The Fossil Record Throws A Curveball: Yi qi

Those crazy, crazy theropods. If there's one thing palaeontology has showed us about dinosaurs, it's that you shouldn't get used to their popular reconstructions because, sooner or later, something will turn up that'll really screw with your mind. And it's not like these events are necessarily rare; Deinocheirus and Spinosaurus got make-overs in the last couple of years, and they're both pretty high-profile.

Less high-profile are the Scansoriopterygidae, small, feathered, theropodan dinosaurs with the long arms you'd expect of an arboreal, aerial-capable dinosaur, but with an immensely-long third digit. A popular notion is that this digit was an adaptation to an arboreal lifestyle, enabling the creature to wrap its arm around tree trunks and branches, like naturalist David Bellamy, just, y' know, sharing the love.

B-b-but - what's this? A new paper by Xu, Zheng, et al, announces the discovery of a new scansoriopterygid, Yi qi, preserving not only the long fingers and feathers, but also a new, hitherto unseen structure. A long bony, or cartilaginous, rod projects backwards from each wrist, and patches of membrane suggest a set-up not totally unlike that of bats or pterosaurs. Or dragons, but I didn't say that. There's still some debate as to how the proximal margins of the wing chord may articulate, i.e., does it merge with the thoraic region or something else. And what is the true arrangement of the manual elements, in particular, the rear-pointing 'prong', referred to in the paper as the styliform element? They offer up a couple of possible arrangements, such as something superficially bat-like, and a set-up where the styliform elements are directed inwards, towards the body, helping to maintain a narrower chord. If this animal did indeed undertake powered flight, it's not too difficult to imagine it 'scooping' the air with its membranous hands, as bats do. Bats' hands' 'palms' form a sort of concave shape as they fly, which looks like a sort of arial butterfly stroke. Their fingers are fully jointed, enabling them to alter the shapes of their manus as required, resulting in a rather effective wing. The paper offers up three potential arrangements for Yi qi's 'wings', the two more plausible (to me) of which are shown here:

Two of three different arrangements proposed in Xu, Zhen, et al (2015), showing a proximally-pointing styliform element running parallel to the forearm (left), and the same feature, free of the forearm, pointing posteriorly and supporting a much-deeper membrane. (Illustrated by Gareth Monger; modified from Xu, Zhen, et al 2015.)

With regards the styliform element, I wonder if, rather than being curved in a horizontal plane (as restored, left) it instead curved ventrally (right), helping to maintain the aerofoil section - and a bat-like scoop. Some time after death, and prior to fossilisation, it has tipped over, rotating approximately 90 degrees, and settling in an unnatural position (left). Compression of the bones and associated remains during preservation could be masking the true shape of this apparently-unique element, but some lateral compression in life would make structural sense in terms of giving it strength during a downstroke. But that's all speculation.

In the paper, the wing reconstructions (shown in dorsal view) show the hind limbs of the animal trailing behind it. Although the main point of the graphic is to demonstrate the possible extent of the membrane, a trailing position for the hind limbs is unlikely; it pushes the centre of gravity back, and increases turbulence. For a volant theropod, it would seem unlikely that it would extend its legs behind it if they're not supporting part of a flight surface, and it also seems unlikely that a volant animal would rely on a narrow wing as suggested in the left-hand diagram. The right-hand diagram shows a deep chord, within which the (estimated) centre of gravity comfortably sits, when the legs are brought up, underneath the body, and out of the airflow.

Speculative illustration showing possible extent of contour feathers on Yi qi, and a possible centre of gravity. Note that the animal brings its legs in under itself, out of the airflow and therefore reduces turbulence. This also maintains a more-central centre of gravity. (Copyright © 2015 Gareth Monger)

Where the trailing edge of the membrane attaches (e.g., the body, or the hind limb) is not clear. Flying dinosaurs which use feathered wings benefit from legs which are independent of the wings. They can run into the airflow to achieve lift-off, or they can jump into the air, with the wings already committed to the flight strokes and not involved in the jump (compare pterosaur quad-launching). Having a skin membrane attached to the leg might be problematic since the legs (if not held out behind) would need to be elevated in order to maintain a level flight surface, and not one which partially faces into the airflow. However, that brings the leg and the membrane attached to it forward, reducing the tautness of membranous wing. Bearing that in mind, one might expect the membrane to attach on the body, somewhere in front of the hip, and not to the leg. The styliform element could work as a means by which the animal adjusts the tautness of the membrane, in a similar way to how a pterosaur is thought to do so with its apparent ankle attachment. Without that extra strut, the animal might enjoy less control and increased flutter in the membranes.

Yi qi in flight. (Copyright © 2015 Gareth Monger)

One of the key questions raised by this is why would a theropod go the route of developing a membranous flight surface when so much experimentation with flight (and there seems to be a lot of it!) is concerned with forming a continuous flight surface from elongated feathers? A major difference between scansoriopterygids and other, flighted, theropods is their elongated third digit. As suggested earlier on, it could be that this is an adaptation towards an arboreal lifestyle, enabling the animal to climb trees and other steep surfaces more easily. And it could be that selective pressures favoured the extension of the postpatagium instead of the feathers present on the arms. Whatever the case, feathers for flight persisted, and the theropodan flight membrane proved an evolutionary dead-end. Hopefully, additional specimens will come to light, adding to our understanding of this weirdo dinosaur.

Many thanks go to Mike Boyd for enabling me to write this particular article.