Birds, in the sense of flying descendants of feathered reptiles (a more expansive group than the "true" birds in today's skies), evolved several times from within the theropods.

Feathered theropods and forearm powered flight in birds

It is widely known that powered flight has evolved at least twice amongst the reptiles, that is in the pterodactyls and the birds. So too it is common knowledge that the birds evolved from the theropod dinosaurs, and this is exemplified by the bird-dinosaur Archaeopteryx from the Upper Jurassic of Germany. What is probably far less well known is that birds, in the broad sense of flying descendants of feathered reptiles (a more expansive group than the “true” birds in today’s skies), evolved at least twice, and possibly several more times, from within the theropods. These insights have come from a diversity of remarkable fossil discoveries, especially a number from northeastern China, which in turn have greatly refined our understanding of theropod evolutionary relationships. Within the reptile family tree, debate around the evolution of flight largely revolves around feathered reptile specimens assigned to the theropod group ‘Paraves’. The Paraves encompasses basal paravians, the dromeosaurs, troodontids and the avialids, within which we also find Archaeopteryx as well as the modern birds.

Among key paravian fossils not only has the bipedal stance been confirmed (a necessary pre-requisite for the fore-arms to be converted into wings), but this group also sported a remarkable variety of integumentary structures, derived from scales, some of which were the precursors of flight feathers. Why these integumentary structures originally evolved is a matter of debate, although sexual selection (analogues of the peacock’s tail) is certainly likely, and in some cases also thermoregulation as at least the birds moved to full warm-bloodedness (which is convergent with that of mammals). It is important also to realize that possession of feathers in reptiles does not automatically mean a capacity for flight. In the dromeosaur Velociraptor, for example, there is good evidence for feathers (in the form of so-called quill knobs on the fore arms), but the animal was far too big to fly. 

So what is the evidence for multiple origins of birds?

1. Archaeopteryx and the avialids

Archaeopteryx needs little introduction as an iconic candidate for the most primitive ‘bird’, being the earliest member of the avialids (Avialidae). It is worth mentioning that despite 150 years’ study of this exceptional bird-dinosaur (supplemented by the occasional new find) new observations continue to be made. For example, it is now known that in addition to the well-developed fore-arm feathers that likely enabled a degree of powered flight, its legs also bore feathers.

In addition to Archaeopteryx, however, there has been a series of remarkable discoveries of fossil feathered reptiles, most notably from the Middle-Late Jurassic and Early Cretaceous deposits of Liaoning (China), which have enormously extended our knowledge of early bird evolution and even cast questions over the canonical place of Archaeopteryx at the base of the avialids.

A key case in point is the paper by Pascal Godefroit et al. (Nature 2013, 498, 359-362: A Jurassic avialan dinosaur from China resolves the early phylogenetic history of birds). In their paper they documented the paravian Aurornis xui and suggested that it (and its close relative Anchiornis) were flighted avialids that significantly predate Archaeopteryx. If correct, Aurornis would then displace Archaeopteryx as the earliest bird and the new avialid grouping would mean that powered flight using feathered forearms only evolved once. So much for convergent evolution of birds then? Perhaps, but then again perhaps not. The intense diversification of paravians evident from the Late Jurassic to Early Cretaceous, as well as the existence of species whose affinities are exceedingly difficult to define makes the hunt for solid answers a tough challenge. For example, while Godefroit’s analysis of Aurornis puts all the powered fliers in one group, other analyses place Aurornis and Anchiornis among the troodontids, that is in a sister group to the avialids. Another uncertain link relates to the proposed avialid Xiaotingia, that some believe could be a primitive dromeosaur. The discoverers of Xioatingia (Xu et al. Nature 2011, 475, 465-470)even tentatively place Archaeopteryx, Anchiornis and Xiaotingia in a separate group outside the avialids. If correct, this would support at least two origins of powered flight.

2. Rahonavis, early bird or flying dromeosaur?

Thanks to critical lower Cretaceous fossil finds, two main ‘true bird’ groups are now identified (Enantiornithes and Ornithurines). Some evidence suggests that both were largely ectothermic and full homeothermy (warm-bloodedness) only evolved later. However, important evidence is available from elsewhere, including preservation of feathers in amber (although these are not necessarily from true birds), and also very importantly remains of an upper Cretaceous bird from Madagascar, known as Vorona. From the same deposits another fossil Rahonavis has also been collected. Rahonavis is regarded by some to be an early bird capable of rather clumsy flight, and yet more plausibly it belongs with the dromeosaurs, sharing with many of them the extraordinary sickle-like claws on its feet. If Rahonavis is a genuine dromeosaur, this would represent the independent acquisition of forearm-powered flight and so point to the convergent evolution of ‘birds’.

3. Microraptor, a four-winged dromeosaur

Microraptor fossil” width=”218″ height=”218″>

The genus Microraptor is a basal dromaeosaurid, and it was a considerable surprise that one species (M. gui) was not only feathered, but had feathers on both fore-limbs and hind-limbs. In other words here we have a four-winged dinosaur. This in itself is of great interest because it may indicate that flight in theropods went through an initial four-wing phase. Modelling of this animal suggests, however, that it was not capable of powered flight and is generally considered to be a superb glider. Earlier in reptile history, a similar glider evolved in the form of Longisquama insignis. Longisquama is a Late Triassic archosaur (very distantly related to the dinosaurs) with four ‘wings’ and feathers distinct in structure from those of Microraptor. Longisquama‘s discovery has sparked debate about possible convergent evolution of feathers and even of birds.

Birds: convergence or common ancestor?

Sceptics of convergence might claim that these examples represent evolution within the theropods, and given they are bipedal and possessed complex integumentary structures we should not be too surprised to see that they learnt to fly (and glide) several times. True, but recall first that bipedality long preceded the capacity to fly. Most importantly we see striking convergences between the earlier archosaurs (flourishing in the Triassic) and the somewhat later dinosaurs, including the parallel emergence of theropod-like forms. So if we accept that amongst the reptiles a “theropod” is on the cards, then so too a good argument exists that the evolution of birds, arising from feathered reptiles, was inevitable.