Fujianvenator, Jurassic Avialan

[Sight Reader]

Hey, did do you guys already talk about Fujianvenator? I might have missed the discussion…

https://www.nature.com/articles/s41586-023-06513-7

[erik simpson]

On Sunday, November 5, 2023 at 8:08:57 AM UTC-8, Sight Reader wrote:
> Hey, did do you guys already talk about Fujianvenator? I might have missed the discussion…
>
> https://www.nature.com/articles/s41586-023-06513-7

Looks interesting, but it's paywalled. I'll see if I can find a way to get it.

[Sight Reader]

Oh dang. There are news reports and some preliminary in Wikipedia…
https://www.reuters.com/science/bizarre-long-legged-bird-like-dinosaur-has-scientists-enthralled-2023-09-06/

[erik simpson]

As Wang Min points out, whether this is a "bird" or not is a subjective call. It may not have been flight-capable,
but is clearly in the clade that lead to birds as we know them.

[Peter Nyikos]

Did you succeed? Thanks to my university"s subscription, I saw it right away.

Occasionally, figures are not paywalled. See whether you can see this:

https://www.nature.com/articles/s41586-023-06513-7/figures/1

It includes a fairly detailed phylogenetic tree.

If you can't see it, the following may convey enough information about the key features.

Discussion
Our phylogenetic analyses consistently recovered Fujianvenator within the clade Anchiornithidae, which unites other Anchiornis-like taxa51 and is resolved as the earliest diverging group of the Avialae (Fig. 1b). Archaeopteryx was resolved within the Avialae as the sister to other avialans except the anchiornithids. This result is robust to maximum parsimony and Bayesian analyses using tip-dating methods (Extended Data Figs. 5 and 6). Given that it is at present the southernmost record of Jurassic avialans, and is 10 million years younger than the Yanliao taxa (Fig. 1b), Fujianvenator contributes a great deal of spatio-temporal information about early avialan diversification close to the end of the Jurassic. Fujianvenator is largely comparable with Archaeopteryx in having similar manual phalangeal proportions. On the other hand, its pelvis shows features that are used to diagnose Anchiornis and troodontids (Extended Data Fig. 3): the short ischium is Anchiornis-like, bearing a distally located obturator process that is constricted at its base7,20, whereas the pubes resemble these of troodontids in having a mediolaterally broad, proximodistally elongated imperforated apron39. Notably, the hindlimb of Fujianvenator exhibits mixed morphologies in its finer aspects, including non-arctometatarsal feet like those in Archaeopteryx and Anchiornis9,14; a metatarsal II that is mediolaterally broader than metatarsals III and IV, as in Archaeopteryx31; a non-ginglymoid metatarsal II, as in troodontids47,48; a metatarsal II trochlea that is wider than that of other metatarsals, as in Anchiornis and some dromaeosaurids14,49; and a ginglymoid metatarsal III, as in dromaeosaurids and some troodontids33 (Fig. 2g and Extended Data Figs. 2d and 3). The unique combination of postcranial features shared with early paravians that is preserved in Fujianvenator shows how deeply the avialan phylogeny has been affected by evolutionary mosaicism. That, in turn, partly explains the controversy about the interrelationships among these early-diverging paravians1,2,7,52.

The last two sentences remind me of the time, over two decades ago, when I did a detailed
study of five factors in the blood clotting cascade. All of them are judged to be
closely related to each other. However, instead of a clear
phylogenetic tree of them, I saw a great deal of mosaicism in which
one factor matched another factor for several amino acids, only to switch
its "allegiance" to another factor.

When I emailed one of the authors about this, he replied that my kind of
analysis was more detailed and painstaking than the usual kind, and he
wished that other researchers would exhibit similar patience for details.


Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer--
University of South Carolina
https://people.math.sc.edu/nyikos

[erik simpson]

Yes, I've seen it. My son is a paleobotanist, and has access to Nature.

[Peter Nyikos]

As with most popularizations, one can expect false generalizations. Here is one that immediately
aroused my suspicions:

"Fujianvenator's lower leg bone - the tibia - was twice as long as its thigh bone - the femur. Such dimensions are unique among theropods, a group that includes all the meat-eating dinosaurs such as Tyrannosaurus and various others."

A quick look at Carroll's _Vertebrate Paleontology and Evolution_ confirmed my suspicions.
On page 304, 14-26, the tibia is twice as long as the femur in *Fabrosaurus*, "one of
the most primitive ornithischians." Two pages later, the same is seen to be true
of another primitive ornithischian *Heterodontosaurus*. Both are from the Lower Jurassic.

Lots of later ornithischians had the same dimensions, including the well-known
*Stegosaurus* [p. 313] and the early "ceratopsian" *Psittacosaurus* [p. 310], which had
neither horns nor frill, but it did have the characteristic ceratopsian beak.


A completely different comment from the linked Reuters report flies in the face of
the excuse Harshman gave for feathers being omitted from most phylogenetic analyses
that centered on bird evolution: too few have them to make it worthwhile to include them.
But here we see a contrary statement from the Reuters report:

""The fossil itself does not preserve feathers. However, its closest relatives and nearly all the known avialan theropods have feathers, and feathers are widely distributed among dinosaurs."

Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer--

Univ. of South Carolina at Columbia
http://people.math.sc.edu/nyikos

[John Harshman]

On 11/16/23 2:28 PM, Peter Nyikos wrote:

Let me allow you to reply to yourself. With a bit of snipping, it works
out well.

> As with most popularizations, one can expect false generalizations. Here is one that immediately
> aroused my suspicions:

[Sight Reader]

On Thursday, November 16, 2023 at 2:54:33 PM UTC-7, Peter Nyikos wrote:
> Our phylogenetic analyses consistently recovered Fujianvenator within the clade Anchiornithidae, which unites other Anchiornis-like taxa51 and is resolved as the earliest diverging group of the Avialae (Fig. 1b). Archaeopteryx was resolved within the Avialae as the sister to other avialans except the anchiornithids.

Ok, so I’m trying to wrap my puny little mind around this. Are they saying here that Anchiornithidae (which, I gather, contains Fujianvenator) and whatever group Archeopteryx belongs to are both different “spurs” of the Avialae tree that are sister to birds - but neither of lead directly TO birds?

[erik simpson]

Partly right. Anchiornithids are within Avialae, as is Archeopterix . Avialae has lots of branches, including crown birds.
All the others are extinct.

[John Harshman]

Phylogenetic trees never ascribe ancestry to any taxa. Nothing leads
directly to anything else. And this is because there can be no evidence
for direct ancestry, only that one species is more closely related to
one other species than to some third species. What this tree shows is
that Archaeopteryx is more closely related to birds than to
anchiornithids. How would you distinguish an ancestor from a cousin?

[Sight Reader]

Ok, that’s what I thought: all 3 (birds, anchios, archeo) are on different branches within the Avialae tree, all 3 appear to be sisters of each other, and none of them appear to be ancestors of each other. Sorry, this kid was talking to me as I was hurriedly trying to submit that post and I don’t think I got it to say what I meant it to say…

[erik simpson]

Bear in mind John's reminder; unless you're talking about VERY recent relations, ancestry is impossible
to establish, particularly over millions (or tens of millions) of years.

[John Harshman]

Actually, ancoirnithids are sister to Archaeopteryx + the remaining
avialans, and Archaeopteryx is sister to the remaining avialans. In
phylogenetics, a given group can have only one sister.

[Popping Mad]

On 11/17/23 16:19, Sight Reader wrote:
> all 3 appear to be sisters

brothers.

[John Harshman]

Siblings?

[Sight Reader]

Oh! I didn’t know there was a difference… Do you think they’re mad at me?

[Popping Mad]

On 11/20/23 20:24, Sight Reader wrote:
> Do you think they’re mad at me?

Did you ever see Hitckocks "The Birds"

[Sight Reader]

Oh crap. Now I’m not going to be able to sleep.

Speaking of crap, all this reminds of some ancient wisdom: “some days you’re the pigeon, some days you’re the statue…”

[Peter Nyikos]

Correct. Paleontologists have abandoned Archie as a bird ancestor. But I have no idea
how much of this is due to careful anatomical study and how much due to an ideology
that dominates taxonomy, which claims that there is "no evidence" that any fossil species is ancestral
to any other species, fossil or extant.

"no evidence" = no incontrovertible proof

Meanwhile, loose "sister group" talk is everywhere, as can be seen from the
on-again, off-again hypothesis that the correct grouping is
{Theropods, Ornithischians} Sauropods
rather than the > century old tradition that the following is correct:

{Theropods, Sauropods} Ornithischians.

Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer--

Univ. of South Carolina in Columbia
https://people.math.sc.edu/nyikos

[John Harshman]

On 11/21/23 6:55 AM, Peter Nyikos wrote:
> On Friday, November 17, 2023 at 9:45:04 AM UTC-5, Sight Reader wrote:
>> On Thursday, November 16, 2023 at 2:54:33 PM UTC-7, Peter Nyikos wrote:
>
>>> Our phylogenetic analyses consistently recovered Fujianvenator within the clade Anchiornithidae, which unites other Anchiornis-like taxa51 and is resolved as the earliest diverging group of the Avialae (Fig. 1b). Archaeopteryx was resolved within the Avialae as the sister to other avialans except the anchiornithids.
>
>> Ok, so I’m trying to wrap my puny little mind around this. Are they saying here that Anchiornithidae (which, I gather, contains Fujianvenator) and whatever group Archeopteryx belongs to are both different “spurs” of the Avialae tree that are sister to birds - but neither of lead directly TO birds?
>
> Correct. Paleontologists have abandoned Archie as a bird ancestor. But I have no idea
> how much of this is due to careful anatomical study and how much due to an ideology
> that dominates taxonomy, which claims that there is "no evidence" that any fossil species is ancestral
> to any other species, fossil or extant.

A little of both, really.

> "no evidence" = no incontrovertible proof

What would count as evidence? If I recall, you mostly won't say. I would
suggest that the best evidence would be character optimization on a tree
showing a zero-length branch between a species and the ancestral node.
But would even that be ver good evidence?

> Meanwhile, loose "sister group" talk is everywhere, as can be seen from the
> on-again, off-again hypothesis that the correct grouping is
> {Theropods, Ornithischians} Sauropods
> rather than the > century old tradition that the following is correct:
>
> {Theropods, Sauropods} Ornithischians.

I'm not quite sure if you think that's a bad thing. But if that's the
case, why? Shouldn't we always be prepared to question traditional
wisdom in the light of new evidence? Isn't uncertainty preferable to
false confidence?

[Peter Nyikos]

On Tuesday, November 21, 2023 at 11:15:47 AM UTC-5, John Harshman wrote:
> On 11/21/23 6:55 AM, Peter Nyikos wrote:
> > On Friday, November 17, 2023 at 9:45:04 AM UTC-5, Sight Reader wrote:
> >> On Thursday, November 16, 2023 at 2:54:33 PM UTC-7, Peter Nyikos wrote:
> >
> >>> Our phylogenetic analyses consistently recovered Fujianvenator within the clade Anchiornithidae, which unites other Anchiornis-like taxa51 and is resolved as the earliest diverging group of the Avialae (Fig. 1b). Archaeopteryx was resolved within the Avialae as the sister to other avialans except the anchiornithids.
> >
> >> Ok, so I’m trying to wrap my puny little mind around this. Are they saying here that Anchiornithidae (which, I gather, contains Fujianvenator) and whatever group Archeopteryx belongs to are both different “spurs” of the Avialae tree that are sister to birds - but neither of lead directly TO birds?
> >
> > Correct. Paleontologists have abandoned Archie as a bird ancestor. But I have no idea
> > how much of this is due to careful anatomical study and how much due to an ideology
> > that dominates taxonomy, which claims that there is "no evidence" that any fossil species is ancestral
> > to any other species, fossil or extant.


> A little of both, really.

What would be an apomorphy that disqualifies it from direct ancestry? [See below for
why I am wording it this way.]

> > "no evidence" = no incontrovertible proof

> What would count as evidence? If I recall, you mostly won't say.

Your memory seems to be worsening lately. I've told you umpteen times
about what I call a "prime candidate for ancestry of __________________":
a reasonably complete skeleton [a fraction of missing ribs and vertebrae don't count]
with no apomorphies that could reasonably disqualify it from direct ancestry of __________________.

With basal avialans, I would also include "some pennaceous feathers" if the putative
descendant had them.

> I would suggest that the best evidence

Didn't you mean to say "the only thing *I* might consider to be evidence" instead
of the last two words? See "equality" above.

> would be character optimization on a tree
> showing a zero-length branch between a species and the ancestral node.

That would depend on how complete the skeleton was, and
whether almost ALL representative bones [and, in the case above, feathers]
were included in the analysis.

> But would even that be ver good evidence?

Mine is so demanding, it can only apply in a few places on the tree of
vertebrata. About the only place where it occurs in abundance is in Equidae.

Contrast that with the ubiquitous use of "sister group," where almost
everything is the alleged sister group of something else. This is why I
am annoyed at your doctrinaire claim that "prime ancestor candidate"
is not "objective" enough to be acceptable to a leading peer-reviewed journal.

> > Meanwhile, loose "sister group" talk is everywhere, as can be seen from the
> > on-again, off-again hypothesis that the correct grouping is
> > {Theropods, Ornithischians} Sauropods
> > rather than the > century old tradition that the following is correct:
> >
> > {Theropods, Sauropods} Ornithischians.

> I'm not quite sure if you think that's a bad thing.

It seriously undermines your notion that "sister group" claims are "objective."
At the very least, such claims should be coupled with "based on the
phylogenetic analysis in __________________." That way, people can judge
for themselves how near to being "objective" these claims are.

> But if that's the
> case, why? Shouldn't we always be prepared to question traditional
> wisdom in the light of new evidence?

"Prime ancestor candidate" does that very well, while giving us
a window into what the actual LCA might have been like.

> Isn't uncertainty preferable to
> false confidence?

Certainly, but look at the false confidence you displayed where my then-future
reply to this very post of yours was the issue in talk.origins:

________________________ begin excerpt, my words alternating with your ripostes_______________________

> Full disclosure: Harshman, as might be expected, tried to undermine
> the above at every turn in a reply to that post.

Yes, it might be expected that I would explain how you are wrong about
so much of what you say. Though you never listen, it's possible that
someone else may.

> However, you might be disappointed by the caliber of that reply.
> In any event, he will get a rebuttal tomorrow.

I doubt it. There might be a response, but I don't expect anything
substantive.

++++++++++++++++++++ end of excerpt
from
https://groups.google.com/g/talk.origins/c/yl0TJZ0nueg/m/LR0mNH1OAAAJ
November 22, 2023, 1:01 AM
Re: DNA PROOFREADING AND ?REPAIR MECHANISMS ~ REVISITED
[original subject line: Origin of Life Challenge]

Your cocksure attitude in the excerpt is well suited to talk.origins, where you have
quite a lot of people who look up to you as a "biology guru."
However, it imparts a tinge of hypocrisy to your question,

"Isn't uncertainty preferable to false confidence?"

Peter Nyikos
Professor, Dept. of Mathematics

Univ. of South Carolina -- standard disclaimer--
https://people.math.sc.edu/nyikos

[erik simpson]

We've been over all this before. Who is the intended audience for your
proposed ancestor candidates? It's certainly not paleotologists or
evolutionary biologists.

[John Harshman]

Obviously, no apomorphy could completely disqualify it, since reversals
are known to happen. Then again, all science is provisional. Apomorphies
just make ancestry less likely.

>>> "no evidence" = no incontrovertible proof
>
>> What would count as evidence? If I recall, you mostly won't say.
>
> Your memory seems to be worsening lately. I've told you umpteen
> times about what I call a "prime candidate for ancestry of
> __________________": a reasonably complete skeleton [a fraction of
> missing ribs and vertebrae don't count] with no apomorphies that
> could reasonably disqualify it from direct ancestry of
> __________________.

But wouldn't the same likely apply to a fairly close relative of the
ancestral species?

> With basal avialans, I would also include "some pennaceous feathers"
> if the putative descendant had them.

Why that one in particular?

>> I would suggest that the best evidence
>
> Didn't you mean to say "the only thing *I* might consider to be
> evidence" instead of the last two words? See "equality" above.

There is no such word above, so I'm not clear on what you mean. What
other evidence would you prefer?

>> would be character optimization on a tree
>> showing a zero-length branch between a species and the ancestral node.
>
> That would depend on how complete the skeleton was, and whether
> almost ALL representative bones [and, in the case above, feathers]
> were included in the analysis.

If I understand you, you're saying that a zero-length branch would be
more significant if the taxon in question had more characters scored.
That's not unreasonable.

>> But would even that be ver good evidence?
>
> Mine is so demanding, it can only apply in a few places on the tree of
> vertebrata. About the only place where it occurs in abundance is in
> Equidae.

So you're saying that your criterion is nearly useless?

> Contrast that with the ubiquitous use of "sister group," where
> almost everything is the alleged sister group of something else. This
> is why > I am annoyed at your doctrinaire claim that "prime ancestor
> candidate" is not "objective" enough to be acceptable to a leading
> peer-reviewed journal.

I don't understand what you're trying to say there. And I don't have
enough information to speculate. Sister groups are bad for some reason,
while ancestor candidates are good for some reason, even though based on
the previous statement they can almost never be determined.

>>> Meanwhile, loose "sister group" talk is everywhere, as can be seen
from the
>>> on-again, off-again hypothesis that the correct grouping is
>>> {Theropods, Ornithischians} Sauropods
>>> rather than the > century old tradition that the following is correct:
>>>
>>> {Theropods, Sauropods} Ornithischians.
>
>> I'm not quite sure if you think that's a bad thing.
>
> It seriously undermines your notion that "sister group" claims are
> "objective." At the very least, such claims should be coupled with
> "based on the phylogenetic analysis in __________________." That way,
> people can judge for themselves how near to being "objective" these
> claims are.

You're saying that any doubt about any node on the tree renders all
other nodes equally uncertain? I don't think you actually know what
you're saying, except that it's spinach and to hell with it.

>> But if that's the
>> case, why? Shouldn't we always be prepared to question traditional
>> wisdom in the light of new evidence?
>
> "Prime ancestor candidate" does that very well, while giving us
> a window into what the actual LCA might have been like.

Not sure how that does anything. "Prime anceestor candidate" is a much
more bold claim than "sister group", one for which you are unlikely even
under your expressed criteria to have evidence. Nor is it necessary to
do in order to have an idea of what the LCA would look like; character
optimization at the ancestral node is what you need for that, and it's a
better estimate than what you want would be.

>> Isn't uncertainty preferable to
>> false confidence?
>
> Certainly, but look at the false confidence you displayed where my
> then-future reply to this very post of yours was the issue in
> talk.origins:

That's what we call sea-lioning. Turns out I was right, though.

> ________________________ begin excerpt, my words alternating with
> your ripostes_______________________
>
>> Full disclosure: Harshman, as might be expected, tried to
>> undermine the above at every turn in a reply to that post.
>
> Yes, it might be expected that I would explain how you are wrong
> about so much of what you say. Though you never listen, it's possible
> that someone else may.
>
>> However, you might be disappointed by the caliber of that reply. In
>> any event, he will get a rebuttal tomorrow.
>
> I doubt it. There might be a response, but I don't expect anything
> substantive.
>
> ++++++++++++++++++++ end of excerpt from
> https://groups.google.com/g/talk.origins/c/yl0TJZ0nueg/m/LR0mNH1OAAAJ
> November 22, 2023, 1:01 AM
> Re: DNA PROOFREADING AND ?REPAIR MECHANISMS ~ REVISITED [original
> subject line: Origin of Life Challenge]
>
> Your cocksure attitude in the excerpt is well suited to
> talk.origins, where you have quite a lot of people who look up to you
> as a "biology guru." However, it imparts a tinge of hypocrisy to your
> question, >
> "Isn't uncertainty preferable to false confidence?"

You would have to first establish that the confidence was false.

[Peter Nyikos]

No, we haven't. Harshman's display of arrogance over in talk.origins is new even for him,
but it is true that your failure to bat an eye at it is the same old same old with you.

> Who is the intended audience for your
> proposed ancestor candidates?

(1) Anyone who wants to make sense of Kathleen Hunt's excellent FAQ on
the horse superfamily:
http://www.talkorigins.org/faqs/horses/horse_evol.html

In it, Kathleen Hunt displays, not a phylogenetic tree, but what Richard Zander
and some others have called an "evolutionary tree," which shows hypothesized
direct ancestry, with real animals known from fossils occupying the nodes.

In contrast, phylogenetic trees even avoid giving names to the LCA's that
form the nodes in them. Even ichnofossils are given names, but the
doctrines to which Harshman subscribes rate ichnofossils higher than actual
species whose existence AND main suite of characters is not in doubt.

Now, what Kathleen's actual text conveys is that every genus-to-genus line in that evolutionary tree represents
a prime ancestor candidate in the genus at the lower end for the genus at the upper end.
Whether she was justified in it, it provides a rationale for the existence of the tree.

(2) Evidence that creationists cannot wave away that actual evolution has occurred.
Trying to use a phylogenetic tree for the same end invites a huge amount
of debate from more knowledgeable creationists who know that it is impossible
to even approximate direct ancestry from them. For instance, the echidnas are the
sister group of the platypus, but it would be madness to claim that one is descended from the other.

> It's certainly not paleotologists or
> evolutionary biologists.

Kathleen Hunt was both. Who are you trying to kid?


Peter Nyikos

Professor, Dept. of Mathematics -- standard disclaimer--

University of So. Carolina at Columbia
http://people.math.sc.edu/nyikos

[John Harshman]

You seem to have a habit lately of mentioning my posts and attacking
them in replies to other people, but not responding to me. This is not
healthy.

You keep returning to Hunt's FAQ. Is it in fact the only example you
would consider useful? If so, of what value is this whole idea of
"ancestor candidates"?

> In it, Kathleen Hunt displays, not a phylogenetic tree, but what Richard Zander
> and some others have called an "evolutionary tree," which shows hypothesized
> direct ancestry, with real animals known from fossils occupying the nodes.

Is that actually true, or are the nodes "genera"? Those are two
different things.

> In contrast, phylogenetic trees even avoid giving names to the LCA's that
> form the nodes in them. Even ichnofossils are given names, but the
> doctrines to which Harshman subscribes rate ichnofossils higher than actual
> species whose existence AND main suite of characters is not in doubt.

I'm not seeing the point here. Ichnofossils are physical objects. Nodes
are not. Nodes don't need names, nor do they have to be identified with
actual fossils. And this has nothing to do with how things "rate".

> Now, what Kathleen's actual text conveys is that every genus-to-genus line in that evolutionary tree represents
> a prime ancestor candidate in the genus at the lower end for the genus at the upper end.
> Whether she was justified in it, it provides a rationale for the existence of the tree.

Does the tree need that rationale? Is really necessary to claim that the
fossils are direct ancestors in order to justify the tree? No, in both
cases.

> (2) Evidence that creationists cannot wave away that actual evolution has occurred.
> Trying to use a phylogenetic tree for the same end invites a huge amount
> of debate from more knowledgeable creationists who know that it is impossible
> to even approximate direct ancestry from them. For instance, the echidnas are the
> sister group of the platypus, but it would be madness to claim that one is descended from the other.

You underestimate creationists. They can wave away anything. You
overestimate them too, claiming that there were any who were more
knowledgeable. I don't think any of them are more knowledgeable than you
are, which, on the subject of phylogenetic trees, is quite inadequate.

Imaginary lines from one fossil to another are not evidence, and there
probably are creationists who know that. This is manufacturing evidence,
in essence lying for evolution. That's not a good strategy.

>> It's certainly not paleotologists or
>> evolutionary biologists.
>
> Kathleen Hunt was both. Who are you trying to kid?

What he presumably meant was modern paleontologists or evolutionary
biologists. Science advances, even if you don't.