Ichthyostega and Acanthostega: secondarily more aquatic

More heresy here
as the large reptile tree (LRT, 1036 taxa) flips the traditional order of fins-to-feet upside down. Traditionally the late Devonian Ichthyostega and Acanthostega, bridge the gap between lobe-fin sarcopterygians, like Osteolepis.

In the LRT
Acanthostega, ‘the fish with limbs’, nests at a more derived node than its precursor, the more fully limbed, Ossinodus (Fig. 1). Evidently neotony, the retention of juvenile traits into adulthood, was the driving force behind the derived appearance of Acanthostega, with its smaller size, stunted limbs, smaller skull, longer more flexible torso and longer fin tail.

Likewise
Ichthyostega is more derived than both fully-limbed Ossinodus and Pederpes, which had five toes. As in Acanthostega, the return to water added digits to the pes of Ichthyostega. In both taxa the interosseus space between the tibia and fibula filled in to produce a less flexible crus.

So, Acanthostega and Ichthyostega were not STEM tetrapods.
Instead, they were both firmly nested within the clade Tetrapoda. Ossinodus lies at the base of the Tetrapoda. The proximal outgroups are similarly flattened Panderichthys and Tiktaalik. The extra digits displayed by Acanthostega and Ichthyostega may or may not tell us what happened in the transition from fins to feet. We need to find a derived Tiktaalik with fingers and toes.

In cases like these
it’s good to remember that ontogeny recapitulates phylogeny. Today and generally young amphibians are more fish-like (with gills and fins) than older amphibians.

It’s also good to remember
that the return to the water happened many times in the evolution of tetrapods. There’s nothing that strange about it. Also the first Devonian footprints precede the Late Devonian by tens of millions of years.

Phylogenetic analysis teaches us things
you can’t see just by looking at the bones of an individual specimen. A cladogram is a powerful tool. The LRT is the basis for many of the heretical claims made here. You don’t have to trust these results. Anyone can duplicate this experiment to find out for themselves. Taxon exclusion is still the number one problem that is largely solved by the LRT.

You might remember
earlier the cylindrical and very fish-like Colosteus and Pholidogaster convergently produced limbs independently of flattened Ossinodus, here the most primitive taxon with limbs that are retained by every living tetrapod. By contrast, the Colosteus/Pholidogaster experiment did not survive into the Permian.

References
Ahlberg PE, Clack JA and Blom H 2005. The axial skeleton of the Devonian trtrapod Ichthyostega. Nature 437(1): 137-140.
Clack JA 2002. Gaining Ground: The origin and evolution of tetrapods. Indiana University Press.
Clack JA 2002. An early tetrapod from ‘Romer’s Gap’. Nature. 418 (6893): 72–76. doi:10.1038/nature00824
Clack JA 2006. The emergence of early tetrapods. Palaeogeography Palaeoclimatology Palaeoecology. 232: 167–189.
Jarvik E 1952. On the fish-like tail in the ichtyhyostegid stegocephalians. Meddelelser om Grønland 114: 1–90.
Jarvik E 1996. The Devonian tetrapod Ichthyostega. Fossils and Strata. 40:1-213.
Säve-Söderbergh G 1932. Preliminary notes on Devonian stegocephalians from East Greenland. Meddelelser øm Grönland 94: 1-211.
Warren A and Turner S 2004. The first stem tetrapod from the Lower Carboniferous of Gondwana. Palaeontology 47(1):151-184.
Warren A 2007. New data on Ossinodus pueri, a stem tetrapod from the Early Carboniferous of Australia. Journal of Vertebrate Paleontology 27(4):850-862.

wiki/Ichthyostega
wiki/Acanthostega
wiki/Ossinodus
wiki/Pederpes