Hastily 'updated' version of a Wessex Formation scene, now with ornithomimosaurs! 'Angloposeidon' in the background, Eotyrannus (with an Ornithodesmus) on a buttress root. Image by Darren Naish.

Regular readers of Tet Zoo might know that Ive published extensively on the theropod dinosaurs of a famous and much-studied Lower Cretaceous rock unit known as the Wealden Supergroup. Wealden rocks range in age from Berriasian to Aptian (that is, they were deposited between about 145 and 120 million years ago) and are most famous for their exposures on both the Isle of Wight and the south-eastern part of the English mainland. The Wealden spans such a huge range of time that whole dynasties of animals came and went over the course of its deposition: there was no Wealden age with a single set of contemporaneous animals.

Swamps, deltas, floodplains and semi-arid woodlands covered much of western Europe during Early Cretaceous times*, and among the many animals that lived here were such iconic Early Cretaceous dinosaurs as iguanodontians, polacanthids and titanosauriform sauropods. Wealden theropods include the spinosaurid Baryonyx, the carcharodontosaurian Neovenator and the tyrannosauroid Eotyrannus.

* For those who still find it confusing, the terms Early and Late mean different things from Lower and Upper when discussing the geological past. Early and Late refer to time (e.g., Iguanodon lived during the Early Cretaceous) whereas Lower and Upper refer to stratigraphy (e.g., Iguanodon is a Lower Cretaceous fossil, found in Lower Cretaceous rocks).

If you're interested in Wealden dinosaurs, it might be a good idea to own these two books. They're both surprisingly affordable.

The stratigraphy of the Wealden is complicated and its taken decades for us to devise a system that experts agree on. The oldest strata belong to whats known as the Hastings Group (deposited between the Berriasian and Valanginian, right at the start of the Cretaceous), while the younger ones at the top belong variously to the Weald Clay Group or the Wealden Group. Familiar Isle of Wight Wealden dinosaurs like Neovenator, Polacanthus and Mantellisaurus come from the Wessex Formation, the most fossiliferous and accessible Wealden unit. Deposited predominantly during the Barremian, the Wessex Formation is part of the Wealden Group. Note that Wealden Group and Wealden Supergroup refer to very different things. The stratigraphic chart below might help. If you really want to understand all of this in greater depth, see the several relevant chapter in Batten (2011).

My handy chart of Wealden stratigraphic terminology. Most of the animals discussed in this article are from the Wessex Formation, deposited during Barremian times. Valdoraptor (discussed below) is most likely from the Grinstead Clay Formation, and hence is Valanginian in age.

Theropods in the Wealden... again

Anyway, theropod fossils are not common in the Wealden theyre rare enough that virtually every single record is worth writing up and interpreting (err, in my opinion, anyway). And while Neovenator, Eotyrannus and Baryonyx are all known from decent associated partial skeletons, the majority of Wealden theropod records are scrappy, consisting of the odd vertebra, limb bone, or tooth.

Harry Seeley (1839-1909) and Richard Lydekker (1849-1915): two gentlemen who left us with some grand taxonomic messes... like that depicted in the scribbled flow-chart shown below.

When you combine this scrappy record with the fact that British scientists were interpreting these fossils at a time when virtually nothing was known of theropod anatomy, diversity or phylogeny, its not surprising that a confusing and very messy picture developed. Victorian scientists like Richard Owen, Harry Seeley and Richard Lydekker proposed new names for these fossils, synomised the names proposed by their colleagues, un-synonymised the proposed synonomisations of others, referred new specimens to certain taxa theyd named beforehand, and generally left a gargantuan mess that few people understand and even fewer care about.

What sorts of theropods do these scrappy specimens represent? Since finishing my 1999 MPhil work on this particular set of dinosaurs, Ive made a cottage industry of interpreting and reinterpreting them, variously suggesting the presence of abelisauroids, indeterminate tetanurans, non-carcharodontosaurian allosauroids, non-neovenatorid carcharodontosaurians, compsognathids, tyrannosauroids, oviraptorosaurs, dromaeosaurids, troodontids and birds (Hutt et al. 2001, Naish et al. 2001, Naish & Martill 2002, 2007, Naish 2002a, 2003, 2011, Naish & Sweetman 2011). Of course, Im hardly the only author to have worked on, or expressed opinions on, these fossils (e.g., Harrison & Walker 1973, Norman 1990, Howse & Milner 1993, Norell & Makovicky 1997). Anyway, some of these proposed identifications are almost certainly wrong, and I increasingly think that they serve as cautionary tales with respect to how confident we should be when interpreting scrappy fossils (and... cough Samrukia cough).

The 'Holy-Grail-o-saurus Hypothesis', with Eotyrannus as the template. It has failed: those different Wealden theropods really are from different Wealden theropods. This illustration is a slide from a talk I gave in 2013.

One of my great hopes for these scrappy Wealden theropods is that some, most or even all of them might prove to be different parts of the same one animal, a sort of Holy-Grail-o-saurus that combines the foot named Valdoraptor, the neck vertebrae named Calamosaurus and Thecocoelurus, the sacrum named Ornithodesmus, and so on. It was hoped that Eotyrannus, discovered in 1997, preliminary described in 2001 (Hutt et al. 2001), and described in full in a soon-to-be-published monograph*, would be that Holy Grail but, nope, it proved to be a wholly new taxon, definitely distinct from the other Wealden taxa. Then theres That Which Cannot Be Named, another associated Wessex Formation coelurosaur, currently in private hands and unavailable for study. Contra expectations, it, again, doesnt tie together any of the named bits and pieces, at least not so far as I can tell from my two brief examinations of it.

More proof that the Eotyrannus monograph really exists. This montage shows some (not all) of the illustrations included in the final work (Naish & Cau, submitted).

* People who follow me on twitter (@TetZoo), or who follow research on tyrannosauroids or Wealden theropods specifically, will know that (with co-author Andrea Cau) I finished the Eotyrannus monograph some considerable time ago. Its since sat in limbo. However, things are progressing and it will appear in print at some stage.

The Angeac ornithomimosaur

Jean-Franois Tournepiche and the gigantic Angeac sauropod femur reported by Nraudeau et al. (2012). Image from ouest-paleo.net.

Finally, we come to the reason for this article. Those of you who keep up with the happening world of Lower Cretaceous European coelurosaurs will know that Ronan Allain and colleagues have recently discovered a theropod bonebed in the Hauterivian or Barremian strata of Angeac, in Charente, southwestern France. The site, already known for the remains of a gargantuan sauropod and assorted other Cretaceous vertebrates (Nraudeau et al. 2012), has yielded more than 850 bones, representing at least 20 individuals, of the same coelurosaurian theropod taxon [adjacent image from this article]. A long paper describing this animal in detail is, so I understand, somewhere in the works. Meanwhile, a new publication has just appeared that presents a preliminary investigation of this animal and its implications (Allain et al. 2014). Ive known about this work for a while and am very pleased to see it out.

What sort of coelurosaur is this animal? A toothless, anteriorly downturned dentary, ventrally flattened pedal unguals and other characters show that its an ornithomimosaur, an ostrich dinosaur (Allain et al. 2014). It doesnt yet have a taxonomic name, so for now were just going to have to call it the Angeac ornithomimosaur.

An illustration from 2002, depicting (at top) the smaller theropods then thought to be known from the Wessex Formation and - at bottom - those suggested to perhaps be present and awaiting discovery. The ornithomimosaur is based on the Spanish Pelecanimimus. Image from Naish (2002b). It was 2002, so the animals are not shown as feathery enough (except the bird).

So, we have a definite ornithomimosaur in the Hauterivian or Barremian of France. Ornithomimosaurs are already known from the Lower Cretaceous of western Europe, as since 1994 weve been aware of the remarkable Pelecanimimus polyodon from the Barremian Calizas de la Hurguina Formation of Las Hoyas, Spain. The Calizas de la Hurguina Formation seemingly includes at least a few taxa that are also present in the similarly-aged Wessex Formation of England (including Baryonyx), so for this reason I once suggested that Pelecanimimus or a similar taxon might await discovery in the Wessex Formation too (Naish 2002b, Naish et al. 2001). That prediction hasnt been borne out (yet), but the raison detre of the new paper is to show that several Wessex Formation theropod elements, and specimens from elsewhere in the Wealden succession, also represent ornithomimosaurs, and, furthermore, ornithomimosaurs that are just about identical with the Angeac ornithomimosaur (Allain et al. 2014). Note, incidentally, that the Angeac ornithomimosaur is a toothless ornithomimosaur, and thus not a close relative of Pelecanimimus.

Is the hypothesis of an oviraptorosaurian Thecocoelurus completely dead? Maybe, maybe not. Calamosaurus and a Wessex stegosaur play bit-parts. Image by Darren Naish, colouring by Patrick Murphy. The illustration is from 2001, hence the sparse feathering.

To reiterate, the reason that this French discovery remember, it is not a Wealden dinosaur is a big deal as goes Wealden dinosaurs is that Allain et al. (2014) argue that several Wealden theropod specimens are so similar to the Angeac ornithomimosaur that they very probably represent the same sort of dinosaur. In fact, Allain et al. (2014) argue that a whole load of previously problematic Wealden theropod remains all belong to this new kind of ornithomimosaur. Valdoraptor (named for a metatarsus from the Hastings Group) is an ornithomimosaur very similar to the Angeac taxon, so is Thecocoelurus (named for a cervical vertebrae from the Wessex Formation), and so is the famous Calamosaurus tibia and several other Wessex Formation hindlimb elements (Allain et al. 2014).

The important message is that these fossils were all previously interpreted as belonging to very different sorts of theropods. In the most recent review of Wealden theropods, I have Thecocoelurus as a possible abelisauroid, Valdoraptor as an indeterminate neotetanuran, and the Calamosaurus tibia as a possible tyrannosauroid or compsognathid (Naish 2011).

Problems from the Wealden 1: Thecocoelurus

Thecocoelurus compared with a cervical vertebra of Falcarius: an image prepared for a manuscript I ended up abandoning. aas = anterior articular surface; apf = anterior peduncular fossa; ilf = interspinous ligament fossa; nc = neural canal; vr = ventrolateral ridge. Scale bar = 20 mm. UPDATE: this figure might be misleading, in that the Thecocoelurus image A is of the anterior end, but I think the Falcarius image B is of the posterior end!

If this is right, then the Angeac ornithomimosaur really is the Holy-Grail-o-saurus of the sort mentioned above... though note that there are still other Wealden theropods that arent part of this story and so far as we can tell do still represent other clades: Becklespinax, Calamosaurus, Aristosuchus, the Ashdown maniraptoran, Wyleyia and Ornithodesmus, for example.

But... is it right? Allain et al. (2014) regard Thecocoelurus (remember: based on half of a single cervical vertebra) as morphologically identical to vertebrae of the Angeac ornithomimosaur. I was initially happy to take their word for this, but a detailed look at their figures shows that, while similar, its difficult to be convinced that the vertebrae really are identical. The problem is that the cervical vertebrae of quite a few distinct theropod groups are highly similar, similar enough that you can be semi-convinced by general comparisons, but less satisfied when all the anatomical minutiae are compared in detail. Hence Thecocoelurus has found itself purportedly sharing specific characters with caenagnathid oviraptorosaurs (Naish & Martill 2002), Falcarius the therizinosaur (Kirkland et al. 2004), noasaurid abelisauroids (Naish 2011), and now with ornithomimosaurs. Mickey Mortimer compared the cervical vertebrae of these groups in his blog article here. The takehome?...

Composite produced by Mickey Mortimer, and showing Thecocoelurus compared with the cervical vertebrae of select other theropods (ornithomimosaurs, therizinosaurs, noasaurids and oviraptorosaurs). Anterior half of posterior cervical vertebrae in (descending) right lateral, anterior, ventral and dorsal views. From left to right: Thecocoelurus holotype (after Naish & Martill 2002); Angeac taxon ANG 10/175 (after Allain et al. 2014); Falcarius UMNH VP 14657 (after Zanno 2010b); Masiakasaurus FMNH PR 2481 (after Carrano et al. 2011); Chirostenotes or Epichirostenotes ROM 43250 (after Sues 1997); Similicaudipteryx holotype (after He et al. 2008).

Yes, Thecocoelurus is similar to the cervical vertebrae of the Angeac ornithomimosaur, but it isnt definitely more similar to them than it is to the vertebrae of caenagnathids and so on, and in fact Thecocoelurus shares more characters with caenagnathids and some other taxa than it does with the Angeac ornithomimosaur. Its a frustrating situation, and again Id repeat the general idea that we may be expecting too much of a single fragmentary vertebra.

Problems from the Wealden 2: Valdoraptor

Im more impressed by Allain et al.s (2014) argument that Valdoraptor based only on a partial metatarsus and originally confused with the ankylosaur Hylaeosaurus and then regarded as a species of Megalosaurus is also highly similar to the Angeac ornithomimosaur, and I agree that Valdoraptor should now be reinterpreted as a member of this group. Thats an interesting turn of events for an animal long imagined as a megalosaur, later suggested to be an allosauroid, and universally considered to be a non-coelurosaur.

Features of the Valdoraptor metatarsus suggested (by Naish 2011) to be diagnostic for this taxon. Photo (c) Natural History Museum, London.

Allain et al. (2014) note that two of three diagnostic characters I suggested for Valdoraptor (Naish 2011) are now invalid, since the characters concerned (a medially projecting distal end of the second metatarsal and a metatarsal III that projects notably further distally than the other metatarsals) are present in other ornithomimosaurs and other coelurosaurs, like tyrannosaurs. Because at that time I had Valdoraptor pegged as a non-coelurosaurian tetanuran (Naish 2011), I wasnt too worried about the form of the metatarsus among coelurosaurs (how you interpret characters is always partially dependent on where you think your taxon fits in the tree of life: a feature thats unique to a taxon in one part of the tree may be a commonplace and ubiquitous character elsewhere).

The Valdoraptor foot (from Owen's original monograph of 1858, where it's shown as a mirror image relative to the real thing) compared with an ornithomimid ornithomimosaur foot (image by Darren Naish, after Osborn).

However, Allain et al. (2014) agree that one of Valdoraptors supposedly diagnostic characters (its mediolaterally compressed metatarsal II) is, among coelurosaurs, uniquely shared by Valdoraptor and the Angeac ornithomimosaur. Heres where we come to the part where I find myself disagreeing with what the authors have done in their paper.

Youll recall from above that the Angeac ornithomimosaur is not named, the strong implication being that it represents a wholly new taxon thatll be named elsewhere. Yet, arent the authors on dangerous ground if theyre also saying that Thecocoelurus and Valdoraptor are extremely similar (if not identical) to the Angeac ornithomimosaur? What I mean is: if the Angeac ornithomimosaur shares diagnostic characters with Thecocoelurus or Valdoraptor, and if those taxa are also ornithomimosaurs (as Allain et al. argue they are), then by definition the Angeac ornithomimosaur has to be assumed to be synonymous with either or both of those taxa. Weve already seen above that Thecocoelurus is not obviously closer to the Angeac ornithomimosaur than it is to other theropods, so lets ignore it for now.

Top left: an ornithomimid. Top right: Thecocoelurus montage (from Naish & Martill 2002). Bottom: noasaurid Masiakasaurus. Ornithomimid by Sebastian Bergmann, licensed under Creative Commons Attribution-Share Alike 2.0 Generic license; Masiakasaurus by Esv, licensed under Creative Commons Attribution-Share Alike 3.0 Unported license.

But Valdoraptor? Allain et al. (2014) say that Valdoraptor and the Angeac ornithomimosaur both share a character previously mooted (Naish 2011) as diagnostic for Valdoraptor (this being the compressed form of metatarsal II). But they go on to say that, since this character is also present in other theropods (specifically, noasaurids), it cant be used to diagnose Valdoraptor. Err, nope.

Recall what I said above about the need to evaluate characters on the basis of where your taxon fits in the tree of life? Allain et al. (2014) are stating that Valdoraptor is an ornithomimosaur: ergo, its mediolaterally compressed metatarsal II has to be evaluated within the context of how this character is distributed within Ornithomimosauria and those lineages closest to it. Noasaurids are nowhere close to ornithomimosaurs in the theropod tree. Theyre so distant that any characters shared by noasaurids and ornithomimosaurs (and not present across the intervening nodes of Theropoda) have to be interpreted as obvious convergences [ornithomimid image above by Sebastian Bergmann; Masiakasaurus by Esv]. In other words, if a character present in Valdoraptor is not present in other ornithomimosaurs, and is not primitive for the clade that includes ornithomimosaurs and other coelurosaurs, it has to be interpreted as a unique, diagnostic character for Valdoraptor. So, if the Angeac ornithomimosaur possesses a character thats otherwise diagnostic for Valdoraptor, then Angeac ornithomimosaur = Valdoraptor.

Lets see what happens next.

For previous Tet Zoo articles on theropods and other dinosaurs and animals from the Wealden see...

Refs - -

Allain, R., Vullo, R., Le Loeuff, J. & Tournepiche, J.-F., 2014. European ornithomimosaurs (Dinosauria, Theropoda): an undetected record. Geologica Acta 12 (2), in press.

Batten, D. J. (ed.) 2011. English Wealden Fossils. The Palaeontological Association, London.

Harrison, C. J. O. & Walker, C. A. 1973. Wyleyia: a new bird humerus from the Lower Cretaceous of England. Palaeontology 16, 721-8.

Howse, S. C. B. & Milner, A. R. 1993. Ornithodesmus - a maniraptoran theropod dinosaur from the Lower Cretaceous of the Isle of Wight, England. Palaeontology 36, 425-37.

Hutt, S., Naish, D., Martill, D. M., Barker, M. J. & Newbery, P. 2001. A preliminary account of a new tyrannosauroid theropod from the Wessex Formation (Early Cretaceous) of southern England. Cretaceous Research 22, 227-242.

Kirkland, J., Zanno, L., Deblieux, D., Smith, D. & Sampson, S. 2004. A new, basal-most therizinosauroid (Theropoda: Maniraptora) from Utah demonstrates a Pan-Laurasian distribution for Early Cretaceous (Barremian) therizinosauroids. Journal of Vertebrate Paleontology 24 (supplement to 3), 78A.

Naish, D. 2002a. The historical taxonomy of the Lower Cretaceous theropods (Dinosauria) Calamospondylus and Aristosuchus from the Isle of Wight. Proceedings of the Geologists Association 113, 153-163.

- . 2002b. Thecocoelurians, calamosaurs and Europes largest sauropod: the latest on the Isle of Wights dinosaurs. Dino Press 7, 85-95.

- . 2003. A definitive allosauroid (Dinosauria; Theropoda) from the Lower Cretaceous of East Sussex. Proceedings of the Geologists Association 114, 319-326.

- . 2011. Theropod dinosaurs. In Batten, D. J. (ed.) English Wealden Fossils. The Palaeontological Association (London), pp. 526-559.

- . & Martill, D. M. 2001. Saurischian dinosaurs 1: Sauropods. In Martill, D. M. & Naish, D. (eds) Dinosaurs of the Isle of Wight. The Palaeontological Association (London), pp. 185-241.

- . & Martill, D. M. 2002. A reappraisal of Thecocoelurus daviesi (Dinosauria: Theropoda) from the Early Cretaceous of the Isle of Wight. Proceedings of the Geologists Association 113, 23-30.

- . & Martill, D. M. 2007. Dinosaurs of Great Britain and the role of the Geological Society of London in their discovery: basal Dinosauria and Saurischia. Journal of the Geological Society, London 164, 493-510.

- . & 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.

Nraudeau, D., Allain, R., Ballvre, M., Batten, D.J., Buffetaut, E., Colin, J.-P., Dabard, M.-P., Daviero-Gomez, V., El Albani, A., Gomez, B., Grosheny, D., Le Loeuff, J., Leprince, A., Martn-Closas, C., Masure, E., Mazin, J.-M., Philippe, M., Pouech, J., Tong, H., Tournepiche, J.-F., Vullo, R., 2012. The HauterivianBarremian lignitic bone bed of Angeac (Charente, SW France): stratigraphical, palaeobiological and palaeogeographical implications. Cretaceous Research 37, 1-14.

Norman, D. B. 1990. Problematic Theropoda: coelurosaurs. In Weishampel, D. B., Dodson, P. & Osmlska, H. (eds) The Dinosauria. University of California Press (Berkeley), pp. 280-305.

Norell, M. A. & Makovicky, P. J. 1997. Important features of the dromaeosaur skeleton: information from a new specimen. American Museum Novitates 3215, 1-28.