Turtles have a goofy body plan. Unlike all other four-limbed critters, which have their shoulder blades riding on the outside of their ribs, the turtle's ribs are outside of its shoulder blades. This allows turtles to make their shell out of fused bones--the only animal to do so. Now a new study uses embryology to look at how this odd skeleton develops, and how the turtle could have evolved.
Turtles appear to spring up fully formed in the fossil record, which has led to a lot of debate about who their closest relatives are. Some paleontologists say that turtles are a primitive reptile; some say they're related to snakes; and molecular phylogenies have lumped them in with birds and crocodiles. So scientists have turned to comparative embryology to see how turtles' development diverges from that of related animals. In the new study, researchers at the RIKEN Center for Developmental Biology in Kobe, Japan, compared turtle embryos to those of a close relative, chickens, and a more distant relative, mice.
The researchers bought fertilized eggs of Chinese soft-shelled turtles--a Japanese delicacy that tastes like chicken, says evolutionary developmental biologist and team member Shigeru Kuratani--from local farms and followed their development. They used several approaches to look at the embryos, including staining them with dyes, scanning them using computed tomography, and tracking them with RNA probes.
In today's issue of Science, Kuratani and his colleagues describe how turtles' ribs take off in a different direction from those of chicken and mice. A young turtle embryo looks a lot like a young chicken embryo. But at about day 11 of its 30-day incubation, the turtle embryo folds at the carapacial ridge, a structure at the edge of what will be the shell that was already known to be involved in getting the shell to develop (see image). The ribs grow toward this fold, shoot over the scapulae, or shoulder blades, and go on to form the shell. Chicken and mice have no such fold. Their ribs grow inside the scapulae and around to the front, making a ribcage. "A turtle is a weird case of origami, a new folding," says Kuratani. The turtle's muscles follow the ribs; in some cases, they attach in the same ways as in mice, chicken, and, therefore, their ancestors, but in others, they form new attachments.
All of these observations suggest that, at some point in turtle evolution, there were probably animals that had their ribs and shoulder blades in some intermediate arrangement. And, indeed, the researchers point to the oldest known fossilized turtle, described last year. It apparently had no top shell, but seems to have had its shoulder blades in front of its ribs, a lot like the arrangement in modern turtle embryos.
"[The study] is a really wonderful example of classical comparative embryology," says evolutionary morphologist Ann Burke of Wesleyan University in Middletown, Connecticut, who did her dissertation on turtle development. "It's very consistent with the work that's come before looking at turtle evolution, but it adds this extra level of detail in terms of the musculature of the shoulder girdle, which is where all the action is."
Robert Reisz, a vertebrate paleontologist at the University of Toronto, Mississauga, in Canada, says that it's particularly helpful to see how turtles' muscles must have shifted through evolution from the arrangement in their ancestors to the unusual body plan they have now: "It's a wonderful design."