All reviews of published articles are made public. This includes manuscript files, peer review comments, author rebuttals and revised materials. Note: This was optional for articles submitted before 13 February 2023.
Peer reviewers are encouraged (but not required) to provide their names to the authors when submitting their peer review. If they agree to provide their name, then their personal profile page will reflect a public acknowledgment that they performed a review (even if the article is rejected). If the article is accepted, then reviewers who provided their name will be associated with the article itself.
Thank you to the authors for carefully incorporating suggestions made by the reviewers. I have assessed the revision and I am happy with the current version. The manuscript is now ready for publication.
[# PeerJ Staff Note - this decision was reviewed and approved by Michael Wink, a PeerJ Section Editor covering this Section #]
Overall the reviewers were satisfied with the revisions of the original submission. The reviewers do highlight a few minor changes that the authors should consider before the manuscript can be accepted.
Yes, this is well done. However, I noticed one issue with the reporting of numbers in table 1: there are a lot of zeros for SS and MS in some fields and some really small numbers in others. All of these are some small non-zero numbers and are the result of the scale of the x and y variables in the analysis. for example, they would be ten times bigger if the y-variable were multiplied by ten, but the p-values, R^2, etc. would stay the same. As it is, these fields do not really contain useful information. I would suggest either changing these columns to scientific notation or adding more decimal places so that you have at least two with non-zero digits (yet keeping number of decimals standardized within each column. Since MS=SS/DF, the MS column is actually redundant with the SS & DF columns. You could delete either the MS or SS without loss of information to make space for more precise number reporting.
Yes, this is also well done, but I suggest one clarification:
L158-163: I generally like the improved description of the full and reduced limb datasets. What gives me pause is that in a limb, the carpals/tarsals come between the radius+ulna/tibia+fibula and the metacarpals/metatarsals. Therefore, the measurement of “full” fore or hind limb is actually discontinuous. I don’t find this methodologically suspect, but biologically confusing. I would suggest simply addressing what happened to the carpals/tarsals. For example, maybe they weren’t measured because they are frequently missing.
No concerns or comments.
The authors addressed the reviewer comments quite well and I have no further major concerns that need to be addressed. I do have two more suggestions, in addition to the comments above:
Abstract: The revised and added text is great. I suggest putting the word “body” in front of each instance of the word “size” or “sizes”.
Figure 2: I think that including the phylogeny is a major improvement. A suggestion might be to include some graphics of what these animals look like to make it less utilitarian and illustrate some of the variation for non-squirrel specialists. A suggestion might be an image of a gliding squirrel, a chipmuck, two ground squirrels (one for each clade), and three or four tree squirrels, again, one for each clade. This would also help illustrate whether and how the ground and tree squirrel clades may or may not look convergent over all.
This is a revised submission of manuscript 78381 titled “Scaling patterns of body plans differ among squirrel ecotypes” by Linden et al. I reviewed the original submission of this manuscript and only minor comments/concerns. This revised version addresses my concerns raised previously. The manuscript is written clearly, with relevant figures, tables, and citations provided. Raw data and R code are also now provided (something I failed to notice was missing in the previous submission).
no comment
I did not have many concerns here, and the concerns raised were addressed. One suggestion I have to follow up on my comment about correcting for multiple comparisons is that I think it should be stated somewhere that p-values are corrected (in the main text or in the table header/figure captions).
(Abstract) Line 23: remove “on these data”
(Methods) Line 182: what version of R? Also, R should be cited.
(Methods) Line196: use the Greek letter since you do so elsewhere in the manuscript.
(Discussion) Line 455: spelling—“shapre” to “shape”
(Literature Cited) Line 696: spelling— “Experimentl” to “Experimental”
As I mentioned at the top, these are minor comments and concerns.
Thank you for adding the raw data and code to your submission.
Research questions were very well identified and the methods were used well to address them. The methods are reproducible with the provided code.
I appreciate the in-depth discussion of the findings and how they relate to the central questions of the manuscript.
Thank you for allowing me to review this again. I appreciate the authors' work in editing their manuscript based on the feedback they were given. Thank you for providing the supplementary code and raw data files for reproducibility. I especially like the addition of the phylogeny with mapped ecotypes. This was an enjoyable read and I think this study makes a great contribution to the study of ecomorphs, body size, body shape, and limb lengths in mammals.
The reviewers were generally very positive about the study but had some critique regarding data analysis and presentation. I would suggest that the author carefully address each of the reviewers' comments in revising the manuscript.
[# PeerJ Staff Note: Please ensure that all review and editorial comments are addressed in a response letter and any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate. #]
Article meets these criteria, although a few terms should be bused more consistently and made more precise (see other comments).
These elements are met, but one set of analyses needs to be described in more detail (see additional comments).
These criteria are mostly met. Data that were analyzed have been provided, but these are not the raw data. The raw data would be the vertebral lengths and heights that were used to calculate the elongation ratios. What the authors provide are the data that were calculated from the raw data and analyzed. The data appear robust.
Conclusions are reasonably well stated, but I disagree with the interpretation of one set of analyses (See Additional Comments).
The authors present a study of body shape evolution and evolutionary allometry for 87 species of squirrels belonging to four different ecotypes. It is a great dataset and a great study with interesting biological questions. The manuscript is also well written and interesting. I do have a number of comments that are important to address before publication, but I think that all of them are easily addressed. I then also provide detailed comments by line number. My larger comments are:
1. Some terms should be used more precisely and consistently. The most confusing term is body shape. When the authors use “body shape”, they actually mean “head-body elongation ratio”. Body shape is used in various ways in the literature and head-body elongation ratio is one aspect of body shape. For clarity, the authors should refer to head-body elongation ratio throughout because body shape is vague and certainly encompasses more that elongation. Similarly, the authors should be consistent in the use of “body size” or “skeletal size”, but not use them interchangeably. I don’t think it matters which they use. For example, “body size” would be fine, and then just define it as the geometric mean of all the skeletal elements, as is done in the methods.
2. The ANCOVAs need to be described in more detail and more clearly. It is unclear exactly what the design of the ANCOVAs is and how they are presented is inconsistent with most statistical outputs of an ANCOVA. For example, the authors present slopes an intercepts for each ecotype in Tables S2 and S3, but most of the time, ANCOVA results in statistical software present one group (ecotype) as a reference line with a slope and intercept, and the tests are whether that slope and intercept are different from zero. Then, for all other groups, the slope and intercept estimates that are presented are differences from the reference line/group, and the tests are whether each is different from the reference intercept and slope. This is not how the authors present their results. It is important to know if the package they used presents ANCOVA results in a different way or what the authors did to present the results that they do in their tables. I also describe this in my comments for lines 183 to 195.
3. The interpretation of the RRPP multiple regression results is not correct. The authors state that they are testing which regions “contributed the most to overall body shape variation” (L263), and the results section is entitled “Predictors of body shape variation”. This suggests that the authors are studying variation as a response variable, implying some study of diversity. This is not the case. Instead, the authors are testing whether the elongation ratio for each body region is related to overall head-body elongation. For these analyses, they focus on the R^2 values for each variable, which might tell them how much of the variation in the response is explained by the elongation of each body region, but this is different from them being predictors of body shape variation or understanding how much they contribute to body shape variation. To illustrate, their analyses show that species with relatively longer ribs tend to be less elongate and those with relatively more elongate thoracic and sacral regions tend to be more elongate overall. This distinction is important and should be made. This comment can be addressed with some changes to the phrasing around those analyses.
Abstract:
- The abstract is well written and informative. However, given that the focus of the manuscript based on the title and the first part of the abstract is body size and allometry, it would be good to see a little more information about the findings with regard to these factors in the abstract. Currently the authors list three findings, two of which deal with relationships between particular body parts. The first finding they list deals with allometry, but is extremely general (morphological components scale allometrically with body size and this differs between ecotypes). I would like to see this expanded a little in the abstract because it is the primary focus.
Introduction:
- Generally well written and informative.
L40: “facilitates” should be “facilitate” to agree with plural changes. Also, this sentence should be supported by some citations. It is a bold statement that body size facilitates evolution of new traits because, given homology and phylogenetic constraints, truly new traits aren’t that common. Furthermore, what about body size facilitates this? Is it large body size, small, or evolutionary changes in body size. Body size just is – every animal has a body size. That in itself does not facilitate anything.
L55: For a broad audience, I suggest changing “squamates” to “squamate reptiles”, and maybe saying what those are.
L68-69: Bergmann & Morinaga (2019 – cited in the manuscript) and Kohlsdorf & Wagner (2006. Evolution 60: 1896-1912) provide examples of clades where the hind limbs a reduced before the front.
L98: Is it body shape or body shape variation? With body shape variation, it sounds like the authors will test for a relationship between of body size and some sort of measure of diversity in body shape. If they are testing for a relationship between body size and body shape (for example, an elongation ratio or a PC of morphometrics), then this should say “body shape” without the word “variation”.
L100 and elsewhere: The paragraph started with references to “body size”, but here switches to “skeletal size”. I suggest being consistent to avoid confusion.
L108: Should be “remains”.
Materials and Methods:
L128: a reference should be provided for head-body elongation ratio, as this has been used extensively in the past.
L146: Why not include the limb lengths in the calculation of the multivariate skeletal size? Inclusion of limbs would provide the maximally holistic measure of size.
L161-162: What datasets give the same major patterns? Do the authors mean that they have the dataset that they describe and a smaller one that also includes the autopodial lengths? If so, then how autopodial lengths were measured should also be indicated, and this sentence should be clarified.
L165-175: It would be helpful to have a figure that provides the phylogeny, with branches color-coded by ecotype. It sounds like chipmunks are a single clade, but it is unclear how the other ecotypes are distributed on the phylogeny. Such a figure could be supplementary or in the main text. In the main text, it could be a graphical centerpiece of the work.
L183-184: I suggest moving the information about what package and function was used to do PGLS regressions and ANCOVAs be moved up to here. It is also important to make clear that size was used as the x-variable because a regression of x on y and a regression of y on x give different slopes.
L184, 195: More details should be provided about the ANCOVA analyses. In particular, did these ANCOVAs test for differences in slope and intercept between ecotypes or just the intercepts? I assume that it is the former, which would be correct, but some ANCOVAs (those that are an ANOVA plus a covariate) test only for differences in intercept and are invalidated when slopes differ. The presentation of these results in Tables S2 and S3 don’t quite shed light on this because although those tables provide different slopes and intercepts for each ecotype, they imply different analyses for each ecotype because there is also a row for all species, which wouldn’t come from an ANCOVA. Furthermore, most ANCOVA outputs use one ecotype as a reference line and then present the differences in slope and intercept between the ecotype that was the reference line and each other ecotype. Then the tests of significance are for these differences. What is presented here is unclear to me in terms of what these analyses look like.
L206-210: In is unclear how doing more regressions using RRPP allows the authors to quantify which variables contribute most variation in body shape. This should be explained. Specifically, what metric is then used to quantify the variation contributed by each variable? Also, and this applies to all of the analyses, the authors should more clearly articulate what response and explanatory variables were included in each analysis.
Results:
L215: “body shape” The text should be more precise about what relationships are being presented. From the supplementary materials and especially Figure 2, it seems that these are relationships between body size and the head-body elongation ratio. All the references to body shape should be replaced with head-body elongation ratio (or its abbreviation) so that it is clear what the results refer to. Other than this, the text is clear.
L217: The interaction/model formula used should be in the methods.
L217-218 & elsewhere: Please explain in the first instance what is presented by “(0.12 [0.06:0.19])”. I assume this is the slope and its 95% CI, but this needs to be stated.
L229-237: What is striking to me from Figure 3 and Table S2 is how ground squirrels are doing their own thing and have significant scaling relationships for every region except sacral. I’d recommend adding a phrase to the ground squirrel sentence to draw attention to this.
L240: Again, it is unclear what “body shape” is. Does this mean the head-body elongation ratio? Please be precise.
L242-243: The description of the full limb should be in the methods. Also, technically, it is not the full limb if it excludes the phalanges. It is also unclear if the carpals and tarsals are included, and this should be indicated.
L253: I suggest adding a topic sentence to tell the reader what this paragraph is about.
L256-258: I find a big disconnect between the statements here and the graphs and statistics. For hindlimb, figure 5 shows non-overlapping data between flying squirrels and the other ecotypes, suggesting that it is quite clear that they are different from all other ecotypes. The text says that flying squirrels are different from ground squirrels but not chipmunks and tree squirrels. However, the 95% confidence intervals are almost identical for ground squirrels and chipmunks and they all overlap with the intervals for gliding squirrels, suggesting no differences. This needs to be resolved.
L260-270: Again, “body shape” is the overall elongation of the body (ERhb). Also, I question the interpretation of the authors in saying how some variables contribute to the overall variation of ERhb and others do not. I simply do not think that the analyses show this. There are a few parts to this. First, all of the variables except rib length are actually components of ERhb: the length of the head, neck, thorax, lumbar, and sacral are parts of the head-body length. Therefore, they all contribute to it. Second, the analyses are testing how much of the variation in ERhb is explained by these other variables. Importantly, this is subtly different (contribution versus explanation). How I would describe these results is that species with relatively longer ribs were less elongate or more robust, and that species with more elongate thoracic and sacral regions were more elongate overall.
Discussion:
- The discussion is interesting, but excessively long. For example, the Intro, Methods, and Results are 11 pages together and the Discussion is 9 pages on its own – almost half of the manuscript is discussion. I recommend shortening the discussion. Every single result does not need to be discussed. I would suggest focusing on the big picture from a synthesis of the results.
L280: Using positive and negative allometry when considering a ratio is a bit confusing because the ratio’s meaning is not like another measurement of a body part. Instead, here positive allometry actually means getting relatively more elongate as size increases. That would be a more direct and interpretable way to describe this. Writing about this as suggested in this comment then eliminates the need for the parenthetical information in line 288.
L294: Dorsoventral bending might be clearer than sagittal bending.
L308-310: I don’t recall the result that forelimbs were reduced before hindlimbs in the current study. Looking at the limb length section of the results doesn’t seem to reveal this result.
L353: This statement should be supported by a citation.
L364-376: Indeed maybe what is surprising here is that not only do chipmunks not show the same relationships as ground squirrels, but they have the steepest relationship of all the ecotypes where larger species are most elongate (Fig. 1).
L391: “increase”, not “increases”.
L402-422: This is an interesting finding because most vertebrates develop anterior to posterior, resulting in negative allometry of the anterior regions and structures. Looking at the graphs in Figure 3, what is interesting is the pretty much all the ecotypes have positive allometry of both the head and the neck, but not of further posterior regions. Given this pattern, it seems like although there may be an adaptive explanation for this, it should apply to all of the ecotypes. Another group of mammals with positive allometry of the head is whales, supposedly to help consume enough food for their body size (Goldbogen et al. 2009, Proc. R. Soc. B 277: 861-868).
L437-460: Similar to the comment I made in the results, I disagree with the interpretation that the multiple regressions between regions and head-body elongation show which regions contribute the most variation to ERhb. They do explain a certain amount of variation, but that is different. This part of the discussion should be modified to account for this.
Figures 2-4: I suggest slightly increasing the axis ranges to provide a little more space in each plot around the data. I think this will make the patterns pop out a little more. But this is a matter of personal preference.
Figure 6: In the caption, I disagree that the authors analyzed body shape diversity. None of the metrics have to do with diversity. These analyses have to do with relationships between pairs of variables that describe body shape. I also find it strange that the authors provided Z-scores over p-values in the graphs. I’m glad that everything is provided in the table, and understand the limitations of p-values, but that still seems like the way to go when presenting only a small set of values.
Table 1: This table spans three pages and for ease of reading, I wonder if it should be broken into multiple tables. The column headings should be explained, and the 2 in R2 should be superscripted.
References:
- The journal is not provided for the Marroig & Cheverud reference.
Supplementary Material.
Table S1: I suggest putting the column headings at the top of each page of this table.
Tables S2 & S3: The tables contain good information but should be formatted more professionally. Start Table S3 on a new page. Start each column heading with a capital letter. Standardize the number of digits after the decimal in each column. Adjust the column widths to be more equal and so that headings aren’t cut (e.g., the e in “slope” spills onto a new line). Justify column headings so that they are above the column data. In table S3 have the subheadings stay on one line. These details matter, especially when journals do not typeset supplementary material.
Linden et al., study scaling relationships among different squirrel ecomorphs. They used linear morphometrics of skeletal specimens to quantify body shape and size, paying particular attention to variation in body elongation and limb reduction in relation to size across different ecomorphs. They find that elongation ratios of chipmunks and gliders show positive size allometry, while ground squirrels tend to exhibit negative allometry, and tree squirrels show isometry. They also find that ground squirrels exhibit negative forelimb length allometry while all other ecomorphs exhibit isometry.
The manuscript was generally well-written and a pleasure to read.
I think the authors ask a fascinating set of questions regarding body shape evolution and use a largely under-utilized resource to answer them, and I commend their efforts having done a fair bit of it myself.
I think their conclusions are reasonable and only have minor concerns and comments. I detail them below.
1. I am not familiar with the system, so this may be off-base, but I find the categorization of “chipmunk” as an ecomorph difficult to reconcile. I don’t doubt that many (or none) can’t be considered fully fossorial or arboreal, but I wonder if that isn’t an impetus to consider a category of semi-fossorial and semi-arboreal ecomorphs in general. Or, perhaps another way to consider this is whether other squirrels fit a generalist category (and replace “chipmunk” with “generalist”). This may simply be a semantics issue more than anything else.
2. What are the adjusted R2 for the results shown in figure 4? All the slopes in this figure look very subtle and I would hesitate to draw any conclusions from this.
3. P-values derived from the multiple regressions (i.e., analyses that resulted in the figure 6/table 1) should be corrected for multiple comparisons. I suggest using FDR/Benjamini-Hochberg correction to do this.
Line-by-line comments:
(Abstract) Line 19: Change “locomotory” to “locomotor”
(Introduction) Line 62–65: I think the message here would be clearer if the order was switched from terrestrial to aquatic, to aquatic to terrestrial, since the example that precedes this is regarding fishes. For example: “…these allometric patterns: aquatic carnivorans exhibit patterns similar to fishes (Friedman et al., 2019), but terrestrial carnivorans instead evolve more robust bodies as size increases (Law, 2021b).”
(Introduction) Line 66: Re-word to: Elongate body shapes are associated with fin and limb size reduction.
(Introduction) 71: change locomotory to locomotor.
(Introduction) Line 79: I believe the second author’s last name for the citation here is Kuratani, not Kurutani
(Results) Line 214–217: Are the R2 values reported here adjusted? If not, I would suggest reporting adjusted R2, which accounts for the number of terms in the model.
(Figures 2 and 3): I’m assuming “skeletal size” on the x-axes for these figures is the same as body size. For the sake of consistency, I would use skeletal size or body size throughout the manuscript (I’d suggest latter).
(Figures 2–4): I think it would these figures can be improved by
(Figure 5): X-axis label shows flying which I’m assuming is the same as gliding. This should be changed to gliding for sake of consistency.
(Table 1): I prefer consistency in the number of digits displayed after the decimal. Additionally, I’m assuming the residuals of 0.00 is actually some exceptionally small value, rather than zero. I suggest the authors display these as <0.001 or some reasonably small value.
My concern with this submission was the lack of resources relating to reproducibility. While two supplementary files were provided, no code or tree files were available. I would request that the authors provide access to all of their code and any files (such as a phylogeny) that they used to create their analyses.
I recommend that the authors review the formatting of their citations and double check some spelling. Here are specific comments:
Line 228: “skeltal” should be skeletal
Line 499: "ofecholocation" needs a space
Line 500: "ofskull" needs a space
Line 508: Giraffa camelopardalis should be italicized (same with Line 559)
Line 548: "an-urans" should be anurans
Line 574: "evolu-tionary" should be evolutionary
Line 575: "in-novation" should be innovation
Line 591: "conse-quences" should be consequences
Line 617: Spermophilus citellus should be italicized
Line 705: "pres-ent" should be present
The authors effectively show how allometric patterns in body shape in squirrels differ by ecotype, and that the variation in body shape is most influenced by the elongation or shortening of the thoracic region. I found this manuscript to be an enjoyable read with clearly stated objectives, straightforward analyses that support those objectives, and in-depth exploration of the results.
The authors had three objectives to explore body size and shape in 87 species of squirrels (Sciuridae). Their objectives were to: 1) to explore the relationships between body shape variation and body size, and body shape variation and limb length; 2) test how the aforementioned relationships were influenced by ecotype or habitat; and 3) to determine which axial components were most responsible for body shape variation across all squirrels and within ecotypes.
To test their first two objectives, the authors used Phylogenetic Generalized Least Squares regressions (PGLS) to test relationships between body size and shape, and body shape and limb length in their sample, then used a phylogenetically framed ANCOVA to explore these patterns in each ecotype (ground squirrels, tree squirrels, chipmunks, and gliding squirrels). The third objective was addressed by using values of length and width of the head and each vertebral region in phylogenetic multiple regressions. They found that most patterns in body shape, size and limb length were best understood in the context of habitat/ecotype, and that the thoracic region in particular (its elongation/shortening and relative length of the ribs) explained most of the body shape variation in their sample of Sciuridae.
Again, the code used for all analyses (and supporting files) should be submitted with the manuscript.
No comment.
I have a few minor topics of interest that the authors might consider exploring in their manuscript:
1) Does the sampling of Sciuridae in this study effectively capture the breadth of ecotypes and taxonomic diversity in the family? The sample is composed of 87 species with Sciuridae having well over 200 species total. I would be interested to know if there are ecotypes not represented by this sample, or how well this sample captures overall diversity in Sciuridae.
2) With the interest of examining body shape and size as well as limb length, is there reason to expect that Allen’s Rule (that endotherms in cold climates have relatively shorter appendages) and/or Bergmann's Rule (that body size in endotherms increase with increasing latitude) may apply to Sciuridae? While the authors overall did not find a relationship between body shape and size-corrected forelimb or hindlimb length (with the exception of forelimb length in ground squirrels, and some patterns in relative lengths among ecotypes), is it possible that these rules could provide a useful framework for interpreting the results in this paper? This would be particularly interesting if there is a bias in the distribution of ecotypes at different latitudes.
Overall, I appreciate this work and its contributions to understanding how body shape and size relate to locomotion and limb lengths and hope that future research will be undertaken to examine these patterns in other clades.
All text and materials provided via this peer-review history page are made available under a Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.