Issues with the Linnean System - Reptiles and Birds
What makes a bird a bird? - In the Linnean view, at some point, creatures cease being reptiles and evolve into birds. Looking at living organisms, this is a clear-cut distinction. Birds could be distinguished by a long list of list of novel characteristics. When we consider extinct organisms, however, the picture blurs.
Here is a graphic representation of the tree of theropod dinosaur evolution as many biologists envision it, with both modern birds and a representative sampling of non-avian members. These critters pose difficulties for our bird identification scheme.
- Allosaurus. doesn't look like a bird, but it definitely has two of our birdy characteristics: Hollow limb bones and the furcula.
- Now look at Sinosauropteryx. Fossils of this small animal preserve what look like filamentous integumentary structures resembling downy feathers. Such structures are also observed in other non-avian theropods, like the early "raptor" Sinornithosaurus and the early tyrannosaurid Dilong.
- It gets worse. Some workers don't think that the filaments of Sinosauropteryx are feathers, but the fossil remains of the oviraptorosaur Caudipteryx are universally accepted as proper contour feathers. Is it a bird or a reptiles?
- Now look at Archaeopteryx, a creature that textbooks traditionally describe as the "first bird": Like Caudipteryx, it has contour feathers, but unlike it, the forelimbs are modified for flight. Is Archaeopteryx a bird? Before you answer, remember that it still has a long unfused tail and a mouth full of teeth.
- Finally, the fossil record gives us very birdy looking creatures like Hesperornis and Ichthyornis, with short fused tails. Even these, however, retain toothy "reptilian" mouths. Are they, perhaps, still reptiles?
Alas, we can't. No uniform criterion exists for choosing key characteristics in such circumstances. Consequently, whichever characteristic we pick, be it furculae, feathers, or toothless beaks, must inevitably be arbitrary and subjective, in other words, unscientific. Better to base our definitions on something more concrete - the fact that birds are descended from the common ancestor of reptiles. To say that birds are members of Reptilia reflects this pattern of descent without invoking arbitrary criteria.
Situations in which one group disappears by evolving into something different is termed a pseudoextinction. As a practical matter, if we wish to count either extinctions or numbers of Linnean taxonomic groups we must somehow factor these out.
By the way, if you don't buy the dinosaurian origin of birds, it really doesn't matter. Regardless of which reptiles one thinks birds descend from, one faces analogous problems.