Where Do Turtles Go?

By Erica Garcia

Turtles may not be the “living fossils” they were thought to be.

The first time that paleontologist Olivier Rieppel presented his findings on turtles, before 200 people at a meeting last year sponsored by the Society of Vertebrate Paleontology, a presenter prefaced his talk with, “And now everybody may hiss as much as you like.” Venomous commentary did not ensue, but a bit of a murmur must have lingered as Rieppel announced that he believed turtles had been classified in the wrong branch of the reptile family tree.

Rieppel, from the Field Museum in Chicago, and Ph.D. student Michael deBraga of Erindale College in Ontario, knew they were proposing a maverick theory. Turtles had long been deemed to be “living fossils,” the only surviving member of a primitive reptile subclass, the anapsids, which originated some 325 million years ago in the Paleozoic era. Now, these two researchers were proposing that turtles belonged to the modern reptilian lot — the diapsids, which first emerged about 230 million years ago in the Triassic and include present-day lizards, snakes and crocodiles.

The team came to that conclusion using cladistics, a generally well-accepted way of figuring evolutionary relations. It relies on the muddy task of identifying so-called homologous characteristics shared by certain groups. But they decided to challenge the long-held belief that anapsid and diapsid skulls are the ultimate defining characteristics. So unlike earlier cladistic modeling of turtles, Rieppel and deBraga examined numerous features and included taxa from outside the Paleozoic. DeBraga says, “We decided to look at everything…[to give turtles] a whole new approach, and lo and behold look at what happens.”

Their work hinged on computer analyses of huge sets of data. Although more data would seem to buttress the validity of their work, it intensified the predicament of deciding which characteristics are appropriate. Among the 168 characteristics studied, what really convinced Rieppel that turtles are diapsids were their ankles. He says the morphological similarity among the ankles of turtles, lizards and the tuatara, a lizard from New Zealand, is too strong to be denied.

Although some of the morphological evidence presented is quite sound, Rieppel and deBraga hear some hissing. Gene Gaffney, curator of vertebrate paleontology at the American Museum of Natural History (and orator of the snide opening statement before Rieppel’s talk), believes the “evidence presented was somewhat skewed.” He notes that some of the characteristics used in the study — bone ossification, for one — are not particularly reliable for all amniotes (reptiles, birds, mammals). Other critics assert that comparing different groups can spawn erroneous, or at least highly questionable, theories. Such was the case in the 1800s, when the theory of hematothermia arose. It stated that birds and mammals are closely related because they are both warm-blooded, even though the overwhelming evidence suggests that birds are more reptilian.

What particularly distresses some researchers about the turtle debate is that it takes only a few additional characteristics in the data matrix to move turtles again, back into anapsids. Rieppel counters by insisting that evolutionary trees tend to become unstable when they become too heavy with characteristics.

Although their work may not have received the sanctification of colleagues, neither has it been discounted. Other animals may be misplaced and thus may force a reworking, or at least a rethinking, of various evolutionary paths. Never mind the hissing, at least they’re talking.

Leave a Comment