Evolution during Three Ripening Stages of Évora Cheese

Round 1

Reviewer 1 Report

Work is relevant but presents several methodological pitfalls that should be revised.

The scientific outputs must be more explicite and I recommend to modify the article to be less descriptive.

Please find in the pdf document all the detailed remarks.

Comments for author File: Comments.pdf

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

 The paper entitled “Differences in chemical and sensory parameters in 2 three ripening stages of Évora cheese PDO” aims to deepen the influence of ripening time of Évora cheese on physic-chemical properties and sensory changes.

Despite the few and fragmentary information on this traditional Portuguese cheese, the authors fail to expand its state of the art. In fact, they do not explain why it is important to study the differences of some pivotal physical-chemical variables and specific sensory attributes in relation to the chosen maturing times. Another critical point is sensory analysis, which is used more as a mere training exercise among panelists than to reveal the appearance of sensory notes specific to the maturation time of Evora cheese. This generates a lot of confusion with many parts of the text that need to be reorganized and some rewritten. Somewhat it seemed to me  to read 2 separate manuscripts and neither was intended to achieve the purpose of the work.

The correlations and statistical analysis deserve separate consideration. In fact, the authors confused the significance of a correlation coefficient with its strength. In fact almost all correlations showed low Spearman coefficients and therefore they were absolutely irrelevant. This should have suggested to the authors to use exploratory techniques such as the analysis of the main components in order to search for the variables most associated with ripeningg times.

Finally, the manuscript presents numerous statements that are not directly related to the Evora cheese, generating a misinterpretation of the data.

L21-22. Move this sentence or aggregate it to L23-24.

L21. Which?

L23. Aw (with w as subscript)

L24. What does it mean?

L35-38. This sentence is useless. Please, describe the impact of ewe's milk production on Évora cheese.

L40. “sold from 6 to”.....

L42. How so? Generic sentence.

L46-47. See L42.

L48. Please replace with "coagulated with thistle (Cynara cardunculus L.) aqueous extract".

L49. This coagulant contains cardosins, an aspartic proteinase with specific... and so on.

L50-51. Generic statement. Which specific and technological consequences?

L47-L55. Please, revise it deleting generic sentences in order to improve the knowledge of state of art.

L71-72. Again. Generic and obvious statement. Of course organoleptic attributes are driven by sensory analysis.

L79-80. The aim is not clear. Do the ripening stages represent the ripening time or process phases? If this is true, you should explain it before in the Introduction why these factors are so important for Évora cheese quality and the consumers' purchase intentions.

L101. Please, indicate more details of the electrode

L102. Five (not Fire) Easy Plus pHmeter. Which kind model? Please, also indicate its resolution.

L105-112. The authors should report if the panelist are regular consumers of Évora cheese (I guess so).

L116. In my opinion piquant attribute is trivial for cheese. Indeed, piquant and pungent are very different sensations. It was necessary routinely to train panel using flavour references. Please see this book Improving the Flavour of Cheese. B C Weimer ed.,  Woodhead Publishing, 2007. eBook ISBN: 9781845693053.  What cheese compounds would the pungency associate with?

L120. wide or thick?

L122-126. If I correctly understood the in each session all judge always tasted 3 cheese samples from 1, 2 and 3 groups (different ripening times) in addition to one randomly chosen from the three types of cheese. Well, in this way the authors neglected nuisance factors such as dairy factories or panelist performance, session... and added a further factor (blind replication). This latter might gain significance in order to rightly preserve the anonymity of the samples (In other words, the judges ate 4 samples but actually tasted 3 types).

 Since the experimental hypothesis was to evaluate sensory differences among Évora cheeses during ripening period, a randomized block design for the replicates (sessions) containing the cheese groups from a specific dairy farm (one for each session) would have been more appropriate. Once the basic assumptions for processing the data with ANOVA have been verified, the authors should verify that the mean square of residual effects is very low.

L127-136. Statistics should be reported under a specific paragraph.

L128. Mann-Whitney U test are usually used for non parametric pairwise comparisons. Are you sure?

L134. Why?

L13-195. The order of Result paragraphs lead to confusion. Please, first report physico-chemical data; then the ones from sensory analysis (descriptive following by ICC) and finally describe the correlations (only higher and significant !!!)

L139. Apparently appositive to the previous sentence L125-126.

L149-150. Figure 1 seems to show no difference between group 2 and 3. Please, evaluate the weight of outliers.

L153-164. First of all, tell about the sensory attributes to which the panelists have given a high score. Are they typical of cheese? Then, report the main differences among the samples across ripening.

L162. Table 1. Since the high mean square error  among panelists about a the same attribute, I suggest to use median with IQR. Please, use only the first decimal places. No information about output of K-W test. Mean rank attribute score among groups? Did you refer P-values to Mann-Whitney U Test outputs?  Which group pair did you consider?

L162. Table 1. Piquant remembers a sensory note found in cheeses enriched with hot chili pepper.

L172. Table 2. What kind of statistical test did you use? I suppose you compared the values from groups 1 and 2 to group 3 one by one. Thus, the letters can cause a bit of confusion, because they refer to a multi-comparison test. Furthermore, are you sure that the reported pH values are not at odds with the resolution of the pHmeter?

S/W is expressed as percentage.

L175-195. The authors should put in better evidence the few stronger monotonous relationships regardless their significance level. By contrast, the significance test aims at rejecting or failing to reject the null hypothesis, namely  if the null hypothesis were true, there is less than a 5% chance that the strength of the relationship you found (your ρ coefficient) happened by chance.

L186. harness?

L194. Table 4. Two details of this table immediately jump out at me: the very weak Spearman's coefficients of the associations lactic-pH and salty-NaCl%. Why? Maybe the recorded variations are not so evident?

L198-203. This is not the aim of the work.

L202. 6 or 12? (see L105).

L203-212. These are statements which are irrelevant to the purpose of the work and the results obtained. I invite authors to consider only those aspects that can really explain the progress compared to the state of the art.

L218-221. Generic statement! In my opinion this consideration represent a weak and critical point of the entire work.  Either data tell a great obviousness (ripened cheeses have surely higher intensity of sensory notes), or the panel test was oriented to irrelevant results for the work. I expected that a trained panel should have indicated which attributes were most influenced by ripening time.

L232-233. The putative water loss should be negatively correlated with moisture content. On the other hand, the authors failed to report the weight loss of the cheese during ripening. The change in shape cheese might be due to variations of cheese consistence associated with proteolyisis. Did you also find this in cheese curdled with thistle pistil extract?

L238-245. Trivial considerations. The authors did not use different levels of salt.    They measured chloride content expressing the total content fo chloride salts.

L238. Relationship between firmness adn saltiness or pungency are very weak.

L242-243. This is not true. You did not add to cheese different levels of salt. Anyway, you measured chloride percentage. Moisture and chloride changes are two dependent variables. In my opinion, you are much more interested in the effect of ripening time on these variable changes

L272. I found no data about this in the manuscript.

L272. thistle extract

L285-287. Generic considerations

L300-302. Did you verify this pH behavior in Évora cheese?

L303-307. I found no proteolysis data for Évora cheese

L323-334. The conclusion reflects the ambiguous approach of the manuscript. On the one hand Évora cheese was evaluated at 3 different advanced stages of ripening, on the other hand the authors insisted on stating that the agreement among panelists increased session after session causing the improvement of panel test. It seems 2 works  that lose sight of the purpose of the work. I suggest the authors to direct the sensory analysis more towards the purpose of the work by highlighting what really emerged from the comparison between Évora cheese at different stages of maturation.

L324. "physico-chemical"

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The authors have substantially improved the manuscript and responded to all my comments and suggestions. Some grammatical errors should still be corrected through a thorough revision of the manuscript; in addition, I found some minor critical points that need to be examined.

L89-93 "This study focused the influence of ripening time on chemical properties and sensory features of Évora cheese in order to contribute to increase the knowledge about the quality of this cheese." In this way it flows better, doesn't it?

L116 Please, delete "pH/mV benchmeter with a"

L122. Replace "that" is with "therefore".

L146-147. Kruskal-Wallis tests for independent samples followed by multicomparison post-hoc tests (considering Bonferroni correction of significance) were used...."

Please see also the comment to your point-by-point replies for L128: What were "the respective post-hoc tests"? Maybe, non parametric one way ANOVA? In this case you properly write: "We used the Kruskal Wallis H test followed by pairwise comparisons of mean rank values. Asymptotic significances (2-sided test) were adjusted with Bonferroni correction for multiple test".

L162-164. Caption of Table 1. Medians (IQR) of physico-chemical determinations among three different ripening groups (why not stages???)

Table 1 Change subtitle "Items" with "Parameters"

L166. In the footnotes (please, also adjust the size): "Kruskall-Wallis test X2 (chi squared) were reported for each parameter (* and **: P < 0.05 and < 0.01) followed by a multiple post-hoc test among pairwised mean ranks. Different superscript letters represent significant differences (P < 0.05).

L196-201. Please, consider the comments for Table 1 and change caption and footnotes accordingly.

L214-216. If I have right understood, I suggest to replace the sentence with "It is important to highlight that due to the outliers presented by the group 2 an apparent low difference between groups 2 and 3 was found."

L421. Please revise the format of references list.

In spite of the excellent bibliographic reference (nevertheless I suggest you to cite Kilcawley K.N. 2017. Cheese Flavour. In: Fundamentals of Cheese Science. P.F. Fox et al., Eds; 2017 Springer, Boston, MA, ch 13, pp. 443-474), I am still not convinced that the piquant is a pertinent attribute for a cheese. Indeed, in the same reference you mentioned, Professor Kieran Kilcawley used only 3 times the term “piquant” in relation to over-ripened hard cheeses. He associated this note to “the short-chain fatty acids (especially butanoic acid) from milk fat triglycerides during maturation; a high level of butanoic acid (1500–2000 mg/kg cheese) is responsible for the peppery, piquant flavour of Romano cheese”. In other words and according to me, he used piquant to explain better the cheese pungency or tangy sensation, although butanoic acid is commonly associated with rancid. a By contrast, pungent (strong irritant) was plenty included among sensory attributes showed in Table 13.2 of the Kilcawley’s chapter and associated with propionic and caprylic acids. Kilcawley found the pungent attribute in surface-ripened cheeses among which I would include Évora cheese.

Thus, after making the required corrections, the manuscript will be ready to be published.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf