x Abu Dhabi, UAEWednesday 26 July 2017

How the turtle got his shell

The evolution of the turtle shell has intrigued scientists for centuries but Chinese researchers think they have solved the mystery.

An artist's impression of an ancestral turtle from the Triassic Period whose 220 million-year-old fossil was found in Guizhou Province, China.
An artist's impression of an ancestral turtle from the Triassic Period whose 220 million-year-old fossil was found in Guizhou Province, China.

Nature has a habit of creating puzzles that tease the imagination as much as they do the intellect. There's the question of how the elephant got its trunk and which came first, the chicken or the egg. Equally intriguing is the issue of how the turtle's shell came about, although this is one conundrum that looks as though it has now been resolved. During two centuries of study, scientists have been torn between two competing hypotheses. One school of thought suggests the shell of these creatures, which are familiar to divers and snorkellers off the coast of the UAE, is formed from skin plates that hardened to ward off enemies.

Crocodiles, some other reptiles and many non-turtle fossil vertebrates, including dinosaurs, have such body armour plates, known as osteoderms, and were possibly related to the turtle. Others believed that the turtle shell formed from ribs that grow out and broaden to protect the animal from danger. Three fossils of a turtle with an incomplete shell, recently discovered in China and described in the journal Nature, appear to have finally resolved the debate in favour of the second hypothesis.

The 220million-year-old fossil species is named Odontochelys semitestacea, which translates as a turtle with teeth and half a shell. This creature is the first fossil turtle with an incomplete shell, because even fossils of turtles from 210million years ago have been fully protected by a shell. While the long-tailed O semitestacea has protective armour or a plastron on its underside, it lacks the hard shell or carapace seen on the back of modern turtles and other fossil turtles.

The presence of breast plates on the creatures indicates they were aquatic, since if they walked on land their front side would be less vulnerable and therefore armour would not have developed there first. While the protection is much more extensive on the animal's front than its back, the animal's back does show some broadening of the ribs, according to Dr Xiao-Chun Wu, a scientist at the Canadian Museum of Nature in Ottawa who examined the fossilised remains.

Dr Wu and his collaborators, who were led by Chun Li of the Chinese Academy of Sciences in Beijing, believe this is conclusive evidence that the modern turtle's shell is made from an extension of its backbone and ribs rather than the hardening of skin plates. Evidence from developmental biology seems to support their conclusions, according to Dr Wu, who is originally from China but has spent the past two decades in Canada.

"From developmental studies, the turtle ribs become broader and fuse with each other and invade into the top layer to co-ossify as a single layer of armour - the carapace," he says. "Our specimen shows exactly the intermediate pattern, similar to the early stage of the shell development in the living turtle embryo. "It does not have dermal armour to form the top layer. It only has rudimentary broadening of the ribs. The ribs broaden but do not connect with each other."

Even the beginnings of a shell on the back would have afforded these early turtles some protection from predators, so it is easy to understand how the full shell came about by extensions of the vertebrae and further broadening of the ribs. The apparent breakthrough represented by the discovery of this fossil species could easily have been missed, since initially scientists believed they had stumbled upon an entirely different type of creature.

The first of three O semitestacea fossils found was thought to represent the remains of a placodont, a type of marine reptile from the triassic period of geological history with a heavily armoured back. However, several key features have made it possible to identify what the fossil really is. Perhaps most important is the presence of the hypoischeum, a bone in the pelvic girdle that is unique to early turtles.

"We figured out it was a turtle rather than any known placodont because it had this extra bone that only occurs in the primitive turtle," says Dr Wu, who examined this first specimen in July last year. Once the first fossil had been correctly identified as an early turtle, one of the group of researchers decided to follow up a lead from two years earlier. He contacted a farmer in the Chinese province of Guizhou in the country's south-west, where the first example had been found, who had previously sent in a photograph of a similar-looking creature. The scientist hoped this other fossil would yield further clues.

"He went back to find the farmer, who still kept the specimen," explains Dr Wu. It turned out this second fossil, which Dr Wu himself got to grips with midway through last year, yielded everything the biologists could have hoped for. It gave better information about what the back of the creature was like, as in the first specimen there was only a good ventral, or front, view. Another important feature it showed was the presence of teeth. Scientists were excited by this because teeth indicated the specimen was primitive - an ancestor of modern turtles - rather than an evolutionary offshoot.

Over time, these teeth were lost, so that in turtles that exist today there are horny ridges covering the upper and lower jaws instead of teeth. The relationship between the ribs and the vertebrae was another indicator that the fossil represented a primitive pre-turtle. In most reptiles there is a 1:1 relationship between the ribs and the vertebrae, while in turtles there is one rib for every two vertebrae.

"All [previous] fossil turtles and living turtles have such a relationship - the rib is articulated between two vertebrae," he said. The fossils, however, had the 1:1 relationship seen in other reptiles, again indicating it was a primitive specimen that existed before other turtle specimens discovered so far. A third specimen of O semitestacea was also found in the same area, although this was in slightly poorer condition.

But by now the scientists had all the information they needed to resolve the puzzle of the turtle's shell. dbardsley@thenational.ae