Could the “flying beasts” have actually flown?


Two life restorations of volaticothere mammals as volant critters: Ichthyoconodon jaworowskorum by Dylan Bajda and Argentoconodon fariasorum by Ceci Thomas ( @alphynix ).

Paleontology, even by the standards of science in general, tends to be prone to extensive and radical paradigm shifting. After all, when all you have are bits and pieces of animals long gone, you’re bound to have an incomplete image to begin with, and any further complexity will be undeniable surprising and expectation-shattering. Functional biology is always one step being caution, after all.

Such a shift has happened to our understanding of Mesozoic mammals. Historically, they have been mostly dismissed as small, shrew-like insectivores living in the shadow of the dinosaurs and other Mesozoic ‘reptiles’. At best, one could hope that the “rodent-like” multituberculates would be accepted as  token of ecomorphological diversity, but this could be seen as mattering relatively little to even most professionals. The idea, simply put, was romantic: that our distant ancestors were small, indistinct and humbled in the face of a more attractive animal group.

But in recent years, various discoveries have shown to not be the case. Large eutriconodonts like Gobiconodon, Jugulator and the infamous Repenomamus have turned out to be specialised carnivores and certainly predators of small dinosaurs, the latter one of the largest of all Mesozoic mammals and comparable in size (and most certainly in habits) to modern wolverines, capable of hunting dinosaurs. This inversion is shocking as it is, but other equally spectacular ones have taken place as well: Castorocauda, a scaly-tailed swimmer, just joined by the paddle-limbed Liaoconodon and Yanoconodon; Vintana, a groundhog-size herbivore; the south american mesungulatoids, also herbivores and some of which as large as a dog; Lotheridium and Cronopio, “sabertoothed” mammals (albeit the former a carnivore and the latter a herbivore); Fruitafossor, an anteater-like form; various types of tree-climbers and diggers, among countless others.

Simply put, Mesozoic mammals were a highly diverse clade of amniotes, occupying a vast array of ecological niches. As far as we know they did not achieve sizes as large as other groups (both lizards, terrestrial crocodylomorphs and pterosaurs have all managed to produce large and imposing forms, for example), but it is clear that virtually every ecological niche under 50 kg was experimented upon, and it would not be surprising if there were even larger species waiting to be discovered.

Even with this in mind, however, one can’t help but be shocked by one particular possibility: that a group of mammals in the ‘Age of the Dinosaurs’, long before there were any bats, truly took to the air, and developed powered flight.

The Flying Beasts

Volaticotherium antiquus holotype.

All the way back into 2006, a particular Mesozoic mammal discovery made the headlines. In the chinese mid-Jurassic Daohugou Beds a spectacular new discovery was found: a mammal fossil bearing imprints of a patagia. Christened Volaticotherium antiquus, it was unambiguously regarded as a glider, the first mammal to take tentative steps into the air 70 million years before placentals and marsupials did so.

Though frequently presented in articles (including in professional publications) as a bizarre evolutionary dead-end, the truth is that Volaticotherium was considered in its debut paper (Meng et al 2006) to be closely related to another Mesozoic mammal, the Cretaceous-aged Ichthyconodon jaworowskorum. Hailing from the Berriasian Anoual Syncline Beds of Morocco, this taxon was first described 11 years prior by Sigogneau-Russell. Its remains are drastically less complete than Volaticotherium’s, being composed of just two molars. However, they are enough to paint a rather strange picture of the animal’s habits: occuring in what was once a littoral environment their cusp sharpness and preservation of several details entirely rules out aquatic transportation. Combined with a rather atypical cusp-morphology (now known to be actually les extreme than in other volaticotherians; Meng 2006 and Gaetano 2011) Sigogneau-Russell interpreted this as a signs of an aquatic lifestyle.

Not too soon after Volaticotherium’s description came the discovery of yet another related taxon, Argentoconodon fariasorum. Occuring in the Toarcian-aged Cañadon Asfalto Formation in Patagonia, it was originally known only from a single molariform tooth, but in 2011 more complete material was discovered, including several sections of the skull and some post-cranial remains. Of most importance is a femur, which is remarkably similar to Volaticotherium’s and outrightly indicative that it too was capable of aerial locomotion.

Since then, these three taxa have been displayed as a sort of “trinity”, forming the clade known as Volaticotheria, Volaticotheridae or Volaticotherini depending on whereas the group is considered independent from other mammals (Meng 2007), a distinct family within Eutriconodonta (example: Averianov 2012), or as a tribe within the eutriconodont family Triconodontidae, and specifically within the subfamily Alticonodontinae (Gaetano 2011). All post-Meng 2007 papers have considered volaticotherines to be within Eutriconodonta, and most recent phylogenetic studies favour the nature of this group as the sister-group to Triconodontidae (examples: Averianov 2011 and Martin 2015). I will default to the original term, Volaticotheria, for the rest of this article.

Other two taxa have also been attributed to this clade, both found in North America: Jugulator amplissimus and Triconolestes curvicuspis. The former dates to the mid-Cretaceous Cedar Mountain Formation and is recovered as a sister-group to the rest of Volaticotheria in both Gaetano 2011 and Rougier 2012; it is by far both the largest taxon both in this clade as well as in its environment, estimated to weight at least 750 g, and bears some speciations in its incisors that suggest predatory habits (Cifelli 1998). The latter is a single half of a molar from the Jurassic Morrison Formation, originally classified as an amphilestid but more reently recovered as a volaticotherine (Averianov 2011).

Long-lived and far-travelled

Localities with volaticotherine taxa in time and space.

As you can see on the above map, volaticotherines were a rather long lived, borderlinely cosmopolitian clade. Just including Argentoconodon + Ichthyoconodon + Volaticotherium we get a clade spanning at least 40 million years in both Gondwanna and Laurasia; the Daohugou Beds are uncertainly placed, with some sources favouring an Aptian Cretaceous age (Ren 2002), in which case it could potentially be a span of 61 million years. If Jugulator is a volaticotherine, then it becomes a range of 78 million years.

This is rather unusually successful for gliding mammal clades, which are generally more specialised and local forms. Gliding is, after all, a rather specialised ecological niche, relying on rather stable forest environments to thrive; this is why most mammalian gliders tend to be relatively restricted ‘oddities’.

The closest in terms of global and temporal distribution are flying squirrels, spanning across Laurasia for about 20 million years. Eomyidae, a lineage of rodents spanning from the Eocene to Pleistocene, has produced more than one gliding form, but it is unclear if they form a consistent lineage or several independent jabs at gliding; same for Gliridae, of which several species are also gliders (Jackson et al 2012). Eocene relatives of the anomalures are known, but it is unclear if they were already gliders or if this is a more recent speciation. Same for colugos; though Plagiomenidae and Mixodectidae were once cited as gliding colugo relatives spread across the North Hemisphere, they are now considered to be unrelated, terrestrial mammals (Yapuncich 2011).

Of course, in and of itself this may not be an indicator for powered flight. After all, Mesozoic ecosystems tended to be more stable than the rapidly shifting Cenozoic, and another lineage of arboreal synapsids, the haramiyidans, was exceptionally long lived. However, Jurassic mammal faunas were rather segregated, and the widely disparate nature of the volaticotherian geographical range is matched by relatively few other mammal groups in this epoch.

Where they should not be


Ichthyoconodon jaworowskorum molar.

The wide spatio-temporal range is also complimented by some rather unusual localities for volaticotherian fossils. Two of the unambiguous volaticothere taxa, Ichthyoconodon and Argentoconodon, occur in places where one would not expect eutriconodont mammals to exist, and in the former’s case any mammals at all.

Ichthyoconodon teeth are present in the littoral sediments of the Anoual Syncline Beds. Though other mammal remains do in fact exist on these beds, Ichthyoconodon is notable because it is one of the only two which probably died in situ. Its molars preserve both the sharpness of their cusps as well as other fragile elements, ruling out the possibility that the teeth underwent long aquatic transportation (Sigogneau-Russel 1995). As such, Ichthyocondon’s remains were not carried downriver; the animal had to have died in the sea.

This has been interpreted by Sigogneau-Russell as evidence of aquatic and piscivorous habits, in particular in regards to the morphology of the cusps, interpreted as similar to those of cetaceans and seals. Posterior writers have doubted that eutriconodont teeth were adapted for piscivory, however, and indicate that their occluding function would have lead to a more carnassial-lik shearing function rather than the interlocking molars of piscivorous mammals (Kielan-Jaworowska et al 2004). Nonetheless, actual aquatic eutriconodonts have been found (Chen 2015).

The hypothesis that Ichthyoconodon was a piscivore can therefore not be fully disregarded. However, in spite of the fact that volaticotherian molars are notable for being rather unusual, possessing strangely recurved cusps which would have generated a rather odd occlusion pattern (Meng 2006), the most recent consensus appears to be that they had the same shearing function as in other eutriconodonts (Butler et al 2016). At any rate, Ichthyoconodon’s molar cusps are noted as being less recurved and deep than those of Volaticotherium and Argentoconodon (Meng 2006, Gaetano 2011), so they could have been functionally different.

Argentoconodon, though not occuring in a marine site, similarly shows up in a rather unusual locality. It is one of the only two south american eutriconodonts – the other being Condorodon spanios, which occurs in younger deposits -, the mammalian fauna in its environment being otherwise composed of australosphenidans and an undetermined ‘allothere’ (Gaetano 2012).

The fact that it occurs in a fauna otherwise unpopulated by other eutriconodonts is rather suspicious, but its early age is the truly intriguiging part. Eutriconodont fossils are known only from the latest part of the Early Jurassic, and it could suggest that volaticotherians dispersed worldwide rather quickly. This is not very consistent with gliding mammal distribution ranges – overwater dispersal is pratically unknown in gliding mammals -, but would fit a true flying mammal.

Ecological range


Jugulator amplissimus by Julio Lacerda.

Eutriconodonts as a whole were animalivorous mammals, possessing features such as long canines (or canine-like incisors), premolars with trenchant main cusps, shearing molariforms, a strong mandibular abductor muscle and what Kielan-Jaworowska called the “bone-crushing ability” (Kielan-Jaworowska et al 2007). Depending on the animal’s size, this would presumably range from small insects to dinosaurs, and in spite of their appearently hypercarnivorous habits eutriconodonts were highly successful.

Volaticotheres are thought to have had the same type of molar function in spite of their cusp weirdness (Butler et al 2016), and both Volaticotherium and Argentoconodon certainly do possess large canines; the original paper even compared Volaticotherium‘s canines to those of insectivorous bats (Meng 2006). Most volaticotheres probably hunted insects and other invertebrate prey, perhaps hard-shelled beetles and other such species based on their deep cusps, but at least Ichthyoconodon and Jugulator were large enough to have eaten vertebrate prey. The latter in particular preserves its medial incisors, which possess mitten-like cutting edges, suggesting speciation towards proportionally larger prey (Cifelli 1998).

Gliding mammals as a whole feed on stationary food sources; gliding is in fact thought to be a mechanism to cover large areas in search of stable food sources (Jackson 2012). Gliding mammals alive today are mostly either arboreal herbivores (anomalures, colugos, some marsupial gliders, some flying squirrels), nectar-drinkers (sugar-gliders), or fungivores (most flying squirrels); evidence suggests that extinct gliding rodents and metatherians probably did not diverge much in this regard.

Gliding dedicated animalivores are absent in mammals; gliding frogs and reptiles are insectivorous, but they are ectothermic, so they might not be a good functional model. As such, it is possible that powered flight would have been directly implicated by their carnivorous habits.

Assuming that they are flyers, volaticotherians follow more or less the same size range as modern predatory bats, and could have been similar ecologically. Competition with pterosaurs would probably have been minimal, as these did not overlap ecologically with birds and other flying dinosaurs, and were generally larger. Interestingly, volaticotherians disappear from the fossil reccord roughly at the same time as non-pterodactyloid pterosaurs, which depending on the exact date (either earliest Cretaceous or Cenomanian) could coincide with several floral upheavals; but like them they could have simply not have been preserved and instead survived much longer.

What can Volaticotherium itself say?

Discussing potential flying habits on volaticotherians as a whole would be rather pointless without the most complete volaticotherian fossil, Volaticotherium itself. In particularly, a number of aspects are rather relevant to this discussion, and could be a deciding factor on whereas volaticotheres were truly volant or not.

First off is the patagium itself. The original paper describes it as being proportionally rather large in relation to the body, considerably more so than in gliding eomyid fossils (Meng 2006). Considering gliding eomyids were probably comparable in propotions to extant gliding squirrels (Storch 1996), this would have been fairly extensive indeed, perhaps comparable to a colugo assuming the animal was a glider and not a flyer. Inspite of this, the patagium extends to the proximal vertebrae of the tail, though the rest of the tail vertebrae are flatened and it bears long, thick hair (this inspired the strange tail shape of the above depictions), which would fit well for either a glider or a flyer, as both would need a balance between lift and hindlimb mobility. The patagia are appearently covered in fine, even fur; though most bats and pterosaurs have naked wings, anurognathids set a precedent in having fuzzy patagia. Unlike pterosaur membranes there are no actinofibrils, though these are sometimes only detectable by UV scans.

Second is the lack of a styliform element, at least as far as examinations of the holotype have shown. Styliforms, though not obligatory in mammalian gliding, are very widespread to the point that it’s hard to find gliding mammals without one. In most species it is located in the elbow (anomalures, colugos, marsupial gliders, eomyids) or on the wrist (flying squirrels, as well as on the dinosaur Yi). By contrast, true flyers do not have styliforms on the brachiopatagium, though pterosaurs and bats have similar elements on “secondary” membranes, the pteroid on the propatagium and the calcar on the uropatagium respectively. It is possible that the absent of a styliform could be a factor weighting on a volant interpretation of volaticotherians, and it should be the subject of research in the future. In the Ichthyoconodon depiction, the “wing finger” can be interpreted as a rather elongate styliform akin to that of a flying squirrel, while the uropatagium features a speculative one derived from the tarsal spur omnipresent in non-therian mammals.

Finally, and perhaps most importantly, there is the hand. The hand of Volaticotherium is notorious for being “poorly preserved” (Meng 2006; see side materials), preserving only four metacarpals and three phalange bases. These are compared to the hindlimb phalanges, assumed to be rather similar due to shared adaptations towards flexing, but the poor state of preservation could mean that parts of the latter are absent, and at any rate they are noted as being proportionally long in relation to the metacarpals. A pterosaur-like wing finger is noted to be absent, and the flexing digits could suggest a more bat-like wing structure, but again the missing elements could simply be due to the preservation status.

In summary, current evidence shows proportionally large wings and an incomplete hand, which is pretty fittingly balanced in regards to the ambivalence of the animal’s flight capacities. Further research on the structure of the membrane, and whereas it has a styliform and/or actinofibrils (or other stiffening agents, and whereas they would be compatible with a flight stroke), should be a priority in future research.

An important aside

The range of forelimb diversity within early mammals is a periodically debated issue, mostly in relationship to modern marsupials. It is clear that most if not all non-placental mammals produced parembrionic, mobile young, due to the presence of epipubic bones and their prevention of abdomen expansion and thus true pregnancy (Schulkin 2012, Kielan-Jaworowska 2004). Due to this, the forelimb develops rather early and requires a function in grasping; this is thought to restrict the forelimb range in marsupials, and thus prevent the presence of true powered flight (Kelly 2011).

However, this can be put into question when one examines both living and extinct non-placental mammals. Monotremes, like marsupials, also produce fetal young and crawl around; in their case there is in fact a greater pressure for a grasping limb, since at least juvenile echidnas are forced to constantly move about in search of milk glands (Schneider 2011). Yet, monotremes display quite derived forelimbs, both in the webbed, long-digit appendages of the platypus and on the stout, fossorially-inclined forelimbs of the echidnas.

Likewise, a particular extinct mammaliform, Docofossor, shows phalange reduction similar to that of the placental golden moles (Luo 2015). Yet, docodonts, like most Mammaliaformes aside from placentals, had epipubic bones (Rose 2009). This seems to suggest that digit speciation may not be as limited among non-placentals as previously thought.

There is clear evidence of volaticotherine patagia, which in Volaticotherium is shown to at least “sandwich” the digits, implicating their role in its support (Meng 2006). Therefore, the possibility of actual wings is not too far off, at least potentially.


Extraordinary claims require  extraordinary evidence, and I gladly admit that the above are currently just mere suggestions, and may probably forever be so. However I do feel that there are issues worth addressing in regards to the current interpretations of volaticotherians as gliding rather than actually volant mammals, and hopefully this will be inspire discussion in some way.

Either option is amazing in and of itself: either a group of highly tenacious, able-to-disperse-overwater and carnivorous gliding mammals like no other that came after, or a fourth group of truly flying vertebrates. Gliding or flying, volaticotherians display the sheer ecological diversity of not just Mesozoic mammals, but specifically of Eutriconodonta, already a fairly successful and divergent group spanning over 120 milion years of evolution on land, tree-canopies, waters and, perhaps, air.


Meng, J.; Hu, Y.; Wang, Y.; Wang, X.; Li, C. (2006). “A Mesozoic gliding mammal from northeastern China”. Nature. 444: 889–893. doi:10.1038/nature05234.

Sigogneau-Russell, Denise (1995). “Two possibly aquatic triconodont mammals from the Early Cretaceous of Morocco” (PDF). Acta Palaeontologica Polonica. 40 (2): 149–162.

Gaetano, L.C.; Rougier, G.W. (2011). “New materials of Argentoconodon fariasorum (Mammaliaformes, Triconodontidae) from the Jurassic of Argentina and its bearing on triconodont phylogeny”. Journal of Vertebrate Paleontology. 31 (4): 829–843. doi:10.1080/02724634.2011.589877.

A. O. Averianov and A. V. Lopatin. 2011. Phylogeny of Triconodonts and Symmetrodonts and the Origin of Extant Mammals. Doklady Biological Sciences 436:32-35 [M. Uhen/M. Uhen]

Thomas Martin, Jesús Marugán-Lobón, Romain Vullo, Hugo Martín-Abad, Zhe-Xi Luo & Angela D. Buscalioni (2015). A Cretaceous eutriconodont and integument evolution in early mammals. Nature 526, 380–384. doi:10.1038/nature14905

L. C. Gaetano and G. W. Rougier. 2012. First amphilestid from South America: a molariform from the Jurassic Cañadón Asfalto Formation, Patagonia, Argentina. Journal of Mammalian Evolution

Cifelli, Richard L.; Madsen, Scott K. (1998). “Triconodont mammals from the medial Cretaceous of Utah”. Journal of Vertebrate Paleontology. 18 (2): 403–411. doi:10.1080/02724634.1998.10011068.

Ren, D.; et al. (2002). “On the biostratigraphy of the Jurassic fossil beds at Daohugou near Ningcheng, Inner Mongolia”. Geol. Bull. China. 21: 584–591.

Jackson, Stephen Matthew and Schouten, Peter. Gliding Mammals of the World, Csiro Publishing, 2012

Gabriel S Yapuncich , The first dentally associated skeleton of Plagiomenidae (Mammalia, ?Dermoptera) from the late Paleocene of Wyoming , Conference Paper · November 2011 DOI: 10.13140/2.1.1302.4322 Conference: Society of Vertebrate Paleontology 71st Annual Meeting, At Las Vegas, NV

Zofia Kielan-Jaworowska, Richard L. Cifelli, Zhe-Xi Luo (2004). “Chapter 7: Eutriconodontans”. Mammals from the Age of Dinosaurs: origins, evolution, and structure. New York: Columbia University Press. pp. 216–248. ISBN 0-231-11918-6.

Meng Chen, Gregory Philip Wilson, A multivariate approach to infer locomotor modes in Mesozoic mammals, Article in Paleobiology 41(02) · February 2015 DOI: 10.1017/pab.2014.14

Percy M. Butler; Denise Sigogneau-Russell (2016). “Diversity of triconodonts in the Middle Jurassic of Great Britain” (PDF). Palaeontologia Polonica 67: 35–65. doi:10.4202/pp.2016.67_035.

G. Storch, Oldest fossil record of gliding in rodents, Nature 379, 439 – 441 (01 February 1996); doi:10.1038/379439a0

Nanette Yvette Schneider, The development of the olfactory organs in newly hatched monotremes and neonate marsupials, J Anat. 2011 Aug; 219(2): 229–242. Published online 2011 May 17. doi: 10.1111/j.1469-7580.2011.01393.x

Kenneth D. Rose, The Beginning of the Age of Mammals , JHU Press, 31/08/2009 – 448 páginas

Author: Carlos Albuquerque

Bisexual, portuguese and proud. Interested in paleobiology, esoterism/occultism and other stuff.

14 thoughts on “Could the “flying beasts” have actually flown?”

  1. We then could say that mesozoic mammal diversity could be equivalent of modern mammalian diversity minus large ungulates, carnivores, cetaceans, bats, primates and many rodents. Still pretty high. Or am I mistaken?
    What lizards surpassed the largest mammalian sizes in the Mesozoic? I mean only terrestrial lizards.
    Sugar gliders still are quite enthusiastic animal eaters, and can even eat rodents. It is not the primary part of their diet, but I want to say that there exist modern gliders with partly animalivorous diets.
    Last, isn’t it more plausible that the gliding mammal could be blown at sea by a strong wind, and subsequently fall in the water alive to die there?

    1. Yes, the largest Mesozoic lizards surpassed contemporary mammals in size. Some lizards like polyglyphanodontids also had chewing teeth, so we can assume lizards were among the chief competitors of mammals in the Mesozoic. Nonetheless, mammals appearently had an edge in terms of ecomorphological spaces.

      Sugar gliders are in fact capable of taking vertebrate prey, and if volaticotherians were gliders they’d be among the first modern analogues to consider. However, gliding marsupials are still primarily herbivorous.

      Finally, maybe, but I’m not aware of any gliders being dragged by winds. Most gliding mammals choose their landing location prior to launching, so they are very precise.

  2. The problem is that gliders aren’t adapted to counter storms, which claim the lives of many actively flying animals as well. That might be a reason why many terrestrial resident birds and insects lost flight ability in remote islands, discussed from Darwin’s time, who also observed that phenomenon of birds being blown out to sea by air currents. He also hypothesised that terrestrial birds and bats reached remote islands by the same roote.
    Isn’t modern teiid dentition quite similar to polyglyphanodonts?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: