Earliest-known turtle had only half a shell

Fresh in this week's Nature: a newly-discovered fossil turtle, the oldest ever found, has a lower shell, but no upper one [$-a]. Odontochelys semitestacea, as it's called, is a really neat potential transitional fossil - the ribs are flattened like butter knives, but not fused into an upper shell. Apparently, this is suggestive of the way in which the upper shell forms in embryonic modern turtles, and the authors are careful to point out that, in other respects, the fossil is clearly an adult.

See that thing on its back? Its
ancestors may not have had one.
Photo by raceytay.

In an accompanying News and Views piece, Reisz and Head suggest that the lower half of a shell would be quite useful [$-a] if Odontochelys lived mostly in the water, where predators are more likely to attack from below than from above. They argue, though, that Odontochelys may not represent a transitional step between shell-less ancestors and full-shelled modern turtles, but a case of "secondary loss," in which a full-shelled turtle took to the water and subsequently lost its unnecessary and cumbersome upper shell. I'm no turtle anatomist, but this sounds like a plausible alternative hypothesis. The only way to test it is is to dig up an even older turtle, and see what its shell looks like.

(See also coverage by All Things Considered, which is pretty good if unnecessarily snarky about the degree to which paleontologists specialize. It's not like it's that odd to think someone might build a career comparing birds' beaks to turtles' beaks.)

References

C. Li, X.-C. Wu, O. Rieppel, L.-T. Wang, L.-J. Zhao (2008). An ancestral turtle from the Late Triassic of southwestern China. Nature, 456 (7221), 497-501 DOI: 10.1038/nature07533

R.R. Reisz, J.J. Head (2008). Palaeontology: Turtle origins out to sea. Nature, 456 (7221), 450-1 DOI: 10.1038/456450a