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.
I confirm that your paper is now accepted for publication.
[# PeerJ Staff Note - this decision was reviewed and approved by Kenneth De Baets, a PeerJ Section Editor covering this Section #]
Dear Dr Woodward,
Your manuscript has been re-evaluated by one of the original reviewers. Please address the first three comments in their attached PDF, and I will make my decision thereafter.
Best regards,
Fabien Knoll
Acceptable. See attached review.
Acceptable. See attached review.
Acceptable. See attached review.
Dear Dr Woodward,
I have now received three reviews of your manuscript. Both the reviewers and I find the study interesting and commend the efforts behind it. However, the reviewers have also raised a number of issues that will need to be addressed in your revision.
There is evidence that at least two tyrannosaur species, potentially differing in adult size, may have coexisted in the Maastrichtian of North America. If you disagree, please provide your reasoning. If not, please explain how you can be confident that the 17 individuals analyzed in your study all pertain to the same species. While this issue is briefly noted in the “Problematic specimens” section, it should be elaborated more explicitly in the Materials and Methods.
I also draw your attention to a few minor points:
-I noticed two typographical errors: “Bordy” (line 1108) and “Circumefence” (Figure 18);
-please avoid judgmental wording (e.g., “cleverly”);
-in the reference list, most words in titles are capitalized, when they should not be.
When submitting your revision, please provide:
-an unmarked revised manuscript;
-a marked-up version showing changes;
-a detailed response document explaining how each of the reviewers’ comments has been addressed.
Thank you for your careful attention to these matters. I look forward to receiving your revised manuscript.
Best regards,
Fabien Knoll
[# PeerJ Staff Note: Reviewer 1 declared a potential Conflict of Interest, and the Editor was aware of this when making their decision #]
No comment
See attached review.
See attached review.
See attached review.
The primary focus of this manuscript is a revised examination of the histological data for growth in Tyrannosaurus rex (or perhaps which might be called the “Tyrannosaurus rex species complex”, depending on any given researcher’s taxonomic opinions of the specimens involved.) This greatly adds to the literature of this subject, both by including several new specimens to the study and by reexamining previous ones in a coherent framework for all included individuals.
The primary conclusions refine some previous research results: a somewhat slower growth rate and a later achievement of maximal body size. Additionally, some specimens considered by various researchers as not belonging to Tyrannosaurus rex do indeed show growth patterns distinct from the others in this analysis: the authors note that this is consistent with the hypothesis that these belong to a separate taxon but also note several alternative working hypotheses that cannot be immediately rejected with current data.
The manuscript is consistent with the professional standards of the field.
The authors are extremely thorough in detailing the specimens they used, the particular method of histological sampling employed, their criteria for measuring cyclical growth markers and how they resolve these as indicators of age-at-time-of-death, and their alternative mathematical approaches to creating a growth curve for Tyrannosaurus rex under various scenarios.
The degree of detail of these are sufficient that other researchers could potentially reproduce these results while examining these specimens or by following these protocols add new individuals to these to expand the data base.
The conclusions are intriguing and are well-supported by the analytical results. The authors document the aspects of their methodology with some previous work in the field and thus indicate possible sources of discrepancy between their results and those of previous studies. It will be intriguing to see these methods applied to other taxa in future studies.
Of particular broader implication is their results showing that inclusion of double growth marks and marks only visible in cross-polarized light shows better fit for the data. Some early studies dismissed the former as possibly non-homologous to standard lines of arrested growth and literally ignored the latter, potentially leading to erroneous conclusions about the growth patterns of the specimens studied.
What might be intriguing in future experiments would be comparing the results of the methods here to some of the alternative approaches (for instance, the traditional “whole bone” approach treating specimens as single final points and core-sampling rather than complete transverse cross-sections) for a broad sampling of taxa, to better ascertain how much they disagree and if there is a general bias (e.g., “whole bone” underestimating the ages as calculated using the techniques here) or not in the older studies.
In terms of narrowly focused implications, the results for the BMRP as being inconsistent with definite Tyrannosaurus rex specimens is intriguing. They do acknowledge the distinct possibility that these forms might well belong to another tyrannosaurid taxon (i.e., Nanotyrannus lancensis). However, they do also indicate that there are alternative explanations consistent with observations in extant archosaurs (Alligator): namely, that the ancestral archosaurian growth plasticity might result in animals in particularly stressed environments to have growth patterns distinct from those in less challenging situtations. The observation that these are the same specimens which show an abundance of multiplets is independent evidence that these individuals did experience conditions different from the other specimens. (Both of these specimens are from relatively close to each other geographically, in southeastern Montana: however, determining the stratigraphic level of these and other finds from that region is an on-going project for which the results are not yet known. It would be interesting to find out if a) both of these are from the same restricted stratigraphic interval and b) if that interval shows paleoenvironmental indicators of conditions different from the more typical Hell Creek Formation.)
Finally, a major conclusion here is that despite common interpretation to the contrary, the particular inflection points in growth of extant animals do not typically indicate biological phenomena such as onset of sexual maturity, and thus should not be interpreted as such in fossil forms.
Line 105 The Institutional Abbreviation for “BMRP, Burpee Museum of Natural History” is lacking (ironically, perhaps one of the more important ones for this manuscript!)
Line 1046 In addition to Carr (2020), the online Supplementary Information of Voris et al. (2025) examines the morphological evidence for ‘Nanotyrannus lancensis’ and find it inadequate at present. Notably, this discussion specifically addresses the work of Longrich and Saitta (2024), the most comprehensive argument so far for the validity of Nanotyrannus.
Voris, J.T., Zelenitsky, D.K., Kobayashi, Y., Modesto, S.P., Therrien, F., Tsutsumi, H., Chinzorig, T. & Tsogtbaatar, K. 2025. A new Mongolian tyrannosauroid and the velution of Eutyrannosauria. Nature 642: 973-979. Doi: 10.1038/s41586-025-08964-6
See below for my comments.
See below for my comments.
See below for my comments.
This manuscript represents a large, long-term undertaking to understand growth in one of the most famous fossil species known, T. rex. The paper is interesting and certainly there are shortcomings in the current state of dinosaur growth modeling and physiology, and I appreciate the authors’ statistical attempts to explore these issues. However, I unfortunately have several major concerns with the study and cannot recommend its publication until they are remedied. This is a major undertaking that has resulted in a somewhat unwieldy manuscript. Presenting dozens of models and variants, many of which are biologically untenable (60 kg hatching mass?) or not ground-truthed with a survey of extant histology (XPL-only annuli), invites confusion in the literature. Upon revision and streamlining, this can be a valuable and lasting contribution to the literature. On to my specific concerns:
First, I have a major, not-so-scientific-but-practical concern that the authors may be unhappy with, so please bear with me. There are several studies in progress or nearing publication on Tyrannosaurus histology, which the authors must be aware of via SVP abstracts or their own collaborations. So, my question is: why publish this now? If there are going to be new datasets relevant to Nanotyrannus, mid-sized or baby T. rex individuals, etc., why not wait 6 months or a year and publish something more comprehensive, rather than something that will immediately be out of date?
The authors invoke a new growth mark category, only visible in cross polarized light, but don’t do any ground truthing around it. This is one of the cardinal sins of many high-impact paleontology papers – develop a metric or method for fossil animals without vetting it in extant animals. To me, introducing the idea of an XPL growth mark without vetting it is irresponsible and will just cause confusion in the literature.
On the same topic, the authors say XPL only annuli have not been reported in the literature (e.g., line 286), except one figured in a recent paper by Curry Rogers and others. Only a handful of studies are cited that “do not have” XPL only annuli—a table of studies the authors have examined and which do not have XPL annuli should be presented. Hopefully it is more exhaustive than the few vague references presented currently.
A simple test of the utility of XPL annuli is that in the femur and tibia of a given individual, the number of XPL annuli should be the same between equivalent lines of arrested growth. Moving deep to the periosteum, the authors could count the number of lines of arrested growth (including or skipping the EFS) and see if the number of XPL annuli interspersed between them is the same. The growth of the femur and tibia mirrors one another, as evidenced by the R-squared value close to 1 in the relationship between their circumferences, and of course, intuitively based on biomechanics. So, the number of pauses in growth should be essentially identical between the two bones. The same can be done for multiplets.
The precision of (i.e., narrower confidence bands) or statistical support for (i.e., “better statistical fits”) a given model has no bearing on its accuracy or its biological meaning, full stop. The data exploration and simulation is nice, but falls short of a test of the utility of XPL annuli for growth reconstruction.
Figure 6 shows that the two Burpee specimens fit in your confidence bands, while the CCM specimen lies outside it. However, the manuscript states that these two individuals “do not fit with other Tyrannosaurus specimens.” The authors present so many variations of their analysis, that they seem to then be free to choose which variant or model they want to believe in. This is a central issue with this paper.
Many of the growth curves in the supplemental information have inferred starting masses of 60-80 kg at hatching. This should indicate to the authors that their modeling approach is biologically unrealistic, and so should be modified.
Line 88 claims that this study uses “full cross-sections”, when it does not always do so.
Line 93 claims that T. rex ‘experienced’ (best word choice?) bone apposition rates between 25 and 100 microns per day. But the presence of a LAG indicates a bone apposition rate of 0 microns per day. Please resolve.
In the abstract, the claim is made that T. rex attained asymptotic size in 35-40 years, but on line 813, the range is 27-42 years. I think the uncertainty in which variant of the dataset to choose means that all results should be presented in the abstract, i.e., the full range.
Table 1 belongs in the main text – this is peerJ, you aren’t limited for space, and the dataset is the foundation of the study, not something to be relegated to supplemental information.
Line 133: rephrase “Photogrammetry surface scans” – photogrammetry doesn’t scan, it aligns photos. Fix throughout please.
Line 145: the thickness range of 80 to 200 microns for slides is concerning. Please measure the thickness of each slide down to the nearest 10 microns and report it in a table. Thicker slides can give misleading results when LAGs split, merge, or run oblique to the cut plane.
Line 161: change “full transverse” to “full resolution transverse”; “full” could refer to an entire versus a partial cross section. Check throughout.
Line 183: cementum omitted for some reason.
Line 224: the word “doublet” does not appear in all of these citations – this is sloppy citing, please resolve.
Line 249: data = plural (data were, not data was).
Line 281: the authors casually use the word “diapause” rather than “pause” – this term has specific connotations in other groups of animals and has not been used in dinosaurs to my knowledge. Please use the more simple “pause”.
Line 324: can you be more precise about the frequency of secondary osteons internally? “Higher” could mean 1% higher or 100% higher…can you give a ballpark?
Line 680: again, the authors are sloppy with their citations, since some of these studies do explicitly identify their criteria for defining a multiplet. Mis-citation like this makes this paper appear biased and not carefully written.
Line 719: some large mammals hibernate. How big does an animal have to be for the authors to find pauses implausible?
Line 742: be careful with calling one model “better” than another. Models can be precise and have good statistical fit while representing biological nonsense.
Line 749: “the archosaur extant phylogenetic bracket” cannot be “directly observed”, at best it is a set of nested hypotheses, at worst a misleading house of cards. I’m not saying the method is useless, but the extremely long branches usually employed when using it make it less useless. Example: use a flightless cormorant and Deinonychous as your phylogenetic bracket and then reconstruct the ancestral avian...ridiculous, right? Long branches from ecologically derived individuals give misleading reconstructions–this is why it is silly to predict things about Mesozoic dinosaur growth from crocodiles and birds living today.
Line 776: the reasoning is circular here: there is no way to tell that there was a decrease in apposition rate at these bands without fluorescent labeling studies on an extant animal. The rate of deposition varies greatly with the same collagen organization.
Line 806: Mass estimates should always be reported with confidence intervals.
Line 817: why didn’t the authors use developmental mass extrapolation (DME, Erickson and Tumanova, 2000)? It is well known that tetrapods “grow into” their bodies, varying bone safety factors as mass catches up to size during ontogeny. The training datasets used by Campione and others to create circumference-mass equations are based on adults, and so are not valid for juveniles. DME accounts for this issue when estimating mass from cortical growth marks.
Line 862: “a welcome addition of data” – this is true, but editorializing like this doesn’t really belong in a scientific paper. This sort of expression of the authors’ personal opinion is peppered throughout the manuscript (e.g., “absurdly” on line 1144, “cleverly” on “1146”) and should be removed.
Line 958: “does not match” – this needs to be elaborated – what is the discrepancy?
Lines 995-1006: there is some typo or misunderstanding here. First, I would argue (see above) that the authors should be using developmental mass extrapolation. Putting that aside, there is still something amiss. The authors state “If the percentage mass increase is regarded as the proper metric then the inflection point occurs at age 0 for all of the variants.” But this cannot be true, just look at your own figures 6 or 9 for example. The inflection point on figure 6, measured either as circumference or mass increase, is near the middle of the curve. When comparing percent increase, the largest increase will be at the steepest part of the curve. Are the authors conflating mass and circumference, and percent and absolute values? Is the large increase between age 0 and 1 an artifact of the starting size? I’ll also note again that these same curves (e.g., Fig S17) reconstruct neonatal mass at around 60-80 kg–this is an indication that these are not good models.
The entire section about prediction of the onset of sexual maturity is interesting and discussion is warranted, but it does not belong in this paper. That is a separate paper. Here it feels tacked on and distracts from the data and results at hand.
The colors used across the age-circumference plots must be made identical…same specimen, same color.
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.