These numerical studies are part of a new wave of computational research in macroevolution. The methods are tough to learn, but my students absorb them like the air they breathe. These methods have opened the floodgates to enable new explorations of the various phases of the evolution of dinosaurs, and how different groups came and went, and are providing ways to tackle big evolutionary questions that were thought impossible to answer when I began to work on these questions back in the 1980s.
The origin of dinosaurs as a three-step process
So, where do the new discoveries of much older dinosaurs and the new computational studies leave us? Do we have a proper understanding of the dynamics of the origin of the dinosaurs? Who was right –was the Romer–Colbert–Charig ecological-relay model, with a drawn-out, competitive rise of dinosaurs correct? Or was my 1983 mass extinction-opportunism model correct?
In fact, we all got it wrong in different ways. I was wrong to assert that dinosaurs had emerged and exploded in the Late Triassic, as we now know their first steps took place 245 million years ago, in the Early and Middle Triassic. Romer and Colbert were right that the origin and initial rise of the dinosaurs lasted throughout some 40 million years of the Triassic, although they had not based this on any knowledge of more ancient dinosaur specimens.
The erect posture of dinosaurs was clearly key to their success, again as Romer and Colbert had said, but in a somewhat different sense from the one they had argued. In other words, if the first dinosaurs arose much earlier in the Triassic than anyone had known back in the 1970s and 1980s, then they did not out-compete the synapsids, early archosaurs, rhynchosaurs, or any other groups. In evolution, organisms mostly avoid competition by shifting their ecological niches – they choose a different diet or geographic range. Discretion is the better part of valour; successful organisms live to fight another day. Evolution probably isn’t exactly ‘red in tooth and claw’, as Alfred Lord Tennyson suggested; more pinkish in tooth and claw.
What set off the big burst, though, the second stage? If this hadn’t happened, dinosaurs might have continued to be rather rare animals, maybe 10 per cent of their faunas. I might have got it right back in 1983. However, new evidence is showing an ever-closer link between the dinosaur explosion and the so-called Carnian Pluvial Episode. This event had been noted and named by Mike Simms and Alistair Ruffell back in 1989. They had observed in Late Triassic rock sections around the UK, and other parts of Europe, that something unusual was happening. The generally dry climates were interrupted by a pluvial phase, when rainfall increased substantially, and then conditions returned to dry. The evidence for the climate change came from the rocks themselves and from the plant remains, which can be classified as wet-loving (mosses, liverworts, horsetails) or dry-loving (conifers).
There things rested until 2012. The dinosaur explosion–Carnian Pluvial Episode link was emphasized from time to time by me, by Simms, and by Ruffell, but nobody else paid much attention. Then, independent work published in 2012 by Italian geologist Jacopo Dal Corso changed everything. Dal Corso found that the rocks documenting the Carnian Pluvial Episode reflected the results of major volcanic eruptions in western North America. There, about 232 million years ago, great eruptions produced huge volumes of volcanic lava, the Wrangellia basalts, seen today around Vancouver and northwards along the coast of British Columbia.
Dal Corso argued that the eruptions had been so huge that they had caused a shock climate change worldwide. Just as at the end of the Permian 252 million years ago, huge volumes of carbon dioxide pumped out of the volcanic vents caused global warming, as well as acid rain. The warming and acid rain killed life on land, and led to ocean acidification and loss of oxygen in bottom waters, and Dal Corso noted extensive evidence of extinctions in the sea in localities across Europe and North America. The warming also led to mega-monsoonal conditions around the broad equatorial belt, which at that time encompassed all the dinosaur localities in North and South America, Europe, and India. After the eruptions ceased, conditions returned to hot and dry, and this was the killer.
Changing climate belts in the Late Triassic, and the movement of dinosaurs from the southern continents to the whole world in the later period.
The break point in ecosystem evolution was triggered by the Carnian Pluvial Episode 232 million years ago.
These new studies stimulated me to look again at the ecological data I had collected back in 1983. I was aided by students Cormac Kinsella and Massimo Bernardi. We checked all the data, and improved the sample of specimens to over 7,700, and worked out the per centages of all the key groups – rhynchosaurs, dinosaurs, and the others. Then, I plotted the ecological ratios as a so-called bubble plot, where each bubble represents a different fauna, with the centre dated correctly according to the geological time scale. The size of each bubble represents the number of specimens in the sample, and I plotted each one against the value of a ratio that essentially gives us a measure of how much dinosaurs dominated the faunas.
There did seem to be a step up from values around 20 per cent to values around 70 per cent, but such a guestimate wasn’t good enough. Critics might say we were just imagining things. So, we applied a numerical method called ‘breakpoint analysis’ to the data. This method attempts to calculate the line that best explains all the data, but you allow it to make one (or more) breaks. We programmed the model to make one break, and set the calculations running. After a short time, the answer came back – the best-fitting line had a break at exactly 232 million years ago.
This we took as independent evidence that something happened at this time that fundamentally reset the nature of Triassic reptile ecosystems. Dinosaurs had indeed originated long before the Carnian Pluvial Episode, but they had not succeeded in taking over the faunas – in fact they continued to do what they had done before, and would have seemed inconsequential to any observer. The key jump happened 232 million years ago, and the new study provided a marker that linked the ecological revolution when dinosaurs exploded in significance to the environmental upheavals in the middle of the Carnian.
Cementing the link between the Wrangellia eruptions and the Carnian Pluvial Episode was a powerful idea, and it confirmed the environmental shock that perhaps killed off the rhynchosaurs and other dominant animals, so providing the opportunity for dinosaurs to diversify explosively into empty ecospace. Still, the dating needed to be right, and, usefully, there had been step-change improvements in our understanding in recent years.
Dating dinosaurian diversification
So far, I have been quoting geological ages (see Timeline, pp. 6–7) such as 230, 232, and 245 million years ago. Geologists do that. But how do we know? This is crucial for all hypotheses we might have about dinosaurian origins and their eventual disappearance. We have to be able to set the time of ancient events in terms of millions of years, and also match or correlate the rocks from continent to continent to test whether a dramatic event in, say, Argentina is matched by a similar-looking crisis in the Italian rock sections.
Dating the rocks is a core activity by geologists. The roots of the science of dating the rocks, called stratigraphy, were profoundly practical. When the humble English surveyor William Smith began his efforts in the 1790s, he was self-taught. He was one of the first economic geologists and he was paid for results. At that time, there was no evidence to suggest that the rocks under our feet were anything other than a complete jumble. The idea of a geological map that showed the orderly arrangement of different rock formations, and of a time scale that placed them in sequence and could be used across various locations, was unheard of. Smith laboured through his life to establish both principles.
Smith practised in the early years of the Industrial Revolution in Britain, when every land owner was delving for coal – often completely at random. The reasoning was something like this: my neighbour Arkwright finds coal at twenty yards beneath his fields, so I should also find coal at the same depth. Sometimes it worked, and sometimes not. If the two spots were
separated by a fault, the rock succession could be entirely different. Smith used his skills in mapping and stratigraphy to work wonders: he could tell people where to dig, and – importantly – where not to dig. If the rocks were dated as Jurassic, there might be coal below because the Jurassic period is younger than the Carboniferous, whose forests and swamps became the coal seams of the future. If your neighbour’s rocks were Silurian in age, however, then there would be no coal; the Silurian is older than the Carboniferous. The Jurassic and Silurian rocks might both be rather similar-looking, dark grey limestones, but the fossil content told Smith the age, and he could then turn this knowledge into cash.
Since Smith’s day, and thanks to efforts in every land, the geological time scale with its main divisions, the geological eras and periods, were all more or less named by 1840. The geological time scale worked everywhere, and correlations were made regionally by walking across country and mapping, and then internationally by comparing fossils. Smith’s Jurassic faunas of ammonites and bivalves could be matched around the world. This marked a line of equivalent age from England to France to Russia to Argentina…Such efforts in the identification of rock ages by fossil assemblages are just as commercially valuable today as in Smith’s day, now especially in the oil industry, where companies spend billions of dollars drilling. They want to know ahead of time what they are going to be drilling through, and whether they have to drill 50 metres or 5 kilometres to reach oil.
Stratigraphy using fossils does not give exact ages. These come from radioisotopic dating. Soon after the discovery of radioactivity in the 1890s, the Nobel-prize-winning physicist Ernest Rutherford suggested in 1905 that radioactive decay could provide an exact chronometer for dating rocks, and so for dating the origin of the Earth, which then points to the date of the origin of the universe. The eager young geologist Arthur Holmes seized on the idea, and he had compiled a list of key dates by 1911, at the age of twenty-one. Since 1911, radioisotopic dating has become an important part of laboratory-based geology, with ever more powerful mass spectrometers being deployed. A great strength of the approach is that the same rock can be dated by different means and in different laboratories to cross-check the estimates.
There is an extensive international endeavour to improve precision (tightness of the estimate; size of error bars) and accuracy (is it right?) of exact rock dates, and the standard geological time scale is revised in detail every few months, as dates are tuned to be sharper and sharper, and more comparable. When I started my studies of geology in the 1970s, we were told to assume an error of plus or minus 5 per cent on any radioisotopic date. In some cases now, precision has improved a hundredfold, to an error of plus or minus 0.05 per cent. So, the dating of the Carnian Pluvial Episode might have improved from 232 ± 11.6 million years ago (Mya) to 232 ± 0.116 Mya. An error of 116,000 years may still sound ridiculously long – but for a geologist it’s a miracle!
The Wrangellia basalts can be dated directly because they are igneous rocks. They were molten and they solidified, and crystals within them can thus be dated to give the time at which they solidified. The sedimentary rocks that document the episode are harder to date directly in this way. However, in the Dolomites in northern Italy, there are wonderful sequences of marine sediments that have been dated into small zones of less than 1 million years each. Interleaved between these are terrestrial sediments with footprints, and these document how dinosaurs were absent before the Carnian Pluvial Episode, but then present in abundance after the event. We used these amazing fossil sites to argue that the Carnian Pluvial Episode triggered the second phase in the origin of the dinosaurs, their great explosion 232 million years ago, in a 2018 paper, led by my former doctoral student Massimo Bernardi, now curator of geology at the Museum of Sciences in Trento in north Italy.
The cross-dating between marine and non-marine sediments in north Italy is confirmed by a remarkable, and independent, method called magnetostratigraphy. This method relies on the fact that the Earth’s direction of magnetization has flipped from north to south, and back again, many dozens of times through Earth’s history. Nobody can quite explain why north flips to south, and what happens during the flip. Yet the record is there in magnetic minerals in the rocks, and the alignment of the striped barber’s poles of normal-reversed-normal-reversed can fix the relative ages of rocks of all kinds.
Earlier, I suggested that the diversification of dinosaurs occurred in three phases. We’ve looked at the first two – their origin about 245 million years ago in the maelstrom of recovery from the Permian–Triassic mass extinction, and their explosive diversification 232 million years ago following the Carnian Pluvial Episode. The third phase followed the end-Triassic mass extinction 201 million years ago, and we will explore this a little more in the next chapter.
How can we identify ancient climates?
In describing the origin of dinosaurs, I have freely talked about arid and monsoonal climates. How are these important conclusions reached? The basics go back to the beginnings of geology. Sedimentology is the science of understanding sediments and reconstructing ancient environments. First-year geology students learn to distinguish marine and non-marine rocks. Marine rocks, for example, uniquely contain minute fossils of plankton, and larger fossils of animals that only lived in the sea, such as brachiopods, sea urchins, or sea lilies. Rocks deposited in lakes or rivers may, on the other hand, contain leaves, insects, or dinosaurs. Of course, leaves, insects, and dinosaurs can be washed down rivers into the sea, but it’s the predominant fossils, not the rarities, that count. The rocks of the Ischigualasto Formation, for example, include tree trunks and leaves from conifers and other plants, which confirms they were deposited on land. The sediments are red-coloured mudstones and sandstones. In places, there are large channels formed by ancient rivers, and muds deposited in temporary lakes. There are also burrows constructed by some of the smaller reptiles, and this confirms chemical evidence that the area experienced really hot conditions at times, which drove the little beasts underground for protection.
There are all sorts of other clues in the rocks. For example, sand dunes of certain types mean deserts. Channels and stacked sediments can identify meandering rivers, flipping from side to side through time. Layers of salt can indicate coastal pools drying under a hot sun.
There are also chemical indicators of ancient conditions. For example, isotopes of oxygen measured through whole rock sections can track rising and falling temperatures. The ratio of oxygen isotopes varies with temperature because of differential effects during evaporation from the surface of a pond or as rain falls, and the isotope signals can also reflect salinity and the volume of water locked up in ice sheets.
Environments of deposition can be identified locally, but what about the worldwide picture?
How different was the Triassic world from ours?
The Earth is composed of numerous great tectonic plates, which are constantly in motion. Some of the plates underlie the continents, and others make up the sea floor. The engine for plate movements comes from the molten mantle of the Earth, which lies below the solid plates. Great convection currents rotate within the magma, transferring their lateral motion to the solid crust. In places, molten material from the mantle comes to the surface, such as along the great mid-ocean ridges. Down the centre of the North and South Atlantic is a continuous system of fissures through which basalt lava bubbles from time to time. The mid-Atlantic ridge surfaces on Iceland. The constant supply of fresh crust in the centre of the Atlantic, and in similar ridge systems in the Pacific and Indian oceans, drives the ocean floor plates apart at the rate of about 1 centimetre (3/8 inch) per year. Where plates move past each other, there can be great faults, such as the San Andreas Fault through California that periodically judders into life. This is real-time evidence that the Earth’s crust is in constant motion. In other places, oceanic plates dive beneath continental plates, such as along the Pacific coast of South America.
During the Triassic, all continenta
l plates were fused together as the supercontinent Pangaea. Also, there was no land across the poles, so there were no ice caps. This means that temperatures from equator to pole varied less than they do today, and the climate is often said to have been equable. Uniform climates over a single land mass meant that the early dinosaurs and other land animals and plants could be much more widely distributed then than now.
Continental drift from the Permian to the present day.
In the Late Triassic, as we have seen, there was a violent set of huge volcanic eruptions along the west coast of Canada, pouring out the magma that solidified as the Wrangellia basalts. Ten million years later, another set of similarly huge volcanic eruptions were set off in the middle of Pangaea, along the rift line of a new ocean that was jerking fitfully into being. These eruptions also produced great thicknesses of basalt, most famously seen today in the Palisades along the Hudson River between New York and New Jersey.
All down the coastal strip of eastern North America, there are rift valleys, extending from Nova Scotia in the north and east to North Carolina in the south. These rift valleys formed as the Earth’s crust was splitting by the force of great convection cells in the mantle that were pulling what is now Europe and North Africa eastwards, and what is now North America westwards. Even though the rate of movement was only 1 centimetre per year, over thousands of years the tension became too great, and the crust ripped apart – just as is happening today in the Great Rift Valley of east Africa, as that continent tears apart too.
Dinosaurs Rediscovered Page 4
