NOTE: Michael Benton usually posts free links to his papers, but this one is not up yet...
Feathers are epidermal appendages comprising mostly corneous Î-proteins (formerly Î-keratins), and are characteristic of birds today.
There are close connections in terms of genomic regulation between numerous regularly arrayed structures in the epidermis, including denticles in sharks, dermal scales in teleost fish, epidermal scales in reptiles, feathers in birds, and hairs in mammals.
The discovery that genes specific to the production of feathers evolved at the base of Archosauria rather than the base of Aves or Avialae (birds) is matched by fossil evidence that feathers were widespread among dinosaurs and pterosaurs, the flying reptiles.
This suggests that feathers arose first, as simple monofilaments, probably for insulation in the archosaurian ancestors of birds and dinosaurs during the Early Triassic, a time when land vertebrates were speeding up in terms of physiology, with erect gaits and endothermy.
Feathers have long been regarded as the innovation that drove the success of birds. However, feathers have been reported from close dinosaurian relatives of birds, and now from ornithischian dinosaurs and pterosaurs, the cousins of dinosaurs. Incomplete preservation makes these reports controversial. If true, these findings shift the origin of feathers back 80 million years before the origin of birds. Gene regulatory networks show the deep homology of scales, feathers, and hairs. Hair and feathers likely evolved in the Early Triassic ancestors of mammals and birds, at a time when synapsids and archosaurs show independent evidence of higher metabolic rates (erect gait and endothermy), as part of a major resetting of terrestrial ecosystems following the devastating end-Permian mass extinction.
In a recent paper, the contention that spinosaurine theropods were semi-aquatic was supported by Arden et al., (2018) and they provided a hypothetical sequence of acquisition of traits that had evolved in line with this lifestyle. However, we find that the presented traits were either loosely defined and / or are clearly distinct from those traits seen in extant animals with adaptations to life in water. Some spinosaurs may have spent extensive time in water, but the data to support this is currently insufficient and other hypotheses for their behaviour also fit the available data.