ÐÓ°ÉÔ­´´

It is now more than ten years since the late Luis Alvarez and his colleagues came forward with an entrancing hypothesis for the demise of the dinosaurs. Some 66 million years ago, they suggested, a giant asteroid hit the Earth, destroying the dinosaurs and much other life. According to their original scenario, the impact produced a worldwide dust cloud, which blocked photosynthesis leaving plants, then higher members of the food chain, to die. In the end, all species died. All this would have happened in the space of one to ten years, a nearly instantaneous event in geological terms.

Alvarez was a Nobel laureate and a respected physicist. His theory, published in 1980, was not based on fantasy but followed the discovery of unusually large, or anomalous, amounts of a rare element, iridium, in rocks in Italy, Denmark and New Zealand, that straddled the boundary between Cretaceous and Tertiary Periods (known as the K/T boundary). Iridium is very rare in the Earth’s crust, but is much more common in meteorites and Alvarez inferred that the extra iridium in the rocks came from outer space. Not everyone agreed.

Battle was joined as soon as Alvarez published, with researchers arguing over both the record of the extinctions and what could have caused them. William Clemens of the University of California at Berkeley and David Archibald of San Diego State University, as well as other vertebrate palaeontologists, argued that the geological record showed species of dinosaurs disappearing one after the other over 500 000 years or so, (as Archibald described in last week’s issue). They found that the last appearance of the last dinosaur fossil, the ubiquitous Triceratops, came in rocks that lay below the rocks that contained the iridium anomaly. Because rocks lower in the stratigraphic column are older, Clemens and Archibald deduced that the dinosaurs died out before a meteorite impact spread iridium around the world.

Palaeontologists were not the only people to object to Alvarez’s interpretation. The huge deadly dust cloud was another sticking point. Alvarez and his colleagues estimated that it was 1000 times bigger than the cloud that came from the 1883 eruption of the volcano Krakatoa in Indonesia, with its dramatic effects on sunsets and climate the world over. But Dennis Kent, of Columbia University, noted that an even larger eruption, that of Toba some 75 000 years ago, released perhaps 400 times more dust than Krakatoa without species of plants and animals dying by the thousand.

I became interested in the controversy shortly after the original presentation by Alvarez in 1980 and the discussion raised by Clemens, Archibald and Kent. My investigations have led me to conclude that the causes of these extinctions are earthly rather than extraterrestrial. I feel that sea level changes and large volcanic eruptions originating from deep within the Earth could have disrupted the ecosystem enough to speed up the rate of extinctions, at the same time supplying the iridium from the Earth’s mantle, which contains far more than the crust. But the controversy about the K/T boundary is an illustration of how hypotheses develop in an observational science where hard facts are rare.

Certainly few facts are available to explain whatever did happen at the K/T boundary. Research in this area is like following a detective story in which everything happened long ago. The trail is littered with dead bodies but few clues as to how and why the victims died. The scattered hints that do exist are ambiguous. The dinosaurs disappeared, as did much of the ocean plankton and many shellfish that lived in shallow water, but about half the species on Earth came through the transition relatively unscathed, including birds, turtles and the mammals such as, presumably, our distant ancestors.

Although the dinosaurs have received the greatest amount of public attention in the present controversy, they are not ideal as sources of information about the extinction process. Only a few species of dinosaurs existed at any particular time, so geologists generally have to work with a limited number of specimens from isolated localities. Plankton and shellfish are more useful subjects for study, because the geological record contains many species and specimens of them.

Gerta Keller of Princeton University in New Jersey has made a detailed study of the extinction record of ocean plankton in two stratigraphic sections – one at the Brazos River in Texas, the other at El Kef in Tunisia. In both places she found that Cretaceous plankton took a few hundred thousand years to disappear. A stratigraphic section on Seymour Island in Antarctica gave William Zinsmeister of Purdue University in Indiana and his colleagues, a similar view of the disappearance of shallow water shellfish. At the Antarctic site, the shellfish died out over an even longer period of time and this event was followed by the demise of the plankton. Such palaeontological data are among the hard facts that any extinction scenario must accommodate if it is to be satisfactory.

The geochemistry that is so central to the extraterrestrial argument is another area where firm evidence is in short supply. Ever since Alvarez and his colleagues first published their ideas, controversy over the iridium layer has proliferated. Geologists have found, for example, that there could be another source for the high levels of iridium in K/T boundary sediments. Fine airborne particles from volcanic eruptions have turned out to carry far more iridium than scientists had previously expected, so high levels of this element in sediments are not categorical evidence of an asteroid hitting Earth.

Furthermore, the investigations of James Crocket of McMaster University and Robert Rocchia and colleagues of the National Centre for Scientific Research in France, based on the original sections in Italy, Denmark, New Zealand and elsewhere, have shown that the iridium does not appear as an isolated peak in the otherwise smooth distribution of the element in the geological record – the pattern expected for an impact. It appears to be spread through rocks formed over a few hundred thousand years, during which it built up to a peak concentration and then subsided.

Objections to the Alvarez hypothesis arose not only out of detailed analysis of the fossil and stratigraphic record but also from evidence that large-scale terrestrial processes were happening at the time of the K/T boundary. But the evidence for an asteroid hitting Earth had to be assessed before it could be ruled out – and the data is scanty. To start with, no impact crater has been positively identified, although several locations have been suggested, including Iceland, the Andaman Basin in the Indian Ocean, the Manson structure in Iowa, Hudson Bay, the Gulf of St Lawrence, the Colombian Basin in the Caribbean Sea and the Chicxulub structure on the Yucatan Peninsula in Mexico . There has been a certain territorial imperative behind these suggestions – Iceland was proposed by a European, the Andaman Basin by a South African and the others by North Americans.

There is also the iridium anomaly, cited as categorical evidence for an asteroid strike. But what if the iridium could have come from the Earth’s mantle, through a volcanic eruption at the time of the K/T boundary? This was a time when volcanoes were especially active worldwide. There were enormous outpourings of lava in the Deccan region of western India, the largest eruption of flood basalts in the past 200 million years. Radiometric and palaeomagnetic dating studies by Vincent Courtillot and his colleagues at the Institute of Global Physics in Paris, and by Robert Duncan and Douglas Pyle of Oregon State University, have clearly demonstrated that more than a million cubic kilometres of basalt erupted in little more than a few hundred thousand years at the K/T boundary. There was also intense volcanism in the western US and the southeast Atlantic at the same time. Antony Hallam of the University of Birmingham is among those who have pointed out that the sea level also fell significantly at the time of the K/T transition.

These and other advocates of earthly causes for the extinctions base their arguments on the climatic effects of the falling sea level and the environmental effects of volatile substances emitted in volcanic eruptions. These include acid rain (from sulphur dioxide), greenhouse warming (from extra carbon dioxide) and depletion of the ozone layer (from chlorine). All these effects are familiar as current environmental concerns but 66 million years ago they would have happened on a much larger scale. The resulting effect on the environment would have lasted for a few hundred thousand years, a short period in geological terms.

The impact theory has not remained static in the face of this opposition. Iridium has been a particular bone of contention. In the past decade, techniques for measuring the iridium content of rocks have improved, more sequences from the crucial time have been found and existing outcrops have been studied in ever more detail. Geological understanding of the processes involved in this and competing theories has advanced, too, giving both sides more ammunition to fling at their opponents.

Before 1980 there were few research reports on the geochemistry of iridium; much of the literature arose from exploration and mining for platinum group elements – iridium together with ruthenium, rhenium, palladium, osmium and platinum. Little of the work was focused on iridium and how it accumulates at the surface – crucial to the impact debate.

Levels of iridium at the K/T boundary are much too low to be worth mining. A substantial iridium anomaly in rocks of this age is somewhere between 5 and 10 parts per billion (ppb). Finding this is the equivalent of seeking out the one or two people in Western Europe who have blue hair and wear purple socks. But such levels are not beyond the reach of modern detection techniques, which can measure trace elements in parts per trillion.

The events suggested by the impact theory have come under closer scrutiny as scientists’ understanding of the Earth’s history has improved. There is no reliable record of anyone dying from a meteorite impact and the earliest documented impact crater of any size is 25 000 years old – the Barringer Crater, better known as Meteor Crater, in Arizona.

Climatology and environmental science have advanced in recent years, alerting scientists to how an impact might have affected Earth. Alvarez’s hypothesis suggested that the impact resulted in a global wildfire, which in turn caused the many millions of deaths that made up the mass extinction. The germ of this idea was the finding of minute amounts of carbon in a form like soot, at some K/T sections. This explanation is, however, problematic, because green, growing trees do not burn well. To get around this objection the impact hypothesis suggests that the world’s trees died and dried out as a result of the impact before the wildfire took hold.

Although it is difficult to see how an impact in, say, China, could kill and dry trees in Europe, recent major volcanic eruptions have shown that single events can disrupt the world’s climate zones. For example, the eruption of Tamboro in Indonesia in 1815 sent an aerosol of sulphur dioxide into the stratosphere, leading to a worldwide cooling of the atmosphere in the following year; there were cold spells, frosts and crop failures in New England and the period is still referred to as ‘the year without a summer’.

Other scientists have also looked at the unusual carbon at the K/T boundary. As for the iridium anomaly, they found that the high levels are spread over a significant stratigraphic interval of 50 000 years or so. But in addition, the ratios of different isotopes of carbon suggest that it came from a volcanic eruption, not a wildfire.

The overall result of the impact debate is the understanding on all sides that the sequence of events, spread over many thousands of years, is difficult to reconcile with a single event, whether it originated on Earth itself or in outer space.

How has the impact hypothesis adapted? It has certainly evolved beyond its initial form: its adherents now suggest that there were several impacts, possibly of comets, spread over a period of between 1 and 3 million years. But even this alteration is in itself debatable. No one is sure how much iridium there is in a comet. No crater has yet been found that fully fits the hypothesis of a single impact; if there were several bodies, then several craters remain to be found. To paraphrase Oscar Wilde, losing one crater may be a misfortune, but losing several looks like carelessness. In addition, the various strands of geological evidence have yet to support the hypothesis. Perhaps the most difficult hurdle to overcome is how the impact or impacts did in first the dinosaurs, then the shallow water shellfish and finally the ocean plankton, as the geological record shows.

Geophysical and geochemical data have provided another problem for the impact group: did the asteroid land with a thud or a splash? The type of rock that the asteroid hit and then, presumably, spattered around the world, is different in each case. A feature of the K/T boundary sediments are grains of quartz that show signs of high stress, very rapidly imposed. These are called shocked quartz grains and have been assumed to record the effects of the impact on the rock it hit. Quartz grains suggest a continental or shallow sea landing: the deep ocean floor is made of basalt, which does not contain quartz. But along with the extra levels of iridium at the K/T boundary there are unusually large amounts of elements such as arsenic, antimony and rhenium. These are all more common in basalt than in the granitic rocks characteristic of the continents. They are unlikely to have come from the asteroid itself, because meteorites have less than average amounts of these elements. Some of the proponents of an impact have got around this objection by calling on several smaller impacts over a year or so, landing both on land and in the sea.

Before Alvarez proposed the impact hypothesis in 1980 it was presumed that meteorites were the only significant source for the iridium found at the Earth’s surface. This was the cornerstone for the original hypothesis and a logical deduction at the time. But if the extinction turned out to have been caused by a volcano, for example, then the unavoidable conclusion is that, contrary to common wisdom, iridium must result from an eruption. William Zoller of the University of Nottingham and his colleagues have subsequently confirmed that this is the case. They found very high levels of iridium in airborne particles collected from recent eruptions of the Hawaiian volcano, Kilauea. Had Zoller’s findings been available prior to 1980, Alvarez’s original hypothesis might have been somewhat different.

There have been three additional studies that add to these findings and demonstrate that iridium enhancements are not only associated with the emission of fine particles from volcanic eruptions but they are also deposited and can become part of the geological record; this could be a way of producing the extra iridium.

The shocked quartz grains in rocks at the K/T boundary were also supposed to have formed during an impact. But if a volcanic eruption could explain the extinctions, then perhaps it could also produce shocked mineral grains. Neville Carter and Alan Huffman of Texas A & M University, among others, have confirmed that this is the case. Proponents of the impact idea have countered that because the shocked grains found so far at volcanic sites do not have as complex a deformation of the crystal lattice as those from K/T boundary sites, the latter must still be considered the result of an impact. Yet the more complex type of deformation has since been found in rocks formed in explosions within volcanoes.

These same deformation features are also found in the pumice – a frothy glass made from rock – associated with the glide surface of a landslide that took place at Koefels, near Innsbruck in the Alps. This is an immense landslide of crystalline rocks; the sheet that slipped is around 500 metres thick and about 3 kilometres by 6 kilometres in area. The slip probably moved in a minute or less, with a gravitational energy equivalent to 4.5 megatonnes of TNT. Friction at the glide surface generated so much heat that the rock melted and quickly solidified again as pumice. Unmelted quartz grains are held in the pumice; they contain both the simple and more complex forms of lattice deformation.

How has this debate affected researchers interested in how the dinosaurs met their end? A number of planetary geologists were initially attracted to the impact hypothesis: it added a significant dimension to the potential importance of extraterrestrial events in shaping the history of the Earth. By and large, the palaeontologists, sedimentologists and stratigraphers did not think much of the idea, because it could not account for the sequential and extended nature of the extinctions that they saw in the geological record.

Whether Alvarez’s hypothesis is right or wrong, he deserves a great deal of credit for reviving interest in one of the more fundamental geological problems – the causes of the mass extinctions that have happened from time to time in the 600-million-year history of life on Earth. Time – and future research – will judge how near he came to the truth.

Charles Officer is a research professor in the Earth Science Department at Dartmouth College, New Hampshire; he is interested in geophysics, environmental science and mass extinction events.

* * *

MYSTERIOUS HOLES IN THE GROUND

The two sites most recently suggested as likely craters from an asteroid or comet impact at the K/T boundary – Chicxulub on the Yucatan Peninsula and the Colombian Basin – are by no means as straightforward craters as they might seem.

At Chicxulub there is a circular area of unusual gravitational and magnetic properties. These qualities depend on the type of rock that lies in the few kilometres below the surface – density affects the gravity readings and iron-bearing minerals the magnetic properties. Variations in the readings suggest different types of rock below the surface, as could arise from layers of rock disrupted by an impact.

The Chicxulub structure is some 200 kilometres across – the right size for the sort of body initially suggested by Alvarez. But although such geophysical anomalies could form in a sizable impact, they do not necessarily indicate a crater; in this case they could also arise from a body of the volcanic rock andesite lying between 1500 and 2000 metres below ground. Reports of the rocks found in boreholes that were drilled into Chicxulub in the search for oil during the 1960s seem to reveal the most useful information about the structure. The logs show an orderly sequence of 350 metres of Cretaceous rocks immediately below the K/T boundary, lying above the andesite. The andesite itself is interlayered with thin Cretaceous limestones and includes layers of anhydrite – a mineral formed when shallow seas evaporate. The regularity of the layers appears to preclude the sort of disruption an impact would have had on the rocks existing at the time – those same Cretaceous beds. If an impact had carved out a crater 200 kilometres wide, the hole would also have been 10 kilometres deep, vaporising all the sediments that existed there. But as Nicola Swinburne points out in ‘It came from outer space’ (New ÐÓ°ÉÔ­´´, this issue), the geological records are not straightforward and the structure merits reinvestigation.

There are also problems with the idea that the western portion of the Colombian Basin is an impact structure. Because the old ‘basement’ rock in this area is semi-cir-cular in shape, the area has been viewed as a crater. But the Cretaceous sediments that lie on the basement consist of flat-lying beds that fill it rather like water in a wine – glass. The curve of the basement beneath must have formed at some earlier time.

The data that first led to the search for impact sites in the Caribbean came from geological sections through sedimentary rocks in two sites in Cuba and Haiti. The Cuban section is known as the Big Boulder Bed; it consists of a layer of big boulders, at first thought to be the blanket of debris thrown out of the crater during the impact. But Manuel Iturralde-Vinent of the National Museum in Havana and his colleagues have examined the critical section in the field and concluded that the big boulders result from an unusual type of weathering. Whether this explanation can also apply to other sites where such ejecta layers have been found remains to be seen.

The Haiti site was where spherules of glass in sedimentary rocks were thought to be tektites – droplets of molten rock formed in an impact. But my colleague John Lyons of Dartmouth College and Celestine Jehanno and colleagues of the National Centre for Scientific Research in France believe that they are the decomposed products of a volcanic eruption, on the basis of their composition and structure.