A Giant Step in Evolution: Discovering the Link Between Fish and Limbed Animals

Ted Daeschler (Academy of Natural Sciences, Philadelphia, Pennsylvania, USA)

Denison University, Granville, Ohio, USA

12 April 2007


All of vertebrate history has occurred since the Cambrian (from ~500 m.y. to now).


Devonian - the time of the highest diversity for jawless fish, placoderms, acanthodians, and lobe-finned fish.  The Devonian had a different assemblage of vertebrates compared to now.  Now, we’ve got sharks & their relatives, ray-finned fish, and tetrapods.  Tetrapods had just started in the Devonian.


Lobe-finned fish are represented today by lungfish & coelacanths.  The precursors to limbed animals are found in the lobe-finned fish.


Tetrapods were not evolving in a vacuum.  Consider vascular plants - many modern plant groups were just getting started in the Late Devonian.  The Early Devonian was the time of the ~first invasion of land by plants.


Look at the Catskill Formation (Devonian) of the Pocono Plateau & Allegheny Plateau of Pennsylvania - there are many old & new roadcuts.  Ex: just north of Williamsport, Pennsylvania.

Sauripterus taylori - a right pectoral fin with long dermal rods from the Catskill Formation of Lycoming County, Pennsylvania.  The fossil has internal bony structure - gave the fin strength & provided much more mobility than we’d expect in a lobe-finned fish.


Red Hill - a large Upper Devonian roadcut in Clinton County, Pennsylvania that has produced 1000s of fossils from ancient stream deposits.  The site also has paleosols (floodplain soils).  These rocks represent an ancient well-watered floodplain.  Lots of plant fossils and freshwater fish fossils (placoderms, acanthodians, lobe-finned fish, ray-finned fish) plus early tetrapods.  Also millipedes, arachnids, plants.

See May 1999 National Geographic for a Red Hill reconstruction.

Freshwater ecosystems are where the fish-tetrapod transition occurred.

See 2 April 2004 issue of Science - article about an important find from Red Hill, even though it’s only one element.

Looking at the lineage to tetrapods (lobe-finned fish to tetrapods):






            Rhizodont fish


There’s still a morphological gap between Panderichthys and Acanthostega.

Saw Dott & Batten’s historical geology book Evolution of the Earth, 2nd Edition & noticed a map of North America showing the Devonian redbeds of the Catskill of Pennsylvania & Devonian redbeds in Greenland & Devonian redbeds in the Canadian Arctic Islands.

Well, Pennsylvania and Greenland were already well studied.  But, the Canadian Arctic islands hadn’t been explored for fossils.  Daeschler went to the Canadian Arctic in 1999 (then just established as Nunavut).  Went to Ellesmere Island & islands west of there (Mellville Island, Bathurst Island).  Returned in 2000, 2002, 2004, and 2006.

The rocks there are part of a clastic wedge - there were mountains to the east - meandering stream, alluvial plains - similar environments then as in Pennsylvania & Ohio, but the Canadian Arctic sediments are ~5 million years older.

So, looked at the Middle-Upper Devonian of Arctic Canada.

Visited areas on Ellesmere Island (in the east) to areas on Banks Island (to the west).

The sections have 8000’ worth of clastic sediments (Frasnian).

Southeastern Melville Island, Nunavut - good exposures in general; no vegetation, but some places have deeply weathered rocks.  There are polar bears in the area.  The Sun is up 24 hours a day.

Fram Formation (Frasnian) - 1800 meters thick, 2-3 m.y. worth of time.  Alluvial, channels, floodplain sediments.  Found good material in the Fram Formation in an unnamed valley near Bird Fiord, southern Ellesmere Island, including lungfish plate fragments.  Did excavation and found more material.  Found skulls, fish bodies, and bodies with fins.

In 2002, got lots of Laccognathus, a large lobe-finned fish.  Laccognathus is not close to the origin of tetrapods, though.  Also got an Elpistostege palette.  Elpistostege is a flat-headed fish with orbits placed high.  It is high on the lobe-finned fish tree, close to tetrapods.

The elpistostegians are known from the Gaspe and from Latvia - Elpistostege + Panderichthys (close to early tetrapods, but not really well preserved).  So, they are from the Euramerican continent.  Canadian Arctic samples are consistent with this distribution.

Revisited in 2004 - found more Laccognathus & Elpistostege.  Also got Eusthenopteron and placoderms and lobe-finned fish and lungfish.  Also found a new elpistostegian that we named Tiktaalik roseae.

Found an articulated specimen - saw snout and a lower jaw in the field of what would become Tiktaalik.  Collected a block of rock with some of the Tiktaalik animal in it & another block with Tiktaalik material.  Did excavation back in the lab - saw orbits raised up.   Saw the back end of the skull.  Got a ~complete elpistostegian skull - different from known elpistostegians.  Got a shoulder girdle, fins, scales.  Fred Mullison is the fossil prepator of this material - he saw this stuff first during excavation in the lab.

Got a skull articulated with a post-cranial skeleton.  Got three skulls in total, each associated with shoulder girdles and fins.  This is now called Tiktaalik.

Tiktaalik is quite different from PanderichthysTiktaalik has a shortened back of skull and deep notches at the back of the skull (like early tetrapods).  The genus Tiktaalik was named by Inuit residents - it’s their word for a freshwater fish.

One skull shows gill supports (brachial supports).  Got pectoral fin material with humerus + ulna/radius.  Also got complete fins.  Are they fins or limbs?

The pectoral fin of Tiktaalik is like that of lobe-finned fish, but it has features of tetrapods as well.

The animal could extend its wrist region, but couldn’t flex it far.  Fish can’t do this (bending at the wrist zone).

Muscle attachments allow us to determine that it could flex at the wrist area.

Trunk of Tiktaalik - under the scales are flattened ribs - this is seen in early tetrapods.  Flattened ribs provide support for respiration.

Tiktaalik had two respiration systems - had lungs & gills (like other lobe-finned fish).



  no neck







  neck - can turn head without turning whole body, as fish have to

  no scales



Tiktaalik has a mixture of fish & tetrapod features.  Under the fish scales are tetrapod ribs.  It has a neck, has scales, has fins, a flat head, and dorsal eyes.  It has a primitive jaw.

Tiktaalik is a mosaic form.  This is a good transitional form.  Couldn’t predict this exact mosaic beforehand, though.  Tiktaalik fits in the gap we mentioned before (the Panderichthys to Acanthostega gap).

Every fossil is a transitional fossil.  But Tiktaalik was a transition between major groups.



Home page