Dude, where's my astrapothere?

Anyone even vaguely interested in fossil mammals has the same problem at some time or another. You repeatedly encounter the same bizarre and fascinating beasts, long to know more about them, and yet have to endure a lifetime of frustration in the absence of any good, comprehensive information. Typically, one species in the group - usually, the most geologically recent, or the biggest, or the first-named - is figured and mentioned in all of the books, while all of the others languish in obscurity and may as well not exist.

i-12e0c95d953f20923bd4382c2b764a2f-Wikipedia Astrapotherium.jpg

An excellent example of this sort of thing is provided by the astrapotheres: a poorly known, rarely discussed group of obscure South American quadrupedal herbivores that don't quite resemble any living mammals, but would perhaps have combined elements of proboscideans, rhinos, tapirs, pigs and primates. They've been mentioned on and off here at Tet Zoo for, I dunno, probably more than a year now, and - finally - their time has come...

Astrapothere fossils are presently known from the Upper Palaeocene (Riochican*) to the Middle-Upper Miocene (Friasian*), and their diversity peaked in the Lower and Middle Miocene (Colhuehuapian and Santacrucian*). Conventionally assumed to have been restricted to the island continent of Cenozoic South America, we now know that they were also present on Antarctica during the Eocene (Hooker 1992), as were other 'South American' mammal groups like sloths and litopterns. The best known astrapothere is Astrapotherium magnum, named in 1879 for material from the Miocene Santa Cruz Formation ash beds of Patagonia. Astrapotherium remains proved to be not uncommon in the Santa Cruz Formation, and an excellent, complete skeleton discovered by Elmer Riggs of Chicago's Field Museum, and described by him (Riggs 1935) and later by William Berryman Scott (1937), makes this the best known astrapothere, and the only one for which a full skeletal reconstruction exists [Scott's reconstruction shown below: image at top is life restoration of A. magnum from wikipedia]. Astrapotherium is also known from the Miocene of Chile.

* These names are South American land mammal ages (SALMAs).

i-a90dee2feccefc7ef0fc4f299d9c54ab-Astrapotherium Astrapotherium always bloody Astrapotherium.gif

Astrapotherium means 'lightning beast': as Scott (1937) wrote, 'A more inappropriate and infelicitous name could hardly have been selected, for anything less lightning-like than this strange beast cannot be imagined' (p. 309). Hermann Burmeister, the person responsible for the name, wanted to mirror the etymology of the brontotheres ('thunder beasts'), a group that Burmeister imagined as superficially similar to (and possibly allied to) the astrapotheres. Simpson (1980) suggested that Burmeister chose this name because astrapotheres were 'big and probably noisy', but this doesn't match with Burmeister's explanation. Florentino Ameghino, recently described as 'undoubtedly the leading figure in Argentinean paleontology' (Salgado 2007, p. 4), named the new 'family' Astrapotheriidae in 1887 and Richard Lydekker coined the new higher group Astrapotheria in 1894 (Ameghino came up with Astrapotheroidea in the same year, but it failed to gain wide usage).

Key astrapothere characters include the presence of enlarged canine tusks (see skull and teeth of Astrapotherium shown below), the form of the lower molar cusps, and a highly distinctive, flattened astragalus with a bowl-like medial articular facet. We infer from reasonably good samples of what appear to be individuals of the same species that astrapotheres exhibited sexual dimorphism in tusk size, shape and configuration. In one sex (inferred to be male), the lower jaw tusks are proportionally longer and thicker, and project directly forward. In the other sex, the shorter, slimmer tusks curve outwards somewhat. The short wrist, foot and ankle bones, columnar limb bones and long proximal limb segments of astrapotheres indicate that they were graviportal, like elephants or sauropods: suited for weight bearing and walking, and not good at running. Their hands were digitigrade but the small hindfeet were plantigrade (more on the feet later). It's difficult to say whether astrapotheres were incapable of running entirely: it used to be assumed that graviportal animals literally couldn't run, but this is no longer thought correct (Hutchinson et al. 2003).

i-b324c8cc406875e633d55ae095f391c7-Astrapotherium skull.jpg

Astrapotheres appear to have been superficially rhino-like, but several of their skull characteristics recall those seen in tapirs and proboscideans, and demonstrate that - in astrapotheriid astrapotheres at least - a proboscis was present. The narial opening is enlarged, the nasal bones are small and retracted, and the lower jaw is longer than the upper jaw. The skull (which, in Astrapotherium, is about 60 cm long) is tall with a domed forehead; sinuses apparently occupy much of the space here. In another parallel with proboscideans, both the upper and lower tusks were enlarged (living elephants are unusual among proboscideans as a whole in lacking mandibular tusks). These tusks appear to have grown continuously throughout life, and a shearing contact between the tips of the upper tusks and the back surface of the lower tusks kept the upper tusks sharp (hippo tusks are kept sharp in the same way).

An obvious diastema separates the tusks from the molars (premolars are absent in all but some of the most basal taxa). The tusks of some species were up to about 1 m long, and with wear marks indicating that they functioned in foraging and feeding. Upper incisors (and lower incisors in one species) are absent and it's been suggested that a horny pad might have been present at the front of the mouth, similar to the one possessed by bovids and other artiodactyls. The incredibly broad, massive cervical vertebrae show that astrapotheres had a disproportionately thick neck. More on these features and their significance later on.

So, we've all heard of Astrapotherium magnum, type species of the astrapothere groups Astrapotheriidae and Astrapotheriinae. But there's so much more than this. Ameghino named Parastrapotherium in 1895, and Astraponotus, Albertogaudrya and Astrapothericulus in 1901. An assemblage of taxa now widely regarded as basal astrapotheres, the (probably paraphyletic) trigonostylopids, were recognised by Ameghino during or prior to 1901. During the 1920s, the description of Uruguaytherium and Xenastrapotherium led to the recognition of the new astrapotheriid group Uruguaytheriinae. New taxa, like basal Eoastrapostylops and Tetragonostylops (both from the Palaeocene), several new species of Xenastrapotherium, and the particularly big, long-trunked Granastrapotherium, were named between the 1960s and 1990s. It actually makes me cross that the authors who write those big overview-type books never even mention any of this diversity; it strikes me as lazy, but I try not to dwell on it.

i-e4669e0af995e2c26b4fb0bec916c18b-Simpson crappy trigonostylopid life.jpg

Phylogenetic studies indicate that several taxa traditionally grouped together as the 'trigonostylopids' are the most basal known astrapotheres (Cifelli 1993). These tend to be smaller than astrapotheriids (the most derived astrapothere clade): Tetragonostylops, for example, has been described as 'fox-terrier size' (that term might sound familiar if you know fossil horses and Stephen Jay Gould's articles). These basal forms have much smaller tusks, lower-crowned teeth which are triangular rather than quadrangular, and a lower forehead with a less retracted nasal region, than astrapotheriids. The skull of the best known form, Trigonostylops wortmani, does have a somewhat domed cranium containing large sinuses, but it's sufficiently different from that of Astrapotherium for Simpson (1933) to conclude that 'trigonostylopids' might be more to do with litopterns than astrapotheres. Simpson in fact thought that 'trigonostylopids' should be placed in their own higher taxon, Trigonostylopoidea [adjacent picture is Judy Spencer's less than excellent life restoration of the head of Trigonostylops from Simpson (1980). I never liked the doggy little ears. Image below shows palate of Trigonostylops. Note absence of upper incisors and tusk-like canines].

i-9a2038f48178af05e23cf8ed5ec39bd0-Trigonostylops palate.jpg

Eoastrapostylops (originally described as worthy of its own 'family', Eoastrapostylopidae Soria & Powell, 1981) was found by Cifelli (1993) to be the most basal astrapothere, and it differs from most other basal forms in still possessing premolars (Scaglia from the Early Eocene also still had premolars). Another basal form, Albertogaudrya from the Early Eocene (known only from teeth), was also originally described in its own family (Albertogaudryidae) and has even been suggested at times to not be an astrapothere, but some sort of astrapothere-mimicking (!) 'condylarth' (Fittkau 1969).

Ok, a lot more to get through yet, but it'll have to wait. Astrapothere lifestyles and affinities next, as well as astrapotheriines and uruguaytheriines.

Refs - -

Cifelli, R. L. 1993. The phylogeny of the native South American ungulates. In Szalay, F. S., Novacek, M. J. & McKenna, M. C. (eds) Mammal Phylogeny: Placentals. Springer-Verlag (New York), pp. 195-216.

Fittkau, E. J. 1969. Biogeography and Ecology in South America. Volume II. Springer-Verlag, Berlin.

Hooker, J. J. 1992. An additional record of a placental mammal (Order Astrapotheria) from the Eocene of west Antarctica. Antarctic Science 4, 107-108.

Hutchinson, J. R., Famini, D., Lair, R. & Kram, R. 2003. Are fast-moving elephants really running? Nature 422, 493-494.

Riggs, E. S. 1935. A skeleton of Astrapotherium. Geological Series of Field Museum of Natural History 6, 167-177.

Salgado, L. 2007. Patagonia and the study of its Mesozoic reptiles: a brief history. In Gasparini, Z., Salgado, L. & Coria, R. A. (eds) Patagonian Mesozoic Reptiles. Indiana University Press (Bloomington & Indianapolis), pp. 1-28.

Scott, W. B. 1937. The Astrapotheria. Proceedings of the American Philosophical Society 77, 309-393.

Simpson, G. G. 1933. Structure and affinities of Trigonostylops. American Museum Novitates 608, 1-28.

- . 1980. Splendid Isolation: the Curious History of South American Mammals. Yale University Press (New Haven and London).


More like this

Darren, you wrote they

combined elements of proboscideans, rhinos, tapirs, pigs and primates

but you missed diprotodontoid marsupials - the plantigrade feet, overall size and proportions suggest Diprotodon or Zygomaturus, while the proboscis is apparently shared with Palorchestes. The loss of upper incisors and hippo-like canines are something else though. Nothing so far suggests primates, that must be coming next.

By John Scanlon, FCD (not verified) on 03 Feb 2008 #permalink

Great article! The weird mammals from down south always seem to get the short shrift.
My question relates to the common artistic restorations of astrapotheres and their fellow weirdos the pyrotheres. They always seem to be restored as hairless with a mostly-bare, elephant-type skin (as seen in the wiki picture). Is this really just an artistic choice based upon their resemblance to the elephants?

By Adam Pritchard (not verified) on 03 Feb 2008 #permalink

Astrapotheres at last :-)! Pyrotheres next?

Thanks for comments. John you're very right about similarities with diprotodontoids - in fact some astrapothere restorations are very, very similar to those depicting palorchestids. The primate similarity comes from the hindlimbs: they vaguely recall those of hominids. Vaguely.

and their fellow weirdos the pyrotheres.

And the xenungulates... and who knows what else...

By David Marjanović (not verified) on 04 Feb 2008 #permalink

Great post! I've been interested in South American "ungulates" for some time, and I still know far too little about them.
Apparently the naming of Astrapotherium also involved something similar to the recent aetosaur events. If I remember this correctly, Burmeister had a look on the specimens that Moreno was studying, and then named the animal before Moreno could. Moreno later named the same animal Mesembriotherium, but couldn't convince anyone else to use that name.
As for Albertogaudrya, Simpson later classified it as a "trigonostylopoid" (i. e. not an astrapothere in his view). I've never heard about the "condylarth" idea before, but of course "condylarths" are a wastebasket, and it could mean essentially the same thing.

By Lars Dietz (not verified) on 04 Feb 2008 #permalink

Finally astrapotheres get their due! When I was a PhD student at the UCMP I came across some big bones in the collection that I wanted to CT scan, and Pat Holroyd had to explain to me what the hell an astrapothere was. Lovin em ever since.

Good point about the hindlimbs, astrapothere/other "graviportal" mammal (incl. elephant) limbs do have some primate similarities, a point to be addressed in a coming paper if it ever gets published...

Once I showed a pic of a frozen elephant foot that was bandsawed sagittally in half to my orthopaedic surgeon and he didn't believe me that it wasn't a photoshopped human foot for a while, then demanded a copy to stump his students with. Good times, good times.

Okay, Darren, these animals are officially awesome. Now I have a question. In one of my older "prehistoric encyclopedia" tomes, there was a bugger whose name now escapes me, but I shall try to describe it.
Long, hippo-like body, long head, short trunk. All four canines are long and thin, but all point forward. The animal was described in the text as being semi-aquatic.
I really thought that bugger was cool, but I've since sold the book. It looked like these astrapotheres, but with more forward-pointing canines.

Zach - are you sure it wasn't a life restoration of Pyrotherium? Google-image the name and you'll get hits for the few published restorations.

In response to earlier comments from Adam and others on grey, mostly-bare skin - I suppose artists normally depict the animals this way by analogy with elephants and rhinos. There are, however, a few life restorations showing astrapotheres with Malayan tapir-like colours. A well known painting of Trigonostylops shows it as furry and nicely patterned in tan and dark brown. But yes, mostly grey, mostly.

Darren, thanks. That's definately not the name, but it looked just like Pyrotherium. I will try to find that book at the used bookstore (where I sold the book to--it's probably still there) and get the name. I remember it fondly, which is why I'm frustrated that my mind refuses to relenquish its name!

Darren, Dude, that's a funky-feeted hippir~lipped elephino, but where's the burnt-orange mohawk mane? Oh, and on both 'live' pics, the "ear-lets" are killing me; but I guess they aren't too far off.
Great post, love those flat feet, wondering if they had dense rear limbs and if they waded bipedally sometimes.

Zach - could your mystery beast have been one of the desmostylians? Desmostylus and Palaeoparadoxia are the most commonly-illustrated ones, and I've seen some restorations that might fit your description.
Come to think of it, I think the desmostylians are weird enough to merit a post of their own sometime!

By Dave Hughes (not verified) on 04 Feb 2008 #permalink

Zach, not an expert, but it sounds a bit like Moeritherium. Possible?

Dammit Dave, you beat me to it! I was going to suggest Palaeoparadoxia

Im going to sulk all day now.

(this must be the coolest tetrapod name ever IMHO)

But we must have a post on the Desmotylians; an elephant that thinks its a walrus is pretty cool.


Could it have been Gomphotherium (alias Trilophodon, Tetrabelodon)?

Great post, love those flat feet, wondering if they had dense rear limbs and if they waded bipedally sometimes.

With those hips? You're kidding. The legs are simply what graviportal mammalian hindlegs look like -- see the comment above on elephants. We aren't graviportal because we can still bend our ankles, but most of the time we act as if we were graviportal anyway.

For wading, I'd recommend a quite different anatomy with less water resistance, as seen in wading birds.

By David Marjanović (not verified) on 05 Feb 2008 #permalink

Although not in the top 5, this post is now featured in "Now on ScienceBlogs".

By David Marjanović (not verified) on 05 Feb 2008 #permalink

Amazing and mysterious creatures!!It's a pitty that there is few palaeo-art of them! Hope to see more!

By Georgios Georgalis (not verified) on 05 Feb 2008 #permalink

> Zach--

> Could it have been Gomphotherium (alias Trilophodon, Tetrabelodon)?


I would suggest a gomphothere, too - but in this case the tusks would be modified incisors rather than canines. BTW, I don't think gomphotheres were more aquatic than modern elephants.

Dave! Yes! Desmostylus!!!

*looks it up on Google Image Search*

Well, the picture I remember isn't there, but that's definately him. It's a SIRENIAN?!?

Pyrotheres ASAP!!!

What? Desmostylus is not a sirenian. Nobody knows what it is, but it's not a sirenian.

By David Marjanović (not verified) on 05 Feb 2008 #permalink

This post has reminded me of a critter from early Cenozoic South America. A sort of false horse. looked like a horse, but died out when grasslands appeared. Can't remember the dang name however.

I mention this animal because I've gotten to thinking about an alternate world where it did adapt to the new grasslands, and was subsequently discovered by and domesticated by the human tribes migrating into the Pampas from the Andes oh so long ago. Think of the possibilities. :)

Alan -- some kind of litoptern, maybe?

Right now, because of Walking With Prehistoric Beasts, the most common litoptern image you'll find on the Internet is of the "camel with a trunk" Macrauchenia type. But I seem to recall many sorts, and I know I've seen some reconstructed in horselike form.

By Stevo Darkly (not verified) on 05 Feb 2008 #permalink

The litoptern Thoatherium was decidedly horse-like. In fact, I've heard it being described as even more horsey than a horse, because while Equus retains the residual second and fourth toes as the splint bones, Thoatherium didn't possess a trace of any toes other than the third.

I'm not sure whether desmostylians are supposed to be more closely related to sirenians or proboscideans, but they seem to be universally accepted as closely related to those two orders.

This post has reminded me of a critter from early Cenozoic South America. A sort of false horse. looked like a horse, but died out when grasslands appeared. Can't remember the dang name however.

The litoptern Thoatherium was decidedly horse-like.


I suspect you're either thinking of T. or Diadiaphorus, both of them prototheriid litopterns.

Grasslands appeared in South America first, during the Eocene apparently. Didn't the prototheriids die out sometime around the Great American Interchange?

By David Marjanović (not verified) on 06 Feb 2008 #permalink

Apparently the proterotheriid Neolicaphrium was still around in the Lujanian (Late Pleistocene/Early Holocene). The side toes were apparently not as reduced as in Thoatherium. By the way, Thoatherium was much too small for anyone (even a hobbit, I guess) to ride on it.

By Lars Dietz (not verified) on 06 Feb 2008 #permalink

Apparently the proterotheriid Neolicaphrium was still around in the Lujanian (Late Pleistocene/Early Holocene).

!!! And I thought Macrauchenia was the last of the litopterns...

"!!! And I thought Macrauchenia was the last of the litopterns..."

It also came as a surprise to me. Macrauchenia is also known from the Lujanian, but I don't know which one of them is known from the youngest remains. The macraucheniid Windhausenia also was one of the last survivors, but I haven't been able to find out when it became extinct.

By Lars Dietz (not verified) on 07 Feb 2008 #permalink

My thanks to those who replied. It reinforces my understanding that people are the best search engine.

I saw my first litopterns in a recent article on ancient South American mammals in Scientific American. The subject was on an area with a wealth of fossil remains going back to early in the Cenozoic. Darren's post on Astropotheres reminded me of them, and together they inspired a temporal displacement scenario.

Namely... A midwest American town is transported in time and space to the northern Argentina of way long ago. The people adjust, adapt, and discover the local flora and fauna. Among the latter are litopterns, a few species very horse-like.

One species they then proceed to domesticate and there descendants subsequently use to settle the world.

Another scenario has a few speeches of pseudo horses surviving up to the arrival of the Indian on the Argentine pampas. At first the new arrivals hunt the animals for food, but then a 'horse cult' arises in which riding the animal becomes part of a ritual observation. To make matters easier and safer for the participants the animals are bred to be easier to handle; which means they become domesticated. (There are those who say this is how Central Eurasian tribes domesticated the horse.)

Thanks to their 'horse' these tribes are able to advance faster in the civilized arts than their neighbors (wagons and roads and stuff like that), and so are able to spread into the Andes and Amazonia (to a lesser degree). By the time the Europeans show up these 'horsemen' have spread their influence, and steeds, as far as North America, and thanks to happy accidents South Africa.

To make this scenario even more fun, when the first Spanish explorers arrive the false horses are spreading into tse tse fly territory, and they are immune to enchephalitis. European horses versus American 'horses', think of the possibilities. :)

Alan: your second alternate-history scenario doesn't require horse-sized surviving proterotheriids. There were real horses in South America when the Indians arrived. They had crossed the land bridge some time earlier. So you just need people to notice that horses could be domesticated before all of them got eaten (the horses, not the people).
Of course, we don't know whether the South American horses were easily domesticable.

By Lars Dietz (not verified) on 09 Feb 2008 #permalink

a few speeches of pseudo horses

Never use a spellchecker. I repeat: Never use a spellchecker.

and they are immune to enc[...]ephalitis.


By David Marjanović (not verified) on 09 Feb 2008 #permalink


Darn. Guess they'd be doing a lot of breeding for size. :)


I also wanted something that I could reasonably say was immune to encephalitis. The 'alienness' of 'lorns' helps.

Pundit: Horses whinny and nicker, lorns chuff and grunt. That's how you tell the difference.


You ougtha see the typoes when I don't use a spellchecker. :)

Must admit, E. Gary Gygax and The Epic of Ærth helped inspire a bit of it. Only his alternate Earth Atlantlans used magick on zebras to get their horse equivalents.

"I also wanted something that I could reasonably say was immune to encephalitis."
Would a macraucheniid work? Those that survived were about the size of camels, so they would have been large enough. They wouldn't have looked like horses, though. More like weird humpless camels with a short trunk.
I've encountered a claim that the creatures used by Edgar Rice Burroughs's Martians to ride on were based on Thoatherium. I haven't read the books, but according to wikipedia they were called "Thoats".

By Lars Dietz (not verified) on 10 Feb 2008 #permalink

CANINE tusks.
Proboscidean tusks are incisors (I1? I2? I can't remember).
Whenever I try to describe South American native ungulates to my less paleontologically-oriented friends, I fall back on calling them "fever dream animals"! Thanks, Darren! I think this post all by itself is as much on these facinating animals as any popular book I know contains!

By Allen Hazen (not verified) on 11 Feb 2008 #permalink

Antarctic big mammals? Does this mean the Elder Things ate astrapotheres?

By Weatherfac (not verified) on 12 Apr 2008 #permalink