A team of researchers in Britain did a new analysis of hundreds of dinosaur skeletons and have concluded that the current classification of dinosaurs into two major groups called the Saurischia (“lizard-hipped”) and Ornithischia (“bird-hipped”), first proposed in 1888, is not correct.
Traditionally, the Saurischia grouped the bipedal, mainly meat-eating theropods (including Allosaurus, Tyrannosaurus, and, through evolution, birds) with the giant, mainly quadrupedal long-necked plant-eating sauropods (Diplodocus, Apatosaurus, etc.). Both groups have some bones that are hollow or have other evidence of air sacs and pneumatic features, and both share a similar basic arrangement of the pelvic bones (although some theropods later developed a birdlike arrangement, including, of course, bird themselves). The other group of dinosaurs called the Ornithischia (including duckbill dinosaurs (Edmontosaurus), horned dinosaurs (Triceratops), and armored dinosaurs (Stegosaurus, Ankylosaurus) apparently lacked air sacs or pneumatic openings in their bones and had a birdlike arrangement of their pelvic bones–but they were not directly related to birds. Most ornithischians were plant-eaters, although some forms small may have been omnivores.
The new study found evidence that the theropods (Theropoda) and Ornithischia belong together in an evolutionay group (clade), which they call the Ornithoscelida, while the sauropods and a primitive group of Triassic meat-eating dinosaurs known as herrerasaurs should be grouped together as a separate evolutionary branch, which they gave the existing name Saurischia (now excluding the Theropoda and birds).
This new reclassification is controversial, but may find possible support from the recent discovery that some ornithischians had feather-like or bristle-like features on their skin. Fossil evidence of primitive feather-like coverings has now been documented for a wide range theropod dinosaurs, beginning with simple, primitive “dino-fuzz” up to the development of true feathers in theropods related to birds, including the highly specialized feathers found on modern birds, used in flight. Even relatives of Tyrannosaurus found in China show clear traces of primitive fuzzy feathers on their bodies. Some small ornithischians found in China and recently in Siberia also have featherlike coverings, although these structures are still under study and have not been fully described. Such feather-like features may be a shared ancestral feature that unites theropods and ornithischians. So far, there is no evidence of feather-like features on sauropods and their early relatives. No feather-like features have been confirmed for the primitive herrerasaurs as well.
The study also suggests that dinosaurs must have evolved earlier than now widely assumed, perhaps as early as the Middle Triassic. Although some dinosaur-like fossil tracks from that period might come from primitive dinosaurs, the skeletal fossil evidence for dinosaurs before the last part of the Late Triassic is sparse. Adding to the difficulties, paleontologist now recognize that there were different “dinosauromorph” groups in the Triassic Period that shared some features with dinosaurs but are usually classified as outside the Dinosauria proper. Because the fossil evidence for the earliest dinosaurs is often fragmentary or open to interpretation, the full history of dinosaurs may await the discovery of better fossils from earlier in the Triassic Period.
Matthew G. Baron, David B. Norman & Paul M. Barrett (2017) A new hypothesis of dinosaur relationships and early dinosaur evolution. Nature 543: 501–506 doi:10.1038/nature21700
Free pdf of supplementary information: