Archaeopteryx: Bird or not?

[rogue06]

The more I think of this the more I think it is deserving of its own thread so I'm going to repost this from the "Fossil Finds" thread:

A collector in western Liaoning in northeastern China found a fossil from deposits dated at around 155 myo (Late Jurassic) and sold it to the Shandong Tianyu Museum of Nature that is causing quite the commotion as it has led to a reassessment of just where in the evolutionary transition of dinosaur to bird does Archaeopteryx actually fit.

Archaeopteryx has long been accepted as being the most basal bird on the evolutionary tree although discoveries, primarily from China, over the past few decades has led to this status being questioned.

This new discovery, which the researchers describe as being "a new Archaeopteryx-like theropod" (an "archaeopterygid") has appeared to push the situation to the boiling point.

The chicken-sized fossilized creature in the center of all the controversy has been given the name of Xiaotingia zhengi and analysis of it suggests that Archaeopteryx was actually just a relative of the lineage that ultimately gave rise to birds.

Essentially what happened is that the researchers, led by paleontologist Xing Xu from Linyi University in Shandong and the Chinese Academy of Sciences in Beijing, listed all of Xiaotingia's birdlike and dinosaurlike features and fed the data into a computer along with measurements from 89 fossilized dinosaur and bird species including ancestors of modern birds, deinonychosaurs and Archaeopteryx.

Without Xiaotingia the computer's phylogenetic analysis put Archaeopteryx on the evolutionary line leading to modern-day birds. But when it was included the resulting family tree clustered the two of them together - out of the avian category and into the birdlike dinosaurs called deinonychosaurs (the same group that contains Velociraptors).

Apparently the reclassification was due to details concerning the shape of the furcula (wishbone) and pelvis as well as a large hole in the skull's snout (the “premaxillary fenestra”) above the nose. All of these things have been noted before.

Xu as his colleages conceded that their reclassification was "only weakly supported by the available data," and has “tentative statistical support,” but declared that this kind of fuzziness was to be expected when the fossils being analyzed are close to the common ancestor of now-extinct dinosaurs and modern birds. "This phenomenon is also seen in some other major transitions, including the origins of major mammalian groups," they noted.

Still others see the conclusions as being nothing really new and some have even been arguing that deinonychosaurs are birds themselves, although flightless ones - sort of like today's ostriches.

It should probably be noted that all of this doesn't in any way change the fact that birds evolved from dinosaurs. Rather, if correct, it only moves Archaeopteryx over into a side branch of that process. And as noted, many have been arguing for years now that Archaeopteryx wasn't a direct ancestor but belonged in some sister group.

There is one thing that all sides can readily agree to though: "If this new phylogenetic hypothesis can be confirmed by further investigation, current assumptions regarding the avialan ancestral condition will need to be re-evaluated."

Attachment 68314

Attachment 68313
Reconstruction*

Further Reading:

An Archaeopteryx-like theropod from China and the origin of Avialae Abstract

An icon knocked from its perch pdf.

Archaeopteryx no longer first bird

Happy 150th birthday, Archaeopteryx…you’re not a bird after all! Maybe

Earliest bird was not a bird? New fossil muddles the Archaeopteryx story

'Oldest bird' knocked off its perch

Flap Flop: Earth's First Bird Not a Bird After All

Now I'm sure we'll hear plenty from the YECs chortling about how this demonstrates that 'Archie' wasn't a bird like those evilootionists kept saying, completely ignoring all of their own leaders when they continued to swear up and down that it was a "true bird."

An excellent example is David Menton who has a heading (Archaeopteryx, a True Bird, Is Older than the “Feathered” Dinosaurs) in an article on AnswersinGenesis (AiG) called Did Dinosaurs Turn Into Birds?

Jonathan Sarfati at Creation Ministries International (CMI) has an article titled Archaeopteryx (unlike Archaeoraptor) is NOT a hoax—it is a true bird, not a “missing link”

Duane Gish of Institute for Creation Research (ICR) unhesitantly calls Archaeopteryx a bird in the opening sentence of his article As a Transitional Form Archaeopteryx Won't Fly. Further down he states that, "Archaeopteryx had an impressive array of features that immediately identify it as a bird, whatever else may be said about it" and "No doubt Archaeopteryx was a feathered creature that flew. It was a bird!"

In another article (Startling Discoveries Support Creation) Gish also declares that 'Archie' is a bird and that "many of the alleged reptile-like features of Archaeopteryx are actually bird-like."

Other YEC websites generally follow their cues from them.





*click on to see images

[phank]

I think understand what the creationists were trying to say. Since they deny that evolution happens, when it clearly has been happening for billions of years, they had a dilemma. Here we have something which is clearly transitional, with several key features of its precedessors and several key features of its descendents, both at once. BUT if you disallow transitions for theological reasons, you are limited to two options: you can claim the fossils are frauds, OR you can pick either MODERN birds or MODERN reptiles and classify it as one or the other, simply ignoring the mix of features. And the listed creationists apparently decided there were simply too many examples of these fossils to dismiss ALL of them as frauds, so they flipped a coin and it came up birds.

But the current attention being given to them really has nothing to do with creationist taxonomy. Instead, it seems we're getting yet another lesson in how the tree of life is a whole lot like a very dense bush, and not at all like a ladder. If the bush model is correct in this instance (as it tends to be generally), we would expect to see a great many branching lineages of "birdlike dinosaurs". And each lineage could (and sometimes would) produce child lineages, which would produce other children. I doubt paleontologists would be very surprised to find half a dozen major lines of development, each with multiple subdivisions, leading to many dozens of "leaf" species which represented the end of each line.

One and ONLY one of these many many branches was the "actual" ancestral line leading to modern birds. (And it might be noted here that very probably a great many lineages that would qualify as modern birds, wherever and however we draw this line, also never made it to the present.) So what's been discovered here is an early line of birdlike dinosaurs that seems like it could more probably have been a direct ancestor of modern birds, while the line including Archaeopteryx may have been abundant and very successful at one time, but wasn't directly on the road to modern birds.

What's confusing all these creationists is, back in the Jurassic there were no modern birds. It's really more a matter of idle curiosity which of many developmental lines MIGHT have been the one that survived 150 million years of evolution to the present day (and equally a matter of curiosity which of today's many birds, if any, might be on the "direct line" to whatever passes for birds 150 million years from now). Creationists had something their theology forces them to fit into a currently existing box, even though the organism doesn't exist today, and their box didn't exist then!

And there's only one good path out of this confusion, which is to recognize that there was a period of 150 million years, at the start of which there wasn't anything like a modern bird, and at the end of which (after much trial and error) evolution had resulted in something much like we see living today. But if one denies Deep Time, and one denies evolution, and one denies transitions, then one has categorical problems, struggling to decide if green is "really" black or "really" white, these being the only two allowed colors. And if it turns out this fossil was orange instead, aha! See, it wasn't black after all, it must have been white, evolution is all wrong!

[ericmurphy]

Whether a particular specimen belongs to a particular group depends on how you diagnose inclusion in that group. What features are required for an organism to be called a bird? If Archaeopteryx has all those features, then it's a bird. If it's missing some, then it's not.

[phank]

Whether a particular specimen belongs to a particular group depends on how you diagnose inclusion in that group. What features are required for an organism to be called a bird? If Archaeopteryx has all those features, then it's a bird. If it's missing some, then it's not.

Eric, I'm surprised at you. You usually do so much better that I wonder what I'm missing. But pressing onward anyway:

Presumably, at the time of Archaeopteryx there WERE NIO MODERN BIRDS. What this means, in practice, is that there were no groups that had all the features of modern birds. None. It's kind of silly to try to classify organisms from 100 million years ago according to whether or not they possess ALL the features that hadn't yet evolved!

I would suppose that if biological taxonomists had existed at that time, and knew about all the species extant at that time, and were charged with producing a Linnaean classification, then Archaeopteryx would be a leaf species, whose lineage would be straightforward. They'd cheefully give it a species name, and name the genus of similar species to which it belonged, etc.

Accordingly, NOTHING alive at that time qualifies as a "bird". It's a basic mistake to try to cram organisms from that far into the past into modern cubbyholes. The more appropriate question is, how closely related is the Archaeopteryx lineage to the one that eventually produced modern birds? Most likely, that specific lineage, like nearly all lineages, faded out before today.

[ericmurphy]

Eric, I'm surprised at you. You usually do so much better that I wonder what I'm missing.

Well, let's find out.

But pressing onward anyway:

Presumably, at the time of Archaeopteryx there WERE NIO MODERN BIRDS.

I'm with you so far, with the caveat that whether or not this is true depends on how you define "modern birds." What we call a group is pretty much up to us. Whether or not a group is a natural group is not.

What this means, in practice, is that there were no groups that had all the features of modern birds. None.

Well, sure. At the time of Archaeopteryx, it's unlikely any theropod existed with a pygostyle, a feature of modern birds. Unless we just haven't found one yet, but based on what we have found, it's extraordinarily unlikely such an organism existed.

It's kind of silly to try to classify organisms from 100 million years ago according to whether or not they possess ALL the features that hadn't yet evolved!

Right; but you're missing my point. Archaeopteryx certainly wasn't a "modern" bird in that it was missing features all modern birds (let's say, features diagnostic of Neoaves) have. But was Archaeopteryx a "bird," is the question posed by the OP. Well, that depends on what you mean by "bird." What are the features diagnostic of inclusion in the group "birds" (or, shall we say, "Aves")? Does Archaeopteryx have all those features? If it does, then yes. If not, well, no.

But whether any organism belongs to a particular group depends on how you decide to diagnose inclusion in that group. You can make up any list of features that diagnose inclusion in a natural group, give that group a name, and then based on that list of features determine whether or not a given organism belongs in that group.

I would suppose that if biological taxonomists had existed at that time, and knew about all the species extant at that time, and were charged with producing a Linnaean classification, then Archaeopteryx would be a leaf species, whose lineage would be straightforward. They'd cheefully give it a species name, and name the genus of similar species to which it belonged, etc.

Accordingly, NOTHING alive at that time qualifies as a "bird".

Wait. Stop. How are you defining "bird"? Unless and until you take a position on that question, the truth of that statement is undecidable. For the longest time, possession of feathers was diagnostic of inclusion in the group "birds." You got feathers? Then you're a bird. You don't? Then, sorry.

Obviously, by that definition, a lot of theropods that look nothing like modern birds would be considered "birds." Tyrannosaurus might have been a "bird" under that definition.

It's a basic mistake to try to cram organisms from that far into the past into modern cubbyholes. The more appropriate question is, how closely related is the Archaeopteryx lineage to the one that eventually produced modern birds? Most likely, that specific lineage, like nearly all lineages, faded out before today.

I'm not sure I'd go with the word "mistake." I'd say it's more important to give names to natural groups, diagnosed by shared characteristics, and not paraphyletic or polyphyletic groups, if what you're concerned with is evolutionary relationships. Where we draw the line between "birds" and "non-birds" is largely a matter of convenience, so long as what we are talking about is a natural group. We call monotremes "mammals" even though they lay eggs. Why? Because they have features clearly diagnostic of mammals: fur and mammary glands. Mammary glands, in particular, are what the group is named for, so inclusion of the platypus in the group of organisms diagnosed by the possession of of mammary glands is not even really optional. Sure, they're not eutherians; they lack a placenta and certain other features diagnostic of eutherians. According to some phylogenies, Archaeopteryx is indeed a member of Aves, and therefore is arguably a bird. Is it a modern bird, i.e., a member of Neoaves? No. But what we call a group is largely a matter of choice and nomenclatural convenience. Whether it's actually a natural group is not.

[ericmurphy]

Let me expand a bit on what I mean:

We could, if we wanted to, define "birds" as "the last common ancestor of Archaeopteryx and Columba and all of its descendants." Such a group definitely exists, just as the group defined as "the last common ancestor of Archaeopteryx and Homo and all of its descendants" exists as a natural group (assuming, of course, that common descent is true).

Defining a group this way isn't all that helpful, though, because it doesn't give you any way to determine whether or not a given organism belongs to the group. What you really need is a list of character states that you believe is diagnostic of inclusion in that group. The hard part, of course, is coming up with a list that actually diagnoses a natural group. For example, a list of features like "eyes, wings, and bilateral symmetry" does not define a natural group, because such a group would place butterflies and bats together in a group that excluded dogs. But a list like "feathers, unidirectional lungs, pygostyle, opposed hallux, furcula, fused trunk vertebrae, etc." may and probably does diagnose a natural group. Such a group would exclude Archaeopteryx, which has free trunk vertebrae, but it's largely a matter of choice as to which assemblage of traits diagnoses inclusion in a given group, provided that such an assemblage really does define a natural group.

Whether we choose to classify Archaeopteryx as a "bird" or not is really up to us. Whether it belongs within a group we call "modern birds" which we then define as those organisms possessing a particular list of character states is not. If it has those traits, it's a modern bird. If it doesn't, it isn't. It can be argued (and I believe this is your argument) that it's not very helpful to classify Archaeopteryx as a "modern bird" (and it may even be argued that it's not very helpful to classify it as a "bird" at all). But doing so would not be wrong, unless and until you've come up with a list a character states diagnosing inclusion in either "modern birds" or, more inclusively, "birds," with at least one character state Archaeopteryx lacks.

[phank]

This is, I think, turning into a discussion of the utility and limits of taxonomy. When you emphasize a "natural group", I'm thinking you are talking about a clade - some organism and all those organisms that descended from it. And accordingly, if we deem the progenitor organism as a "bird", then every member of the clade ipso facto is a bird. This makes sense to me. It would also mean that the common ancestor or mammals, monotremes and marsupials be classified as a mammal.

Now, the problem with the clade approach is, as you go back in time to more and more distant common ancestors, you find "leaf" species of that clade diverging more and more. And so eventually, inclusion in the clade doesn't tell us all that much. A clade whose common ancestor was very similar to a modern rodent, might be useful. A clade starting with the common ancestor of rodents and rutabagas isn't so helpful.

I think the question in the OP is inherently misleading, because it's based on assumptions I don't think are useful - namely that the term "bird" can be extended indefinitely into the past, and that all past species can be classified into a bird/not bird division (even as the notion of "bird" itself evolves). I think it would be better to ask whether Archaeopteryx lies on the direct line to some modern birds, or rather is a member of some branch off of the line that eventually became modern birds, that did not survive to the present. And if the latter, WHEN did it branch off? Otherwise, the answer to the OP question is "somewhat".

[ericmurphy]

This is, I think, turning into a discussion of the utility and limits of taxonomy.

Well, any discussion of whether or not Archaeopteryx was a "bird" or not necessarily impinges on the subject of taxonomy, don't you think?

When you emphasize a "natural group", I'm thinking you are talking about a clade - some organism and all those organisms that descended from it.

That is in fact exactly what I'm talking about.

And accordingly, if we deem the progenitor organism as a "bird", then every member of the clade ipso facto is a bird. This makes sense to me. It would also mean that the common ancestor or mammals, monotremes and marsupials be classified as a mammal.

Yes, by the definition of "clade" (aka, "natural group"): all of the descendants of a common ancestor, including that ancestor.

Now, the problem with the clade approach is, as you go back in time to more and more distant common ancestors, you find "leaf" species of that clade diverging more and more. And so eventually, inclusion in the clade doesn't tell us all that much. A clade whose common ancestor was very similar to a modern rodent, might be useful. A clade starting with the common ancestor of rodents and rutabagas isn't so helpful.

Well, I think it tells you a great deal about common descent. But, that might not be what you want to know. You might want to know, for example, whether or not Archaeopteryx resembled modern birds. In many ways it does not. No modern bird has teeth, for one thing.

But I think it's very useful to talk about, say, the common ancestor of rodents and rutabagas. We would expect that the last common ancestor of, say, dogs and coyotes would look quite a bit like both dogs and coyotes. The common ancestor of rodents and rutabagas certainly looked nothing like either one, yet it must have had features that were common to both. I think there's probably a lot more to be learned about the natural history of life on earth by investigating the common ancestry of rodents and rutabagas than there is investigating the common ancestor of dogs and coyotes.

But I guess it depends on where your interests lie.

I think the question in the OP is inherently misleading, because it's based on assumptions I don't think are useful - namely that the term "bird" can be extended indefinitely into the past, and that all past species can be classified into a bird/not bird division (even as the notion of "bird" itself evolves).

I don't see that assumption in the OP, but maybe it's a matter of emphasis. Obviously we can extend further into the past the notion of "bird" than we can, say, "chicken," but we can't extend it as far as we can the notion of "vertebrate."

I think it would be better to ask whether Archaeopteryx lies on the direct line to some modern birds, or rather is a member of some branch off of the line that eventually became modern birds, that did not survive to the present.

Except that that question is probably unanswerable. Identifying direct ancestry in the fossil record is at best problematic, and is very likely impossible. How do we know that Archaeopteryx, meaning any member of that genus, left any descendants at all? How do we tie any particular later fossil to Archaeopteryx? It's always possible that what we think is a direct descendant of Archaeopteryx is actually instead descended from a close relative (same family, different genus, for example) that simply doesn't appear in the fossil record.

Stratocladists apparently believe that at least in some cases we can, in fact, identify direct ancestors in the fossil record, but I think this is only possible in extremely well-represented taxa, such as some marine invertebrates. The avian fossil record is pretty sparse, for fairly obvious taphonomical reasons. The chances that the actual common ancestor of all modern birds even appears in the fossil record seems to me fantastically unlikely.

I think the best we can say is that Archaeopteryx is more, or less, closely related to modern birds than some other taxon. Even if it were actually true that Archaeopteryx were the last common ancestor of all modern birds—how would we know that? The only claim we could ever make that is actually based on evidence is that we have not found any fossil that is arguably more closely related to (i.e., groups more closely based on character states) modern birds than Archaeopteryx so far.

[phank]

OK, this all sounds fine to me. We can say that Archaeopteryx represents a solid indication that some lineage of dinosaurs evolved into modern birds, and that Archaeopteryx falls somewhere in that development path. Not real helpful if we're trying to reconstruct direct ancestry, but very helpful if we're trying to show that birds weren't poofed into existence supernaturally.

[rogue06]

I think the question in the OP is inherently misleading,

Considering the OP consists of copying an earlier post of mine concerning the discovery and the controversy surrounding the analysis of it along with a prediction that some YECs will jump on this announcement when they've been calling it a bird for years I'm not sure what you mean by a misleading question in the OP.

[ericmurphy]

OK, this all sounds fine to me. We can say that Archaeopteryx represents a solid indication that some lineage of dinosaurs evolved into modern birds, and that Archaeopteryx falls somewhere in that development path. Not real helpful if we're trying to reconstruct direct ancestry, but very helpful if we're trying to show that birds weren't poofed into existence supernaturally.

Yes; there is a clear progression of traits over time, with more and more traits common to modern birds appearing as we get closer and closer to the present. None of this is proof of common descent, of course; science rarely deals with proof. But it is certainly overwhelmingly powerful evidence supporting common descent, and no competing hypothesis has ever been proposed that isn't riddled with ad-hoc-ery.

I think most evolutionary systematists have given up trying to establish direct ancestry, simply because in the vast majority of cases there's simply no rigorous way to do it. the only truly rigorous way of determining ancestor-descendant relationships is to witness a birth.

[phank]

Considering the OP consists of copying an earlier post of mine concerning the discovery and the controversy surrounding the analysis of it along with a prediction that some YECs will jump on this announcement when they've been calling it a bird for years I'm not sure what you mean by a misleading question in the OP.

What I mean is, at least as I read it, the question presumes a solid unambiguous agreement on the definition of "bird" at the time of Archaeopteryx. Because without such a definition, as I wrote, the closest we can come to answering it is to say "sort of".

Creationists, of course, are obliged to labor under a model which doesn't allow birds to evolve, only to be created recently in modern form. This model implies that very distant ancestors of modern forms aren't actually distant at all (the time not having itself been poofed into existence), and therefore all fossils can be neatly cubbyholed into current classifications. Since Archaeopteryx fit into a cubbyhole that no longer exists, which can't be admitted, and won't fit any current slot, which also can't be admitted, we see questions like the OP, which make sense ONLY within the creationists' artificial taxonomy.

[ericmurphy]

What I mean is, at least as I read it, the question presumes a solid unambiguous agreement on the definition of "bird" at the time of Archaeopteryx.

I'd say the definition of "bird" isn't all that solid or unambiguous today. "Aves" is frequently defined as a crown group, i.e., all the descendants of the last common ancestor of Archeopteryx and Columba, including that ancestor. By that definition, Archaeopteryx is definitely a "bird," although likely many other decidedly un-birdlike organisms would also be "birds." But, more importantly, stem-group definitions aren't really all that informative. Defining such groups by a list of diagnostic character states probably tells us more about the evolutionary history of such groups.

[phank]

Defining such groups by a list of diagnostic character states probably tells us more about the evolutionary history of such groups.

I think that's true, but begs the question a bit. Let's say we have a set of diagnostic characters. So would inclusion in that set mean that an organism has ALL those characters? Most of them? The most important ones? What about additional characters? So I understand what you mean about ambiguity. If our list is too small (so that many organisms can have all of them), we risk including things that clearly should not be included. Conversely, if the list tries to be exhaustive, we risk omitting critters that clearly should belong. In either case, we risk falling into the creationist trap of regarding "kinds" as mutually exclusive, and trying to draw borders through gray areas.

My view is that while the biosphere is "lumpy", the lumps aren't entirely without overlap and they tend to shift around with time.

[ericmurphy]

I think that's true, but begs the question a bit. Let's say we have a set of diagnostic characters. So would inclusion in that set mean that an organism has ALL those characters? Most of them? The most important ones? What about additional characters?

Well, this is where you come down to various phylogenetic techniques. An argument would have to be made in any case that whatever character states are in your list actually do diagnose monophyletic groups. And generally the way you make such a determination is just to try to construct such groups using your proposed list of diagnostic character states. A valid list will result in one, and only one (or, statistically speaking, a number small enough to be indistinguishable from one; for more info, check this out) phylogeny.

So I understand what you mean about ambiguity. If our list is too small (so that many organisms can have all of them), we risk including things that clearly should not be included. Conversely, if the list tries to be exhaustive, we risk omitting critters that clearly should belong. In either case, we risk falling into the creationist trap of regarding "kinds" as mutually exclusive, and trying to draw borders through gray areas.

Well, what you're really looking for, ultimately, is a single, consistent, non-arbitrary nested hierarchy of groups within groups. Such groups, ultimately, represent pretty much all of the evidence supporting common descent. If such groups did not exist, common descent would be unlikely to be true (or, at least, the argument in favor of it would be very weak).