The finds of very bird-like dinosaurs seem to reinforce the hypothesis that birds began to fly after fast running against the tree-dwelling and gliding hypothesis, since these animals have very long legs and no adaptation for climbing. Nevertheless, their behaviour is still not certain, and the family tree of early birds and bird-like dinosaurs is subject to caution.
Within the cursorial hypothesis for the origin of flight, some fast running animals with long tails would have used their arms to keep balance while running. Increasing the surface of the 'wings' could have helped them. The feathers could also have been used as a trap to catch insects or other preys. Progressively, the animals would have sprung on longer distances, helped by their forecoming wings.
The arboreal hypothesis states that the ancestors of birds lived in trees. They would have sprung from branch to branch, progressively increasing the surface of their 'wings' to develop a good gliding ability. After gliding, they would have begun to flap to increase their flying efficiency.
There is little evidence for tree-climbing dinosaurs (Microraptor is the only one known so far), but forest sediments are only rarely preserved. On the other hand, all known feathered theropods, except Microraptor have long and slim legs, and seemingly ran fast. Furthermore, they lack an opposable first toe, hence they could not perch.
The opposable first toe is found in Archaeopteryx, showing that it could have lived in trees (the fossils were found in lacustrian rather than forest sediments, but they could have been carried there after the death). However, Unenlagia, which is more distant to birds than Archaeopteryx, shows a shoulder allowing flapping, and no perching ability. This means that flapping probably occurred before tree dwelling.
This practically excludes a gliding phase. Gliding is the less probable as gliding and flapping require very different flight anatomical structures, not easily changed into the other.
If Archaeopteryx could climb, it could also run, as is shown by its long tail (for balance) and long legs. A young Archaeopteryx is known, with even proportionally longer legs than the adults.
Archaeopteryx had asymmetrical feathers, clearly an aerodynamic feature, hence it was almost certainly able to fly. More primitive bird-like dinosaurs such as Protarchaeopteryx have symmetrical feathers; it is not known whether it is possible to fly with such feathers.
Thus, the existence of the feathered theropods seems to strengthen the cursorial hypothesis against the arboreal one, even if, of course, some could have climbed trees.
Relationships among early birds and bird-like dinosaurs
It is really difficult to establish reliable relationships among these animals. This is mainly due to the fact that a unique adaptive character, namely flight, has lots of correlated anatomical consequences, that bird-like dinosaurs show some but not all flight features, and that they probably did not fly. Moreover, Rahonavis, found in the Late Cretaceous of Madagascar, is probably a remnant of an old group (due to geographical isolation); and Unenlagia (Late Cretaceous too) seems unable to fly, but had probably had flying ancestors...
The most widely accepted phylogeny is the following one (not every group is mentioned). Groups containing feathered species appear with a *. But as new discoveries occur at a high rate, this is subject to much caution.
|
+-Sinosauropteryx*
+-Tyrannosaurs
+-Ornithomimosaurs (Ostrich dinosaurs)
+-Troodontids
`-Maniraptors
|-+-Caudipteryx*
| +-Oviraptors
| `-Therizinosaurs* (segnosaurs)
`-+-Protarchaeopteryx*
+-Deinonychosaurs*
+-Megaraptor (?=Unenlagia)
+-Microraptor*
`-+-Rahonavis*
+-Unenlagia
`-Birds
+-Archaeopteryx*
`-+-Alvarezsaurs*
`-Modern birds*
In fact, Archaeopteryx is older than lots of genera coming before it in this tree (it is Late Jurassic): deinonychosaurs, for example, are known only during the Cretaceous. But there is only negative evidence that they did not exist at the end of the Jurassic. All the groups above are probably older than what appears in the fossil data.
This tree is subject to much caution. Troodontids could be inside Deinonychosauria. Deinonychosaurs could be a paraphyletic group, some of them closer to birds than others. Therizinosaurs could be close to ornithomimosaurs. Alvarezsaurs, a group of strange sprinters with long legs and short, some with one-fingered arms, could be close to ornithomimosaurs. Etc.
It has also been hypothesised that deinonychosaurs (and some other bird-like dinosaurs, if not all of them) were secondarily flightless: they could be descended from some primitive birds. This would explain why Archaeopteryx occurs earlier. It may be the case, even if it is thought that having a flying ancestor should leave some particular anatomical features that are not found in deinonychosaurs. The discovery of Sinornithosaurus bolsters this theory: it is considered as a primitive deinonychosaur but shows more characters related to flight than later dromaeosaurids. The page of John Jackson is devoted to a detailed argumentation in favour of this theory.