There was snow on the mountains nearby, and in the dark gorge the winter water was icy. We sloshed up the meandering stream bed, examining the rocky walls as we went. Parallel layers of whitish rock containing darker interior bands plunged down to meet the stream ahead of us. The formation continued for hundreds of metres up the steep cliffs above us, up into patches of sunlight. But down where we were, there was no hint of sun.
Joy Lines, my fossicking companion, and I were on a journey back to the end of an epoch. Sixty-five million years ago, life on Earth faced an apocalypse. Half of all organisms were rendered extinct, including the dinosaurs. Most ecosystems suffered massive destruction, and it took a million years or more for life to return to a semblance of order, though it was never the same as it had been. All that is left of that cataclysmic change is a scar of clay-like rock exposed in a few dozen locations around the world.
I peered at the sheaf of papers in my hand, looking for congruence between the detail that appeared in the grainy pictures and the rocks beside us. Judging from the rock formations, we were certainly in the right area, but finding the exact spot—a layer of rock just centimetres wide laid down 65 million years ago—could prove tricky.
When I’d spoken with Chris Hollis, a geologist from GNS Science, on the phone a few days earlier, I could hear doubt in his voice that we’d be able to find the spot. After all, we were well off the beaten track.
We’d travelled with Hamish Murray (one of the family who owns 40,000 ha Bluff Station at Kekerengu, midway between Blenheim and Kaikoura) for some 30 km along a farm road in his ute, climbing over a low saddle near the eastern end of the Seaward Kaikouras and winding west towards the lower ramparts of the Inland Kaikouras. Murray pointed us in the direction of the south branch of Mead Stream and wished us luck.
Geologists come from all over the globe to see the rocks in Mead Stream. Sixty-five million years ago, these formations were sludge on the seafloor off the coast of a very different New Zealand. Dinosaurs probably roamed the land and erosion was steadily nibbling away at the hills. Rivers bore the debris back to the seafloor in the process of relentless levelling that nature undertakes, gradually building up a geological record in the sediment that accumulated close to the coast.
But then occurred an event that changed life here and everywhere else on Earth. It wiped out pretty much every animal on land larger than a rat, and also put paid to much that lived in the sea as well.
For instance, the ammonites—a group of cephalopod molluscs resembling the chambered nautilus still found in the tropical Pacific—completely disappeared despite dominating the oceans for more than 200 million years. Most large marine reptiles vanished. Half of all organisms became extinct.
So major was the abrupt change recorded in fossils at this time that it split geological time, ending the 180 million-year-long Mesozoic era and beginning the Tertiary, also known as the Cenozoic era. This has been commonly abbreviated to the K–T boundary—K is from the German name for the Cretaceous period (Kreidezeit meaning chalk age), the last of the three periods of the Mesozoic. These days, however, the use of T for Tertiary is being discouraged by the International Commission on Stratigraphy, and “Paleogene” is replacing it—so it is now more accurately termed the K–Pg boundary.
What actually caused this sudden extinction is the subject of much debate and suggestions have been many and varied—including terminal hay fever brought on by the evolution of flowering plants!
But the most persuasive notion has been a theory posited in 1980 by Nobel Prize-winning physicist Luis Alvarez, his son, geologist Walter Alvarez, and two chemists. In New Zealand and elsewhere, the K-Pg boundary is marked by a layer of clay-like sediment, a couple of centimetres thick, sandwiched between layers of harder rocks. Alvarez and his team discovered that this layer contains high levels of iridium—an element that is extremely rare in the Earth’s crust, but abundant in most asteroids and comets. These extraterrestrial bodies also possess oddities in the abundance of other heavy metals, such as chromium and nickel, that Alvarez also found in the boundary rocks. The layer also contained tektite glass spherules and tiny quartz grains bearing stress lines. All this is evidence, the Alvarez team claimed, of a massive asteroid impact on the Earth 65 million years ago. The asteroid would have exploded on impact, spreading iridium-laden dust around the globe.
Estimating the total amount of iridium in the K–Pg layer (and assuming that the asteroid contained the normal percentage of iridium), the Alvarez team calculated that the asteroid must have been about 10 km in diameter. Such a massive impact would have released energy equivalent to exploding two million of the largest nuclear bombs ever tested, throwing 90,000 cubic kilometres of rock and dust into the atmosphere. In comparison, the moderately large eruption of Mt Pinatubo—ejecting 10 cubic kilometres—in the Philippines in 1991 would have appeared a minor sideshow. Even so, ash and sulphuric acid aerosols from the Pinatubo eruption reduced sunlight reaching Earth’s surface by 10 per cent for close to three years and reduced global temperatures by 0.4°C.
But the actual impact of the asteroid was merely the beginning—the environmental holocaust that followed was catastrophic. A dust cloud would have enveloped the globe, blocking sunlight and therefore inhibiting photosynthesis for months or years. Temperatures would have plummeted.
While most plants, phytoplankton and organisms dependent on plants would have perished—along with the carnivores that preyed on the herbivores—scavenging animals that fed on detritus could have survived, as would fungi that thrive on dead wood.
Fragments from the asteroid impact—heated by being blasted right up through the atmosphere and then falling back to Earth—may have triggered widespread fires. The oxygen content of the atmosphere was very high (30–35 per cent) during the late Cretaceous, supporting intense combustion, but the geological record demonstrates that oxygen levels plummeted in the early Paleogene. It’s possible that the excess of oxygen was literally burnt up by K–Pg fires, becoming carbon dioxide in the process.
Once the atmosphere cleared, the elevated carbon dioxide levels would likely have caused considerable global warming—perhaps by as much as 10°C—a hot spell which could have lasted for centuries and killed off some of what had survived the dark and cold.
The Alvarez paper created a furore at the time, and one problem with the hypothesis was that there was no crater of suitable size known—something 200 km in diameter and the right age. However, a decade later, in 1990, a 180 km-wide crater of suitable age was discovered on the Yucatan Peninsula of Mexico’s east coast.
The impact site was a gypsum (calcium sulphate) bed on the coast, indicating the presence of an ancient sea and raising the prospect that a devastating tsunami had added to the destruction. Vaporised by the force of the impact, the gypsum would also have produced massive amounts of carbon dioxide and given rise to a vast sulphur dioxide aerosol that would have further diminished sunlight and eventually fallen back to Earth as acid rain—making a catastrophic situation even worse.
Since the Yucatan discovery, several other possible impact craters have been discovered that are about the same age. It’s conceivable that a larger asteroid broke up and peppered the Earth with fragments in much the same way that more than 20 pieces of the comet Shoemaker-Levy 9 smacked into Jupiter over a few days in July 1994. (Although the largest fragments there were only 2 km, the impacts produced large visible effects in the Jovian atmosphere.)
The largest of these other terrestrial craters is 600 km x 400 km and lies in the seafloor west of Mumbai, India. It was named Shiva by Texas paleontologist Shankar Chatterjee, although there remains some uncertainty as to whether it really is an impact crater. Lesser craters include the 24 km-wide Boltysh crater in Ukraine and the 20 km-wide Silverpit crater in the North Sea. All are dated at about 65 million years ago.
A rival explanation for the K–Pg extinction has been a sudden increase in volcanic activity at the Deccan Traps, a series of vast volcanic outpourings that cover half a million square kilometres of western India to a depth of up to two kilometres. Before erosion took its toll, lava probably covered three times that area.
This series of eruptions are thought to have occurred about 65 million years ago, over a span of between two million and 30,000 years. Gases released by the eruptions would likely have caused dramatic global warming—perhaps by as much as 8°C. Given the proximity of Shiva and the fact that it’s a similar age, it’s been suggested that it was the asteroid impact that triggered the Deccan eruptions.
If an asteroid was the sole destroyer, extinction should have had a very abrupt onset in the fossil record. If volcanism or some other mechanism lay behind the mass extinction, a more gradual timetable might be found, including extinctions before the iridium anomaly. And to distinguish between those possibilities, we need to turn our attention to the rocks themselves.
Sedimentary rocks laid down at the K–Pg boundary in the New Zealand area have been uplifted by tectonic forces and exposed in a number of places around the country—the best collection of such exposures in the Southern Hemisphere. And since New Zealand lies on the other side of the globe from the Yucatan Peninsula, the records from our rocks can provide an insight into the global dimensions of what happened 65 million years ago. Furthermore, our rocks were deposited in a range of different environments—some in deep water, others at shallower depths or even on the floor of a lake in one instance.
While most of the sites do not contain large fossils, microfossils are abundant. They are the minute shell-like remains of plankton—unicellular organisms such as radiolarians, foraminifera and dinoflagellates. Paleontologists can break down or dissolve most sedimentary rocks and release a cargo of these microfossils, and since many species are restricted to a particular environment (oceanic, coastal waters or estuaries), the conditions under which the rocks were deposited can often be determined. Many species occur in the fossil record for only 2–10 million years and are globally distributed, making microfossils invaluable for assigning accurate ages to sedimentary rocks and also providing information about the climate of the time.
In the Clarence River region of south-eastern Marlborough, a nearly complete 650-metre-thick sequence of sedimentary rocks laid down from 70 to 40 million years ago lies exposed in several gullies. And the place where the most complete sequence can be seen is Mead Stream, where each millennium has been recorded in 0.7–3.5 cm of rock for geologists to read.
Once Joy and I found it, the K–Pg boundary seemed completely unremarkable—a two centimetre-wide crevice between harder rocks. A tiny turquoise cross was painted onto the rock where the drawings and photos indicate the boundary lies, and Joy pointed out a couple of inconspicuous bore holes nearby made by geologists.
Younger rocks higher on the wall contain records of three other globally significant climate events—the late Paleocene carbon isotope maximum (59.5–56 million years ago), the initial Eocene thermal maximum (55.5 million years ago), and the early Eocene climate optimum (53–50.2 million years ago)—a volume of the planet’s tumultuous history cast in stone which is being closely studied by scientists concerned about climate change.
But unlike these other events, the K-Pg boundary is obvious even from hundreds of metres away. Above the boundary the rocks form a conspicuously darker band, 20 metres wide, of hard, glassy and splintery chert—the fine-grained sedimentary rock derived from the silica shells of diatoms and radiolarians—while the surrounding white rock is softer, rich in lime or calcium carbonate.
The change in colour and geology represents a total revolution in the ecology of the ocean following the K-Pg event. It scribes a line clean through the geological record, through mountains and gullies; an unmistakable pattern spelling out a change of guard in black and white. White, where the sediment laid down in the Cretaceous was rich in foraminifera which had shells made of pale carbonate; and black, where those micro-organisms were rendered all but extinct in an instant and rapidly replaced by diatoms and radiolarians, with shells made of dark silica, which went on to dominate the marine ecosystem during the cool conditions that prevailed for a million years following the event.
It is odd to think that this unremarkable, dull layer of sediment was formed at a time of unimaginable catastrophe. It seems too bland, too commonplace, altogether too nondescript, but it is all that remains of one of the most destructive events in Earth’s history.
Later that day, Joy and I drove a little north along the Kaikoura coast and walked back down to Wharanui Point into a blustery southerly. A narrow reef, resembling a set of decayed molars, marched down the beach and out into the surf, resisting the gnashing waves. The strata here run almost vertically, tipped on end by the geological forces that have shaped New Zealand over the past 65 million years.
Woodside Creek, the original New Zealand K–Pg boundary site, flows into the sea just a few hundred metres north of the point, and its gorge, where the boundary rocks are exposed, is in low hills just a kilometre or two inland. Between 10 and 20 km to the north lie three further exposures of the boundary, at Needles Point, Flaxbourne River and Chancet Rocks.
By analysing the minerals and microfossils found in the boundary rocks of the area that is now eastern Marlborough, geologists have learned much about what went on 65 million years ago.
But first, it’s important to understand that much of modern Marlborough was once under a good deal of water, to the north of a very different proto-New Zealand near the eastern edge of what was termed the East Coast Basin. The land once stretched out east to west over 3000 km and lay at 60° South. For 1200 km east of the East Coast Basin stretched the land of the Chatham Platform (which is now largely an undersea ridge), and south of this long finger of land lay the Canterbury Basin.
Water upwelled, bringing nutrients to the surface to stimulate blooms of the microscopic plankton whose shells dominate the seafloor deposits that would become today’s limestone hills.
The site of the Flaxbourne River today was once some 250 km from shore—with Woodside Creek, Mead Stream and Branch Stream representing progressively shallower coastal zones. Even the site that Branch Stream now occupies was perhaps 50 km out to sea and in 500 m of water. Because of the distance of the sites from land, most of what accumulated on the seafloor was the remains of minute organisms that lived in the water above or in the seafloor itself—radiolarians, foraminifera (forams), diatoms, coccolithophores and dinoflagellates (see sidebar). Farther west in the basin—today’s coastal Wairarapa—the seafloor was closer to the ancient shoreline, meaning that the sediments which accumulated were richer in mud and detritus derived from the land.
There is no doubt that a major cataclysm occurred at the K-Pg boundary in Marlborough. It’s evident in the darkening of the rock, which signals the almost complete and instant disappearance of the rich calcareous-shelled plankton (foraminifera and coccolithophores) that had dominated Cretaceous seas. It’s a different story for the silica-shelled microfossils, such as the radiolarians and diatoms. Remarkably, there appear to be no extinctions at all in these groups although, at least for the radiolarians, there is a major rearrangement in the relative abundance of groups of species, implying a dramatic change in ocean conditions. Larger, more complex species were rapidly replaced by smaller, simpler bacteria-feeders, and diatoms (very simple phytoplankton with cell walls made of silica) suddenly became extremely abundant. After about a million years, as the seas and atmosphere recovered, there was a return to the deposition of carbonate-rich rocks (limestone), signalling the recovery of the calcareous plankton that had dominated the late Cretaceous.
While the rocks in Marlborough give us a good sense of changes to life in the open ocean 65 million years ago, to understand the effects on and closer to land, Joy and I headed into the Waipara River gorge in north Canterbury, another place where the K–Pg boundary is exposed. Last year, we came here looking for fossils of marine reptiles, but this time it was later in the season following a lot of rain. The river that we waded along easily last year was now deep, muddy, swift and uninviting. We clambered over hills and through scrub to get down to where the boundary was exposed on a bank above the water. Here there were no clearly defined strata, just a rubbly grey cliff bearing assorted yellowish streaks and smears like faded graffiti. In Marlborough locations, the rocks were laid down in deep water, but here the rock was formed in shallower waters close to the coast with a lot of sediment present. Along with marine microfossils, pollen and spores from land plants are abundant in the deposits.
And here at Waipara in late Cretaceous and early Paleocene rock samples, spores and pollen from some 82 plant taxa have been found. Seventy-one species were found in the top seven metres of Cretaceous rocks (perhaps representing half a million years of deposition)—45 per cent of them conifers, 40 per cent ferns and 10 per cent flowering plants (including beeches). The species present indicate that the climate was mild and temperate, with little ice and frost but sufficient rain for the large range of ferns.
But from the K–Pg boundary, for the next 20 cm of rock there is evidence of an abrupt change in the flora. Fern spores increased dramatically in abundance and conifer and flowering plant pollen halved, representing a major shake-up in the structure of plant communities—forests seem to vanish in the blink of a geological eye, to be replaced by little more than ferns.
The types of fern present also differed from those in the Cretaceous. The very earliest Paleogene was dominated by species of forked ferns (gleichenids), which are pioneers in the wake of environmental disturbance such as burning. Later, these gave way to tree ferns.
It was not until hundreds of thousands of years after the event that the number of fern spores dropped to about half their original abundance, whereas conifers and flowering plants resurged to much greater populations than in the Cretaceous, suggesting that the climate was both cooler and drier.
It was a damp afternoon on the West Coast as I tried to find my way to the old Moody Creek coal mine in Dunollie, north of Greymouth. I parked in the Solid Energy Spring Creek car park and entered an old house that served as the office. Inside was a small pot-belly stove exuding warmth. I came upon two men who confessed to being mining geologists, and asked if they had ever heard of the K–Pg boundary.
“Well, I hadn’t until January,” answered one of the geologists. “But then I took a party of geologists up there to find it.”
We hopped into a ute and headed up a narrow road into the hills. “Down here,” said Michael Nickerson as he clambered out and headed down into a dark gully. We descended to a murky-looking stream.
“The boundary layer would be right on the water level over there. In January, when I came with a party from a geological conference, some knew all about it.”
Again, a completely inconspicuous site. You wonder how much field work it took GNS paleontologist Ian Raine to locate the site some 20 years ago. Here the rocks looked quite shaly and were a sombre grey—a soft carbonaceous mudstone—and there was a vague outline of the coal seam at stream level that contained the boundary. Above the boundary was a fern-covered bank, dark in the fading light.
Despite its dour appearance, this was another K–Pg boundary site of great international significance. Unlike any other site in the country, the rock here was deposited in the bed of a lake or within a boggy mire, and the fossils were pollen, spores and leaves from plants growing on the surrounding hills.
Here, analyses have given evidence of even more dramatic changes than at Waipara. The iridium is up to 71 parts per billion, the highest value found in non-marine rocks anywhere in the world. In the late Cretaceous, conifers dominated the ecosystem here, but at the K–Pg boundary, fern spores increased to 90 per cent of the plant community. Flowering plants disappeared completely from the deposits above it and represent less than five per cent of the flora for perhaps a million years afterwards.
Recently, a four millimetre-thick layer containing only fungal spores and cells has been recognised at Moody Creek above the last remains of Cretaceous plants and before the spike in ferns. Fungi don’t depend on photosynthesis at all but get nutrients from decomposing wood and other plant remains—exactly the sort of materials that would have been abundant in the years following the asteroid impact. The fungal layer here is yet more evidence of extensive forest destruction—and about as far as you can get from Yucatan and other possible impact sites.
The rocks in New Zealand bear compelling evidence for Luis Alvarez’s asteroid hypothesis, an event so catastrophic that it abruptly and radically altered the forests and seas of our remote archipelago half a world away from the site of impact—rather than slow changes that might have been wrought by an effusion of gases from the Deccan Traps. And by studying the effects of the K-Pg event here, it now seems unlikely that any part of the world escaped devastation.
The fungal and fern spikes immediately after the iridium anomaly tell us that the forests here died off, either through lack of light, acid rain or burning. Thirty thousand years after the impact, tree ferns replaced more opportunistic fern species and then conifer forests dominated for the next million years.
And in the seas, as in other parts of the world, the New Zealand K–Pg boundary is defined by mass extinctions among the calcareous-shelled plankton. However, in marked contrast to other regions, there was no general collapse in plankton production, rather an abrupt reorganisation in silica-shelled plankton that begins directly above the boundary clay and lasts for at least a million years, indicative of a prolonged period of cool climatic conditions in the New Zealand region. How and why this happened, and whether it occurred in other regions, is still under study.
And though it now seems that Alvarez and colleagues had it right, the details of what happened after the K-Pg boundary, one of the most significant events in the history of our planet, remain lost in pre-history. However, relics of the creatures and forests of 65 million years ago remain preserved in the sediment in which they perished, a geological tomb that may yet hold more clues to their demise, but one that does not give up its secrets easily.