An Asteroid Ended the Age of the Dinosaurs. But How Did Their Reign Begin? Mysterious Early Reptiles May Hold the Answer

Researchers are uncovering the evolutionary steps that set the stage for dinosaurs to rule the planet

Amy McDermott, PNAS Front Matter

February 11, 2026 10:10 a.m.

This piece first appeared in the Front Matter section of the Proceedings of the National Academy of Sciences.

A small but fierce jawbone sits in Argentina’s natural science museum in Buenos Aires. Six inches long and studded with backward-curving fangs that would have hooked into flesh to rip it open, the fossil contains teeth not unlike those of the fearsome Komodo dragon of today, says paleontologist Martín Ezcurra, a researcher at the museum.

But this jaw is no Komodo dragon bone: The 236-million-year-old fossil belonged to an extinct reptile called Lewisuchus admixtus, which ran across Triassic northern Argentina, a world of cycads and ferns that thrived long before grass existed. Lewisuchus stood about knee-high, was roughly five feet long from nose to tail and hunted the cousins of early mammals. It lived about 30 million years before the well-known dinosaurs first roamed the Argentine plains.

So was Lewisuchus a very early dinosaur? Or was it a cousin of the dinosaurs, one of many reptile groups skittering over Earth at the time?

Lewisuchus is embroiled in an ongoing debate that could help resolve enduring mysteries about the dawn of the dinosaurs. Placing it and related reptiles on the tree of life could reveal the early steps in dinosaur evolution, how their rise was shaped by their body plans and how they interacted with other species. Ultimately, says vertebrate paleontologist Sterling Nesbitt at Virginia Tech, research could answer some of the most fundamental questions about the creatures that dominated our world for 180 million years: “What is a dinosaur? And why did they take over the planet?”

A mysterious group of reptiles

We know the main reason that the age of the dinosaurs came to an end: an asteroid impact on the Yucatán Peninsula some 66 million years ago. But how the dinosaurs’ reign began is far less clear—and that’s where Lewisuchus and its relatives can help.

To determine whether these species were early dinosaurs, paleontologists need to compare their skeletons with other reptile fossils from the same geological period. By interpreting features found on hundreds of bones, they can construct evolutionary trees that reveal how dinosaurs were related to similar species. Yet answers have been elusive, in part because complete skeletons are rare and because of disagreements about how to interpret the bones.

“I’ve lost a reasonable amount of hair trying to work out some of the things going on here,” says researcher Paul Barrett, a dinosaur specialist at the Natural History Museum in London. Part of the challenge, he adds, is that there’s no simple way to define a dinosaur. Paleontologists generally agree that an open hip socket—with a hole in the middle, like a turkey’s—is a feature common to all dinosaurs. And a host of other skeletal features, such as the number of vertebrae connecting to the hip and large muscle scars on the limbs, are also used to distinguish dinosaurs.

By the Early Jurassic, some 200 million years ago, dinosaurs with these traits dominated the planet. The very first indisputable dinosaurs turned up in the prior geological period, during the early stages of the Late Triassic, Barrett says, though there aren’t many known species from that time. But other reptiles lived during the Late Triassic, too, and a handful of somewhat mysterious fossils, such as those from Lewisuchus, may or may not be classified as dinosaurs, depending on how researchers interpret their anatomy.

One example of a Triassic early dinosaur is the small bipedal reptile Saturnalia tupiniquim. It lived 233 million years ago in what is now Brazil. Saturnalia had a small head for its body, a very long neck, teeth for slicing plants and an arrangement of bones in its hands such that paleontologists generally agree, Barrett says, that it’s an early member of the lineage that leads to long-necked dinosaurs. Barrett estimates that there are probably about 20 similar examples from the early stages of the Late Triassic for which “the majority of specialists would put their hands on their hearts and say, ‘I believe this is a dinosaur.’”

Complicating matters further, all of these confirmed early dinosaurs belong to only two of the three major dinosaur groups: the long-necked sauropods (like Diplodocus) and the meat-eating theropods (like Tyrannosaurus). The forebears of the third big group—the bird-hipped dinosaurs, or ornithischians, which includes such famous herbivores as Stegosaurus and Triceratops—are conspicuously absent from the fossil record when other early dinosaurs appear. That’s “a real problem,” Barrett says, because the branching patterns of evolutionary trees suggest that ornithischians should have evolved around the same time as the other dinosaurs. “So it’s either that the trees are wrong, or we have a big gap in the fossil record,” he says.

One possibility is that Lewisuchus and other ambiguous reptile fossils from the Middle and Late Triassic were actually the first bird-hipped dinosaurs. If so, “it completely closes that gap,” Barrett says. Certainly, some aspects of Lewisuchus’ skull and teeth look dinosaur-like, hinting that it could represent an early glimmer of the dinosaur line. But it also has traits that are not typical of dinosaurs, such as too few vertebrae connecting to its hips, suggesting that it could instead be a close cousin on a neighboring branch of the tree of life, an experiment in evolution that eventually met a dead end. The Late Triassic was a reptile-dominated world, and determining whether these hard-to-interpret fossils represent additional early dinosaur species or close relatives will help fill in paleontologists’ view of the opening millennia of dinosaur Earth.

In search of the earliest dinosaurs

Lewisuchus is part of a wider discussion about the origins of the ornithischians that paleontologists have been wrestling with for two decades. That debate began with the discovery of another animal, Silesaurus opolensis, which was among the first dinosaur-like reptiles found from the Triassic. In a paper published in 2003, paleontologists described the creature based on surprisingly complete skeletons embedded in claystone in Poland. The fossils dated back about 230 million years, to when Europe was part of the supercontinent Pangea.

Silesaurus has some dinosaur-like features, including a narrow pelvis, leaf-shaped teeth and a beak at the end of its lower jaw—traits shared by the bird-hipped ornithischians. But it also has features that dinosaurs don’t have, including a rounded hinge attaching its neck to its head (in most early dinosaurs, this feature is heart-shaped) and, most important, a closed hip socket. Silesaurus’ eclectic combination of traits put it near the base of the dinosaur tree, but its exact position was uncertain.

Although the Silesaurus discovery was published in a relatively niche journal, the findings quickly grabbed attention in the paleontology world. Nesbitt, of Virginia Tech, remembers sitting at a bar when he was 20 or 21, poring over the Silesaurus paper “and thinking, ‘Oh, this is going to change everything.’” The discovery was groundbreaking because Silesaurus was only the fourth Triassic dinosaur-like animal ever found. The first three were all unearthed in the 1960s from the same layer of rocks in Argentina, and all dated to the early mid-Triassic. Finding another fossil so far from Argentina suggested that Triassic dinosaur-like reptiles had spread across Pangea. “If you draw a line from where Poland was to where Argentina was, it covered most of our planet at the time,” Nesbitt says.

Realizing that those reptiles were either the first dinosaurs or their cousins, and that their bones were probably scattered worldwide, Nesbitt’s next thought was “Let’s go find them!” Five years later, he did. On the second day of a dig in Tanzania, a colleague picked up a bone from the grassy plains and asked: “What do you think this is?” “They were turning it in their hands, and I saw a feature, a flat head on the femur,” Nesbitt recalls. The moment seemed to unfold in slow motion, he says, because that femur shape is found only in dinosaurs and their close relatives. When Nesbitt looked down, he realized “we were standing on all kinds of bones of this animal.” He named it Asilisaurus kongwe: another six-foot-long, knee-high reptile from the Triassic with a head-scratching mix of skeletal traits.

Nesbitt wasn’t the only paleontologist hunting dinosaur cousins in those years. Riding the groundswell of interest, a handful of labs dug for the bones of Triassic reptiles and browsed the dusty shelves of museum collections for any fossils defying categorization. Today, about 13 more species have been found in a group broadly called the silesaurs, to which Silesaurus, Asilisaurus and Lewisuchus are generally thought to belong. There are species from South America, Europe, Africa—“everywhere,” Nesbitt says.

Paleontologists are still debating whether these silesaurs were the first ornithischians or a closely related group of reptiles outside the dinosaur line. But in the past few years, they’ve found new fossils and published new evolutionary trees that aim to resolve the question.

Finding clues in prehistoric anatomy

Paleontologist Rodrigo Temp Müller created one such tree in 2020. “We have no DNA in the fossils,” he explains, so researchers can’t use genetics to infer relatedness. Instead, in the dim light of museum corridors, he leans close to the bones with a hand lens and a tape measure to analyze hundreds of traits.

On a video call from his office at the Federal University of Santa Maria in Brazil, Müller explains how these traits are measured. He pulls a small replica of a femur out of his desk drawer and holds the tiny plastic bone between his thumb and forefinger. “There is a trait around here,” he says, pointing to the top of the bone with a pen. Most dinosaurs have a well-developed crest at the top of this bone, called an anterior trochanter. If paleontologists find it on a femur, they rank the bone with a “one.” If not, they give it a “zero.” For other characteristics, the value is not a simple binary—it could be the ratio between the length and width of the humerus, for example, or the number of serrations in the teeth.

Paleontologists do this for hundreds of traits, in dozens of bones, in every skeleton. Then, they plug the values into a data matrix that includes all the species being considered for a given evolutionary tree. Statistical software pinpoints the simplest hypothesis for the evolutionary relationships between species.

Müller’s 2020 study included dozens of Jurassic dinosaurs, as well as three earlier dinosaur species (such as Saturnalia), and about a dozen silesaurs. It placed the silesaurs firmly within the bird-hipped dinosaur category, suggesting that silesaurs were very early ornithischians. Follow-up studies in 2022 and 2025 added more Jurassic ornithischians and four more silesaurs to the dataset and came to the same conclusion, lending “further support to our hypothesis,” Müller says.

If he’s right, then specific traits shared by some silesaurs and later ornithischians, like teeth for eating plants and a lower jaw with a beak, “probably helped the groups evolve and establish in ecosystems,” Müller says.

However, some researchers, including Nesbitt and Ezcurra, remain unconvinced that silesaurs are dinosaurs. They point out that earlier phylogenies had placed silesaurs outside the dinosaur line.

Max Langer, a paleontologist at the University of São Paulo in Brazil, worked on one of those phylogenies. But as more silesaur fossils were unearthed, Langer revised his view. In 2007, he published the first work that nested the group within the ornithischians, based on evidence in a fossil jaw. Langer explains that “the tip of the ornithischian lower jaw ends in a single pointed toothless bone, the predentary,” whereas those of silesaurs end in a pair of pointed toothless bones. The thinking is that, over time, those two bones fused to each other, says Langer, whose subsequent studies have continued to identify silesaurs as ornithischians.

Experts continue to disagree, however, because of nuances in the interpretation and codification of each skeletal trait in the data matrices, says Ezcurra. Various skeletal features can be interpreted differently, especially when paleontologists categorize a trait using a more subjective grading system, such as large or small. Different researchers might disagree about how to interpret and score traits, leading to different outcomes in evolutionary trees. And although hands and skull cases are particularly useful for revealing close evolutionary relationships, those parts of the skeleton are rarely found in silesaur fossils. “The problem is, in some cases, some structures are poorly preserved,” Ezcurra explains. The result, Müller says, is that “the phylogenetic or evolutionary tree of dinosaurs is in a state of flux.”

Slow and steady

Paleontologists could spend another 20 years scoring and rescoring skeletons, trying to get to the bottom of this puzzle. But even if they never agree on a single answer, they have already narrowed it down.

Two possibilities remain the most likely. One is that Lewisuchus and other silesaurs were not actually dinosaurs, but rather the closest relatives of dinosaurs. That would mean dinosaurs weren’t so different from their cousins, Nesbitt says, implying that both groups evolved similar body plans to survive the conditions of the day. For example, developing elongated hindlegs and feet likely helped both groups move faster, perhaps to evade predators, catch prey or simply forage more efficiently over larger ranges.

The other option is that silesaurs were indeed dinosaurs that “represent an early step in a gradation of forms toward the famous bird-hipped dinosaurs like Stegosaurus,” Nesbitt says. That could push back the date of the earliest dinosaurs to as far as 240 million years ago, some seven million years earlier than the current estimate. That may be a drop in the bucket on geologic timescales, but importantly, it moves the dawn of the dinosaurs ever closer to the largest mass extinction in history—the Permian-Triassic extinction event about 252 million years ago—which could have played a key role in opening new ecological opportunities for the rise of dinosaurs, Nesbitt says.

If more silesaur specimens confirm that timeline, the research may force paleontologists to rethink how dinosaurs evolved and the date of the earliest dinosaur common ancestor. Paleontologists have long assumed that the open hip socket and other telltale features shared by dinosaurs were passed down from their first common ancestor. But if the closed-hipped silesaurs were the first ornithischians, then it’s not so simple. Instead, the three major dinosaur groups—theropods, sauropods and ornithischians—would have independently evolved some of their shared traits, like the open hip, rather than inheriting those traits from a common ancestor.

But the silesaurs may not cause such disruption. Jack Lovegrove, a PhD student in the fossil reptiles research group at London’s Natural History Museum, argues that many of the recent papers that judge whether silesaurs were early ornithischians ultimately depend on tinkering with the same few datasets that most paleontologists employ. Each of these matrices has plenty of data points but needs to be cleaned up and made more precise by removing or de-emphasizing poorly defined and ambiguous traits, Lovegrove says.

In June 2024, he published a review that called for researchers to “focus on building new matrices, rather than bolting bits onto the old ones,” he says. That would provide an opportunity to re-evaluate dozens of traits in hundreds of species and to exclude particularly hard-to-place or confounding skeletal features that bias a tree toward one hypothesis or another. Lovegrove is now working on a new matrix that offers clearer definitions of some key characteristics.

Ezcurra, too, is working on a brand-new matrix that scores all these skeletal traits afresh to reassess the position of the silesaurs. His matrix includes 1,200 traits and almost 400 species of dinosaurs and dinosaur precursors, and although the work is not yet published, “the preliminary results are that silesaurs are outside dinosaurs,” he says.

“There’s always an idea in paleontology that out there is a magical missing link, and we’ll find it and everything will slot into place,” Lovegrove says. But in this case, the field needs fewer “big explosive new results,” he says, and more focus on the slow, steady work of classifying traits.