How Quickly Do They Tip?
Recently, I have been writing a lot about the political aspects of climate change, so today we get back to environmental and ecological science. I’m sure there will be more politics in the future, but not today. There has been lots of speculation in recent months about tipping points – is our climate approaching a tipping point, will we be able to recognize it before it happens, how close can we be to it before it’s too late to avoid going over it? We know we have to change our behavior in some fairly radical ways and some of this speculation is part of trying our best to avoid the bad medicine until the very last moment. Some of the speculation is based more on fear – we sense tipping points may happen and they likely will not be pleasant. Maybe there are ways to anticipate them, giving us a bit more warning?
Our planet has been through many ecological tipping points in the past – periods of unusually intense volcanic activity, sudden and dramatic changes in sea level, or onsets or cessation of periods of glaciation each of which led to radical reorganization of ecosystems or the biosphere. The “mass extinction events” surely qualify as tipping points. Paleontologists count five major extinction events in the last 542 million years, periods when a substantial proportion of living species became extinct in a relatively short space of time, dramatically reducing global biodiversity. Subsequent to each, biodiversity slowly rebounded, but the structure of the resulting ecosystems was radically changed because entire major groups of species were no longer present. How sudden were these tipping points, and what kind of roller coaster ride did the biosphere go on in the years that followed them? Last week, a couple of papers in Science gave us a somewhat better glimpse of the most recent mass extinction, the end-Cretaceous, or KT, event that occurred about 65 million years ago. So, I figured it might be fun to look at this event, and see if we can picture what happened.
The popular fable concerning the end-Cretaceous mass extinction event is that a 10km diameter asteroid (the correct term is apparently ‘bolide’, but ‘asteroid’ conveys ‘great big rock’ a lot more clearly to non-initiates like me) hit the earth creating a now buried crater in Mexico’s Yucatan peninsula, and that this single, catastrophic event caused the extinction of all the dinosaurs then living, except the birds, as well as large numbers of other creatures around the world. The arrival of an asteroid certainly ranks as a discrete event, although in judging its effects on the biosphere it is appropriate to include the impact and the subsequent period of many years when the perturbation would slowly die down. The fable reports that all the species that became extinct did so during this period of perhaps 25 or perhaps several thousand years.
Mass extinction of the dinosaurs due to the arrival of an asteroid?
What would such an event be like? On entering the atmosphere the asteroid would heat up and begin to break apart. Its impact would blow a mass of vaporized rock and steam high into the atmosphere, and if the asteroid was rich in sulfur this vaporized material would include sulfate aerosols, a potent source of cloud and acid rain. The impact would release energy equivalent to perhaps as much as 100 teratonnes of TNT, and the immense blast would spread dust and debris worldwide, while also mowing down trees and larger living things (like dinosaurs) in at least the immediate vicinity. (‘Immediate’ here is on a continental scale; these effects would extend for thousands of kilometers from ground zero.) The dust and other substances would shield the earth from sunlight cutting light by 10-20% and ushering in a period of twilight that would persist some months or years, slowing or shutting down photosynthesis. The gases released from all the vaporized rock would likely include CO2 and other greenhouse gases, so the initial cooling following the impact could be followed by some warming over the next 1000 years or so.
There is solid evidence that such an asteroid did hit the earth very late in the Cretaceous. In 1980, Luis Alvarez, a physicist, and colleagues including his brother, Walter, deduced the existence of a massive asteroid strike when they detected an unusually high concentration of iridium exactly at the boundary between Cretaceous and Paleocene rock in outcrops around the world. Iridium is rare on Earth, but more abundant in asteroids, and they deduced that a massive, explosive impact could have showered the Earth in iridium-rich dust. Almost a decade later, the 180 km diameter Chicxulub crater in the north-west Yucatan was identified as the right age and size for the impact site, and with traces of materials that tied it to the same asteroid that yielded the Iridium. Over time the geophysical evidence for the asteroid impact, at Chicxulub, at or close to the end of the Cretaceous, has only grown stronger. This part of the fable is a fact, and it must have been a massive disruption to the planet. And last week in Science, Paul Renne, of the Berkeley Geochronology Center and UC Berkeley, along with 8 colleagues added some new evidence tying the impact to the Cretaceous-Paleocene boundary more tightly than ever.
The Alvarez brothers made almost as big an impact on the scientific world in 1980 as the asteroid had made on the Earth. They brought some showmanship and flair, as well as a radically new story. Scientific establishments react to sudden impacts in their own way – bold new ideas are often met with intense opposition from lots of quarters, especially if there are hints of showbusiness – and you have to admit that an asteroid hurtling out of the sky and killing off the dinosaurs is a Grade B movie script of the first order. Some paleontologists disputed the idea of a single, relatively brief event as responsible for a complex, global pattern of extinction. Some geologists disputed the idea that the asteroid hit on the last day of the Cretaceous and not hundreds of thousands of years earlier or later. Some others thought that massive volcanism in the Indian subcontinent might be more important in causing the mass extinction, or rapid and extreme drops in sea level, or cooler climates. Renne and colleagues have brought new, state-of-the-art, more precise dating methods to bear on the question of timing. Earlier, less sensitive methods had dated the impact as occurring 180 thousand years later, or several hundred thousand years earlier than the Cretaceous-Paleocene boundary. With the new procedures, the impact is now dated as occurring 66.038 ± 0.049 million years ago, and the Cretaceous-Paleocene boundary 66.043 ± 0.043 million years ago.
Yes, those error measurements represent 43-49 thousand years, but that is quite precise for something measured so long ago, and the two estimates are only 5000 years apart giving me confidence that the two events are quite close together. The asteroid landed at the time the Earth was transitioning from the Cretaceous to the Paleocene. With their more precise dating methods, Renne and colleagues are also able to comment on dates for some events surrounding the asteroid impact. Their better resolution of ages in sediments at the Hell Creek, Montana site show that immediately after the impact a relatively brief period of low primary productivity lasted from about 5 to 13 thousand years. This is generally shorter than the periods of reduced productivity at various marine sites that range from somewhat under 100 thousand to several million years before productivity is restored to pre-impact rates.
The Controversy around the Mass Extinction
But what about the extinction event? When did it start, how severe was it, and how long did it take? Did the asteroid cause all of the loss? As I delved into the recent literature, I discovered that on this front the controversy sparked by the Alvarez discovery remains alive and well. Here the fable appears to be just a story. Controversies in science tend to result in more detailed study, and there has been intense study of the pattern of extinction over the Cretaceous-Paleocene boundary since 1980. This was the second largest mass extinction on record, with somewhere between 50% and 80% of extant species disappearing (the variation is partly because different things happened in different places, and partly because different taxa suffered very differently). As a lowly ecologist, I am not in a position to discriminate among the claims by different paleontologists and geologists in quality academic institutions, but I can present a sense of the present state of play. The two camps, if we can call them that, are a) a group of scientists who are thoroughly satisfied that the asteroid occurred at the KT boundary and was the primary cause of the mass extinction, and b) an equally eminent group of scientists who claim that the mass extinction was actually a series of separate smaller extinction events that happened over an extended period through the mid to late Cretaceous and into the Paleocene. This second group disputes the existence of a single cause, and some seem determined to demonstrate that the asteroid was not even an important cause among several. Take that, Dr. Alvarez.
End of dinosaurs due to massive volcanic eruptions? Figure © George Arthur Bush
To understand why such diametrically held views can persist, despite intensive study, think for a moment about the difficulty of pin-pointing the time of extinction of a species. Even today we have species on the IUCN Red List that are probably already extinct, but we do not know for sure. Because one thing that happens when a species dwindles to extinction is that it first dwindles. Rare species can be incredibly hard to find, but extinction formally occurs when the last living individual of the species dies, so identifying when this occurs is very difficult even today. Now go back 65 million years (that is over eight thousand times as long ago as the dawn of civilization in the Middle East, or 274 thousand times as long ago as the US war of independence in 1776). The only way you can determine what species exist is to find their fossils, and as we all know, lots of individuals die but fossil-hood, a bit like sainthood, is reserved for a tiny minority. So what is the chance of finding rare fossils of a species which is dwindling towards extinction? Then too, not every fossil stays put. They get moved about and if they get moved up in the sediments it will appear that that species survived longer than it really did. If they get moved down it might appear that the species originated earlier, or died out sooner than it really did. A large part of the controversy over the end-Cretaceous extinction revolves around the adequacy of the fossil record at different sites around the world. In what follows, I am setting out what I think is a middle-ground view, but I suspect plenty of geologists and paleontologists will disagree with the details. I’m the lowly ecologist doing his best here.
Before we can talk about mass extinction, we need to reflect on the immense time periods involved. As well as being a long time ago, the Cretaceous was an immensely long time itself – about 80 million years from its beginning at the end of the Jurassic 145 million years ago to its end at the Paleocene 66 million years ago. The uppermost part of the Cretaceous is the Maastrichtian stage, which lasted about 5 million years (that’s only 2.5 times as long as humans have existed). Now, when did the extinctions occur? As I delved into the literature, I learned that there was a lot of extinction going on in the Cretaceous, especially towards its close, but there were only a few groups of species that were diverse up until the late Maastrichtian and then disappeared completely.
If you do not want to read about the details of who went extinct when, skip these next few bits and jump ahead to: So What DID The End-Cretaceous Mass Extinction Look Like?
The Marine Microfossils
Diagrams showing percentage of extant families of each group of microfossils that went extinct during each stage of the Cretaceous (note that the bars represent extinctions of whole families, not number of families, and not extinctions of species). Diagram from MacLeod 2005 (sorry about the quality)
The marine microfossils include several groups that died out in large numbers during the Maastrichtian. The coccolithophores, benthic foraminifera, and radiolarians each show a peak of extinctions restricted to this period, while the planktonic foraminifera and dinoflagellates show a period of high rates of extinction that extends from the Maastrichtian through into the Paleocene. Gerta Keller of Princeton University, one of the main promoters of a multiple-cause extinction pattern has recently published detailed results for planktonic foraminifer showing that they gained their maximum diversity early in the Maastrichtian, and declined sharply, with only one new species evolving, in the final 500 thousand years of the Cretaceous.
In this figure, dates and names of time-periods are shown in the left-hand panel. The two middle panels reveal declining water temperatures in both surface (open circles) and deeper waters (closed circles), and coincident increasing rates of productivity. The right-hand panel shows increasing diversity (number of species), at least in surface and intermediate depth foraminifera until the late Maastrichtian, followed by a slow loss of diversity culminating in a collapse at the close of the Cretaceous. Figure taken from G. Keller, 2001, Planetary and Space Science v49
The Marine Invertebrates
The marine invertebrates provide several different stories depending on which group is examined. McLeod’s family-level surveys show that the sponges (Porifera), and to a lesser extent, the molluscs, exhibit a pronounced increase in extinctions during the Maastrichtian, while the corals and arthropods (lobsters, shrimps, crabs) showed no peak whatever in loss of families. The echinoderms (sea urchins, starfish, crinoids) and lophophorates (brachiopods and bryozoans) give an intermediate picture with a slight increase in rate of extinction of families towards the final stages of the Cretaceous. Among the mollusc families that became extinct during the Maastrichtian, there were 14 families of clams and other bivalves, 8 families of gastropods (snails) and 12 families of cephalopods (octopus, squid, ammonites and others), but most of these extinctions appear to have occurred during the early or mid-Maastrichtian rather than at its end.
Another figure from MacLeod 2005 showing the percentages of extant families of marine invertebrates going extinct in each stage of the Cretaceous and immediately after.
The ammonites are often held up as an example of a group that disappeared at the end of the Cretaceous. These cephalopods, with massive spiral shells, were abundant throughout the Cretaceous and fossilized well. Detailed studies have shown they proliferated through the early and mid-Cretaceous, reaching their greatest diversity of species in the Maastrichtian (although at the family level they were most diverse about 30 million years earlier in the mid-Cretaceous). However, extinctions increased and they became progressively less diverse through the upper Maastrichtian, with only a few species surviving through to the end of the Cretaceous.
While the corals displayed negligible rates of extinction at the family level, about 98% of reef-building species and 83% of genera present in the Maastrichtian did not survive into the Tertiary. Indeed, the role of corals in reef-building had been declining progressively during the Cretaceous and the reefs of the Maastrichtian were predominantly built by rudists, a strange group of bivalve molluscs that enjoyed considerable Cretaceous success, only to become extinct at or close to its close. Reefs built by corals did not reappear in any abundance before the Eocene, about 10 million years after the end of the Cretaceous. Incidentally, deep-water corals that do not possess symbiotic algae (zooxanthellae) and do not build reefs, include most of the families that occur among shallow, zooxanthellate, reef-building forms – thus the loss of nearly all reef-building species resulted in few extinctions of families.
The Marine Vertebrates
Among marine vertebrates, the prevailing Cretaceous pattern is that larger, epipelagic, top carnivores are the species that became extinct. These include marine reptiles such as the mosasaurs and plesiosaurs which became extinct in the late Maastrichtian as well as a number of sharks and teleost fishes. The fishes, as a group, however, did not show any real sign of an end-Cretaceous mass extinction, and a recently published analysis (Tom Near and others, 2012 Proceedings of the National Academy of Sciences) which used molecular and fossil data to fix the times of major speciation events in the evolution of the fishes shows that the Cretaceous and early Tertiary was a period of intense diversification in this fauna. Except for the larger, open-ocean, epipelagic, top carnivores, that is. Another recent paper by Matt Friedman (2009 in Proceedings of the National Academy of Sciences) shows that during the Maastrichtian there was a near-complete extinction of “large-bodied fishes with biomechanically fast jaws” – fishes that “appear to be the ecological analogues of modern, large-bodied predatory teleosts such as scombroids (tunas, mackerels, cutlassfishes, and the wahoo), xiphioids (billfishes), sphyraenids (barracudas), and carangoids (jacks and dolphinfishes)”. These modern groups are all among the types of fish that evolved during the early Tertiary. There is a similar pattern of loss and subsequent replacement of large-bodied sharks at the end of the Cretaceous.
Taken together, the disappearance of large, carnivorous fish and reptiles supports the idea that the surface waters of the world’s oceans became a lot less productive; this reduced abundances (and diversity) of marine plankton such as coccolithophores, radiolarians and dinoflagellates, disrupted food webs and reduced diversity of larger carnivorous species present.
The dinosaurs show a very pronounced peak of extinctions during the Maastrichtian although it is not at all clear that these extinctions all came at the end of that period. The mammals also experienced a peak of extinctions at that time, perhaps earlier than that of the dinosaurs (extinctions center on the Maastrichtian and the immediately prior Campanian stage). By contrast, the amphibians (frogs and allies) showed essentially no extinctions at the family level during the entire Cretaceous, and the freshwater fishes also showed little evidence of any effects of an asteroid or anything else. Non-dinosaur reptiles showed yet another pattern – relatively high levels of extinction but throughout the Cretaceous.
This diagram, from O’Leary et al, Science, 2013, shows the phylogenetic relationships among mammals, including fossil groups. Orange lines show phylogenetic linkages inferred on morphological or molecular evidence but for which fossil confirmation is currently lacking. The time scale of the 200 to 400 thousand years immediately after the KT boundary is expanded to show the pronounced diversification that took place during this time. Among placental mammals, only a single stem taxon crossed the KT boundary, but most placental orders had evolved before the end of the Paleocene. Monotremes and marsupials also experienced a burst of speciation during the Paleocene.
Maureen O’Leary of Stony Brook University and colleagues have just published a revised phylogenetic tree for the placental mammals using fossil and molecular data (the tree is here, but the article is not on open access). They have paid particular attention to dates of origin and show that the entire group of placental mammals evolved from a single ancestral form during the Paleocene. In other words, the mammals that were around in the time of the dinosaurs were exclusively monotremes, marsupials and some extinct precursors of placentals. Astoundingly, all but one or two of the orders of placental mammals that existed as fossils, or exist today, including the primates, evolved during the Paleocene with much of this diversification taking place during the first 200 to 400 thousand years of this period.
So What DID the End-Cretaceous Mass Extinction Look Like?
Summarizing this quick romp through the animal kingdom, it’s clear that the mass extinction was not nearly as quick as the arrival of the asteroid in the Yucatan. For some species, it began long before the asteroid was in sight, and for many, the die-off was completed also long before the asteroid arrived. For other groups there was pronounced extinction during the Maastrichtian, and for some of these there is evidence that much of this extinction came at or very close to its end. Clearly, factors beyond the asteroid were operating before the asteroid arrived to cause high rates of extinction in some fauna.
What might those other factors be? There is abundant evidence that the Cretaceous was a time of cooling temperatures, developing polar ice caps, and falling sea levels. The extensive shallow seas of the earlier Mesozoic basically dried up as sea level fell to expose most continental shelves. Something must have caused these changes, and many geologists believe it was extensive volcanism in the form of continental flood eruptions in India that formed what is known as the Deccan Traps. Continental flood basalts form when eruptions result in extremely fluid lava that flows laterally rather than building volcanoes. The Deccan Trap floods include the longest flows recorded on Earth, extending over 1500 km across much of India.
Map from G. Keller, In Earth and Life, Springer, 2012, showing the Deccan Traps formed in the late Cretaceous and early Paleocene, and likely a major cause of the mass extinction.
Building of the Deccan Traps occurred in three main phases: a small pulse around 67.5 million years ago, the second, largest pulse immediately before and ending at the KT boundary, in which about 80% of all the basalt was formed, and a third pulse around 64.5 million years ago. The volumes of basalt and the sizes of the rock thus formed are amazing. The second, largest pulse, which occurred over a few thousand years, consisted of several major eruptive events with volumes of each ranging from 20,000 km3 to 120,000 km3, attaining thicknesses up to 200 m and extending over hundreds of kilometers. All told, these flows built plateaus that are 3500 meters high today (taller before being eroded).
The Deccan eruptions released copious quantities of SO2 into the atmosphere in multiple events over several thousand years. Quantities released are estimated to have been around 150 gigatonnes SO2 per event for about 30 eruptive events overall. This would have resulted in significant cooling, acid rain, and perhaps ocean acidification globally. The marine extinction patterns suggest a collapse of production and food webs, and the loss of larger carnivorous species along with losses of certain types of smaller organisms. And the Deccan Traps were being formed earlier in the Cretaceous, so could have caused the extinctions which seem to have happened well before the KT boundary.
Let’s not forget the asteroid. It still landed. It landed at the same time as the KT boundary, and the major Deccan eruptions. It would have had sudden and devastating effects on larger terrestrial species, while the dust it kicked up would have added to the cooling being caused by the Deccan eruptions.
So What Does All This Tell Us About Tipping Points?
I started looking at the end Cretaceous mass extinction to gain an idea of just how fast tipping points tipped. We have learned several things. First, the fable is very incomplete – reality is far more complex, and it is unrealistic to tie the entire end-Cretaceous mass extinction to a single asteroid. Tying it to the asteroid exclusively demands that the extinctions occurred over a very short period of time right at the KT boundary. While the asteroid came in rapidly, and caused disruptions that likely lasted decades, the eruptions of the Deccan Traps were a series of events that continued over several tens of thousands of years divided among three pulses spanning perhaps three million years. These eruptions were certainly capable of bringing about the global cooling, and changes in ocean chemistry seen during the late Cretaceous. These changes, in turn, seem likely to have been the causes of many of the extinctions that occurred on land and in the oceans during the late Cretaceous, and particularly within the 5 million year-long Maastrichtian stage at its end. The arrival of the asteroid at the end of the Cretaceous provided the final cymbal crash to close a 30 million year-long period in which life became substantially more difficult for many species. This radical reorganization of the biosphere took place over many millions of years.
With this knowledge, we can look again at our present circumstances. The main conclusion I reach is that the tipping point we fear – the point at which anthropogenic impacts change the biosphere sufficiently that it reorganizes in a radically different form to what existed before – is already upon us. The world is changing as rapidly or more rapidly now as it was at any time during the late Cretaceous. The global climate is likely to warm at least 4oC over the 200 years from 1900 to 2100. Temperature fell 3-4oC during the entire 5 million years of the Maastrichtian. Our rate of global temperature change is about 25 thousand times more rapid than this average rate, and certainly as fast, or faster, than any of the most rapid temperature shifts within the Maastrichtian. That we are warming rather than cooling is likely immaterial – it is the pace of change that should alarm us.
Similarly, our rate of loss of species appears to most experts to be on track for a mass extinction if it continues. Anthony Barnosky, of UC Berkeley, and colleagues, published a provocative review of this question in Nature in 2011. They discussed the difficulties of using current information on rates of extinction and speciation in comparisons with records of extinction in the distant past. But they drew a
Figure from Barnosky et al, Nature, 2011, showing that current rates of extinction for birds, mammals and amphibians are on track to produce a mass extinction if they endure for at least 500 years. E/MSY = extinctions per million species-years, vertical lines and circles at right side represent, for each mass extinction, the range of rates (E/MSY) that would have produced that event given the range in estimates of its duration, and the rate needed if it were to have taken place over just 500 years.
firm conclusion to the question of whether current rates of extinction would result in a mass extinction (more than 75% of species lost) if they continued for a length of time equal to the duration of past mass extinctions. Their answer: “the answer is yes”. They report that the five past mass extinctions were believed to occur mostly over periods of about 1 to 3 million years (upper limit for end-Devonian event is 29 million years, lower limit for end-Cretaceous event is less than one year), and that modern rates of loss for mammals, birds and amphibians would all lead to 75% loss within 2 thousand years.
Elsewhere I have talked about the current rate of ocean acidification being greater than at any time in the past 65 million years. Rate of rise of sea level is as great as at any time since the end of the Pleistocene. Rate of removal of land from natural ecosystems to monoculture agriculture or urban settlements is clearly unprecedented. The tipping point is here, and while the roller-coaster may be less steep than as imagined in some movies, we are riding it, and it may be getting beyond our ability to control it.