Thanks to Brilliant for supporting this episode of SciShow.

Brilliant courses can help you cultivate your math and scientific thinking skills. Head to Brilliant.org/SciShow for a special offer on their annual premium subscription. [♪ INTRO].

The origin story of turtles is a mystery that has perplexed scientists for literally centuries. Because turtles are -- and this is a technical term in evolutionary biology -- really weird. At least as far back as the 1700s, scientists have hypothesized and debated about how and when turtles evolved their unique bodies, and what ancestors they even came from.

Since then, we’ve learned a lot about evolution and the history of life in general, but scientists are still trying to answer many of those pesky old turtle questions. But things have become much clearer over the last couple decades. With new insights from genetics, developmental biology, and the fossil record, our understanding of turtle evolution is improving fast.

Well, faster than tortoise-speed, anyway. One major reason why turtles are so confusing is that they’re just so darn different from everything else. I mean, look at that shell.

It’s a mosaic of bones, including expanded portions of ribs, vertebrae, and shoulder bones, plus bones that grow within the skin, all fused together into a portable shelter. They’re all but wearing their skeletons on the outside. Nothing else on Earth has a structure quite like that.

Nothing with bones, anyhow. Their skulls are bizarre, too. They’re missing a lot of the joints and open spaces we see in most other vertebrates, either now or in the past, not to mention their lack of teeth.

Now, when scientists sort out animal relationships, they compare similar features to identify which groups of living things share the most recent common ancestry. But turtles have a lot of derived features: that is, traits that weren’t present in the ancestors they share with other animals, and that only evolved in the turtle lineage. That makes it tricky to know exactly how to compare their bodies with other animals.

Normally, in cases like this, the fossil record can be a lot of help. That’s because when you can get a look at earlier stages in an evolutionary line, you can sometimes identify ancestral features that have since been lost. That would help us connect turtles to other groups that also have those features.

But, while we have great fossil records showing the early evolution of animals like birds, whales, humans, and many more, the record of early turtles is sparse. And until recently, the oldest known turtles looked mostly like turtles. They had already evolved their derived features.

What all that adds up to is that there must have been an early evolutionary history of turtles, but they didn’t have the decency to leave any trace of it. These complications have led to a whole list of differing hypotheses. Over the decades, various studies have placed turtles in vastly different parts of the vertebrate family tree.

Some have suggested turtles are not-too-distant cousins of other modern reptiles. Others have said they represent an ancient branch that predates the split between reptiles and mammals, and many other hypotheses have fallen somewhere in between. But in the past couple decades, the evolutionary story of turtles is finally starting to come into clearer focus, thanks to some excellent new scientific insights.

For one thing, there’s DNA! Genetic comparisons are often useful when anatomy is confusing, so scientists have used DNA to explore turtle genealogy. And, well, the results disagree with almost all the previous hypotheses, which is surprising.

But they’re very consistent in their disagreement, which is exciting! See, other modern reptiles -- lizards, snakes, crocodilians, and technically birds -- are united in a group called diapsids. While most anatomical studies identify turtles outside this group, genetic studies tend to place them inside it, often as close relatives of archosaurs, that is, crocs and birds.

Now, DNA isn’t perfect, and there are still complications that could be leading these results off course, but the fact that studies keep agreeing with each other is definitely intriguing. Another exciting new source of information is developmental biology. Recent research has been able to closely follow turtle embryos as they develop, which helps researchers pin down how different body parts form.

This can give us clues to the ancestral origins of some of those weird derived features, helping us understand how to hone our anatomical comparisons and showing us what to look for in fossils. Speaking of fossils, our paleontological understanding of turtles has been getting better, too. For a long time, the oldest known turtle was Proganochelys from the.

Triassic Period, about 210 million years old. It has some teeth in its mouth, but other than that, it looks a lot like modern turtles. But a bunch of recent discoveries have given us peeks earlier in turtle evolution.

Slightly older than Proganochelys are Odontochelys and Eorhynchochelys, both from China. These two also have teeth, and they have an incomplete shell comprising broad ribs that don’t fuse into a full dome like modern turtles. Even older than that is Pappochelys from Germany, which also has an incomplete shell and extra spaces in the skull, similar to diapsid reptiles.

And even older still is an animal named Eunotosaurus, a possible turtle ancestor that has turtle-like ribs and signs of a turtle-like breathing system. But this animal has no true shell, and extra gaps in its skull. Considered together, these fossils give us a hint of the turtle evolutionary story.

It seems the broad ribs and stiffened bodies evolved early, possibly to support a digging lifestyle. Over time, those reinforced bones gave rise to a protective shell, the extra holes in the skull gradually disappeared, and eventually, so did teeth. And the best part is that these fossils match what we’d expect from other data!

Those skull-holes are an anatomical clue linking turtles to diapsid reptiles, which lends support to those DNA studies. And the shapes of the ribs in those early turtles, before the full shell evolved, match what we see in turtle embryos as they develop their shells. Now, there’s still plenty left to learn.

The fossil record is still pretty sparse, DNA analyses still need refinement, and there’s lots we still don’t fully understand about turtle development. But with all this new data, it seems a scientific consensus is slowly forming. Not only around the steps it took for turtles’ unique bodies to develop, but also around their position on the tree of life.

Which appears to be closer to other living reptiles than we used to think! Not only are we on track to answering these long-standing turtle questions, these analyses can help guide our approach to other big evolutionary questions. It’s a great model to help scientists improve the ways they integrate data from different disciplines.

On top of that, a better understanding of turtle origins can contribute to our understanding of the evolution of their reptilian cousins, and that includes crocs, snakes, dinosaurs, and more. So maybe it was worth the time it took to understand turtles. Slow and steady wins the race, or at least helps us understand the rest of evolution!

By and large, though, biology today is a fast-paced world. And to understand it all, more and more scientists turn to computers to help them handle large amounts of data. You can learn how with Brilliant’s course on Computational Biology.

Brilliant is an online learning platform with courses about science, engineering, computer science and math. The courses are designed for people of all skill levels, so you can jump in at any point and work your way to mastery. Plus, they are super hands-on with interactive quizzes and guided problems with explanations.

If you’re interested, you can sign up at brilliant.org/scishow to get 20% off an annual Premium subscription. So thanks for your support!