The world’s smallest dinosaur?

The ‘Ashdown maniraptoran’

Most palaeontologists now accept that birds are descended from non-avian dinosaurs and are therefore the only living representatives of otherwise terrestrial animals that ruled the world during the Mesozoic. That being the case, the Cuban bee hummingbird, which measures just 5cm in length and weighs about 1.8g, is arguably the world’s smallest dinosaur. However, if birds are excluded and looking only at the non-avian dinosaurs, all of which became extinct in the often-discussed mass extinction event at the end of the Cretaceous, what is the world’s smallest dinosaur? Until very recently and considering only animals that are thought to be fully grown, the answer is Anchiornis huxleyi (Hu et al., 2009; Fig. 1) from the earliest Late Jurassic of China. This tiny, feathered troodontid has been estimated to be between 34 to 40cm long (excluding the tail feathers). In the ever-changing environment of what represents the world’s largest and what the smallest, Anchiornis’ position as the world’s smallest dinosaur has now been threatened and quite possibly overturned. The fossil concerned is not another wonderful specimen from the Late Jurassic or Early Cretaceous of China, but a single, tiny cervical (neck) vertebra (Fig. 2) discovered at a brickworks near Bexhill, in East Sussex.

For more than 20 years, brickworks employee and keen fossil collector, Dave Brockhurst (Figs. 3 and 4), has been collecting vertebrate remains from Ibstock Brick Limited’s Ashdown works (Fig. 4). He first noticed large bone fragments in piles of clay destined for the brick kilns and, having rescued these, set about trying to locate the horizons from which they were derived. Several bone-bearing layers were identified and, between then and now, a valuable collection has been amassed. Apart from keeping an eye on the workings, both during his working hours and during his free time, Dave also organises collecting expeditions to the pit (which is not open to the general public) attended by Bexhill Museum staff and a small band of carefully vetted volunteers (Figs. 4 and 5). This dedicated approach has led to some remarkable discoveries, all of which have been donated to Bexhill Museum. For a summary of these, see Peter Austen’s account in Wealden News, No. 8 (http://www.kentrigs.org.uk/w8.pdf). This selfless approach is of huge benefit to science because, once part of a registered collection, specimens can be formally described and, where appropriate, named. According to internationally accepted rules, this cannot be done if specimens are held in private collections.

Many of the smaller specimens collected by Dave and colleagues come from a bed known locally as the ‘Conglomerate Bed’. This is of variable lateral extent and thickness, and occurs within the Valanginian Wadhurst Clay Formation of the Hastings Group (Wealden Supergroup) exposed in one of the two pits at the site. It contains abundant, water-worn fish remains, including hybodont shark teeth, and the teeth and scales of the large mollusc-eating fish, Lepidotes. Having heard of my work on small vertebrate remains from the slightly younger (Barremian) Wessex Formation of the Wealden Group of the Isle of Wight (see Issue 19 of Deposits, pp. 8-13) Dave contacted me to ask if I would be interested in processing samples from vertebrate bearing horizons at the pit, using the sieving techniques outlined in Issue 18 of Deposits, pp. 41-45. Surface prospecting and excavation of these had yielded bones of at least one salamander and a lizard, and it was hoped that by adopting a microvertebrate approach, yet more new small taxa would be found. This did indeed prove to be the case and a combined list of taxa discovered ‘by eye’ and using a sieve and microscope can be found in Naish and Sweetman (2011). The most notable discoveries were several new species of salamanders (the atlas vertebra of one is shown in Fig. 5), frogs and lizards. Outline descriptions of these are provided in a forthcoming book provisionally called, English Wealden Fossils: Field Guides to Fossils, 14, to be published by the Palaeontological Association later this year.


Fig. 6. A salamander atlas vertebra from the ‘Conglomerate Bed’. Unusually, and like the maniraptoran vertebra, the neural arch is also preserved in this specimen. A, dorsal; B, ventral; C, anterior; D, left lateral; and E, posterior views.

While this work is ongoing, so are Dave’s excavations and not long after he first approached me, he called to say that he had found a most unusual small vertebra, which he couldn’t identify. It stayed in a draw at Dave’s home for some time and when he eventually donated it to Bexhill Museum, a provisional identification of “snake vertebra” was attached to it. If this was correct, it would have represented the earliest record of a snake by some 25myrs or more and I was intrigued. Therefore, I found the first opportunity to visit the museum to have a look (and collect more samples from the pit). My first reaction when I saw the specimen was one of disappointment – it was clearly not the vertebra of a snake. However, closer examination revealed that this was indeed a most remarkable specimen. Although somewhat polished and abraded by being transported in a stream or river, the specimen is largely complete (Fig. 2). Commonly, the top part of vertebrae (the neural arch) breaks away from the body of the vertebra (the centrum) during transport before final burial. In this specimen, it is preserved and what was particularly interesting is the fact that no suture between the neural arch and centrum can be seen. This is a clear indication that the animal from which it came was fully mature at the time of its death. Further examination also revealed that this was, in fact, the vertebra of a theropod dinosaur and that being the case, with a centrum length of just 7.1mm, a very small one indeed. Excited by this, I arranged to borrow the specimen from the museum for further study. I am not a theropod expert (I prefer to study all the little things that lived with the dinosaurs, not the dinosaurs themselves) but luckily a University of Portsmouth colleague of mine, Dr Darren Naish, is. He quickly concluded that the specimen is a posterior cervical (neck) vertebra of a tiny maniraptoran dinosaur. Therefore, it represented a new discovery and, although we could not safely put a name to it (more fossils will be required before we can name the new species), we decided that it should certainly be reported in the scientific literature. It was only when this process was nearly complete that the possibly record-breaking small size of the animal the vertebra represents fully dawned on us.


Fig. 7. Ashdown maniraptoran cervical vertebra shown in approximate life position within a generalised maniraptoran silhouette. By duplicating and manipulating images of the specimen to reflect changes in bone shape along the neck, a schematic cervical column was generated and scaled to approximate size within the silhouette. From Naish and Sweetman (2011).

Almost as an afterthought, it was decided to make an estimate of the total size of the animal from which the vertebra came. But how do you do that based on a single neck vertebra? We decided to try two different methods and see how closely the results compared. The first involved building a digital model of the dinosaur’s neck and then fitting this into a silhouette of a generic maniraptoran (Fig 7). A considerable amount of art (some would say pure guesswork) rather than science was involved in this process, but we felt that the result might be interesting when compared to a slightly more mathematical approach. In the latter, we used neck-to-body ratios of other related dinosaurs to calculate the Ashdown maniraptoran’s length. Somewhat surprising to us, both methods yielded  similar results and suggested a skeletal length of between 16 and  40cm, with the most likely length being somewhere in the middle of this range. If so, it would have been slightly smaller that the minimum length suggested for Anchiornis (34cm) and certainly no bigger than the maximum estimate (40cm) (Fig. 8).


Fig. 8. Anchiornis and the Ashdown maniraptoran shown at the same scale to illustrate differences in size, based on an estimate for the latter as discussed in the text. Image of Anchiornis and outline shape of the Ashdown maniraptoran by Matthew Martyniuk.

So what was the Ashdown maniraptoran like? For reasons set out in our paper, Darren and I have concluded that the Ashdown vertebra is that of a non-avian dinosaur, rather than that of a bird (although birds are also part of Maniraptora), despite it having some bird-like characters. Our view is that, like other maniraptoran theropods, this would have been a small, feathered, bird-like bipedal dinosaur, with a fairly short tail (referring here to the skeleton, not the feathers), a long neck, long slim hind legs and feathered, clawed forelimbs (Fig. 9). It would probably have walked with its body and tail held in a horizontal position (Fig. 7) and, as in other small maniraptorans, it was perhaps an omnivore, eating small animals, including insects, but also perhaps leaves and fruit. To draw firmer conclusions, we need more fossils, so Dave continues digging and I continue sieving.


Fig. 9. The Ashdown maniraptoran – an artist’s impression by Matthew Martyniuk.

Hu, D., Hou, L., Zhang, L. and Xu, X. 2009. A pre-Archaeopteryx troodontid theropod from China with long feathers on the metatarsus. Nature 461 (7264), 640–643.

Naish, D and Sweetman S. C. 2011. A tiny maniraptoran dinosaur in the Lower Cretaceous Hastings Group: Evidence from a new vertebrate-bearing locality in south-east England. Cretaceous Research 32, 464–471.


♦ Buy Fossils, Crystals, Tools
♦ Subscribe to Deposits
♦ Join Fossil Hunts
♦ UK Fossil Locations