Round 1

Reviewer 1 Report

This is an interesting paper. I would like to see this study published, however, in its current form, the article is not ready for publication. While the results are fairly well-elaborated, the text requires additional descriptions to the text. Some main points are detailed below.

I suggest changing the title. Currently, the title is very general.

The abstract also needs modification. There is no description of the method used and the results obtained.

General Comments:
1. Section 1.2: Describe in detail the research conducted in this area. Include those that have helped you in your research. This may be of interest to the readers of the MDPI journal.

2. Section 2.1: Describe in more detail where the remains used come from (archaeological research, when carried out and for what purpose etc.). This is an important chapter in the article.

3. Section 3.1 (and to the entire manuscript ): Is it not worth using additional Latin inscriptions of bone fragments. I use: Buikstra, J.E.; Ubelaker, D.H. Standards for Data Collection from Human Skeletal Remains; Arkansas Archaeological Survey: Fayetteville, AR, USA, 1994.

4. p. 5 – Figure S2 ?

5. Figure 3. Very poor description of the figure. What do the marked groups of poinst mean? What do the colors mean? There are no axis captions. You can explain it in the text.

6. p. 7 – (OnlineMatherials3b)  - this figure should be included in the main text, not as an additional attachment. Additional, similar drawings (fragments of the skeleton) should be added to section 2.

7. p. 7- The scatterplot (Figure4) … - maybe it's worth reminding in the text what the numbered points mean.

8. p. 11 - OnlineMatherials3e-f - I don't have access to this file

9. Figure 7 and 8 – comments as above (my comments in point 7).

10. Conclusions Section: The conclusions section contains no conclusions. In general, the conclusions should be entirely rewritten.

Also, address how the research may be used in other contexts and how, more specifically, future work can improve upon the current results. Could additional human skeletal from diffrent or new archaeological prospection bring something new? Can other methods of comparing data / 3D models used (imaging recognition) ?

Author Response

Estimate Reviewer,

We have finished modifying the paper according to the comments and the questions of all the five reviewers. The new version of the paper (text, figures and tables) is so the result of the changes requested.

Here we report all our answers point by point moreover changes and answers were reported in the comments placed into the track changes file.

We remain at your complete disposal for further clarifications.

Sincerely

The Authors

 

I suggest changing the title. Currently, the title is very general.

AUTHORS: We changed according as requested by the Reviewer1 and Reviewer2.

<< Human Peopling and Population Dynamics in Sicily. Preliminary analysis of the Craniofacial Morphometric Variation from the Paleolithic to the Contemporary Age>>.

The abstract also needs modification. There is no description of the method used and the results obtained.

AUTHORS: We expanded by adding a brief description of the method used and the results obtained, and the simple/size composition as also requested by Reviewer3.

<<…... The approach used allows to identify the main micro-anatomical and micro-evolutionary features. Despite sample size/composition it has been possible to discriminate prehistorical and historical populations.  This study work includes a diachronic morphometrics study of 3D Models of 95 Sicilian skulls coming from 19 populations (from the Paleolithic to the Contemporary Age), providing an overview of human biodiversity and variability in Sicily. To achieve this a geometric morphometrics analyses of the facial features of adult human skulls was performed..........>>

 

 

General Comments:

1. Section 1.2: Describe in detail the research conducted in this area. Include those that have helped you in your research. This may be of interest to the readers of the MDPI journal.

AUTHORS: We expanded the introduction by adding a new dedicated paragraph to mention previous studies and the recognition of the sample.

<<1.2 Backgrounds Studies and Sample recognition. Sicily, in the last century, was characterized by dozens of excavations and archae-ological studies that, shedding to light several finds, signed out significative infor-mation about the history of the island. These studies created important references for the successive paleontological, palaeoecological and osteological studies that in the last 40 years increased a multidis-ciplinary view of Sicily, from its early colonization until nowadays. The sample of skulls analyzed was selected on the basis of some of these studies (Bechtold et al. 1999, Belvedere et al. 2017,…………..>>

2. Section 2.1: Describe in more detail where the remains used come from (archaeological research, when carried out and for what purpose etc.). This is an important chapter in the article.

AUTHORS: In addition to Table 2 the previous archaeological researches were mentioned in the “Backgrounds Studies and Sample recognition” paragraphs reported above.

3. Section 3.1 (and to the entire manuscript ): Is it not worth using additional Latin inscriptions of bone fragments. I use: Buikstra, J.E.; Ubelaker, D.H. Standards for Data Collection from Human Skeletal Remains; Arkansas Archaeological Survey: Fayetteville, AR, USA, 1994.

AUTHORS: All the terms reported use the terminology proposed by Buikstra and Ubelaker 1994 and bone fragments were used only when Buikstra and Ubelaker do not label the points. However we added the refernce, and the pages considered.

4. p. 5 – Figure S2 ?

AUTHORS: Figure S2 display the soft-wireframe but being similar to Figure2 it was removed as also requested by Reviewer4.

5. Figure 3. Very poor description of the figure. What do the marked groups of poinst mean? What do the colors mean? There are no axis captions. You can explain it in the text.

AUTHORS: We extended the captions by adding the legend of the colours.

6. p. 7 – (OnlineMatherials3b) - this figure should be included in the main text, not as an additional attachment. Additional, similar drawings (fragments of the skeleton) should be added to section 2.

AUTHORS:. The OnlineMatherials3b, now renamed Supplementary2, includes tables and graphs distributed in all the results, therefore not having a precise location in the text we prefer positioning it only in the downloadable files.

7. p. 7- The scatterplot (Figure4) … - maybe it's worth reminding in the text what the numbered points mean.

AUTHORS: The Legend was completed to remember at the readers what the numbered points mean.

8. p. 11 - OnlineMatherials3e-f - I don't have access to this file

AUTHORS: Supplementary Matherials3e-f, renamed Supplementary2 after the Reviewer4 comments, is present in the heritage-2129631 word’s file and in the upload files.

9. Figure 7 and 8 – comments as above (my comments in point 7).

AUTHORS: We completed the legend to remember at the reader what the numbered points mean.

10. Conclusions Section: The conclusions section contains no conclusions. In general, the conclusions should be entirely rewritten.

AUTHORS: We expanded the conclusions.

<< The timing and dynamics of the human peopling of Sicily and the island-related issues are subjects of continuous debate and are currently of great interest. The pro-posed approach based on GM, allows to identify the main micro-anatomical and mi-cro-evolutionary features. Despite sample size/composition it was possible discriminate prehistorical and historical populations. GM also allowed topoint out the importance of the coexistence of microevolutionary patterns, population dynamics, and migrations with the latter being one of the main causes of morphological variations (genetic variability - Matsumura et al. 2018)…………..>>

Also, address how the research may be used in other contexts and how, more specifically, future work can improve upon the current results. Could additional human skeletal from diffrent or new archaeological prospection bring something new? Can other methods of comparing data / 3D models used (imaging recognition) ?

AUTHORS: We expanded the Conclusion by adding the proposal for further comprehensive studies a also requested by Reviewer4.

Author Response File: Author Response.docx

Reviewer 2 Report

See pdf file.

Comments for author File: Comments.pdf

Author Response

Estimate Reviewer,

We have finished modifying the paper according to the comments and the questions of the five reviewers. The new version of the paper (text, figures and tables) is so the result of the changes requested.

Here we report all our answers point by point for, moreover changes and answers were reported in the comments placed into the track changes file.

We remain at your complete disposal for further clarifications.

Sincerely

The Authors

 

1. Regarding the sample: it needs to be better described. It is not indicated whether reconstruction/estimation of missing landmarks was necessary, and if it was necessary, it is not indicated which methods were used (mirroring/TPS…). It would be helpful to have a more direct access to the total number of specimens analysed in the study and to the number of specimens per group, this could easily be achieved by turning the table in Figure 1 into a full table. I would also like to have access to more biological information regarding the specimens: gender, age at death…

AUTHORS: The state of sample was already clarified. As requested Reviewer1, Reviewer2, Reviewer4 and Reviewer 5 we have modified and expanded the related sentence by adding the number of specimens and the periods covered, explaining the spatial bias and repositioning the Tables. Moreover the status of preservation and biological information in a dedicated sentence. The paragraphs was also rewritten considering the requests of the Reviewers. Map and Table that constitute Figure1 were separated and the new table modified by removing the unnecessary information and proper relabeling the columns.

<< A set of Sicilian human skeletal remains from different chronologies (Table1) were selected to carry out a 3D craniofacial geometric morphometrics (GM) analysis. Considering the preservation status of the skulls, a dataset of  3D models of 95 human skulls was built representing 19 populations from the Paleolithic to……….>>

2. Regarding the raw data: the landmarks coordinates are not made available with the manuscript, making reproducing the study virtually impossible. The raw data should be published as a table (e.g. excel sheet) in the supplementary.

AUTHORS: Considering that raw coordinates will be used by the authors for further analysis and different aims and that the available literature (i.g. vonCramon, Harvati, Friess, D'Amore, Galland, Bruner etc etc.) do not include this type of information we have decided to not upload a supplementary file containing the raw coordinates.

3. Regarding methodology: I am struggling to understand what the authors intended to do. I am not sure I can follow the step by step protocol used by the authors as information is missing (e.g. what type of data are analysed in each analysis, what are the different shape changes “analyses” referring to -I think they mostly refer to the morphospace from the PCA analysis, but some shape changes plots (Fig. 2) are presented before the PCA in the manuscript. It could refer to the aligned shapes, but then the authors need to specify what it actually refers to (mean vs most deviant specimen from the mean?). This situation is also due to the fact that the authors seem to over interpret the shape change plots of Fig. 2 by giving an “historical” direction to the presented morphological changes, which cannot be done based on such a figure. Additionally, some of the methodology used (MANOVA & CVA) are very poorly introduced and the conditions necessary to obtain robust results are ignored and not tested for (i.e. homoscedasticity, normality, number of variables). Finally, the authors are trying to test complex hypothesis while relying almost exclusively on visual assessments of morphospaces. I feel they could have used additional metrics (e.g. mahalanobils and Euclidean distances that are being calculated when the CVA is run) to further test their hypotheses. In the present version of the manuscript I feel their results are not enough to support most of the claims they are making.

AUTHORS: The methodology was rewritten and expanded clarifying each step of the protocol describing all the analyzes carried out motivating the choice. As usual in all of these studies, we are interested to perform and discuss the best fit between the specimens. Residuals allow the identification of the procustes errors highlighting the worst fit between the specimens that are not part of the aim of our study.

<<…………Raw coordinates were initially subjected to a generalized Procrustes analysis (GPA) (Dryden & Mardia 2016). GPA removes the effects of translation and rotation in the raw coordinates data and standardizes each specimen to unit centroid size (Gower & Payne, 1975; . Rohlf & Slice, 1990; Goodall, 1991). Procrustes fitted coordinates were initially visualized by performing the Shape Changes Graphs displaying the Lollipop Graph and Wireframe Graphs that illustrates the shape changes from a starting shape (the mean shape in the sample) (Klingenberg 2013). These graphs make the resulting three..……….>>

4. Regarding the results: this section should only describe the results. In the present version of the manuscript the authors are interpreting and discussing their results in this section, which in my opinion render makes the text more difficult to follow as some interpretations are made too early in the results while they would need the full results to be, at least partly, supported. The quality of the figures should be improve as it is very difficult to read (understand part of the problem is the formatting for the “reviewing” version of the manuscript).

AUTHORS: As Requested by Reviewer1, Reviewer2, Reviewer4 and Reviewer5 we relabeled the whole “Result” Section deleting any interpretation making it easier to read. Moreover the quality of the figures were increased up to 600 dpi.

Additional comments

Introduction

4. 2, l. 4. “the research goal are: (…)”

AUTHORS: As requested by Reviewer2, Reviewer4 and Reviewer5 research goals were expanded reporting, in the text, intents and hypotheses in dedicated sentence.

9. 2, l. 9. “analyses of facial features”, yet the authors are analysing the upper face along with oart of the calvarium (to the exclusion of the basicranium).

AUTHORS: We have specified in the text that we analyzed the upper face and only part of the calvarium.

2. 2. Table 2. It is not clear what the references are referring to. Additionally, for to sites there are no reference, the authors would need to explain why (have those specimens never been published?). I would actually merge this table with the table present inside Fig. 1.

AUTHORS: As requested by Reviewer1 and Reviewer5 Table2 was modified and repositioned in “Materials” section. Instead of  merging, the new table was positioned into a proper sentence in the “Materials” section. The two sites not reporting references, because not published yet, were specified.

3. 3. Figure 1. The table needs to be removed from the figure.

AUTHORS: The comment reiterates Point1 of the same reviewer:, so related answer are already reported in Point1.

Methods

1. 4, l. 3: replace “though” by through

AUTHORS: The word “though” was replaced with "through".

8. 4, l. 8. Throughout the text: replace “the software “Landmark3.6” (IDAV: Institute (…))” by “the

landmark software (v. 3.6, IDAV)” and cite (Wiley, 2005).

AUTHORS: The sentences were replaced with the suggested reference.

11. 4, l. 11. I would use (Bookstein, 1991) for the types of landmarks. The MorphoJ software is called “MorphoJ” and the reference for the software is (Klingenberg, 2011).

AUTHORS: The sentence was modified with the suggested references.

3. 4, Table 3. The presentation of the numbering of the landmarks according to MorphoJ and to

Landmark software is irrelevant. The landmark software is a tool to collect the data while MorphoJ is a tool to analyse the data. The numbering in Landmark software can also be adjusted to start at 1 not at 0.

AUTHORS: As is possible to see in Supplementary File1 “Landmark3.6” starts numbering from 0 while “MorphoJ” starts with 1. The same picture reports the draw of a skull displaying the landmarks according “MorphoJ” and the captured screen-shot of “Landmark3.6” so we could not delete the column.

17. 4, l. 17. For GPA cite (Gower, 1975; Rohlf and Slice, 1990; Goodall, 1991). (Freidline et al., 2012) is a good paper however it is not where the GPA was described.

AUTHORS: The suggested references were added.

22. 4, l. 19-22. What are a Shape Deformations Graph and a PC Shape Changes Graph? After a GPA it is possible to project the aligned shapes of the analysed sample into a morphospace. Generally, a Principal Component Analysis is first used in order to reduce the dimensionality of the data and to identify axes bearing the most variation within the data. Those axes are the Principal Components. After PCA, each individual/shape included in the analysis obtain a new set of coordinates corresponding to each axis. It is then possible to project each individual on each axis of the PCA, forming morphospaces. To understand and interpret those morphospaces, it is possible to compute the shape change on each of the axis. Generally, along with the graph of the morphospace, the shapes corresponding to the maximum and minimum values of each axis are displayed. To do so, it is possible to use wireframes, or shape deformation (directly using a 3D models being warped into different shapes) or lollipop graphs which are what the authors display in the middle of their Figure 2 (where the letter is missing). Therefore, I am not sure I understand what the authors mean, but it seems to me that different ways of visualising the same shape changes are described as being different methods that are displaying different kind of shape changes. An alternative would be that Fig. 2 wireframes shape changes only refer to the aligned shapes before PCA and that the light blue shape could be the mean and the dark blue shape could be the most extreme of the specimens of the sample. I am not sure this type of plots would bring anything to the manuscript and I am most definitely sure that it cannot be interpreted in terms of chronological variation as the authors do, unfortunatly there is no information explaining this aspect. This paragraph needs to be rewritten.

AUTHORS: Being the shape deformation graphs (soft wireframe) very similar to the standard wireframe, as requested by Revierwer1 and Reviewer5 they were deleted (FigureS2 and FigureS4) and not discussed. According to all Reviewers the captions were completed and modified. Moreover, Wireframe Graphs result, and interpretations were explained in the “Materials” section. The whole paragraph was profoundly modified.

23. 4, l. 23. What do the authors mean by “a covariance-based PCA? Did they compute a covariance matrix from their aligned shape data (i.e. after GPA) before running the PCA? The most common workflow in 3D GMM is to run the PCA from the aligned 3D coordinates right after the PCA.

AUTHORS: The workflow applied  (being already in accordance with the  most common 3D GMM) was clarified in the “Methods” section that , as reported before,  was  significantly modified.

<<…… A Principal Component Analysis (PCA) based on a Covariance Matrix was per-formed on the Procrustes fitted coordinates to study the landmarks positions in a collection of specimens (Hammer & Harper 2008).…….>>

26. 4, l. 25-26. The authors state that they run a MANOVA/CVA analysis in order to “show the differences and similarities between the elements graphically”. MANOVA (Mutlivariate analaysis of variance) and CVA (Canonical Variate Analysis) are not designed to show differences/similarities graphically, although it is possible to plot the results of a CVA. A CVA analysis will try to discriminate predeterminate groups. For the results to be interpreted with confidence, the variance-covariance matrices of the different groups that will be analysed in the CVA must be (more or less) equal: homoscedasticity that can be tested with a Box’s M test for which the data should meet the assumption of multivariate normality. It does not appear in the manuscript that those tests were run. Additionally, the authors do not specify which variables were used to run the MANOVA, the CVA and the neighbour-joining analyses. I assume they have used the PC coordinates, but it is not specified anywhere.

AUTHORS: As written before the new revised and expanded version of the “Matherials” section clarify the MANOVA/CVA and the NJ analyses and the data used.

<<…….PCA, is however an exploratory analysis (Le Maître and Mitteroecker, 2019) based on the visualization, only sufficient to support hypothesis. Theories should be sup-ported by multivariate tests (i.g. MANOVA) and other studies. Following a MANOCA/CVA was carried out on Procrustes fitted coordinates.  Canonical variate analysis (CVA) is a type of discriminant analysis, used with more than two groups, that maximizes separation between the given groups (Bronstein et al. 2006). CVA is mathematically closely tied to the Multivariate ANalysis Of VAriance (MANOVA) that is a multivariate procedure to test the equality of the multivariate means of several multivariate samples (Hammer & Harper 2008). PAST Software (as others software) present the MANOVA as part of the CVA output. Moreover, PAST directly…..>>

Craniofacial Geometric Morphometrics in Sicily

2. 5, 3 first paragraphs & Figure 2. What do the authors mean by “Overall cranial shape variation”? What do the author mean by “PC shape changes” for Fig. 2b? It looks like fig. 2a represents the same shape change as Fig. 2b (mislabelled in Figure 2) except that Fig. 2b is a lollipop type of graphic representation. My interpretation of this graph is that the light blue represent the mean shape of the data and the dark blue represent an extreme of the first PC (either towards the minimum or towards the maximum values of PC1). Fig. 2b is the same but using a different way of showing the shape change. Alternatively, it could be that the light blue represent an extreme of a PC (e.g. negative values) and the dark blue another extreme (e.g. positive values). Looking at supplementary figure 2 which is supposed to display shape change on PC1, it looks possible that the “overall cranial shape variation” would be the shape variation expressed on PC1 (unless it represents the mean shape vs one extreme specimen of the sample after GPA, see supra). Finally, in the last paragraph the authors state that the shape changes displayed on Fig. S2 indicate morphological change from prehistoric populations to Mesolithic populations. The figure does not indicate any of this as it only displays shape change on axis 1 of a PCA which have not been shown yet in the paper and it is not explained to what the displayed shapes in dark and light blue correspond.

AUTHORS: As already explained above, FigureS2 and FigureS4 representing the soft wireframe have, was deleted because they are very similar to the standard wireframe. As concern the Lollipop and Wireframe, dedicated sentences were added on “Methods” section further expanding the captions;  moreover all the related paragraphs in “Results” section were rewritten as requested by Reviewer2, Reviewer4 and Reviewer5. Figure2 and its caption were modified according to the new text.

<< The Lollipop Graph (Figure2a) show the direction and magnitude of the shape variations highlighting that the changes are located mainly on the landmarks placed on the superior jaw (1, 2, 7, 8, 9 and 10), on the left and right sides of the face (landmarks 19, 20, 21 and 22), and on the cranial vault (landmarks 5, 23, 24, 25 and 26). Specifically, the first group of landmarks moves inwards while the others two move outward. With regards to the cranial vault, the parietals and the occipital all change in the direction of the mesocephalization with the parietal bones presenting the wider variation…..>>

3. 5, paragraph 3.2. Here, I am really struggling to follow. The authors are showing shape changes in paragraph 3.1, which I believe are linked to the PCA, and they state at the beginning of paragraph 3.2 that the “analysis was continued by performing PCA and MANOVA/CVA”. The second sentence is also really unclear “considering that sample size and composition are always included in the analysis, the study produced tenable results”.

AUTHORS: Also in this case, according Reviewer2, Reviewer4 and Reviewer5 the section were modified rewriting and clarifying the suggested points.

<<Futher analysis were conducted by performing PCA and MANOVA/CVA of the procustes fitted coordinates. Considering that always sample size and composition (number of specimens and sexual dimophism) influence the analysis the plots obtained produced plausible results ….>>

3. 6, l. 2-3. The eigenvalues and % of variance of the PCA indicates that the first three PCs represent

more of the variation of the data then the rest. Agreed, but why do the authors state that it “denotes

significant variation only between the components of the prehistoric specimens”? I do not understand this statement.

AUTHORS: The sentence were rewritten explaining the requests of the reviewer.

13. 6, l. 7-13. This part should be in the discussion section (although I am not sure the results of the present study are supporting the statement). Figure 3. What do the authors mean by “PC1 vs PC2 of Procrustes Coordinates”? They mean they ran the PCA using the aligned coordinates after GPA? What type of data did they use for the CVA? Did they use the PCs coordinates? Or did they use the aligned 3D coordinates after GPA? If this is the case, I am quite surprised as to run a CVA one should include less variables than the number of specimens in the smallest group. In that case the smallest group is composed of 2 (San Teodoro) and the CVA should be run using only one PC (PC1). Which seems quite unrealistic. I know that MorphoJ will run a CVA no matter what the user uses in terms of data, number of individuals, number of groups and/or variables. Nevertheless, CVA should be run with n variable = n-1 smallest group size and it is difficult to interpret results where those conditions are not met (including normality and homoscedasticity). An alternative would be to use between group PCA which does not need the data to follow these requirements.

AUTHORS: PC1 vs PC2 refer to the two axes and means that the linear combination of the variables is done by PC1 that is the extracts of the maximum variance that is loss and PC2 that is the extracts the maximum variance that is loss. PCA3 is another reduction of the dimensionality of the data but reducing further the information is possible to perform analyses PC1 vs PC2 or PC2 vs PC3 etc. etc. PC3 and other PCS are  not significative as displayed in SupplementaryFiles2. PC1 vs PC2 is a technical abbreviation used in the captions. 

PC1 vs PC2 of Procrustes Coordinates means that the PCA (also MANOVA/CVA) were performed after the GPA and this was explained by expanding the dedicated sentence in the text.

MprphoJ was used only for the Shape Changes Analysis, PCA, MANOVA/CVA and NJ were performed with PAST software.

PAST perform the MANOVA/CVA analysis that  was possible only in this case with 78 variables and 94 specimens. In the latter analyses, performed with averages of the coordinate for each group, specimens are less than variables so MANOVA/CVA not will be performed.

AUTHORS: p. 6, l. 27. “with progressively increasing variability”. The authors do not test for this. There are test that can be run to compare variation between groups (Le Maître and Mitteroecker, 2019).

AUTHORS: PAST (as other software) presents the MANOVA as part of the CVA output and the software automatically perform the Wilk’s lambda Test and the Pillai trace Test (Pillai 1967). Descriptions of both the test, references and the related result were argued in  “Methods”, “Results” and ‘Discussion” sections explaining why these two test were performed and clarifying the meanings of the values obtained. Moreover we clearly state that PCA is exclusively an explorative analysis and that the “progressively increasing variability” is an hypothesis based only on the visualization of the plot interpreted as used in literature (vonCramon, Harvati, Friedline, Bruner etc. etc.). Finally The word “Theories” were removed from “Results”, “Discussion” and “Conclusion” sections and replaced, when appropriate, with the word “Hypothesis”.

7. 7. Most of the text on page seven should be actually moved to the discussion section.

AUTHORS: According to Reviewer2, Reviewer4 and Reviewer5 the “Result” section was rewritten removing any possible discussion about the presented data, and  moving part of the text on the “Discussion” section.

1. 7, 2nd paragraph. “Finally, contemporary specimens (…)” This statement does not seem to be

supported by the data.

AUTHORS: Contemporary specimens are placed in a wide morphospace due to the increase of specimens available, the historical context and modern dispersal patterns. All of these three factors were reported by relabeling the sentence in the text.

1. 7, 5th paragraph. “the previous eigenvalue and percent variance gradually decrease on the PC axes only after PC4” I feel this should be rewritten, as it stands it is not true. The eigenvalue, as the % of variance, always decreases from PC1 to the last PC of the analysis.

AUTHORS: The comment reiterates a previous comment of the same reviewer and, as reported above, the sentence was rewritten.

1. 7, 6th paragraph. “This support the theory that the first significant migratory flow on the island was seen only at the end of the Stone Age”. I do not think that the data supports this statement. There should also be references regarding this hypothesis.

AUTHORS: As already requested “Theory” was replaced with “Hypothesis” and interpreting the plots as usual data displayed clearly supports our statement. As requested proper references were added.

Neighbor-Joining of Procrustes coordinates.

8. 8. Did the authors used the aligned coordinates to run the NJ analysis? Or did they use the PCs? They should also clarify that they are using the mean shape for each represented population.

AUTHORS: NJ as all the analysis were performed whit the Procrustes fitted coordinates and it were specified in “Materials” and “Result” section for all the analyses reported.

8. 8. “the outgroup”. There is no outgroup in this kind of analysis. This is phenetic and an outgroup is flagged as such before the beginning of a cladistics analysis (it used to help providing a “direction” to the morphological changes in the data).

AUTHORS: PAST automatically recognize the oldest group of  San Teodoro as the outgroup of the first two NJ and in the same way the software recognize the Antiquity group as the outgroup of the last NJ.

Prehistory and history: Geometric morphometrics focus

9. 9. I am surprised to find Bronze age and Iron age people grouped into a group called “prehistory”. To my knowledge they are definitely part of history.

AUTHORS: According all the internationals’ conventions the chronological limits of Prehistory’s Periods vary considerably according to the geographical and cultural context. Considering the island’s cultural and geographical context, clearly stated in Table2, Bronze Age is definitely part of Prehistory while Iron Age is borderline between Prehistory and Protohistory (but in Sicily far from the History). Considering the specimens available and the choice to divide our sample into two coherent groups Bronze Age and iron Age were preventively grouped with the prehistorical samples.

11. 11. “inhomogeneous morphospace supports the theory of a discontinuity genetic flows carried by settlers during prehistory”. I disagree with this statement, PCA is an exploratory analysis and it is rarely sufficient to support any hypothesis (the authors should change from theory to hypothesis throughout the text). The authors could have relied on additional metrics (distances between groups within groups, whether Euclidean or Mahalanaobis that are actually calculated when running a CVA) to further test their hypotheses.

AUTHORS: As already requested and subsequently modified the word “Theory” were deleted and replaced in the whole text with the word “Hypothesis” moreover In the “Methods” section we clearly state that PCA is an exploratory analysis, that NJ was performed with Euclidean distances and additionally we report the test automatically performed by PAST after the MANOVA/CVA>

Author Response File: Author Response.docx

Reviewer 3 Report

Dear Editor of Heritage and Authors,

Upon reading the manuscript “The Human Peopling of Sicily. Craniofacial Morphometric Variation from the Paleolithic to the Contemporary Age” I found it both theoretically and technically substantiated, providing an interesting and original perspective regarding the peopling of the island of Sicily, Italy. The authors make a good point regarding the association between cranial shape and successive migrations but I would like a theoretical description regarding other influences on cranial shape, including random genetic drift (see, e.g., Smith. 2011. The Role of Genetic Drift in Shaping Modern Human Cranial Evolution: A Test Using Microevolutionary Modeling. Journal of Evolutionary Biology. https://doi.org/10.4061/2011/145262).

Best regards.

Author Response

Estimate Reviewer,

We have finished modifying the paper according to the comments and the questions of the five reviewers. The new version of the paper (text, figures and tables) is so the result of the changes requested.

Here we report all our answers point by point for, moreover changes and answers were reported in the comments placed into the track changes file.

We remain at your complete disposal for further clarifications.

Sincerely

The Authors

 

Reviewer3

Yes        Can be improved            Must be improved         Not applicable

Is the content succinctly described and contextualized with respect to previous and present theoretical background and empirical research (if applicable) on the topic?

( )          (x)         ( )          ( )

Are all the cited references relevant to the research?

(x)         ( )          ( )          ( )

Are the research design, questions, hypotheses and methods clearly stated?

(x)         ( )          ( )          ( )

Are the arguments and discussion of findings coherent, balanced and compelling?

(x)         ( )          ( )          ( )

For empirical research, are the results clearly presented?

(x)         ( )          ( )          ( )

Is the article adequately referenced?

( )          (x)         ( )          ( )

Are the conclusions thoroughly supported by the results presented in the article or referenced in secondary literature?

(x)         ( )          ( )          ( )

 

Comments and Suggestions for Authors

Dear Editor of Heritage and Authors,

Upon reading the manuscript “The Human Peopling of Sicily. Craniofacial Morphometric Variation from the Paleolithic to the Contemporary Age” I found it both theoretically and technically substantiated, providing an interesting and original perspective regarding the peopling of the island of Sicily, Italy. The authors make a good point regarding the association between cranial shape and successive migrations but I would like a theoretical description regarding other influences on cranial shape, including random genetic drift (see, e.g., Smith. 2011. The Role of Genetic Drift in Shaping Modern Human Cranial Evolution: A Test Using Microevolutionary Modeling. Journal of Evolutionary Biology. https://doi.org/10.4061/2011/145262).

Best regards.

AUTHORS: We have found very interesting the suggested paper so, always considering that we do not have analyzed the cranial vault, the upper face, the masticatory apparatus and the climate diet, which mainly carries the neutral changes we have expanded the text focusing on the influence of these force adding appropriate references.

Discussion <<… When evaluating the variation between human groups arriving to a localized geo-graphical region (like an island), it is important to consider that the genetic pool is often stressed by genetic drift phenomena like bottle neck and founder effect (Manica et al. 2007). In addition to. these stochastic forces, adaptive changes (like masticatory-inducted phenotype) are in parallel impacted with cultural variations with the same plasticity (Harvati & Weaver 2006) but with a slow degree of diversification. In particular, patterns of the cranial vault and the upper face are evolving largely neutrally (Smith 2011) nevertheless the large differentiation of facial shapes during….>>

Author Response File: Author Response.docx

Reviewer 4 Report

The paper presents the results of geometric morphometric study of Sicily population from Paleolithic to the modern times. The research is generally well-conceived with and adequate geometric morphometric approach. However, sample size is not appropriate because authors aimed to describe craniofacial variation from Paleolithic to the modern times with n < 100. Therefore, this study can only be considered “a preliminary” study which limits strong conclusions and hampers full interpretations. Besides, although most particular topics in the manuscript are well described, the manuscript as a whole lacks standard structure of a research paper and should be almost completely rewritten.

Therefore, I suggest that authors revise the manuscript paying attention to the issues raised below.

Title

-        Please include in the title “a preliminary study”. This is necessary because of the sample size and temporal frame covered by the study.

 

Abstract

-        The abstract is written in the form that is too general. What was the material, what was the sample size? I suggest authors to enhance it, provide some “exact” results, and after that main conclusion and limitations*

 

Introduction

-        The introduction section lacks real background. I understand that the present study is probably the first of this type on population samples of Sicily. However, authors should mention some previous studies on other populations that used same approach and explain why it is appropriate and relevant and provide more information of historical context. Tables 1 and 2 must be described with few sentences and provided with more context.

-        The “aim” part should be much shorter, clearer, and simpler, and the reasons why authors conduct study should be completely clear from the text that is written before (First part of the introduction should contain background and contextual information). So, tables 1 and 2 should not be in the aim section. What and how was done should not be in the introduction section but in Materials and methods.

2. Materials and Methods

- Authors should clearly state sample size. It is visible in the table, but it must be summed up in the text.

Title 3 - Results

-        The title 3 should be “Results” and authors can use appropriate subtitles.

-        This section needs and extensive cleaning. It contains relevant presentation of the results, but interpretations are too extensive for results. It is OK that authors explain some results but interpretations and conclusions such as “dolichocephaly that characterized prehistoric populations of northern Europe (like the British and Scandinavians), southern Europe (like the southern Iberian Peninsula and southern Italy) and the Mediterranean islands (such as Sardinia, Corsica and Sicily) changed in direction to a progressive meso-cephalization during the Mesolithic.”could only belong to the discussion. I will list few similar examples below.

-         “Considering that sample size and composition are always included in the analysis, the study produced tenable results.”

-        “The scatter plots confirm the previous theories about the first human peo-pling during the Upper-Palaeolithic which have identified the Würm-Settlers of San Teo-doro as the first humans to enter and establish a permanent settlement in Sicily (D’Amore et al. 2009; Galland et al. 2019)”

-        These were examples, but I suggest clearing all conclusions and comparisons that are not necessary to present the results. All other interpretations should be placed in the discussion section.

-        Figures 3 and 7 lack appropriate legend. It could be added in the figure or described in the text.

-        Authors constantly mention “previous assumptions”, “theories”, and “hypotheses”. If they really exist, they should be elaborated in the introduction, not necessarily as formal hypotheses but they should be mentioned in some form.

 

Discussion

-        The discussion section should begin with most important findings and scientific contribution of the study, please adjust it. The discussion should be more oriented from specific study results, comparisons, and interpretations to general conclusions.

-        All detailed interpretations and comparison to other studies and the previous knowledge should be placed here.

-        Discussion does not mention study limitations, particularly sample size, and proposals for more comprehensive studies. Please add this part.

Conclusion

-        This part is written to generally. It should be more focused to the main findings and what they practically mean.

 

Minor issues

-        Please write full abbreviation first time you mention something in the text, e. g. PCA

-        p 10. PC1 and PC2 of both the groups cover around 90% of the percentage while… Did authors mean 90% of the variance?

-        In many places authors mention “significant”. If it is “statistical significance” it is OK, but, in other instances, I suggest different phrasing, e.g., remarkable changes or similar.

Author Response

Estimate Reviewer,

We have finished modifying the paper according to the comments and the questions of the five reviewers. The new version of the paper (text, figures and tables) is so the result of the changes requested.

Here we report all our answers point by point for, moreover changes and answers were reported in the comments placed into the track changes file.

We remain at your complete disposal for further clarifications.

Sincerely

The Authors

 

Reviewer4

Yes        Can be improved            Must be improved         Not applicable

Is the content succinctly described and contextualized with respect to previous and present theoretical background and empirical research (if applicable) on the topic?

( )          ( )          (x)         ( )

Are all the cited references relevant to the research?

( )          (x)         ( )          ( )

Are the research design, questions, hypotheses and methods clearly stated?

( )          ( )          (x)         ( )

Are the arguments and discussion of findings coherent, balanced and compelling?

( )          ( )          (x)         ( )

For empirical research, are the results clearly presented?

( )          ( )          (x)         ( )

Is the article adequately referenced?

( )          (x)         ( )          ( )

Are the conclusions thoroughly supported by the results presented in the article or referenced in secondary literature?

( )          (x)         ( )          ( )

 

Comments and Suggestions for Authors

The paper presents the results of geometric morphometric study of Sicily population from Paleolithic to the modern times. The research is generally well-conceived with and adequate geometric morphometric approach. However, sample size is not appropriate because authors aimed to describe craniofacial variation from Paleolithic to the modern times with n < 100. Therefore, this study can only be considered “a preliminary” study which limits strong conclusions and hampers full interpretations. Besides, although most particular topics in the manuscript are well described, the manuscript as a whole lacks standard structure of a research paper and should be almost completely rewritten.

Therefore, I suggest that authors revise the manuscript paying attention to the issues raised below.

Title

-        Please include in the title “a preliminary study”. This is necessary because of the sample size and temporal frame covered by the study.

AUTHORS: The temporal frame is complete, temporally the sample covers the whole colonization of the Island. Sicily was colonized at the end of the Upper Palaeolithic and our sample comprehends specimens from the first sentiment of SanTeodoro to the early years after 2.000.

According to the reviewer, the sample size allows a preliminary study, and the word was included in the text rewritten as also requested by Reviewer1.

<< Human Peopling and Population Dynamics in Sicily. Prelimi-nary analysis of the Craniofacial Morphometric Variation from the Paleolithic to the Contemporary Age>>

Abstract

-        The abstract is written in the form that is too general. What was the material, what was the sample size? I suggest authors to enhance it, provide some “exact” results, and after that main conclusion and limitations*

AUTHORS: The abstract was expanded by adding information about the materials, the hypothesis requested, and the results as also asked by Reviewer 1. Limitations of simple size/composition were added in the text.

Introduction

-        The introduction section lacks real background. I understand that the present study is probably the first of this type on population samples of Sicily. However, authors should mention some previous studies on other populations that used same approach and explain why it is appropriate and relevant and provide more information of historical context. Tables 1 and 2 must be described with few sentences and provided with more context.

AUTHORS: The introduction was expanded by adding a new dedicated paragraph to mention previous studies and the recognition of the sample.

<<.1.2 Backgrounds Studies and Sample recognition. Sicily, in the last century, was characterized by dozens of excavations and archaeological studies that, shedding to light several finds, signed out significative information about the history of the island. These studies created important references for the successive paleontological, palaeoecological and osteological studies that in the last 40 years increased a multidisciplinary view of Sicily, from its early colonization until nowadays. The sample of skulls analyzed was selected on the basis of some of these studies (Bechtold et al. 1999, Belvedere et al. 2017, Brea & Cavalier 1965, Cavalier 1995, De Miro 1988, Di Stefano 1997 - 1998, Fama’ & Toti 2019, Griffo 1997 – 2008, Hodos 2010, Kistler 2012 – 2013, La Duca 2000, Mannino 2016, Mölk …………..>>

-        The “aim” part should be much shorter, clearer, and simpler, and the reasons why authors conduct study should be completely clear from the text that is written before (First part of the introduction should contain background and contextual information). So, tables 1 and 2 should not be in the aim section. What and how was done should not be in the introduction section but in Materials and methods.

AUTHORS: Considering the previous information and studies available and evaluating the request of Reviwer4 as concern this we formulated, in a dedicated paragraph, specific research hypotheses treated and discussed in the paper.

<<………All data have been collected and analyzed taking into consideration all historical, archaeological, geological, palaeoecological and paleoanthropological information available.  The intents are to verify the hypotheses that the cranial morphology was modified by the various genetic contributions and more or less related to the intense and significant migratory flows that follow one another from the end of the Upper Paleo-lithic to the Contemporary and clarify the population dynamics through the ages………..>>

 

 

2. Materials and Methods

Authors should clearly state sample size. It is visible in the table, but it must be summed up in the text.

AUTHORS: The state sample was already clarified but we modify the sentence by adding the number of specimens, the periods covered and repositioning Table1 and Table 2. The new paragraph was modified also considering the request of Reviewer1 and Reviewer4.

<<2.1 Cranial Sample << 2.1. Cranial samples A set of Sicilian humans skeletal remains from different chronologies (Table1) were selected to carry out a 3D craniofacial geometric morphometrics (GM) analysis. Considering the preservation status of the skulls, a dataset of 3D models of 95 human skulls was built representing 19 populations from the Paleolithic to the Contemporary Age (Table2). Using the standards proposed by Buikstra & Ubelaker (Buikstra et al. 1994; Ubelaker 1989), were selcted only adult crania also considering their integrity; broken or incomplete specimens (bones not in anatomical connection and/or lacking landmarks necessary to take the anthropometric measurements mentioned below have been excluded before the analysis). Characteristics …...>>

Title 3 - Results

-        The title 3 should be “Results” and authors can use appropriate subtitles.

AUTHORS: The Section3 was relabeled Results and some sentences of the section were rewritten, simplified or deleted.

-        This section needs and extensive cleaning. It contains relevant presentation of the results, but interpretations are too extensive for results. It is OK that authors explain some results but interpretations and conclusions such as “dolichocephaly that characterized prehistoric populations of northern Europe (like the British and Scandinavians), southern Europe (like the southern Iberian Peninsula and southern Italy) and the Mediterranean islands (such as Sardinia, Corsica and Sicily) changed in direction to a progressive meso-cephalization during the Mesolithic.”could only belong to the discussion. I will list few similar examples below.

AUTHORS: Considering the request of the reviewer and the comments of the other reviewer that describe the results “fairly well-elaborated” modified the whole section. “dolicocephaly” or others were moved and discussed in the appropriate section. “Considering that sample size and composition are always included in the analysis, the study produced tenable results.”

AUTHORS: The sentences were rewritten.

<<….. PCA and MANOVA/CVA have been interpreted always considering that sample size and sample composition (number of specimens and sexual dimorphism) can in-fluence the analysis. Nevertheless, the obtained plots produced plausible results..…..>>

-        “The scatter plots confirm the previous theories about the first human peopling during the Upper-Paleolithic which have identified the Würm-Settlers of San Teodoro as the first humans to enter and establish a permanent settlement in Sicily (D’Amore et al. 2009; Galland et al. 2019)”.

AUTHORS: The sentence was modified.

<<….. The pattern obtained for the PCA (Figure3a-b) and the MANOVA/CVA (Fig-ure3c-d) highlights a sudden clear separation between the Würm-Settlers of San Teodoro and the hunter-gatherers of the Mesolithic, themselves separated from the other prehistorical populations. In fact, the some archaic characters….>>

-        These were examples, but I suggest clearing all conclusions and comparisons that are not necessary to present the results. All other interpretations should be placed in the discussion section.

AUTHORS: Considering the request of the reviewer comparisons were discussed in detail only in the discussion.

-        Figures 3 and 7 lack appropriate legend. It could be added in the figure or described in the text.

AUTHORS: The captions of the figure were completed by adding legend, symbols and key.

-        Authors constantly mention “previous assumptions”, “theories”, and “hypotheses”. If they really exist, they should be elaborated in the introduction, not necessarily as formal hypotheses but they should be mentioned in some form.

AUTHORS: Hypotheses were clarified in the Introduction and the terms “previous assumptions” and “theories” deleted and remanded to the discussion.

Discussion

-        The discussion section should begin with most important findings and scientific contribution of the study, please adjust it. The discussion should be more oriented from specific study results, comparisons, and interpretations to general conclusions.

AUTHORS: The section was significantly changed and modified as requested by the reviewer.

<< GM allows to highlight morphological differences (micro-anatomical and micro-evolutionary features) that can be related to the time frame and migratory phenomena that occurred in Sicily from Prehistory to Contemporary age. GM has therefore been used on Sicilian remains to assess the human peopling and the morphological relationships (population influx) (D’Amore et al. 2009 – 2010b; Galland et. Al. 2009) among human groups during Prehistory and History (Reyes-Centeno et al. 2017) …..>>

-        All detailed interpretations and comparison to other studies and the previous knowledge should be placed here.

AUTHORS: The type of analysis and the approach was never applied exhaustively in Sicily for this reason all the possible comparative work are limited but cited in the text.

-        Discussion does not mention study limitations, particularly sample size, and proposals for more comprehensive studies. Please add this part.

AUTHORS: We have already reported in the discussion the lack of adequate findings in some parts of the sample. Proposals for more comprehensive studies were added in the “Conclusion”.

Conclusion

-        This part is written to generally. It should be more focused to the main findings and what they practically mean.

AUTHORS: The section has expanded focusing the main results and proposal for further studies.

<< The timing and dynamics of the human peopling of Sicily and the island-related issues are subjects of continuous debate and are currently of great interest. The pro-posed approach based on GM, allows to identify the main micro-anatomical and mi-cro-evolutionary features. Despite sample size/composition it was possible discriminate prehistorical and historical populations. GM also allowed to point out the importance of the coexistence of microevolutionary patterns, population dynamics, and migrations with the latter being one of the main causes of morphological variations (genetic variability - Matsumura et al. 2018)……..>>

Minor issues

-        Please write full abbreviation first time you mention something in the text, e. g. PCA

AUTHORS: The abbreviation were explained once at the beginning of the text and after standardized in the whole paper.

-        p 10. PC1 and PC2 of both the groups cover around 90% of the percentage while… Did authors mean 90% of the variance?

AUTHORS: The wrong word “percentage was replaced with the correct. “variance”.

-        In many places authors mention “significant”. If it is “statistical significance” it is OK, but, in other instances, I suggest different phrasing, e.g., remarkable changes or similar.

AUTHORS: The word “significant” was replaced with the word  “remarkable” o similar word.

Author Response File: Author Response.docx

Reviewer 5 Report

The paper under review uses geometric morphometrics to examine the population history of Sicily. The paper is interesting but there are certain issues that should be addressed before publication.

1. At the moment the research question is open-end. The authors perform GM to see what patterns emerge. Instead, they should formulate some more specific research hypotheses based on archaeological/historical/geographic data, e.g. ‘we anticipate greater heterogeneity/gene flow in Hellenistic and Roman times’ etc.

2. Table 2 fits better in the Materials section.

3. b.C.E. should be BCE

4. BCE does not stand for before Christian era but for Before Common Era

5. Similarly, CE stands for Common Era

6. The data of Table 1 should be aligned to the left

7. In Table 2, what do the two columns with references represent? Is there a reason the references are given in two different columns instead of one?

8. What does ethnology mean in the context of Figure 1?

9. The key of Figure 1 should be given in Figure 4.

10. The symbols of the Historical Period in Figure 1 are not used in the map. They should be given as a key in Figure 3.

11. In the Figure 1 table some numbers are used multiple times; for example, No 9 is Caserna Tukory, Phoenician of Palermo, Castel San Pietro and Rotoli. The dates of these assemblages are very different. Are these 4 assemblages in the exact same location of the map?

12. Is there a reason why there are no samples from the east part of the island? The authors should comment on this spatial bias.

13. Are the data collected by CT directly comparable with those obtained from photogrammetry?

14. Table 3: nasospinal should be nasospinale

15. Why did the authors not convert the Procrustes coordinates to Procrustes residuals?

16. Section 3 should be labeled Results.

17. What about the potential effects of sexual dimorphism in the authors' results? These would mostly relate to size differences which have been eliminated after GPA but there may also be some residual shape differences.

18. Isn't Figure S2 the same as Fig. 2 without the right part of (a)?

19. What do the light blue and dark blue lines represent in the wireframes?

20. In the legend of Figure 3, the (c) is missing.

21. What percentage of the variation in the dataset do PC1 and PC2 explain in the various figures?

22. Isn't Figure S4 the same as Figure 6 without the c and d parts?

23. The quality of most figures is very poor. Most symbols are illegible.

 

 

 

Author Response

Estimate Reviewer,

We have finished modifying the paper according to the comments and the questions of the five reviewers. The new version of the paper (text, figures and tables) is so the result of the changes requested.

Here we report all our answers point by point for, moreover changes and answers were reported in the comments placed into the track changes file.

We remain at your complete disposal for further clarifications.

Sincerely

The Authors

 

Comments and Suggestions for Authors

The paper under review uses geometric morphometrics to examine the population history of Sicily. The paper is interesting but there are certain issues that should be addressed before publication.

1. At the moment the research question is open-end. The authors perform GM to see what patterns emerge. Instead, they should formulate some more specific research hypotheses based on archaeological/historical/geographic data, e.g. ‘we anticipate greater heterogeneity/gene flow in Hellenistic and Roman times’ etc.

AUTHORS: Considering the previous information and studies available we formulate, in a dedicated paragraph, specific research hypotheses treated and discussed in the paper.

<< ……………. The intents are to verify the hypotheses that the cranial morphology was modified by the various genetic contributions and more or less related to the intense and significant migratory flows that follow one another from the end of the Upper Paleolithic to the Contemporary and clarify the population dynamics through the ages…………..>>.

2. Table 2 fits better in the Materials section.

AUTHORS: Table 2 was moved to the materials section.

3. b.C.E. should be BCE

AUTHORS: b.C.E. was replaced to BCE in the whole table.

4. BCE does not stand for before Christian era but for Before Common Era.

AUTHORS: before Christian era was replaced to Before Common Era.

5. Similarly, CE stands for Common Era

AUTHORS: Christian era was replaced to Common Era.

6. The data of Table 1 should be aligned to the left

AUTHORS: all the data of Table1 were aligned to the left.

7. In Table 2, what do the two columns with references represent? Is there a reason the references are given in two different columns instead of one?

AUTHORS: there is not a reason to represent the references in two different columns. The version in the paper presents references in one columns aligned to left.

8. What does ethnology mean in the context of Figure 1?

AUTHORS: the word “ethnology” was removed.

9. The key of Figure 1 should be given in Figure 4.

AUTHORS: The key, not originally added to Figure 4 and into the caption to not weight the content, was also added in the caption of Figure 4.

10. The symbols of the Historical Period in Figure 1 are not used in the map. They should be given as a key in Figure 3.

AUTHORS: Considering that sometimes from the same location comes from findings belonging to different places and periods the map of Figure1, the map intentionally does not report both symbols of the Prehistorichal as a historical period but symbols were added to the captions of Figure 3, 4 and Figure 7. The table associated to the map was modified by removing the unnecessary information and proper relabeling the columns.

11. In the Figure 1 table some numbers are used multiple times; for example, No 9 is Caserna Tukory, Phoenician of Palermo, Castel San Pietro and Rotoli. The dates of these assemblages are very different. Are these 4 assemblages in the exact same location of the map?

AUTHORS: The sample sometimes contains findings coming from the same place but belonging to different locations or periods so different populations so were analysed as different groups. Key and dating reported in Figure 1 specify the criteria used for the assemblage.

12. Is there a reason why there are no samples from the east part of the island? The authors should comment on this spatial bias.

AUTHORS: We added dedicated sentence, in paragraph 2.1. Cranial samples, to explain the spatial bias.

<< ……The number of specimens and the spatial bias in part of the map (lack of available samples) are related with the funerary practices used in the island (incineration and inhumation in ossuaries) and the lack of conservation of the osteological finds during the excavations until the end of the 70s, the modern urbanization and the excavations in progress.>>

 

13. Are the data collected by CT directly comparable with those obtained from photogrammetry?

AUTHORS: CT and photogrammetric models were scaled and exported in the same format (.PLY), before marking the landmark moreover GPA standardizes each specimen to unit centroid size. Despite the GPA standardisation, we have avoided the regression analysis of the centroid size that is available only if data are acquired with the same instrument.

14. Table 3: nasospinal should be nasospinale

AUTHORS: Nasospinal was replaced with the word Nasospinale.

15. Why did the authors not convert the Procrustes coordinates to Procrustes residuals?

AUTHORS: As usual in all of these studies, we are interested to perform and discuss the best fit between the specimens. Residuals allow the identification of the procustes error highlighting the worst fit between the specimens that are not part of the aim of our study. Moreover, as requested by Reviewer2 the methodology was rewritten and expanded clarifying each steps of the protocol describing all the analyzes carried out motivating the choice.

16. Section 3 should be labeled Results.

AUTHORS: Section 3 was relabeled Results.

17. What about the potential effects of sexual dimorphism in the authors' results? These would mostly relate to size differences which were eliminated after GPA but there may also be some residual shape differences.

AUTHORS: The simple/size composition always influences the results and this was highlighted argued once on page 6 and twice on page 8.

18. Isn't Figure S2 the same as Fig. 2 without the right part of (a)?

AUTHORS: As written in the text and the captions Figure 2 is the “standard” wireframe and S2 is the soft-wireframe but being similar Figure S2 was removed.

19. What do the light blue and dark blue lines represent in the wireframes?

AUTHORS: Light blue and dark blue lines represent the direction and the magnitude of the changes that move from light blue to dark blue. Captions of Figures 2 and 6 were modified by specifying this representation.

20. In the legend of Figure 3, the (c) is missing.

AUTHORS: The legend of Figure3 was completed.

21. What percentage of the variation in the dataset do PC1 and PC2 explain in the various figures?

AUTHORS: The percentage of variation of all the figures is reported in Supplementary Files 2.

22. Isn't Figure S4 the same as Figure 6 without the c and d parts?

AUTHORS: As written in the text and the captions Figure 6 is the “standard” wireframe and S4 the soft-wireframe but being similar Figure S4 was removed.

23. The quality of most figures is very poor. Most symbols are illegible.

AUTHORS: The quality depends on the original size of the plots. Unfortunately, both MorphoJ and PAST allow exporting of only plots of low quality but the dpi of all the figures were increased up to 600dpi.

Author Response File: Author Response.docx

Round 2

Reviewer 4 Report

I thank the authors for considering the comments and adjusting the manuscript accordingly. However, in my opinion, there are still issues that should be resolved prior to the acceptance of the manuscript.

Abstract

-        The Abstract has been improved with new information. However, it is not organized in logical order. I know that the journal does not have a strictly structured form, but, for example, the sentence “Considering sample size/composition it has been possible to discriminate prehistorical and historical populations” before mentioning the sample, methods, and results is not logical. I suggest that authors follow a logical order in the abstract: background/aim, materials, methods, results, and conclusions.

Backgrounds Studies and Sample recognition

-        “The sample of skulls analyzed was selected on the basis of some of these studies”. First, how was sample selected belongs to the materials section. Second, this sentence is imprecise as it does not explain which criteria were used.

Aim of the study

-        It is still not written well and contains parts that belong to M&M, e.g., “To achieve this, geometric morphometrics (GM)…”, “All data have been collected and…”

-        The intents are to verify the hypotheses – did you think to test the hypotheses?

-        The hypothesis is also not clear considering the study type. It must be shorter and more précised and focused to questions that can be answered by study materials and methods. For example, this is not a genetic study that can really prove genetic contributions (although it can be a valid interpretation when discussing results).

Methods

-        Authors wrote in the pdf comment, “CT and photogrammetric models were scaled and exported in the same format (.PLY)…” but. I think it is relevant and should be included in the manuscript.

-        “It however is an exploratory analysis (Le Maître and Mitteroecker, 2019) based on the visualization, only sufficient to support hypothesis. Theories should be supported by mul-tivariate tests (i.g. MANOVA) and other studies. “ – It is written generally. I suggest connecting it with what was done in the study. For example “ Since ….. is an exploratory analysis, we conducted also…. to….”

Results

-        “Finally, the multivariate comparison tests of variance (Wilk’s lambda and Pillai trace) returned, respectively p(same) values of 1,69E-06 and 0,0001486 in a significance level of α=0,05” – The level of statistical significance should be stated in the methods section along with statistical analyses. P-value should be reported to three decimal places; in this case, P < 0.001.

-        Please delete the sentence:” The complex, varied and dynamic “populations influx” in the Sicilian setting will be clarified and discussed in detail in the following paragraphs.”

3.3. Neighbour-Joining of Procrustes Coordinates

- Please delete from here that something will be explained in the discussion.

 

4. Prehistory and History: Geometric Morphometrics focus

- This should also be part of the results section.

 

Discussion

-        The authors have modified the first paragraph of the discussion, but it is again general, not specific. The main fining and contributions are not what was done but what was obtained (most important findings) and why it is that important.

-        In the second paragraph, it is unclear if authors wanted to say that their multivariate analysis was provided by their results or results from previous studies. For discussion, it is crucial to start from specific study findings and put them in the broader context with explanations, previous studies (when available), and general knowledge.

Conclusions

-        I have commented previously that “This part is written to generally. It should be more focused to the main findings and what they practically mean.” because it was too general. I wanted authors to be more précised and start from their findings to the wider conclusion. I suggest that authors briefly employ some of these ideas in the first part of the discussion because that part should contain the main results, their meaning, and contribution. So, if needed, conclusions can be slightly adjusted to avoid repeating.

Author Response

Estimates Reviewer

We have finished modifying the paper according to your comments.

Here we report all our answers point by point moreover changes and answers were reported in the comments placed into the track changes file as “2nd-Round Reviewer4”.

We remain at full complete disposal.

Sincerely

The Authors

 

Reviewer4

Comments and Suggestions for Authors

I thank the authors for considering the comments and adjusting the manuscript accordingly. However, in my opinion, there are still issues that should be resolved prior to the acceptance of the manuscript.

Abstract

-        The Abstract has been improved with new information. However, it is not organized in logical order. I know that the journal does not have a strictly structured form, but, for example, the sentence “Considering sample size/composition it has been possible to discriminate prehistorical and historical populations” before mentioning the sample, methods, and results is not logical. I suggest that authors follow a logical order in the abstract: background/aim, materials, methods, results, and conclusions.

Authors: We have rearranged following the order of the abstract as requested by the reviewer moreover the aim has expanded by adding the intent.

<<…………………region. The research goals are the identification of temporal trends in facial morphology in order to assess the adaptations and the microevolutionary trends to verify the hypotheses that the cranial morphology was modified by the various genetic contributions and more or less related to the intense and significant migratory flows. This work includes a diachronic morphometrics study of 3D Models of 95 Sicilian skulls coming from 19 populations (from the Paleolithic to the Contemporary Age) providing an overview of human biodiversity and variability in Sicily.  To achieve this a geometric morphometrics…………….>>

Backgrounds Studies and Sample recognition

-        “The sample of skulls analyzed was selected on the basis of some of these studies”. First, how was sample selected belongs to the materials section. Second, this sentence is imprecise as it does not explain which criteria were used.

Authors: We clarified, adding new sentences, how we use the previous publications to identify the sample analyzed and the criteria used have been explained in the “Methods” section.

<<…………Based on the previous publications was possible to identify the osteological mate-rial containing skulls and the related ubications. After recognition and a preliminary evaluation of all the osteological materials skulls were selected according to the characteristics reported in the "Materials" section. Only after the selected skulls were moved to the laboratory. In detail the considered references were……………>>

Aim of the study

-        It is still not written well and contains parts that belong to M&M, e.g., “To achieve this, geometric morphometrics (GM)…”, “All data have been collected and…”

Authors: We moved into M&M the sentence.

-        The intents are to verify the hypotheses – did you think to test the hypotheses?

Authors: Despite the two-test provided after the request of Reviewer4 we have removed each sentences containing the hypothesis in the whole article.

<<…………….. Nevertheless, this is not a genetic study the intents are to verify if the cranial mor-phology underwent modifications after the various population contributions (genetic variability - Matsumura et al. 2018) more……………..>>

-        The hypothesis is also not clear considering the study type. It must be shorter and more précised and focused to questions that can be answered by study materials and methods. For example, this is not a genetic study that can really prove genetic contributions (although it can be a valid interpretation when discussing results).

Authors: In the sentence reported above as in the whole text we remove the hypothesis and moreover that our is not a genetic study was clarified in the same sentence. In addition, we added a reference related to the morphological changes and the variability (Matsumura et al. 2018).

Methods

-        Authors wrote in the pdf comment, “CT and photogrammetric models were scaled and exported in the same format (.PLY)…” but. I think it is relevant and should be included in the manuscript.

Authors: We included in the text the comment that regards the scale and the format.

<<…………..Photogrammetric and CT models were scaled and exported in .PLY format……..>>

-        “It however is an exploratory analysis (Le Maître and Mitteroecker, 2019) based on the visualization, only sufficient to support hypothesis. Theories should be supported by mul-tivariate tests (i.g. MANOVA) and other studies. “ – It is written generally. I suggest connecting it with what was done in the study. For example “ Since ….. is an exploratory analysis, we conducted also…. to….”

Authors: As suggested we rewrote the sentence.

Results

-        “Finally, the multivariate comparison tests of variance (Wilk’s lambda and Pillai trace) returned, respectively p(same) values of 1,69E-06 and 0,0001486 in a significance level of α=0,05” – The level of statistical significance should be stated in the methods section along with statistical analyses. P-value should be reported to three decimal places; in this case, P < 0.001.

Authors: As suggested P-values were reported in decimal places.

-        Please delete the sentence:” The complex, varied and dynamic “populations influx” in the Sicilian setting will be clarified and discussed in detail in the following paragraphs.”

Authors: As requested we deleted the sentence.

3.3. Neighbour-Joining of Procrustes Coordinates

- Please delete from here that something will be explained in the discussion.

Authors: As requested we removed any possible sentences explained in the discussion.

<<The NJ tree obtained with the Procrustes coordinates of the entire Sicilian sample (Figure5) shows the Outgroup separated into two different main clusters, one that has Mesolithic roots and the second that generates all the other groups. This split with the Bronze and Iron root preceding the historical one. Finally, the more recent groups (Middle Ages and Contemporary) cluster relatively closer to each other showing a certain similarity.>>

4. Prehistory and History: Geometric Morphometrics focus

- This should also be part of the results section.

Authors: The results comprehend the analysis of what we can define as “three blocks”: 1-all the specimens, 2-Prehostoric specimens and 3-Historical Specimens. However we removed the word ‘‘Focus” from the title, we replaced the first sentences to a proper position into the “Methods” and we deleted any possible interpretation and discussion of the data the were only presented without any explanation.

<<……… The comparison between the Shape Variation and Changes in Direction graphs compare the direction and magnitude of the shape variations between the two groups. Among the prehistoric specimens is possible to notice that the the landmarks placed on the superior jaw (landmarks 1, 7, 8, 13 and 14) (Figure6a-b) slightly move forward and inwards. At the same time, the facial and frontal bones (Figure6c) do not undergo any change. In that period, the main changes occur in the cranial…………>>>

Discussion

-        The authors have modified the first paragraph of the discussion, but it is again general, not specific. The main fining and contributions are not what was done but what was obtained (most important findings) and why it is that important.

Authors: We re-organize the section modifying and adding sentences in order to present: intents – materials – methods – main results – point by point discussion – few sentences to summarize the discussion.

-        In the second paragraph, it is unclear if authors wanted to say that their multivariate analysis was provided by their results or results from previous studies. For discussion, it is crucial to start from specific study findings and put them in the broader context with explanations, previous studies (when available), and general knowledge.

Authors: We specified that we are presenting our results of our multivariate analyses with the few previous studies that are clearly explained and differenced by our data.

Conclusions

-        I have commented previously that “This part is written to generally. It should be more focused to the main findings and what they practically mean.” because it was too general. I wanted authors to be more précised and start from their findings to the wider conclusion. I suggest that authors briefly employ some of these ideas in the first part of the discussion because that part should contain the main results, their meaning, and contribution. So, if needed, conclusions can be slightly adjusted to avoid repeating.

Authors: We rewrite the conclusions structuring the section with the main result, meaning and contribution.

Author Response File: Author Response.docx

Reviewer 5 Report

The authors have addressed all my comments. The paper can be published in its current form.

Author Response

We thanks the reviewer to the priceless advices and suggestions.