Takins, and another what-the-hell-is-it

The small arboreal squirrel gorgonopsian *Sciurops*.

This picture looks familiar! If I remember correctly, it's from a paper that came out maybe a decade or so ago, published in either Nature or Science, although I can't at the moment remember the exact reference or names of the authors. It proposed the rather (to say the least!) controversial notion that mammals and birds are each other's closest relatives within the amniotes,based on shared endothermy, as opposed to the standard model that sees birds as archosaurs, grouped with crocs and dinosaurs, and mammals as synapsid descendants. The picture claims to show what a VERY hypothetical common Triassic ancestor of birds/mammals might have looked like, with some kind of intermediate fur/feather insulation, incipient birdy wing-flaps, mammalian whiskers, Archaeopteryx-like tail etc. As I remember the idea caused a bit of a flap (no pun intended!) at the time, then sank without trace - presumably because it's total bollocks and would require us to discard a mountain of fossil, molecular and morphological evidence that underlies the 'traditional' version of amniote relationships.

I'll try and find the full ref. so that we can all check out this wacky idea for ourselves.

Dammit dave, you're spoiling it for everyone, can't you just say it's a Venusian squirolizard or something and be done with it. Me, I think it's a squabrat from the Jim Hensen movie The Dark Crystal.

Whatever it is, I bet it would make a great pet!

Hmmm - before finding out, I was going to guess that it was something cryptozoological. It looks like an arborial viverrid or mustellid to me.

Namesake Hughes is almost certainly right.

Note the animal's BANDit features: impressive sprawling capabilities, a reduced thumb next to three long fingers, and feet that are completely ridiculous for a climbing animal that doesn't have very recent cursorial ancestors.

Well now, this is clearly a proto-ropen. Which was a hypothetical dimly-bioluminescent ptero-bat-sloth-bird of three and a half thousand years ago that microevolved into todays ropen.

Just another not-iguanian.
Next?

Still haven't managed to dig up this damn paper. Maybe I just imagined it....?
Birds and mammals as sister-groups - would anyone really suggest that???

SINOSAUROPTERYX????????

> Birds and mammals as sister-groups - would anyone really suggest
> that???

This has been suggested in the late eighties and early nineties, on the basis of the strong - but superficial - similarity between some pelycosaurs and erythrosuchians. I remember that I have seen cladograms that lumped birds and mammals together in a clade "Homeotherma" or "Homoiotherma", but I don't remember the author. Of course, its bollocks - all the claimed synapomorphic traits shared between sphenacodonts and erythrosuchians are either plesiomorphic features primitive for all amniotes, or the result of convergent evolution (crocodiles don't fit into the picture, either).

That looks like the climbing duck-possum of Brazil;)

Darren, we don't believe it was your beloved pet.

John,

I believe your conclusion of a proto-ropen is wrong. This is clearly the shrew-like ancestor of Rods that we talked about in Old Street.

> Birds and mammals as sister-groups - would anyone really suggest
> that???

Obvious. Every dumb publication makes few citations. Why do you think, when somebody discovers perfect fossil hominid, another group jumps up saying it was malformed modern human despite every child seeing it's totally unlike human skull? Every sh*t generates scientific buzz.

I would have said an unsual reconstruction of a climbing dinosaur or a proto-bat/bat ancestor

Jerzy; I never thought of it that way, but I'm sure you're right. I must bear that principle in mind next time I write a paper. Have to boost the citation rate somehow!

Brian Gardiner has argued that birds and mammals are one another's closest living relatives, making synapsids the descendants of primitive archosaurs. Dinosaurs and pterosaurs are therefore closer to mammals than they are to crocodiles. It seems a bit odd to say the least, but if it is wrong it is wrong in an interesting manner.

I hope you don't, or may flock of ropen descend upon you ;-)

The Darren's ex pet might be:
- early example of pterosaur clade,
- early gliding marsupial,
- some recently discovered mesosoic mammal remarkable mostly for its discoverers.

Or illustrator might:
- got vibrissae wrong,
- got hair wrong,
- got number of fingers and toes wrong,

An archaeoptoryx reconstruction from the 1900s?

SINOSAUROPTERYX????????

Not at all. The arms and hands are completely different, the tail is too small, and S. was completely unable to sprawl.

on the basis of the strong - but superficial - similarity between some pelycosaurs and erythrosuchians.

No, here you're talking about Reig's idea from the 1960s that archosaurs are theropsids rather than sauropsids. I've read Romer's paper shredding this hypothesis to very fine pieces. <sound of name dropping on floor>

The idea that mammals and birds are sister-groups was thought up by neontologists (Løvtrup, Gardiner) in the late 70s through early 90s. They were among the first cladists in the field. Unfortunately, they 1) believed that fossils should not be used for phylogeny reconstruction at all because they're so incomplete (missing data, shock horror), 2) honestly believed in the results of their teensy-weensy matrices, and 3) made coding mistakes.

If you ignore all fossils, mammals and birds do indeed share lot of anatomical details, and the same often comes out of primitive phenetic molecular analyses. Løvtrup and Gardiner called this clade Haemothermia and later corrected it to Haematothermia, which is the name Sir Richard Owen had given to it when the relevant fossils were only being discovered.

In his last paper (1993), Gardiner tried to insert the fossils into the tree he got, but didn't use cladistics for this.

An archaeoptoryx reconstruction from the 1900s?

The wings are well preserved in the London specimen and were AFAIK never denied.

Its a ferret specialy bred to catch squirrels.

Oh. This is the third most active ScienceBlogs post at the moment. I'll stop making such reports -- Tet Zoo now clearly has so many regular readers that we'll see it contributing to the top 5 for evermore.

I remember a paper from the early 90s called something like 'Is a chicken more like a dog or a lizard?'. I can't remember the author. It took the hypothesis that birds and mammals were sister groups - the homeothermia - and then pointed out why it couldn't be true. What I particularly remember was the argument that the common ancestor would have had to have all the purported synapomorphies of the two groups, but none of the derived uniting characteristics of each group. For example, it would be homeothermic, but couldn't have insulation from either hair or feathers, so would have died of cold. I think there was an illustration of the resulting ridiculous creature that looked a bit like the one shown here. ALthough of course this one has fur, so probably I am wrong.

Isn't this an imaginary creature taken from a book that imagined what dinosaurs would look like today if they hadn't disappeared and had kept on evolving? Can't remember the title, though. Must be some 20 years ago.

Hypothetical bat ancestor? Having seen the recent "old bat" I imagine something like this could precede it.

Gregory Paul mentions the Haem(at)othermia concept in Dinosaurs of the Air, and cites an illustration that sounds a lot like this one from the description. He gives the reference:
Janvier, P. (1983): La divorce de l'oiseau et du crocodile. La Recherche 14:1430-1432.
Paul admits that at least Haematothermia is not as bad as the idea that birds evolved from flying fish, which has also been seriously proposed.

Christophe, you're thinking of The New Dinosaurs: An Alternate Zoology by Dougal Dixon, one of my favorite books as a teenager. What Darren posted does indeed look like one of the "arbrosaurs" in that book, and the dinosaurs in that book were rather mammal-like, but it's not an illustration from that book.

Not that it has four finger, so it can't possibly be a tetanuran (unless it's an extremely basal one). The proto-haematotherm idea sounds plausible, more plausible than my guess (basal pterosauromorph).

It's still not as bad as the hypothesis that Iguanodon evolved flippers and flew away.

Another wonderful book by Dougal Dixon was After Man, outlining the zoology 50 million years hence.

The beetle is saying "Do you want me to talk?", and the squabrat is saying "No, I want you to die!"

ARGH! I didn't even notice the first toe is in the place of the fifth!!!

Hypothetical bat ancestor? Having seen the recent "old bat" I imagine something like this could precede it.

No, not with those hands or those feet or the complete lack of external ears. It must be Janvier's proto-haematotherm.

(BTW, that means "blood-warm" -- "warm like blood" -- rather than the presumably intended "warm-blooded".)

That's a hypothetical pterosaur ancestor. I saw it in Unwin's "Pterosaurs: From Deep Time." I was reproduced from an article in some magazine.

Of course, my first guess is always "It's a deceptive image of an ordinary house cat." But even I can see that this mystery animal is clutching an insect in a grasping hand. So I'm going to have to guess that it's a deceptive image of an ordinary house cat with polydactyly. And since the bug in question looks remarkably like a palmetto bug, I'm going to go way out a limb here and suggest that it's a deceptive image of one the famous Hemingway cats in Key West, Florida.

Looks a bit like a colugo to me.

A tree shrew perhaps?

Clearly that is a climbing hellasaur, direct ancestor of the better-known flying hellasaur. The well-known tendency of the flying hellasaur to "gank" cavegirls in fur bikinis is an exaptation of the beetle-ganking behavior of the climbing hellasaur shown here. The picture above is figure 8b in Romer, 1949.

Refs:

Romer, A.S. 1949. The evolution of ganking behavior in the Hellasauria. Quarterly Bulletin of the South Bavarian Ichthyosaurological Society, Supplement 7, III(2):1-42.

The sketch is not the cover art from "La Recherche" (the 1982 or 1983, I think, issue that had the story on Haemothermia had a lurid red and blue painting, in a different pose), but seems to be the same imaginary animal.

"Haemothermia" was defended by Gardiner in ZJLS 74 (1982) and Cladistics 9 (1993). The argument did depend on excluding fossil taxa as a matter of principle, and then counting synapomorphies in the living taxa. The methodology was called "pattern cladistics," but maybe came closer to numerical taxonomy in practice: recall that this was a LONG time ago, before the use of computers to construct phylogenetic trees!

The idea was criticized by Gauthier et al in Cladistics 4 (1988), who found the (sensible) phylogeny associating extant mammals with earlier synapsids to the exclusion of birds. What was interesting in their critique was the lesson that excluding fossils was a BAD idea: when they redid their analysis restricting attention to extant taxa, they recovered Haemothermia! (Or: fossils help you tell the difference between convergent evolution and common ancestry.)

I'm going to pretend that I didn't read any of the comments above, and go with my original hypotheses:

1) A very inaccurate drawing of a colugo ("flying lemur"), by an early naturalist, who only had a badly decomposed corpse to go on. (I'm assuming the gliding membrane was rotted away, and the tail too, so our naturalist was ignorant of the former and assumed the latter was longer than it was.)

2) A bad drawing of a tayra. I don't know why this popped into my head. It just did.

I'm pretty sure David Hughes is right, believe it or not. I seem to remember seeing a report with this pic in some newspaper and wondering how it got published for long enough for it to register in my long-term memory.

The methodology was called "pattern cladistics,"

No, pattern cladistics was the idea that evolution is not a necessary assumption for doing cladistics -- that cladistics could be done as if it were phenetics. Has died out, too, though.

The idea that fossils are second-class OTUs, however, goes all the way back to Hennig, who was -- surprisingly -- naïve enough to simply assert that phylogenetics starts with the extant taxa and then inserts the fossils into that tree as a second step. (Not "must start" or something, but just "starts", as in "this is how it's done (the right way)".

when they redid their analysis restricting attention to extant taxa, they recovered Haemothermia!

Not even: they recovered (Mammalia (Crocodylia, Aves)).

The paper by Gauthier et al. (1988) is also noteworthy for being among the very first cladistic analyses that used a data matrix of a halfway reasonable size. It was probably around the limits of computer power at the time...

Gardiner's "reply" from 1993 is really pathetic. It shouldn't have passed peer-review.

Hate to break this to you Matt but you're clearly misreading Romer, 1949. The origin of the "gank-stroke" cannot be disentangled from cavegirl capture behavior. Insect-ganking in the haematotherms is fundamentally a manus-closing phenomenon and is certainly not a viable exaptive route to cavegirl-ganking. I have some preliminary FEA models that prove this which I will be presenting at Cal Paleo next month.

Well, I hate to break this to you, neil, but I've conclusively proven the opposite with my new method of one-dimensional geometric morphometrics, which I call Extremely Thin-Plate Spline Analysis. It's all in my in-progress book, Marcel Delgado's Lost World: Toward a Transformative Hermeneutics of Quantum Dinosaurs. I will also present it next month, when I'll be putting the Cal in Cal Paleo. What's up now?

David Marjanovic--
Sorry. I have only hazy recolletions of the Gauthier et al. paper. I ***thought*** I remembered that they had found that the inclusion of fossil taxa changed the recovered tree, and since (Mammalia(Crocodylia,Aves)) is what you get when you DO include fossils... I should do more homework before posting.
Darren--
If your French is better than your Spanish, I suspect the Janvier article in "La Recherche" is the original of the Spanish article you cite. (But isn't Amniote phylogeny a bit modern for Janvier? I mean, I associate him with stem Chordates and what, if people will permit me to use a word for a group of organisms that doesn't denote a clade, I would like to call fossil "fish.")

(Mammalia(Crocodylia,Aves)) is what you get when you DO include fossils...

Yes, but I meant that, without fossils, Gauthier et al. found the lizards and turtles outside that clade; with fossils included, they found the traditional (Mammalia (Testudinata (Lepidosauria (Crocodylia, Aves)))).

what, if people will permit me to use a word for a group of organisms that doesn't denote a clade, I would like to call fossil "fish."

Cladists, of course, use terms for non-clades all the time, like "non-avian dinosaurs". The term you're after is "non-tetrapodan fish".

;)

David Marjanovic--
Thanks for the clarification. I obviously hadn't remembered the Gauthier et al. article very clearly. Their actual finding (that fossils allow the recovery of "Reptilia" which is lost when the analysis is restricted to extant taxa) is, of course, as good as the one I misremembered for making the methodological point.

(I like it because, as a philosopher, I don't like it when people try to make an a priori decision about what is and is not empirical: exclude or include fossils in your analysis is a methodological policy decision-- the sort of thing some 1930s positivists and Wittgensteinians would have said was purely a priori-- and Gauthier et al showed that empirical results were relevant to the question of whether or not it is a GOOD policy. Another thing that I think some 1980s cladists would have claimed was a purely methodological question is whether or not you should try to identify likely convergences before doing a phylogenetic analysis or whether it is more "objective" to avoid conjectures about functional complexes: one of my favorite papers is one by Christine Janis and someone, in a Mammal phylogeny collection edited Szalay, Novacek et al, demonstrating that the decision does serious work in the case of the phyl. of ruminant artiodactyls.)

Thanks for clarification on such things as toes David M. The neck was too long anyway.

Use fossils too!

The Hemingway cat thing is hilarious if obscure.

Arboreal coastal Columbian Colugodile?

it looks like my brother, only not as fat, and with more hands.