When I logged on to Facebook Wednesday morning, one of the first things I saw was a cryptic status update from University of Maryland paleontologist Thomas Holtz. He speculated that the paleo community at large would be "duly impressed" by something set to debut later in the day, but what was it? I jokingly replied that it would have to be something pretty impressive to outshine the weird raptor Balaur bondoc, but Holtz was right. Described in this week's issue of Nature, Concavenator corcovatus is one of the strangest dinosaurs ever found, and possibly one of the most significant.
On a superficial level, Concavenator looks very familiar. Discovered in the approximately 130-million-year-old rock of Spain, this dinosaur was a carcharodontosaurid, or an early relative of the giant Giganotosaurus and a somewhat distant cousin of Allosaurus. What made it unique, however, were a series of elongated, upward-pointing neural spines near its hips. This dinosaur did not have a sail running the length of its back, like Spinosaurus, nor did it have a more uniform set of elongated neural spines, like the carcharodontosauid Acrocanthosaurus, but instead had two neural spines that jutted up high right in front of its hips followed by a series of shorter—but still elongated—spines at the base of its tail. This kind of arrangement—a short, tall sail near the hips—had been proposed for a very incompletely known dinosaur named Becklespinax before, but with the mostly complete remains of Concavenator we now know that at least some predatory dinosaurs had this weird decorative arrangement.
With two (and possibly three, if Becklespinax turns out to belong to the same group) carcharodontosaurids with strange structures on their backs, sails, humps, or fins may very well be found on other members of this group. But, despite its flashy sail, the most impressive aspect of Concavenator is much more subtle. Arrayed in a line along its ulna—one of the two bones that make up the forearm—were a series of round, raised bumps. This is not the first time paleontologists have seen such a feature. In 2007 it was announced that Velociraptor had these same structures, and they looked identical to the quill knobs on the arms of birds where arm feathers attached. The question was: what was a dinosaur so far removed from the origin of birds doing with quill knobs?
During the past two decades, a flood of new fossils has confirmed that birds evolved from one lineage within the diverse, feather-covered group of theropod dinosaurs called coelurosaurs. Almost every lineage within this group has at least one feathered representative, but Concavenator was not a coelurosaur. As a carcharodontosaurid, its lineage last shared a common ancestor with the coelurosaurs back in the Middle Jurassic, and the knobs on its arms represent the first evidence of a body covering other than scales on a theropod outside the coelurosaurs. Just what these knobs supported is as yet unknown. Perhaps they were feathers, or maybe they were a kind of bristle that was structurally similar to feathers. Skin impressions from other parts of the dinosaur show that it was not entirely covered by such structures, meaning that Concavenator may have had a mosaic of scales and feather-like structures on its body.
Even better, the discovery that Concavenator had a type of filamentous body covering reinforces the emerging hypothesis that dinosaurs as a whole may have sported a variety of such structures. Within the past decade paleontologists have found at least two examples of ornithischian dinosaurs (Psittacosaurus and Tianyulong) with feather-like bristles on their backs. These animals were all the way on the other side of the major evolutionary divide in the dinosaur family tree—about as distantly related to birds as possible while still being dinosaurs—yet they, too, had unique body coverings that were similar in structure to the fuzzy precursors of feathers seen in some coelurosaurs. If ornithischians had bristles, coelurosaurs had feathers, and carcharodontosaurids had similar structures, then it is possible that feather-like body coverings were a common trait for dinosaurs that might go all the way back to their origins. Either that, or these structures independently evolved in different lineages multiple times during evolutionary history. Regardless of which hypothesis turns out to be correct, we need to rethink what we thought dinosaurs looked like, and I expect that we are going to see the discovery of further evidence in the years to come that many dinosaurs were feathery, bristly creatures.
Post-script: There is some debate as to whether the knobs on the ulna of Concavenator are truly quill knobs or are another feature associated with muscle attachments. For more details on this angle of the story, see the posts by Darren Naish and Mickey Mortimer.
Ortega, F., Escaso, F., & Sanz, J. (2010). A bizarre, humped Carcharodontosauria (Theropoda) from the Lower Cretaceous of Spain Nature, 467 (7312), 203-206 DOI: 10.1038/nature09181