LIMUSAURUS

Not to be outdone by its trail-blazing "dino death trap" bedfellows Guanlong (Asia's oldest tyrannosaur), Yinlong (the oldest ceratopsian) and Jiangjunosaurus (the first Jurassic representative of Stegosauridae from the Shishugou Fm. of Xinjiang), Limusaurus is the first definitive member of its family known from Eastern Asia and one of the earliest known from anywhere. It's a ceratosaur, but choc-a-bloc with weird features.

Typically, ceratosaurs are large-headed, long-toothed, stocky-legged critters with fused ankles and short arms, and most have a fondness for anything made of flesh, even members of their own kind. By comparison, Limusaurus had a small head and long slender legs, and it lacked claws and teeth but preserved evidence of a beak and gastroliths which all point to a plant-based diet. In a nutshell, Limusaurus appears to be the only beaked, herbivorous theropod known from the Jurassic. But it's superficially similar to (and sandwiched between) the Triassic shuvosaurids and the Cretaceous ornithomimids which is a remarkable case of convergent evolution among three distinct groups of archosaurs that are separated by an enormous chunk of time. However, it's the hands of Limusaurus that get all the attention.

Based on fossil evidence, paleontologists knew that tetanurans — the branch of theropod dinosaurs closest to birds — arrived at tridactyl (three-fingered) hands from five fingered ancestors, but assumed that they lost digits 4 and 5 (ring finger and pinky) because digit one looked like a thumb. However, super speed "evolution" witnessed by a chosen few in the embryos of modern birds showed that they began with five fingers then lost digits one and five (the thumb and pinky). This difference in hand form has long been the last holdout for the "birds are not dinosaurs, their hands are different" brigade. But Limusaurus has all but lost digits one and five too: its thumb has shrunk to a mere nub, and the pinky is gone completely.

Because of its position way down on the theropod family tree, some experts now suspect that ceratosaurs such as Limusaurus may be ancestral to Tetanurae rather than a sister group, which would demolish the discrepancy between theropod and bird hands but also mean that tetanuran digits have been misnumbered for a century and a half! Of course, it's equally plausible that Limusaurus just started a trend of "finger shrinking" that culminated in the Cretaceous abelisaurid ceratosaur Carnotaurus whose fingers were all shrunk to fused, virtually useless, unclawed stumps. Or that it was closer to the slender-footed noasaurid abelisauroid ceratosaurs and evolved bird-like fingers to exploit a niche. Heck, it may just be an oddball that did its own thing and its weird hands, perhaps undergoing a "frameshift" whereby digits two, three and four took over the identities of one, two and three (index finger became thumb, and so on), had no bearing on the design of any other dinosaur.