A fish is a fish, right? They’re just a blur of aquatic beasties that most people distinguish by flavor, rather than morphology or descent. But fish are incredibly diverse, far more diverse than terrestrial vertebrates, and there are significant divisions within the group. Most people know of one big distinction, between the Chondrichthyes (fish with cartilaginous skeletons, like sharks and rays) and the Osteichthyes (fish with bony skeletons), but there’s another particularly interesting split within the Osteichthyes: the distinction between Sarcopterygians (the word means “fleshy fins”, and we call them lobe-finned fishes colloquially) and the Actinopterygians, the ray-finned fishes. The lobe-finned fishes most distinctive feature is the muscular and bony central core of their fins — extant forms are the coelacanth and lungfish. It is this lineage that led to us terrestrial tetrapods, but other than that successful invasion of the land, the sarcops were something of an aquatic failure, with only a few genera surviving. The ray-finned fishes, on the other hand, are a major success story, with more than 28,000 species today. To put that in context, there are only about 5,500 species of mammals.

The Sarcopterygii and the Actinopterygii must have begun diverging a long time ago, and a couple of questions of interest are a) when did the last common ancestor of both groups live, and b) what did it look like? We don’t have a good and specific answer yet, because Osteichthyes origins are lost far, far back in time, over 400 million years ago, but every new discovery edges us a little closer. What we now have is a well-preserved fossil of a fish that has been determined to be an early sarcopterygian, and it tells us that a) the last common ancestor had to have lived over 419 million years ago, the age of this fossil, and the divergence probably occurred deep in the Silurian, and b) this animal has a mosaic of primitive Osteichthyan features, which tells us that that last common ancestor may well have shared some of these elements. It is another transitional fossil that reveals much about the gradual separation of two great vertebrate groups.

And here it is:

(Click for larger image)
a, b, A near-complete fish in part and counterpart. c, Close-up view of the anterior portion of the trunk shield in dorsal view, showing MD1 and MD2 flanked by rhomboid scales. d, Close-up view of the dorsal fin spine. MD1, first median dorsal plate; MD2, second median dorsal plate. Scale bar, 1 cm.

That may be a bit disappointing at first — it looks like Silurian road-kill — but really, that’s a remarkable good and useful specimen. The animal was covered with thick bony scales, and the skull was built of thick bony plates, and so while it was squashed flat by pitiless geology, the pieces are all there, and it can be reassembled into a much more fishy state. This drawing may be more satisfying:

(Click for larger image)
a, Restoration of the entire fish in lateral view. b, Interpretive drawing of the holotype V15541. Areas shaded in grey are unknown, and are reconstructed from other early osteichthyans. ano, anterior nostril; br, branchiostegal ray; cla, clavicle; cle, cleithrum; drs, dorsal ridge scale; dsp, dorsal fin spine; et, extratemporal; eta, accessory extratemporal; f.add, adductor fossa; f.gl, glenoid fossa; gu, gular; ju, jugal; l.ext, lateral extrascapular; lj, lower jaw; m.ext, median extrascapular; mx, maxillary; n.sp., spiracular notch; op, opercular; pa, parietal shield; pcl, postcleithrum; pop, preopercular; ppa, postparietal shield; psc, presupracleithrum; pt, post-temporal; scl, supracleithrum; sop, subopercular; sp., pectoral spine; tr, lepidotrichia; vrs, ventral ridge scale.

Now it looks like a kind of armored, spiky salmon with a thick muscular body (and yes, I too wonder about flavor, and would like to taste a slab of that). It’s definitely not a salmon, though — the bony structure is a curious set of compromises where some features are distinctly sarcopterygian, some look like they belong on a primitive actinopterygian, and others are unique or show affinities to characters of ancient extinct fishes, like rhipidistians. This is very cool. What we see here are relics of an ancient common osteichthyan ancestor, which are being honed into the specific characteristics of the Sarcopterygii. The analysis of the totality of the animal’s features, though, place it more in the lobe-finned than the ray-finned clade. That places it on a branch of the line leading to us…a very, very old branch, making this your many-times-great grand uncle, or cousin only a few million times removed. Now my curiosity about a taste-test is making me feel mildly cannibalistic.

(Click for larger image)
The topology is the most parsimonious tree arising from a matrix of 23 taxa coded for 153 morphological characters (tree length = 292, consistency index = 0.572, retention index = 0.737, rescaled consistency index = 0.421). The numbers at nodes indicate bootstrap support (where the value is greater than 50%) and Bremer decay index (bottom and top numbers, respectively). Eif., Eifelian; Ems., Emsian; Fam., Famennian; Fras., Frasnian; Giv., Givetian; Gor., Gorstian; Loch., Lochkovian; Lud., Ludfordian; Prag., Pragian.

When you look at that diagram, what should jump out at you is all the diversity in the Devonian, the so-called Age of Fishes, and the paucity of representative fossils from the Silurian…which is exactly where all the interesting branch points in the fish family tree are located. Once again, paleontology is a predictive science, and this tells us where to look for the next batch of exciting and informative fossils.

Zhu M, Zhao W, Jia L, Lu J, Qiao T, Qu Q (2009) The oldest articulated osteichthyan reveals mosaic gnathostome characters. Nature 458:469-474.