In the shadow of the Isle of Wight dinosaurs

Dr Steven C Sweetman (UK)

Bones from the Early Cretaceous Wessex Formation of the Isle of Wight in southern England have been at the forefront of dinosaur research since before the term ‘Dinosauria’ was invented the following is a summarhy of the significant events in terms of dinosaurs and the Isle of Wight.

1824 – William Buckland is the first to scientifically describe and name what was later to be known as a dinosaur: Megalosaurus from the Jurassic of Oxfordshire.
1825 – Gideon Mantell describes and names Iguanodon from the Wealden Supergroup of mainland Britain.
1829 – Buckland provides the first description of a dinosaur bone from the Isle of Wight: a pedal phalanx of Iguanodon found at Yaverland (Figs. 1).
1841 – Richard Owen coins the name Dinosauria meaning, in his translation, ‘fearfully great lizards’ and publishes the name in 1842. He visits the Isle of Wight shortly afterwards and maintains correspondence with local collector, Rev William Fox.

Fig. 1. OUM K859. A pedal phalanx of *Iguanodon* sp. The first dinosaur bone from the Isle of Wight to be described and figured.

The island is Europe’s richest source of dinosaur fossils, although most species are known only from fragmentary remains. New discoveries continue to be made despite the very small outcrop area (Fig. 2), largely due to the fact that the cliffs in which the Wessex Formation is exposed (Figs. 2 to 6) are subject to continual erosion.

Fig. 2. Outline geological and location maps of the Isle of Wight. Dimensions: E-W 37km; N-S 21.5km. Total length of Wealden Group cliff exposures: about 1km, south-east coast; about 10km, south-west coast.

Wessex Formation strata are the oldest exposed on the island and are of Barremian age. This makes the vertebrate fossils obtained from them of particular significance, because few sites of similar age (about 125 to 130 million years ago), which also yield terrestrial vertebrate remains, are known elsewhere in the world.

Fig. 3. Wealden Group (foreground) and middle and upper Cretaceous strata (middle and far distance) exposed in a cliff section north-east of Sandown on the south-east coast of the Isle of Wight.

Based on studies of macro skeletal remains collected from the Wessex Formation over the past 180 years or so, some 22 dinosaur species are currently considered to be valid, occurring together with one pterosaur and perhaps as many as 12 crocodiles and four turtles (the crocodiles and turtles are in need of review). In the 1960s and 70s, there was an explosion of interest in Mesozoic mammals and this prompted a number of private collectors to sieve samples from horizons in the Wessex Formation and the overlying Vectis Formation in the hope of finding mammal teeth.

Fig. 4. Cliff exposures on the south-west coast, showing the Vectis and Wessex Formations at Barnes High.

Of these, only one was successful. He recovered two multituberculate teeth from the Wessex Formation in Compton Bay. He also found previously unrecorded, isolated teeth of the crocodilian Bernissartia, adding another to the list.

Fig. 5. The Wealden Group seen from the cliff top at Barnes High looking south-east. This remarkable section has produced a wealth of dinosaur fossils, including partial skeletons of a brachiosaurid sauropod, iguanodontids and of the ankylosaur *Polacanthus*. It has also produced numerous fragmentary remains of other dinosaurs, crocodiles and fishes, and contains the famous *Hypsilophodon* Bed.

A while ago, scientists involved with the BBC television series, Live From Dinosaur Island, (during which there were, first and foremost, looking for the small dinosaur, Hypsilophodon; Fig. 6) sieved small samples from the excavation sites for dinosaur bones hoping to find the remains of small animals that must surely have lived alongside the giants of their day. Results were disappointing, but they did succeed in finding very fragmentary and indeterminate remains of two lizards, an albanerpetontid, a frog and a salamander. Adding these to the list of previously recorded tetrapods (four legged animals) recovered from the Wessex Formation gave a total of about 46 species in all.

Fig. 6. The *Hypsilophodon* Bed (red bed adjacent to scale bar), exposed near Cowleaze Chine with the White Rock and Vectis Formation above.

The apparent rarity of small vertebrate remains in the Wessex Formation was puzzling and suggested to some workers that the environment of deposition was unsuitable for their preservation. The bulk of the succession comprises mudstones. These were subject to prolonged periods of wetting and drying, weathering, and soil-forming processes prior to final burial, and these conditions are certainly not favourable for the preservation of bones, large or small.

Where large bones are encountered in the mudstones, they are often demineralised and fragile, suggesting that small vertebrate remains would not be preserved. To see if this was the case, I took samples from the mudstones and sieved them using the techniques outlined in Deposits (Sieving out the big picture). Results were much as expected with no microvertebrate remains being recovered from the majority of samples. However, small bones and teeth were found in matrix associated with some large dinosaur bones excavated from one of the mudstones. Here, it appears that the large bones acted as a physical and chemical shield, protecting the small vertebrate remains in their immediate vicinity.

In contrast to the preservation seen in the mudstones, preservation of large bones in the so-called plant debris beds of the Wessex Formation is often very good and it is from these beds that the majority of dinosaur and other vertebrate remains have been recovered. These distinctive units (Fig. 7) occur randomly throughout the succession.

Fig. 7. A plant debris bed exposed in Compton Bay.

They are usually no more than a metre thick and are very variable in the details of their sedimentology. However, the majority comprise a thin basal conglomerate overlain by very poorly sorted plant remains often containing logs (Fig. 8) in a fine-grained matrix.

Fig. 8. A log from one of the plant debris beds exposed at Yaverland.

The constituents of all beds either fine upward to become indistinguishable from the overlying mudstone or to a thin zone representing either a zone of oxidation or a soil horizon. Most of the plant debris beds also contain substantial amounts of fossil charcoal. While the origin of the plant debris beds remains the subject of some debate, it appears that they represent debris flows generated by flood waters flowing over a landscape denuded of vegetation following a wildfire ignited by lightning strike.

The occurrence in the same plant debris beds of remains of animals derived from both aquatic and terrestrial environments appears to support this hypothesis. Their good state of preservation relates to anoxic conditions within the beds after deposition caused by the decay of the large amounts of plant material contained within them. These conditions were also favourable for the formation of pyrite and most plant debris beds contain large amounts of this mineral. It occurs as both discrete nodules and as infill of pore spaces in plants and bones, and as encrustations on them (Fig. 9).

Fig. 9. Plant remains from the Wessex Formation, some with typical pyrite encrustation.

The good preservation of large bones and some of the plant material in the plant debris beds, coupled with the very poor sorting of plant and animal remains in them, suggested that well-preserved small vertebrate remains should also be present. However, if that was the case, why had they not been found before? Four years of extensive sampling eventually provided an answer – the sieves used by workers in the 1970s had a mesh size too large to retain most microvertebrate remains present in the Wessex Formation and the samples taken in 2001 by scientists working on the BBC television series were simply not large enough to ensure that determinate remains were recovered. In light of discoveries made in 2002, when comprehensive sampling of the productive horizon sampled during Live From Dinosaur Island was undertaken, they could have struck lucky, but did not. However, that is absolutely typical of fossil hunting in the Wessex Formation, whatever technique is adopted.

So, what has been found so far? In short, results have far exceeded expectations. At least as many new species were found in four years using the techniques outlined in the earlier article referred to above, as had been found in the previous 180 years of surface collecting. The key was the ability to process large amounts of matrix (getting on for four tonnes now), using a sieve with a mesh size small enough to retain the fossils and having the patience to pick through every grain of residues after sieving, even when they were very poor in fossils.

Selecting horizons to sample was easy – the plant debris beds are readily identifiable in the field. Selecting a place within each bed from which to take samples was less easy because it soon became apparent that fossil abundance at different exposures of the same bed was very variable and unpredictable. Also, microvertebrate remains are usually impossible to see in the field being, by definition, very small and, in most cases, indistinguishable from the plant material with which they are associated.

This is where the pot luck approach previously described came in – if the lithology looked right, samples were taken at random in the hope that they would produce fossils. It was also necessary to sample as many beds as possible (27 in total), despite the often arduous task of getting samples from the point of collection to the processing site (Fig. 10). Samples weighing over 50kg often had to be carried long distances along shingle beaches and up steep cliffs to the nearest point of vehicular access, but it was worth it in the end.

Fig. 10. Sample collection from a bed below Barnes High. Each bucket weighs about 25kg and the pack contains another 10kg.

Looking at the dinosaurs and other large vertebrates in isolation provides a sketch of life on the Wessex Formation floodplain. However, it leaves many of the questions relating to their palaeoecology and, indeed, that of the ecosystem as a whole unanswered. Study of the smaller creatures that lived in their shadow paints a much broader and more colourful picture. It also facilitates a detailed understanding of community structure, both within animal groups such as the dinosaurs and as a whole.

So far as the dinosaurs are concerned, sampling for small vertebrates resulted in the recovery of remains of at least nine new taxa, excluding avian theropods of which more than one may be present. These include several small ornithischians (Fig. 11) and a number of theropods, including a velociraptorine dromaeosaur and a troodontid, all of which (with the possible exception of the troodontid) are currently unknown from macro skeletal remains.

Additional specimens of previously recorded taxa also shed new light on the tooth morphology of some of these (Fig. 12). The relative abundance of these remains is also interesting. Counterintuitively (one would expect more herbivores then carnivores to be present) but in common with a number of other localities, remains of theropods are both more abundant and diverse than those of ornithischians. The reasons for this are, as yet, unclear.

Teeth of two previously unrecorded crocodilians were also obtained (Figs. 13 and 14) the smaller of which occurs in most beds, often in abundance.

Fig. 13. A, unusual serrated tooth of a relatively large and uncommon crocodile of uncertain affinities; B, enlargement to show serrations.

Fig. 14. Tooth of a previously unrecorded small atoposaurid (a type of crocodile).

The remainder of the assemblage comprises:

Frogs (Fig. 15).

Salamanders (Fig. 16).

Fig. 16. Salamander atlas vertebrae in: A, rostral; B, caudal; C, dorsal; D, ventral views.

An albanerpetontid (Fig 17).

Fig. 17. Fused frontals of the Wessex Formation albanerpetontid in: (left) dorsal and (right) ventral view.
albanerpetontid in: left, dorsal and right, ventral view.

Probably the most diverse Early Cretaceous lizard assemblage yet discovered (Fig. 18).

Fig. 18. Lizard premaxillae from a plant debris bed at Yaverland.

Pterosaurs (Fig. 19).

Fig. 19. Teeth of an istiodactylid pterosaur.

Mammals (Fig. 20).

Fig. 12. Holotype partial dentary of a spalacolestine spalacotheriid mammal in: A, lingual and B, labial view. Scale bar represents 2mm.

A diverse fish assemblage, including both sharks and bony fishes, is also present.

Steven C Sweetman currently works at the School of the Environment, Geography and Geosciences, University of Portsmouth. Steven is a palaeontologist. His primaray research interest concerns Early Cretaceous microvertebrate remains. He is currently working on microvertebrate projects involving the the Purbeck Group of Dorset and the Wealden Supergroup of south-east England and the Isle of Wight and the Salema Formation of Portugal.