The Question of Segnosaurs

From: "Tim Williams" <T.Williams@cclru.randwick.unsw.edu.au>

While on the subject of segnosaurians - WHAT ON EARTH ARE THEY? Since their discovery (what - ?twenty years ago) they've been regarded as (1) aberrant theropods, (2) relicts of a prosauropod-ornithischian transition, (3) saurischians close to sauropodomorphs. Now, as I understand it, segnosaurians are regarded as advanced coelurosaurians, with even a few bird-like characters found in the braincase. What's the consensus on the Segnosauria? (Is there one??) These critters fascinate me.

From: Dinogeorge@aol.com

Segnosaurian remains have proved to be fairly common in Cretaceous rocks of the eastern republics of the former USSR, according to Nessov's 1995 monograph on Russian dinosaur localities. The earlier remains represent rather small forms, the later remains include Plateosaurus-size segnosaurids and the even larger therizinosaurids. Nessov suggests that the smaller, earlier segnosaurs lived in trees eating wasp's nests (a form taxon he calls Desertiana; the wasp larvae would provide a protein-rich source of nourishment), and the larger, later segnosaurs on the ground. The analogy he pursues is tree sloth and megathere ground sloth. This is an elaboration of a hypothesis put forward around 1972 by Rozhdestvensky, before segnosaurs were recognized as a distinct dinosaur group, to account for the large-clawed, large-forelimbed therizinosaurids and deinocheirids being discovered in eastern Asia: The large claws were for holding the creatures upside-down, hanging sloth-like from trees. Most paleontologists have ignored Rozhdestvensky's paper, or have cited it as some kind of fringe paleontology.

From: "Tim Williams" <T.Williams@cclru.randwick.unsw.edu.au>

I'm not surprised - the idea still sounds a bit loopy. I didn't think there were too many trees in the Gobi Desert, even back in the Cretaceous. And any tree capable of supporting a fully-grown Deinocheirus or Therizinosaurus must have been HUGE!! But maybe Nessov is onto something when he suggests that the earlier segnosaurians were insectivores. (Didn't someone else once suggest that Therizinosaurus used its huge arms and claws for ripping open giant termite nests?). I'd love to see Nessov's work.

Has anyone found a skull for Therizinosaurus (or for Alxasaurus) yet?

What about the phylogeny of the Segnosauria? Have the segnosaurians, like the tyrannosaurians, now been classified in the Coelurosauria?

From: egood@ALPHA2.CURTIN.EDU.AU

I find it hard to imagine dinosaurs as large as the segnosaurs existing on insects. Unless of course they were as big as Meganeura was! Most insectivorous creatures today are small, with the exceptions of the aardvark, and the giant anteater.

I find the notion of the Coelurosauria a bit dated nowadays, since reading Paul's PDW. personally think they were a group somewhere near the ancestral ornitho/sauris split. Most people that think of the predatory dinosaurs as big carnosaurs, or small coelurosaurs tend to overlook creatures like Nanotyrannus, and creatures like Dilophosaurus (more akin to Coelophysis).

From: "Tim Williams" <T.Williams@cclru.randwick.unsw.edu.au>

>I find it hard to imagine dinosaurs as large as the segnosaurs
> existing on insects.

Yep, so do I. Unless Cretaceous termites came extra-large, I don't think a hungry Therizinosaurus could have gone too far on a diet consisting solely of white-ants.

>I find the notion of the Coelurosauria a bit dated nowadays,
> since reading Paul's PDW.

A revised Coelurosauria proposed by Tom Holtz is similar to Greg Paul's Avetheropoda - but it includes both large and small theropods. It includes most of the "traditional" coelurosaurs, as well as segnosaurs and tyrannosaurs.

Segnosaurs were at first lumped among the carnosaurs, but this was an unashamedly provisional classification. But I thought Paul's view of the Segnosauria as ornithischian-sauropodomorph transitional forms was also a bit dated. As I understand it, the inclusion of segnosaurs in the Coelurosauria was prompted (at least partly) by the discovery of a good skull from Erlikosaurus, which shows several coelurosaur-like features.

From: Dinogeorge@aol.com

>I find it hard to imagine dinosaurs as large as the segnosaurs
>existing on insects. Unless of course they were as big as Meganeura was!
>Most insectivorous creatures today are small, with the exceptions of the
>aardvark, and the giant anteater.

Nessov only had the smaller segnosaurs (known only from fragmentary material from the Lower Cretaceous and, perhaps, footprints from the Late Jurassic; none named) up in the trees eating wasp's nests. The larger forms (which are known from better material, and include Alxasaurus, Segnosaurus, Erlikosaurus, Enigmosaurus, Nanshiungosaurus, and Therizinosaurus), he said, were terrestrial rather than arboreal--like megatheres vis a vis tree sloths.

From: Stan Friesen <swf@ElSegundoCA.ATTGIS.COM>

It is interesting to note that these, the largest modern insectivores, are slow-moving, and specialize in insect that can be found in large concentrations (eusocial ones, that is).

Of course the suggestion that the segnosaurs may have been insectivores comes from the general similarity of the Therizinosaurus claws to those of anteaters - which it uses to open termite mounds and ant hills.

Still, an anteater the size of Therizinosaurus rather boggles the mind. Those ants must sure have been large :-)

From: Stang1996@aol.com

>with even a few bird-like characters found in the braincase.

What? You know of a study on segnosaur brains? Last I heard, no one had done anything on segnosaur brains. I hope you are not confusing this with my suggestion that segnosaur brains be studied to see if the V1 nerve is in the place that it is in birds and arctometatarsalians, that would prove that they were indeed bullatosaurs.

>These critters fascinate me.

They fascinate everyone, but clearly, there are too many features that segnosaurs possess that are also possessed by prosauropods, and too many features they possess that are also possessed by coelurians to really make any definite conclusions, though I side for the prosauropod end myself. I am serious, I tried to construct a cladogram using all of Paul's; Russel & Dong's; and some of my own characters, but I found it a waste of time because either way it goes there are about 50 synapomorphies and 50 convergent characters.

From: "Tim Williams" <T.Williams@cclru.randwick.unsw.edu.au>

> What? You know of a study on segnosaur brains? Last I heard, no one had
> done anything on segnosaur brains.

The article I read was an abstract in JVP, which said that there were similarities regarding certain elements of the braincase between segnosaurs and coelurosaurs (including birds). I can't remember the details though. I'm pretty sure it didn't mention anything about the brain itself.

> They fascinate everyone, but clearly, there are too many features that
> segnosaurs possess that are also possessed by prosauropods, and too many
> features they possess that are also possessed by coelurians to really make
> any definite conclusions, though I side for the prosauropod end myself.

To my mind, the segnosaurian foot is one sticking point for theropod affinities. In all theropods (including Eoraptor) the hallux is reversed and assumed to be nonfunctional, producing a functionally tridactyl foot. Segnosaurs have no reversed hallux and have a functionally tetradactyl foot, which (unless this feature is a reversal) makes the foot more primitive than even Eoraptor and the herrerasaurids.

However, Erlikosaurus is said to have an intramandibular "hinge" in the lower jaw, like all theropods (including herrerasaurids, but not Eoraptor according to Paul Sereno). But, despite the cranial evidence, I also lean towards sauropodomorph affinities for the Segnosauria.

From: ornstn@inforamp.net (Ronald Orenstein)

>Yep, so do I. Unless Cretaceous termites came extra-large, I don't
>think a hungry Therizinosaurus could have gone too far on a diet
>consisting solely of white-ants.

Surely it is the number of termites ingested rather than the size of each individual termite that matters! Anyway, if segnosaurs were termite-eaters their hyoid apparatus (is it known?) might show evidence of an elongate tongue.

The largest living termite-eaters that I know of are the giant anteater (18-39 kg) which has been reported to eat 30,000 termites (and other insects) a day; the giant armadillo (up to 60 kg, at least in captivity) whose primary food is termites; and the aardvark (40-100 kg, usually 50-70 kg). How does this compare with segnosaur sizes?

More to the point - were mound-building termites coeval with segnosaurs?

From: "Tim Williams" <T.Williams@cclru.randwick.unsw.edu.au>

> Surely it is the number of termites ingested rather than the size of each
> individual termite that matters! Anyway, if segnosaurs were termite-eaters
> their hyoid apparatus (is it known?) might show evidence of an elongate
>tongue.

I guess I was being a bit flippant when I made that comment. A good therizinosaurid skull would help resolve the issue of their eating habits. Which brings me back to a previous question - is any skull material known for Therizinosaurus or Alxasaurus?

I don't know if mound-building termites existed in the Mesozoic (are there any palaeoentomologists out there? :-)). In my opinion, ant-eating therizinosaurids are certainly feasible. But I like the chalicothere analogy better - that therizinosaurids used their claws to hook branches down to within reach of the mouth.

From: Thomas_R_HOLTZ@umail.umd.edu (th81)

The former's skull is unknown, only the dentary is known of Alxasaurus. Of course, Erlicosaurus's skull is the best preserved of any Mesozoic theropod!! The teeth of Erlicosaurus are "troodontiform" (pinched based and big denticles), which are morphometrically distinct from serration densities associated with vertebrate hypercarnivores (some sharks, basal synapsids, sabrecats of all stripes, varanoid lizards, pseudosuchians (including rauisuchids and xiphodont crocs), and most theropods). This may indicate a non-hypercarnivorous diet (i.e., a diet which includes things other than vertebrate flesh, such as arthropods, molluscs, fungi, plants, etc.)

From: Stan Friesen <swf@ElSegundoCA.ATTGIS.COM>

> The largest living termite-eaters that I know of are the giant anteater
> (18-39 kg) which has been reported to eat 30,000 termites (and other
> insects) a day; the giant armadillo (up to 60 kg, at least in captivity)
> whose primary food is termites; and the aardvark (40-100 kg, usually 50-70
> kg). How does this compare with segnosaur sizes?

Small. Segnosaurs were in the thousand kg and up size range.

Scaling up from a giant anteater this comes to hundreds of thousands of termites per day.

From: "Darren Naish" <dwn194@soton.ac.uk>

Someone asked about the fossil record of termites. While I believe this to be pretty much irrelevant to the evolution of segnosaurs, termites were certainly around in the Cretaceous: this has been known for a while but a partial nest reported from Triassic strata last year showed that some were essentially modern in habits and architecture. If this was the state of play in the Triassic, I would expect Cretaceous tropical parkland environs to sport large termitaria of an essentially modern appearance. Bees and wasps are well known from the Cretaceous at least - many of these bugs would be making papery tree-nests similar or very similar to those we can see today.

From: "Darren Naish" <dwn194@soton.ac.uk>

Looks like we're back in familiar territory people - not long before someone launches into an 'Alxasaur? Oviratorosaur? Since when?' thread (guilty party?).

With regard to speculation on the diet of these dinosaurs, ascribing an insectivorous diet when the animals have edentate-type claws is probably missing the point when edentates that hang around trees (the sloths - NPI) are herbivorous. This is entirely compatible, however, with Nessov's notions of initial insectivory (and, dare I say, fits in nicely with a certain hypothesis of dinosaurian evolution [one that ain't 'alf badd]). It is also interesting that this hypothetical situation represents the reverse of what is supposed to have happened in pangolin evolution - they also have recurved claws for climbing (and, of course, were also figured by Rozhdestvensky).

One question that arises, however, is 'Are segnosaur claws like those of megatheres?'. Though I've seen a fair few megathere claws, I'm none too sure here, but my recollection is that they are far more hook-like than those of Segnosaurus et al. Though this certainly doesn't rule out the analogy, it has opened different interpretations for others. Uncle Bob, for example, insists that segnosaurs have ideal digging claws - long and straight, like those of a badger, he says. Somehow I just can't imagine a 2 ton dinosaur merrily tunneling its way beneath the Cretaceous topsoil.

From: Dinogeorge@aol.com

Except for their relative size, segnosaur claws aren't very similar to edentate claws. For example, I don't think segnosaurs walked on all fours with their manual claws turned under. Also, the manual claws for Alxasaurus and Therizinosaurus, the only two taxa for which the manual claws are known, are quite different in proportions, curvature, and absolute size. One hallmark is that they are strongly compressed transversely (both manual and pedal claws, actually), which is one way Nessov was able to identify them in the fossil record.

I can't be certain, but the longest and strongest manual digit seems to be the first (the "brontosaur" pattern), even though the corresponding metacarpal is shortest. This seems to be the case with the type specimen of Alxasaurus--though it was not restored that way in the mounted cast in the Greatest Show Unearthed traveling dinosaur exhibition. Alxasaurus is the only segnosaur known with a more or less complete manus.

Burrowing seems out of the question, not only because of the sheer size of the animals but because transversely compressed claws would be rather useless for digging with--like trying to dig a hole with a scythe. Claws designed for digging would probably be shorter and thicker, so as to get some shoveling action and leverage. But I can see segnosaur claws as useful for ripping open termite and ant colonies (or marauding tarbosaurinids).

From: Thomas_R_HOLTZ@umail.umd.edu (th81)

>It is interesting to note that these, the largest modern insectivores,
>are slow-moving, and specialize in insect that can be found in large
>concentrations (eusocial ones, that is).

Not to say that I support the idea of an arboreal ancestor of the therizinosauroids (aka segnosaurians), but it IS interesting that, of all mammals, ground sloths seem to be the best analogue for these strange dinosaurs. And ground sloths, after all, were simply the giant, herbivorous, secondarily-terrestrial descendants of small, once-insectivorous, then-herbivorous, arboreal forms...

(The two "D. Russells", Don E. and Dale A., consider therizinosaurs to be chalicothere analogues.)

From: Dinogeorge@aol.com

Heh, heh! You quick, Kemo Sabe...

Arboreality could explain why segnosaurians are essentially absent from the fossil record prior to the evolution of large, ground-dwelling forms in the Early and Late Cretaceous. Unfortunately, there are also dozens of other reasons. Footprint experts: Are trackways known in which the hind-foot impressions are tetradactyl with long, very narrow unguals? These would be good segnosaurian trackway candidates. Nessov (1995) noted "segnosaur" trackways as early as Late Jurassic in central Asia, but he didn't figure them, but I haven't as yet had a chance to really dive into the paper and read about them in detail.