Under a Green Sky: Global Warming, the Mass Extinctions of the Past, and What They Can Tell Us About Our Future

Book Review: Under A Green Sky: Global Warming, the Mass Extinctions of the Past, and What They Can Tell Us About Our Future – by Peter Ward, Ph.D. (Harper Collins Publishers, 2007)

This is a fun foray into the scientific worlds of paleontology, paleoclimatology, geology, and mass extinctions. The book reads like an adventure story, or rather a detective story – trying to piece together geologic clues from the past to determine what caused the mass extinctions of the past and what processes were involved in the preceding and subsequent years and how they compare to today’s global warming challenges. The author and his colleagues visit outcrops and sedimentary sequences all over the world, sometimes in isolated places and harsh environments. 

The first place worked is the Muller Canyon area of Nevada where rocks at the end of the Triassic period are exposed. I have done some rock-hounding and geologic mapping in central Nevada in the much older rocks of the Valley and Ridge in the sagebrush desert areas. It’s a great place to look at rocks. At the time geologists were looking for evidence of asteroid impact at the end of the Triassic as evidence was found at the end of the Cretaceous in the mass extinction that wiped out the dinosaurs. No convincing evidence has been found for impact at the end of the Triassic there, only a loss of many fossil species and a thick siltstone nearly bereft of fossils. If it wasn’t asteroid impact was it climate change, he considers. Eventually he builds up a model, a case, that it was indeed fast climate change, with rapid global warming and strong positive feedbacks that led to massive amounts of CO2, methane, and eventually other toxic gases like H2S bubbling out of the ocean and accumulating in the atmosphere, raising temperatures and making it hard to exist for many species. About 60% of all species on earth were lost in the mass extinction event at the end of the Triassic.

Next he ends up in the summer of 1982 in the Basque region, in the Pyrenees Mountains between France and Spain. Here he meets up with another geologist, Jost Wiedmann, a biostratigrapher cataloging, correlating, and dating fossil assemblages throughout the world. He noted that the extinction of ammonites in the fossil record near the K-T (Cretaceous-Tertiary) boundary was gradual, lasting about 20 million years, rather than immediate. Ward, with a fresh Ph.D., was interested in why the ammonite cephalopods went extinct at the K-T event after a 360 million year biological success and their cousins, the chambered nautilus, survived. He also studied wild nautilus by diving in the Pacific off the coasts of New Caledonia and Fiji.

A paper came out in 1980 by Luis and Walter Alvarez, a father and son team from the University of California, Berkeley that strongly advocated that the K-T extinction event was the result of an asteroid impact. Catastrophic environmental changes, particularly a long lasting “blackout” from massive amounts of particulate matter in the air, they proposed, were the mechanism of the mass extinction. Ward and Wiedmann found no ammonites within 15 meters of the proposed impact layer.

Mass extinctions were recognized in the fossil record in the 19^th century but were attributed to “catastrophism,” typically worldwide floods like the biblical flood. Such ideas were tossed as the science of paleontology developed further. The two largest mass extinctions divide the stratigraphic record into three main eras: the Paleozoic, the Mesozoic, and the Cenozoic. There are five main mass extinction events noted in the geologic record - from oldest to youngest: 1) Ordovican, 2) Devonian, end of Permian (Permian-Triassic), end of Triassic (Triassic-Jurassic), and end of Cretaceous (Cretaceous-Tertiary, or K-T).

Ward talks about a split among vertebrate and invertebrate paleontologists in the 1970’s where views on mass extinction was a factor: the vertebrate paleontologists did not think the mass extinctions occurred, only that the fossil record was missing. Evidence is now much stronger that the mass extinction indeed did occur and there is little dissent from that view. Two types of mass extinction were proposed: slow and gradual ones due to climate change, changing sea levels, disease, and predation; and rapid catastrophic ones characterized by the sudden disappearance of a large number of fossil biota in the record. The slow extinctions could not really be tested, only theorized. When asteroid impact became seen as a plausible mechanism for extinction there was at least something to look for – iridium and altered quartz that is associated with impacts. 

The Alvarez’s paper began a new paradigm, or revolution, in thinking about mass extinctions, that they weren’t slow and gradual and due to climate change but fast and due to asteroid impact and its after-effects which include climate change. He puts this in the contexts of Thomas Kuhn’s “structure of scientific revolutions.” Much evidence for a K-T boundary impact was accumulated: iridium, “shocked quartz,” spherules, and carbon isotope ratio changes which indicated a rapid loss of plant life presumably due to fire. However, some other geologists had another explanation: volcanism involving “flood basalts” and associated ash and lava flows. The impact vs. volcanism battle went on for over a decade. Flood basalts strongly correlated to all mass extinctions and even minor extinctions. Iridium, shocked quartz and spherules could also be associated with volcanism. Ward suggests that the geochemical evidence for impact was strong because they found what they were looking for in the impact layer but the fossil evidence required looking before and after in different places where the intervals were preserved. 

He tells of an odd experience stalking the Cretaceous-Tertiary boundary in France at a beachside outcrop where there was a large group of tanned naked frolicking gay men while he hammered rocks in geologist garb! Here he finds 12 species of ammonites in abundance near the boundary where in other places they seemingly died off gradually – here they did not until the actual boundary layer, which is further evidence of the asteroid impact. Ward proves that impact cannot kill off just what would become microfossils but macrofossils as well. He presents his findings at a conference where Jost Wiedmann was in attendance, after Wiedmann asserted that impact was not the cause and that the extinction of the ammonites came slowly. Wiedmann listened to his talk then left and never spoke to Ward again – dying a few years later, as Ward explains, his life’s work disproved by an apprentice. Science can indeed be a sad world. By the end of the 1980’s the evidence for impact as the cause of the K-T extinction was very strong. The 120-mile wide impact crater was found (in the Yucatan peninsula of Mexico) and both the geochemical and paleontological evidence supported a very rapid mass extinction. The problem, notes Ward, is that now all the other mass extinctions were assumed to have been caused by impact, as the new “paradigm” took hold.

Ward’s further studies in the French Pyrenees examined the quick (geologically speaking) recovery of life in the Late Paleocene of the Tertiary Period the first 5 million years after the K-T extinction event. The new fossils are of species still around today and indicate the area was warmer as they were tropical species. Oxygen isotope ratios found in shell material provide a very good record of temperatures when they were made. Analysis of oxygen isotope ratios from bottom-dwelling (benthic) organisms from the Antarctic a few million years after the K-T boundary showed that the basal ocean water there had anomalously warmed over a short period of time. The warmer water in the polar high latitudes (both Arctic and Antarctic) was also found to be more depleted of oxygen which caused an extinction of benthic organisms here at the Paleocene-Eocene boundary a few million years after the K-T asteroid impact boundary. The benthic organisms were not affected directly by the impact. The suggestion was that the oceanic conveyor belt which transfers heat to and from depth in the ocean was somehow shut down – presumably by the warm surface temperatures. This became known as the Paleocene thermal event. The event was confirmed to have occurred on land also by compared patterns of carbon and oxygen isotope ratios in well-measured fossil assemblage sections in Wyoming. Here many exotic forms of mammals were found, many now extinct. The Paleocene thermal event is considered a minor extinction event. More evidence was searched for in Aeolian (wind) deposits – basically dust that made it to the ocean floor. The amount was reduced and extremely reduced at the point of the event suggesting low wind conditions – typically as a result of prolonged arid weather. Also found was volcanic ash and indeed a great uptick in volcanic activity 58-56 million years ago. Estimates of seawater temperature differences from equator to poles (now 45 deg C) then shifted from 17 deg C to a mere 6 deg C, suggesting a quite unusual homogeneous ocean temperature. The basic mechanism of the Paleocene thermal event is thought to have been volcanoes spewing carbon dioxide with the CO2 heating up the surface of the planet and later the ocean, shutting down the deep-water circulation conveyor belt. The event ended after the volcanism subsided and later when the CO2 levels finally dropped. By 2000, other minor extinctions began to show similarities to the Paleocene event.

He ends up in the Southern Tunisian Desert in 2000 at one of the best exposures of the K-T boundary. This time they took small cores with the goal of discovering the magnetic stratigraphy as Alvarez and colleagues did in other sections. Here there is a six-foot layer of black rock in an otherwise 100ft thick cliff of white limestone. This black layer can also be found in Italy, England, Wyoming, Colorado, California, offshore British Columbia, and Alaska. This represents an abrupt change to anoxic (oxygen depleted) water. This extinction and others were now firmly linked to warming oceans. 

Next he explores the Permian mass extinction, the “mother of all extinctions” and the Great Dying, along the Caledon River in South Africa. After ten years of studying the K-T boundary, Ward was now fossil hunting near the Permian-Triassic boundary for land animals, terrestrial fossils. The P-T mass extinction resulted in the loss of up to 90% of species on earth. He found one of the best outcrop sections of the transition and noted the difference between the K-T and P-T boundaries’ fossil losses – The P-T losses were more gradual and seemed to be the result of many small events and one big one, rather than one abrupt big one as in the K-T asteroid impact. No asteroid impact was implicated here even though at the time he was looking for one. The P-T boundary was associated with global warming, an anoxic ocean, and volcanic activity via flood basalts from the massive Siberian Traps – a source of CO2 to heat everything up. However, the impact advocators also found what they thought was evidence – so-called “bucky balls” or “fullerenes,” geodesic-dome shaped carbon molecules named after Buckminster Fuller, that were thought to be of extra-terrestrial origin – thus suggesting impact. However, no iridium was found. NASA scientists reported that they may have found an impact crater that caused the P-T extinction in 2003. In 2006, scientists at Ohio State University reported a large impact crater deep in Antarctic ice detected with gravity anomaly measurements but it could not be seen or dated. Ideas of a comet impact also came about with that impact initiating volcanism but these ideas were all vague and difficult to confirm. Eminent paleontologists and geochemists got together to discuss the ideas and re-examine the evidence. They later found that the bucky balls did not come from the Permian but from much younger rocks in the Triassic and so did not correlate to the loss of species.

Another aspect of P-T boundary time was increased atmospheric methane, a greenhouse gas which would have heated things up. Extinction of many plant species occurred and subsequent increases in sedimentation rates. Tropical species appeared where there were previously temperate species. Increased volcanism, repeated changes in oceanic circulation, and presumed methane hydrate melting impulses are also in evidence. Impact as a possible cause for the Permian extinction has been rejected by the majority of scientists.

 A group led by Harvard paleobotanist Andrew Knoll beginning in 1996 proposed that the Permian extinction was similar to the Precambrian extinction of 600 million years ago. Similarities were a stratified ocean with oxygen near the surface but depleted at depth and large amounts of organic material as bottom sediments. When this changed, possibly due to plate tectonics, the deep ocean carbon began to be liberated to surface water and then to the atmosphere through large bubbles. The P-T boundary isotope changes showed a series of perturbations rather than a single one as the K-T had shown. This suggested multiple events over several million years.

In 2001 Ward ends up in the Queen Charlotte Islands off of the coast of British Columbia to study well-exposed sections of the Triassic-Jurassic boundary, the T-J extinction being responsible for the loss of about half of earth’s species. They wanted to get auger cores stratigraphically through the boundary to compare isotope signatures to the other extinctions. They did so in 1996 and found a main single event but did not get very far into the Jurassic section where they now hoped to see if there were multiple perturbations as there had been in the Permian. That is indeed what they found. However, iridium was recently found in several localities in some of the best T-J boundary exposures in the Newark Basin and Connecticut River valley areas of New Jersey, which suggested impact. However, the amount of iridium was quite small compared to the K-T iridium. The proposed impact crater in Quebec was later dated to be about 15 million years too soon to have caused the event. The impact from that massive crater apparently did not cause any significant extinctions – which suggests that the effects of asteroid impact may have been overestimated. 

In 2004 he returns to the Queen Charlotte Islands to look at older rocks on the distant islands to see if extinction was single or multiple, gradual or abrupt. He digs ammonites beginning about 12 million years before the extinction and notes a classic slow gradual decrease in species of them and other fossils. He notes that while his early career was involved in showing what was once thought to be a gradual extinction at the K-T boundary was actually abrupt, now he was showing what was presumed by many to be a sudden extinction at the T-J boundary was actually a slow gradual one. The progression seemed to be that ammonites first reduced their variety as some species died out then a new species of clam, Monotis, appeared in abundance, only to be reduced as the extinction got worse. Monotis might possibly have been adapted to lower oxygen sea bottoms. Better dating techniques by finding a volcanic ash bed to date revealed that the Rhaetian stage of the late Triassic, with low oxygen seas largely devoid of life lasted up to 11 million years. After the Rhaetian stage came the Norian stage when the rest of the bivalves and ammonites died out so Ward sees this as two extinctions, one quite gradual and culminating at the end of the Rhaetian and one more abrupt but still gradual ending at the end of the Norian stage. Subsequent fossil work in other places showed extinction pulses occurring into the Jurassic as well. To sum up it was now thought that most extinctions were gradual and only one, the K-T, was definitively associated with impact, the others being logically ruled out. Thus the ‘extinctions were caused by asteroids’ paradigm was given up except for K-T.

The next chapter finds Ward diving in a pristine coral reef near Palau in tropical Pacific Micronesia. This was back in 1983. Ward was a long-experienced diver. He lost a fellow diver in the past who had passed out during a deep dive and Ward got a serious case of the bends attempting to save his live by bringing him up fast. His friend died but Ward suffered chronic bodily pains and a permanent limp from his own injuries. Here they were studying the nautiluses, along with the ammonites, another cephalopod. The ammonites survived many extinctions but were wiped out at the K-T boundary in the Cretaceous. They tagged the nautiloids and found that they dived deep during the day and came closer to surface at night. That may have been why they survived the K-T and the ammonites who stayed in shallow water did not. It seems that while the Permian, Paleocene, and Tertiary extinctions wiped out bottom dwellers the K-T extinction wiped out the surface dwellers.

It was still unclear exactly how a slow gradual change of climate could have killed so many species several times in the past. New ideas were forming. Microbiologists studying anoxic lakes found some new fossils, chemical fossils, known as biomarkers. They did not leave behind skeletal remains but chemical remains in the lake sediment. Toxic hydrogen sulfide gas (H2S) was one chemical marker and calculations by one author, Kump, suggested that the amount of H2S was significant in the Permian – 2000 times that produced by volcanoes. The Kump Hypothesis also noted that the H2S would have destroyed the ozone layer and evidence from Greenland of fossils damaged by ultraviolet light suggests this may have occurred. Destruction of the ozone layer would mean a decrease in phytoplankton, the base of the food chain. Another hypothesis suggests the ozone layer could have been destroyed by particles from a supernova. With increased CO2 and methane bubbling up from the sea in a hot Permian the H2S would have been more toxic as it is in a warmer environment. Evidence was found of H2S –producing microbes in the Permian throughout the world. Since sea level was low at the time they also looked for evidence of eroding phosphorous which would have been a nutrient for microbes to accelerate their growth.

Next he ends up near his hometown, Seattle, looking at fossils in non-bedded limestones deposited in a “mixed” ocean of little oxygen variation with cold areas at the poles and warm ones at the tropics, as now, or since the Oligocene, about 30 million years ago. Older rocks show black bedded rocks deposited in an anoxic ocean bottom. Pyrite is common in these rocks.  Anoxic bottoms are filled with black shales, around since 3.5 billion years ago and sometimes with very-well preserved fossils of life forms that fell into the sediment with their forms preserved. The famous Burgess Shale is one example. There are two types of stratified oceans, he notes: one with low-oxygen bottoms which supports some life, mostly microbial; and one entirely devoid of oxygen which supports only microbes that utilize sulfur for food and give off H2S as a waste product. The latter is known as a Canfield ocean. Canfield oceans were toxic to life. They are thought to have been around in the Precambrian inhibiting the development of life. The eukaryotes require microbes to fix nitrogen, a needed nutrient, for them. The sulfur-imbibing microbes do not fix nitrogen, instead inhibiting it. Chemical biomarkers also suggest that the T-J extinction is associated with pulses of short-lived Canfield ocean conditions. The oceanic circulation, the conveyor belt, may be the key to the changing ocean states. There is strong evidence that the conveyor belt shut down (or shifted) in the Paleocene and now it appears that this happened in the Permian as well. Of course, the continents were in different places in these past times due to plate tectonics so the actual circulation patterns were different than today but a similar mechanism is still likely to have been in play. The shift in ocean circulation in the Permian was thought to have brought anoxic water to the deep ocean which allowed the H2S-producing microbes to thrive and upwelling of poisonous bottom-waters. If the Paleocene had H2S-producing microbes they were at far lower concentrations than in the Permian. He compares extinctions from Anthony Hallam’s and Paul Wignall’s 1997 book, Mass Extinctions and Their Aftermath, which was written when impact was still thought to be associated with most or all extinctions. Even so, their data revealed that of the 14 mass extinctions that were catalogued, 12 were associated with poorly oxygenated oceans as a major cause. The three “kill mechanisms” are now thought to be heat, low oxygen, and perhaps H2S.

Next he ends up in Namibia in Southern Africa where the scorching hot Kalahari Desert is flanked by a foggy Atlantic Ocean that is very cold. Models of atmospheric CO2 and O2 concentrations of the past can be made using changes in sedimentation burial rates. One of the main modeling setups for paleoclimatological studies is GEOCARB for CO2 and GEOCARBSULF for oxygen. Modeling indicates that CO2 levels were very high from the Precambrian to the lower Permian – from about 5000 then down to about 300 PPM, rising back up to 3000 near the Permian extinction. Modeling also indicates that all of the mass extinctions of the past with the exception of the K-T impact-caused extinction, are associated with maximum or ‘rising toward maximum’ atmospheric CO2 concentrations. Thus rapid rises in CO2 correlate strongly to mass extinctions. This implicates our anthropogenic CO2 increase as a potential cause as well – if it were to rise ever higher – though likely far beyond current projections. Another way to estimate past CO2 concentrations is through fossil plant leaves. These readings on leaf stomata confirmed the CO2 estimates modeled.

Ward summarizes the sequences of events that are thought to have taken place in these mass extinctions: 1) world warms due to increase in greenhouse gases, initially from volcanoes; 2) The ocean circulation system is disrupted or shut down; 3) the deep ocean becomes de-oxygenated then shallow water suffers the same fate; 4) deoxygenated shallow water bottoms with some light penetration allow green sulfur bacteria to grow and produce H2S which rises in the atmosphere and breaks down the ozone layer with the UV light killing off phytoplankton. – The high heat and H2S also cause mass extinction on land. He notes significant variability in each extinction and calls the model the ‘conveyor disruption hypothesis.’ He envisions seas full of gelatinous bacterial mats, stromatolites which would later become food for terrestrial herbivores as (very slow and weak) waves brought them in. The ocean would look serene and waveless and be purple due to floating bacteria. Thick bubbles of various sizes filled with poisonous H2S would belch from the sea giving the sky a green tint – thus the book’s title. The bottom line is perhaps the realization that it is mainly increased atmospheric CO2 and other greenhouse gases like methane that serve as the trigger for mass extinctions. 

Next he talks about bridging all the varying scientific disciplines involved in modern climatology and paleoclimatology. For much of the book he also addresses motivations for reward and prestige among scientists and how that can affect their work. 

He goes into the carbon dating work of Minze Stuiver of the Quaternary Research Institute. He dated the Greenland ice cores year-by-year dating back 200,000 years. Using mass spectrometers they were able to accurately approximate temperatures and CO2 levels. What they found is that the current climate on Earth is quite aberrant even for recent geological history. Temperature changes of up to 18 deg F over a few decades were more common in the past.  Before 10,000 years ago it is thought that storms the size of the major hurricanes occurred several times a year. At about 10,000 years ago a period of unprecedented calm apparently set in. Humans settled and mastered agriculture during this new period of calm. The records of the ice cores match quite well the planetary and orbital cycles proposed by  Milankvitch with those cycles being the triggers for glacial and interglacial periods. One of the unknowns that Ward emphasizes is how much CO2 and global warming would it take to alter the oceanic circulation system. Wally Broecker thinks it could slow down but is unlikely to shut down with even say 1000 PPM CO2. It may be changing now. Fresh water from melting northern ice could be a prime trigger for changing the conveyor belt. Ward goes through smaller time period climate cycles like the Dansgaard-Oeschger cycles and the cycles of floating melting ice dropping cobbles they were carrying, now called Heinrich events – seen in the ocean floor sediments. For 90% of the last 100,000 years the earth has been in an ice age so these are anomalous times indeed. Before 8000 years ago the conveyor belt is thought to have been less stable. The current stable period is a precarious stability, scientists suggest. Biodiversity strongly correlates to this stability. The implication is that the “on-off” conveyor belt tips the earth’s climate to one of two stable states: the cold one that takes up 90 % of the last 100,000 years and the warm one we are in now.

He next visits Manua Loa in Hawaii where atmospheric CO2 has been dutifully measured since the 1950’s – as part of a Canadian TV documentary about climate change. In addressing climate history of the last 8000 years Ward gives the data from William Ruddiman which shows that humans have been affecting CO2 and methane levels since the advent of agriculture, forest burning to clear land, flood for rice paddies (which is major source of methane), and livestock agriculture (another major source of methane). The CO2 range of the last 200,000 years has been between 180 and 280 PPM with most of it in the small end of the range since most times were ice age times. At the beginning of the Industrial Age CO2 levels were at 280 PPM and now they are above 400 PPM, a level unprecedented in the last 200,000 years. CO2 can also directly cause limited extinctions of certain species in the form of increases in ocean acidity and this is happening now in cases of coral bleaching. The changes in ocean pH will likely persist for thousands of years, he notes, thus changing life patterns. While there may have been times of high ocean acidity in the past he suggests that they have not been as high as they are expected to get soon for quite some time – perhaps 100 million years – since certain species were more adapted in the past to higher acidity – however, the abrupt changes now due to anthropogenic CO2 are too fast for many species to evolve adaptations. The present rate in the rise of CO2 seems to be faster than at any period in the past and global average temperatures have not been this warm since the Eocene epoch 60 million years ago which followed a mass extinction. 

Next he delves into the Eocene epoch looking at fossils along the Pacific coast of North America. He notes that this hot time was a time of very high sea levels compared to today. This area was tropical during the Eocene as evidenced by abundant palm and crocodile fossils found as far north as the Arctic Circle. He explores the climatic features of the Eocene and compares them to what a 1000 PPM atmospheric CO2 level world might be like as after we humans create it. First he notes that the tropics are the source of many of the human diseases that affect us. He suggests that tropical peoples in particular have developed coping mechanisms for the heat in the form of various local drugs. I am not so sure they have the monopoly on that. He mentions widespread use of betel nut, kava root, and khat. Of course, the same could be said for alcohol and cannabis in the temperate climes. The prevalence of mosquitoes makes malaria and other diseases more likely as well. He goes through all the typical scenarios of global warming effects: melting ice, rising sea levels, changing weather patterns, submerged cities, storm surges, changes in habitat patterns, etc. He notes that the temperature rise in the Arctic has been 20 times that of other places on earth and is quite worrying to scientists. Are effects underestimated? Overestimated? No one knows for sure but some attribute a significant amount of deaths now to global warming in the form of malaria and malnutrition. He invokes the view that hurricanes will worsen in both magnitude and frequency, popular at time of publication. However, that has not occurred and may end up being a misattributed global warming affect. The increase in hurricanes from 1990-2004 may be part of a natural cycle. Heat waves are another effect that has increased. Suggestions of war and famine are speculative. Cereal grain crops may not yield well in a more tropical climate. 

Next he discusses climate and the possibility of re-entering an Eocene-like epoch with famed University of Washington climate scientist David Battista. Windless tropical conditions in some temperate areas with super hurricanes pounding the equatorial tropics. The conveyor might change into a form where warm water from the tropics sinks much further south in the Atlantic which would freeze Western Europe perhaps giving the false impression to some of an impending ice age. Then when the sinking low salinity freshwater did not sink deep enough a situation of lower oxygen could develop at ocean depths resulting in the next chain in the link of mass extinctions that have occurred in the past.

He goes through some more speculative scenarios at different CO2 levels but it really is hard to know how things will play out and there are still uncertainties about that. 

Great book overall by a geologist who wears the scars of his work and his craft through an adventurous but often lonely existence in far off corners of the world as well as in the academic realms.