The original 'four-legged' fish: the first Devonian tetrapod to be found

If a non-palaeontologist has heard of a Devonian tetrapod, it will probably be Ichthyostega, the 'four-legged fish'. Our 1987 expedition had recovered some excellent material from Stensiö Bjerg of this animal as well as Acanthostega, which revealed its curious complement of seven toes on the hind limb.

However, many questions remain unanswered about its anatomy, body form, and environment. One of the main objectives of the 1998 expedition was to collect material that would try to answer some of these questions. For example, three papers have placed Ichthyostega and Acanthostega in their stratigraphical and sedimentological context (Blom et al. 2005, 2007; Larsen et al. 2009). Others have concentrated on the anatomy of Ichthyostega.

Skeletal Anatomy

Although several reconstructions had been made of whole animal, of both its skeleton and the 'living' animal, in fact, there were many problems with them. For one thing, most of the articulated fossils showed either just a head and thorax or a tail, pelvis and hind limb. None really showed how those two halves meshed together. The 1998 material included some that did. In 2000, then, we received the first of our grants from the Natural Environment Research Council to work up this material alongside the previously collected specimens. Major discoveries resulted from these.

That study concentrated on the skeletal anatomy of Ichthyostega. Using mainly the original material, we found that the axial skeleton showed many differences from previous reconstructions: the tail was shorter, the shoulder girdle was larger and there were fewer vertebrae in front of the sacrum. Most surprisingly, the pre-sacral column showed distinct regionization along its length, including neural spines whose orientation varied, and in the immediate presacral region, they alternated in size and shape between broad and narrow as seen in side view. We made an new reconstruction, published in 2005 (Ahlberg et al. 2005). However, this was only a two-dimensional reconstruction: it remained to be tested in three-dimensions.

In 2009, a new study also funded by the NERC, aimed to resolve some of the puzzles set by the early tetrapods, especially Ichthyostega. Using the latest computer techniques, in collaboration with the Royal Veterinary College, London, (see Stephanie Pearce and John Hutchinson in the Collaborators section), we have built a realistic virtual 3-D model of the skeleton of Ichthyostega, to see how the bits really did fit together. Thus we corroborated most of the findings of the 2-D model from 2005, with some subtle differences.

We were then able to study the potential mobility of the axial skeleton and limbs of both Acanthostega and Ichthyostega, and compared these with analagous modern tetrapods. Our work on the range of motion of the limbs of Ichthyostega was published in 2012. We showed that Ichthyostega did not 'walk' in a conventional sense, but probably used its forelimbs in a mud-skipper-like manner, with the hind limbs contributing stability but not power to the movement. The hind limbs were more likely to have been used in swimming.

We have also discovered that its vertebrae are not formed in the way expected of an early tetrapod, but the centra were made of open-topped rings composed of a ventral intercentrum, combined with anterordorsally sutured or fused pleurocentra. This is unusual, and is the reverse positioning of the inter-and pleuro-centra from what was considered normal for tetrapods. We also found that the ribs were not bicipital, contrary to expectations, but had elongate articulations with the transverse processes. Such a structure rules out their use in breathing motions. See Professor John Hutchinson's RVC web page about it.

Jenny and Stephanie were interviewed by Richard Hollingham for the NERC website Planet Earth Online and the result broadcast recently in their podcast at http://planetearth.nerc.ac.uk/multimedia/story.aspx?id=1369&cookieConsent=A.

We have also looked at Pederpes in our study, and our colleague Julia Molnar has been working on this.

Braincase and Ear Region

Clack et al. (2003) showed that its braincase and ear region were unique, and were probably adapted for underwater audition. Those parts of its anatomy had previously been difficult to interpret, and earlier studies by Jarvik had effectively opted out of trying to do so. We used microCT scanning of our new material as well as fresh preparation of original material to come to our conclusions. We also discovered that, contrary to previous observations, Ichthyostega possessed ossified and grooved gill-bars, like Acanthostega.

The Humerus

Our 1998 material included sub-adult Ichthyostega specimens, whose humeri showed distinct differences from those of the adult material from the original collections. Some of these differences suggested that the sub-adults retained more primitive and fish-like characters than the adults. The equivalent stages in Acanthostega humeri did not show these differences. Secondly, both sub-adult and adult humeri of Ichthyostega lacked several derived characters that were shared by Acanthostega and other, Carboniferous, early tetrapods. These findings argued against the general and widely accepted idea that Acanthostega was the more primitive of the two taxa. Phylogenetic analysis was unable to resolve this conundrum unequivocally, and we concluded that there must have been a lot more diversity among Devonian tetrapods, and probably several ways of being a tetrapod with limbs during this period. This study was published in 2009 (Callier et al. 2009).

Last updated 5th November, 2013 by Rob Clack