The rhynchosaur Hyperodapedon from Elgin in Scotland – a page from my doctoral thesis.
Genus:
Hyperodapedon
Species:
gordoni
Named by:
Thomas Huxley, 1859
Age:
Late Triassic, 237–227 million years ago
Fossil location:
Scotland
Classification:
Archosauromorpha: Rhynchosauria
Length:
1.3 m (4¼ ft)
Weight:
50 kg (110 lbs)
Little-known fact:
Hyperodapedon lived worldwide and is known from Argentina, Brazil, India, and Tanzania.
Understanding the adaptations and the world of the rhynchosaurs was important, as they had been the dominant plant-eaters before the dinosaurs exploded onto the scene. How quickly were they replaced, and were they pushed out by the dinosaurs, or did they simply die out for other reasons?
As I came to finish my doctorate, I faced a dilemma. Rhynchosaurs were lovely, or so I thought, with their happy smiles and precise shearing jaw action, but they were all the same. Hundreds of skeletons of rhynchosaurs had been found, not only in Scotland, but also in Triassic rocks of similar age in Brazil, Argentina, India, Tanzania, Zimbabwe, Canada, and the United States. At first, these different finds accumulated several different names, but they have all been re-studied, by me and others, and we struggled to highlight any differences. The rhynchosaur Hyperodapedon lived worldwide in the Late Triassic at the same time as what were then the world’s oldest dinosaurs.
What was the first dinosaur?
Up to the year 2000, the oldest known dinosaurs were all from the Late Triassic, and dated at about 230 million years ago. The oldest decent dinosaur specimens were found in the Ischigualasto Formation of Argentina in the 1950s and 1960s, when Al Romer from Harvard and local Argentinian geologists began excavating. The Ischigualasto country lies in sight of the Andes Mountains, and has been uplifted on the flanks of that great mountain range. Geologists travelled 200 kilometres (124 miles) north of the regional city of Mendoza in San Juan Province, first on passable roads, and then on dirt tracks as they got close to the dinosaur sites. The Ischigualasto landscape consists of wide, bare valleys, eroded by seasonal floods tearing down the eastern flank of the Andes, and exposing broad strips of badland scenery, with sharp ravines cut in the rock, revealing the mix of red- and grey-coloured sandstones. The fossil beds are in the Ischigualasto Provincial Park, located in the romantically named Valle de la Luna, the Valley of the Moon. Picking over these barren landscapes is hard work, but it’s ideal fossil-hunting territory as there is no soil or vegetation: the fossils stand out as white and purplish bones in the rock.
The fossil collections made by Romer went back to Harvard and he and his students published a series of papers describing the fossils, including the dinosaur Herrerasaurus (see overleaf). This was named by Osvaldo Reig, a famed Argentinian palaeontologist, in 1963. Herrerasaurus was a large animal, some 6 metres (20 feet) long, with great meat-cutting jaws. It was a biped, clearly capable of fast movement on its powerful, upright hind legs, each equipped with broadly spreading toes. It had long, powerful arms, and could have used these to grab prey. Its jaws were lined with twenty-five scimitar-like teeth, each with a serrated edge, like a steak knife. It pains me to report that Herrerasaurus was probably large enough to feed on the most abundant animals of its day, the rhynchosaurs. Other animals in the Ischigualasto rocks include smaller animals such as the dinosaurs Eoraptor and Panphagia, each about 1 metre (3 feet) long, as well as armoured plant-eating early archosaurs such as the aetosaurs, and some smaller carnivorous synapsids that probably looked like partly hairy, partly bald rats.
Scientific expeditions to Ischigualasto Provincial Park in the 1990s revealed dozens more dinosaur skeletons, including fairly complete skeletons of Herrerasaurus and Eoraptor. The Ischigualasto dinosaurs, dated as about 230 million years old, are matched by smaller faunas of similar dinosaurs from rock formations of the same age in Brazil, India, and North America, which is why I took them as the marker of an explosive worldwide diversification of dinosaurs following a major environmental crisis.
Then, after the year 2000, a series of new discoveries rather suddenly pushed the date of the origin of dinosaurs back by 15 million years, and placed it in an entirely new and unexpected context.
The first hint of the revolution in our understanding came from Poland. In 2003, Jerzy Dzik, Director of the Palaeontological Institute in Warsaw, reported a skinny reptile from southern Poland, called Silesaurus (see overleaf). The fossil was remarkably complete, some 2 metres (6½ feet) in length, with a long, slender body, long, thin arms and legs definitely held in the erect posture, and a long neck and sleek head. It looked as if it mainly ran as a biped, and could then use its long, slender arms to walk slowly on all fours. The jaws are lined with peg-like teeth, and there is a bony lip at the front of the jaws, so it seems that Silesaurus was a plant-eater that nipped at leaves with the bony tips of its jaws, and chopped the food further back in the jaws. Silesaurus sort of looks like a dinosaur, but not quite. Could it point back to the ancestry of dinosaurs?
Genus:
Herrerasaurus
Species:
ischigualastensis
Named by:
Osvaldo Reig, 1963
Age:
Late Triassic, 237–227 million years ago
Fossil location:
Argentina
Classification:
Dinosauria: Saurischia: Herrerasauridae
Length:
6 m (20 ft)
Weight:
270 kg (595 lbs)
Little-known fact:
Herrerasaurus looks like a theropod, but appears to be an early saurischian, neither theropod nor sauropodomorph.
Genus:
Silesaurus
Species:
opolensis
Named by:
Jerzy Dzik, 2003
Age:
Late Triassic, 227–201 million years ago
Fossil location:
Poland
Classification:
Dinosauromorpha: Silesauridae
Length:
2.3 m (7½ ft)
Weight:
40 kg (88 lbs)
Little-known fact:
The fossils come from a clay pit used by a cement company.
The second Polish surprise came in 2011, when fossilized slender, three-toed footprints from several localities were reported as definitely dinosaurian by Steve Brusatte, Grzegorz Niedźwiedzki, and Richard Butler. Their discovery was disputed – can we be sure these little footprints were really made by a dinosaur, or could they perhaps have been made by something like a dinosaur, maybe even a silesaurid? Well, yes, possibly, but in a way it doesn’t matter.
Silesaurus, a member of the group that is closest to the dinosaurs in evolutionary terms.
The clincher came in 2010, when Sterling Nesbitt reported a Middle Triassic silesaurid from the Manda Formation of Tanzania, called Asilisaurus. The Manda Formation consists of red-coloured sandstones laid down in ancient rivers, and the rocks are now exposed under the thin soil and the burning hot sun of southwestern Tanzania, near the shores of Lake Malawi. The first fossils were found there 100 years ago, but renewed explorations by Sterling Nesbitt and his team have revealed many remarkable new specimens.
The discovery of Asilisaurus unequivocally re-dated the origin of dinosaurs back from 230 to 245 million years ago, or older. The point was that the Polish slender dinosaur-like animal Silesaurus was not alone. In fact, it turns out that Silesaurus is the exemplar of a whole new group, named the Silesauridae in 2010. Half a dozen little animals from the Middle and Late Triassic of South and North America were assigned to this group…and then along came the oldest silesaurid, Asilisaurus. All these little animals looked somewhat like dinosaurs because it turned out that the Silesauridae were th
e nearest relatives of the Dinosauria (the formal name for dinosaurs, invented in 1842, as we shall see in Chapter 2), meaning they shared an immediate common ancestor. If the Silesauridae had originated by 245 million years ago, then their immediate relatives, the Dinosauria, must have too. There is even a possible dinosaur, Nyasasaurus, from the Manda Formation, but it is known only from isolated bones.
The macroecology of dinosaur origins
If the dinosaurs originated in the Early Triassic rather than the Late Triassic, then this shifts their time of origin back into one of the most turbulent periods in the history of life. This was when life was recovering from near-complete annihilation, and Earth environments were convulsed repeatedly by terrible episodes of acid rain, global warming, and loss of oxygen from the ocean floors. It all started 252 million years ago during the largest mass extinction of all time: the Permian–Triassic mass extinction.
The extinction had been triggered by huge volcanic eruptions in what is now Siberia, which drove an episode of profound environmental destruction – acid rain and extreme warming on land destroyed the forests, and plants and soil were washed into the sea, leaving gaunt, rocky, and baked landscapes. Shallow seas were swamped with the debris and acid and warming, and this perturbed normal ocean cycles. Life on land and in the sea was destroyed, and only 5 per cent of species survived.
Normally, after a mass extinction event, life can recover in a reasonably benign world. The Early Triassic world was far from benign, however. For 6 million years after the crisis, repeated paroxysms of eruptions and environmental destruction occurred. Life would recover for half a million years, and then it would be set back again. Into this perturbed world came the first dinosaurs, taking their chance against other groups in the bleak post-extinction environment.
In my 1983 paper, I had challenged the competitive model for dinosaur success and suggested an alternative extinction model. To test this hypothesis, I had documented occurrences of fossils, identified what they were, and matched them to geological time as best I could. This at least allowed me to test the pattern of change. My data showed that there had been a pretty sharp change in the composition of reptile faunas through the Triassic. Romer, Colbert, and Charig had been right that we started the Triassic with faunas of synapsids, we passed through faunas dominated by rhynchosaurs, and ended the Triassic with dinosaurs everywhere. But the change happened quickly, in one event, some time around 230 million years ago.
My model was explicitly ecological. This meant I didn’t simply record the presence or absence of different species, but I also wanted to document how ecologically important they were. This required some knowledge of their size and likely diets, but also of how abundant they were. In other words, in any particular location, out of 100 specimens, how many belonged to each group? I found to my surprise that, when there were rhynchosaurs, they typically made up 50 per cent or more of the fauna. Indeed, at Elgin, and in some other locations, they could represent 80 per cent or more of all specimens.
This attempt to represent ecological importance in terms of relative abundances showed that rhynchosaurs were the dominant herbivores all round the world, and then at a particular point, 230 million years ago, they disappeared. This was the point. They didn’t dwindle from 80 per cent to 40 per cent to 20 per cent as you passed up through the rock sections. One minute they were there, and then a few metres higher in the rock section they were gone. It might have been recorded simply as the loss of one or two species worldwide, but ecologically those few species of rhynchosaurs had dominated their ecosystems, and their loss must have left a large gap. (We shall return later to a likely reason as to why the rhynchosaurs suddenly vanished.)
That was the crux of my argument in 1983. Sudden death of the dominants, ecologically speaking, followed by the rise of the dinosaurs. The dinosaurs were already there, as we have seen in the Ischigualasto Formation, and they were diverse and important, but they made up only 5–10 per cent of the fauna. After the disappearance of the rhynchosaurs, dinosaurs switched from representing 5–10 per cent of their faunas to more than 50 per cent in many parts of the world.
New methods and new models
Advances in palaeontology don’t all come from new fossils. There are also advances in computational methods and, although they might seem less exciting because they do not involve Land Rovers, sweat, and exotic desert locations, they can be crucial in resolving questions.
When I did my early studies on Triassic ecosystems, the numerical tools were quite feeble. I could only use simple statistics, such as counting up proportions of specimens to describe what was going on. Now, we have access to new mathematical methods that were developed to allow biologists to make comparisons between modern species while properly taking account of their evolutionary relationships. They also allow biologists to work out the so-called ancestral state of any character or trait, which can be a physical feature, such as body size or leg length, or a behavioural feature, such as egg clutch size or feeding ritual. By plotting the known data onto an evolutionary tree, they can estimate down the tree what the ancestors would have been like, and then use this knowledge of calculated ancestral states to look at rates and types of change through geological time.
In an example of the application of these new numerical methods, the Colbert–Romer model for Triassic ecological relay between synapsids and archosaurs (including dinosaurs) could be explored. Roland Sookias did this as part of his doctoral work in 2012. He documented the body size of several hundred synapsids and archosauromorphs (archosaurs, rhynchosaurs and relatives), and tracked the changing sizes through time. He found that archosauromorphs became larger through the Triassic, mainly represented by the dinosaurs, and synapsids became smaller as they shrank down to the tiny shrew-like mammals at the end of the Triassic. But was this a driven trend or just passive change?
Working out the ancestral states of plants that are pollinated by insects, birds, or the wind.
Sookias was able to fit different models of evolution to his data, and mostly it seemed that evolution was simply proceeding in a random way. That is, the changes in body size were surely happening, but with sufficient local variations, and sufficiently slowly, that it was not possible to say that the changes in body size were driven by some powerful evolutionary force. If the size change through the Triassic had been a driven trend, it would have been possible to make a case that natural selection was in play, and there was selective pressure for larger size. All Sookias could say was that the synapsids were getting smaller, on the whole, and the archosaurs were getting larger, but they could have been changing size in more or less random ways. Therefore, this did not disprove the Romer–Colbert model in which dinosaurs outcompeted their precursors, but it didn’t lend the idea any support either.
In an earlier study, Steve Brusatte, then a Masters student in Bristol, studied the same question, but looking more closely at the first dinosaurs and the early archosaurs they replaced. He decided to measure rates of evolution not just from one trait, such as body size, but from all aspects of their anatomy. He made a huge table of 500 traits for each beast, and used standard statistical methods to reduce this great mass of data into something more manageable.
One way of visualizing such huge and complex data sets is to seek some main directions of variation, and extract these and plot them up as a so-called morphospace, meaning literally ‘shape space’. A morphospace is a graphical way to show how the morphology, or external appearance and physical characters, of organisms vary, and it has the great strength of summarizing huge amounts of information into something we can more easily understand. Species that are most similar plot out close together, and those that are most different in life are plotted far apart in the morphospace.
While archosauromorphs (top line) became larger during the Triassic period, synapsids (lower line) became smaller.
A diagram representing the morphospace of dinosaurs and other Triassic reptiles, corresponding to their adaptations.
&nb
sp; The morphospace for the dinosaurs and early archosaurs showed areas of morphology occupied by each group. They did not overlap – possibly, but not conclusively, suggesting they were not directly competing. When Brusatte extracted rates of change, he found that dinosaur morphological variation expanded as the group diversified in the Late Triassic, but so too did that of their supposed competitors, the crocodile-like archosaurs, or crurotarsans. There was no sign that the explosion of dinosaurs was hitting the other archosaurs and forcing them into retreat. In fact, they all seemed to be diversifying and occupying new morphospace, in parallel; far from being crushed by the new dinosaurs, the other archosaurs were apparently flourishing. In their ecosystems, the new dinosaurs such as Herrerasaurus and Plateosaurus were still preyed upon by some of their crurotarsan cousins.
Dinosaurs Rediscovered Page 3