bird evolution

A couple of years ago an article on this group said that birds evolved from reptiles. I thought they came from dinosaurs. Which is correct?

At another point, it stated that all dinosaurs were bipeds. Is that correct? If so where does that leave such animals as brontosaurus, triceratops etc.? George Prehmus

Both. Dinosaurs are reptiles. Birds are reptiles. Birds are dinosaurs.

No. Dinosaurs were primitively bipedal, but as you note several groups returned to all fours.

Reptilia is a polyphyletic taxon ? Well: that would imply that the amniotic egg appeared several times independently during the evolution of tetrapods, which seems (to me) very unlikely... Reptilia is undoubtely a paraphyletic clade, but not polyphyletic.

Another question. I have a huge 'theoretical' problem with paraphyletic taxa and their treatment within the cladistic paradigm. Can anyone here give to me at least one (non-circular) reason to methodologically eliminate paraphyletic groups from our systematic *and* phylogenetic classifications? Why paraphyletic taxa should be estimated to be useless for our understanding of evolution? For me, taxa such as Amphibia, Reptilia, or Condylarthra, are very usefull systematic *and* phylogenetic concepts. Of course they are 'evolutionnary grades', but everything

Terminological quibble: there is no such thing as a paraphyletic clade, clades being defined as monophyletic groups.

Because they are arbitrary in a way that no clade is. Clades have objective existence. All clades on a given tree can be identified by reference to a particulare node, and all can be named in a single, consistent classification. However the number of paraphletic groups on the same tree is much larger (equal to the number of combinations of one ancestral node and one excluded, descendant node), and many of them are contradictory to each other and to some clades. So you have to choose among possible paraphyletic groups and clades if you want a hierarchical classification. If, for example, you pick 'Reptilia' (amniotes minus birds and mammals) as a group to name, you can't also pick 'Archosauria' (a clade including some reptiles and all birds). The criteria by which some paraphyletic groups are accepted and others rejected are unclear and arbitrary.

Why? What's useful about them?

Not true. Many clades are not grades, or at least it would be hard to come up with their defining characters. Is Sarcopterygia (the clade) a grade?

Only if you define your taxa in some essentialist fashion, in which loss of the defining character causes exclusion from the taxon. But we don't do that these days. And in fact we never really did. If we did then whales and snakes wouldn't belong to Tetrapoda.

Because there is a difference. Because paraphyletic taxa are arbitrarily delimited. Because they create the illusion that you have said something about evolution when you have just created a wastebasket 'not-group'. Because they encourage the impression that such groups are objective, in contrast to all the other paraphyletic and monophyletic groups that are not recognized, and can't be because they conflict with your favorite paraphyletic group (as in the Reptilia/Archosauria example above).

<snip>

Chasmatosaurus is a member of a group of animals that likely contained the LCA of dinosaurs but it is most definitely not the LCA.

Australopithecus is a genus composed of several species. One of those species may have been ancestral to Homo Sapiens but there is considerable debate on which species.

No, on obvious morphological details. Other archosaurs had equal length hind and fore legs while all dinosaurs have or have ancestors who had longer hind legs than forelegs. This a pretty clear indicator of bipedal ancestry. Coupled with the evidence that all known dinosaur groups evolved from bipedal ancestors the theory that the LCA of dinosaurs was a biped is unchallenged by anyone I know of who knows anything about the subject.

As far as I'm aware there isn't any suggestion that the Reptilia as a class are polyphyletic - that is, that synaspids, diapsids and the rest are independently derived from separate amphibian ancestors - the problem comes rather from the class's unhelpfully broad scope (essentially, any terrestrial vertebrate that can breed in the absence of water is a reptile or derived from reptiles) and the difficulty in defining exactly where the descendant classes Mammalia and Aves should be recognised. The bone of contention mostly involves these cases rather than the origin of the reptiles themselves - if birds are recognised as a separate class, should dinosaurs be given their own class? Are the 'mammal-like reptiles' (therapsids) really reptiles, very early mammals or a class of their own? And so on and so forth.

Philip Bowles

The fact that the first crocodilians (eg, Protosuchus) were bipeds strongly resembling early theropods is also suggestive of a bipedal common ancestor for both crocodiles and saurischian dinosaurs.

Philip Bowles

Why 'saurischian' here?

Ok, I'm convinced, I think.