Table 2: mention in the caption that the environments were selected for their known broomrape incidence and explain the abbreviation “Oc”. The “A” (in superscript) in the column “soil type“ is not explained. In the caption, make reference to Table S1 to explain that more detailed climatic data are given, and remove sentence “more detailed data used in Figs 9 and 10 are not listed here”.

- In “assessments”. Mention in this section if some traits were not recorded for all 9 environments. Powdery mildew assessments is not described here.

In “statistical analysis”, the sentence line 149-150 is not clear. Please rephrase. For the calculation of the WASSBY index, give the weight for the response variable and for the WASSB that were used. What are the weight chosen for performance and stability? Cite the formula.

Results:

- In this section each part should be better introduced. A short sentence summarizing the results for each paragraph would facilitate the reading of the manuscript. There are also too many Tables and Figures (11 of each!). I suggest to move Tables 4 to 9 and Table 11 to the supp data. Correlation between traits (lines 379-384, Table 11) should be introduced at the beginning of the results as there are mentioned at the beginning of this section (ex: line 200).

- Line 179, cite table S2 instead of S1.

- In Table S2:  I could not find the “c” superscript within the table. For grain yield, use kg instead of Kg. For powdery mildew infection, the sum square for the G is not correctly placed. Add a columns with the (G+G*E)/(E+G+G*E) values and then make reference to Table S2 on line 187. Give the unit for broomrape infection (and remove the “s”).

- Figure 1: please add a title for the horizontal axis and clarify “eta2%” in the title.

- Line 210: Figure 5 is not appropriately cited.

- Line 255, make reference to the actual data (Table 11).

- Table 10 and line 365: why only six environments were used to calculate the genetic gain? What about the MTSI?

- Line 390, better introduce the NMDS analysis (one sentence to explain the purpose of the analysis; for instance by moving the sentence on line 165 from materials and methods).

Discussion

Please see my comments above to improve the discussion.

Other remarks:

The paragraph on Aschochyta blight (lines 527-539) is quite long, especially regarding the few results presented. Consider.

In one of their last article evaluating the adaptation of grasspea to Mediterranean environments, the authors used AMMI analysis instead HA-GGE biplots. Please comment the differences between both method and why using one or the other regarding adaptability and phenotypic stability.

Other minor comments:

Line 54: sown instead of shown

Line 80: autumn instead autum

Line 83: six contrasting environment? I guess it is nine…

Line 139, explain HA

Check the whole manuscript for “kg ha-1” that should be replaced by “kg ha^-1” (in particular check the paragraph, starting on line 228 + Figure3)

Line307: “It” instead of “it”

Line 329: explain DS

Table 11: “Pearson correlations among assessed traits” instead of “accessed”.

Line 394 and 411: length instead of “lenght”

Line398: remove coma after “whereas”.

Line 420: a parenthesis is missing.

Author Response

Overall the paper is good, and the results showing genetic gain for the assessed traits are interesting and promising for pea breeding adapted for Mediterranean climate. However the format and structure of the paper must be improved, and the results better discussed. The discussion is currently too long considering that a substantial part of it consists of general considerations on pea breeding that have more their place in an introduction. In this context, the “past history” of the breeding lines evaluated in the paper should be described in the introduction (and/or materials and methods) rather than in the discussion (see paragraph starting on line 476). On the other hand, results should be better discussed and the findings described in the article should be better highlighted (and the corresponding figures/tables cited, e.g. lines 465, 473, 513, 517 etc…). I would also appreciate a more detailed discussion on the MTS index. I was surprised to see that genotype NS22 was ranked almost last (among the breeding lines), while it ranked first for yield and dry biomass… Also discuss how to use this genotype ranking for further selection/advice to farmer (this aspect is –too- briefly addressed on line 542-545) ALL THIS WAS REVISED

The lines were selected for broomrape resistance (Rubiales et al., 2021), and further crossed to elite cultivar to improved standing ability and plant vigour. Did the authors achieve their goal? Could they comment on this? ALTHOUGH RECENT, THE PAPER MENTIONED REPORTED THE ACHIEVEMENTS OF ACTIVITIES PERFORMED BEFORE 2010. BEST LINES REPORTED THERE WERE J4 AND J20. AS SEEN FROM THE TRIALS (2018-2020) REPORTED HERE, THEY ARE INDEED HIGHLY RESISTANT TO BROOMRAPE, BUT AMONG THE WORSE AGRONOMICALLY. AS EXPLAINED IN THE INTRODUCTION (REVISED NOW FOR CLARITY) THIS IS WHAT WE TRIED TO IMPROVE DURING LAST 10 YEARS AND THE RESULTING LINES ARE REPORTED HERE. AS SEEN IN COLUMN TWO OF TABLE 1, J4 AND J20 ARE IN THE PEDIGREE OF THE NEW NS-LINES REPORTED HERE. DID WE SUCCEED IN IMPROVING STANDING ABILITY?. THIS IS WHAT WE DESCRIBE IN THIS PAPER: JUST SEE PERFORMANCE OF RESULTING NS-LINES DESCRIBED HERE, COMPARED TO RESISTANT PARENTS J4 AND J20

Abstract is too long (it should be a total of about 200 words maximum) and should be a single paragraph. Both background and results could be reduced. Explain the term “crop appearance” (which is not intuitive neither common) or remove. DONE. ABSTRACT HAS BEEN SHORTENED

Figures and tables should have a short explanatory and informative title. Abbreviations and/or extra explanations should be given in the caption and not in the title (example of Figure S1). Please check also the title of your figures and tables for consistency (e.g. number of environments and/or pea lines studied in Figures 7 and 8). ALL TABLES AND FIGURES WERE REVISED

Introduction: Better introduce the breeding program that led to the breeding lines tested here (some of these information are currently in the discussion). REVISED

 Alternatively Introduce the WASSBY/MTSI index. CLARIFIED

Materials and Methods: - In “plant material and experimental design”: the 19 breeding lines should be better described (see above). Consider to give extra explanations here and/or in the introduction. SOME ADDITIONAL DETAILS ADDED. WHAT WE WANTED WAS TO COMPARE THEM IN MULTIPLE ENVIRONMENTS, AS DONE IN ALREADY QUITE A LONG NUMBER OF TABLES. IN THIS PAPER WE JUST WANTED TO PROVIDE THE PEDIGREE, AS DONE IN TABLE 1, AND THEN COMPARE THEIR PERFORMANCE IN THE REPORTED ENVIRONMENTS. IN A NEAR FUTURE, WE INTEND TO SUBMIT FOR REGISTRATION (PROBABLY JOURNAL OF PLANT REGISTRATION) SOME OF THE LINES, AND THAT WILL BE THE TIME TO DESCRIBE THE BREEDING PROCESS

- Table 2: mention in the caption that the environments were selected for their known broomrape incidence and explain the abbreviation “Oc”. The “A” (in superscript) in the column “soil type“ is not explained. In the caption, make reference to Table S1 to explain that more detailed climatic data are given, and remove sentence “more detailed data used in Figs 9 and 10 are not listed here”. OK, DONE

- In “assessments”. Mention in this section if some traits were not recorded for all 9 environments. Powdery mildew assessments is not described here. OK, DONE

In “statistical analysis”, the sentence line 149-150 is not clear. Please rephrase. For the calculation of the WASSBY index, give the weight for the response variable and for the WASSB that were used. What are the weight chosen for performance and stability? Cite the formula. CLARIFIED

Results: - In this section each part should be better introduced. A short sentence summarizing the results for each paragraph would facilitate the reading of the manuscript. There are also too many Tables and Figures (11 of each!). I suggest to move Tables 4 to 9 and Table 11 to the supp data. Correlation between traits (lines 379-384, Table 11) should be introduced at the beginning of the results as there are mentioned at the beginning of this section (ex: line 200). OK, DONE

- Line 179, cite table S2 instead of S1. OK, DONE

- In Table S2: I could not find the “c” superscript within the table. For grain yield, use kg instead of Kg. For powdery mildew infection, the sum square for the G is not correctly placed. Add a columns with the (G+G*E)/(E+G+G*E) values and then make reference to Table S2 on line 187. Give the unit for broomrape infection (and remove the “s”). OK, DONE

- Figure 1: please add a title for the horizontal axis and clarify “eta2%” in the title.  OK, DONE

- Line 210: Figure 5 is not appropriately cited. OK, CORRECTED

- Line 255, make reference to the actual data (Table 11). DONE

- Table 10 and line 365: why only six environments were used to calculate the genetic gain? What about the MTSI? BECAUSE SOME TRAITS COULD NOT BE ASSESSED AT ALL ENVIRONMENTS

- Line 390, better introduce the NMDS analysis (one sentence to explain the purpose of the analysis; for instance by moving the sentence on line 165 from materials and methods). OK, DONE

The paragraph on Aschochyta blight (lines 527-539) is quite long, especially regarding the few results presented. Consider. OK, SHORTENED

In one of their last article evaluating the adaptation of grasspea to Mediterranean environments, the authors used AMMI analysis instead HA-GGE biplots. Please comment the differences between both method and why using one or the other regarding adaptability and phenotypic stability. WE PREFER HA-GGE BIPLOT. HOWEVER, IT IS MORE RESTRICTIVE AND CAN ONLY BE USED WHEN SOME REQUIREMENTS ARE MET (YANG ET AL 2009: THE fiRST TWO PCS SHOULD BE HIGHER THAN 60% AND (G + G*E) / (E + G + G*E) RATIO SHOULD BE HIGHER THAN 10%). WHEN THESE REQUIREMENTS ARE NOT MET, WE CAN NOT USE HA-GGE AND USE AMMI

Other minor comments:

Line 54: sown instead of shown. OK, CORRECTED

Line 80: autumn instead autum. OK, CORRECTED

Line 83: six contrasting environment? I guess it is nine… OK, CORRECTED

Line 139, explain HA. OK, DONE. HERITABILITY-ADJUSTED GGE BIPLOT (HA-GGE)

Check the whole manuscript for “kg ha-1” that should be replaced by “kg ha^-1” (in particular check the paragraph, starting on line 228 + Figure3). OK, CORRECTED

Line307: “It” instead of “it”. OK, CORRECTED

Line 329: explain DS. DISEASE SEVERITY, CLARIFIED

Table 11: “Pearson correlations among assessed traits” instead of “accessed”. OK, CORRECTED

Line 394 and 411: length instead of “lenght”. OK, CORRECTED

Line 398: remove coma after “whereas”. OK, CORRECTED

Line 420: a parenthesis is missing. OK, CORRECTED