This, then, is the origin of the model that a 7-kilometre-wide (4-mile) asteroid hit the Earth, vaporized, and threw an immense cloud of ash into the upper atmosphere, which encircled the globe and blocked out sunlight, so photosynthesis in green plants ceased, and life consequently died on land and in the oceans.
The publication caused uproar. Geologists were incensed – who was this crazy physicist telling us what to think? We all know asteroids never hit the Earth; this goes against Lyell and uniformitarianism. Palaeontologists, with their hangups, were defiant too. Bob Bakker said what many were thinking: ‘The arrogance of those people is simply unbelievable. They know next to nothing about how real animals evolve, live, and become extinct. But despite their ignorance, the geochemists feel that all you have to do is crank up some fancy machine and you’ve revolutionized science.’ Well, Bakker was wrong, as were many (perhaps most) other palaeontologists and geologists at the time. The impact really did happen, as we now know, based on hard evidence from field research, as we shall see.
Periodicity and nuclear winter
The idea of an asteroid impact 66 million years ago immediately spawned another, perhaps more startling consequence. If an impact happened once, why not many times? The suggestion was made in 1984 by David Raup and his colleague Jack Sepkoski, based on their preliminary analysis of the fossil record. They focused on the past 250 million years, and plotted a measure of extinction through time. They expected to see times of low and high extinction, but were surprised to see what looked like a regularly repeating signal of extinction peaks. The raw measurements suggested that the peaks repeated every 26 million years, but Raup and Sepkoski had to test this, so they applied a numerical analysis to assess whether the pattern could have arisen by chance or not – the repeat pattern, termed periodicity, was significant to a high degree of probability.
Astronomers were excited by the Raup and Sepkoski data because any phenomenon with a periodicity, or repeated pattern, on the scale of 26 million years was likely to have an astronomical driver. Astronomers considered three main theories: that the entire solar system was tilting up and down like a wobbly plate; or that there was a sister star to the sun called Nemesis; or that there was a tenth planet, called Planet X, lying at the edges of the solar system. In all three cases, the perturbations affected the outer fringes of the solar system, where the Oort cloud of comets is located. The perturbation sent comets flying into the heart of the solar system, and one or more would have hit the Earth.
The periodicity in mass extinctions of 26 million years, as proposed by Raup and Sepkoski.
If the periodic pattern were true, the next flurry of impacts could then be predicted. The last one happened 14 million years ago, so the next one would be in 12 million years’ time – a great way to test the hypothesis. I remember, with fellow palaeontologists, being amazed, and slightly awed, by the fact that a relatively straightforward synoptic diagram taken from our knowledge of the fossil record could have set loose such amazing speculations about the functioning of the Earth and the universe.
From the start, geologists and palaeontologists pointed out flaws in the line of reasoning. The periodic signal depended on a particular dating of the geological record, and slight revisions would break up the tightness of the 26-million-year period. They also noted that the last event is barely supported by any data – and yet that should have been the clearest to see in the rock record, because it is nearest to the present day. Further, the matching of events to the regular periodicity broke down in the Jurassic and Cretaceous. The debate rumbles on, with revivals of the idea in papers in 2016 and 2017, but most have abandoned the idea of periodicity.
One other consequence of the Alvarez model has had more traction, and that is the idea of nuclear winter. Three years after publication of the Alvarez paper, several climatologists began to speculate about the effects of all-out nuclear war. Richard P. Turco coined the term ‘nuclear winter’ in 1983 to describe the main outcome of mass bombing, which would be the lofting of ash into the upper atmosphere that would blot out the sun, leading to freezing conditions as the warming effect of sunlight was removed. Quickly the climatologists, modellers, and futurists saw the parallels with the Alvarez extinction model, and the assumptions are all now accepted, both for the impact at the end of the Cretaceous, and for the consequences of a similarly massive energy release from explosion of the Earth’s nuclear arsenals. Periodicity may have bitten the dust, but nuclear winter and impact killing of dinosaurs survived scrutiny. Then the crater was found.
The killer crater
The impact theory, periodicity, and nuclear winter idea set scientists, and the public, talking, and in 1985 the BBC made a Horizon programme about the proposed end-Cretaceous asteroid impact. The journalists asked the rather obvious question: where was the crater, the smoking gun? At the time, the geologists could say little more than that the crater might well have been lost somehow – which wasn’t a hugely satisfactory answer. Even then, however, the trail of detective work was pointing to where the crater must be.
Geologists had noticed that there were strangely perturbed rock units at the Cretaceous–Palaeogene boundary in rock sections throughout coastal areas in Mexico and along the Brazos River in Texas. In the midst of orderly, flat-lying beds of limestone and mudstone were levels where the limestones had apparently been ripped up and dumped higgledy-piggledy. These were called storm beds, or even tsunami beds, or tsunamites. The idea was that the rocks had been torn asunder and dumped along the coastline of the proto-Caribbean, which lay inland along a line arcing through Mexico and the southern United States. If these geologists were right, then it implied there had been an impact out in the ocean, and a huge shock wave radiating outwards, in the form of a tsunami front, many tens of metres high, that beat on the shores and smashed up the freshly deposited rock layers.
The next clue came in 1991 from close study of a rock section across the Cretaceous–Palaeogene boundary at Beloc on the Caribbean island of Haiti by geologists Florentin Maurrasse and Gautam Sen. They noted that the boundary bed was 72.5 centimetres (28½ inches) thick, not 1 centimetre, as at Gubbio and Stevns Klint in Europe. Geologists read this great thickness to indicate that the source of the asteroid impact was not so far away. In the lower layers, the sediment was stuffed full of glassy spherules of exotic geochemical composition – they were interpreted as impact glasses, which were formed by high-pressure and high-temperature effects at an impact site some distance away. These small glass beads were thrown high in the air, and carried in the atmosphere, together with other impact debris, for some 1,000 kilometres (620 miles). Glass spherules are commonly thrown out by volcanoes, but they have the chemistry of an igneous rock, such as basalt or andesite, matching the molten lava. The Beloc glass beads, bizarrely, had the chemistry of limestone and natural rock salt – in other words, they had come from the melting of such rocks, and they provided a direct clue to the nature of the then-unknown impact site.
The Cretaceous–Palaeogene section at Beloc on Haiti, showing spherules (bed h), tsunami beds (beds b–g), and the iridium clay (bed a).
Higher in the Beloc section, the researchers noted a layer of tsunamite, with perturbed limestone rocks thrown up, and finally the uppermost, 1-centimetre-thick, dust layer, with enriched iridium. This top layer is called the impact layer, and this is all that is found further from source, such as at Gubbio. The researchers identified that the Beloc boundary beds showed two thick layers at the base that could only have been generated in a location close to the impact, and they interpreted these as showing two separate phenomena, multiple glass beads hurled through the air and falling rapidly into the sea or onto land, and the tsunami beds, later. This indicates that one impact occurred and it generated two shock waves, one that rushed through the air first with the glass beads, and the second moving more slowly through the water.
In fact, the geologist Alan Hildebrand and colleagues had already identified the actual crater and they published it
a few months later, in 1991. Hildebrand had located it in old borehole records made in the 1960s by Pemex, a Mexican oil company, drilling into an anomalous structure they detected deep beneath the Yucatán peninsula near the village of Chicxulub. Pemex quickly discovered this was not an oil trap when they hit melt rocks, and so they abandoned their efforts. Hildebrand, however, was looking for the crater, and he identified the melt rocks as typical of high-pressure impacts, where the meteorite hits the Earth, and smashes deep into the crust, vaporizing as it goes, melting the bedrock.
Hildebrand’s initial geophysical survey was confirmed by later work. Samples of meltrock gave an exact age matching the Cretaceous–Palaeogene boundary, and further geophysical survey and drilling in 1997, 2002, and 2016 have given the detail. The crater is indeed formed in latest Cretaceous limestones and rock salt, as predicted from the Beloc glassy spherules. The seismic profiles show that the centre of the crater comprises an inner ring of shattered rock with a diameter of about 80 kilometres (50 miles). This is a so-called ‘peak ring’, an inner mountainous ring seen in large craters on other planets. Round the edges of this crater is a zone of slumped rock, the terraced zone, that extends to a diameter of 130 kilometres (81 miles), marking the limit of collapse of the walls of the original crater. Further out still, at a diameter of 195 kilometres (120 miles), the investigators found a major slope that extends over 35 kilometres (21 miles) deep into the Earth’s crust and into the underlying mantle of the Earth. This outer ring is like features seen in craters on other planets, such as Venus, and it’s the first time such a feature has been identified in a crater on Earth.
The sequence of events when the asteroid hit the Earth, first penetrating and vaporizing (1), and then bouncing back, leaving a crater and debris fields (2).
The sequence of events documented by the Chicxulub crater, and predicted by Alvarez and colleagues in 1980, is that the 7-kilometre asteroid smashed into the Earth’s crust, burrowed deep, and vaporized. Within seconds, the equal and opposite recoil happened, sending vast energy vertically upwards, and radiating outwards to form the outer crater wall. The recoil headed upwards, bringing the edges of the crater hundreds of metres into the air. Because of the size of the crater and the height of its walls, it collapsed rapidly under gravity, forming the outer and inner walls. The peak ring in the middle is part of the recoil. Just as the rebound of a water droplet hitting the surface of water sends a little spout of water upwards, which then falls back, so too with the rebound phase and formation of a crater.
The classic droplet rebound, showing how the rebound from an impact can lead to a similar central structure that then collapses.
The physics of the crater and the physical consequences, in terms of the blocking out of the sun, and consequent ending of photosynthesis, and global freezing, are pretty clear. How life was killed is less certain: but these effects would be enough to finish off the dinosaurs. In addition, there were huge volcanic eruptions in India, the so-called Deccan Traps, which began half a million years before the asteroid impact, and these eruptions would have driven regional-scale warming and acid rain. Also, climate cooling had begun about 30 million years before the impact, and this would have put pressure on the survival ability of dinosaurs, which probably preferred warm-climate conditions. The interplay of these longer- and short-term crises before the impact is still debated.
But, as noted at the start of this chapter, there is no debate about when the asteroid hit – even down to the actual month.
How do we know the impact happened in June?
Jack Wolfe, a veteran palaeobotanist working for the United States Geological Survey, had studied Late Cretaceous fossil plants all his career. He was looking at a rock section in Wyoming at a site called Teapot Dome in the 1980s, when he realized he was seeing the exact minute-by-minute story of the whole end-Cretaceous impact. The Teapot Dome section spanned the boundary; there was the boundary clay, only 2 centimetres thick or so, but Wolfe found he could pull it apart millimetre by millimetre to see what had happened. And, with his unrivalled knowledge of fossil plants, he could also pin down the temperatures throughout the event.
The Teapot Dome section records events in an ancient lily pond. In the latest Cretaceous, the lilies seemed to be flourishing, and Wolfe found dozens of leaves and stems that were buried as the pond silted up. At the Cretaceous–Palaeogene boundary, he found first a thin layer with glass spherules and dust. This marked the arrival of the first phase after the impact blast. After a few millimetres of this, there was a layer with dead lily leaves. Under the microscope, Wolfe saw that the cells in the leaves had burst. This was unequivocal proof of freezing – the sap had been shock-frozen. Ice occupies more space than water, so the ice crystals had pierced the cell walls.
Above the freezing layer was a second dust layer, and this one contained the iridium spike. Then sedimentation returned to normal, the lilies recovered, and temperatures returned to 25 degrees or so. Palaeobotanists can measure ancient temperatures quite accurately if the fossil plants have modern relatives, because plants have quite specific temperature and water requirements, and these can be assumed for their ancient relatives. It is a very accurate way of determining Cretaceous climates. Still, how did Wolfe know the impact happened in June?
He used Lyell’s principle of uniformitarianism, comparing the fossil case with the modern world. The fossil lilies at Teapot Dome happen to be close relatives of the modern pond lily Nuphar of the family Nymphaeaceae. The lilies were instantly frozen at a particular stage of their development, and from the state of their buds and flowers, and by comparison with modern lilies of the same genus, Wolfe saw that the freezing had happened in June. It’s a neat example of impressive detective work by a scientist using the principle of uniformitarianism.
Was the death of the dinosaurs sudden or gradual?
We now know that all the dinosaurs – except for some bird species – died out after the cataclysm of 66 million years ago. Did they vanish with a bang or a whimper? In other words, were they in fine fettle before the asteroid struck and wiped them out, or are there signs they were in decline anyway? In support of the sudden-disappearance model are well-studied rock units that go right to the end of the Cretaceous, such as the famous Hell Creek Formation of Montana, where dinosaurs including Tyrannosaurus, Triceratops, and Ankylosaurus are found right up to the very end of the Cretaceous (see pl. iii). There’s no sign that individual faunas were becoming less diverse, that species were dropping out. Is it, however, simply a choice between a long decline or instant disappearance, or could there be a third model?
Genus:
Triceratops
Species:
horridus
Named by:
Othniel Marsh, 1889
Age:
Late Cretaceous, 68–66 million years ago
Fossil location:
United States, Canada
Classification:
Dinosauria: Ornithischia: Ceratopsia
Length:
8 m (26 ft)
Weight:
14 tonnes (30,865 lbs)
Little-known fact:
Triceratops is the official state fossil of South Dakota, and the official state dinosaur of Wyoming.
Our evidence is that it was a bit of both. We found a first clue in a study in 2008, led by my then student Graeme Lloyd, and introduced in Chapter 2. Our aim was to build a ‘supertree’ of dinosaurs, a fancy term for a phylogenetic tree of all dinosaur species that summarizes the best of current knowledge. Graeme crunched data from 200 published trees, and we produced a supertree of some 420 species. We then calculated rates of diversification, and found to our surprise that dinosaurs had done most of their fast evolving in the first 60 million years of their history. During the Late Jurassic and Cretaceous, nothing out of the ordinary happened, and we argued that dinosaurs had more or less run out of steam in their last 50 million years, except for two specialized plant-eating groups, the duck-billed hadrosaurs and the horn-f
aced ceratopsians.
Genus:
Ankylosaurus
Species:
magniventris
Named by:
Barnum Brown, 1908
Age:
Late Cretaceous, 68–66 million years ago
Fossil location:
United States, Canada
Classification:
Dinosauria: Ornithischia: Thyreophora: Ankylosauridae
Length:
7 m (23 ft)
Weight:
4.8 tonnes (10,584 lbs)
Little-known fact:
The tail club weighed about 20 kg (44 lbs) and the impact force could be as much as 2,000 newtons, equivalent to a mass of 200 kg (440 lbs).
We picked up this idea again in a further investigation in which I was involved, in 2016. We wanted to find out the deep-seated evolutionary dynamics of dinosaurs throughout their history. With colleagues Manabu Sakamoto and Chris Venditti from the University of Reading, we put together an even larger supertree of all dinosaur species, and dated it as accurately as we could. We then ran calculations to work out whether speciation and extinction rates were stable, rising, or falling through the Mesozoic. We were looking for one of three possible outcomes: that overall the balance of speciation and extinction gave ever-rising values, or levelling off, or declining values.
Dinosaurs Rediscovered Page 23
