When I was a teenager at the turn of the millennium, right around the time I became smitten with fossils, the Field Museum in Chicago dismantled its Brachiosaurus and installed a Tyrannosaurus rex. In essence, the institution was trading one dinosaur icon for another. Out went the plant-eating colossus, heavier than 10 elephants, its neck arcing gracefully far above the museum’s second-floor viewing gallery. In came the biggest, baddest predator of all time: a bus-sized brute with railroad-spike teeth that shattered the bones of its prey.

These were the dinosaurs that fired my imagination as I grew up 75 miles down the road from Chicago, in a flat expanse of Midwestern corn and bean fields. I visited them as often as I could convince my parents to make the drive. Standing underneath their skeletons was hypnotic: their size, their strength, their bodies so alien compared with those of any animals alive today. No wonder they ruled the earth for more than 150 million years. They were magnificent.

But how did dinosaurs get this way? It was a question I rarely contemplated during those obsessive years. In the same way that it was hard to envision that my parents were once my age, I just assumed dinosaurs materialized at some point in the deep past as fully formed long-necked and sharp-toothed giants. I didn’t know it at the time, but that wasn’t too far off from the scientific consensus for much of the late 20th century. Dinosaurs were special, this convention held, endowed with such superior speed, agility and metabolism that they quickly and easily outcompeted their early rivals, spread across the planet and established an empire.

Over the past 15 years, however, a wealth of new fossil discoveries from around the globe, fresh insights into the physical world the first dinosaurs inhabited, and novel approaches to building family trees and analyzing evolutionary trends have challenged that long-standing view. From these advances, a rather different story has emerged: the rise of dinosaurs was gradual, and for the first 30 million years of their history they were restricted to a few corners of the world, outpaced by other species. Only after catching a couple of lucky breaks did they rise up to take over the planet.

Humble Origins

Like many successful organisms, dinosaurs were born of catastrophe. Around 252 million years ago, at the tail end of the Permian Period, a pool of magma began to rumble underneath Siberia. The animals living at the surface—an exotic menagerie of large amphibians, knobby-skinned reptiles and flesh-eating forerunners of mammals—had no inkling of the carnage to come. Streams of liquid rock snaked through the mantle and then the crust, before flooding out through mile-wide cracks in the earth’s surface. For hundreds of thousands, maybe millions, of years the eruptions continued, spewing heat, dust, noxious gases and enough lava to drown several million square miles of the Asian landscape. Temperatures spiked, oceans acidified, ecosystems collapsed and up to 95 percent of the Permian species went extinct. It was the worst mass extinction in our planet’s history. But a handful of survivors staggered into the next period of geologic time, the Triassic. As the volcanoes quieted and ecosystems stabilized, these plucky creatures now found themselves in a largely empty world. Among them were various small amphibians and reptiles, which diversified as the earth healed and which later diverged into today’s frogs, salamanders, turtles, lizards and mammals.

Scientists know about these animals from the handprints and footprints they left in layer after layer of river and lake sediments now exposed in the Holy Cross Mountains in Poland. For more than 20 years Grzegorz Niedwiedzki, who grew up in these hills and is now a paleontologist based at Uppsala University in Sweden, has meticulously collected these fossil tracks, occasionally with me by his side. In 2005, while fossil hunting near the village of Stryczowice, along a narrow stream tangled with brambles, Niedwiedzki discovered an unusual type of track that did not seem to match any of the more common reptile and amphibian traces. The strange prints are about the size of a cat’s paw, arranged in narrow trackways, with the five-fingered handprints positioned in front of the slightly larger footprints, which have three long central toes flanked by a toe nubbin on each side. The tracks go by the genus name Prorotodactylus. All that we know about this creature comes from these prints—there are no known fossils of the animal itself.

Fossil tracks of Prorotodactylus show that around 250 million years ago dinosaur precursors called dinosauromorphs roamed what are now the Holy Cross Mountains in Poland. Credit: Grzegorz Niedwiedzki Uppsala University

The Prorotodactylus tracks date to about 250 million years ago, just one or two million years after the volcanic eruptions that brought the Permian to a close. Early on it was clear from the narrow distance between the left and right tracks that they belonged to a specialized group of reptiles called archosaurs that emerged after the Permian extinction with a newly evolved upright posture that helped them run faster, cover longer distances and track down prey with greater ease. The fact that the tracks came from an early archosaur meant that they could potentially bear on questions about the origins of dinosaurs. Almost as soon as the archosaurs originated, they branched into two major lineages, which would grapple with each other in an evolutionary arms race over the remainder of the Triassic: the pseudosuchians, which led to today’s crocodiles, and the avemetatarsalians, which developed into dinosaurs. Which branch did Prorotodactylus belong to?

To find out, I conducted a study with Niedwiedzki and Richard J. Butler, now at the University of Birmingham in England. Our analysis of the prints, published in 2011, revealed peculiarities of the footprints that link them to signature features of the dinosaur foot: the digitigrade arrangement of the bones, in which only the toes make contact with the ground while walking, and the very narrow foot with three main toes. Prorotodactylus is therefore a dinosauromorph: not a dinosaur per se but a primitive member of the avemetatarsalian subgroup that includes dinosaurs and their very closest cousins. Members of this group had long tails, big leg muscles, and hips with extra bones connecting the legs to the trunk, which allowed them to move even faster and more efficiently than other archosaurs.

These earliest dinosauromorphs were hardly fearsome, however. Fossils indicate that they were only about the size of a house cat, with long, skinny legs. And there were not very many of them either: less than 5 percent of all Stryczowice tracks belong to Prorotodactylus, which is far outnumbered by tracks of small reptiles, amphibians and even other archosaurs. The dinosauromorphs’ time had not come. Yet.

The First Dinosaurs

Over the next 10 million to 15 million years the dinosauromorphs continued to diversify. The fossil record from this time period shows an increasing number of track types in Poland and then around the world. The tracks get larger and develop a greater variety of shapes. Some trackways stop showing impressions of the hand, a sign the makers were walking only on their hind legs. Skeletons start to turn up as well. Then, at some point between 240 million and 230 million years ago, one of these primitive dinosauromorph lineages evolved into true dinosaurs. It was a radical change in name only—the transition involved just a few subtle anatomical innovations: a long scar on the upper arm that anchored bigger muscles, some tablike flanges on the neck vertebrae that supported stronger ligaments, and an open, windowlike joint where the thighbone meets the pelvis that stabilized upright posture. Still, modest though these changes were, they marked the start of something big.

The oldest unequivocal dinosaur fossils, which date to around 230 million years ago, come from the otherworldly landscapes of Ischigualasto Provincial Park in Argentina. Scientists have collected there for decades, beginning with legendary American paleontologist Alfred Romer in the 1950s and continuing with Argentine researchers Osvaldo Reig and José Bonaparte in the 1960s. More recently, Paul Sereno of the University of Chicago and Ricardo N. Martínez of the National University of San Juan in Argentina led expeditions to Ischigualasto in the 1980s and 1990s. Among the fossils they found there were those belonging to Herrerasaurus, Eoraptor and other creatures representing all three of the main branches of the dinosaur family: the meat-eating theropods; the long-necked, plant-eating sauropodomorphs and the beaked, plant-eating ornithischians.

By the middle part of the Triassic, around 230 million to 220 million years ago, these three main dinosaur subgroups were on the march, siblings setting out to form their own broods in a world we would barely recognize. Back then a single supercontinent called Pangea stretched from pole to pole, surrounded by a global ocean called Panthalassa. It was not a safe place to call home. The earth was much warmer, and because Pangea was centered on the equator, half the land was always scorching in the summer while the other half was cooler in the winter. These marked temperature differences fueled violent “mega monsoons” that divided Pangea into environmental provinces characterized by varying degrees of precipitation and wind. The equatorial region was unbearably hot and muggy, flanked by subtropical deserts on both sides. The midlatitude regions were slightly cooler and much wetter.

Herrerasaurus, Eoraptor and the other Ischigualasto dinosaurs were ensconced in the comparatively hospitable midlatitudes. So were their counterparts from Brazil and India, known from exciting recent fossil discoveries. But what about other parts of the supercontinent? Did early dinosaurs colonize these harsher regions just as capably, as the conventional wisdom about them suggests? In 2009, a few months after our first jaunt together in Poland, Butler and I teamed up with Octávio Mateus of the Museum of Lourinhã in Portugal to test this hypothesis by exploring a remnant of the subtropical arid belt of northern Pangea in what is now southern Portugal. We were hoping to find dinosaurs, but what we found instead was a mass graveyard of hundreds of Smart car–sized amphibians that we assigned to a new species, Metoposaurus algarvensis. These rulers of the Triassic lakes and rivers had been victims of a freak shift in the capricious Pangean weather that probably caused their lakes to dry up. We returned later to excavate the bone bed and started to also find fossils of various fishes, poodle-sized reptiles and archosaurs from the line leading to crocodiles. But still, to this day, we have yet to come across even a scrap of dinosaur bone.

We probably never will. Spain, Morocco and the eastern seaboard of North America have stellar fossil sites from this same time between 230 million and 220 million years ago that show the same pattern we saw in Portugal: plenty of amphibians and reptiles but nary a dinosaur. All these places were in the arid sector of Pangea. Together these sites indicate that during the formative years of their evolution, dinosaurs were slowly diversifying in the humid temperate regions but were seemingly unable to colonize the deserts. It is an unexpected story line: far from being superior creatures that swept across Pangea the moment they originated, dinosaurs could not handle the heat. They were geographically localized—mere bit players in the drama playing out across a world still recovering from the great End Permian extinction.

But then, just when it seemed that dinosaurs would never escape their rut, they received two lucky breaks. First, in the humid zone, the dominant large herbivores of the time—reptiles called rhynchosaurs and mammal cousins called dicynodonts—went into decline, disappearing entirely in some areas for reasons still unknown. Their fall from grace between 225 million and 215 million years ago gave primitive plant-eating sauropodomorphs such as Saturnalia, a dog-size species with a slightly elongated neck, the opportunity to claim an important niche. Before long these sauropod precursors were the main herbivores in the humid parts of the Northern and Southern Hemispheres. Second, around 215 million years ago dinosaurs finally broke into the deserts of the Northern Hemisphere, probably because shifts in the monsoons and the amount of carbon dioxide in the atmosphere made differences between the humid and arid regions less severe, allowing dinosaurs to migrate between them more easily.

They still had a long road ahead of them, however. The best records of these first desert-dwelling dinosaurs come from areas that are once again deserts today, in the colorful badlands of the southwestern U.S. For more than a decade a team of young researchers has been methodically excavating the Hayden Quarry, a fossil-rich locality in artist Georgia O’Keeffe’s much loved retreat of Ghost Ranch in New Mexico. Randall Irmis of the University of Utah, Sterling Nesbitt of Virginia Tech, Nathan Smith of the Natural History Museum of Los Angeles County, Alan Turner of Stony Brook University and Jessica Whiteside of the University of Southampton in England have found a bounty of skeletons: monster amphibians closely related to our Portuguese Metoposaurus, primitive crocodile relatives, and a host of curious swimming and tree-hopping reptiles. There are also dinosaurs in the Hayden Quarry, though not many of them: only a few species of predatory theropods, each represented by a few fossils. There were no plant eaters: none of the ancestral long-necked species so common in the humid zones, none of the ornithischian forebears of Triceratops. The team argued that, once again, the paucity of dinosaurs came down to the weather: these deserts were unstable environments of fluctuating temperatures and rainfall, with raging wildfires during some parts of the year and humid spells in others. Plants had difficulty establishing stable communities, which meant that plant-eating dinosaurs did not have a steady source of food. Thus, some 20 million years after they had originated and even after they had taken over the big herbivore role in humid ecosystems and started to settle the tropical deserts, dinosaurs had yet to mount a global revolution.

Stiff Competition: For much of the Triassic period dinosaurs were a marginal group, overshadowed by the likes of crocodile relatives such as Saurosuchus (1) and giant amphibians such as Metoposaurus (2). Credit: Ricardo N. Martínez Institute and Museum of Natural Sciences, National University of San Juan (1); Tomasz Sulej Institute of Paleobiology, Polish Academy of Sciences (2)

Croc Competition

No matter which interval you look at in the Triassic, from the time the first dinosaurs appeared around 230 million years ago until the period ended 201 million years ago, the story is the same. Only some dinosaurs were able to live in some parts of the world, and wherever they lived—humid forests or parched deserts—they were surrounded by all kinds of bigger, more common, more diverse animals. In Argentina’s Ischigualasto, for instance, those earliest dinosaurs made up only about 10 to 20 percent of the total ecosystem. The situation was similar in Brazil and, millions of years later, at the Hayden Quarry. In all cases, the dinosaurs were vastly outnumbered by mammal forerunners, giant amphibians and eccentric reptiles.

More than anything, however, Triassic dinosaurs were being outgunned by their close cousins the so-called pseudosuchians, on the crocodile side of the archosaur family. At Ischigualasto, a crocodile-line archosaur called Saurosuchus ruled the food chain, with its sharp teeth and muscular jaws. Hayden Quarry harbored numerous pseudosuchian species: semiaquatic ones with long snouts, armored ones that ate plants, and even toothless ones that sprinted on their hind legs and bore a striking resemblance to some of the theropod dinosaurs they lived alongside.

As a master’s student in the late 2000s, around the time many of these fossils were being discovered, I found this pattern peculiar. At the same time I was following the onslaught of new fossils, I started reading classic studies by giants in the field of paleontology, including Robert Bakker and Alan Charig, who effusively argued that dinosaurs were so perfectly adapted, with speed and endurance and smarts, that they quickly took out their crocodile cousins and other competitors during the Triassic. But this idea did not seem to jibe with the fossil record. Was there some way I could test it?

After immersing myself in literature on statistics, I realized that two decades earlier paleontologists who study invertebrate animals had come up with a method for measuring anatomical diversity in a group of species, which had so far been ignored by dinosaur researchers. This measurement is called morphological disparity. If I could track the disparity of dinosaurs and pseudosuchians over the Triassic, I could see whether they were becoming more or less diverse and at what rate—which would indicate whether they became successful gradually or abruptly—and whether one group was pulling ahead of the other.

Working with my then supervisors at the University of Bristol in England—Michael Benton, Marcello Ruta and Graeme Lloyd—I compiled a large data set of Triassic dinosaurs and pseudosuchians, assessing more than 400 characteristics of their anatomy. When we analyzed it statistically, we came up with a startling result that we published in 2008 in Science. All throughout the Triassic the pseudosuchians were significantly more anatomically diverse than the dinosaurs, which indicates that they were experimenting with more diets, more behaviors and more ways of making a living. Both groups were becoming more diverse as the Triassic unfolded, but the pseudosuchians always outpaced the dinosaurs. Contrary to the leading view of dinosaurs as superior soldiers slaying their rivals, they were actually losing to the pseudosuchians for most of their long coexistence.

Carpe Diem

Our statistical analysis led us to an iconoclastic conclusion: the first dinosaurs were not particularly special, at least compared with the variety of other animals they were evolving alongside during the Triassic. If you were around back then to survey the Pangean scene, you probably would have considered the dinosaurs a fairly marginal group. And if you were of a gambling persuasion, you would probably have bet on some of the other animals, most likely those hyperdiverse pseudosuchians, to eventually become dominant, grow to massive sizes and conquer the world. But of course, we know that it was the dinosaurs that became ascendant and even persist today as more than 10,000 species of birds. In contrast, only two dozen or so species of modern crocodilians have survived to the present day.

How did dinosaurs eventually wrestle the crown from their crocodile-line cousins? The biggest factor appears to have been another stroke of good fortune outside the dinosaurs’ control. Toward the end of the Triassic great geologic forces pulled on Pangea from both the east and west, causing the supercontinent to fracture. Today the Atlantic Ocean fills that gap, but back then it was a conduit for magma. For more than half a million years tsunamis of lava flooded across much of central Pangea, eerily similar to the enormous volcanic eruptions that closed out the Permian 50 million years prior. Like those earlier eruptions, the End Triassic ones also triggered a mass extinction. The crocodile-line archosaurs were decimated, with only a few species—the ancestors of today’s crocodiles and alligators—able to endure.

Dinosaurs, on the other hand, seemed to have barely noticed this fire and brimstone. All the major subgroups—the theropods, sauropodomorphs and ornithischians—sailed into the next interval of geologic time, the Jurassic Period. As the world was going to hell, dinosaurs were thriving, somehow taking advantage of the chaos around them. I wish I had a good answer for why—was there something special about dinosaurs that gave them an edge over the pseudosuchians, or did they simply walk away from the plane crash unscathed, saved by sheer luck when so many others perished? This is a riddle for the next generation of paleontologists to solve.

Whatever the reason dinosaurs survived that disaster, there is no mistaking the consequences. Once on the other side, freed from the yoke of their pseudosuchian rivals, these dinosaurs had the opportunity to prosper in the Jurassic. They became more diverse, more abundant and bigger than ever before. Completely new dinosaur species evolved and migrated widely, taking pride of place in terrestrial ecosystems the world over. Among these newcomers were the first dinosaurs with plates on their backs and armor covering their bodies; the first truly colossal sauropods that shook the earth as they walked; carnivorous ancestors of T. rex that began to get much bigger; and an assortment of other theropods that started to get smaller, lengthen their arms and cover themselves in feathers—predecessors of birds. Dinosaurs were now dominant. It took more than 30 million years, but they had, at long last, arrived.

Family Feud

Perhaps the most heated debate in contemporary dinosaur research concerns how the theropods, sauropodomorphs and ornithischians are arranged on the family tree. In 1887 British paleontologist Harry Govier Seeley surveyed the flood of new fossils from Europe and the American West and argued that dinosaurs could be separated into two distinct types, based on the structure of their hip bones. Theropods and sauropodomorphs both have a pubis bone pointing forward, as modern lizards do, so he placed them together in a group he called Saurischia—the “lizard-hipped” species. Ornithischians, with their pubis projecting backward like that of modern birds, were deemed a separate branch of “bird-hipped” dinosaurs. This basic dichotomy persists today as the standard dinosaur classification scheme that I and all my fellow dinosaur hunters learned as students.

It might be incorrect, however. In a bombshell study published in Nature early last year, University of Cambridge Ph.D. student Matthew Baron and his colleagues presented a new dinosaur genealogy based on an analysis of an expansive data set of early dinosaurs and their anatomical features. Their tree links together theropods and ornithischians into a group they call Ornithoscelida, with sauropodomorphs perched outside on a separate limb. Instead of saurischians versus ornithischians, the new dinosaur dichotomy is ornithoscelidans versus sauropodomorphs.

Or maybe not. Soon after Baron’s study was released, I was approached by Max C. Langer, a Brazilian paleontologist who has described a slew of new Triassic dinosauromorphs and dinosaurs from his homeland over the past decade, including Ixalerpeton (a dinosaur precursor very similar to the one that left the Prorotodactylus tracks from Poland) and Saturnalia (a dog-sized protosauropodomorph). He was skeptical of the new genealogy and recruited a team of experts on early dinosaurs to pore over Baron’s data set. Because I had studied the Polish trackways and other key Triassic fossils, Langer asked me to be part of the group. For a month we carefully picked through the data set and noted our various disagreements about how the other team had characterized certain features. We then reran the analysis of the traits with our corrections. The resulting family tree shifted back to saurischians versus ornithischians, although statistical tests showed that this arrangement was not a significantly better fit to the data than Baron’s ornithoscelidan versus sauropodomorph tree. We presented our results in a follow-up Nature paper in the autumn of 2017.

What this ambivalence in the results means is that paleontologists do not currently have a good understanding of the basic shape of the dinosaur tree. It seems that the rush of new discoveries in Argentina, Brazil, Poland and elsewhere over the past 15 years has muddied the picture. We now realize that the earliest members of the three major dinosaur lineages were remarkably similar in body size and anatomy, which makes untangling their relationships difficult. This puzzle is ripe to be solved by the next generation of paleontologists, probably the way these arguments are usually settled: with new fossils. —S.B.

Credit: Portia Sloan Rollings, Graphics by Jen Christiansen; Sources: “Untangling the Dinosaur Family Tree,” by Max C. Langer et al., in Nature, Vol. 551. Published Online November 1, 2017; “Baron et al. Reply,” by Matthew G. Baron et al., in Nature, Vol. 551. Published Online November 1, 2017