Early evidence of molariform hypsodonty in a Triassic stem-mammal

Hypsodonty, the occurrence of high-crowned teeth, is widespread among mammals with diets rich in abrasive material, such as plants or soil, because it increases the durability of dentitions against wear. Hypsodont postcanine teeth evolved independently in multiple mammalian lineages and in the closely related mammaliaforms since the Jurassic period. Here, we report the oldest record, to our knowledge, of hypsodont postcanines in the non-mammaliaform stem-mammal, Menadon besairiei, from the early Late Triassic. The postcanines are long and columnar, with open roots. They were not replaced in older individuals and remained functional after the total wear of the crown enamel. Dental histology suggests that, convergently to hypsodont mammals, wear was compensated by the prolonged growth of each postcanine, resulting in dentine hypsodont teeth most similar to extant xenarthran mammals. These findings highlight the constraints imposed by limited tooth replacement and tooth wear in the evolutionary trajectories of herbivorous mammals and stem-mammals.


Comment to Melo et al.:
Earliest evidence of molariform hypsodonty in a Triasic stem.mammal. Teeth are essential for animals to disintegrate food items for acquiring the needed energy for their life. Mammals evolved highly differentiated dentitions, whereas teeth of non-mammalian dentitions are often regarded as less sophisticated. Melo et al. describe a highly derived tooth structure in the traversodont cynodont Menadon from the Late Triassic of Brazil. These teeth show -surprisingly -a hypsodont structure. In mammalian dentitions hypsodont teeth were evolved several times independently and in different constructions. The applied methods brought excellent result. In contrast to some other students, they restricted their studies not on the SEM only but combined light microscope and SEM work in a very efficient way. Their finds are original and well documented by the fossil material. The descriptions are clear focusing on the subject. The figures are very informative. The paper of Melo et al. do not only state the hypsodonty for Menadon, but discuss its construction with those of mammals. Therefore, the article provides much more that new data but a very interesting contribution to the problem how different animals cope with the problem of destructive wear during mastication. The careful analysis shows that the teeth of Menadon have some enamel in their early stage, but continue as dentine teeth, thus the hypsodont teeth of Menadon are comparable to the mammalian type of dentine hypsodonty. To aspects, seem to me of special interest. 1) The very early occurrence of hypsodonty already in the Late Triassic. 2) The impressive case of parallel evolution among non related vertebrates. Therefore, I recommend the publication of the paper and its supplementary information warmly in the form they are now.
As requested, we have reorganized and rewritten the manuscript, focusing more on the evolutionary context and making it clearer and more accessible to readers. We looked to reduce the number of specialized terms throughout the text, defined the ones present at their first use and included clarification on the terminology in the Methods section. The Introduction was almost completely rewritten, with some information added from the Discussion section. The Results were streamlined by removing the interpretative parts and adding them to the Discussion. The Discussion section was restructured, with interpretation of the Results and comparations to other taxa in the beginning (three paragraphs), an expanded discussion on the evolution of tooth replacement in gomphodonts (partly from the Supplementary Material), followed by discussions on early instances of hypsodonty in the fossil record and convergences between gomphodontians and extant mammals (slightly modified from the previous version). The final part concerns the palaeoecological context of the period, to which we added clearer statements about climate and a small section regarding age and the Carnian Pluvial Episode. We slightly modified the concluding paragraph to define more explicitly the conditions that probably let to hypsodonty in Menadon. The Methods section was expanded and subdivided to include information on dental terminology, background of the fossil material, and histological methodology. Finally, we added tooth orientations to the figures, and tentatively included taxonomic information and a legend box to fig. 3 for clarity.
We have followed most the specific suggestions and adressed the points raised by the reviewer #1 in the text and figures. Below are the answers to the questions and corrections we. If we overlooked any comment or failed to address it properly, please let us know so that we can correct it.
Lines 6-8 -Abstract: "Hypsodont cheek-teeth evolved independently in multiple mammalian lineages and in the closely related mammaliaforms, but have never been described in other vertebrates, or from before the Jurassic period." "Any evolutionary or environmental explanations? Limited set of teeth in hypsodont taxa (except mole-rats)? ..." I hope to have made this point more clearly in the revised manuscript, but as it is a central aspect of the article, it always merits explanation… In my opinion, the most likely explanation for the timing of the evolution of hypsodonty is more related to developmental constraints than to specific environmental changes. In mammaliaforms, dental occlusion, diphyodonty and oral processing of food are given, and I would expect hypsodonty to evolve when tooth wear becomes too high, as it has happened in so many mammal groups, be it because of drying climates, dietary changes or tectonics.
As most other tetrapods are less reliant on oral processing and can dispose of an indeterminate number of teeth (or do not use teeth at all for this purpose), wear is not such an important limiting factor. Gomphodont cynodonts are an exception in this regard, showing extreme wear in their postcanines, despite their (modified) polyphyodonty. Thus, it makes sense to me that, in this group, it might be advantageous to make the teeth more durable, instead of making more teeth that have to be moved around the mouth in order to be shed. Or at least, the balance between the two evolutionary strategies (durable X abundant teeth) might be more susceptible to influences by environmental factors or evolutionary novelties.
If diphyodonty first evolved during the latest Triassic-Early Jurassic in Mammaliaformes, earlier forms would more probably respond to these issues in other ways, rather than prolonging dental growth. Later (diphyodont) species would be more evolutionarily constrained and hypsodonty might be a more viable "option". Of course, by the Jurassic, the ecological role of "grazer", or medium to large-sized vertebrate herbivore, had been occupied by dinosaurs in most ecosystems, which probably limited the opportunities of hypsodonty to evolve in mammaliaforms, specially until the appearance of social insects and termitophagy/myrmecophagy (possibly around the time of Fruitafossor).
Lines 79-80 -Results: Histology. "Meaning that you only sectioned adult teeth ? But what about the juvenile specimen described in supplementary data?" We only sectioned isolated teeth. The age of the individual can be hard to assess, since adult teeth vary in size depending on their position, but are otherwise very similar. The juvenile lower jaw is, so far, the only one ever found, so destructive analyses could not be performed. We maintained the use of "fibre", as opposed to the American form "fiber", because the rest of the text is written in British English. We actually were in doubt about the preferred variant of English, as we could not find this information in the Nature Communications website (and some Nature journals ask specifically for American or British English).
Lines 153-156 -Discussion. "Studying the morpho-functional properties of the masticatory apparatus of Menadon compared to its sister taxa could be of high interest to understand the presence of hypsodonty (…)" Yes, I agree. Even though detailed morpho-functional analyses are beyond the scope of our study, a small amount of work has already been done in this regard (e.g. Ranivoharimanana, 2012), and I believe some more is currently underway by other research groups, hopefully they will bring better information on the sister taxa, which is also scarce. However, I can remark that the overall cranial/mandibular morphology in Menadon is quite similar to other traversodontids (e.g. fig. 6 of Ray, 2015). One difference seems to be the deeper skull (Melo et al., 2015), likely to accommodate the elongated upper postcanines, but the dentary, for example, is not as deep as you could expect based on the different root shapes of other species, which are much shorter and narrower. Therefore, it would be interesting to understand the morpho-functional characteristics of other (more classical) gomphodonts in order to determine which parameters might be relevant for comparison, and what are their palaeobiological meaning.
Lines 153-156 -Discussion. "(…) And you should slighlty improve the quality of your diagrams (e.g. replacement of teeth in Diademodon, Cricodon...), which are not very clear." I was not completely sure about how to improve the quality, so I updated the diagram and enlarged the teeth of some species, in case they were too small, and added a legend for the symbols. It also seems that the figure quality in the pdf is even more reduced than in the doc file, so I annexed the figures as separate files during submission.
We hope that we were able to improve the manuscript in relation to the last version and look forward to further feedback from the referees.
The authors significantly improved their manuscript, especially the introduction, and they included most of my previous comments and suggestions. I think that even if this paper could be then accepted for publication, the abstract should be a bit more convincing for a wider audience. It also remains some imprecisions in text, and precisions are needed at some point for non-familiar readers, especially regarding the taxonomy. Most of my other comments and suggestions are listed in the pdf, but they are mainly related to minor issues.
I also wish to answer one comment from the authors regarding the evolution of hypsodonty in the present taxon: Authors: "In my opinion, the most likely explanation for the timing of the evolution of hypsodonty is more related to developmental constraints than to specific environmental changes. In mammaliaforms, dental occlusion, diphyodonty and oral processing of food are given, and I would expect hypsodonty to evolve when tooth wear becomes too high, as it has happened in so many mammal groups, be it because of drying climates, dietary changes or tectonics…" My answer: I rather think that increasing the longevity of the dentition in synapsids is tightly related to environmental factors in relation to increasing dental wear. Herbivores generally have more complex teeth to improve plant comminution, but these teeth are difficult to be continuously vertically replaced in relation to developmental constraints or issues (e.g. high energetic cost?). As a result, there are two possibilities to increase the efficiency of the dentition: hypsodonty or continuous horizontal replacement. This latter alternative is generally frequent in polyphyodont synapsids such as traversodontids and tritylodontids, whereas hypsodonty is widespread in mammals. However, some studies, including the present one, showed that hypsodonty can occur in synapsids previously polyphyodont, and continuous horizontal replacement in diphyodont mammals, and both strategies can also occur in the same species (e.g. the silvery mole-rat). The evolutionary and developmental explanations regarding the unexpected presence of these innovations in this group remains to be discovered. The best example is the sirenians, in which the dugong have a degenerate hypsodont dentition, while the manatees have a continuous horizontal replacement. It is thus difficult to explain why close taxa, with a very similar developmental background, present two different dental strategies in relation to grazing. The same question arises concerning Menadon compared to other traversodontids, meaning that there is no simple adaptive explanation, and that should be further investigated.
I would be glad to further discuss these aspects if needed.

Response to Referees Letter
Again, we would like to thank the referee for the helpful suggestions and discussion.
We have followed the suggestions and addressed the points raised by the reviewer in the text and figures. Below are all the reviewer's comments (in bold) and the answers to the questions and corrections that demanded explanation or were not promptly altered in the manuscript. If we overlooked any comment or failed to address it properly, please let us know so that we can correct it.

"The authors significantly improved their manuscript, especially the introduction, and they included most of my previous comments and suggestions. I think that even if this paper could be then accepted for publication, the abstract should be a bit more convincing for a wider audience. It also remains some imprecisions in text, and precisions are needed at some point for non-familiar readers, especially regarding the taxonomy. Most of my other comments and suggestions are listed in the pdf, but they are mainly related to minor issues.
I also wish to answer one comment from the authors regarding the evolution of hypsodonty in the present taxon: Authors: "In my opinion, the most likely explanation for the timing of the evolution of hypsodonty is more related to developmental constraints than to specific environmental changes. In mammaliaforms, dental occlusion, diphyodonty and oral processing of food are given, and I would expect hypsodonty to evolve when tooth wear becomes too high, as it has happened in so many mammal groups, be it because of drying climates, dietary changes or tectonics…" My answer: I rather think that increasing the longevity of the dentition in synapsids is tightly related to environmental factors in relation to increasing dental wear. Herbivores generally have more complex teeth to improve plant comminution, but these teeth are difficult to be continuously vertically replaced in relation to developmental constraints or issues (e.g. high energetic cost?). As a result, there are two possibilities to increase the efficiency of the dentition: hypsodonty or continuous horizontal replacement. This latter alternative is generally frequent in polyphyodont synapsids such as traversodontids and tritylodontids, whereas hypsodonty is widespread in mammals. However, some studies, including the present one, showed that hypsodonty can occur in synapsids previously polyphyodont, and continuous horizontal replacement in diphyodont mammals, and both strategies can also occur in the same species (e.g. the silvery mole-rat). The evolutionary and developmental explanations regarding the unexpected presence of these innovations in this group remains to be discovered. The best example is the sirenians, in which the dugong have a degenerate hypsodont dentition, while the manatees have a continuous horizontal replacement. It is thus difficult to explain why close taxa, with a very similar developmental background, present two different dental strategies in relation to grazing. The same question arises concerning Menadon compared to other traversodontids, meaning that there is no simple adaptive explanation, and that should be further investigated.

I would be glad to further discuss these aspects if needed."
Yes, I absolutely agree. I was addressing the specific points of the rarity/absence of hypsodonty before the rise of mammaliaforms, and what are the prerequisites or important factors for the evolution of this trait. Simple adaptative explanations are often not satisfactory, even for modern groups with known diets and physiologies, and with better temporal and phylogenetic resolution. I was assuming that, although tightly linked to the evolution of hypsodonty, the environment changed considerably over geological time, only leading to hypsodonty in few clades, and was speculating on what these groups have in common.

Line 8 -Not necessary
Lines 13-14 -You should rather emphasize the convergence with crown mammals (i.e. xenarthrans) based on dental histology.

Lines 22-23 -Not necessary.
Line 24 -I would have said "increased food processing efficiency via improved mastication".

Line 30 -Which adaptations?
Line 31 -to->of Line 44 -Do you mean large batteries of occluding teeth and thick enamel? Precise, please.

Line -51-52 You should add precisions on the mode of replacement.
Explained in end of the same paragraph.

Line 56 -Why the occlusion would be disrupted by vertical replacement of teeth?
Is not the case in mammals? (see other comment p.7) Yes, but in the case of mammals, it happens once per locus (at most), usually early in life, with many lineages losing the deciduous dentition altogether, or shedding it in utero. Also, the fast and determinate growth of mammaliaforms means that a permanent dentition will not become obsolete as the skull continues to grow (Luo, et al., 2004;O'Meara & Asher, 2016). This is not true for the majority of vertebrates, which are polyphyodont, with continuous alternate replacement that does not allow for the establishment of occlusal relations between the teeth, affecting the performance in occlusion (Crompton & Jenkins 1968). For comparison, an iguana will replace its teeth about five times per year, while replacement in the basal cynodont Thrinaxodon is estimated to have occurred four to seven times during life (Abdala et al. 2013).
In lineages that developed a form of dental occlusion that depends on correspondence between opposing teeth, it is frequent the suppression of replacements. There are several instances of squamates and sphenodontians with acrodont teeth, displaying occlusion and cessation of replacement (Nydam et al., 2000;Zaher & Rieppel, 1999), as the agamid lizards, with distally erupting acrodont teeth that are never replaced (Cooper et al 1970).
To my understanding, this is the most accepted hypothesis for the evolution of diphyodonty in mammaliaforms. Please see my reply to the comments of p.7.

Lines 61-62 -Precise please, for non-familiar readers.
Detailed in the previous paragraph.
Lines 87-88 -Ok, you are right. But for more accuracy, the dental height vs length should be measured.
These measurements are presented in the Supplementary Tables. As for the degree of convergence between groups (early gomphodonts and early mammaliaforms), I agree that the convergence is not exact in the details, but the underlying evolutionary strategies seem to be very similar. This idea has been pointed out in the literature, for example, Crompton and Jenkins (1968; footnote in page 445) state: "It is interesting to note that in both the gomphodont cynodonts and mammals which develop complex occlusion, alternate tooth replacement was independently lost. This, in part, accounts for the similarities and differences between dental succession in gomphodonts and mammals." I apologise for the long answer, specially if I misunderstood your comment.
Line 217 -The reverse hypothesis could also be suggested: the development of hypsodonty favour a reduction of the dentition and limit the replacement as in some extant mammals.
Yes, the reverse hypothesis is not unfeasible. However, I believe it is more unlikely.
Of the many groups of mammals that developed hypsodonty, only a fraction became monophyodont afterward. On the other hand, many brachyodont species reduced replacements or became monophyodont, for various reasons. Therefore, at least for mammals, that already have a plesiomorphic determinate number of teeth, it does not seem that one is a prerequisite for the other, even there is some influence. In the mammalian case, the determinate dentition undoubtedly evolved first, in Triassic mammaliaforms.
In traversodontids, most of the history of this transition is still unknown.
Considering its rarity, the presence of hypsodonty outside Mammaliaformes seems to demand more explaining than the comparingly mundane limitation of replacement/eruption. In analogy with extant mammals, the reverse hypothesis would be like a species with CDR (e.g. a manatee) evolving some degree of hypsodonty and later reversing to a determinate number of teeth, while my original hypothesis would be more similar to a mammal attaining a determinate number of teeth (as occurred) and then evolving some degree of hypsodonty.  It was removed from a previous version. I have corrected the legend. Line 236 -Do you know if a fossorial hypothesis has been proposed for this taxa based on cranial remains? Hypsodont could also represent an adaptation related to this specialization.
Not for Menadon and not on cranial remains, but for other gomphodont cynodonts, it was proposed based on postcranial material and small cynodonts have been found inside burrows, including the gomphodontian Trirachodon (Groenewald et al., 2001). Menadon and gomphodontians, in general, tended to be larger than other cynodonts, and the known burrows from the Triassic are usually too small. There is no record of burrows in the Santacruzodon Assemblage Zone.
Lines 237-240 -You can also suggest further analysis (FEA, dental microtextures...) to explain the different dental "strategies" between traversodontids, in relation to the morpho-functional aspects of their masticatory apparatus and the putative partioning (or differences) of food resources. But see also my general comment concerning occurrence of hypsodonty vs continuous horizontal replacement in synapsids (in response to your previous comment).
Line 248 -Especially in the case of dentine hypsodonty.
We hope to have further improved the manuscript based on your suggestions.