Thursday, May 04, 2006

The "other end" of the fish-tetrapod transition

I guess one can't really talk about an evolution transition having 'ends', since one might even consider the origin of vertebrates as a whole to be part of the origin of tetrapods story. However, we focus a lot on the transformation from sarcopterygian (lobe-finned) fishes to the earliest tetrapods like Acanthostega and Ichthyostega. Lest we forget the origin of the sarcopterygians themselves. This is a murky and poorly understood part of vertebrate evolution where the fossils are limited to what we affectionately call "thumbnail skulls". These are usually braincases and associated snouts that aren't much bigger than, well, your thumbnail (and kind of shaped like them too).

In recent years, we have seen the publication of a number of really peculiar thumbnail skulls from China, nearly all of which belong to some kind of sarcopterygian. One of the defining features of the sarcopterygian skull is a joint that runs right through the braincase. The only living sarcopt to retain such a feature is the modern coelacanth. However, once upon a time in the Devonian, most sarcopterygians were built this way. We know these skulls are sarcopterygian because of this division of the skull and braincase, as well as the histology of the dermal bone. Sarcopterygians are unique in having had a system of pores and canals running through an enamel-like layer on the outside. This tissue matrix is called cosmine. One will often hear of "cosmine-covered osteolepiforms" or "cosmine covered sarcopterygians", to distinguish these forms from taxa that later lose cosmine (for instance, the loss of cosmine is a character that Tiktaalik shares with tetrapods).

But apart from these defining features of sarcopterygians, the thumbnail skulls from China also have a strange mélange of characters from all kinds of other fish groups: actinopterygian (ray-finned) fishes, sharks, placoderms, and acanthodians. For instance, it was recently noted that several of these taxa had the bony facet for the attachment of an eyestalk -- a cartilage rod found in some modern sharks that supports the eyeball. Similar facets have been seen in fossil shark braincases and later in ray-finned fish braincases. Psarolepis, for instance, was shown to have a cheek very similar to that of an actinopterygian and it even had a spine on a shoulder girdle that was associated with the skull. The actual identity of some of these parts, however, may be disputed because we still haven't got whole skeletons of these fishes.

Today in Nature a new piece is added to this puzzle: Meemannia eos. This new taxon is peculiar as far as sarcopts go. It has something sort of like cosmine, but somewhat more rudimentary. However, another interesting aspect is the absence of an intracranial joint, like the one described above.

This 405-Myr-old fish shows a mixture of basal actinopterygian and sarcopterygian features. a, b, Dorsal view of skull roof (a, holotype, V14536.1; b, V14536.2). e, f, Ventral view of posterior portion of the skull roof with incompletely preserved oto-occipital structures (e, V14536.4; f, illustrative drawing). c, d, g, h, Reconstruction of skull roof (c) compared with two actinopterygians, Dialipina (d) and Cheirolepis (g), and one sarcopterygian, Powichthys (h). Abbreviations: am.a, am.e, anterior and external ampullae;, cranial cavity;, supraotic cavity; Dsp, dermosphenotic; It, intertemporal; lc, otic portion of the main lateral line canal; lcc, lateral cranial canal; P, parietal; pdf, posterior dorsal fontanelle; pl.m, pl.p, middle and posterior pit-lines; Pp, postparietal; re.u, utricular recess; sac, sacculus; sca, anterior semicircular canal; soc, supraorbital canal; St, supratemporal; T, tabular; IX, exit of the ninth cranial nerve. Open arrow in c, g and h indicates the position of the orbit. Scale bar, 5 mm (a, b, e, f). From Zhu et al. 2006.

Immediately, one is faced with two options: was the joint lost or was it simply primtive? The answer appears to be the latter, and that Meemannia is the most primitive sarcopterygian known to date. The skull roofing bones of Meemannia are curiously similar to those of primitive ray-finned fishes. In fact, the authors note, the snout appears to have been loosely attached to the rest of the skull, a feature also seen in placoderms. The weight of the evidence appears to support the conclusion that Meemannia is a rather 'primitive' animal.

The figure is based on two most parsimonious trees that differ in the positions of Ligulalepis. Bremer support values are shown at nodes. Tree length 222, consistency index 0.6216, homoplasy index 0.3784, retention index 0.7807, rescaled consistency index 0.4853. See Supplementary Information for details. Insets compare the histological features of Meemannia (b) with those found in actinopterygians (Andreolepis, a) and crown-group sarcopterygians (Porolepis, c). From Zhu et al. 2006

Furthermore, I mentioned that it had cosmine or something very much like it. The authors note that what we see in Meemannia the cosmine structure is unique among sarcopterygians and offers some clues about its natural history. In previous decades, the pore-canals have been suggested to represent a system something like the ampullae of Lorenzini in sharks. This is a network of electrosensory organs that sharks have spread over their faces, and the networks of the ampullae are quite similar to that seen in cosmine. Of course, cosmine-covered sarcopterygians are all extinct and we have no way of knowing what the function of the cosmine canals and pores were for. However, the new fossil suggests a rather different function: growth. Unlike the cosmine of other sarcopterygians, but somewhat like the enamel tissue of actinopterygians, the cosmine of Meemannia grew by the addition and expansion of layers. This suggests that Meemannia's cosmine growth is more primitive. Furthermore, this unique mode of growth suggests that he pore-canals may be a vascular system that supprted this, and may not necessarily have been for sensory purposes. However, I don't see these options as neccessarily exclusive, so I'm not sure if we can really rule out a sensory function just yet.

Meemannia is another addition to the story of the early evolution of bony vertebrates. It's another clue about how this event proceeded and how the two major lineages of bony vertebrates emerged. We still need a lot more clues and a lot more work to figure out the details, but the fossils from China are (once again) helping to disperse the fog.

Also worth a mention here is that Meemannia eos is named after the renowned vertebrate palaeontologist Chang Mee-mann for her enormous contribution to vertebrate palaeontology, especially in China. Her influence on palaeoichthyology has been enormous and she is easily one of the most respected contributers to the field of early vertebrate and early fish evolution.

Zhu, M. et al. 2006. A primitive fish provides key characters bearing on deep osteichthyan phylogeny. Nature 441: 77-80. <link>

See also:

Basden, A.M. et al. 2000. The most primitive osteichthyan braincase? Nature 403: 185-188. <link>

Zhu, M. et al. 1999. A primitive fossil fish sheds light on the origin of bony fishes. Nature 397: 607–610. <link>

Zhu, M. et al. 2001. A primitive sarcopterygian fish with an eyestalk. Nature 410: 81–84. <link>

No comments: